diff --git "a/checkpoint-45000/trainer_state.json" "b/checkpoint-45000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-45000/trainer_state.json"
@@ -0,0 +1,31534 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998222538215429,
+  "eval_steps": 500,
+  "global_step": 45000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00022218272307145397,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.9996000000000003e-05,
+      "loss": 1.2837,
+      "step": 10
+    },
+    {
+      "epoch": 0.00044436544614290794,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.9991555555555558e-05,
+      "loss": 0.6861,
+      "step": 20
+    },
+    {
+      "epoch": 0.0006665481692143619,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.9987111111111113e-05,
+      "loss": 0.6204,
+      "step": 30
+    },
+    {
+      "epoch": 0.0008887308922858159,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.9982666666666668e-05,
+      "loss": 0.5587,
+      "step": 40
+    },
+    {
+      "epoch": 0.0011109136153572697,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.9978222222222222e-05,
+      "loss": 0.5842,
+      "step": 50
+    },
+    {
+      "epoch": 0.0013330963384287238,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.9973777777777777e-05,
+      "loss": 0.6044,
+      "step": 60
+    },
+    {
+      "epoch": 0.0015552790615001777,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9969333333333335e-05,
+      "loss": 0.5807,
+      "step": 70
+    },
+    {
+      "epoch": 0.0017774617845716318,
+      "grad_norm": 2.625,
+      "learning_rate": 1.996488888888889e-05,
+      "loss": 0.5589,
+      "step": 80
+    },
+    {
+      "epoch": 0.0019996445076430856,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.9960444444444445e-05,
+      "loss": 0.5105,
+      "step": 90
+    },
+    {
+      "epoch": 0.0022218272307145395,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9956000000000003e-05,
+      "loss": 0.5402,
+      "step": 100
+    },
+    {
+      "epoch": 0.002444009953785994,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.9951555555555558e-05,
+      "loss": 0.5461,
+      "step": 110
+    },
+    {
+      "epoch": 0.0026661926768574476,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9947111111111113e-05,
+      "loss": 0.5681,
+      "step": 120
+    },
+    {
+      "epoch": 0.0028883753999289015,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.9942666666666668e-05,
+      "loss": 0.5401,
+      "step": 130
+    },
+    {
+      "epoch": 0.0031105581230003554,
+      "grad_norm": 2.5,
+      "learning_rate": 1.9938222222222223e-05,
+      "loss": 0.525,
+      "step": 140
+    },
+    {
+      "epoch": 0.0033327408460718097,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9933777777777778e-05,
+      "loss": 0.5138,
+      "step": 150
+    },
+    {
+      "epoch": 0.0035549235691432635,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9929333333333332e-05,
+      "loss": 0.5688,
+      "step": 160
+    },
+    {
+      "epoch": 0.0037771062922147174,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.992488888888889e-05,
+      "loss": 0.5616,
+      "step": 170
+    },
+    {
+      "epoch": 0.003999289015286171,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9920444444444446e-05,
+      "loss": 0.5798,
+      "step": 180
+    },
+    {
+      "epoch": 0.004221471738357625,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.9916e-05,
+      "loss": 0.5198,
+      "step": 190
+    },
+    {
+      "epoch": 0.004443654461429079,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.991155555555556e-05,
+      "loss": 0.5195,
+      "step": 200
+    },
+    {
+      "epoch": 0.004665837184500533,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9907111111111113e-05,
+      "loss": 0.5163,
+      "step": 210
+    },
+    {
+      "epoch": 0.004888019907571988,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.990266666666667e-05,
+      "loss": 0.5526,
+      "step": 220
+    },
+    {
+      "epoch": 0.0051102026306434414,
+      "grad_norm": 2.375,
+      "learning_rate": 1.9898222222222223e-05,
+      "loss": 0.5381,
+      "step": 230
+    },
+    {
+      "epoch": 0.005332385353714895,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9893777777777778e-05,
+      "loss": 0.4959,
+      "step": 240
+    },
+    {
+      "epoch": 0.005554568076786349,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.9889333333333336e-05,
+      "loss": 0.5625,
+      "step": 250
+    },
+    {
+      "epoch": 0.005776750799857803,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.988488888888889e-05,
+      "loss": 0.5574,
+      "step": 260
+    },
+    {
+      "epoch": 0.005998933522929257,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9880444444444446e-05,
+      "loss": 0.5417,
+      "step": 270
+    },
+    {
+      "epoch": 0.006221116246000711,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9876e-05,
+      "loss": 0.5873,
+      "step": 280
+    },
+    {
+      "epoch": 0.006443298969072165,
+      "grad_norm": 2.75,
+      "learning_rate": 1.9871555555555556e-05,
+      "loss": 0.5572,
+      "step": 290
+    },
+    {
+      "epoch": 0.006665481692143619,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9867111111111114e-05,
+      "loss": 0.4952,
+      "step": 300
+    },
+    {
+      "epoch": 0.006887664415215073,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.986266666666667e-05,
+      "loss": 0.5766,
+      "step": 310
+    },
+    {
+      "epoch": 0.007109847138286527,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9858222222222224e-05,
+      "loss": 0.5868,
+      "step": 320
+    },
+    {
+      "epoch": 0.007332029861357981,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.985377777777778e-05,
+      "loss": 0.531,
+      "step": 330
+    },
+    {
+      "epoch": 0.007554212584429435,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9849333333333337e-05,
+      "loss": 0.5477,
+      "step": 340
+    },
+    {
+      "epoch": 0.007776395307500889,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.984488888888889e-05,
+      "loss": 0.5688,
+      "step": 350
+    },
+    {
+      "epoch": 0.007998578030572343,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9840444444444446e-05,
+      "loss": 0.5066,
+      "step": 360
+    },
+    {
+      "epoch": 0.008220760753643797,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.9836e-05,
+      "loss": 0.5124,
+      "step": 370
+    },
+    {
+      "epoch": 0.00844294347671525,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9831555555555556e-05,
+      "loss": 0.5109,
+      "step": 380
+    },
+    {
+      "epoch": 0.008665126199786705,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.982711111111111e-05,
+      "loss": 0.5672,
+      "step": 390
+    },
+    {
+      "epoch": 0.008887308922858158,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.982266666666667e-05,
+      "loss": 0.5503,
+      "step": 400
+    },
+    {
+      "epoch": 0.009109491645929613,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9818222222222224e-05,
+      "loss": 0.5321,
+      "step": 410
+    },
+    {
+      "epoch": 0.009331674369001066,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.981377777777778e-05,
+      "loss": 0.4862,
+      "step": 420
+    },
+    {
+      "epoch": 0.00955385709207252,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.9809333333333337e-05,
+      "loss": 0.4908,
+      "step": 430
+    },
+    {
+      "epoch": 0.009776039815143975,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.9804888888888892e-05,
+      "loss": 0.5749,
+      "step": 440
+    },
+    {
+      "epoch": 0.009998222538215428,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.9800444444444447e-05,
+      "loss": 0.532,
+      "step": 450
+    },
+    {
+      "epoch": 0.010220405261286883,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9796e-05,
+      "loss": 0.5432,
+      "step": 460
+    },
+    {
+      "epoch": 0.010442587984358336,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9791555555555556e-05,
+      "loss": 0.5124,
+      "step": 470
+    },
+    {
+      "epoch": 0.01066477070742979,
+      "grad_norm": 2.375,
+      "learning_rate": 1.978711111111111e-05,
+      "loss": 0.5398,
+      "step": 480
+    },
+    {
+      "epoch": 0.010886953430501244,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.978266666666667e-05,
+      "loss": 0.5623,
+      "step": 490
+    },
+    {
+      "epoch": 0.011109136153572698,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9778222222222224e-05,
+      "loss": 0.4947,
+      "step": 500
+    },
+    {
+      "epoch": 0.011331318876644151,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.977377777777778e-05,
+      "loss": 0.5006,
+      "step": 510
+    },
+    {
+      "epoch": 0.011553501599715606,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.9769333333333337e-05,
+      "loss": 0.5097,
+      "step": 520
+    },
+    {
+      "epoch": 0.01177568432278706,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9764888888888892e-05,
+      "loss": 0.4702,
+      "step": 530
+    },
+    {
+      "epoch": 0.011997867045858514,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9760444444444447e-05,
+      "loss": 0.5285,
+      "step": 540
+    },
+    {
+      "epoch": 0.012220049768929968,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9756000000000002e-05,
+      "loss": 0.522,
+      "step": 550
+    },
+    {
+      "epoch": 0.012442232492001421,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9751555555555557e-05,
+      "loss": 0.4878,
+      "step": 560
+    },
+    {
+      "epoch": 0.012664415215072876,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.974711111111111e-05,
+      "loss": 0.5499,
+      "step": 570
+    },
+    {
+      "epoch": 0.01288659793814433,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9742666666666666e-05,
+      "loss": 0.487,
+      "step": 580
+    },
+    {
+      "epoch": 0.013108780661215784,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9738222222222225e-05,
+      "loss": 0.5365,
+      "step": 590
+    },
+    {
+      "epoch": 0.013330963384287239,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.973377777777778e-05,
+      "loss": 0.5492,
+      "step": 600
+    },
+    {
+      "epoch": 0.013553146107358692,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9729333333333334e-05,
+      "loss": 0.4887,
+      "step": 610
+    },
+    {
+      "epoch": 0.013775328830430146,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9724888888888893e-05,
+      "loss": 0.5365,
+      "step": 620
+    },
+    {
+      "epoch": 0.0139975115535016,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9720444444444447e-05,
+      "loss": 0.5356,
+      "step": 630
+    },
+    {
+      "epoch": 0.014219694276573054,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9716000000000002e-05,
+      "loss": 0.5229,
+      "step": 640
+    },
+    {
+      "epoch": 0.014441876999644507,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9711555555555557e-05,
+      "loss": 0.4801,
+      "step": 650
+    },
+    {
+      "epoch": 0.014664059722715962,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9707111111111112e-05,
+      "loss": 0.5292,
+      "step": 660
+    },
+    {
+      "epoch": 0.014886242445787415,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9702666666666667e-05,
+      "loss": 0.4828,
+      "step": 670
+    },
+    {
+      "epoch": 0.01510842516885887,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9698222222222222e-05,
+      "loss": 0.5338,
+      "step": 680
+    },
+    {
+      "epoch": 0.015330607891930324,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.969377777777778e-05,
+      "loss": 0.5348,
+      "step": 690
+    },
+    {
+      "epoch": 0.015552790615001777,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9689333333333335e-05,
+      "loss": 0.5324,
+      "step": 700
+    },
+    {
+      "epoch": 0.01577497333807323,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.968488888888889e-05,
+      "loss": 0.4862,
+      "step": 710
+    },
+    {
+      "epoch": 0.015997156061144685,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9680444444444448e-05,
+      "loss": 0.4729,
+      "step": 720
+    },
+    {
+      "epoch": 0.01621933878421614,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9676000000000003e-05,
+      "loss": 0.5072,
+      "step": 730
+    },
+    {
+      "epoch": 0.016441521507287594,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9671555555555558e-05,
+      "loss": 0.4851,
+      "step": 740
+    },
+    {
+      "epoch": 0.016663704230359046,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9667111111111112e-05,
+      "loss": 0.5163,
+      "step": 750
+    },
+    {
+      "epoch": 0.0168858869534305,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9662666666666667e-05,
+      "loss": 0.5307,
+      "step": 760
+    },
+    {
+      "epoch": 0.017108069676501955,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9658222222222222e-05,
+      "loss": 0.4971,
+      "step": 770
+    },
+    {
+      "epoch": 0.01733025239957341,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9653777777777777e-05,
+      "loss": 0.5307,
+      "step": 780
+    },
+    {
+      "epoch": 0.017552435122644865,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.9649333333333335e-05,
+      "loss": 0.5158,
+      "step": 790
+    },
+    {
+      "epoch": 0.017774617845716316,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.964488888888889e-05,
+      "loss": 0.5216,
+      "step": 800
+    },
+    {
+      "epoch": 0.01799680056878777,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9640444444444445e-05,
+      "loss": 0.544,
+      "step": 810
+    },
+    {
+      "epoch": 0.018218983291859225,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9636000000000003e-05,
+      "loss": 0.5232,
+      "step": 820
+    },
+    {
+      "epoch": 0.01844116601493068,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9631555555555558e-05,
+      "loss": 0.481,
+      "step": 830
+    },
+    {
+      "epoch": 0.01866334873800213,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9627111111111113e-05,
+      "loss": 0.4943,
+      "step": 840
+    },
+    {
+      "epoch": 0.018885531461073586,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9622666666666668e-05,
+      "loss": 0.5153,
+      "step": 850
+    },
+    {
+      "epoch": 0.01910771418414504,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9618222222222222e-05,
+      "loss": 0.4575,
+      "step": 860
+    },
+    {
+      "epoch": 0.019329896907216496,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.961377777777778e-05,
+      "loss": 0.5078,
+      "step": 870
+    },
+    {
+      "epoch": 0.01955207963028795,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9609333333333336e-05,
+      "loss": 0.4857,
+      "step": 880
+    },
+    {
+      "epoch": 0.0197742623533594,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.960488888888889e-05,
+      "loss": 0.5011,
+      "step": 890
+    },
+    {
+      "epoch": 0.019996445076430856,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9600444444444445e-05,
+      "loss": 0.5439,
+      "step": 900
+    },
+    {
+      "epoch": 0.02021862779950231,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9596e-05,
+      "loss": 0.4918,
+      "step": 910
+    },
+    {
+      "epoch": 0.020440810522573766,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.959155555555556e-05,
+      "loss": 0.4805,
+      "step": 920
+    },
+    {
+      "epoch": 0.020662993245645217,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9587111111111113e-05,
+      "loss": 0.4735,
+      "step": 930
+    },
+    {
+      "epoch": 0.02088517596871667,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9582666666666668e-05,
+      "loss": 0.4953,
+      "step": 940
+    },
+    {
+      "epoch": 0.021107358691788126,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9578222222222223e-05,
+      "loss": 0.4638,
+      "step": 950
+    },
+    {
+      "epoch": 0.02132954141485958,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.957377777777778e-05,
+      "loss": 0.5258,
+      "step": 960
+    },
+    {
+      "epoch": 0.021551724137931036,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.9569333333333336e-05,
+      "loss": 0.4854,
+      "step": 970
+    },
+    {
+      "epoch": 0.021773906861002487,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.956488888888889e-05,
+      "loss": 0.5109,
+      "step": 980
+    },
+    {
+      "epoch": 0.021996089584073942,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9560444444444446e-05,
+      "loss": 0.5116,
+      "step": 990
+    },
+    {
+      "epoch": 0.022218272307145397,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9556e-05,
+      "loss": 0.5149,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02244045503021685,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9551555555555555e-05,
+      "loss": 0.5,
+      "step": 1010
+    },
+    {
+      "epoch": 0.022662637753288303,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9547111111111114e-05,
+      "loss": 0.4927,
+      "step": 1020
+    },
+    {
+      "epoch": 0.022884820476359757,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.954266666666667e-05,
+      "loss": 0.5126,
+      "step": 1030
+    },
+    {
+      "epoch": 0.023107003199431212,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9538222222222223e-05,
+      "loss": 0.5177,
+      "step": 1040
+    },
+    {
+      "epoch": 0.023329185922502667,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.953377777777778e-05,
+      "loss": 0.4775,
+      "step": 1050
+    },
+    {
+      "epoch": 0.02355136864557412,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9529333333333336e-05,
+      "loss": 0.5069,
+      "step": 1060
+    },
+    {
+      "epoch": 0.023773551368645573,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.952488888888889e-05,
+      "loss": 0.4852,
+      "step": 1070
+    },
+    {
+      "epoch": 0.023995734091717028,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9520444444444446e-05,
+      "loss": 0.4755,
+      "step": 1080
+    },
+    {
+      "epoch": 0.024217916814788482,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9516e-05,
+      "loss": 0.4473,
+      "step": 1090
+    },
+    {
+      "epoch": 0.024440099537859937,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9511555555555556e-05,
+      "loss": 0.5012,
+      "step": 1100
+    },
+    {
+      "epoch": 0.024662282260931388,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.950711111111111e-05,
+      "loss": 0.4845,
+      "step": 1110
+    },
+    {
+      "epoch": 0.024884464984002843,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.950266666666667e-05,
+      "loss": 0.4897,
+      "step": 1120
+    },
+    {
+      "epoch": 0.025106647707074298,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9498222222222224e-05,
+      "loss": 0.4571,
+      "step": 1130
+    },
+    {
+      "epoch": 0.025328830430145752,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9493777777777782e-05,
+      "loss": 0.4878,
+      "step": 1140
+    },
+    {
+      "epoch": 0.025551013153217207,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9489333333333337e-05,
+      "loss": 0.5114,
+      "step": 1150
+    },
+    {
+      "epoch": 0.02577319587628866,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.948488888888889e-05,
+      "loss": 0.5107,
+      "step": 1160
+    },
+    {
+      "epoch": 0.025995378599360113,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9480444444444446e-05,
+      "loss": 0.5008,
+      "step": 1170
+    },
+    {
+      "epoch": 0.026217561322431568,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.9476e-05,
+      "loss": 0.5021,
+      "step": 1180
+    },
+    {
+      "epoch": 0.026439744045503023,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9471555555555556e-05,
+      "loss": 0.538,
+      "step": 1190
+    },
+    {
+      "epoch": 0.026661926768574477,
+      "grad_norm": 1.625,
+      "learning_rate": 1.946711111111111e-05,
+      "loss": 0.4946,
+      "step": 1200
+    },
+    {
+      "epoch": 0.02688410949164593,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.946266666666667e-05,
+      "loss": 0.4855,
+      "step": 1210
+    },
+    {
+      "epoch": 0.027106292214717383,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9458222222222224e-05,
+      "loss": 0.5346,
+      "step": 1220
+    },
+    {
+      "epoch": 0.027328474937788838,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.945377777777778e-05,
+      "loss": 0.5336,
+      "step": 1230
+    },
+    {
+      "epoch": 0.027550657660860293,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9449333333333337e-05,
+      "loss": 0.5163,
+      "step": 1240
+    },
+    {
+      "epoch": 0.027772840383931744,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9444888888888892e-05,
+      "loss": 0.483,
+      "step": 1250
+    },
+    {
+      "epoch": 0.0279950231070032,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9440444444444447e-05,
+      "loss": 0.5556,
+      "step": 1260
+    },
+    {
+      "epoch": 0.028217205830074654,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.9436e-05,
+      "loss": 0.5281,
+      "step": 1270
+    },
+    {
+      "epoch": 0.028439388553146108,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.9431555555555556e-05,
+      "loss": 0.4806,
+      "step": 1280
+    },
+    {
+      "epoch": 0.028661571276217563,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.942711111111111e-05,
+      "loss": 0.5233,
+      "step": 1290
+    },
+    {
+      "epoch": 0.028883753999289014,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9422666666666666e-05,
+      "loss": 0.5057,
+      "step": 1300
+    },
+    {
+      "epoch": 0.02910593672236047,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9418222222222224e-05,
+      "loss": 0.4829,
+      "step": 1310
+    },
+    {
+      "epoch": 0.029328119445431924,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.941377777777778e-05,
+      "loss": 0.4969,
+      "step": 1320
+    },
+    {
+      "epoch": 0.02955030216850338,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9409333333333334e-05,
+      "loss": 0.5034,
+      "step": 1330
+    },
+    {
+      "epoch": 0.02977248489157483,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9404888888888892e-05,
+      "loss": 0.5165,
+      "step": 1340
+    },
+    {
+      "epoch": 0.029994667614646284,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9400444444444447e-05,
+      "loss": 0.4964,
+      "step": 1350
+    },
+    {
+      "epoch": 0.03021685033771774,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9396000000000002e-05,
+      "loss": 0.4331,
+      "step": 1360
+    },
+    {
+      "epoch": 0.030439033060789194,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9391555555555557e-05,
+      "loss": 0.4651,
+      "step": 1370
+    },
+    {
+      "epoch": 0.03066121578386065,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.938711111111111e-05,
+      "loss": 0.48,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0308833985069321,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9382666666666667e-05,
+      "loss": 0.4874,
+      "step": 1390
+    },
+    {
+      "epoch": 0.031105581230003555,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.937822222222222e-05,
+      "loss": 0.4431,
+      "step": 1400
+    },
+    {
+      "epoch": 0.03132776395307501,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.937377777777778e-05,
+      "loss": 0.5207,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03154994667614646,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9369333333333334e-05,
+      "loss": 0.5071,
+      "step": 1420
+    },
+    {
+      "epoch": 0.03177212939921792,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.936488888888889e-05,
+      "loss": 0.5093,
+      "step": 1430
+    },
+    {
+      "epoch": 0.03199431212228937,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9360444444444448e-05,
+      "loss": 0.4823,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03221649484536082,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9356000000000002e-05,
+      "loss": 0.4658,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03243867756843228,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9351555555555557e-05,
+      "loss": 0.5124,
+      "step": 1460
+    },
+    {
+      "epoch": 0.03266086029150373,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9347111111111112e-05,
+      "loss": 0.513,
+      "step": 1470
+    },
+    {
+      "epoch": 0.03288304301457519,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9342666666666667e-05,
+      "loss": 0.4456,
+      "step": 1480
+    },
+    {
+      "epoch": 0.03310522573764664,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9338222222222225e-05,
+      "loss": 0.5171,
+      "step": 1490
+    },
+    {
+      "epoch": 0.03332740846071809,
+      "grad_norm": 1.875,
+      "learning_rate": 1.933377777777778e-05,
+      "loss": 0.4782,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03354959118378955,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9329333333333335e-05,
+      "loss": 0.4689,
+      "step": 1510
+    },
+    {
+      "epoch": 0.033771773906861,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.932488888888889e-05,
+      "loss": 0.506,
+      "step": 1520
+    },
+    {
+      "epoch": 0.03399395662993246,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9320444444444445e-05,
+      "loss": 0.5159,
+      "step": 1530
+    },
+    {
+      "epoch": 0.03421613935300391,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9316000000000003e-05,
+      "loss": 0.504,
+      "step": 1540
+    },
+    {
+      "epoch": 0.03443832207607536,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9311555555555558e-05,
+      "loss": 0.5043,
+      "step": 1550
+    },
+    {
+      "epoch": 0.03466050479914682,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9307111111111112e-05,
+      "loss": 0.4947,
+      "step": 1560
+    },
+    {
+      "epoch": 0.03488268752221827,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.9302666666666667e-05,
+      "loss": 0.4833,
+      "step": 1570
+    },
+    {
+      "epoch": 0.03510487024528973,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9298222222222226e-05,
+      "loss": 0.4666,
+      "step": 1580
+    },
+    {
+      "epoch": 0.03532705296836118,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.929377777777778e-05,
+      "loss": 0.4763,
+      "step": 1590
+    },
+    {
+      "epoch": 0.03554923569143263,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9289333333333335e-05,
+      "loss": 0.519,
+      "step": 1600
+    },
+    {
+      "epoch": 0.03577141841450409,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.928488888888889e-05,
+      "loss": 0.5078,
+      "step": 1610
+    },
+    {
+      "epoch": 0.03599360113757554,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9280444444444445e-05,
+      "loss": 0.5186,
+      "step": 1620
+    },
+    {
+      "epoch": 0.036215783860647,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9276e-05,
+      "loss": 0.4755,
+      "step": 1630
+    },
+    {
+      "epoch": 0.03643796658371845,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9271555555555558e-05,
+      "loss": 0.4768,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0366601493067899,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9267111111111113e-05,
+      "loss": 0.5004,
+      "step": 1650
+    },
+    {
+      "epoch": 0.03688233202986136,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.9262666666666668e-05,
+      "loss": 0.5269,
+      "step": 1660
+    },
+    {
+      "epoch": 0.03710451475293281,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9258222222222226e-05,
+      "loss": 0.4504,
+      "step": 1670
+    },
+    {
+      "epoch": 0.03732669747600426,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.925377777777778e-05,
+      "loss": 0.5188,
+      "step": 1680
+    },
+    {
+      "epoch": 0.03754888019907572,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9249333333333336e-05,
+      "loss": 0.4665,
+      "step": 1690
+    },
+    {
+      "epoch": 0.03777106292214717,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.924488888888889e-05,
+      "loss": 0.5029,
+      "step": 1700
+    },
+    {
+      "epoch": 0.03799324564521863,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9240444444444445e-05,
+      "loss": 0.5019,
+      "step": 1710
+    },
+    {
+      "epoch": 0.03821542836829008,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9236e-05,
+      "loss": 0.4716,
+      "step": 1720
+    },
+    {
+      "epoch": 0.03843761109136153,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9231555555555555e-05,
+      "loss": 0.4626,
+      "step": 1730
+    },
+    {
+      "epoch": 0.03865979381443299,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9227111111111113e-05,
+      "loss": 0.4727,
+      "step": 1740
+    },
+    {
+      "epoch": 0.03888197653750444,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9222666666666668e-05,
+      "loss": 0.5317,
+      "step": 1750
+    },
+    {
+      "epoch": 0.0391041592605759,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9218222222222226e-05,
+      "loss": 0.4401,
+      "step": 1760
+    },
+    {
+      "epoch": 0.03932634198364735,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.921377777777778e-05,
+      "loss": 0.4829,
+      "step": 1770
+    },
+    {
+      "epoch": 0.0395485247067188,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9209333333333336e-05,
+      "loss": 0.4393,
+      "step": 1780
+    },
+    {
+      "epoch": 0.03977070742979026,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.920488888888889e-05,
+      "loss": 0.4402,
+      "step": 1790
+    },
+    {
+      "epoch": 0.03999289015286171,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9200444444444446e-05,
+      "loss": 0.491,
+      "step": 1800
+    },
+    {
+      "epoch": 0.04021507287593317,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.9196e-05,
+      "loss": 0.487,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04043725559900462,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9191555555555555e-05,
+      "loss": 0.5038,
+      "step": 1820
+    },
+    {
+      "epoch": 0.04065943832207607,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9187111111111114e-05,
+      "loss": 0.4536,
+      "step": 1830
+    },
+    {
+      "epoch": 0.04088162104514753,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.918266666666667e-05,
+      "loss": 0.5185,
+      "step": 1840
+    },
+    {
+      "epoch": 0.04110380376821898,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9178222222222223e-05,
+      "loss": 0.4818,
+      "step": 1850
+    },
+    {
+      "epoch": 0.041325986491290434,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.917377777777778e-05,
+      "loss": 0.4608,
+      "step": 1860
+    },
+    {
+      "epoch": 0.04154816921436189,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9169333333333336e-05,
+      "loss": 0.4942,
+      "step": 1870
+    },
+    {
+      "epoch": 0.04177035193743334,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.916488888888889e-05,
+      "loss": 0.4829,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0419925346605048,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9160444444444446e-05,
+      "loss": 0.4504,
+      "step": 1890
+    },
+    {
+      "epoch": 0.04221471738357625,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.9156e-05,
+      "loss": 0.491,
+      "step": 1900
+    },
+    {
+      "epoch": 0.042436900106647704,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9151555555555556e-05,
+      "loss": 0.486,
+      "step": 1910
+    },
+    {
+      "epoch": 0.04265908282971916,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.914711111111111e-05,
+      "loss": 0.4814,
+      "step": 1920
+    },
+    {
+      "epoch": 0.042881265552790614,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.914266666666667e-05,
+      "loss": 0.4853,
+      "step": 1930
+    },
+    {
+      "epoch": 0.04310344827586207,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9138222222222224e-05,
+      "loss": 0.5006,
+      "step": 1940
+    },
+    {
+      "epoch": 0.04332563099893352,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.913377777777778e-05,
+      "loss": 0.5142,
+      "step": 1950
+    },
+    {
+      "epoch": 0.043547813722004974,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9129333333333337e-05,
+      "loss": 0.4764,
+      "step": 1960
+    },
+    {
+      "epoch": 0.04376999644507643,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.912488888888889e-05,
+      "loss": 0.4343,
+      "step": 1970
+    },
+    {
+      "epoch": 0.043992179168147884,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9120444444444446e-05,
+      "loss": 0.5007,
+      "step": 1980
+    },
+    {
+      "epoch": 0.04421436189121934,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.9116e-05,
+      "loss": 0.5271,
+      "step": 1990
+    },
+    {
+      "epoch": 0.04443654461429079,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9111555555555556e-05,
+      "loss": 0.4994,
+      "step": 2000
+    },
+    {
+      "epoch": 0.044658727337362245,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.910711111111111e-05,
+      "loss": 0.4418,
+      "step": 2010
+    },
+    {
+      "epoch": 0.0448809100604337,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9102666666666666e-05,
+      "loss": 0.4824,
+      "step": 2020
+    },
+    {
+      "epoch": 0.045103092783505154,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9098222222222224e-05,
+      "loss": 0.4577,
+      "step": 2030
+    },
+    {
+      "epoch": 0.045325275506576605,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.909377777777778e-05,
+      "loss": 0.4742,
+      "step": 2040
+    },
+    {
+      "epoch": 0.04554745822964806,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9089333333333334e-05,
+      "loss": 0.4691,
+      "step": 2050
+    },
+    {
+      "epoch": 0.045769640952719515,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9084888888888892e-05,
+      "loss": 0.4703,
+      "step": 2060
+    },
+    {
+      "epoch": 0.04599182367579097,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9080444444444447e-05,
+      "loss": 0.4721,
+      "step": 2070
+    },
+    {
+      "epoch": 0.046214006398862424,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9076e-05,
+      "loss": 0.4702,
+      "step": 2080
+    },
+    {
+      "epoch": 0.046436189121933875,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9071555555555557e-05,
+      "loss": 0.4956,
+      "step": 2090
+    },
+    {
+      "epoch": 0.046658371845005334,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.906711111111111e-05,
+      "loss": 0.4889,
+      "step": 2100
+    },
+    {
+      "epoch": 0.046880554568076785,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.906266666666667e-05,
+      "loss": 0.4642,
+      "step": 2110
+    },
+    {
+      "epoch": 0.04710273729114824,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9058222222222224e-05,
+      "loss": 0.4673,
+      "step": 2120
+    },
+    {
+      "epoch": 0.047324920014219694,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.905377777777778e-05,
+      "loss": 0.478,
+      "step": 2130
+    },
+    {
+      "epoch": 0.047547102737291146,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9049333333333334e-05,
+      "loss": 0.4863,
+      "step": 2140
+    },
+    {
+      "epoch": 0.047769285460362604,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.904488888888889e-05,
+      "loss": 0.4782,
+      "step": 2150
+    },
+    {
+      "epoch": 0.047991468183434055,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9040444444444447e-05,
+      "loss": 0.4336,
+      "step": 2160
+    },
+    {
+      "epoch": 0.04821365090650551,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9036000000000002e-05,
+      "loss": 0.4448,
+      "step": 2170
+    },
+    {
+      "epoch": 0.048435833629576965,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9031555555555557e-05,
+      "loss": 0.4361,
+      "step": 2180
+    },
+    {
+      "epoch": 0.048658016352648416,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9027111111111112e-05,
+      "loss": 0.4494,
+      "step": 2190
+    },
+    {
+      "epoch": 0.048880199075719874,
+      "grad_norm": 1.75,
+      "learning_rate": 1.902266666666667e-05,
+      "loss": 0.4557,
+      "step": 2200
+    },
+    {
+      "epoch": 0.049102381798791325,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9018222222222225e-05,
+      "loss": 0.4584,
+      "step": 2210
+    },
+    {
+      "epoch": 0.049324564521862777,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.901377777777778e-05,
+      "loss": 0.5059,
+      "step": 2220
+    },
+    {
+      "epoch": 0.049546747244934235,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.9009333333333335e-05,
+      "loss": 0.4931,
+      "step": 2230
+    },
+    {
+      "epoch": 0.049768929968005686,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.900488888888889e-05,
+      "loss": 0.4557,
+      "step": 2240
+    },
+    {
+      "epoch": 0.049991112691077144,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9000444444444444e-05,
+      "loss": 0.4672,
+      "step": 2250
+    },
+    {
+      "epoch": 0.050213295414148595,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8996000000000002e-05,
+      "loss": 0.4516,
+      "step": 2260
+    },
+    {
+      "epoch": 0.05043547813722005,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.8991555555555557e-05,
+      "loss": 0.4555,
+      "step": 2270
+    },
+    {
+      "epoch": 0.050657660860291505,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.8987111111111112e-05,
+      "loss": 0.453,
+      "step": 2280
+    },
+    {
+      "epoch": 0.050879843583362956,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.898266666666667e-05,
+      "loss": 0.4844,
+      "step": 2290
+    },
+    {
+      "epoch": 0.051102026306434414,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.8978222222222225e-05,
+      "loss": 0.4848,
+      "step": 2300
+    },
+    {
+      "epoch": 0.051324209029505866,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.897377777777778e-05,
+      "loss": 0.4426,
+      "step": 2310
+    },
+    {
+      "epoch": 0.05154639175257732,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.8969333333333335e-05,
+      "loss": 0.4427,
+      "step": 2320
+    },
+    {
+      "epoch": 0.051768574475648775,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.896488888888889e-05,
+      "loss": 0.4774,
+      "step": 2330
+    },
+    {
+      "epoch": 0.051990757198720226,
+      "grad_norm": 1.875,
+      "learning_rate": 1.8960444444444445e-05,
+      "loss": 0.4683,
+      "step": 2340
+    },
+    {
+      "epoch": 0.052212939921791685,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.8956e-05,
+      "loss": 0.4915,
+      "step": 2350
+    },
+    {
+      "epoch": 0.052435122644863136,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8951555555555558e-05,
+      "loss": 0.4971,
+      "step": 2360
+    },
+    {
+      "epoch": 0.05265730536793459,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.8947111111111113e-05,
+      "loss": 0.4544,
+      "step": 2370
+    },
+    {
+      "epoch": 0.052879488091006045,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.894266666666667e-05,
+      "loss": 0.4575,
+      "step": 2380
+    },
+    {
+      "epoch": 0.053101670814077497,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8938222222222226e-05,
+      "loss": 0.5068,
+      "step": 2390
+    },
+    {
+      "epoch": 0.053323853537148955,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.893377777777778e-05,
+      "loss": 0.4863,
+      "step": 2400
+    },
+    {
+      "epoch": 0.053546036260220406,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.8929333333333335e-05,
+      "loss": 0.4527,
+      "step": 2410
+    },
+    {
+      "epoch": 0.05376821898329186,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.892488888888889e-05,
+      "loss": 0.5141,
+      "step": 2420
+    },
+    {
+      "epoch": 0.053990401706363315,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.8920444444444445e-05,
+      "loss": 0.478,
+      "step": 2430
+    },
+    {
+      "epoch": 0.05421258442943477,
+      "grad_norm": 2.125,
+      "learning_rate": 1.8916e-05,
+      "loss": 0.4193,
+      "step": 2440
+    },
+    {
+      "epoch": 0.05443476715250622,
+      "grad_norm": 2.125,
+      "learning_rate": 1.8911555555555555e-05,
+      "loss": 0.4617,
+      "step": 2450
+    },
+    {
+      "epoch": 0.054656949875577676,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.8907111111111113e-05,
+      "loss": 0.4434,
+      "step": 2460
+    },
+    {
+      "epoch": 0.05487913259864913,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.8902666666666668e-05,
+      "loss": 0.456,
+      "step": 2470
+    },
+    {
+      "epoch": 0.055101315321720586,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8898222222222226e-05,
+      "loss": 0.4581,
+      "step": 2480
+    },
+    {
+      "epoch": 0.05532349804479204,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.889377777777778e-05,
+      "loss": 0.5004,
+      "step": 2490
+    },
+    {
+      "epoch": 0.05554568076786349,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8889333333333336e-05,
+      "loss": 0.4618,
+      "step": 2500
+    },
+    {
+      "epoch": 0.055767863490934946,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.888488888888889e-05,
+      "loss": 0.4335,
+      "step": 2510
+    },
+    {
+      "epoch": 0.0559900462140064,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8880444444444445e-05,
+      "loss": 0.4709,
+      "step": 2520
+    },
+    {
+      "epoch": 0.056212228937077856,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.8876e-05,
+      "loss": 0.4395,
+      "step": 2530
+    },
+    {
+      "epoch": 0.05643441166014931,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8871555555555555e-05,
+      "loss": 0.4741,
+      "step": 2540
+    },
+    {
+      "epoch": 0.05665659438322076,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8867111111111113e-05,
+      "loss": 0.4438,
+      "step": 2550
+    },
+    {
+      "epoch": 0.056878777106292217,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8862666666666668e-05,
+      "loss": 0.444,
+      "step": 2560
+    },
+    {
+      "epoch": 0.05710095982936367,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.8858222222222223e-05,
+      "loss": 0.4554,
+      "step": 2570
+    },
+    {
+      "epoch": 0.057323142552435126,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.885377777777778e-05,
+      "loss": 0.507,
+      "step": 2580
+    },
+    {
+      "epoch": 0.05754532527550658,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8849333333333336e-05,
+      "loss": 0.4842,
+      "step": 2590
+    },
+    {
+      "epoch": 0.05776750799857803,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.884488888888889e-05,
+      "loss": 0.4443,
+      "step": 2600
+    },
+    {
+      "epoch": 0.05798969072164949,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.8840444444444446e-05,
+      "loss": 0.4685,
+      "step": 2610
+    },
+    {
+      "epoch": 0.05821187344472094,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.8836e-05,
+      "loss": 0.4668,
+      "step": 2620
+    },
+    {
+      "epoch": 0.05843405616779239,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8831555555555555e-05,
+      "loss": 0.4316,
+      "step": 2630
+    },
+    {
+      "epoch": 0.05865623889086385,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.882711111111111e-05,
+      "loss": 0.4482,
+      "step": 2640
+    },
+    {
+      "epoch": 0.0588784216139353,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.882266666666667e-05,
+      "loss": 0.4789,
+      "step": 2650
+    },
+    {
+      "epoch": 0.05910060433700676,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.8818222222222223e-05,
+      "loss": 0.4504,
+      "step": 2660
+    },
+    {
+      "epoch": 0.05932278706007821,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8813777777777778e-05,
+      "loss": 0.486,
+      "step": 2670
+    },
+    {
+      "epoch": 0.05954496978314966,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.8809333333333336e-05,
+      "loss": 0.4454,
+      "step": 2680
+    },
+    {
+      "epoch": 0.05976715250622112,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.880488888888889e-05,
+      "loss": 0.4169,
+      "step": 2690
+    },
+    {
+      "epoch": 0.05998933522929257,
+      "grad_norm": 1.875,
+      "learning_rate": 1.8800444444444446e-05,
+      "loss": 0.4754,
+      "step": 2700
+    },
+    {
+      "epoch": 0.06021151795236403,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8796e-05,
+      "loss": 0.5018,
+      "step": 2710
+    },
+    {
+      "epoch": 0.06043370067543548,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.8791555555555556e-05,
+      "loss": 0.4931,
+      "step": 2720
+    },
+    {
+      "epoch": 0.06065588339850693,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.8787111111111114e-05,
+      "loss": 0.4725,
+      "step": 2730
+    },
+    {
+      "epoch": 0.06087806612157839,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.878266666666667e-05,
+      "loss": 0.4276,
+      "step": 2740
+    },
+    {
+      "epoch": 0.06110024884464984,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.8778222222222224e-05,
+      "loss": 0.4817,
+      "step": 2750
+    },
+    {
+      "epoch": 0.0613224315677213,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.877377777777778e-05,
+      "loss": 0.4638,
+      "step": 2760
+    },
+    {
+      "epoch": 0.06154461429079275,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.8769333333333333e-05,
+      "loss": 0.4529,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0617667970138642,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.876488888888889e-05,
+      "loss": 0.4779,
+      "step": 2780
+    },
+    {
+      "epoch": 0.06198897973693566,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.8760444444444447e-05,
+      "loss": 0.4788,
+      "step": 2790
+    },
+    {
+      "epoch": 0.06221116246000711,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.8756e-05,
+      "loss": 0.4513,
+      "step": 2800
+    },
+    {
+      "epoch": 0.06243334518307856,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8751555555555556e-05,
+      "loss": 0.4661,
+      "step": 2810
+    },
+    {
+      "epoch": 0.06265552790615002,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.8747111111111114e-05,
+      "loss": 0.4529,
+      "step": 2820
+    },
+    {
+      "epoch": 0.06287771062922147,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.874266666666667e-05,
+      "loss": 0.4556,
+      "step": 2830
+    },
+    {
+      "epoch": 0.06309989335229292,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.8738222222222224e-05,
+      "loss": 0.4645,
+      "step": 2840
+    },
+    {
+      "epoch": 0.06332207607536439,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.873377777777778e-05,
+      "loss": 0.3935,
+      "step": 2850
+    },
+    {
+      "epoch": 0.06354425879843584,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.8729333333333334e-05,
+      "loss": 0.4634,
+      "step": 2860
+    },
+    {
+      "epoch": 0.06376644152150729,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.872488888888889e-05,
+      "loss": 0.4722,
+      "step": 2870
+    },
+    {
+      "epoch": 0.06398862424457874,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.8720444444444447e-05,
+      "loss": 0.4674,
+      "step": 2880
+    },
+    {
+      "epoch": 0.06421080696765019,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8716000000000002e-05,
+      "loss": 0.4621,
+      "step": 2890
+    },
+    {
+      "epoch": 0.06443298969072164,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8711555555555557e-05,
+      "loss": 0.4584,
+      "step": 2900
+    },
+    {
+      "epoch": 0.06465517241379311,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8707111111111115e-05,
+      "loss": 0.4808,
+      "step": 2910
+    },
+    {
+      "epoch": 0.06487735513686456,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.870266666666667e-05,
+      "loss": 0.4729,
+      "step": 2920
+    },
+    {
+      "epoch": 0.06509953785993601,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.8698222222222225e-05,
+      "loss": 0.4282,
+      "step": 2930
+    },
+    {
+      "epoch": 0.06532172058300746,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.869377777777778e-05,
+      "loss": 0.5072,
+      "step": 2940
+    },
+    {
+      "epoch": 0.06554390330607891,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.8689333333333334e-05,
+      "loss": 0.4563,
+      "step": 2950
+    },
+    {
+      "epoch": 0.06576608602915038,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.868488888888889e-05,
+      "loss": 0.4412,
+      "step": 2960
+    },
+    {
+      "epoch": 0.06598826875222183,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.8680444444444444e-05,
+      "loss": 0.4591,
+      "step": 2970
+    },
+    {
+      "epoch": 0.06621045147529328,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.8676000000000002e-05,
+      "loss": 0.4204,
+      "step": 2980
+    },
+    {
+      "epoch": 0.06643263419836473,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.8671555555555557e-05,
+      "loss": 0.4324,
+      "step": 2990
+    },
+    {
+      "epoch": 0.06665481692143618,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.8667111111111115e-05,
+      "loss": 0.434,
+      "step": 3000
+    },
+    {
+      "epoch": 0.06687699964450765,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.866266666666667e-05,
+      "loss": 0.4756,
+      "step": 3010
+    },
+    {
+      "epoch": 0.0670991823675791,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.8658222222222225e-05,
+      "loss": 0.4765,
+      "step": 3020
+    },
+    {
+      "epoch": 0.06732136509065055,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.865377777777778e-05,
+      "loss": 0.4121,
+      "step": 3030
+    },
+    {
+      "epoch": 0.067543547813722,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8649333333333335e-05,
+      "loss": 0.4742,
+      "step": 3040
+    },
+    {
+      "epoch": 0.06776573053679345,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.864488888888889e-05,
+      "loss": 0.4649,
+      "step": 3050
+    },
+    {
+      "epoch": 0.06798791325986492,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.8640444444444444e-05,
+      "loss": 0.472,
+      "step": 3060
+    },
+    {
+      "epoch": 0.06821009598293637,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.8636e-05,
+      "loss": 0.4169,
+      "step": 3070
+    },
+    {
+      "epoch": 0.06843227870600782,
+      "grad_norm": 2.125,
+      "learning_rate": 1.8631555555555557e-05,
+      "loss": 0.4786,
+      "step": 3080
+    },
+    {
+      "epoch": 0.06865446142907927,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8627111111111112e-05,
+      "loss": 0.4465,
+      "step": 3090
+    },
+    {
+      "epoch": 0.06887664415215072,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.862266666666667e-05,
+      "loss": 0.4672,
+      "step": 3100
+    },
+    {
+      "epoch": 0.06909882687522219,
+      "grad_norm": 2.375,
+      "learning_rate": 1.8618222222222225e-05,
+      "loss": 0.4604,
+      "step": 3110
+    },
+    {
+      "epoch": 0.06932100959829364,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.861377777777778e-05,
+      "loss": 0.4389,
+      "step": 3120
+    },
+    {
+      "epoch": 0.06954319232136509,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.8609333333333335e-05,
+      "loss": 0.4301,
+      "step": 3130
+    },
+    {
+      "epoch": 0.06976537504443654,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.860488888888889e-05,
+      "loss": 0.4703,
+      "step": 3140
+    },
+    {
+      "epoch": 0.069987557767508,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.8600444444444445e-05,
+      "loss": 0.4542,
+      "step": 3150
+    },
+    {
+      "epoch": 0.07020974049057946,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.8596e-05,
+      "loss": 0.4301,
+      "step": 3160
+    },
+    {
+      "epoch": 0.07043192321365091,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.8591555555555554e-05,
+      "loss": 0.453,
+      "step": 3170
+    },
+    {
+      "epoch": 0.07065410593672236,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8587111111111113e-05,
+      "loss": 0.4156,
+      "step": 3180
+    },
+    {
+      "epoch": 0.07087628865979381,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.8582666666666667e-05,
+      "loss": 0.4333,
+      "step": 3190
+    },
+    {
+      "epoch": 0.07109847138286526,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8578222222222226e-05,
+      "loss": 0.4398,
+      "step": 3200
+    },
+    {
+      "epoch": 0.07132065410593673,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.857377777777778e-05,
+      "loss": 0.4683,
+      "step": 3210
+    },
+    {
+      "epoch": 0.07154283682900818,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8569333333333335e-05,
+      "loss": 0.4475,
+      "step": 3220
+    },
+    {
+      "epoch": 0.07176501955207963,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.856488888888889e-05,
+      "loss": 0.4453,
+      "step": 3230
+    },
+    {
+      "epoch": 0.07198720227515108,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.8560444444444445e-05,
+      "loss": 0.4922,
+      "step": 3240
+    },
+    {
+      "epoch": 0.07220938499822253,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.8556e-05,
+      "loss": 0.492,
+      "step": 3250
+    },
+    {
+      "epoch": 0.072431567721294,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.8551555555555555e-05,
+      "loss": 0.4376,
+      "step": 3260
+    },
+    {
+      "epoch": 0.07265375044436545,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.8547111111111113e-05,
+      "loss": 0.4818,
+      "step": 3270
+    },
+    {
+      "epoch": 0.0728759331674369,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.8542666666666668e-05,
+      "loss": 0.4779,
+      "step": 3280
+    },
+    {
+      "epoch": 0.07309811589050835,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8538222222222223e-05,
+      "loss": 0.4235,
+      "step": 3290
+    },
+    {
+      "epoch": 0.0733202986135798,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.853377777777778e-05,
+      "loss": 0.4584,
+      "step": 3300
+    },
+    {
+      "epoch": 0.07354248133665126,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.8529333333333336e-05,
+      "loss": 0.4687,
+      "step": 3310
+    },
+    {
+      "epoch": 0.07376466405972272,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.852488888888889e-05,
+      "loss": 0.4481,
+      "step": 3320
+    },
+    {
+      "epoch": 0.07398684678279417,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8520444444444445e-05,
+      "loss": 0.4901,
+      "step": 3330
+    },
+    {
+      "epoch": 0.07420902950586562,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.8516e-05,
+      "loss": 0.4399,
+      "step": 3340
+    },
+    {
+      "epoch": 0.07443121222893707,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.851155555555556e-05,
+      "loss": 0.4367,
+      "step": 3350
+    },
+    {
+      "epoch": 0.07465339495200853,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.8507111111111113e-05,
+      "loss": 0.4724,
+      "step": 3360
+    },
+    {
+      "epoch": 0.07487557767507999,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.8502666666666668e-05,
+      "loss": 0.4534,
+      "step": 3370
+    },
+    {
+      "epoch": 0.07509776039815144,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8498222222222223e-05,
+      "loss": 0.4649,
+      "step": 3380
+    },
+    {
+      "epoch": 0.0753199431212229,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.8493777777777778e-05,
+      "loss": 0.4539,
+      "step": 3390
+    },
+    {
+      "epoch": 0.07554212584429434,
+      "grad_norm": 1.75,
+      "learning_rate": 1.8489333333333336e-05,
+      "loss": 0.4484,
+      "step": 3400
+    },
+    {
+      "epoch": 0.0757643085673658,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.848488888888889e-05,
+      "loss": 0.4544,
+      "step": 3410
+    },
+    {
+      "epoch": 0.07598649129043726,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.8480444444444446e-05,
+      "loss": 0.4431,
+      "step": 3420
+    },
+    {
+      "epoch": 0.07620867401350871,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.8476e-05,
+      "loss": 0.437,
+      "step": 3430
+    },
+    {
+      "epoch": 0.07643085673658016,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.847155555555556e-05,
+      "loss": 0.458,
+      "step": 3440
+    },
+    {
+      "epoch": 0.07665303945965161,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8467111111111114e-05,
+      "loss": 0.4512,
+      "step": 3450
+    },
+    {
+      "epoch": 0.07687522218272307,
+      "grad_norm": 2.375,
+      "learning_rate": 1.846266666666667e-05,
+      "loss": 0.435,
+      "step": 3460
+    },
+    {
+      "epoch": 0.07709740490579453,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.8458222222222223e-05,
+      "loss": 0.4778,
+      "step": 3470
+    },
+    {
+      "epoch": 0.07731958762886598,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.8453777777777778e-05,
+      "loss": 0.446,
+      "step": 3480
+    },
+    {
+      "epoch": 0.07754177035193743,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8449333333333333e-05,
+      "loss": 0.4624,
+      "step": 3490
+    },
+    {
+      "epoch": 0.07776395307500888,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.844488888888889e-05,
+      "loss": 0.4458,
+      "step": 3500
+    },
+    {
+      "epoch": 0.07798613579808034,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.8440444444444446e-05,
+      "loss": 0.4879,
+      "step": 3510
+    },
+    {
+      "epoch": 0.0782083185211518,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.8436e-05,
+      "loss": 0.4646,
+      "step": 3520
+    },
+    {
+      "epoch": 0.07843050124422325,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.843155555555556e-05,
+      "loss": 0.4816,
+      "step": 3530
+    },
+    {
+      "epoch": 0.0786526839672947,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.8427111111111114e-05,
+      "loss": 0.4507,
+      "step": 3540
+    },
+    {
+      "epoch": 0.07887486669036615,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.842266666666667e-05,
+      "loss": 0.425,
+      "step": 3550
+    },
+    {
+      "epoch": 0.0790970494134376,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.8418222222222224e-05,
+      "loss": 0.4619,
+      "step": 3560
+    },
+    {
+      "epoch": 0.07931923213650907,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.841377777777778e-05,
+      "loss": 0.4677,
+      "step": 3570
+    },
+    {
+      "epoch": 0.07954141485958052,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8409333333333333e-05,
+      "loss": 0.4423,
+      "step": 3580
+    },
+    {
+      "epoch": 0.07976359758265197,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.840488888888889e-05,
+      "loss": 0.4354,
+      "step": 3590
+    },
+    {
+      "epoch": 0.07998578030572343,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.8400444444444447e-05,
+      "loss": 0.4159,
+      "step": 3600
+    },
+    {
+      "epoch": 0.08020796302879488,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8396e-05,
+      "loss": 0.4579,
+      "step": 3610
+    },
+    {
+      "epoch": 0.08043014575186634,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.839155555555556e-05,
+      "loss": 0.4623,
+      "step": 3620
+    },
+    {
+      "epoch": 0.08065232847493779,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.8387111111111114e-05,
+      "loss": 0.4846,
+      "step": 3630
+    },
+    {
+      "epoch": 0.08087451119800924,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.838266666666667e-05,
+      "loss": 0.4815,
+      "step": 3640
+    },
+    {
+      "epoch": 0.0810966939210807,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.8378222222222224e-05,
+      "loss": 0.4374,
+      "step": 3650
+    },
+    {
+      "epoch": 0.08131887664415215,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.837377777777778e-05,
+      "loss": 0.454,
+      "step": 3660
+    },
+    {
+      "epoch": 0.0815410593672236,
+      "grad_norm": 1.75,
+      "learning_rate": 1.8369333333333334e-05,
+      "loss": 0.4526,
+      "step": 3670
+    },
+    {
+      "epoch": 0.08176324209029506,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.836488888888889e-05,
+      "loss": 0.4432,
+      "step": 3680
+    },
+    {
+      "epoch": 0.08198542481336651,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8360444444444444e-05,
+      "loss": 0.4295,
+      "step": 3690
+    },
+    {
+      "epoch": 0.08220760753643797,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.8356000000000002e-05,
+      "loss": 0.4415,
+      "step": 3700
+    },
+    {
+      "epoch": 0.08242979025950942,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.8351555555555557e-05,
+      "loss": 0.43,
+      "step": 3710
+    },
+    {
+      "epoch": 0.08265197298258087,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.8347111111111115e-05,
+      "loss": 0.4624,
+      "step": 3720
+    },
+    {
+      "epoch": 0.08287415570565233,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.834266666666667e-05,
+      "loss": 0.4405,
+      "step": 3730
+    },
+    {
+      "epoch": 0.08309633842872378,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.8338222222222225e-05,
+      "loss": 0.4621,
+      "step": 3740
+    },
+    {
+      "epoch": 0.08331852115179524,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.833377777777778e-05,
+      "loss": 0.4183,
+      "step": 3750
+    },
+    {
+      "epoch": 0.08354070387486669,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.8329333333333334e-05,
+      "loss": 0.49,
+      "step": 3760
+    },
+    {
+      "epoch": 0.08376288659793814,
+      "grad_norm": 2.125,
+      "learning_rate": 1.832488888888889e-05,
+      "loss": 0.4697,
+      "step": 3770
+    },
+    {
+      "epoch": 0.0839850693210096,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8320444444444444e-05,
+      "loss": 0.4161,
+      "step": 3780
+    },
+    {
+      "epoch": 0.08420725204408105,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.8316e-05,
+      "loss": 0.4223,
+      "step": 3790
+    },
+    {
+      "epoch": 0.0844294347671525,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.8311555555555557e-05,
+      "loss": 0.4626,
+      "step": 3800
+    },
+    {
+      "epoch": 0.08465161749022396,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.8307111111111112e-05,
+      "loss": 0.4782,
+      "step": 3810
+    },
+    {
+      "epoch": 0.08487380021329541,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.830266666666667e-05,
+      "loss": 0.4219,
+      "step": 3820
+    },
+    {
+      "epoch": 0.08509598293636687,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8298222222222225e-05,
+      "loss": 0.4611,
+      "step": 3830
+    },
+    {
+      "epoch": 0.08531816565943832,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.829377777777778e-05,
+      "loss": 0.4372,
+      "step": 3840
+    },
+    {
+      "epoch": 0.08554034838250978,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8289333333333335e-05,
+      "loss": 0.4556,
+      "step": 3850
+    },
+    {
+      "epoch": 0.08576253110558123,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.828488888888889e-05,
+      "loss": 0.4458,
+      "step": 3860
+    },
+    {
+      "epoch": 0.08598471382865268,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.8280444444444444e-05,
+      "loss": 0.4566,
+      "step": 3870
+    },
+    {
+      "epoch": 0.08620689655172414,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8276e-05,
+      "loss": 0.4049,
+      "step": 3880
+    },
+    {
+      "epoch": 0.0864290792747956,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.8271555555555557e-05,
+      "loss": 0.4401,
+      "step": 3890
+    },
+    {
+      "epoch": 0.08665126199786705,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.8267111111111112e-05,
+      "loss": 0.4426,
+      "step": 3900
+    },
+    {
+      "epoch": 0.0868734447209385,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8262666666666667e-05,
+      "loss": 0.4529,
+      "step": 3910
+    },
+    {
+      "epoch": 0.08709562744400995,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.8258222222222225e-05,
+      "loss": 0.4444,
+      "step": 3920
+    },
+    {
+      "epoch": 0.08731781016708141,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.825377777777778e-05,
+      "loss": 0.4486,
+      "step": 3930
+    },
+    {
+      "epoch": 0.08753999289015287,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.8249333333333335e-05,
+      "loss": 0.4299,
+      "step": 3940
+    },
+    {
+      "epoch": 0.08776217561322432,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.824488888888889e-05,
+      "loss": 0.4385,
+      "step": 3950
+    },
+    {
+      "epoch": 0.08798435833629577,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.8240444444444445e-05,
+      "loss": 0.4493,
+      "step": 3960
+    },
+    {
+      "epoch": 0.08820654105936722,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8236000000000003e-05,
+      "loss": 0.455,
+      "step": 3970
+    },
+    {
+      "epoch": 0.08842872378243868,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.8231555555555558e-05,
+      "loss": 0.4461,
+      "step": 3980
+    },
+    {
+      "epoch": 0.08865090650551014,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8227111111111113e-05,
+      "loss": 0.4066,
+      "step": 3990
+    },
+    {
+      "epoch": 0.08887308922858159,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8222666666666667e-05,
+      "loss": 0.4344,
+      "step": 4000
+    },
+    {
+      "epoch": 0.08909527195165304,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.8218222222222222e-05,
+      "loss": 0.4603,
+      "step": 4010
+    },
+    {
+      "epoch": 0.08931745467472449,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.821377777777778e-05,
+      "loss": 0.4158,
+      "step": 4020
+    },
+    {
+      "epoch": 0.08953963739779595,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.8209333333333335e-05,
+      "loss": 0.4181,
+      "step": 4030
+    },
+    {
+      "epoch": 0.0897618201208674,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.820488888888889e-05,
+      "loss": 0.4191,
+      "step": 4040
+    },
+    {
+      "epoch": 0.08998400284393886,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8200444444444445e-05,
+      "loss": 0.4887,
+      "step": 4050
+    },
+    {
+      "epoch": 0.09020618556701031,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8196000000000003e-05,
+      "loss": 0.4436,
+      "step": 4060
+    },
+    {
+      "epoch": 0.09042836829008176,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.8191555555555558e-05,
+      "loss": 0.4669,
+      "step": 4070
+    },
+    {
+      "epoch": 0.09065055101315321,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8187111111111113e-05,
+      "loss": 0.4529,
+      "step": 4080
+    },
+    {
+      "epoch": 0.09087273373622468,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8182666666666668e-05,
+      "loss": 0.4579,
+      "step": 4090
+    },
+    {
+      "epoch": 0.09109491645929613,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.8178222222222223e-05,
+      "loss": 0.4856,
+      "step": 4100
+    },
+    {
+      "epoch": 0.09131709918236758,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.8173777777777778e-05,
+      "loss": 0.4385,
+      "step": 4110
+    },
+    {
+      "epoch": 0.09153928190543903,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.8169333333333336e-05,
+      "loss": 0.4163,
+      "step": 4120
+    },
+    {
+      "epoch": 0.09176146462851048,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.816488888888889e-05,
+      "loss": 0.4889,
+      "step": 4130
+    },
+    {
+      "epoch": 0.09198364735158195,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.8160444444444445e-05,
+      "loss": 0.4412,
+      "step": 4140
+    },
+    {
+      "epoch": 0.0922058300746534,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.8156000000000004e-05,
+      "loss": 0.5022,
+      "step": 4150
+    },
+    {
+      "epoch": 0.09242801279772485,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.815155555555556e-05,
+      "loss": 0.4063,
+      "step": 4160
+    },
+    {
+      "epoch": 0.0926501955207963,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8147111111111113e-05,
+      "loss": 0.4474,
+      "step": 4170
+    },
+    {
+      "epoch": 0.09287237824386775,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8142666666666668e-05,
+      "loss": 0.4276,
+      "step": 4180
+    },
+    {
+      "epoch": 0.09309456096693922,
+      "grad_norm": 1.875,
+      "learning_rate": 1.8138222222222223e-05,
+      "loss": 0.4464,
+      "step": 4190
+    },
+    {
+      "epoch": 0.09331674369001067,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.8133777777777778e-05,
+      "loss": 0.4756,
+      "step": 4200
+    },
+    {
+      "epoch": 0.09353892641308212,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.8129333333333333e-05,
+      "loss": 0.4513,
+      "step": 4210
+    },
+    {
+      "epoch": 0.09376110913615357,
+      "grad_norm": 2.125,
+      "learning_rate": 1.812488888888889e-05,
+      "loss": 0.4588,
+      "step": 4220
+    },
+    {
+      "epoch": 0.09398329185922502,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.8120444444444446e-05,
+      "loss": 0.424,
+      "step": 4230
+    },
+    {
+      "epoch": 0.09420547458229649,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8116000000000004e-05,
+      "loss": 0.4253,
+      "step": 4240
+    },
+    {
+      "epoch": 0.09442765730536794,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.811155555555556e-05,
+      "loss": 0.4561,
+      "step": 4250
+    },
+    {
+      "epoch": 0.09464984002843939,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.8107111111111114e-05,
+      "loss": 0.3994,
+      "step": 4260
+    },
+    {
+      "epoch": 0.09487202275151084,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.810266666666667e-05,
+      "loss": 0.4645,
+      "step": 4270
+    },
+    {
+      "epoch": 0.09509420547458229,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.8098222222222223e-05,
+      "loss": 0.4633,
+      "step": 4280
+    },
+    {
+      "epoch": 0.09531638819765376,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.809377777777778e-05,
+      "loss": 0.4848,
+      "step": 4290
+    },
+    {
+      "epoch": 0.09553857092072521,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8089333333333333e-05,
+      "loss": 0.4327,
+      "step": 4300
+    },
+    {
+      "epoch": 0.09576075364379666,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.8084888888888888e-05,
+      "loss": 0.4712,
+      "step": 4310
+    },
+    {
+      "epoch": 0.09598293636686811,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.8080444444444446e-05,
+      "loss": 0.4267,
+      "step": 4320
+    },
+    {
+      "epoch": 0.09620511908993956,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.8076e-05,
+      "loss": 0.4325,
+      "step": 4330
+    },
+    {
+      "epoch": 0.09642730181301103,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.807155555555556e-05,
+      "loss": 0.4381,
+      "step": 4340
+    },
+    {
+      "epoch": 0.09664948453608248,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.8067111111111114e-05,
+      "loss": 0.4514,
+      "step": 4350
+    },
+    {
+      "epoch": 0.09687166725915393,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.806266666666667e-05,
+      "loss": 0.4969,
+      "step": 4360
+    },
+    {
+      "epoch": 0.09709384998222538,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.8058222222222224e-05,
+      "loss": 0.4552,
+      "step": 4370
+    },
+    {
+      "epoch": 0.09731603270529683,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.805377777777778e-05,
+      "loss": 0.432,
+      "step": 4380
+    },
+    {
+      "epoch": 0.0975382154283683,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.8049333333333334e-05,
+      "loss": 0.4501,
+      "step": 4390
+    },
+    {
+      "epoch": 0.09776039815143975,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.804488888888889e-05,
+      "loss": 0.4332,
+      "step": 4400
+    },
+    {
+      "epoch": 0.0979825808745112,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.8040444444444443e-05,
+      "loss": 0.4507,
+      "step": 4410
+    },
+    {
+      "epoch": 0.09820476359758265,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8036e-05,
+      "loss": 0.4568,
+      "step": 4420
+    },
+    {
+      "epoch": 0.0984269463206541,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.8031555555555556e-05,
+      "loss": 0.4763,
+      "step": 4430
+    },
+    {
+      "epoch": 0.09864912904372555,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.8027111111111115e-05,
+      "loss": 0.4373,
+      "step": 4440
+    },
+    {
+      "epoch": 0.09887131176679702,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.802266666666667e-05,
+      "loss": 0.4492,
+      "step": 4450
+    },
+    {
+      "epoch": 0.09909349448986847,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.8018222222222224e-05,
+      "loss": 0.454,
+      "step": 4460
+    },
+    {
+      "epoch": 0.09931567721293992,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.801377777777778e-05,
+      "loss": 0.4747,
+      "step": 4470
+    },
+    {
+      "epoch": 0.09953785993601137,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.8009333333333334e-05,
+      "loss": 0.4355,
+      "step": 4480
+    },
+    {
+      "epoch": 0.09976004265908282,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.800488888888889e-05,
+      "loss": 0.4867,
+      "step": 4490
+    },
+    {
+      "epoch": 0.09998222538215429,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.8000444444444444e-05,
+      "loss": 0.4415,
+      "step": 4500
+    },
+    {
+      "epoch": 0.10020440810522574,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.7996000000000002e-05,
+      "loss": 0.4903,
+      "step": 4510
+    },
+    {
+      "epoch": 0.10042659082829719,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.7991555555555557e-05,
+      "loss": 0.4308,
+      "step": 4520
+    },
+    {
+      "epoch": 0.10064877355136864,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.798711111111111e-05,
+      "loss": 0.5011,
+      "step": 4530
+    },
+    {
+      "epoch": 0.1008709562744401,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.798266666666667e-05,
+      "loss": 0.4614,
+      "step": 4540
+    },
+    {
+      "epoch": 0.10109313899751156,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.7978222222222225e-05,
+      "loss": 0.4547,
+      "step": 4550
+    },
+    {
+      "epoch": 0.10131532172058301,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.797377777777778e-05,
+      "loss": 0.4505,
+      "step": 4560
+    },
+    {
+      "epoch": 0.10153750444365446,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7969333333333334e-05,
+      "loss": 0.4285,
+      "step": 4570
+    },
+    {
+      "epoch": 0.10175968716672591,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.796488888888889e-05,
+      "loss": 0.4422,
+      "step": 4580
+    },
+    {
+      "epoch": 0.10198186988979736,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7960444444444447e-05,
+      "loss": 0.4242,
+      "step": 4590
+    },
+    {
+      "epoch": 0.10220405261286883,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.7956000000000002e-05,
+      "loss": 0.4183,
+      "step": 4600
+    },
+    {
+      "epoch": 0.10242623533594028,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.7951555555555557e-05,
+      "loss": 0.4375,
+      "step": 4610
+    },
+    {
+      "epoch": 0.10264841805901173,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.7947111111111112e-05,
+      "loss": 0.4561,
+      "step": 4620
+    },
+    {
+      "epoch": 0.10287060078208318,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7942666666666667e-05,
+      "loss": 0.4326,
+      "step": 4630
+    },
+    {
+      "epoch": 0.10309278350515463,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.7938222222222225e-05,
+      "loss": 0.4414,
+      "step": 4640
+    },
+    {
+      "epoch": 0.1033149662282261,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.793377777777778e-05,
+      "loss": 0.4279,
+      "step": 4650
+    },
+    {
+      "epoch": 0.10353714895129755,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.7929333333333335e-05,
+      "loss": 0.4223,
+      "step": 4660
+    },
+    {
+      "epoch": 0.103759331674369,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.792488888888889e-05,
+      "loss": 0.4071,
+      "step": 4670
+    },
+    {
+      "epoch": 0.10398151439744045,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.7920444444444448e-05,
+      "loss": 0.4339,
+      "step": 4680
+    },
+    {
+      "epoch": 0.1042036971205119,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.7916000000000003e-05,
+      "loss": 0.4988,
+      "step": 4690
+    },
+    {
+      "epoch": 0.10442587984358337,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7911555555555557e-05,
+      "loss": 0.4143,
+      "step": 4700
+    },
+    {
+      "epoch": 0.10464806256665482,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7907111111111112e-05,
+      "loss": 0.4351,
+      "step": 4710
+    },
+    {
+      "epoch": 0.10487024528972627,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.7902666666666667e-05,
+      "loss": 0.4467,
+      "step": 4720
+    },
+    {
+      "epoch": 0.10509242801279772,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7898222222222222e-05,
+      "loss": 0.446,
+      "step": 4730
+    },
+    {
+      "epoch": 0.10531461073586917,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.789377777777778e-05,
+      "loss": 0.4593,
+      "step": 4740
+    },
+    {
+      "epoch": 0.10553679345894064,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.7889333333333335e-05,
+      "loss": 0.4329,
+      "step": 4750
+    },
+    {
+      "epoch": 0.10575897618201209,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.788488888888889e-05,
+      "loss": 0.389,
+      "step": 4760
+    },
+    {
+      "epoch": 0.10598115890508354,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.7880444444444448e-05,
+      "loss": 0.4639,
+      "step": 4770
+    },
+    {
+      "epoch": 0.10620334162815499,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.7876000000000003e-05,
+      "loss": 0.4327,
+      "step": 4780
+    },
+    {
+      "epoch": 0.10642552435122644,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.7871555555555558e-05,
+      "loss": 0.4199,
+      "step": 4790
+    },
+    {
+      "epoch": 0.10664770707429791,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.7867111111111113e-05,
+      "loss": 0.4632,
+      "step": 4800
+    },
+    {
+      "epoch": 0.10686988979736936,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7862666666666668e-05,
+      "loss": 0.449,
+      "step": 4810
+    },
+    {
+      "epoch": 0.10709207252044081,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7858222222222222e-05,
+      "loss": 0.4737,
+      "step": 4820
+    },
+    {
+      "epoch": 0.10731425524351226,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7853777777777777e-05,
+      "loss": 0.4336,
+      "step": 4830
+    },
+    {
+      "epoch": 0.10753643796658371,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.7849333333333335e-05,
+      "loss": 0.4554,
+      "step": 4840
+    },
+    {
+      "epoch": 0.10775862068965517,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.784488888888889e-05,
+      "loss": 0.3901,
+      "step": 4850
+    },
+    {
+      "epoch": 0.10798080341272663,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.784044444444445e-05,
+      "loss": 0.4239,
+      "step": 4860
+    },
+    {
+      "epoch": 0.10820298613579808,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.7836000000000003e-05,
+      "loss": 0.4552,
+      "step": 4870
+    },
+    {
+      "epoch": 0.10842516885886953,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.7831555555555558e-05,
+      "loss": 0.4197,
+      "step": 4880
+    },
+    {
+      "epoch": 0.10864735158194098,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7827111111111113e-05,
+      "loss": 0.404,
+      "step": 4890
+    },
+    {
+      "epoch": 0.10886953430501244,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7822666666666668e-05,
+      "loss": 0.4529,
+      "step": 4900
+    },
+    {
+      "epoch": 0.1090917170280839,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7818222222222223e-05,
+      "loss": 0.4506,
+      "step": 4910
+    },
+    {
+      "epoch": 0.10931389975115535,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.7813777777777778e-05,
+      "loss": 0.4431,
+      "step": 4920
+    },
+    {
+      "epoch": 0.1095360824742268,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.7809333333333332e-05,
+      "loss": 0.4971,
+      "step": 4930
+    },
+    {
+      "epoch": 0.10975826519729825,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.780488888888889e-05,
+      "loss": 0.4553,
+      "step": 4940
+    },
+    {
+      "epoch": 0.1099804479203697,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7800444444444446e-05,
+      "loss": 0.4072,
+      "step": 4950
+    },
+    {
+      "epoch": 0.11020263064344117,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.7796000000000004e-05,
+      "loss": 0.4075,
+      "step": 4960
+    },
+    {
+      "epoch": 0.11042481336651262,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.779155555555556e-05,
+      "loss": 0.4849,
+      "step": 4970
+    },
+    {
+      "epoch": 0.11064699608958407,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.7787111111111113e-05,
+      "loss": 0.4143,
+      "step": 4980
+    },
+    {
+      "epoch": 0.11086917881265552,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.7782666666666668e-05,
+      "loss": 0.4471,
+      "step": 4990
+    },
+    {
+      "epoch": 0.11109136153572698,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.7778222222222223e-05,
+      "loss": 0.4606,
+      "step": 5000
+    },
+    {
+      "epoch": 0.11131354425879844,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7773777777777778e-05,
+      "loss": 0.4429,
+      "step": 5010
+    },
+    {
+      "epoch": 0.11153572698186989,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.7769333333333333e-05,
+      "loss": 0.4411,
+      "step": 5020
+    },
+    {
+      "epoch": 0.11175790970494134,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.7764888888888888e-05,
+      "loss": 0.4839,
+      "step": 5030
+    },
+    {
+      "epoch": 0.1119800924280128,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7760444444444446e-05,
+      "loss": 0.4667,
+      "step": 5040
+    },
+    {
+      "epoch": 0.11220227515108425,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7756e-05,
+      "loss": 0.4371,
+      "step": 5050
+    },
+    {
+      "epoch": 0.11242445787415571,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.775155555555556e-05,
+      "loss": 0.435,
+      "step": 5060
+    },
+    {
+      "epoch": 0.11264664059722716,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.7747111111111114e-05,
+      "loss": 0.426,
+      "step": 5070
+    },
+    {
+      "epoch": 0.11286882332029861,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.774266666666667e-05,
+      "loss": 0.4797,
+      "step": 5080
+    },
+    {
+      "epoch": 0.11309100604337007,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.7738222222222224e-05,
+      "loss": 0.4882,
+      "step": 5090
+    },
+    {
+      "epoch": 0.11331318876644152,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.773377777777778e-05,
+      "loss": 0.4936,
+      "step": 5100
+    },
+    {
+      "epoch": 0.11353537148951298,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7729333333333333e-05,
+      "loss": 0.4728,
+      "step": 5110
+    },
+    {
+      "epoch": 0.11375755421258443,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7724888888888888e-05,
+      "loss": 0.4162,
+      "step": 5120
+    },
+    {
+      "epoch": 0.11397973693565588,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7720444444444446e-05,
+      "loss": 0.4128,
+      "step": 5130
+    },
+    {
+      "epoch": 0.11420191965872734,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.7716e-05,
+      "loss": 0.4388,
+      "step": 5140
+    },
+    {
+      "epoch": 0.11442410238179879,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7711555555555556e-05,
+      "loss": 0.4442,
+      "step": 5150
+    },
+    {
+      "epoch": 0.11464628510487025,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.7707111111111114e-05,
+      "loss": 0.4308,
+      "step": 5160
+    },
+    {
+      "epoch": 0.1148684678279417,
+      "grad_norm": 1.625,
+      "learning_rate": 1.770266666666667e-05,
+      "loss": 0.4208,
+      "step": 5170
+    },
+    {
+      "epoch": 0.11509065055101315,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.7698222222222224e-05,
+      "loss": 0.4284,
+      "step": 5180
+    },
+    {
+      "epoch": 0.1153128332740846,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.769377777777778e-05,
+      "loss": 0.4512,
+      "step": 5190
+    },
+    {
+      "epoch": 0.11553501599715606,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.7689333333333334e-05,
+      "loss": 0.464,
+      "step": 5200
+    },
+    {
+      "epoch": 0.11575719872022751,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.7684888888888892e-05,
+      "loss": 0.4381,
+      "step": 5210
+    },
+    {
+      "epoch": 0.11597938144329897,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7680444444444447e-05,
+      "loss": 0.4139,
+      "step": 5220
+    },
+    {
+      "epoch": 0.11620156416637042,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7676e-05,
+      "loss": 0.4377,
+      "step": 5230
+    },
+    {
+      "epoch": 0.11642374688944188,
+      "grad_norm": 2.0,
+      "learning_rate": 1.7671555555555556e-05,
+      "loss": 0.4682,
+      "step": 5240
+    },
+    {
+      "epoch": 0.11664592961251333,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.766711111111111e-05,
+      "loss": 0.4179,
+      "step": 5250
+    },
+    {
+      "epoch": 0.11686811233558478,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.766266666666667e-05,
+      "loss": 0.438,
+      "step": 5260
+    },
+    {
+      "epoch": 0.11709029505865624,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.7658222222222224e-05,
+      "loss": 0.4374,
+      "step": 5270
+    },
+    {
+      "epoch": 0.1173124777817277,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.765377777777778e-05,
+      "loss": 0.4344,
+      "step": 5280
+    },
+    {
+      "epoch": 0.11753466050479915,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.7649333333333334e-05,
+      "loss": 0.4616,
+      "step": 5290
+    },
+    {
+      "epoch": 0.1177568432278706,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.7644888888888892e-05,
+      "loss": 0.3837,
+      "step": 5300
+    },
+    {
+      "epoch": 0.11797902595094205,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.7640444444444447e-05,
+      "loss": 0.4316,
+      "step": 5310
+    },
+    {
+      "epoch": 0.11820120867401351,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.7636000000000002e-05,
+      "loss": 0.4124,
+      "step": 5320
+    },
+    {
+      "epoch": 0.11842339139708496,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7631555555555557e-05,
+      "loss": 0.4376,
+      "step": 5330
+    },
+    {
+      "epoch": 0.11864557412015642,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.762711111111111e-05,
+      "loss": 0.437,
+      "step": 5340
+    },
+    {
+      "epoch": 0.11886775684322787,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7622666666666666e-05,
+      "loss": 0.4459,
+      "step": 5350
+    },
+    {
+      "epoch": 0.11908993956629932,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.7618222222222225e-05,
+      "loss": 0.4244,
+      "step": 5360
+    },
+    {
+      "epoch": 0.11931212228937078,
+      "grad_norm": 2.5,
+      "learning_rate": 1.761377777777778e-05,
+      "loss": 0.4318,
+      "step": 5370
+    },
+    {
+      "epoch": 0.11953430501244224,
+      "grad_norm": 2.0,
+      "learning_rate": 1.7609333333333334e-05,
+      "loss": 0.4228,
+      "step": 5380
+    },
+    {
+      "epoch": 0.11975648773551369,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.7604888888888893e-05,
+      "loss": 0.4659,
+      "step": 5390
+    },
+    {
+      "epoch": 0.11997867045858514,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7600444444444447e-05,
+      "loss": 0.4646,
+      "step": 5400
+    },
+    {
+      "epoch": 0.12020085318165659,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.7596000000000002e-05,
+      "loss": 0.405,
+      "step": 5410
+    },
+    {
+      "epoch": 0.12042303590472805,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.7591555555555557e-05,
+      "loss": 0.4746,
+      "step": 5420
+    },
+    {
+      "epoch": 0.1206452186277995,
+      "grad_norm": 2.0,
+      "learning_rate": 1.7587111111111112e-05,
+      "loss": 0.4529,
+      "step": 5430
+    },
+    {
+      "epoch": 0.12086740135087096,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7582666666666667e-05,
+      "loss": 0.4221,
+      "step": 5440
+    },
+    {
+      "epoch": 0.12108958407394241,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.757822222222222e-05,
+      "loss": 0.4198,
+      "step": 5450
+    },
+    {
+      "epoch": 0.12131176679701386,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.757377777777778e-05,
+      "loss": 0.4673,
+      "step": 5460
+    },
+    {
+      "epoch": 0.12153394952008532,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.7569333333333335e-05,
+      "loss": 0.4656,
+      "step": 5470
+    },
+    {
+      "epoch": 0.12175613224315678,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7564888888888893e-05,
+      "loss": 0.4143,
+      "step": 5480
+    },
+    {
+      "epoch": 0.12197831496622823,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7560444444444448e-05,
+      "loss": 0.4781,
+      "step": 5490
+    },
+    {
+      "epoch": 0.12220049768929968,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7556000000000003e-05,
+      "loss": 0.4364,
+      "step": 5500
+    },
+    {
+      "epoch": 0.12242268041237113,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.7551555555555558e-05,
+      "loss": 0.4422,
+      "step": 5510
+    },
+    {
+      "epoch": 0.1226448631354426,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.7547111111111112e-05,
+      "loss": 0.4161,
+      "step": 5520
+    },
+    {
+      "epoch": 0.12286704585851405,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7542666666666667e-05,
+      "loss": 0.4024,
+      "step": 5530
+    },
+    {
+      "epoch": 0.1230892285815855,
+      "grad_norm": 2.125,
+      "learning_rate": 1.7538222222222222e-05,
+      "loss": 0.4651,
+      "step": 5540
+    },
+    {
+      "epoch": 0.12331141130465695,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.7533777777777777e-05,
+      "loss": 0.4519,
+      "step": 5550
+    },
+    {
+      "epoch": 0.1235335940277284,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.7529333333333335e-05,
+      "loss": 0.4236,
+      "step": 5560
+    },
+    {
+      "epoch": 0.12375577675079986,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.752488888888889e-05,
+      "loss": 0.4643,
+      "step": 5570
+    },
+    {
+      "epoch": 0.12397795947387132,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.7520444444444448e-05,
+      "loss": 0.3911,
+      "step": 5580
+    },
+    {
+      "epoch": 0.12420014219694277,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.7516000000000003e-05,
+      "loss": 0.4421,
+      "step": 5590
+    },
+    {
+      "epoch": 0.12442232492001422,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.7511555555555558e-05,
+      "loss": 0.4339,
+      "step": 5600
+    },
+    {
+      "epoch": 0.12464450764308567,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.7507111111111113e-05,
+      "loss": 0.4642,
+      "step": 5610
+    },
+    {
+      "epoch": 0.12486669036615712,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.7502666666666668e-05,
+      "loss": 0.4276,
+      "step": 5620
+    },
+    {
+      "epoch": 0.12508887308922859,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.7498222222222222e-05,
+      "loss": 0.405,
+      "step": 5630
+    },
+    {
+      "epoch": 0.12531105581230004,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.7493777777777777e-05,
+      "loss": 0.4389,
+      "step": 5640
+    },
+    {
+      "epoch": 0.1255332385353715,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.7489333333333332e-05,
+      "loss": 0.4203,
+      "step": 5650
+    },
+    {
+      "epoch": 0.12575542125844294,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.748488888888889e-05,
+      "loss": 0.4125,
+      "step": 5660
+    },
+    {
+      "epoch": 0.1259776039815144,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7480444444444445e-05,
+      "loss": 0.4202,
+      "step": 5670
+    },
+    {
+      "epoch": 0.12619978670458584,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.7476000000000003e-05,
+      "loss": 0.4097,
+      "step": 5680
+    },
+    {
+      "epoch": 0.1264219694276573,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.7471555555555558e-05,
+      "loss": 0.4353,
+      "step": 5690
+    },
+    {
+      "epoch": 0.12664415215072877,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.7467111111111113e-05,
+      "loss": 0.4267,
+      "step": 5700
+    },
+    {
+      "epoch": 0.12686633487380022,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.7462666666666668e-05,
+      "loss": 0.4148,
+      "step": 5710
+    },
+    {
+      "epoch": 0.12708851759687168,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7458222222222223e-05,
+      "loss": 0.4694,
+      "step": 5720
+    },
+    {
+      "epoch": 0.12731070031994313,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7453777777777778e-05,
+      "loss": 0.4575,
+      "step": 5730
+    },
+    {
+      "epoch": 0.12753288304301458,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.7449333333333333e-05,
+      "loss": 0.4206,
+      "step": 5740
+    },
+    {
+      "epoch": 0.12775506576608603,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.744488888888889e-05,
+      "loss": 0.4359,
+      "step": 5750
+    },
+    {
+      "epoch": 0.12797724848915748,
+      "grad_norm": 2.25,
+      "learning_rate": 1.7440444444444446e-05,
+      "loss": 0.4633,
+      "step": 5760
+    },
+    {
+      "epoch": 0.12819943121222893,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.7436e-05,
+      "loss": 0.4273,
+      "step": 5770
+    },
+    {
+      "epoch": 0.12842161393530038,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.743155555555556e-05,
+      "loss": 0.4296,
+      "step": 5780
+    },
+    {
+      "epoch": 0.12864379665837183,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.7427111111111114e-05,
+      "loss": 0.4253,
+      "step": 5790
+    },
+    {
+      "epoch": 0.12886597938144329,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.742266666666667e-05,
+      "loss": 0.4561,
+      "step": 5800
+    },
+    {
+      "epoch": 0.12908816210451476,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7418222222222223e-05,
+      "loss": 0.4183,
+      "step": 5810
+    },
+    {
+      "epoch": 0.12931034482758622,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.7413777777777778e-05,
+      "loss": 0.4209,
+      "step": 5820
+    },
+    {
+      "epoch": 0.12953252755065767,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.7409333333333336e-05,
+      "loss": 0.4257,
+      "step": 5830
+    },
+    {
+      "epoch": 0.12975471027372912,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.740488888888889e-05,
+      "loss": 0.4621,
+      "step": 5840
+    },
+    {
+      "epoch": 0.12997689299680057,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.7400444444444446e-05,
+      "loss": 0.4583,
+      "step": 5850
+    },
+    {
+      "epoch": 0.13019907571987202,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.7396e-05,
+      "loss": 0.4717,
+      "step": 5860
+    },
+    {
+      "epoch": 0.13042125844294347,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.7391555555555556e-05,
+      "loss": 0.405,
+      "step": 5870
+    },
+    {
+      "epoch": 0.13064344116601492,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.7387111111111114e-05,
+      "loss": 0.4456,
+      "step": 5880
+    },
+    {
+      "epoch": 0.13086562388908637,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.738266666666667e-05,
+      "loss": 0.4651,
+      "step": 5890
+    },
+    {
+      "epoch": 0.13108780661215783,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7378222222222224e-05,
+      "loss": 0.3938,
+      "step": 5900
+    },
+    {
+      "epoch": 0.1313099893352293,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.737377777777778e-05,
+      "loss": 0.4213,
+      "step": 5910
+    },
+    {
+      "epoch": 0.13153217205830076,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7369333333333337e-05,
+      "loss": 0.438,
+      "step": 5920
+    },
+    {
+      "epoch": 0.1317543547813722,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.736488888888889e-05,
+      "loss": 0.4435,
+      "step": 5930
+    },
+    {
+      "epoch": 0.13197653750444366,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.7360444444444446e-05,
+      "loss": 0.4096,
+      "step": 5940
+    },
+    {
+      "epoch": 0.1321987202275151,
+      "grad_norm": 1.75,
+      "learning_rate": 1.7356e-05,
+      "loss": 0.4139,
+      "step": 5950
+    },
+    {
+      "epoch": 0.13242090295058656,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.7351555555555556e-05,
+      "loss": 0.3856,
+      "step": 5960
+    },
+    {
+      "epoch": 0.132643085673658,
+      "grad_norm": 1.625,
+      "learning_rate": 1.734711111111111e-05,
+      "loss": 0.4187,
+      "step": 5970
+    },
+    {
+      "epoch": 0.13286526839672946,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.734266666666667e-05,
+      "loss": 0.4594,
+      "step": 5980
+    },
+    {
+      "epoch": 0.13308745111980091,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.7338222222222224e-05,
+      "loss": 0.4173,
+      "step": 5990
+    },
+    {
+      "epoch": 0.13330963384287237,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.733377777777778e-05,
+      "loss": 0.4522,
+      "step": 6000
+    },
+    {
+      "epoch": 0.13353181656594384,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.7329333333333337e-05,
+      "loss": 0.421,
+      "step": 6010
+    },
+    {
+      "epoch": 0.1337539992890153,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.7324888888888892e-05,
+      "loss": 0.4665,
+      "step": 6020
+    },
+    {
+      "epoch": 0.13397618201208675,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.7320444444444447e-05,
+      "loss": 0.4417,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1341983647351582,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7316e-05,
+      "loss": 0.3872,
+      "step": 6040
+    },
+    {
+      "epoch": 0.13442054745822965,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.7311555555555556e-05,
+      "loss": 0.4432,
+      "step": 6050
+    },
+    {
+      "epoch": 0.1346427301813011,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.730711111111111e-05,
+      "loss": 0.4623,
+      "step": 6060
+    },
+    {
+      "epoch": 0.13486491290437255,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.7302666666666666e-05,
+      "loss": 0.4505,
+      "step": 6070
+    },
+    {
+      "epoch": 0.135087095627444,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.7298222222222224e-05,
+      "loss": 0.4451,
+      "step": 6080
+    },
+    {
+      "epoch": 0.13530927835051546,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.729377777777778e-05,
+      "loss": 0.4378,
+      "step": 6090
+    },
+    {
+      "epoch": 0.1355314610735869,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7289333333333337e-05,
+      "loss": 0.4491,
+      "step": 6100
+    },
+    {
+      "epoch": 0.13575364379665839,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.7284888888888892e-05,
+      "loss": 0.421,
+      "step": 6110
+    },
+    {
+      "epoch": 0.13597582651972984,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7280444444444447e-05,
+      "loss": 0.4395,
+      "step": 6120
+    },
+    {
+      "epoch": 0.1361980092428013,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7276000000000002e-05,
+      "loss": 0.4214,
+      "step": 6130
+    },
+    {
+      "epoch": 0.13642019196587274,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7271555555555557e-05,
+      "loss": 0.4264,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1366423746889442,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.726711111111111e-05,
+      "loss": 0.3855,
+      "step": 6150
+    },
+    {
+      "epoch": 0.13686455741201564,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.7262666666666667e-05,
+      "loss": 0.4683,
+      "step": 6160
+    },
+    {
+      "epoch": 0.1370867401350871,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.725822222222222e-05,
+      "loss": 0.4176,
+      "step": 6170
+    },
+    {
+      "epoch": 0.13730892285815854,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.725377777777778e-05,
+      "loss": 0.4196,
+      "step": 6180
+    },
+    {
+      "epoch": 0.13753110558123,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7249333333333334e-05,
+      "loss": 0.4262,
+      "step": 6190
+    },
+    {
+      "epoch": 0.13775328830430145,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.7244888888888893e-05,
+      "loss": 0.4148,
+      "step": 6200
+    },
+    {
+      "epoch": 0.1379754710273729,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7240444444444448e-05,
+      "loss": 0.398,
+      "step": 6210
+    },
+    {
+      "epoch": 0.13819765375044438,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.7236000000000002e-05,
+      "loss": 0.425,
+      "step": 6220
+    },
+    {
+      "epoch": 0.13841983647351583,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.7231555555555557e-05,
+      "loss": 0.4279,
+      "step": 6230
+    },
+    {
+      "epoch": 0.13864201919658728,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.7227111111111112e-05,
+      "loss": 0.4503,
+      "step": 6240
+    },
+    {
+      "epoch": 0.13886420191965873,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.7222666666666667e-05,
+      "loss": 0.445,
+      "step": 6250
+    },
+    {
+      "epoch": 0.13908638464273018,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.7218222222222222e-05,
+      "loss": 0.4387,
+      "step": 6260
+    },
+    {
+      "epoch": 0.13930856736580163,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.7213777777777777e-05,
+      "loss": 0.4213,
+      "step": 6270
+    },
+    {
+      "epoch": 0.13953075008887308,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.7209333333333335e-05,
+      "loss": 0.4127,
+      "step": 6280
+    },
+    {
+      "epoch": 0.13975293281194454,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.720488888888889e-05,
+      "loss": 0.4115,
+      "step": 6290
+    },
+    {
+      "epoch": 0.139975115535016,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.7200444444444448e-05,
+      "loss": 0.4444,
+      "step": 6300
+    },
+    {
+      "epoch": 0.14019729825808744,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.7196000000000003e-05,
+      "loss": 0.4442,
+      "step": 6310
+    },
+    {
+      "epoch": 0.14041948098115892,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.7191555555555558e-05,
+      "loss": 0.4216,
+      "step": 6320
+    },
+    {
+      "epoch": 0.14064166370423037,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.7187111111111112e-05,
+      "loss": 0.4539,
+      "step": 6330
+    },
+    {
+      "epoch": 0.14086384642730182,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.7182666666666667e-05,
+      "loss": 0.4116,
+      "step": 6340
+    },
+    {
+      "epoch": 0.14108602915037327,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.7178222222222222e-05,
+      "loss": 0.4251,
+      "step": 6350
+    },
+    {
+      "epoch": 0.14130821187344472,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7173777777777777e-05,
+      "loss": 0.4355,
+      "step": 6360
+    },
+    {
+      "epoch": 0.14153039459651617,
+      "grad_norm": 2.0,
+      "learning_rate": 1.7169333333333335e-05,
+      "loss": 0.409,
+      "step": 6370
+    },
+    {
+      "epoch": 0.14175257731958762,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.716488888888889e-05,
+      "loss": 0.4058,
+      "step": 6380
+    },
+    {
+      "epoch": 0.14197476004265908,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7160444444444445e-05,
+      "loss": 0.4102,
+      "step": 6390
+    },
+    {
+      "epoch": 0.14219694276573053,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.7156000000000003e-05,
+      "loss": 0.4403,
+      "step": 6400
+    },
+    {
+      "epoch": 0.14241912548880198,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.7151555555555558e-05,
+      "loss": 0.4555,
+      "step": 6410
+    },
+    {
+      "epoch": 0.14264130821187346,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7147111111111113e-05,
+      "loss": 0.4371,
+      "step": 6420
+    },
+    {
+      "epoch": 0.1428634909349449,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.7142666666666668e-05,
+      "loss": 0.4506,
+      "step": 6430
+    },
+    {
+      "epoch": 0.14308567365801636,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7138222222222222e-05,
+      "loss": 0.402,
+      "step": 6440
+    },
+    {
+      "epoch": 0.1433078563810878,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.713377777777778e-05,
+      "loss": 0.4712,
+      "step": 6450
+    },
+    {
+      "epoch": 0.14353003910415926,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.7129333333333336e-05,
+      "loss": 0.4108,
+      "step": 6460
+    },
+    {
+      "epoch": 0.14375222182723071,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.712488888888889e-05,
+      "loss": 0.4138,
+      "step": 6470
+    },
+    {
+      "epoch": 0.14397440455030217,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7120444444444445e-05,
+      "loss": 0.4218,
+      "step": 6480
+    },
+    {
+      "epoch": 0.14419658727337362,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.7116e-05,
+      "loss": 0.3997,
+      "step": 6490
+    },
+    {
+      "epoch": 0.14441876999644507,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.711155555555556e-05,
+      "loss": 0.3893,
+      "step": 6500
+    },
+    {
+      "epoch": 0.14464095271951652,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7107111111111113e-05,
+      "loss": 0.453,
+      "step": 6510
+    },
+    {
+      "epoch": 0.144863135442588,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7102666666666668e-05,
+      "loss": 0.4047,
+      "step": 6520
+    },
+    {
+      "epoch": 0.14508531816565945,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.7098222222222223e-05,
+      "loss": 0.3939,
+      "step": 6530
+    },
+    {
+      "epoch": 0.1453075008887309,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.709377777777778e-05,
+      "loss": 0.4508,
+      "step": 6540
+    },
+    {
+      "epoch": 0.14552968361180235,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7089333333333336e-05,
+      "loss": 0.429,
+      "step": 6550
+    },
+    {
+      "epoch": 0.1457518663348738,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.708488888888889e-05,
+      "loss": 0.4291,
+      "step": 6560
+    },
+    {
+      "epoch": 0.14597404905794525,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.7080444444444446e-05,
+      "loss": 0.4572,
+      "step": 6570
+    },
+    {
+      "epoch": 0.1461962317810167,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.7076e-05,
+      "loss": 0.4417,
+      "step": 6580
+    },
+    {
+      "epoch": 0.14641841450408816,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.7071555555555555e-05,
+      "loss": 0.4619,
+      "step": 6590
+    },
+    {
+      "epoch": 0.1466405972271596,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.7067111111111114e-05,
+      "loss": 0.4344,
+      "step": 6600
+    },
+    {
+      "epoch": 0.14686277995023106,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.706266666666667e-05,
+      "loss": 0.4255,
+      "step": 6610
+    },
+    {
+      "epoch": 0.1470849626733025,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.7058222222222223e-05,
+      "loss": 0.4519,
+      "step": 6620
+    },
+    {
+      "epoch": 0.147307145396374,
+      "grad_norm": 2.125,
+      "learning_rate": 1.705377777777778e-05,
+      "loss": 0.4239,
+      "step": 6630
+    },
+    {
+      "epoch": 0.14752932811944544,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7049333333333336e-05,
+      "loss": 0.4371,
+      "step": 6640
+    },
+    {
+      "epoch": 0.1477515108425169,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.704488888888889e-05,
+      "loss": 0.4028,
+      "step": 6650
+    },
+    {
+      "epoch": 0.14797369356558834,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7040444444444446e-05,
+      "loss": 0.4049,
+      "step": 6660
+    },
+    {
+      "epoch": 0.1481958762886598,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7036e-05,
+      "loss": 0.4177,
+      "step": 6670
+    },
+    {
+      "epoch": 0.14841805901173125,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7031555555555556e-05,
+      "loss": 0.4601,
+      "step": 6680
+    },
+    {
+      "epoch": 0.1486402417348027,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.702711111111111e-05,
+      "loss": 0.4334,
+      "step": 6690
+    },
+    {
+      "epoch": 0.14886242445787415,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.702266666666667e-05,
+      "loss": 0.4312,
+      "step": 6700
+    },
+    {
+      "epoch": 0.1490846071809456,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.7018222222222224e-05,
+      "loss": 0.4537,
+      "step": 6710
+    },
+    {
+      "epoch": 0.14930678990401705,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.701377777777778e-05,
+      "loss": 0.3942,
+      "step": 6720
+    },
+    {
+      "epoch": 0.14952897262708853,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.7009333333333337e-05,
+      "loss": 0.4384,
+      "step": 6730
+    },
+    {
+      "epoch": 0.14975115535015998,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.700488888888889e-05,
+      "loss": 0.4478,
+      "step": 6740
+    },
+    {
+      "epoch": 0.14997333807323143,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7000444444444446e-05,
+      "loss": 0.4289,
+      "step": 6750
+    },
+    {
+      "epoch": 0.15019552079630288,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6996e-05,
+      "loss": 0.4442,
+      "step": 6760
+    },
+    {
+      "epoch": 0.15041770351937434,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6991555555555556e-05,
+      "loss": 0.4368,
+      "step": 6770
+    },
+    {
+      "epoch": 0.1506398862424458,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.698711111111111e-05,
+      "loss": 0.447,
+      "step": 6780
+    },
+    {
+      "epoch": 0.15086206896551724,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6982666666666666e-05,
+      "loss": 0.4368,
+      "step": 6790
+    },
+    {
+      "epoch": 0.1510842516885887,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6978222222222224e-05,
+      "loss": 0.4489,
+      "step": 6800
+    },
+    {
+      "epoch": 0.15130643441166014,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.697377777777778e-05,
+      "loss": 0.4006,
+      "step": 6810
+    },
+    {
+      "epoch": 0.1515286171347316,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.6969333333333337e-05,
+      "loss": 0.4404,
+      "step": 6820
+    },
+    {
+      "epoch": 0.15175079985780307,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.6964888888888892e-05,
+      "loss": 0.4283,
+      "step": 6830
+    },
+    {
+      "epoch": 0.15197298258087452,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.6960444444444447e-05,
+      "loss": 0.4502,
+      "step": 6840
+    },
+    {
+      "epoch": 0.15219516530394597,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.6956e-05,
+      "loss": 0.3995,
+      "step": 6850
+    },
+    {
+      "epoch": 0.15241734802701742,
+      "grad_norm": 2.0,
+      "learning_rate": 1.6951555555555556e-05,
+      "loss": 0.4811,
+      "step": 6860
+    },
+    {
+      "epoch": 0.15263953075008888,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.694711111111111e-05,
+      "loss": 0.4137,
+      "step": 6870
+    },
+    {
+      "epoch": 0.15286171347316033,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.6942666666666666e-05,
+      "loss": 0.4289,
+      "step": 6880
+    },
+    {
+      "epoch": 0.15308389619623178,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.693822222222222e-05,
+      "loss": 0.4144,
+      "step": 6890
+    },
+    {
+      "epoch": 0.15330607891930323,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.693377777777778e-05,
+      "loss": 0.4346,
+      "step": 6900
+    },
+    {
+      "epoch": 0.15352826164237468,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6929333333333334e-05,
+      "loss": 0.3889,
+      "step": 6910
+    },
+    {
+      "epoch": 0.15375044436544613,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.6924888888888892e-05,
+      "loss": 0.4404,
+      "step": 6920
+    },
+    {
+      "epoch": 0.1539726270885176,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6920444444444447e-05,
+      "loss": 0.4582,
+      "step": 6930
+    },
+    {
+      "epoch": 0.15419480981158906,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.6916000000000002e-05,
+      "loss": 0.4138,
+      "step": 6940
+    },
+    {
+      "epoch": 0.1544169925346605,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.6911555555555557e-05,
+      "loss": 0.4371,
+      "step": 6950
+    },
+    {
+      "epoch": 0.15463917525773196,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.6907111111111112e-05,
+      "loss": 0.4283,
+      "step": 6960
+    },
+    {
+      "epoch": 0.15486135798080342,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.6902666666666667e-05,
+      "loss": 0.4166,
+      "step": 6970
+    },
+    {
+      "epoch": 0.15508354070387487,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.689822222222222e-05,
+      "loss": 0.4245,
+      "step": 6980
+    },
+    {
+      "epoch": 0.15530572342694632,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.689377777777778e-05,
+      "loss": 0.4454,
+      "step": 6990
+    },
+    {
+      "epoch": 0.15552790615001777,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.6889333333333334e-05,
+      "loss": 0.4286,
+      "step": 7000
+    },
+    {
+      "epoch": 0.15575008887308922,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.688488888888889e-05,
+      "loss": 0.429,
+      "step": 7010
+    },
+    {
+      "epoch": 0.15597227159616067,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.6880444444444448e-05,
+      "loss": 0.4676,
+      "step": 7020
+    },
+    {
+      "epoch": 0.15619445431923212,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6876000000000002e-05,
+      "loss": 0.4476,
+      "step": 7030
+    },
+    {
+      "epoch": 0.1564166370423036,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6871555555555557e-05,
+      "loss": 0.4274,
+      "step": 7040
+    },
+    {
+      "epoch": 0.15663881976537505,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6867111111111112e-05,
+      "loss": 0.45,
+      "step": 7050
+    },
+    {
+      "epoch": 0.1568610024884465,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.6862666666666667e-05,
+      "loss": 0.3895,
+      "step": 7060
+    },
+    {
+      "epoch": 0.15708318521151796,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.6858222222222225e-05,
+      "loss": 0.4266,
+      "step": 7070
+    },
+    {
+      "epoch": 0.1573053679345894,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.685377777777778e-05,
+      "loss": 0.41,
+      "step": 7080
+    },
+    {
+      "epoch": 0.15752755065766086,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6849333333333335e-05,
+      "loss": 0.4325,
+      "step": 7090
+    },
+    {
+      "epoch": 0.1577497333807323,
+      "grad_norm": 1.75,
+      "learning_rate": 1.684488888888889e-05,
+      "loss": 0.4437,
+      "step": 7100
+    },
+    {
+      "epoch": 0.15797191610380376,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.6840444444444445e-05,
+      "loss": 0.4044,
+      "step": 7110
+    },
+    {
+      "epoch": 0.1581940988268752,
+      "grad_norm": 2.0,
+      "learning_rate": 1.6836000000000003e-05,
+      "loss": 0.4015,
+      "step": 7120
+    },
+    {
+      "epoch": 0.15841628154994666,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.6831555555555558e-05,
+      "loss": 0.4318,
+      "step": 7130
+    },
+    {
+      "epoch": 0.15863846427301814,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.6827111111111112e-05,
+      "loss": 0.4308,
+      "step": 7140
+    },
+    {
+      "epoch": 0.1588606469960896,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.6822666666666667e-05,
+      "loss": 0.4395,
+      "step": 7150
+    },
+    {
+      "epoch": 0.15908282971916105,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.6818222222222226e-05,
+      "loss": 0.4215,
+      "step": 7160
+    },
+    {
+      "epoch": 0.1593050124422325,
+      "grad_norm": 2.125,
+      "learning_rate": 1.681377777777778e-05,
+      "loss": 0.4215,
+      "step": 7170
+    },
+    {
+      "epoch": 0.15952719516530395,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.6809333333333335e-05,
+      "loss": 0.4018,
+      "step": 7180
+    },
+    {
+      "epoch": 0.1597493778883754,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.680488888888889e-05,
+      "loss": 0.4546,
+      "step": 7190
+    },
+    {
+      "epoch": 0.15997156061144685,
+      "grad_norm": 1.875,
+      "learning_rate": 1.6800444444444445e-05,
+      "loss": 0.4275,
+      "step": 7200
+    },
+    {
+      "epoch": 0.1601937433345183,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.6796e-05,
+      "loss": 0.4418,
+      "step": 7210
+    },
+    {
+      "epoch": 0.16041592605758975,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.6791555555555558e-05,
+      "loss": 0.3983,
+      "step": 7220
+    },
+    {
+      "epoch": 0.1606381087806612,
+      "grad_norm": 1.875,
+      "learning_rate": 1.6787111111111113e-05,
+      "loss": 0.4474,
+      "step": 7230
+    },
+    {
+      "epoch": 0.16086029150373268,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.6782666666666668e-05,
+      "loss": 0.4463,
+      "step": 7240
+    },
+    {
+      "epoch": 0.16108247422680413,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.6778222222222226e-05,
+      "loss": 0.4474,
+      "step": 7250
+    },
+    {
+      "epoch": 0.16130465694987559,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.677377777777778e-05,
+      "loss": 0.4243,
+      "step": 7260
+    },
+    {
+      "epoch": 0.16152683967294704,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6769333333333336e-05,
+      "loss": 0.4301,
+      "step": 7270
+    },
+    {
+      "epoch": 0.1617490223960185,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.676488888888889e-05,
+      "loss": 0.4086,
+      "step": 7280
+    },
+    {
+      "epoch": 0.16197120511908994,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6760444444444445e-05,
+      "loss": 0.4076,
+      "step": 7290
+    },
+    {
+      "epoch": 0.1621933878421614,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.6756e-05,
+      "loss": 0.4533,
+      "step": 7300
+    },
+    {
+      "epoch": 0.16241557056523284,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6751555555555555e-05,
+      "loss": 0.3845,
+      "step": 7310
+    },
+    {
+      "epoch": 0.1626377532883043,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.6747111111111113e-05,
+      "loss": 0.3745,
+      "step": 7320
+    },
+    {
+      "epoch": 0.16285993601137574,
+      "grad_norm": 1.875,
+      "learning_rate": 1.6742666666666668e-05,
+      "loss": 0.3866,
+      "step": 7330
+    },
+    {
+      "epoch": 0.1630821187344472,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6738222222222223e-05,
+      "loss": 0.447,
+      "step": 7340
+    },
+    {
+      "epoch": 0.16330430145751867,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.673377777777778e-05,
+      "loss": 0.4418,
+      "step": 7350
+    },
+    {
+      "epoch": 0.16352648418059013,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6729333333333336e-05,
+      "loss": 0.3611,
+      "step": 7360
+    },
+    {
+      "epoch": 0.16374866690366158,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.672488888888889e-05,
+      "loss": 0.4063,
+      "step": 7370
+    },
+    {
+      "epoch": 0.16397084962673303,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.6720444444444446e-05,
+      "loss": 0.4245,
+      "step": 7380
+    },
+    {
+      "epoch": 0.16419303234980448,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.6716e-05,
+      "loss": 0.434,
+      "step": 7390
+    },
+    {
+      "epoch": 0.16441521507287593,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.6711555555555555e-05,
+      "loss": 0.3997,
+      "step": 7400
+    },
+    {
+      "epoch": 0.16463739779594738,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.670711111111111e-05,
+      "loss": 0.4215,
+      "step": 7410
+    },
+    {
+      "epoch": 0.16485958051901883,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.670266666666667e-05,
+      "loss": 0.4563,
+      "step": 7420
+    },
+    {
+      "epoch": 0.16508176324209028,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.6698222222222223e-05,
+      "loss": 0.4351,
+      "step": 7430
+    },
+    {
+      "epoch": 0.16530394596516174,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.6693777777777778e-05,
+      "loss": 0.4374,
+      "step": 7440
+    },
+    {
+      "epoch": 0.16552612868823322,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6689333333333336e-05,
+      "loss": 0.4122,
+      "step": 7450
+    },
+    {
+      "epoch": 0.16574831141130467,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.668488888888889e-05,
+      "loss": 0.4547,
+      "step": 7460
+    },
+    {
+      "epoch": 0.16597049413437612,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.6680444444444446e-05,
+      "loss": 0.4414,
+      "step": 7470
+    },
+    {
+      "epoch": 0.16619267685744757,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.6676e-05,
+      "loss": 0.4345,
+      "step": 7480
+    },
+    {
+      "epoch": 0.16641485958051902,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6671555555555556e-05,
+      "loss": 0.4033,
+      "step": 7490
+    },
+    {
+      "epoch": 0.16663704230359047,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.666711111111111e-05,
+      "loss": 0.4159,
+      "step": 7500
+    },
+    {
+      "epoch": 0.16685922502666192,
+      "grad_norm": 2.75,
+      "learning_rate": 1.6662666666666665e-05,
+      "loss": 0.4305,
+      "step": 7510
+    },
+    {
+      "epoch": 0.16708140774973337,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.6658222222222224e-05,
+      "loss": 0.4756,
+      "step": 7520
+    },
+    {
+      "epoch": 0.16730359047280483,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.665377777777778e-05,
+      "loss": 0.4448,
+      "step": 7530
+    },
+    {
+      "epoch": 0.16752577319587628,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.6649333333333337e-05,
+      "loss": 0.3996,
+      "step": 7540
+    },
+    {
+      "epoch": 0.16774795591894776,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.664488888888889e-05,
+      "loss": 0.3974,
+      "step": 7550
+    },
+    {
+      "epoch": 0.1679701386420192,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.6640444444444446e-05,
+      "loss": 0.4326,
+      "step": 7560
+    },
+    {
+      "epoch": 0.16819232136509066,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6636e-05,
+      "loss": 0.4018,
+      "step": 7570
+    },
+    {
+      "epoch": 0.1684145040881621,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.6631555555555556e-05,
+      "loss": 0.4749,
+      "step": 7580
+    },
+    {
+      "epoch": 0.16863668681123356,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.662711111111111e-05,
+      "loss": 0.4123,
+      "step": 7590
+    },
+    {
+      "epoch": 0.168858869534305,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.6622666666666666e-05,
+      "loss": 0.4242,
+      "step": 7600
+    },
+    {
+      "epoch": 0.16908105225737646,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6618222222222224e-05,
+      "loss": 0.4025,
+      "step": 7610
+    },
+    {
+      "epoch": 0.16930323498044791,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.661377777777778e-05,
+      "loss": 0.3951,
+      "step": 7620
+    },
+    {
+      "epoch": 0.16952541770351937,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6609333333333334e-05,
+      "loss": 0.4199,
+      "step": 7630
+    },
+    {
+      "epoch": 0.16974760042659082,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6604888888888892e-05,
+      "loss": 0.4041,
+      "step": 7640
+    },
+    {
+      "epoch": 0.1699697831496623,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.6600444444444447e-05,
+      "loss": 0.4304,
+      "step": 7650
+    },
+    {
+      "epoch": 0.17019196587273375,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.6596000000000002e-05,
+      "loss": 0.4008,
+      "step": 7660
+    },
+    {
+      "epoch": 0.1704141485958052,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.6591555555555557e-05,
+      "loss": 0.3829,
+      "step": 7670
+    },
+    {
+      "epoch": 0.17063633131887665,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.658711111111111e-05,
+      "loss": 0.4511,
+      "step": 7680
+    },
+    {
+      "epoch": 0.1708585140419481,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.658266666666667e-05,
+      "loss": 0.4274,
+      "step": 7690
+    },
+    {
+      "epoch": 0.17108069676501955,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.6578222222222224e-05,
+      "loss": 0.426,
+      "step": 7700
+    },
+    {
+      "epoch": 0.171302879488091,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.657377777777778e-05,
+      "loss": 0.4066,
+      "step": 7710
+    },
+    {
+      "epoch": 0.17152506221116245,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.6569333333333334e-05,
+      "loss": 0.412,
+      "step": 7720
+    },
+    {
+      "epoch": 0.1717472449342339,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.656488888888889e-05,
+      "loss": 0.4298,
+      "step": 7730
+    },
+    {
+      "epoch": 0.17196942765730536,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.6560444444444447e-05,
+      "loss": 0.4233,
+      "step": 7740
+    },
+    {
+      "epoch": 0.1721916103803768,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.6556000000000002e-05,
+      "loss": 0.4258,
+      "step": 7750
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.6551555555555557e-05,
+      "loss": 0.4082,
+      "step": 7760
+    },
+    {
+      "epoch": 0.17263597582651974,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6547111111111112e-05,
+      "loss": 0.4062,
+      "step": 7770
+    },
+    {
+      "epoch": 0.1728581585495912,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.654266666666667e-05,
+      "loss": 0.397,
+      "step": 7780
+    },
+    {
+      "epoch": 0.17308034127266264,
+      "grad_norm": 2.375,
+      "learning_rate": 1.6538222222222225e-05,
+      "loss": 0.4277,
+      "step": 7790
+    },
+    {
+      "epoch": 0.1733025239957341,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.653377777777778e-05,
+      "loss": 0.4284,
+      "step": 7800
+    },
+    {
+      "epoch": 0.17352470671880554,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.6529333333333335e-05,
+      "loss": 0.4554,
+      "step": 7810
+    },
+    {
+      "epoch": 0.173746889441877,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.652488888888889e-05,
+      "loss": 0.4404,
+      "step": 7820
+    },
+    {
+      "epoch": 0.17396907216494845,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.6520444444444444e-05,
+      "loss": 0.4119,
+      "step": 7830
+    },
+    {
+      "epoch": 0.1741912548880199,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.6516000000000002e-05,
+      "loss": 0.4272,
+      "step": 7840
+    },
+    {
+      "epoch": 0.17441343761109135,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.6511555555555557e-05,
+      "loss": 0.4163,
+      "step": 7850
+    },
+    {
+      "epoch": 0.17463562033416283,
+      "grad_norm": 1.75,
+      "learning_rate": 1.6507111111111112e-05,
+      "loss": 0.4186,
+      "step": 7860
+    },
+    {
+      "epoch": 0.17485780305723428,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.650266666666667e-05,
+      "loss": 0.438,
+      "step": 7870
+    },
+    {
+      "epoch": 0.17507998578030573,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.6498222222222225e-05,
+      "loss": 0.4119,
+      "step": 7880
+    },
+    {
+      "epoch": 0.17530216850337718,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.649377777777778e-05,
+      "loss": 0.4031,
+      "step": 7890
+    },
+    {
+      "epoch": 0.17552435122644863,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.6489333333333335e-05,
+      "loss": 0.4007,
+      "step": 7900
+    },
+    {
+      "epoch": 0.17574653394952008,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.648488888888889e-05,
+      "loss": 0.4203,
+      "step": 7910
+    },
+    {
+      "epoch": 0.17596871667259154,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.6480444444444445e-05,
+      "loss": 0.4436,
+      "step": 7920
+    },
+    {
+      "epoch": 0.176190899395663,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6476e-05,
+      "loss": 0.415,
+      "step": 7930
+    },
+    {
+      "epoch": 0.17641308211873444,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.6471555555555558e-05,
+      "loss": 0.4047,
+      "step": 7940
+    },
+    {
+      "epoch": 0.1766352648418059,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6467111111111113e-05,
+      "loss": 0.4092,
+      "step": 7950
+    },
+    {
+      "epoch": 0.17685744756487737,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.6462666666666667e-05,
+      "loss": 0.3885,
+      "step": 7960
+    },
+    {
+      "epoch": 0.17707963028794882,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.6458222222222226e-05,
+      "loss": 0.424,
+      "step": 7970
+    },
+    {
+      "epoch": 0.17730181301102027,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.645377777777778e-05,
+      "loss": 0.3977,
+      "step": 7980
+    },
+    {
+      "epoch": 0.17752399573409172,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6449333333333335e-05,
+      "loss": 0.417,
+      "step": 7990
+    },
+    {
+      "epoch": 0.17774617845716317,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.644488888888889e-05,
+      "loss": 0.3975,
+      "step": 8000
+    },
+    {
+      "epoch": 0.17796836118023462,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.6440444444444445e-05,
+      "loss": 0.3919,
+      "step": 8010
+    },
+    {
+      "epoch": 0.17819054390330608,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6436e-05,
+      "loss": 0.4242,
+      "step": 8020
+    },
+    {
+      "epoch": 0.17841272662637753,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6431555555555555e-05,
+      "loss": 0.4207,
+      "step": 8030
+    },
+    {
+      "epoch": 0.17863490934944898,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.6427111111111113e-05,
+      "loss": 0.3976,
+      "step": 8040
+    },
+    {
+      "epoch": 0.17885709207252043,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.6422666666666668e-05,
+      "loss": 0.4321,
+      "step": 8050
+    },
+    {
+      "epoch": 0.1790792747955919,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.6418222222222223e-05,
+      "loss": 0.4196,
+      "step": 8060
+    },
+    {
+      "epoch": 0.17930145751866336,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.641377777777778e-05,
+      "loss": 0.4491,
+      "step": 8070
+    },
+    {
+      "epoch": 0.1795236402417348,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.6409333333333336e-05,
+      "loss": 0.4271,
+      "step": 8080
+    },
+    {
+      "epoch": 0.17974582296480626,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.640488888888889e-05,
+      "loss": 0.4652,
+      "step": 8090
+    },
+    {
+      "epoch": 0.1799680056878777,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.6400444444444445e-05,
+      "loss": 0.422,
+      "step": 8100
+    },
+    {
+      "epoch": 0.18019018841094916,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6396e-05,
+      "loss": 0.3766,
+      "step": 8110
+    },
+    {
+      "epoch": 0.18041237113402062,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6391555555555555e-05,
+      "loss": 0.3981,
+      "step": 8120
+    },
+    {
+      "epoch": 0.18063455385709207,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.638711111111111e-05,
+      "loss": 0.4288,
+      "step": 8130
+    },
+    {
+      "epoch": 0.18085673658016352,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6382666666666668e-05,
+      "loss": 0.4185,
+      "step": 8140
+    },
+    {
+      "epoch": 0.18107891930323497,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6378222222222223e-05,
+      "loss": 0.4175,
+      "step": 8150
+    },
+    {
+      "epoch": 0.18130110202630642,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.637377777777778e-05,
+      "loss": 0.4535,
+      "step": 8160
+    },
+    {
+      "epoch": 0.1815232847493779,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.6369333333333336e-05,
+      "loss": 0.4269,
+      "step": 8170
+    },
+    {
+      "epoch": 0.18174546747244935,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.636488888888889e-05,
+      "loss": 0.4164,
+      "step": 8180
+    },
+    {
+      "epoch": 0.1819676501955208,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.6360444444444446e-05,
+      "loss": 0.3888,
+      "step": 8190
+    },
+    {
+      "epoch": 0.18218983291859225,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.6356e-05,
+      "loss": 0.4377,
+      "step": 8200
+    },
+    {
+      "epoch": 0.1824120156416637,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.6351555555555555e-05,
+      "loss": 0.4383,
+      "step": 8210
+    },
+    {
+      "epoch": 0.18263419836473516,
+      "grad_norm": 2.375,
+      "learning_rate": 1.634711111111111e-05,
+      "loss": 0.461,
+      "step": 8220
+    },
+    {
+      "epoch": 0.1828563810878066,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.634266666666667e-05,
+      "loss": 0.4201,
+      "step": 8230
+    },
+    {
+      "epoch": 0.18307856381087806,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.6338222222222223e-05,
+      "loss": 0.4292,
+      "step": 8240
+    },
+    {
+      "epoch": 0.1833007465339495,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.6333777777777778e-05,
+      "loss": 0.4403,
+      "step": 8250
+    },
+    {
+      "epoch": 0.18352292925702096,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.6329333333333336e-05,
+      "loss": 0.3969,
+      "step": 8260
+    },
+    {
+      "epoch": 0.18374511198009244,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.632488888888889e-05,
+      "loss": 0.4344,
+      "step": 8270
+    },
+    {
+      "epoch": 0.1839672947031639,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.6320444444444446e-05,
+      "loss": 0.3917,
+      "step": 8280
+    },
+    {
+      "epoch": 0.18418947742623534,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6316e-05,
+      "loss": 0.4092,
+      "step": 8290
+    },
+    {
+      "epoch": 0.1844116601493068,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6311555555555556e-05,
+      "loss": 0.4259,
+      "step": 8300
+    },
+    {
+      "epoch": 0.18463384287237825,
+      "grad_norm": 2.25,
+      "learning_rate": 1.630711111111111e-05,
+      "loss": 0.4341,
+      "step": 8310
+    },
+    {
+      "epoch": 0.1848560255954497,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.630266666666667e-05,
+      "loss": 0.4374,
+      "step": 8320
+    },
+    {
+      "epoch": 0.18507820831852115,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.6298222222222224e-05,
+      "loss": 0.4293,
+      "step": 8330
+    },
+    {
+      "epoch": 0.1853003910415926,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.629377777777778e-05,
+      "loss": 0.4398,
+      "step": 8340
+    },
+    {
+      "epoch": 0.18552257376466405,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.6289333333333333e-05,
+      "loss": 0.4505,
+      "step": 8350
+    },
+    {
+      "epoch": 0.1857447564877355,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.628488888888889e-05,
+      "loss": 0.3886,
+      "step": 8360
+    },
+    {
+      "epoch": 0.18596693921080698,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6280444444444447e-05,
+      "loss": 0.3695,
+      "step": 8370
+    },
+    {
+      "epoch": 0.18618912193387843,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6276e-05,
+      "loss": 0.4525,
+      "step": 8380
+    },
+    {
+      "epoch": 0.18641130465694988,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.6271555555555556e-05,
+      "loss": 0.4285,
+      "step": 8390
+    },
+    {
+      "epoch": 0.18663348738002133,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.6267111111111114e-05,
+      "loss": 0.4404,
+      "step": 8400
+    },
+    {
+      "epoch": 0.18685567010309279,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.626266666666667e-05,
+      "loss": 0.405,
+      "step": 8410
+    },
+    {
+      "epoch": 0.18707785282616424,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.6258222222222224e-05,
+      "loss": 0.4259,
+      "step": 8420
+    },
+    {
+      "epoch": 0.1873000355492357,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.625377777777778e-05,
+      "loss": 0.4658,
+      "step": 8430
+    },
+    {
+      "epoch": 0.18752221827230714,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.6249333333333334e-05,
+      "loss": 0.4148,
+      "step": 8440
+    },
+    {
+      "epoch": 0.1877444009953786,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.624488888888889e-05,
+      "loss": 0.4423,
+      "step": 8450
+    },
+    {
+      "epoch": 0.18796658371845004,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6240444444444447e-05,
+      "loss": 0.4113,
+      "step": 8460
+    },
+    {
+      "epoch": 0.18818876644152152,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6236000000000002e-05,
+      "loss": 0.4338,
+      "step": 8470
+    },
+    {
+      "epoch": 0.18841094916459297,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.6231555555555557e-05,
+      "loss": 0.4146,
+      "step": 8480
+    },
+    {
+      "epoch": 0.18863313188766442,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6227111111111115e-05,
+      "loss": 0.399,
+      "step": 8490
+    },
+    {
+      "epoch": 0.18885531461073587,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.622266666666667e-05,
+      "loss": 0.4089,
+      "step": 8500
+    },
+    {
+      "epoch": 0.18907749733380733,
+      "grad_norm": 2.0,
+      "learning_rate": 1.6218222222222225e-05,
+      "loss": 0.421,
+      "step": 8510
+    },
+    {
+      "epoch": 0.18929968005687878,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.621377777777778e-05,
+      "loss": 0.4608,
+      "step": 8520
+    },
+    {
+      "epoch": 0.18952186277995023,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.6209333333333334e-05,
+      "loss": 0.4154,
+      "step": 8530
+    },
+    {
+      "epoch": 0.18974404550302168,
+      "grad_norm": 2.0,
+      "learning_rate": 1.620488888888889e-05,
+      "loss": 0.394,
+      "step": 8540
+    },
+    {
+      "epoch": 0.18996622822609313,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6200444444444444e-05,
+      "loss": 0.4051,
+      "step": 8550
+    },
+    {
+      "epoch": 0.19018841094916458,
+      "grad_norm": 1.875,
+      "learning_rate": 1.6196000000000002e-05,
+      "loss": 0.3784,
+      "step": 8560
+    },
+    {
+      "epoch": 0.19041059367223603,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.6191555555555557e-05,
+      "loss": 0.4235,
+      "step": 8570
+    },
+    {
+      "epoch": 0.1906327763953075,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.6187111111111112e-05,
+      "loss": 0.4422,
+      "step": 8580
+    },
+    {
+      "epoch": 0.19085495911837896,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.618266666666667e-05,
+      "loss": 0.387,
+      "step": 8590
+    },
+    {
+      "epoch": 0.19107714184145042,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.6178222222222225e-05,
+      "loss": 0.4141,
+      "step": 8600
+    },
+    {
+      "epoch": 0.19129932456452187,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.617377777777778e-05,
+      "loss": 0.4182,
+      "step": 8610
+    },
+    {
+      "epoch": 0.19152150728759332,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.6169333333333335e-05,
+      "loss": 0.3894,
+      "step": 8620
+    },
+    {
+      "epoch": 0.19174369001066477,
+      "grad_norm": 1.875,
+      "learning_rate": 1.616488888888889e-05,
+      "loss": 0.4464,
+      "step": 8630
+    },
+    {
+      "epoch": 0.19196587273373622,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.6160444444444444e-05,
+      "loss": 0.4745,
+      "step": 8640
+    },
+    {
+      "epoch": 0.19218805545680767,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.6156e-05,
+      "loss": 0.4128,
+      "step": 8650
+    },
+    {
+      "epoch": 0.19241023817987912,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.6151555555555557e-05,
+      "loss": 0.4236,
+      "step": 8660
+    },
+    {
+      "epoch": 0.19263242090295057,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.6147111111111112e-05,
+      "loss": 0.4418,
+      "step": 8670
+    },
+    {
+      "epoch": 0.19285460362602205,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.6142666666666667e-05,
+      "loss": 0.4106,
+      "step": 8680
+    },
+    {
+      "epoch": 0.1930767863490935,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.6138222222222225e-05,
+      "loss": 0.4154,
+      "step": 8690
+    },
+    {
+      "epoch": 0.19329896907216496,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.613377777777778e-05,
+      "loss": 0.4608,
+      "step": 8700
+    },
+    {
+      "epoch": 0.1935211517952364,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.6129333333333335e-05,
+      "loss": 0.4428,
+      "step": 8710
+    },
+    {
+      "epoch": 0.19374333451830786,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.612488888888889e-05,
+      "loss": 0.4212,
+      "step": 8720
+    },
+    {
+      "epoch": 0.1939655172413793,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6120444444444445e-05,
+      "loss": 0.4195,
+      "step": 8730
+    },
+    {
+      "epoch": 0.19418769996445076,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.6116e-05,
+      "loss": 0.4108,
+      "step": 8740
+    },
+    {
+      "epoch": 0.1944098826875222,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6111555555555554e-05,
+      "loss": 0.3959,
+      "step": 8750
+    },
+    {
+      "epoch": 0.19463206541059366,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.6107111111111113e-05,
+      "loss": 0.4463,
+      "step": 8760
+    },
+    {
+      "epoch": 0.19485424813366511,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6102666666666667e-05,
+      "loss": 0.4036,
+      "step": 8770
+    },
+    {
+      "epoch": 0.1950764308567366,
+      "grad_norm": 2.0,
+      "learning_rate": 1.6098222222222222e-05,
+      "loss": 0.3891,
+      "step": 8780
+    },
+    {
+      "epoch": 0.19529861357980804,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.609377777777778e-05,
+      "loss": 0.5055,
+      "step": 8790
+    },
+    {
+      "epoch": 0.1955207963028795,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.6089333333333335e-05,
+      "loss": 0.4405,
+      "step": 8800
+    },
+    {
+      "epoch": 0.19574297902595095,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.608488888888889e-05,
+      "loss": 0.4027,
+      "step": 8810
+    },
+    {
+      "epoch": 0.1959651617490224,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.6080444444444445e-05,
+      "loss": 0.4373,
+      "step": 8820
+    },
+    {
+      "epoch": 0.19618734447209385,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.6076e-05,
+      "loss": 0.4376,
+      "step": 8830
+    },
+    {
+      "epoch": 0.1964095271951653,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.6071555555555555e-05,
+      "loss": 0.4551,
+      "step": 8840
+    },
+    {
+      "epoch": 0.19663170991823675,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.6067111111111113e-05,
+      "loss": 0.4178,
+      "step": 8850
+    },
+    {
+      "epoch": 0.1968538926413082,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.6062666666666668e-05,
+      "loss": 0.403,
+      "step": 8860
+    },
+    {
+      "epoch": 0.19707607536437965,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.6058222222222223e-05,
+      "loss": 0.4423,
+      "step": 8870
+    },
+    {
+      "epoch": 0.1972982580874511,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.605377777777778e-05,
+      "loss": 0.4375,
+      "step": 8880
+    },
+    {
+      "epoch": 0.19752044081052259,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6049333333333336e-05,
+      "loss": 0.4115,
+      "step": 8890
+    },
+    {
+      "epoch": 0.19774262353359404,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.604488888888889e-05,
+      "loss": 0.4264,
+      "step": 8900
+    },
+    {
+      "epoch": 0.1979648062566655,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6040444444444445e-05,
+      "loss": 0.4018,
+      "step": 8910
+    },
+    {
+      "epoch": 0.19818698897973694,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.6036e-05,
+      "loss": 0.4006,
+      "step": 8920
+    },
+    {
+      "epoch": 0.1984091717028084,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6031555555555555e-05,
+      "loss": 0.4268,
+      "step": 8930
+    },
+    {
+      "epoch": 0.19863135442587984,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6027111111111113e-05,
+      "loss": 0.4188,
+      "step": 8940
+    },
+    {
+      "epoch": 0.1988535371489513,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.6022666666666668e-05,
+      "loss": 0.401,
+      "step": 8950
+    },
+    {
+      "epoch": 0.19907571987202274,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.6018222222222223e-05,
+      "loss": 0.4369,
+      "step": 8960
+    },
+    {
+      "epoch": 0.1992979025950942,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6013777777777778e-05,
+      "loss": 0.4732,
+      "step": 8970
+    },
+    {
+      "epoch": 0.19952008531816565,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.6009333333333336e-05,
+      "loss": 0.3873,
+      "step": 8980
+    },
+    {
+      "epoch": 0.19974226804123713,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.600488888888889e-05,
+      "loss": 0.4255,
+      "step": 8990
+    },
+    {
+      "epoch": 0.19996445076430858,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.6000444444444446e-05,
+      "loss": 0.4358,
+      "step": 9000
+    },
+    {
+      "epoch": 0.20018663348738003,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.5996e-05,
+      "loss": 0.4104,
+      "step": 9010
+    },
+    {
+      "epoch": 0.20040881621045148,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.599155555555556e-05,
+      "loss": 0.4249,
+      "step": 9020
+    },
+    {
+      "epoch": 0.20063099893352293,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.5987111111111114e-05,
+      "loss": 0.4467,
+      "step": 9030
+    },
+    {
+      "epoch": 0.20085318165659438,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.598266666666667e-05,
+      "loss": 0.4227,
+      "step": 9040
+    },
+    {
+      "epoch": 0.20107536437966583,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5978222222222223e-05,
+      "loss": 0.3909,
+      "step": 9050
+    },
+    {
+      "epoch": 0.20129754710273728,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5973777777777778e-05,
+      "loss": 0.4254,
+      "step": 9060
+    },
+    {
+      "epoch": 0.20151972982580874,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.5969333333333333e-05,
+      "loss": 0.4576,
+      "step": 9070
+    },
+    {
+      "epoch": 0.2017419125488802,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.596488888888889e-05,
+      "loss": 0.4059,
+      "step": 9080
+    },
+    {
+      "epoch": 0.20196409527195167,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5960444444444446e-05,
+      "loss": 0.4446,
+      "step": 9090
+    },
+    {
+      "epoch": 0.20218627799502312,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.5956e-05,
+      "loss": 0.4466,
+      "step": 9100
+    },
+    {
+      "epoch": 0.20240846071809457,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.595155555555556e-05,
+      "loss": 0.4282,
+      "step": 9110
+    },
+    {
+      "epoch": 0.20263064344116602,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5947111111111114e-05,
+      "loss": 0.4246,
+      "step": 9120
+    },
+    {
+      "epoch": 0.20285282616423747,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.594266666666667e-05,
+      "loss": 0.4582,
+      "step": 9130
+    },
+    {
+      "epoch": 0.20307500888730892,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.5938222222222224e-05,
+      "loss": 0.4167,
+      "step": 9140
+    },
+    {
+      "epoch": 0.20329719161038037,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.593377777777778e-05,
+      "loss": 0.4345,
+      "step": 9150
+    },
+    {
+      "epoch": 0.20351937433345182,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.5929333333333334e-05,
+      "loss": 0.4211,
+      "step": 9160
+    },
+    {
+      "epoch": 0.20374155705652328,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.592488888888889e-05,
+      "loss": 0.4329,
+      "step": 9170
+    },
+    {
+      "epoch": 0.20396373977959473,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.5920444444444447e-05,
+      "loss": 0.4123,
+      "step": 9180
+    },
+    {
+      "epoch": 0.2041859225026662,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5916e-05,
+      "loss": 0.4131,
+      "step": 9190
+    },
+    {
+      "epoch": 0.20440810522573766,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.5911555555555556e-05,
+      "loss": 0.3962,
+      "step": 9200
+    },
+    {
+      "epoch": 0.2046302879488091,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5907111111111115e-05,
+      "loss": 0.4045,
+      "step": 9210
+    },
+    {
+      "epoch": 0.20485247067188056,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.590266666666667e-05,
+      "loss": 0.4396,
+      "step": 9220
+    },
+    {
+      "epoch": 0.205074653394952,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5898222222222224e-05,
+      "loss": 0.4224,
+      "step": 9230
+    },
+    {
+      "epoch": 0.20529683611802346,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.589377777777778e-05,
+      "loss": 0.4518,
+      "step": 9240
+    },
+    {
+      "epoch": 0.2055190188410949,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.5889333333333334e-05,
+      "loss": 0.4198,
+      "step": 9250
+    },
+    {
+      "epoch": 0.20574120156416636,
+      "grad_norm": 2.125,
+      "learning_rate": 1.588488888888889e-05,
+      "loss": 0.4272,
+      "step": 9260
+    },
+    {
+      "epoch": 0.20596338428723782,
+      "grad_norm": 1.875,
+      "learning_rate": 1.5880444444444444e-05,
+      "loss": 0.4243,
+      "step": 9270
+    },
+    {
+      "epoch": 0.20618556701030927,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.5876000000000002e-05,
+      "loss": 0.3997,
+      "step": 9280
+    },
+    {
+      "epoch": 0.20640774973338072,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.5871555555555557e-05,
+      "loss": 0.4025,
+      "step": 9290
+    },
+    {
+      "epoch": 0.2066299324564522,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.586711111111111e-05,
+      "loss": 0.3916,
+      "step": 9300
+    },
+    {
+      "epoch": 0.20685211517952365,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.586266666666667e-05,
+      "loss": 0.4201,
+      "step": 9310
+    },
+    {
+      "epoch": 0.2070742979025951,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.5858222222222225e-05,
+      "loss": 0.4181,
+      "step": 9320
+    },
+    {
+      "epoch": 0.20729648062566655,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.585377777777778e-05,
+      "loss": 0.4092,
+      "step": 9330
+    },
+    {
+      "epoch": 0.207518663348738,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.5849333333333334e-05,
+      "loss": 0.4069,
+      "step": 9340
+    },
+    {
+      "epoch": 0.20774084607180945,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.584488888888889e-05,
+      "loss": 0.3788,
+      "step": 9350
+    },
+    {
+      "epoch": 0.2079630287948809,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.5840444444444444e-05,
+      "loss": 0.3968,
+      "step": 9360
+    },
+    {
+      "epoch": 0.20818521151795236,
+      "grad_norm": 1.875,
+      "learning_rate": 1.5836e-05,
+      "loss": 0.4224,
+      "step": 9370
+    },
+    {
+      "epoch": 0.2084073942410238,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5831555555555557e-05,
+      "loss": 0.4123,
+      "step": 9380
+    },
+    {
+      "epoch": 0.20862957696409526,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5827111111111112e-05,
+      "loss": 0.4285,
+      "step": 9390
+    },
+    {
+      "epoch": 0.20885175968716674,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.5822666666666667e-05,
+      "loss": 0.4053,
+      "step": 9400
+    },
+    {
+      "epoch": 0.2090739424102382,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.5818222222222225e-05,
+      "loss": 0.3969,
+      "step": 9410
+    },
+    {
+      "epoch": 0.20929612513330964,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.581377777777778e-05,
+      "loss": 0.4062,
+      "step": 9420
+    },
+    {
+      "epoch": 0.2095183078563811,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5809333333333335e-05,
+      "loss": 0.4248,
+      "step": 9430
+    },
+    {
+      "epoch": 0.20974049057945254,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.580488888888889e-05,
+      "loss": 0.4129,
+      "step": 9440
+    },
+    {
+      "epoch": 0.209962673302524,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.5800444444444444e-05,
+      "loss": 0.4518,
+      "step": 9450
+    },
+    {
+      "epoch": 0.21018485602559545,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5796e-05,
+      "loss": 0.4386,
+      "step": 9460
+    },
+    {
+      "epoch": 0.2104070387486669,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5791555555555557e-05,
+      "loss": 0.42,
+      "step": 9470
+    },
+    {
+      "epoch": 0.21062922147173835,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.5787111111111112e-05,
+      "loss": 0.4053,
+      "step": 9480
+    },
+    {
+      "epoch": 0.2108514041948098,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.5782666666666667e-05,
+      "loss": 0.424,
+      "step": 9490
+    },
+    {
+      "epoch": 0.21107358691788128,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.5778222222222225e-05,
+      "loss": 0.4544,
+      "step": 9500
+    },
+    {
+      "epoch": 0.21129576964095273,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.577377777777778e-05,
+      "loss": 0.4027,
+      "step": 9510
+    },
+    {
+      "epoch": 0.21151795236402418,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.5769333333333335e-05,
+      "loss": 0.4137,
+      "step": 9520
+    },
+    {
+      "epoch": 0.21174013508709563,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.576488888888889e-05,
+      "loss": 0.4224,
+      "step": 9530
+    },
+    {
+      "epoch": 0.21196231781016708,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.5760444444444445e-05,
+      "loss": 0.4379,
+      "step": 9540
+    },
+    {
+      "epoch": 0.21218450053323853,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5756e-05,
+      "loss": 0.4296,
+      "step": 9550
+    },
+    {
+      "epoch": 0.21240668325630999,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.5751555555555558e-05,
+      "loss": 0.4358,
+      "step": 9560
+    },
+    {
+      "epoch": 0.21262886597938144,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5747111111111113e-05,
+      "loss": 0.4208,
+      "step": 9570
+    },
+    {
+      "epoch": 0.2128510487024529,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.5742666666666668e-05,
+      "loss": 0.4214,
+      "step": 9580
+    },
+    {
+      "epoch": 0.21307323142552434,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5738222222222222e-05,
+      "loss": 0.3849,
+      "step": 9590
+    },
+    {
+      "epoch": 0.21329541414859582,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.573377777777778e-05,
+      "loss": 0.4297,
+      "step": 9600
+    },
+    {
+      "epoch": 0.21351759687166727,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5729333333333335e-05,
+      "loss": 0.4194,
+      "step": 9610
+    },
+    {
+      "epoch": 0.21373977959473872,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.572488888888889e-05,
+      "loss": 0.3898,
+      "step": 9620
+    },
+    {
+      "epoch": 0.21396196231781017,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5720444444444445e-05,
+      "loss": 0.4226,
+      "step": 9630
+    },
+    {
+      "epoch": 0.21418414504088162,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.5716000000000003e-05,
+      "loss": 0.4438,
+      "step": 9640
+    },
+    {
+      "epoch": 0.21440632776395308,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5711555555555558e-05,
+      "loss": 0.4744,
+      "step": 9650
+    },
+    {
+      "epoch": 0.21462851048702453,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5707111111111113e-05,
+      "loss": 0.4753,
+      "step": 9660
+    },
+    {
+      "epoch": 0.21485069321009598,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5702666666666668e-05,
+      "loss": 0.4282,
+      "step": 9670
+    },
+    {
+      "epoch": 0.21507287593316743,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.5698222222222223e-05,
+      "loss": 0.4125,
+      "step": 9680
+    },
+    {
+      "epoch": 0.21529505865623888,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.5693777777777778e-05,
+      "loss": 0.3632,
+      "step": 9690
+    },
+    {
+      "epoch": 0.21551724137931033,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.5689333333333336e-05,
+      "loss": 0.4184,
+      "step": 9700
+    },
+    {
+      "epoch": 0.2157394241023818,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.568488888888889e-05,
+      "loss": 0.431,
+      "step": 9710
+    },
+    {
+      "epoch": 0.21596160682545326,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5680444444444445e-05,
+      "loss": 0.4466,
+      "step": 9720
+    },
+    {
+      "epoch": 0.2161837895485247,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5676000000000004e-05,
+      "loss": 0.4611,
+      "step": 9730
+    },
+    {
+      "epoch": 0.21640597227159616,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.567155555555556e-05,
+      "loss": 0.4283,
+      "step": 9740
+    },
+    {
+      "epoch": 0.21662815499466762,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.5667111111111113e-05,
+      "loss": 0.4084,
+      "step": 9750
+    },
+    {
+      "epoch": 0.21685033771773907,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.5662666666666668e-05,
+      "loss": 0.4268,
+      "step": 9760
+    },
+    {
+      "epoch": 0.21707252044081052,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5658222222222223e-05,
+      "loss": 0.3887,
+      "step": 9770
+    },
+    {
+      "epoch": 0.21729470316388197,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.5653777777777778e-05,
+      "loss": 0.3958,
+      "step": 9780
+    },
+    {
+      "epoch": 0.21751688588695342,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.5649333333333333e-05,
+      "loss": 0.4138,
+      "step": 9790
+    },
+    {
+      "epoch": 0.21773906861002487,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.564488888888889e-05,
+      "loss": 0.4339,
+      "step": 9800
+    },
+    {
+      "epoch": 0.21796125133309635,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5640444444444446e-05,
+      "loss": 0.414,
+      "step": 9810
+    },
+    {
+      "epoch": 0.2181834340561678,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.5636e-05,
+      "loss": 0.4075,
+      "step": 9820
+    },
+    {
+      "epoch": 0.21840561677923925,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.563155555555556e-05,
+      "loss": 0.39,
+      "step": 9830
+    },
+    {
+      "epoch": 0.2186277995023107,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5627111111111114e-05,
+      "loss": 0.4215,
+      "step": 9840
+    },
+    {
+      "epoch": 0.21884998222538216,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.562266666666667e-05,
+      "loss": 0.4007,
+      "step": 9850
+    },
+    {
+      "epoch": 0.2190721649484536,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5618222222222223e-05,
+      "loss": 0.391,
+      "step": 9860
+    },
+    {
+      "epoch": 0.21929434767152506,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.561377777777778e-05,
+      "loss": 0.4381,
+      "step": 9870
+    },
+    {
+      "epoch": 0.2195165303945965,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5609333333333333e-05,
+      "loss": 0.4479,
+      "step": 9880
+    },
+    {
+      "epoch": 0.21973871311766796,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.5604888888888888e-05,
+      "loss": 0.4425,
+      "step": 9890
+    },
+    {
+      "epoch": 0.2199608958407394,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5600444444444446e-05,
+      "loss": 0.4015,
+      "step": 9900
+    },
+    {
+      "epoch": 0.2201830785638109,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.5596e-05,
+      "loss": 0.4212,
+      "step": 9910
+    },
+    {
+      "epoch": 0.22040526128688234,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5591555555555556e-05,
+      "loss": 0.4013,
+      "step": 9920
+    },
+    {
+      "epoch": 0.2206274440099538,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.5587111111111114e-05,
+      "loss": 0.4078,
+      "step": 9930
+    },
+    {
+      "epoch": 0.22084962673302524,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.558266666666667e-05,
+      "loss": 0.4728,
+      "step": 9940
+    },
+    {
+      "epoch": 0.2210718094560967,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.5578222222222224e-05,
+      "loss": 0.4154,
+      "step": 9950
+    },
+    {
+      "epoch": 0.22129399217916815,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.557377777777778e-05,
+      "loss": 0.3855,
+      "step": 9960
+    },
+    {
+      "epoch": 0.2215161749022396,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.5569333333333334e-05,
+      "loss": 0.4627,
+      "step": 9970
+    },
+    {
+      "epoch": 0.22173835762531105,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.556488888888889e-05,
+      "loss": 0.4345,
+      "step": 9980
+    },
+    {
+      "epoch": 0.2219605403483825,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.5560444444444443e-05,
+      "loss": 0.4055,
+      "step": 9990
+    },
+    {
+      "epoch": 0.22218272307145395,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5556e-05,
+      "loss": 0.4263,
+      "step": 10000
+    },
+    {
+      "epoch": 0.22240490579452543,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5551555555555556e-05,
+      "loss": 0.3832,
+      "step": 10010
+    },
+    {
+      "epoch": 0.22262708851759688,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.554711111111111e-05,
+      "loss": 0.3835,
+      "step": 10020
+    },
+    {
+      "epoch": 0.22284927124066833,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.554266666666667e-05,
+      "loss": 0.3895,
+      "step": 10030
+    },
+    {
+      "epoch": 0.22307145396373979,
+      "grad_norm": 2.0,
+      "learning_rate": 1.5538222222222224e-05,
+      "loss": 0.4213,
+      "step": 10040
+    },
+    {
+      "epoch": 0.22329363668681124,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.553377777777778e-05,
+      "loss": 0.3773,
+      "step": 10050
+    },
+    {
+      "epoch": 0.2235158194098827,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.5529333333333334e-05,
+      "loss": 0.3892,
+      "step": 10060
+    },
+    {
+      "epoch": 0.22373800213295414,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.552488888888889e-05,
+      "loss": 0.4141,
+      "step": 10070
+    },
+    {
+      "epoch": 0.2239601848560256,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.5520444444444444e-05,
+      "loss": 0.3856,
+      "step": 10080
+    },
+    {
+      "epoch": 0.22418236757909704,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.5516000000000002e-05,
+      "loss": 0.4099,
+      "step": 10090
+    },
+    {
+      "epoch": 0.2244045503021685,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.5511555555555557e-05,
+      "loss": 0.4171,
+      "step": 10100
+    },
+    {
+      "epoch": 0.22462673302523994,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.550711111111111e-05,
+      "loss": 0.4606,
+      "step": 10110
+    },
+    {
+      "epoch": 0.22484891574831142,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5502666666666666e-05,
+      "loss": 0.4292,
+      "step": 10120
+    },
+    {
+      "epoch": 0.22507109847138287,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.5498222222222225e-05,
+      "loss": 0.3862,
+      "step": 10130
+    },
+    {
+      "epoch": 0.22529328119445433,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.549377777777778e-05,
+      "loss": 0.4293,
+      "step": 10140
+    },
+    {
+      "epoch": 0.22551546391752578,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.5489333333333334e-05,
+      "loss": 0.4031,
+      "step": 10150
+    },
+    {
+      "epoch": 0.22573764664059723,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.548488888888889e-05,
+      "loss": 0.4126,
+      "step": 10160
+    },
+    {
+      "epoch": 0.22595982936366868,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.5480444444444444e-05,
+      "loss": 0.4381,
+      "step": 10170
+    },
+    {
+      "epoch": 0.22618201208674013,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.5476000000000002e-05,
+      "loss": 0.4138,
+      "step": 10180
+    },
+    {
+      "epoch": 0.22640419480981158,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5471555555555557e-05,
+      "loss": 0.4036,
+      "step": 10190
+    },
+    {
+      "epoch": 0.22662637753288303,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5467111111111112e-05,
+      "loss": 0.4509,
+      "step": 10200
+    },
+    {
+      "epoch": 0.22684856025595448,
+      "grad_norm": 1.75,
+      "learning_rate": 1.5462666666666667e-05,
+      "loss": 0.4349,
+      "step": 10210
+    },
+    {
+      "epoch": 0.22707074297902596,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5458222222222225e-05,
+      "loss": 0.4387,
+      "step": 10220
+    },
+    {
+      "epoch": 0.22729292570209741,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.545377777777778e-05,
+      "loss": 0.3939,
+      "step": 10230
+    },
+    {
+      "epoch": 0.22751510842516887,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.5449333333333335e-05,
+      "loss": 0.3793,
+      "step": 10240
+    },
+    {
+      "epoch": 0.22773729114824032,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.544488888888889e-05,
+      "loss": 0.4314,
+      "step": 10250
+    },
+    {
+      "epoch": 0.22795947387131177,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5440444444444448e-05,
+      "loss": 0.4209,
+      "step": 10260
+    },
+    {
+      "epoch": 0.22818165659438322,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.5436000000000003e-05,
+      "loss": 0.4622,
+      "step": 10270
+    },
+    {
+      "epoch": 0.22840383931745467,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.5431555555555557e-05,
+      "loss": 0.4193,
+      "step": 10280
+    },
+    {
+      "epoch": 0.22862602204052612,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.5427111111111112e-05,
+      "loss": 0.4271,
+      "step": 10290
+    },
+    {
+      "epoch": 0.22884820476359757,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.5422666666666667e-05,
+      "loss": 0.4034,
+      "step": 10300
+    },
+    {
+      "epoch": 0.22907038748666902,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.5418222222222222e-05,
+      "loss": 0.4247,
+      "step": 10310
+    },
+    {
+      "epoch": 0.2292925702097405,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.541377777777778e-05,
+      "loss": 0.4069,
+      "step": 10320
+    },
+    {
+      "epoch": 0.22951475293281196,
+      "grad_norm": 2.0,
+      "learning_rate": 1.5409333333333335e-05,
+      "loss": 0.4488,
+      "step": 10330
+    },
+    {
+      "epoch": 0.2297369356558834,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.540488888888889e-05,
+      "loss": 0.4234,
+      "step": 10340
+    },
+    {
+      "epoch": 0.22995911837895486,
+      "grad_norm": 1.875,
+      "learning_rate": 1.5400444444444448e-05,
+      "loss": 0.4164,
+      "step": 10350
+    },
+    {
+      "epoch": 0.2301813011020263,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5396000000000003e-05,
+      "loss": 0.4142,
+      "step": 10360
+    },
+    {
+      "epoch": 0.23040348382509776,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5391555555555558e-05,
+      "loss": 0.401,
+      "step": 10370
+    },
+    {
+      "epoch": 0.2306256665481692,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5387111111111113e-05,
+      "loss": 0.3975,
+      "step": 10380
+    },
+    {
+      "epoch": 0.23084784927124066,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.5382666666666668e-05,
+      "loss": 0.4395,
+      "step": 10390
+    },
+    {
+      "epoch": 0.23107003199431211,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.5378222222222222e-05,
+      "loss": 0.4033,
+      "step": 10400
+    },
+    {
+      "epoch": 0.23129221471738357,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.5373777777777777e-05,
+      "loss": 0.4159,
+      "step": 10410
+    },
+    {
+      "epoch": 0.23151439744045502,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.5369333333333335e-05,
+      "loss": 0.3769,
+      "step": 10420
+    },
+    {
+      "epoch": 0.2317365801635265,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.536488888888889e-05,
+      "loss": 0.3959,
+      "step": 10430
+    },
+    {
+      "epoch": 0.23195876288659795,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.5360444444444445e-05,
+      "loss": 0.3965,
+      "step": 10440
+    },
+    {
+      "epoch": 0.2321809456096694,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.5356000000000003e-05,
+      "loss": 0.4088,
+      "step": 10450
+    },
+    {
+      "epoch": 0.23240312833274085,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5351555555555558e-05,
+      "loss": 0.3845,
+      "step": 10460
+    },
+    {
+      "epoch": 0.2326253110558123,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.5347111111111113e-05,
+      "loss": 0.4088,
+      "step": 10470
+    },
+    {
+      "epoch": 0.23284749377888375,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5342666666666668e-05,
+      "loss": 0.4099,
+      "step": 10480
+    },
+    {
+      "epoch": 0.2330696765019552,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.5338222222222223e-05,
+      "loss": 0.4516,
+      "step": 10490
+    },
+    {
+      "epoch": 0.23329185922502665,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.5333777777777778e-05,
+      "loss": 0.4197,
+      "step": 10500
+    },
+    {
+      "epoch": 0.2335140419480981,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5329333333333332e-05,
+      "loss": 0.4455,
+      "step": 10510
+    },
+    {
+      "epoch": 0.23373622467116956,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.532488888888889e-05,
+      "loss": 0.391,
+      "step": 10520
+    },
+    {
+      "epoch": 0.23395840739424104,
+      "grad_norm": 1.875,
+      "learning_rate": 1.5320444444444446e-05,
+      "loss": 0.4127,
+      "step": 10530
+    },
+    {
+      "epoch": 0.2341805901173125,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.5316e-05,
+      "loss": 0.4628,
+      "step": 10540
+    },
+    {
+      "epoch": 0.23440277284038394,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.531155555555556e-05,
+      "loss": 0.414,
+      "step": 10550
+    },
+    {
+      "epoch": 0.2346249555634554,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5307111111111113e-05,
+      "loss": 0.4388,
+      "step": 10560
+    },
+    {
+      "epoch": 0.23484713828652684,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.530266666666667e-05,
+      "loss": 0.4559,
+      "step": 10570
+    },
+    {
+      "epoch": 0.2350693210095983,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.5298222222222223e-05,
+      "loss": 0.429,
+      "step": 10580
+    },
+    {
+      "epoch": 0.23529150373266974,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5293777777777778e-05,
+      "loss": 0.4091,
+      "step": 10590
+    },
+    {
+      "epoch": 0.2355136864557412,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5289333333333333e-05,
+      "loss": 0.3757,
+      "step": 10600
+    },
+    {
+      "epoch": 0.23573586917881265,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.5284888888888888e-05,
+      "loss": 0.4064,
+      "step": 10610
+    },
+    {
+      "epoch": 0.2359580519018841,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.5280444444444446e-05,
+      "loss": 0.4397,
+      "step": 10620
+    },
+    {
+      "epoch": 0.23618023462495558,
+      "grad_norm": 2.0,
+      "learning_rate": 1.5276e-05,
+      "loss": 0.4588,
+      "step": 10630
+    },
+    {
+      "epoch": 0.23640241734802703,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.5271555555555556e-05,
+      "loss": 0.4108,
+      "step": 10640
+    },
+    {
+      "epoch": 0.23662460007109848,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5267111111111114e-05,
+      "loss": 0.4755,
+      "step": 10650
+    },
+    {
+      "epoch": 0.23684678279416993,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.526266666666667e-05,
+      "loss": 0.4257,
+      "step": 10660
+    },
+    {
+      "epoch": 0.23706896551724138,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.5258222222222224e-05,
+      "loss": 0.403,
+      "step": 10670
+    },
+    {
+      "epoch": 0.23729114824031283,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5253777777777778e-05,
+      "loss": 0.4352,
+      "step": 10680
+    },
+    {
+      "epoch": 0.23751333096338428,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5249333333333333e-05,
+      "loss": 0.4014,
+      "step": 10690
+    },
+    {
+      "epoch": 0.23773551368645574,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.524488888888889e-05,
+      "loss": 0.4451,
+      "step": 10700
+    },
+    {
+      "epoch": 0.2379576964095272,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.5240444444444446e-05,
+      "loss": 0.4506,
+      "step": 10710
+    },
+    {
+      "epoch": 0.23817987913259864,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.5236000000000001e-05,
+      "loss": 0.4418,
+      "step": 10720
+    },
+    {
+      "epoch": 0.23840206185567012,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.5231555555555558e-05,
+      "loss": 0.3964,
+      "step": 10730
+    },
+    {
+      "epoch": 0.23862424457874157,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.5227111111111113e-05,
+      "loss": 0.429,
+      "step": 10740
+    },
+    {
+      "epoch": 0.23884642730181302,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.5222666666666667e-05,
+      "loss": 0.3924,
+      "step": 10750
+    },
+    {
+      "epoch": 0.23906861002488447,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.5218222222222224e-05,
+      "loss": 0.4009,
+      "step": 10760
+    },
+    {
+      "epoch": 0.23929079274795592,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5213777777777779e-05,
+      "loss": 0.411,
+      "step": 10770
+    },
+    {
+      "epoch": 0.23951297547102737,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.5209333333333334e-05,
+      "loss": 0.4171,
+      "step": 10780
+    },
+    {
+      "epoch": 0.23973515819409882,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.5204888888888888e-05,
+      "loss": 0.4003,
+      "step": 10790
+    },
+    {
+      "epoch": 0.23995734091717028,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.5200444444444447e-05,
+      "loss": 0.3866,
+      "step": 10800
+    },
+    {
+      "epoch": 0.24017952364024173,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5196000000000002e-05,
+      "loss": 0.4146,
+      "step": 10810
+    },
+    {
+      "epoch": 0.24040170636331318,
+      "grad_norm": 1.75,
+      "learning_rate": 1.5191555555555556e-05,
+      "loss": 0.3649,
+      "step": 10820
+    },
+    {
+      "epoch": 0.24062388908638463,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5187111111111113e-05,
+      "loss": 0.4661,
+      "step": 10830
+    },
+    {
+      "epoch": 0.2408460718094561,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.5182666666666668e-05,
+      "loss": 0.4078,
+      "step": 10840
+    },
+    {
+      "epoch": 0.24106825453252756,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.5178222222222223e-05,
+      "loss": 0.401,
+      "step": 10850
+    },
+    {
+      "epoch": 0.241290437255599,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.517377777777778e-05,
+      "loss": 0.4201,
+      "step": 10860
+    },
+    {
+      "epoch": 0.24151261997867046,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.5169333333333334e-05,
+      "loss": 0.3816,
+      "step": 10870
+    },
+    {
+      "epoch": 0.2417348027017419,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.516488888888889e-05,
+      "loss": 0.4488,
+      "step": 10880
+    },
+    {
+      "epoch": 0.24195698542481336,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5160444444444447e-05,
+      "loss": 0.3732,
+      "step": 10890
+    },
+    {
+      "epoch": 0.24217916814788482,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5156000000000002e-05,
+      "loss": 0.4348,
+      "step": 10900
+    },
+    {
+      "epoch": 0.24240135087095627,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5151555555555557e-05,
+      "loss": 0.4499,
+      "step": 10910
+    },
+    {
+      "epoch": 0.24262353359402772,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.5147111111111112e-05,
+      "loss": 0.4267,
+      "step": 10920
+    },
+    {
+      "epoch": 0.24284571631709917,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.5142666666666668e-05,
+      "loss": 0.4251,
+      "step": 10930
+    },
+    {
+      "epoch": 0.24306789904017065,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5138222222222223e-05,
+      "loss": 0.3651,
+      "step": 10940
+    },
+    {
+      "epoch": 0.2432900817632421,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.5133777777777778e-05,
+      "loss": 0.3934,
+      "step": 10950
+    },
+    {
+      "epoch": 0.24351226448631355,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5129333333333334e-05,
+      "loss": 0.3921,
+      "step": 10960
+    },
+    {
+      "epoch": 0.243734447209385,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.5124888888888891e-05,
+      "loss": 0.4099,
+      "step": 10970
+    },
+    {
+      "epoch": 0.24395662993245645,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.5120444444444446e-05,
+      "loss": 0.4271,
+      "step": 10980
+    },
+    {
+      "epoch": 0.2441788126555279,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5116000000000002e-05,
+      "loss": 0.4147,
+      "step": 10990
+    },
+    {
+      "epoch": 0.24440099537859936,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.5111555555555557e-05,
+      "loss": 0.411,
+      "step": 11000
+    },
+    {
+      "epoch": 0.2446231781016708,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5107111111111112e-05,
+      "loss": 0.3897,
+      "step": 11010
+    },
+    {
+      "epoch": 0.24484536082474226,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5102666666666667e-05,
+      "loss": 0.4195,
+      "step": 11020
+    },
+    {
+      "epoch": 0.2450675435478137,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5098222222222223e-05,
+      "loss": 0.3904,
+      "step": 11030
+    },
+    {
+      "epoch": 0.2452897262708852,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5093777777777778e-05,
+      "loss": 0.3475,
+      "step": 11040
+    },
+    {
+      "epoch": 0.24551190899395664,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.5089333333333333e-05,
+      "loss": 0.4163,
+      "step": 11050
+    },
+    {
+      "epoch": 0.2457340917170281,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.5084888888888891e-05,
+      "loss": 0.4046,
+      "step": 11060
+    },
+    {
+      "epoch": 0.24595627444009954,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.5080444444444446e-05,
+      "loss": 0.4096,
+      "step": 11070
+    },
+    {
+      "epoch": 0.246178457163171,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.5076000000000001e-05,
+      "loss": 0.4577,
+      "step": 11080
+    },
+    {
+      "epoch": 0.24640063988624245,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5071555555555558e-05,
+      "loss": 0.3678,
+      "step": 11090
+    },
+    {
+      "epoch": 0.2466228226093139,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.5067111111111112e-05,
+      "loss": 0.4472,
+      "step": 11100
+    },
+    {
+      "epoch": 0.24684500533238535,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5062666666666667e-05,
+      "loss": 0.4231,
+      "step": 11110
+    },
+    {
+      "epoch": 0.2470671880554568,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5058222222222224e-05,
+      "loss": 0.3996,
+      "step": 11120
+    },
+    {
+      "epoch": 0.24728937077852825,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.5053777777777779e-05,
+      "loss": 0.4105,
+      "step": 11130
+    },
+    {
+      "epoch": 0.24751155350159973,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.5049333333333333e-05,
+      "loss": 0.4256,
+      "step": 11140
+    },
+    {
+      "epoch": 0.24773373622467118,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5044888888888892e-05,
+      "loss": 0.4418,
+      "step": 11150
+    },
+    {
+      "epoch": 0.24795591894774263,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5040444444444447e-05,
+      "loss": 0.4127,
+      "step": 11160
+    },
+    {
+      "epoch": 0.24817810167081408,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5036000000000001e-05,
+      "loss": 0.4138,
+      "step": 11170
+    },
+    {
+      "epoch": 0.24840028439388553,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5031555555555556e-05,
+      "loss": 0.4248,
+      "step": 11180
+    },
+    {
+      "epoch": 0.24862246711695699,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.5027111111111113e-05,
+      "loss": 0.3865,
+      "step": 11190
+    },
+    {
+      "epoch": 0.24884464984002844,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5022666666666668e-05,
+      "loss": 0.4174,
+      "step": 11200
+    },
+    {
+      "epoch": 0.2490668325630999,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5018222222222222e-05,
+      "loss": 0.393,
+      "step": 11210
+    },
+    {
+      "epoch": 0.24928901528617134,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.5013777777777779e-05,
+      "loss": 0.4152,
+      "step": 11220
+    },
+    {
+      "epoch": 0.2495111980092428,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.5009333333333334e-05,
+      "loss": 0.4029,
+      "step": 11230
+    },
+    {
+      "epoch": 0.24973338073231424,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.500488888888889e-05,
+      "loss": 0.3629,
+      "step": 11240
+    },
+    {
+      "epoch": 0.24995556345538572,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5000444444444447e-05,
+      "loss": 0.3671,
+      "step": 11250
+    },
+    {
+      "epoch": 0.25017774617845717,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.4996000000000002e-05,
+      "loss": 0.413,
+      "step": 11260
+    },
+    {
+      "epoch": 0.2503999289015286,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.4991555555555557e-05,
+      "loss": 0.404,
+      "step": 11270
+    },
+    {
+      "epoch": 0.2506221116246001,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4987111111111111e-05,
+      "loss": 0.4044,
+      "step": 11280
+    },
+    {
+      "epoch": 0.2508442943476715,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.4982666666666668e-05,
+      "loss": 0.4293,
+      "step": 11290
+    },
+    {
+      "epoch": 0.251066477070743,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.4978222222222223e-05,
+      "loss": 0.3781,
+      "step": 11300
+    },
+    {
+      "epoch": 0.25128865979381443,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.4973777777777778e-05,
+      "loss": 0.458,
+      "step": 11310
+    },
+    {
+      "epoch": 0.2515108425168859,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4969333333333334e-05,
+      "loss": 0.4059,
+      "step": 11320
+    },
+    {
+      "epoch": 0.25173302523995733,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.496488888888889e-05,
+      "loss": 0.3664,
+      "step": 11330
+    },
+    {
+      "epoch": 0.2519552079630288,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4960444444444446e-05,
+      "loss": 0.4389,
+      "step": 11340
+    },
+    {
+      "epoch": 0.25217739068610023,
+      "grad_norm": 2.625,
+      "learning_rate": 1.4956000000000002e-05,
+      "loss": 0.4298,
+      "step": 11350
+    },
+    {
+      "epoch": 0.2523995734091717,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4951555555555557e-05,
+      "loss": 0.4423,
+      "step": 11360
+    },
+    {
+      "epoch": 0.25262175613224314,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4947111111111112e-05,
+      "loss": 0.4159,
+      "step": 11370
+    },
+    {
+      "epoch": 0.2528439388553146,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.4942666666666668e-05,
+      "loss": 0.3997,
+      "step": 11380
+    },
+    {
+      "epoch": 0.25306612157838604,
+      "grad_norm": 2.0,
+      "learning_rate": 1.4938222222222223e-05,
+      "loss": 0.3869,
+      "step": 11390
+    },
+    {
+      "epoch": 0.25328830430145755,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4933777777777778e-05,
+      "loss": 0.4579,
+      "step": 11400
+    },
+    {
+      "epoch": 0.253510487024529,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4929333333333333e-05,
+      "loss": 0.4292,
+      "step": 11410
+    },
+    {
+      "epoch": 0.25373266974760045,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4924888888888891e-05,
+      "loss": 0.4276,
+      "step": 11420
+    },
+    {
+      "epoch": 0.2539548524706719,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4920444444444446e-05,
+      "loss": 0.3931,
+      "step": 11430
+    },
+    {
+      "epoch": 0.25417703519374335,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.4916e-05,
+      "loss": 0.4296,
+      "step": 11440
+    },
+    {
+      "epoch": 0.2543992179168148,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4911555555555557e-05,
+      "loss": 0.3877,
+      "step": 11450
+    },
+    {
+      "epoch": 0.25462140063988625,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4907111111111112e-05,
+      "loss": 0.4044,
+      "step": 11460
+    },
+    {
+      "epoch": 0.2548435833629577,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4902666666666667e-05,
+      "loss": 0.4242,
+      "step": 11470
+    },
+    {
+      "epoch": 0.25506576608602916,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4898222222222224e-05,
+      "loss": 0.3976,
+      "step": 11480
+    },
+    {
+      "epoch": 0.2552879488091006,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.4893777777777778e-05,
+      "loss": 0.4272,
+      "step": 11490
+    },
+    {
+      "epoch": 0.25551013153217206,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.4889333333333335e-05,
+      "loss": 0.4176,
+      "step": 11500
+    },
+    {
+      "epoch": 0.2557323142552435,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.4884888888888892e-05,
+      "loss": 0.4055,
+      "step": 11510
+    },
+    {
+      "epoch": 0.25595449697831496,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.4880444444444446e-05,
+      "loss": 0.4689,
+      "step": 11520
+    },
+    {
+      "epoch": 0.2561766797013864,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.4876000000000001e-05,
+      "loss": 0.4248,
+      "step": 11530
+    },
+    {
+      "epoch": 0.25639886242445786,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.4871555555555556e-05,
+      "loss": 0.3976,
+      "step": 11540
+    },
+    {
+      "epoch": 0.2566210451475293,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.4867111111111113e-05,
+      "loss": 0.4339,
+      "step": 11550
+    },
+    {
+      "epoch": 0.25684322787060077,
+      "grad_norm": 2.375,
+      "learning_rate": 1.4862666666666667e-05,
+      "loss": 0.3969,
+      "step": 11560
+    },
+    {
+      "epoch": 0.2570654105936722,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.4858222222222222e-05,
+      "loss": 0.4235,
+      "step": 11570
+    },
+    {
+      "epoch": 0.25728759331674367,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.4853777777777779e-05,
+      "loss": 0.4033,
+      "step": 11580
+    },
+    {
+      "epoch": 0.2575097760398151,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4849333333333335e-05,
+      "loss": 0.3886,
+      "step": 11590
+    },
+    {
+      "epoch": 0.25773195876288657,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.484488888888889e-05,
+      "loss": 0.3896,
+      "step": 11600
+    },
+    {
+      "epoch": 0.2579541414859581,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.4840444444444447e-05,
+      "loss": 0.3802,
+      "step": 11610
+    },
+    {
+      "epoch": 0.25817632420902953,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4836000000000002e-05,
+      "loss": 0.4334,
+      "step": 11620
+    },
+    {
+      "epoch": 0.258398506932101,
+      "grad_norm": 2.0,
+      "learning_rate": 1.4831555555555556e-05,
+      "loss": 0.4029,
+      "step": 11630
+    },
+    {
+      "epoch": 0.25862068965517243,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.4827111111111111e-05,
+      "loss": 0.3693,
+      "step": 11640
+    },
+    {
+      "epoch": 0.2588428723782439,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.4822666666666668e-05,
+      "loss": 0.4093,
+      "step": 11650
+    },
+    {
+      "epoch": 0.25906505510131533,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.4818222222222223e-05,
+      "loss": 0.4496,
+      "step": 11660
+    },
+    {
+      "epoch": 0.2592872378243868,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.4813777777777778e-05,
+      "loss": 0.405,
+      "step": 11670
+    },
+    {
+      "epoch": 0.25950942054745824,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4809333333333336e-05,
+      "loss": 0.3803,
+      "step": 11680
+    },
+    {
+      "epoch": 0.2597316032705297,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.480488888888889e-05,
+      "loss": 0.3986,
+      "step": 11690
+    },
+    {
+      "epoch": 0.25995378599360114,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4800444444444445e-05,
+      "loss": 0.4121,
+      "step": 11700
+    },
+    {
+      "epoch": 0.2601759687166726,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4796000000000002e-05,
+      "loss": 0.3735,
+      "step": 11710
+    },
+    {
+      "epoch": 0.26039815143974404,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4791555555555557e-05,
+      "loss": 0.396,
+      "step": 11720
+    },
+    {
+      "epoch": 0.2606203341628155,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4787111111111112e-05,
+      "loss": 0.3862,
+      "step": 11730
+    },
+    {
+      "epoch": 0.26084251688588694,
+      "grad_norm": 2.25,
+      "learning_rate": 1.4782666666666668e-05,
+      "loss": 0.3994,
+      "step": 11740
+    },
+    {
+      "epoch": 0.2610646996089584,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.4778222222222223e-05,
+      "loss": 0.3949,
+      "step": 11750
+    },
+    {
+      "epoch": 0.26128688233202985,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4773777777777778e-05,
+      "loss": 0.4026,
+      "step": 11760
+    },
+    {
+      "epoch": 0.2615090650551013,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.4769333333333336e-05,
+      "loss": 0.4095,
+      "step": 11770
+    },
+    {
+      "epoch": 0.26173124777817275,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.4764888888888891e-05,
+      "loss": 0.4203,
+      "step": 11780
+    },
+    {
+      "epoch": 0.2619534305012442,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.4760444444444446e-05,
+      "loss": 0.3837,
+      "step": 11790
+    },
+    {
+      "epoch": 0.26217561322431565,
+      "grad_norm": 2.25,
+      "learning_rate": 1.4756e-05,
+      "loss": 0.4313,
+      "step": 11800
+    },
+    {
+      "epoch": 0.26239779594738716,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4751555555555557e-05,
+      "loss": 0.4183,
+      "step": 11810
+    },
+    {
+      "epoch": 0.2626199786704586,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.4747111111111112e-05,
+      "loss": 0.4251,
+      "step": 11820
+    },
+    {
+      "epoch": 0.26284216139353006,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.4742666666666667e-05,
+      "loss": 0.4186,
+      "step": 11830
+    },
+    {
+      "epoch": 0.2630643441166015,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4738222222222223e-05,
+      "loss": 0.3983,
+      "step": 11840
+    },
+    {
+      "epoch": 0.26328652683967296,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4733777777777778e-05,
+      "loss": 0.4149,
+      "step": 11850
+    },
+    {
+      "epoch": 0.2635087095627444,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.4729333333333335e-05,
+      "loss": 0.3967,
+      "step": 11860
+    },
+    {
+      "epoch": 0.26373089228581587,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4724888888888891e-05,
+      "loss": 0.4256,
+      "step": 11870
+    },
+    {
+      "epoch": 0.2639530750088873,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4720444444444446e-05,
+      "loss": 0.4262,
+      "step": 11880
+    },
+    {
+      "epoch": 0.26417525773195877,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4716000000000001e-05,
+      "loss": 0.4139,
+      "step": 11890
+    },
+    {
+      "epoch": 0.2643974404550302,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4711555555555556e-05,
+      "loss": 0.4439,
+      "step": 11900
+    },
+    {
+      "epoch": 0.26461962317810167,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.4707111111111112e-05,
+      "loss": 0.4166,
+      "step": 11910
+    },
+    {
+      "epoch": 0.2648418059011731,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.4702666666666667e-05,
+      "loss": 0.4344,
+      "step": 11920
+    },
+    {
+      "epoch": 0.2650639886242446,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.4698222222222222e-05,
+      "loss": 0.3895,
+      "step": 11930
+    },
+    {
+      "epoch": 0.265286171347316,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4693777777777779e-05,
+      "loss": 0.4079,
+      "step": 11940
+    },
+    {
+      "epoch": 0.2655083540703875,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4689333333333335e-05,
+      "loss": 0.4007,
+      "step": 11950
+    },
+    {
+      "epoch": 0.2657305367934589,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.468488888888889e-05,
+      "loss": 0.404,
+      "step": 11960
+    },
+    {
+      "epoch": 0.2659527195165304,
+      "grad_norm": 1.875,
+      "learning_rate": 1.4680444444444447e-05,
+      "loss": 0.3996,
+      "step": 11970
+    },
+    {
+      "epoch": 0.26617490223960183,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4676000000000001e-05,
+      "loss": 0.4015,
+      "step": 11980
+    },
+    {
+      "epoch": 0.2663970849626733,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.4671555555555556e-05,
+      "loss": 0.4102,
+      "step": 11990
+    },
+    {
+      "epoch": 0.26661926768574473,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4667111111111111e-05,
+      "loss": 0.398,
+      "step": 12000
+    },
+    {
+      "epoch": 0.2668414504088162,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.4662666666666668e-05,
+      "loss": 0.4066,
+      "step": 12010
+    },
+    {
+      "epoch": 0.2670636331318877,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.4658222222222223e-05,
+      "loss": 0.4213,
+      "step": 12020
+    },
+    {
+      "epoch": 0.26728581585495914,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4653777777777777e-05,
+      "loss": 0.398,
+      "step": 12030
+    },
+    {
+      "epoch": 0.2675079985780306,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4649333333333336e-05,
+      "loss": 0.3696,
+      "step": 12040
+    },
+    {
+      "epoch": 0.26773018130110204,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.464488888888889e-05,
+      "loss": 0.3524,
+      "step": 12050
+    },
+    {
+      "epoch": 0.2679523640241735,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.4640444444444445e-05,
+      "loss": 0.4166,
+      "step": 12060
+    },
+    {
+      "epoch": 0.26817454674724495,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4636000000000002e-05,
+      "loss": 0.4267,
+      "step": 12070
+    },
+    {
+      "epoch": 0.2683967294703164,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.4631555555555557e-05,
+      "loss": 0.4347,
+      "step": 12080
+    },
+    {
+      "epoch": 0.26861891219338785,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4627111111111112e-05,
+      "loss": 0.436,
+      "step": 12090
+    },
+    {
+      "epoch": 0.2688410949164593,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4622666666666668e-05,
+      "loss": 0.3948,
+      "step": 12100
+    },
+    {
+      "epoch": 0.26906327763953075,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4618222222222223e-05,
+      "loss": 0.4133,
+      "step": 12110
+    },
+    {
+      "epoch": 0.2692854603626022,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.461377777777778e-05,
+      "loss": 0.3911,
+      "step": 12120
+    },
+    {
+      "epoch": 0.26950764308567365,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4609333333333336e-05,
+      "loss": 0.4272,
+      "step": 12130
+    },
+    {
+      "epoch": 0.2697298258087451,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.460488888888889e-05,
+      "loss": 0.4426,
+      "step": 12140
+    },
+    {
+      "epoch": 0.26995200853181656,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.4600444444444446e-05,
+      "loss": 0.4244,
+      "step": 12150
+    },
+    {
+      "epoch": 0.270174191254888,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.4596e-05,
+      "loss": 0.4009,
+      "step": 12160
+    },
+    {
+      "epoch": 0.27039637397795946,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.4591555555555557e-05,
+      "loss": 0.386,
+      "step": 12170
+    },
+    {
+      "epoch": 0.2706185567010309,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.4587111111111112e-05,
+      "loss": 0.4325,
+      "step": 12180
+    },
+    {
+      "epoch": 0.27084073942410236,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4582666666666667e-05,
+      "loss": 0.3962,
+      "step": 12190
+    },
+    {
+      "epoch": 0.2710629221471738,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4578222222222223e-05,
+      "loss": 0.4439,
+      "step": 12200
+    },
+    {
+      "epoch": 0.27128510487024526,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.457377777777778e-05,
+      "loss": 0.4283,
+      "step": 12210
+    },
+    {
+      "epoch": 0.27150728759331677,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4569333333333335e-05,
+      "loss": 0.4495,
+      "step": 12220
+    },
+    {
+      "epoch": 0.2717294703163882,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.4564888888888891e-05,
+      "loss": 0.3652,
+      "step": 12230
+    },
+    {
+      "epoch": 0.2719516530394597,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.4560444444444446e-05,
+      "loss": 0.3978,
+      "step": 12240
+    },
+    {
+      "epoch": 0.2721738357625311,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4556000000000001e-05,
+      "loss": 0.4179,
+      "step": 12250
+    },
+    {
+      "epoch": 0.2723960184856026,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4551555555555556e-05,
+      "loss": 0.4119,
+      "step": 12260
+    },
+    {
+      "epoch": 0.272618201208674,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.4547111111111112e-05,
+      "loss": 0.414,
+      "step": 12270
+    },
+    {
+      "epoch": 0.2728403839317455,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4542666666666667e-05,
+      "loss": 0.4351,
+      "step": 12280
+    },
+    {
+      "epoch": 0.27306256665481693,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.4538222222222222e-05,
+      "loss": 0.3993,
+      "step": 12290
+    },
+    {
+      "epoch": 0.2732847493778884,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.453377777777778e-05,
+      "loss": 0.4078,
+      "step": 12300
+    },
+    {
+      "epoch": 0.27350693210095983,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4529333333333335e-05,
+      "loss": 0.4145,
+      "step": 12310
+    },
+    {
+      "epoch": 0.2737291148240313,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.452488888888889e-05,
+      "loss": 0.4601,
+      "step": 12320
+    },
+    {
+      "epoch": 0.27395129754710273,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.4520444444444446e-05,
+      "loss": 0.4044,
+      "step": 12330
+    },
+    {
+      "epoch": 0.2741734802701742,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.4516000000000001e-05,
+      "loss": 0.4171,
+      "step": 12340
+    },
+    {
+      "epoch": 0.27439566299324564,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4511555555555556e-05,
+      "loss": 0.4093,
+      "step": 12350
+    },
+    {
+      "epoch": 0.2746178457163171,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4507111111111111e-05,
+      "loss": 0.3676,
+      "step": 12360
+    },
+    {
+      "epoch": 0.27484002843938854,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.4502666666666668e-05,
+      "loss": 0.4007,
+      "step": 12370
+    },
+    {
+      "epoch": 0.27506221116246,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.4498222222222222e-05,
+      "loss": 0.3973,
+      "step": 12380
+    },
+    {
+      "epoch": 0.27528439388553144,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.449377777777778e-05,
+      "loss": 0.3983,
+      "step": 12390
+    },
+    {
+      "epoch": 0.2755065766086029,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4489333333333335e-05,
+      "loss": 0.4332,
+      "step": 12400
+    },
+    {
+      "epoch": 0.27572875933167434,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.448488888888889e-05,
+      "loss": 0.4434,
+      "step": 12410
+    },
+    {
+      "epoch": 0.2759509420547458,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.4480444444444445e-05,
+      "loss": 0.4142,
+      "step": 12420
+    },
+    {
+      "epoch": 0.2761731247778173,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.4476000000000002e-05,
+      "loss": 0.4278,
+      "step": 12430
+    },
+    {
+      "epoch": 0.27639530750088875,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4471555555555557e-05,
+      "loss": 0.4008,
+      "step": 12440
+    },
+    {
+      "epoch": 0.2766174902239602,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4467111111111111e-05,
+      "loss": 0.3918,
+      "step": 12450
+    },
+    {
+      "epoch": 0.27683967294703166,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4462666666666668e-05,
+      "loss": 0.3931,
+      "step": 12460
+    },
+    {
+      "epoch": 0.2770618556701031,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.4458222222222223e-05,
+      "loss": 0.4337,
+      "step": 12470
+    },
+    {
+      "epoch": 0.27728403839317456,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.445377777777778e-05,
+      "loss": 0.4374,
+      "step": 12480
+    },
+    {
+      "epoch": 0.277506221116246,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4449333333333336e-05,
+      "loss": 0.4279,
+      "step": 12490
+    },
+    {
+      "epoch": 0.27772840383931746,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.444488888888889e-05,
+      "loss": 0.4163,
+      "step": 12500
+    },
+    {
+      "epoch": 0.2779505865623889,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4440444444444446e-05,
+      "loss": 0.4316,
+      "step": 12510
+    },
+    {
+      "epoch": 0.27817276928546036,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.4436e-05,
+      "loss": 0.4183,
+      "step": 12520
+    },
+    {
+      "epoch": 0.2783949520085318,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.4431555555555557e-05,
+      "loss": 0.4326,
+      "step": 12530
+    },
+    {
+      "epoch": 0.27861713473160327,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.4427111111111112e-05,
+      "loss": 0.4192,
+      "step": 12540
+    },
+    {
+      "epoch": 0.2788393174546747,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4422666666666667e-05,
+      "loss": 0.4446,
+      "step": 12550
+    },
+    {
+      "epoch": 0.27906150017774617,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4418222222222223e-05,
+      "loss": 0.3878,
+      "step": 12560
+    },
+    {
+      "epoch": 0.2792836829008176,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.441377777777778e-05,
+      "loss": 0.4499,
+      "step": 12570
+    },
+    {
+      "epoch": 0.27950586562388907,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.4409333333333335e-05,
+      "loss": 0.4303,
+      "step": 12580
+    },
+    {
+      "epoch": 0.2797280483469605,
+      "grad_norm": 2.25,
+      "learning_rate": 1.4404888888888891e-05,
+      "loss": 0.3757,
+      "step": 12590
+    },
+    {
+      "epoch": 0.279950231070032,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4400444444444446e-05,
+      "loss": 0.3978,
+      "step": 12600
+    },
+    {
+      "epoch": 0.2801724137931034,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4396e-05,
+      "loss": 0.4275,
+      "step": 12610
+    },
+    {
+      "epoch": 0.2803945965161749,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4391555555555556e-05,
+      "loss": 0.407,
+      "step": 12620
+    },
+    {
+      "epoch": 0.2806167792392464,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4387111111111112e-05,
+      "loss": 0.4344,
+      "step": 12630
+    },
+    {
+      "epoch": 0.28083896196231783,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.4382666666666667e-05,
+      "loss": 0.3926,
+      "step": 12640
+    },
+    {
+      "epoch": 0.2810611446853893,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4378222222222222e-05,
+      "loss": 0.4002,
+      "step": 12650
+    },
+    {
+      "epoch": 0.28128332740846074,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.437377777777778e-05,
+      "loss": 0.3673,
+      "step": 12660
+    },
+    {
+      "epoch": 0.2815055101315322,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.4369333333333335e-05,
+      "loss": 0.3803,
+      "step": 12670
+    },
+    {
+      "epoch": 0.28172769285460364,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.436488888888889e-05,
+      "loss": 0.4249,
+      "step": 12680
+    },
+    {
+      "epoch": 0.2819498755776751,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4360444444444446e-05,
+      "loss": 0.4255,
+      "step": 12690
+    },
+    {
+      "epoch": 0.28217205830074654,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.4356000000000001e-05,
+      "loss": 0.4691,
+      "step": 12700
+    },
+    {
+      "epoch": 0.282394241023818,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.4351555555555556e-05,
+      "loss": 0.4481,
+      "step": 12710
+    },
+    {
+      "epoch": 0.28261642374688944,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.434711111111111e-05,
+      "loss": 0.3862,
+      "step": 12720
+    },
+    {
+      "epoch": 0.2828386064699609,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.4342666666666667e-05,
+      "loss": 0.404,
+      "step": 12730
+    },
+    {
+      "epoch": 0.28306078919303235,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4338222222222224e-05,
+      "loss": 0.3762,
+      "step": 12740
+    },
+    {
+      "epoch": 0.2832829719161038,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.433377777777778e-05,
+      "loss": 0.4001,
+      "step": 12750
+    },
+    {
+      "epoch": 0.28350515463917525,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4329333333333335e-05,
+      "loss": 0.4061,
+      "step": 12760
+    },
+    {
+      "epoch": 0.2837273373622467,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.432488888888889e-05,
+      "loss": 0.416,
+      "step": 12770
+    },
+    {
+      "epoch": 0.28394952008531815,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4320444444444445e-05,
+      "loss": 0.3675,
+      "step": 12780
+    },
+    {
+      "epoch": 0.2841717028083896,
+      "grad_norm": 2.625,
+      "learning_rate": 1.4316000000000002e-05,
+      "loss": 0.3726,
+      "step": 12790
+    },
+    {
+      "epoch": 0.28439388553146105,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4311555555555556e-05,
+      "loss": 0.4244,
+      "step": 12800
+    },
+    {
+      "epoch": 0.2846160682545325,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4307111111111111e-05,
+      "loss": 0.4231,
+      "step": 12810
+    },
+    {
+      "epoch": 0.28483825097760396,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4302666666666668e-05,
+      "loss": 0.4255,
+      "step": 12820
+    },
+    {
+      "epoch": 0.2850604337006754,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.4298222222222224e-05,
+      "loss": 0.4275,
+      "step": 12830
+    },
+    {
+      "epoch": 0.2852826164237469,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4293777777777779e-05,
+      "loss": 0.4344,
+      "step": 12840
+    },
+    {
+      "epoch": 0.28550479914681837,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.4289333333333336e-05,
+      "loss": 0.4564,
+      "step": 12850
+    },
+    {
+      "epoch": 0.2857269818698898,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.428488888888889e-05,
+      "loss": 0.3974,
+      "step": 12860
+    },
+    {
+      "epoch": 0.28594916459296127,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4280444444444445e-05,
+      "loss": 0.3708,
+      "step": 12870
+    },
+    {
+      "epoch": 0.2861713473160327,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4276e-05,
+      "loss": 0.4,
+      "step": 12880
+    },
+    {
+      "epoch": 0.28639353003910417,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4271555555555557e-05,
+      "loss": 0.3767,
+      "step": 12890
+    },
+    {
+      "epoch": 0.2866157127621756,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4267111111111112e-05,
+      "loss": 0.4141,
+      "step": 12900
+    },
+    {
+      "epoch": 0.2868378954852471,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4262666666666666e-05,
+      "loss": 0.4175,
+      "step": 12910
+    },
+    {
+      "epoch": 0.2870600782083185,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.4258222222222225e-05,
+      "loss": 0.4612,
+      "step": 12920
+    },
+    {
+      "epoch": 0.28728226093139,
+      "grad_norm": 2.5,
+      "learning_rate": 1.425377777777778e-05,
+      "loss": 0.4161,
+      "step": 12930
+    },
+    {
+      "epoch": 0.28750444365446143,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4249333333333334e-05,
+      "loss": 0.4,
+      "step": 12940
+    },
+    {
+      "epoch": 0.2877266263775329,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4244888888888891e-05,
+      "loss": 0.4594,
+      "step": 12950
+    },
+    {
+      "epoch": 0.28794880910060433,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.4240444444444446e-05,
+      "loss": 0.4308,
+      "step": 12960
+    },
+    {
+      "epoch": 0.2881709918236758,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4236e-05,
+      "loss": 0.4383,
+      "step": 12970
+    },
+    {
+      "epoch": 0.28839317454674723,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.4231555555555555e-05,
+      "loss": 0.4272,
+      "step": 12980
+    },
+    {
+      "epoch": 0.2886153572698187,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.4227111111111112e-05,
+      "loss": 0.4004,
+      "step": 12990
+    },
+    {
+      "epoch": 0.28883753999289014,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4222666666666667e-05,
+      "loss": 0.4038,
+      "step": 13000
+    },
+    {
+      "epoch": 0.2890597227159616,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4218222222222225e-05,
+      "loss": 0.3879,
+      "step": 13010
+    },
+    {
+      "epoch": 0.28928190543903304,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.421377777777778e-05,
+      "loss": 0.3951,
+      "step": 13020
+    },
+    {
+      "epoch": 0.2895040881621045,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4209333333333335e-05,
+      "loss": 0.3893,
+      "step": 13030
+    },
+    {
+      "epoch": 0.289726270885176,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.420488888888889e-05,
+      "loss": 0.4287,
+      "step": 13040
+    },
+    {
+      "epoch": 0.28994845360824745,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.4200444444444446e-05,
+      "loss": 0.3676,
+      "step": 13050
+    },
+    {
+      "epoch": 0.2901706363313189,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.4196000000000001e-05,
+      "loss": 0.4039,
+      "step": 13060
+    },
+    {
+      "epoch": 0.29039281905439035,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.4191555555555556e-05,
+      "loss": 0.4102,
+      "step": 13070
+    },
+    {
+      "epoch": 0.2906150017774618,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.4187111111111112e-05,
+      "loss": 0.4129,
+      "step": 13080
+    },
+    {
+      "epoch": 0.29083718450053325,
+      "grad_norm": 2.0,
+      "learning_rate": 1.4182666666666667e-05,
+      "loss": 0.424,
+      "step": 13090
+    },
+    {
+      "epoch": 0.2910593672236047,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.4178222222222224e-05,
+      "loss": 0.3931,
+      "step": 13100
+    },
+    {
+      "epoch": 0.29128154994667615,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.417377777777778e-05,
+      "loss": 0.4318,
+      "step": 13110
+    },
+    {
+      "epoch": 0.2915037326697476,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.4169333333333335e-05,
+      "loss": 0.3904,
+      "step": 13120
+    },
+    {
+      "epoch": 0.29172591539281906,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.416488888888889e-05,
+      "loss": 0.3825,
+      "step": 13130
+    },
+    {
+      "epoch": 0.2919480981158905,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4160444444444445e-05,
+      "loss": 0.3881,
+      "step": 13140
+    },
+    {
+      "epoch": 0.29217028083896196,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4156000000000001e-05,
+      "loss": 0.3714,
+      "step": 13150
+    },
+    {
+      "epoch": 0.2923924635620334,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4151555555555556e-05,
+      "loss": 0.413,
+      "step": 13160
+    },
+    {
+      "epoch": 0.29261464628510486,
+      "grad_norm": 2.25,
+      "learning_rate": 1.4147111111111111e-05,
+      "loss": 0.4085,
+      "step": 13170
+    },
+    {
+      "epoch": 0.2928368290081763,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4142666666666668e-05,
+      "loss": 0.4194,
+      "step": 13180
+    },
+    {
+      "epoch": 0.29305901173124776,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4138222222222224e-05,
+      "loss": 0.3961,
+      "step": 13190
+    },
+    {
+      "epoch": 0.2932811944543192,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.4133777777777779e-05,
+      "loss": 0.413,
+      "step": 13200
+    },
+    {
+      "epoch": 0.29350337717739067,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.4129333333333335e-05,
+      "loss": 0.414,
+      "step": 13210
+    },
+    {
+      "epoch": 0.2937255599004621,
+      "grad_norm": 2.0,
+      "learning_rate": 1.412488888888889e-05,
+      "loss": 0.4066,
+      "step": 13220
+    },
+    {
+      "epoch": 0.29394774262353357,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.4120444444444445e-05,
+      "loss": 0.4498,
+      "step": 13230
+    },
+    {
+      "epoch": 0.294169925346605,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.4116e-05,
+      "loss": 0.3566,
+      "step": 13240
+    },
+    {
+      "epoch": 0.29439210806967653,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.4111555555555557e-05,
+      "loss": 0.3573,
+      "step": 13250
+    },
+    {
+      "epoch": 0.294614290792748,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4107111111111111e-05,
+      "loss": 0.3703,
+      "step": 13260
+    },
+    {
+      "epoch": 0.29483647351581943,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.4102666666666666e-05,
+      "loss": 0.3922,
+      "step": 13270
+    },
+    {
+      "epoch": 0.2950586562388909,
+      "grad_norm": 2.375,
+      "learning_rate": 1.4098222222222224e-05,
+      "loss": 0.3918,
+      "step": 13280
+    },
+    {
+      "epoch": 0.29528083896196233,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.409377777777778e-05,
+      "loss": 0.4299,
+      "step": 13290
+    },
+    {
+      "epoch": 0.2955030216850338,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4089333333333334e-05,
+      "loss": 0.4128,
+      "step": 13300
+    },
+    {
+      "epoch": 0.29572520440810524,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.408488888888889e-05,
+      "loss": 0.4024,
+      "step": 13310
+    },
+    {
+      "epoch": 0.2959473871311767,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4080444444444446e-05,
+      "loss": 0.3516,
+      "step": 13320
+    },
+    {
+      "epoch": 0.29616956985424814,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.4076e-05,
+      "loss": 0.4223,
+      "step": 13330
+    },
+    {
+      "epoch": 0.2963917525773196,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4071555555555555e-05,
+      "loss": 0.4104,
+      "step": 13340
+    },
+    {
+      "epoch": 0.29661393530039104,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4067111111111112e-05,
+      "loss": 0.4217,
+      "step": 13350
+    },
+    {
+      "epoch": 0.2968361180234625,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.4062666666666668e-05,
+      "loss": 0.4201,
+      "step": 13360
+    },
+    {
+      "epoch": 0.29705830074653394,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.4058222222222225e-05,
+      "loss": 0.3873,
+      "step": 13370
+    },
+    {
+      "epoch": 0.2972804834696054,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.405377777777778e-05,
+      "loss": 0.4223,
+      "step": 13380
+    },
+    {
+      "epoch": 0.29750266619267685,
+      "grad_norm": 2.0,
+      "learning_rate": 1.4049333333333335e-05,
+      "loss": 0.3799,
+      "step": 13390
+    },
+    {
+      "epoch": 0.2977248489157483,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.404488888888889e-05,
+      "loss": 0.4534,
+      "step": 13400
+    },
+    {
+      "epoch": 0.29794703163881975,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.4040444444444446e-05,
+      "loss": 0.3999,
+      "step": 13410
+    },
+    {
+      "epoch": 0.2981692143618912,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4036e-05,
+      "loss": 0.4292,
+      "step": 13420
+    },
+    {
+      "epoch": 0.29839139708496265,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.4031555555555556e-05,
+      "loss": 0.3904,
+      "step": 13430
+    },
+    {
+      "epoch": 0.2986135798080341,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.4027111111111112e-05,
+      "loss": 0.4645,
+      "step": 13440
+    },
+    {
+      "epoch": 0.2988357625311056,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.4022666666666669e-05,
+      "loss": 0.4084,
+      "step": 13450
+    },
+    {
+      "epoch": 0.29905794525417706,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4018222222222224e-05,
+      "loss": 0.437,
+      "step": 13460
+    },
+    {
+      "epoch": 0.2992801279772485,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.401377777777778e-05,
+      "loss": 0.3942,
+      "step": 13470
+    },
+    {
+      "epoch": 0.29950231070031996,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.4009333333333335e-05,
+      "loss": 0.397,
+      "step": 13480
+    },
+    {
+      "epoch": 0.2997244934233914,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.400488888888889e-05,
+      "loss": 0.4051,
+      "step": 13490
+    },
+    {
+      "epoch": 0.29994667614646287,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.4000444444444445e-05,
+      "loss": 0.369,
+      "step": 13500
+    },
+    {
+      "epoch": 0.3001688588695343,
+      "grad_norm": 2.375,
+      "learning_rate": 1.3996000000000001e-05,
+      "loss": 0.3987,
+      "step": 13510
+    },
+    {
+      "epoch": 0.30039104159260577,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3991555555555556e-05,
+      "loss": 0.4054,
+      "step": 13520
+    },
+    {
+      "epoch": 0.3006132243156772,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.3987111111111111e-05,
+      "loss": 0.3922,
+      "step": 13530
+    },
+    {
+      "epoch": 0.30083540703874867,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.3982666666666669e-05,
+      "loss": 0.3975,
+      "step": 13540
+    },
+    {
+      "epoch": 0.3010575897618201,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.3978222222222224e-05,
+      "loss": 0.3955,
+      "step": 13550
+    },
+    {
+      "epoch": 0.3012797724848916,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.3973777777777779e-05,
+      "loss": 0.4114,
+      "step": 13560
+    },
+    {
+      "epoch": 0.301501955207963,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.3969333333333335e-05,
+      "loss": 0.4023,
+      "step": 13570
+    },
+    {
+      "epoch": 0.3017241379310345,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.396488888888889e-05,
+      "loss": 0.3926,
+      "step": 13580
+    },
+    {
+      "epoch": 0.3019463206541059,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3960444444444445e-05,
+      "loss": 0.4103,
+      "step": 13590
+    },
+    {
+      "epoch": 0.3021685033771774,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.3956e-05,
+      "loss": 0.4331,
+      "step": 13600
+    },
+    {
+      "epoch": 0.30239068610024883,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.3951555555555556e-05,
+      "loss": 0.3934,
+      "step": 13610
+    },
+    {
+      "epoch": 0.3026128688233203,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.3947111111111111e-05,
+      "loss": 0.4256,
+      "step": 13620
+    },
+    {
+      "epoch": 0.30283505154639173,
+      "grad_norm": 2.0,
+      "learning_rate": 1.394266666666667e-05,
+      "loss": 0.4013,
+      "step": 13630
+    },
+    {
+      "epoch": 0.3030572342694632,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3938222222222224e-05,
+      "loss": 0.4024,
+      "step": 13640
+    },
+    {
+      "epoch": 0.30327941699253463,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.393377777777778e-05,
+      "loss": 0.3839,
+      "step": 13650
+    },
+    {
+      "epoch": 0.30350159971560614,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3929333333333334e-05,
+      "loss": 0.3905,
+      "step": 13660
+    },
+    {
+      "epoch": 0.3037237824386776,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.392488888888889e-05,
+      "loss": 0.3926,
+      "step": 13670
+    },
+    {
+      "epoch": 0.30394596516174904,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.3920444444444445e-05,
+      "loss": 0.3708,
+      "step": 13680
+    },
+    {
+      "epoch": 0.3041681478848205,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.3916e-05,
+      "loss": 0.3952,
+      "step": 13690
+    },
+    {
+      "epoch": 0.30439033060789195,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3911555555555555e-05,
+      "loss": 0.428,
+      "step": 13700
+    },
+    {
+      "epoch": 0.3046125133309634,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.3907111111111112e-05,
+      "loss": 0.3858,
+      "step": 13710
+    },
+    {
+      "epoch": 0.30483469605403485,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3902666666666668e-05,
+      "loss": 0.4229,
+      "step": 13720
+    },
+    {
+      "epoch": 0.3050568787771063,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.3898222222222225e-05,
+      "loss": 0.4229,
+      "step": 13730
+    },
+    {
+      "epoch": 0.30527906150017775,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.389377777777778e-05,
+      "loss": 0.3765,
+      "step": 13740
+    },
+    {
+      "epoch": 0.3055012442232492,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.3889333333333334e-05,
+      "loss": 0.4365,
+      "step": 13750
+    },
+    {
+      "epoch": 0.30572342694632065,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.388488888888889e-05,
+      "loss": 0.4774,
+      "step": 13760
+    },
+    {
+      "epoch": 0.3059456096693921,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3880444444444446e-05,
+      "loss": 0.4109,
+      "step": 13770
+    },
+    {
+      "epoch": 0.30616779239246356,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.3876e-05,
+      "loss": 0.4125,
+      "step": 13780
+    },
+    {
+      "epoch": 0.306389975115535,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.3871555555555555e-05,
+      "loss": 0.3687,
+      "step": 13790
+    },
+    {
+      "epoch": 0.30661215783860646,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.3867111111111112e-05,
+      "loss": 0.4284,
+      "step": 13800
+    },
+    {
+      "epoch": 0.3068343405616779,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3862666666666669e-05,
+      "loss": 0.3744,
+      "step": 13810
+    },
+    {
+      "epoch": 0.30705652328474936,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.3858222222222223e-05,
+      "loss": 0.4065,
+      "step": 13820
+    },
+    {
+      "epoch": 0.3072787060078208,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.385377777777778e-05,
+      "loss": 0.4315,
+      "step": 13830
+    },
+    {
+      "epoch": 0.30750088873089226,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.3849333333333335e-05,
+      "loss": 0.3864,
+      "step": 13840
+    },
+    {
+      "epoch": 0.3077230714539637,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.384488888888889e-05,
+      "loss": 0.3963,
+      "step": 13850
+    },
+    {
+      "epoch": 0.3079452541770352,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3840444444444444e-05,
+      "loss": 0.3705,
+      "step": 13860
+    },
+    {
+      "epoch": 0.3081674369001067,
+      "grad_norm": 2.0,
+      "learning_rate": 1.3836000000000001e-05,
+      "loss": 0.3835,
+      "step": 13870
+    },
+    {
+      "epoch": 0.3083896196231781,
+      "grad_norm": 2.625,
+      "learning_rate": 1.3831555555555556e-05,
+      "loss": 0.441,
+      "step": 13880
+    },
+    {
+      "epoch": 0.3086118023462496,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.382711111111111e-05,
+      "loss": 0.4058,
+      "step": 13890
+    },
+    {
+      "epoch": 0.308833985069321,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.3822666666666669e-05,
+      "loss": 0.4046,
+      "step": 13900
+    },
+    {
+      "epoch": 0.3090561677923925,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.3818222222222224e-05,
+      "loss": 0.4499,
+      "step": 13910
+    },
+    {
+      "epoch": 0.30927835051546393,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3813777777777779e-05,
+      "loss": 0.4132,
+      "step": 13920
+    },
+    {
+      "epoch": 0.3095005332385354,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3809333333333335e-05,
+      "loss": 0.3895,
+      "step": 13930
+    },
+    {
+      "epoch": 0.30972271596160683,
+      "grad_norm": 2.0,
+      "learning_rate": 1.380488888888889e-05,
+      "loss": 0.417,
+      "step": 13940
+    },
+    {
+      "epoch": 0.3099448986846783,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3800444444444445e-05,
+      "loss": 0.3978,
+      "step": 13950
+    },
+    {
+      "epoch": 0.31016708140774973,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.3796e-05,
+      "loss": 0.4105,
+      "step": 13960
+    },
+    {
+      "epoch": 0.3103892641308212,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.3791555555555556e-05,
+      "loss": 0.3786,
+      "step": 13970
+    },
+    {
+      "epoch": 0.31061144685389264,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.3787111111111113e-05,
+      "loss": 0.4309,
+      "step": 13980
+    },
+    {
+      "epoch": 0.3108336295769641,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.378266666666667e-05,
+      "loss": 0.3947,
+      "step": 13990
+    },
+    {
+      "epoch": 0.31105581230003554,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3778222222222224e-05,
+      "loss": 0.3722,
+      "step": 14000
+    },
+    {
+      "epoch": 0.311277995023107,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.3773777777777779e-05,
+      "loss": 0.4129,
+      "step": 14010
+    },
+    {
+      "epoch": 0.31150017774617844,
+      "grad_norm": 1.875,
+      "learning_rate": 1.3769333333333334e-05,
+      "loss": 0.4232,
+      "step": 14020
+    },
+    {
+      "epoch": 0.3117223604692499,
+      "grad_norm": 2.0,
+      "learning_rate": 1.376488888888889e-05,
+      "loss": 0.4003,
+      "step": 14030
+    },
+    {
+      "epoch": 0.31194454319232134,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.3760444444444445e-05,
+      "loss": 0.4271,
+      "step": 14040
+    },
+    {
+      "epoch": 0.3121667259153928,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.3756e-05,
+      "loss": 0.4188,
+      "step": 14050
+    },
+    {
+      "epoch": 0.31238890863846425,
+      "grad_norm": 2.0,
+      "learning_rate": 1.3751555555555555e-05,
+      "loss": 0.3805,
+      "step": 14060
+    },
+    {
+      "epoch": 0.31261109136153575,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3747111111111113e-05,
+      "loss": 0.4182,
+      "step": 14070
+    },
+    {
+      "epoch": 0.3128332740846072,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3742666666666668e-05,
+      "loss": 0.4018,
+      "step": 14080
+    },
+    {
+      "epoch": 0.31305545680767866,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.3738222222222225e-05,
+      "loss": 0.4084,
+      "step": 14090
+    },
+    {
+      "epoch": 0.3132776395307501,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.373377777777778e-05,
+      "loss": 0.3705,
+      "step": 14100
+    },
+    {
+      "epoch": 0.31349982225382156,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3729333333333334e-05,
+      "loss": 0.413,
+      "step": 14110
+    },
+    {
+      "epoch": 0.313722004976893,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.3724888888888889e-05,
+      "loss": 0.4258,
+      "step": 14120
+    },
+    {
+      "epoch": 0.31394418769996446,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.3720444444444446e-05,
+      "loss": 0.4217,
+      "step": 14130
+    },
+    {
+      "epoch": 0.3141663704230359,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.3716e-05,
+      "loss": 0.428,
+      "step": 14140
+    },
+    {
+      "epoch": 0.31438855314610736,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.3711555555555555e-05,
+      "loss": 0.3744,
+      "step": 14150
+    },
+    {
+      "epoch": 0.3146107358691788,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.3707111111111114e-05,
+      "loss": 0.4357,
+      "step": 14160
+    },
+    {
+      "epoch": 0.31483291859225027,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.3702666666666668e-05,
+      "loss": 0.4138,
+      "step": 14170
+    },
+    {
+      "epoch": 0.3150551013153217,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.3698222222222223e-05,
+      "loss": 0.3804,
+      "step": 14180
+    },
+    {
+      "epoch": 0.31527728403839317,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.369377777777778e-05,
+      "loss": 0.4022,
+      "step": 14190
+    },
+    {
+      "epoch": 0.3154994667614646,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3689333333333335e-05,
+      "loss": 0.4232,
+      "step": 14200
+    },
+    {
+      "epoch": 0.31572164948453607,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.368488888888889e-05,
+      "loss": 0.4258,
+      "step": 14210
+    },
+    {
+      "epoch": 0.3159438322076075,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3680444444444444e-05,
+      "loss": 0.4339,
+      "step": 14220
+    },
+    {
+      "epoch": 0.316166014930679,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.3676000000000001e-05,
+      "loss": 0.3538,
+      "step": 14230
+    },
+    {
+      "epoch": 0.3163881976537504,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.3671555555555556e-05,
+      "loss": 0.4367,
+      "step": 14240
+    },
+    {
+      "epoch": 0.3166103803768219,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3667111111111114e-05,
+      "loss": 0.3788,
+      "step": 14250
+    },
+    {
+      "epoch": 0.3168325630998933,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3662666666666669e-05,
+      "loss": 0.4369,
+      "step": 14260
+    },
+    {
+      "epoch": 0.3170547458229648,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3658222222222224e-05,
+      "loss": 0.4107,
+      "step": 14270
+    },
+    {
+      "epoch": 0.3172769285460363,
+      "grad_norm": 2.375,
+      "learning_rate": 1.3653777777777778e-05,
+      "loss": 0.4378,
+      "step": 14280
+    },
+    {
+      "epoch": 0.31749911126910774,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.3649333333333335e-05,
+      "loss": 0.3985,
+      "step": 14290
+    },
+    {
+      "epoch": 0.3177212939921792,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.364488888888889e-05,
+      "loss": 0.3981,
+      "step": 14300
+    },
+    {
+      "epoch": 0.31794347671525064,
+      "grad_norm": 2.0,
+      "learning_rate": 1.3640444444444445e-05,
+      "loss": 0.3764,
+      "step": 14310
+    },
+    {
+      "epoch": 0.3181656594383221,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.3636e-05,
+      "loss": 0.3738,
+      "step": 14320
+    },
+    {
+      "epoch": 0.31838784216139354,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3631555555555556e-05,
+      "loss": 0.4342,
+      "step": 14330
+    },
+    {
+      "epoch": 0.318610024884465,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.3627111111111113e-05,
+      "loss": 0.4101,
+      "step": 14340
+    },
+    {
+      "epoch": 0.31883220760753644,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.362266666666667e-05,
+      "loss": 0.3919,
+      "step": 14350
+    },
+    {
+      "epoch": 0.3190543903306079,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3618222222222224e-05,
+      "loss": 0.3694,
+      "step": 14360
+    },
+    {
+      "epoch": 0.31927657305367935,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3613777777777779e-05,
+      "loss": 0.4038,
+      "step": 14370
+    },
+    {
+      "epoch": 0.3194987557767508,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3609333333333334e-05,
+      "loss": 0.4014,
+      "step": 14380
+    },
+    {
+      "epoch": 0.31972093849982225,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.360488888888889e-05,
+      "loss": 0.377,
+      "step": 14390
+    },
+    {
+      "epoch": 0.3199431212228937,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3600444444444445e-05,
+      "loss": 0.4202,
+      "step": 14400
+    },
+    {
+      "epoch": 0.32016530394596515,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.3596e-05,
+      "loss": 0.3957,
+      "step": 14410
+    },
+    {
+      "epoch": 0.3203874866690366,
+      "grad_norm": 2.0,
+      "learning_rate": 1.3591555555555555e-05,
+      "loss": 0.4077,
+      "step": 14420
+    },
+    {
+      "epoch": 0.32060966939210805,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3587111111111113e-05,
+      "loss": 0.4165,
+      "step": 14430
+    },
+    {
+      "epoch": 0.3208318521151795,
+      "grad_norm": 2.5,
+      "learning_rate": 1.3582666666666668e-05,
+      "loss": 0.4065,
+      "step": 14440
+    },
+    {
+      "epoch": 0.32105403483825096,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.3578222222222224e-05,
+      "loss": 0.3884,
+      "step": 14450
+    },
+    {
+      "epoch": 0.3212762175613224,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.357377777777778e-05,
+      "loss": 0.403,
+      "step": 14460
+    },
+    {
+      "epoch": 0.32149840028439386,
+      "grad_norm": 2.625,
+      "learning_rate": 1.3569333333333334e-05,
+      "loss": 0.4437,
+      "step": 14470
+    },
+    {
+      "epoch": 0.32172058300746537,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.3564888888888889e-05,
+      "loss": 0.3638,
+      "step": 14480
+    },
+    {
+      "epoch": 0.3219427657305368,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.3560444444444445e-05,
+      "loss": 0.4478,
+      "step": 14490
+    },
+    {
+      "epoch": 0.32216494845360827,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.3556e-05,
+      "loss": 0.3968,
+      "step": 14500
+    },
+    {
+      "epoch": 0.3223871311766797,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.3551555555555555e-05,
+      "loss": 0.4,
+      "step": 14510
+    },
+    {
+      "epoch": 0.32260931389975117,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.3547111111111113e-05,
+      "loss": 0.4057,
+      "step": 14520
+    },
+    {
+      "epoch": 0.3228314966228226,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3542666666666668e-05,
+      "loss": 0.3609,
+      "step": 14530
+    },
+    {
+      "epoch": 0.3230536793458941,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.3538222222222223e-05,
+      "loss": 0.4431,
+      "step": 14540
+    },
+    {
+      "epoch": 0.3232758620689655,
+      "grad_norm": 2.25,
+      "learning_rate": 1.353377777777778e-05,
+      "loss": 0.3886,
+      "step": 14550
+    },
+    {
+      "epoch": 0.323498044792037,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.3529333333333334e-05,
+      "loss": 0.3583,
+      "step": 14560
+    },
+    {
+      "epoch": 0.3237202275151084,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.352488888888889e-05,
+      "loss": 0.3723,
+      "step": 14570
+    },
+    {
+      "epoch": 0.3239424102381799,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.3520444444444444e-05,
+      "loss": 0.4102,
+      "step": 14580
+    },
+    {
+      "epoch": 0.32416459296125133,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.3516e-05,
+      "loss": 0.4085,
+      "step": 14590
+    },
+    {
+      "epoch": 0.3243867756843228,
+      "grad_norm": 2.75,
+      "learning_rate": 1.3511555555555557e-05,
+      "loss": 0.3869,
+      "step": 14600
+    },
+    {
+      "epoch": 0.32460895840739423,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3507111111111114e-05,
+      "loss": 0.4148,
+      "step": 14610
+    },
+    {
+      "epoch": 0.3248311411304657,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3502666666666669e-05,
+      "loss": 0.3853,
+      "step": 14620
+    },
+    {
+      "epoch": 0.32505332385353713,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.3498222222222223e-05,
+      "loss": 0.3862,
+      "step": 14630
+    },
+    {
+      "epoch": 0.3252755065766086,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.3493777777777778e-05,
+      "loss": 0.3866,
+      "step": 14640
+    },
+    {
+      "epoch": 0.32549768929968004,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.3489333333333335e-05,
+      "loss": 0.3989,
+      "step": 14650
+    },
+    {
+      "epoch": 0.3257198720227515,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.348488888888889e-05,
+      "loss": 0.3554,
+      "step": 14660
+    },
+    {
+      "epoch": 0.32594205474582294,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.3480444444444445e-05,
+      "loss": 0.3863,
+      "step": 14670
+    },
+    {
+      "epoch": 0.3261642374688944,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3476e-05,
+      "loss": 0.3912,
+      "step": 14680
+    },
+    {
+      "epoch": 0.3263864201919659,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.3471555555555558e-05,
+      "loss": 0.4403,
+      "step": 14690
+    },
+    {
+      "epoch": 0.32660860291503735,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.3467111111111112e-05,
+      "loss": 0.4324,
+      "step": 14700
+    },
+    {
+      "epoch": 0.3268307856381088,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.3462666666666669e-05,
+      "loss": 0.3766,
+      "step": 14710
+    },
+    {
+      "epoch": 0.32705296836118025,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.3458222222222224e-05,
+      "loss": 0.4056,
+      "step": 14720
+    },
+    {
+      "epoch": 0.3272751510842517,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3453777777777779e-05,
+      "loss": 0.3994,
+      "step": 14730
+    },
+    {
+      "epoch": 0.32749733380732315,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3449333333333334e-05,
+      "loss": 0.4055,
+      "step": 14740
+    },
+    {
+      "epoch": 0.3277195165303946,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.344488888888889e-05,
+      "loss": 0.4287,
+      "step": 14750
+    },
+    {
+      "epoch": 0.32794169925346606,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.3440444444444445e-05,
+      "loss": 0.3927,
+      "step": 14760
+    },
+    {
+      "epoch": 0.3281638819765375,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.3436e-05,
+      "loss": 0.418,
+      "step": 14770
+    },
+    {
+      "epoch": 0.32838606469960896,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3431555555555558e-05,
+      "loss": 0.379,
+      "step": 14780
+    },
+    {
+      "epoch": 0.3286082474226804,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.3427111111111113e-05,
+      "loss": 0.4045,
+      "step": 14790
+    },
+    {
+      "epoch": 0.32883043014575186,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.3422666666666668e-05,
+      "loss": 0.4052,
+      "step": 14800
+    },
+    {
+      "epoch": 0.3290526128688233,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3418222222222224e-05,
+      "loss": 0.4304,
+      "step": 14810
+    },
+    {
+      "epoch": 0.32927479559189476,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.3413777777777779e-05,
+      "loss": 0.4112,
+      "step": 14820
+    },
+    {
+      "epoch": 0.3294969783149662,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3409333333333334e-05,
+      "loss": 0.3764,
+      "step": 14830
+    },
+    {
+      "epoch": 0.32971916103803767,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3404888888888889e-05,
+      "loss": 0.3632,
+      "step": 14840
+    },
+    {
+      "epoch": 0.3299413437611091,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.3400444444444445e-05,
+      "loss": 0.4074,
+      "step": 14850
+    },
+    {
+      "epoch": 0.33016352648418057,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.3396e-05,
+      "loss": 0.3798,
+      "step": 14860
+    },
+    {
+      "epoch": 0.330385709207252,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.3391555555555558e-05,
+      "loss": 0.4275,
+      "step": 14870
+    },
+    {
+      "epoch": 0.33060789193032347,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.3387111111111113e-05,
+      "loss": 0.3696,
+      "step": 14880
+    },
+    {
+      "epoch": 0.330830074653395,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.3382666666666668e-05,
+      "loss": 0.3988,
+      "step": 14890
+    },
+    {
+      "epoch": 0.33105225737646643,
+      "grad_norm": 2.0,
+      "learning_rate": 1.3378222222222223e-05,
+      "loss": 0.3687,
+      "step": 14900
+    },
+    {
+      "epoch": 0.3312744400995379,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.337377777777778e-05,
+      "loss": 0.3799,
+      "step": 14910
+    },
+    {
+      "epoch": 0.33149662282260933,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3369333333333334e-05,
+      "loss": 0.3866,
+      "step": 14920
+    },
+    {
+      "epoch": 0.3317188055456808,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3364888888888889e-05,
+      "loss": 0.3815,
+      "step": 14930
+    },
+    {
+      "epoch": 0.33194098826875224,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3360444444444444e-05,
+      "loss": 0.3885,
+      "step": 14940
+    },
+    {
+      "epoch": 0.3321631709918237,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.3356e-05,
+      "loss": 0.3973,
+      "step": 14950
+    },
+    {
+      "epoch": 0.33238535371489514,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.3351555555555557e-05,
+      "loss": 0.3729,
+      "step": 14960
+    },
+    {
+      "epoch": 0.3326075364379666,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.3347111111111114e-05,
+      "loss": 0.3958,
+      "step": 14970
+    },
+    {
+      "epoch": 0.33282971916103804,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.3342666666666668e-05,
+      "loss": 0.4206,
+      "step": 14980
+    },
+    {
+      "epoch": 0.3330519018841095,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.3338222222222223e-05,
+      "loss": 0.411,
+      "step": 14990
+    },
+    {
+      "epoch": 0.33327408460718094,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3333777777777778e-05,
+      "loss": 0.3738,
+      "step": 15000
+    },
+    {
+      "epoch": 0.3334962673302524,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3329333333333335e-05,
+      "loss": 0.3963,
+      "step": 15010
+    },
+    {
+      "epoch": 0.33371845005332385,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.332488888888889e-05,
+      "loss": 0.4623,
+      "step": 15020
+    },
+    {
+      "epoch": 0.3339406327763953,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.3320444444444444e-05,
+      "loss": 0.4067,
+      "step": 15030
+    },
+    {
+      "epoch": 0.33416281549946675,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.3316e-05,
+      "loss": 0.3753,
+      "step": 15040
+    },
+    {
+      "epoch": 0.3343849982225382,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3311555555555557e-05,
+      "loss": 0.4527,
+      "step": 15050
+    },
+    {
+      "epoch": 0.33460718094560965,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3307111111111112e-05,
+      "loss": 0.3978,
+      "step": 15060
+    },
+    {
+      "epoch": 0.3348293636686811,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3302666666666669e-05,
+      "loss": 0.3915,
+      "step": 15070
+    },
+    {
+      "epoch": 0.33505154639175255,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3298222222222224e-05,
+      "loss": 0.4272,
+      "step": 15080
+    },
+    {
+      "epoch": 0.335273729114824,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.3293777777777779e-05,
+      "loss": 0.3927,
+      "step": 15090
+    },
+    {
+      "epoch": 0.3354959118378955,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.3289333333333333e-05,
+      "loss": 0.4217,
+      "step": 15100
+    },
+    {
+      "epoch": 0.33571809456096696,
+      "grad_norm": 2.25,
+      "learning_rate": 1.328488888888889e-05,
+      "loss": 0.4208,
+      "step": 15110
+    },
+    {
+      "epoch": 0.3359402772840384,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.3280444444444445e-05,
+      "loss": 0.3905,
+      "step": 15120
+    },
+    {
+      "epoch": 0.33616246000710986,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3276e-05,
+      "loss": 0.4056,
+      "step": 15130
+    },
+    {
+      "epoch": 0.3363846427301813,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3271555555555558e-05,
+      "loss": 0.4163,
+      "step": 15140
+    },
+    {
+      "epoch": 0.33660682545325277,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3267111111111113e-05,
+      "loss": 0.4131,
+      "step": 15150
+    },
+    {
+      "epoch": 0.3368290081763242,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.3262666666666668e-05,
+      "loss": 0.3995,
+      "step": 15160
+    },
+    {
+      "epoch": 0.33705119089939567,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.3258222222222224e-05,
+      "loss": 0.4,
+      "step": 15170
+    },
+    {
+      "epoch": 0.3372733736224671,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3253777777777779e-05,
+      "loss": 0.4479,
+      "step": 15180
+    },
+    {
+      "epoch": 0.33749555634553857,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.3249333333333334e-05,
+      "loss": 0.4247,
+      "step": 15190
+    },
+    {
+      "epoch": 0.33771773906861,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3244888888888889e-05,
+      "loss": 0.4119,
+      "step": 15200
+    },
+    {
+      "epoch": 0.3379399217916815,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3240444444444445e-05,
+      "loss": 0.4206,
+      "step": 15210
+    },
+    {
+      "epoch": 0.3381621045147529,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3236000000000002e-05,
+      "loss": 0.3911,
+      "step": 15220
+    },
+    {
+      "epoch": 0.3383842872378244,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3231555555555558e-05,
+      "loss": 0.3802,
+      "step": 15230
+    },
+    {
+      "epoch": 0.33860646996089583,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.3227111111111113e-05,
+      "loss": 0.3835,
+      "step": 15240
+    },
+    {
+      "epoch": 0.3388286526839673,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.3222666666666668e-05,
+      "loss": 0.4121,
+      "step": 15250
+    },
+    {
+      "epoch": 0.33905083540703873,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.3218222222222223e-05,
+      "loss": 0.4217,
+      "step": 15260
+    },
+    {
+      "epoch": 0.3392730181301102,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.321377777777778e-05,
+      "loss": 0.3714,
+      "step": 15270
+    },
+    {
+      "epoch": 0.33949520085318163,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3209333333333334e-05,
+      "loss": 0.4171,
+      "step": 15280
+    },
+    {
+      "epoch": 0.3397173835762531,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3204888888888889e-05,
+      "loss": 0.4203,
+      "step": 15290
+    },
+    {
+      "epoch": 0.3399395662993246,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3200444444444444e-05,
+      "loss": 0.4552,
+      "step": 15300
+    },
+    {
+      "epoch": 0.34016174902239604,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3196000000000002e-05,
+      "loss": 0.3636,
+      "step": 15310
+    },
+    {
+      "epoch": 0.3403839317454675,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.3191555555555557e-05,
+      "loss": 0.414,
+      "step": 15320
+    },
+    {
+      "epoch": 0.34060611446853895,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.3187111111111113e-05,
+      "loss": 0.3958,
+      "step": 15330
+    },
+    {
+      "epoch": 0.3408282971916104,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.3182666666666668e-05,
+      "loss": 0.3564,
+      "step": 15340
+    },
+    {
+      "epoch": 0.34105047991468185,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.3178222222222223e-05,
+      "loss": 0.3744,
+      "step": 15350
+    },
+    {
+      "epoch": 0.3412726626377533,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.3173777777777778e-05,
+      "loss": 0.4155,
+      "step": 15360
+    },
+    {
+      "epoch": 0.34149484536082475,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3169333333333335e-05,
+      "loss": 0.4235,
+      "step": 15370
+    },
+    {
+      "epoch": 0.3417170280838962,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.316488888888889e-05,
+      "loss": 0.4195,
+      "step": 15380
+    },
+    {
+      "epoch": 0.34193921080696765,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.3160444444444444e-05,
+      "loss": 0.3463,
+      "step": 15390
+    },
+    {
+      "epoch": 0.3421613935300391,
+      "grad_norm": 2.375,
+      "learning_rate": 1.3156000000000002e-05,
+      "loss": 0.3943,
+      "step": 15400
+    },
+    {
+      "epoch": 0.34238357625311056,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3151555555555557e-05,
+      "loss": 0.405,
+      "step": 15410
+    },
+    {
+      "epoch": 0.342605758976182,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.3147111111111112e-05,
+      "loss": 0.3886,
+      "step": 15420
+    },
+    {
+      "epoch": 0.34282794169925346,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3142666666666669e-05,
+      "loss": 0.4052,
+      "step": 15430
+    },
+    {
+      "epoch": 0.3430501244223249,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.3138222222222224e-05,
+      "loss": 0.4238,
+      "step": 15440
+    },
+    {
+      "epoch": 0.34327230714539636,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.3133777777777778e-05,
+      "loss": 0.3883,
+      "step": 15450
+    },
+    {
+      "epoch": 0.3434944898684678,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.3129333333333333e-05,
+      "loss": 0.3901,
+      "step": 15460
+    },
+    {
+      "epoch": 0.34371667259153926,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.312488888888889e-05,
+      "loss": 0.3939,
+      "step": 15470
+    },
+    {
+      "epoch": 0.3439388553146107,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3120444444444445e-05,
+      "loss": 0.3887,
+      "step": 15480
+    },
+    {
+      "epoch": 0.34416103803768217,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.3116000000000003e-05,
+      "loss": 0.3727,
+      "step": 15490
+    },
+    {
+      "epoch": 0.3443832207607536,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3111555555555558e-05,
+      "loss": 0.3968,
+      "step": 15500
+    },
+    {
+      "epoch": 0.3446054034838251,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.3107111111111113e-05,
+      "loss": 0.4136,
+      "step": 15510
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3102666666666667e-05,
+      "loss": 0.3942,
+      "step": 15520
+    },
+    {
+      "epoch": 0.345049768929968,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.3098222222222224e-05,
+      "loss": 0.4188,
+      "step": 15530
+    },
+    {
+      "epoch": 0.3452719516530395,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3093777777777779e-05,
+      "loss": 0.3917,
+      "step": 15540
+    },
+    {
+      "epoch": 0.34549413437611093,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3089333333333334e-05,
+      "loss": 0.4118,
+      "step": 15550
+    },
+    {
+      "epoch": 0.3457163170991824,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3084888888888888e-05,
+      "loss": 0.4058,
+      "step": 15560
+    },
+    {
+      "epoch": 0.34593849982225383,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3080444444444445e-05,
+      "loss": 0.3819,
+      "step": 15570
+    },
+    {
+      "epoch": 0.3461606825453253,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.3076000000000002e-05,
+      "loss": 0.4499,
+      "step": 15580
+    },
+    {
+      "epoch": 0.34638286526839673,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.3071555555555558e-05,
+      "loss": 0.3809,
+      "step": 15590
+    },
+    {
+      "epoch": 0.3466050479914682,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.3067111111111113e-05,
+      "loss": 0.3994,
+      "step": 15600
+    },
+    {
+      "epoch": 0.34682723071453964,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.3062666666666668e-05,
+      "loss": 0.4276,
+      "step": 15610
+    },
+    {
+      "epoch": 0.3470494134376111,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3058222222222223e-05,
+      "loss": 0.358,
+      "step": 15620
+    },
+    {
+      "epoch": 0.34727159616068254,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.3053777777777779e-05,
+      "loss": 0.4068,
+      "step": 15630
+    },
+    {
+      "epoch": 0.347493778883754,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3049333333333334e-05,
+      "loss": 0.4042,
+      "step": 15640
+    },
+    {
+      "epoch": 0.34771596160682544,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3044888888888889e-05,
+      "loss": 0.3948,
+      "step": 15650
+    },
+    {
+      "epoch": 0.3479381443298969,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.3040444444444444e-05,
+      "loss": 0.4206,
+      "step": 15660
+    },
+    {
+      "epoch": 0.34816032705296834,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.3036000000000002e-05,
+      "loss": 0.3708,
+      "step": 15670
+    },
+    {
+      "epoch": 0.3483825097760398,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.3031555555555557e-05,
+      "loss": 0.3844,
+      "step": 15680
+    },
+    {
+      "epoch": 0.34860469249911125,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.3027111111111113e-05,
+      "loss": 0.4132,
+      "step": 15690
+    },
+    {
+      "epoch": 0.3488268752221827,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.3022666666666668e-05,
+      "loss": 0.3929,
+      "step": 15700
+    },
+    {
+      "epoch": 0.3490490579452542,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.3018222222222223e-05,
+      "loss": 0.4396,
+      "step": 15710
+    },
+    {
+      "epoch": 0.34927124066832566,
+      "grad_norm": 2.125,
+      "learning_rate": 1.3013777777777778e-05,
+      "loss": 0.4274,
+      "step": 15720
+    },
+    {
+      "epoch": 0.3494934233913971,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3009333333333334e-05,
+      "loss": 0.389,
+      "step": 15730
+    },
+    {
+      "epoch": 0.34971560611446856,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.300488888888889e-05,
+      "loss": 0.4186,
+      "step": 15740
+    },
+    {
+      "epoch": 0.34993778883754,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.3000444444444444e-05,
+      "loss": 0.3345,
+      "step": 15750
+    },
+    {
+      "epoch": 0.35015997156061146,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2996000000000002e-05,
+      "loss": 0.4343,
+      "step": 15760
+    },
+    {
+      "epoch": 0.3503821542836829,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2991555555555557e-05,
+      "loss": 0.392,
+      "step": 15770
+    },
+    {
+      "epoch": 0.35060433700675436,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.2987111111111112e-05,
+      "loss": 0.3902,
+      "step": 15780
+    },
+    {
+      "epoch": 0.3508265197298258,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2982666666666669e-05,
+      "loss": 0.4401,
+      "step": 15790
+    },
+    {
+      "epoch": 0.35104870245289727,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.2978222222222223e-05,
+      "loss": 0.3895,
+      "step": 15800
+    },
+    {
+      "epoch": 0.3512708851759687,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2973777777777778e-05,
+      "loss": 0.3998,
+      "step": 15810
+    },
+    {
+      "epoch": 0.35149306789904017,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.2969333333333333e-05,
+      "loss": 0.3912,
+      "step": 15820
+    },
+    {
+      "epoch": 0.3517152506221116,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.296488888888889e-05,
+      "loss": 0.3873,
+      "step": 15830
+    },
+    {
+      "epoch": 0.35193743334518307,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2960444444444446e-05,
+      "loss": 0.3816,
+      "step": 15840
+    },
+    {
+      "epoch": 0.3521596160682545,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2956000000000003e-05,
+      "loss": 0.3876,
+      "step": 15850
+    },
+    {
+      "epoch": 0.352381798791326,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.2951555555555558e-05,
+      "loss": 0.4435,
+      "step": 15860
+    },
+    {
+      "epoch": 0.3526039815143974,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2947111111111112e-05,
+      "loss": 0.4185,
+      "step": 15870
+    },
+    {
+      "epoch": 0.3528261642374689,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.2942666666666667e-05,
+      "loss": 0.4111,
+      "step": 15880
+    },
+    {
+      "epoch": 0.3530483469605403,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2938222222222224e-05,
+      "loss": 0.3857,
+      "step": 15890
+    },
+    {
+      "epoch": 0.3532705296836118,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.2933777777777779e-05,
+      "loss": 0.4239,
+      "step": 15900
+    },
+    {
+      "epoch": 0.35349271240668323,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.2929333333333333e-05,
+      "loss": 0.4113,
+      "step": 15910
+    },
+    {
+      "epoch": 0.35371489512975474,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2924888888888888e-05,
+      "loss": 0.3829,
+      "step": 15920
+    },
+    {
+      "epoch": 0.3539370778528262,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2920444444444447e-05,
+      "loss": 0.4038,
+      "step": 15930
+    },
+    {
+      "epoch": 0.35415926057589764,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.2916000000000001e-05,
+      "loss": 0.4277,
+      "step": 15940
+    },
+    {
+      "epoch": 0.3543814432989691,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.2911555555555558e-05,
+      "loss": 0.3843,
+      "step": 15950
+    },
+    {
+      "epoch": 0.35460362602204054,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.2907111111111113e-05,
+      "loss": 0.4202,
+      "step": 15960
+    },
+    {
+      "epoch": 0.354825808745112,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2902666666666668e-05,
+      "loss": 0.3613,
+      "step": 15970
+    },
+    {
+      "epoch": 0.35504799146818344,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2898222222222222e-05,
+      "loss": 0.3687,
+      "step": 15980
+    },
+    {
+      "epoch": 0.3552701741912549,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.2893777777777779e-05,
+      "loss": 0.3963,
+      "step": 15990
+    },
+    {
+      "epoch": 0.35549235691432635,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.2889333333333334e-05,
+      "loss": 0.3936,
+      "step": 16000
+    },
+    {
+      "epoch": 0.3557145396373978,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2884888888888889e-05,
+      "loss": 0.3891,
+      "step": 16010
+    },
+    {
+      "epoch": 0.35593672236046925,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2880444444444447e-05,
+      "loss": 0.406,
+      "step": 16020
+    },
+    {
+      "epoch": 0.3561589050835407,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2876000000000002e-05,
+      "loss": 0.401,
+      "step": 16030
+    },
+    {
+      "epoch": 0.35638108780661215,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.2871555555555557e-05,
+      "loss": 0.4107,
+      "step": 16040
+    },
+    {
+      "epoch": 0.3566032705296836,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.2867111111111113e-05,
+      "loss": 0.3754,
+      "step": 16050
+    },
+    {
+      "epoch": 0.35682545325275505,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.2862666666666668e-05,
+      "loss": 0.3904,
+      "step": 16060
+    },
+    {
+      "epoch": 0.3570476359758265,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2858222222222223e-05,
+      "loss": 0.4039,
+      "step": 16070
+    },
+    {
+      "epoch": 0.35726981869889796,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2853777777777778e-05,
+      "loss": 0.3954,
+      "step": 16080
+    },
+    {
+      "epoch": 0.3574920014219694,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2849333333333334e-05,
+      "loss": 0.3688,
+      "step": 16090
+    },
+    {
+      "epoch": 0.35771418414504086,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2844888888888889e-05,
+      "loss": 0.4089,
+      "step": 16100
+    },
+    {
+      "epoch": 0.3579363668681123,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.2840444444444447e-05,
+      "loss": 0.3689,
+      "step": 16110
+    },
+    {
+      "epoch": 0.3581585495911838,
+      "grad_norm": 2.625,
+      "learning_rate": 1.2836000000000002e-05,
+      "loss": 0.3942,
+      "step": 16120
+    },
+    {
+      "epoch": 0.35838073231425527,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2831555555555557e-05,
+      "loss": 0.4092,
+      "step": 16130
+    },
+    {
+      "epoch": 0.3586029150373267,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2827111111111112e-05,
+      "loss": 0.3869,
+      "step": 16140
+    },
+    {
+      "epoch": 0.35882509776039817,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.2822666666666668e-05,
+      "loss": 0.4189,
+      "step": 16150
+    },
+    {
+      "epoch": 0.3590472804834696,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2818222222222223e-05,
+      "loss": 0.4198,
+      "step": 16160
+    },
+    {
+      "epoch": 0.3592694632065411,
+      "grad_norm": 2.125,
+      "learning_rate": 1.2813777777777778e-05,
+      "loss": 0.4085,
+      "step": 16170
+    },
+    {
+      "epoch": 0.3594916459296125,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.2809333333333333e-05,
+      "loss": 0.4151,
+      "step": 16180
+    },
+    {
+      "epoch": 0.359713828652684,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.280488888888889e-05,
+      "loss": 0.3998,
+      "step": 16190
+    },
+    {
+      "epoch": 0.3599360113757554,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2800444444444446e-05,
+      "loss": 0.431,
+      "step": 16200
+    },
+    {
+      "epoch": 0.3601581940988269,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2796000000000003e-05,
+      "loss": 0.4133,
+      "step": 16210
+    },
+    {
+      "epoch": 0.36038037682189833,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2791555555555557e-05,
+      "loss": 0.3874,
+      "step": 16220
+    },
+    {
+      "epoch": 0.3606025595449698,
+      "grad_norm": 2.125,
+      "learning_rate": 1.2787111111111112e-05,
+      "loss": 0.3901,
+      "step": 16230
+    },
+    {
+      "epoch": 0.36082474226804123,
+      "grad_norm": 2.625,
+      "learning_rate": 1.2782666666666667e-05,
+      "loss": 0.4536,
+      "step": 16240
+    },
+    {
+      "epoch": 0.3610469249911127,
+      "grad_norm": 2.5,
+      "learning_rate": 1.2778222222222224e-05,
+      "loss": 0.3992,
+      "step": 16250
+    },
+    {
+      "epoch": 0.36126910771418413,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2773777777777778e-05,
+      "loss": 0.3939,
+      "step": 16260
+    },
+    {
+      "epoch": 0.3614912904372556,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2769333333333333e-05,
+      "loss": 0.3851,
+      "step": 16270
+    },
+    {
+      "epoch": 0.36171347316032704,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.2764888888888888e-05,
+      "loss": 0.4086,
+      "step": 16280
+    },
+    {
+      "epoch": 0.3619356558833985,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.2760444444444446e-05,
+      "loss": 0.4101,
+      "step": 16290
+    },
+    {
+      "epoch": 0.36215783860646994,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.2756000000000001e-05,
+      "loss": 0.4031,
+      "step": 16300
+    },
+    {
+      "epoch": 0.3623800213295414,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2751555555555558e-05,
+      "loss": 0.3995,
+      "step": 16310
+    },
+    {
+      "epoch": 0.36260220405261284,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.2747111111111113e-05,
+      "loss": 0.3927,
+      "step": 16320
+    },
+    {
+      "epoch": 0.36282438677568435,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2742666666666667e-05,
+      "loss": 0.4459,
+      "step": 16330
+    },
+    {
+      "epoch": 0.3630465694987558,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.2738222222222222e-05,
+      "loss": 0.3973,
+      "step": 16340
+    },
+    {
+      "epoch": 0.36326875222182725,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.2733777777777779e-05,
+      "loss": 0.3703,
+      "step": 16350
+    },
+    {
+      "epoch": 0.3634909349448987,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.2729333333333334e-05,
+      "loss": 0.4159,
+      "step": 16360
+    },
+    {
+      "epoch": 0.36371311766797015,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.2724888888888889e-05,
+      "loss": 0.4324,
+      "step": 16370
+    },
+    {
+      "epoch": 0.3639353003910416,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2720444444444447e-05,
+      "loss": 0.4476,
+      "step": 16380
+    },
+    {
+      "epoch": 0.36415748311411306,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2716000000000002e-05,
+      "loss": 0.4055,
+      "step": 16390
+    },
+    {
+      "epoch": 0.3643796658371845,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.2711555555555556e-05,
+      "loss": 0.3853,
+      "step": 16400
+    },
+    {
+      "epoch": 0.36460184856025596,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2707111111111113e-05,
+      "loss": 0.4151,
+      "step": 16410
+    },
+    {
+      "epoch": 0.3648240312833274,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2702666666666668e-05,
+      "loss": 0.3948,
+      "step": 16420
+    },
+    {
+      "epoch": 0.36504621400639886,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2698222222222223e-05,
+      "loss": 0.3796,
+      "step": 16430
+    },
+    {
+      "epoch": 0.3652683967294703,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.2693777777777778e-05,
+      "loss": 0.428,
+      "step": 16440
+    },
+    {
+      "epoch": 0.36549057945254176,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2689333333333334e-05,
+      "loss": 0.4008,
+      "step": 16450
+    },
+    {
+      "epoch": 0.3657127621756132,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.268488888888889e-05,
+      "loss": 0.4065,
+      "step": 16460
+    },
+    {
+      "epoch": 0.36593494489868467,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.2680444444444447e-05,
+      "loss": 0.3712,
+      "step": 16470
+    },
+    {
+      "epoch": 0.3661571276217561,
+      "grad_norm": 2.125,
+      "learning_rate": 1.2676000000000002e-05,
+      "loss": 0.3857,
+      "step": 16480
+    },
+    {
+      "epoch": 0.36637931034482757,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2671555555555557e-05,
+      "loss": 0.3649,
+      "step": 16490
+    },
+    {
+      "epoch": 0.366601493067899,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2667111111111112e-05,
+      "loss": 0.4072,
+      "step": 16500
+    },
+    {
+      "epoch": 0.36682367579097047,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.2662666666666668e-05,
+      "loss": 0.3622,
+      "step": 16510
+    },
+    {
+      "epoch": 0.3670458585140419,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.2658222222222223e-05,
+      "loss": 0.3704,
+      "step": 16520
+    },
+    {
+      "epoch": 0.36726804123711343,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.2653777777777778e-05,
+      "loss": 0.3951,
+      "step": 16530
+    },
+    {
+      "epoch": 0.3674902239601849,
+      "grad_norm": 2.5,
+      "learning_rate": 1.2649333333333333e-05,
+      "loss": 0.3834,
+      "step": 16540
+    },
+    {
+      "epoch": 0.36771240668325633,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.2644888888888891e-05,
+      "loss": 0.3831,
+      "step": 16550
+    },
+    {
+      "epoch": 0.3679345894063278,
+      "grad_norm": 2.125,
+      "learning_rate": 1.2640444444444446e-05,
+      "loss": 0.4009,
+      "step": 16560
+    },
+    {
+      "epoch": 0.36815677212939923,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2636000000000002e-05,
+      "loss": 0.3755,
+      "step": 16570
+    },
+    {
+      "epoch": 0.3683789548524707,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2631555555555557e-05,
+      "loss": 0.3906,
+      "step": 16580
+    },
+    {
+      "epoch": 0.36860113757554214,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.2627111111111112e-05,
+      "loss": 0.386,
+      "step": 16590
+    },
+    {
+      "epoch": 0.3688233202986136,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2622666666666667e-05,
+      "loss": 0.3982,
+      "step": 16600
+    },
+    {
+      "epoch": 0.36904550302168504,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2618222222222223e-05,
+      "loss": 0.3969,
+      "step": 16610
+    },
+    {
+      "epoch": 0.3692676857447565,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2613777777777778e-05,
+      "loss": 0.3981,
+      "step": 16620
+    },
+    {
+      "epoch": 0.36948986846782794,
+      "grad_norm": 2.125,
+      "learning_rate": 1.2609333333333333e-05,
+      "loss": 0.3991,
+      "step": 16630
+    },
+    {
+      "epoch": 0.3697120511908994,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2604888888888891e-05,
+      "loss": 0.3996,
+      "step": 16640
+    },
+    {
+      "epoch": 0.36993423391397084,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2600444444444446e-05,
+      "loss": 0.3984,
+      "step": 16650
+    },
+    {
+      "epoch": 0.3701564166370423,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.2596000000000001e-05,
+      "loss": 0.3777,
+      "step": 16660
+    },
+    {
+      "epoch": 0.37037859936011375,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2591555555555558e-05,
+      "loss": 0.4213,
+      "step": 16670
+    },
+    {
+      "epoch": 0.3706007820831852,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2587111111111112e-05,
+      "loss": 0.3577,
+      "step": 16680
+    },
+    {
+      "epoch": 0.37082296480625665,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.2582666666666667e-05,
+      "loss": 0.4019,
+      "step": 16690
+    },
+    {
+      "epoch": 0.3710451475293281,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2578222222222222e-05,
+      "loss": 0.3884,
+      "step": 16700
+    },
+    {
+      "epoch": 0.37126733025239955,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.2573777777777779e-05,
+      "loss": 0.3848,
+      "step": 16710
+    },
+    {
+      "epoch": 0.371489512975471,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.2569333333333333e-05,
+      "loss": 0.3885,
+      "step": 16720
+    },
+    {
+      "epoch": 0.37171169569854245,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.2564888888888892e-05,
+      "loss": 0.404,
+      "step": 16730
+    },
+    {
+      "epoch": 0.37193387842161396,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2560444444444447e-05,
+      "loss": 0.4297,
+      "step": 16740
+    },
+    {
+      "epoch": 0.3721560611446854,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2556000000000001e-05,
+      "loss": 0.39,
+      "step": 16750
+    },
+    {
+      "epoch": 0.37237824386775686,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.2551555555555556e-05,
+      "loss": 0.3653,
+      "step": 16760
+    },
+    {
+      "epoch": 0.3726004265908283,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2547111111111113e-05,
+      "loss": 0.4062,
+      "step": 16770
+    },
+    {
+      "epoch": 0.37282260931389977,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2542666666666668e-05,
+      "loss": 0.3857,
+      "step": 16780
+    },
+    {
+      "epoch": 0.3730447920369712,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2538222222222222e-05,
+      "loss": 0.3776,
+      "step": 16790
+    },
+    {
+      "epoch": 0.37326697476004267,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2533777777777777e-05,
+      "loss": 0.391,
+      "step": 16800
+    },
+    {
+      "epoch": 0.3734891574831141,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2529333333333334e-05,
+      "loss": 0.4218,
+      "step": 16810
+    },
+    {
+      "epoch": 0.37371134020618557,
+      "grad_norm": 2.75,
+      "learning_rate": 1.252488888888889e-05,
+      "loss": 0.398,
+      "step": 16820
+    },
+    {
+      "epoch": 0.373933522929257,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.2520444444444447e-05,
+      "loss": 0.3775,
+      "step": 16830
+    },
+    {
+      "epoch": 0.3741557056523285,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2516000000000002e-05,
+      "loss": 0.3884,
+      "step": 16840
+    },
+    {
+      "epoch": 0.3743778883753999,
+      "grad_norm": 2.875,
+      "learning_rate": 1.2511555555555557e-05,
+      "loss": 0.4421,
+      "step": 16850
+    },
+    {
+      "epoch": 0.3746000710984714,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.2507111111111111e-05,
+      "loss": 0.3946,
+      "step": 16860
+    },
+    {
+      "epoch": 0.37482225382154283,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2502666666666668e-05,
+      "loss": 0.4128,
+      "step": 16870
+    },
+    {
+      "epoch": 0.3750444365446143,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.2498222222222223e-05,
+      "loss": 0.4037,
+      "step": 16880
+    },
+    {
+      "epoch": 0.37526661926768573,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.2493777777777778e-05,
+      "loss": 0.3919,
+      "step": 16890
+    },
+    {
+      "epoch": 0.3754888019907572,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2489333333333333e-05,
+      "loss": 0.3728,
+      "step": 16900
+    },
+    {
+      "epoch": 0.37571098471382863,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.248488888888889e-05,
+      "loss": 0.3716,
+      "step": 16910
+    },
+    {
+      "epoch": 0.3759331674369001,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2480444444444446e-05,
+      "loss": 0.4272,
+      "step": 16920
+    },
+    {
+      "epoch": 0.37615535015997154,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.2476000000000002e-05,
+      "loss": 0.3697,
+      "step": 16930
+    },
+    {
+      "epoch": 0.37637753288304304,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.2471555555555557e-05,
+      "loss": 0.3907,
+      "step": 16940
+    },
+    {
+      "epoch": 0.3765997156061145,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2467111111111112e-05,
+      "loss": 0.4133,
+      "step": 16950
+    },
+    {
+      "epoch": 0.37682189832918594,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2462666666666667e-05,
+      "loss": 0.4295,
+      "step": 16960
+    },
+    {
+      "epoch": 0.3770440810522574,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.2458222222222223e-05,
+      "loss": 0.3938,
+      "step": 16970
+    },
+    {
+      "epoch": 0.37726626377532885,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2453777777777778e-05,
+      "loss": 0.4059,
+      "step": 16980
+    },
+    {
+      "epoch": 0.3774884464984003,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2449333333333333e-05,
+      "loss": 0.366,
+      "step": 16990
+    },
+    {
+      "epoch": 0.37771062922147175,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.2444888888888891e-05,
+      "loss": 0.4216,
+      "step": 17000
+    },
+    {
+      "epoch": 0.3779328119445432,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.2440444444444446e-05,
+      "loss": 0.3814,
+      "step": 17010
+    },
+    {
+      "epoch": 0.37815499466761465,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.2436000000000001e-05,
+      "loss": 0.3967,
+      "step": 17020
+    },
+    {
+      "epoch": 0.3783771773906861,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2431555555555557e-05,
+      "loss": 0.399,
+      "step": 17030
+    },
+    {
+      "epoch": 0.37859936011375755,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2427111111111112e-05,
+      "loss": 0.3599,
+      "step": 17040
+    },
+    {
+      "epoch": 0.378821542836829,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.2422666666666667e-05,
+      "loss": 0.4512,
+      "step": 17050
+    },
+    {
+      "epoch": 0.37904372555990046,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.2418222222222222e-05,
+      "loss": 0.4067,
+      "step": 17060
+    },
+    {
+      "epoch": 0.3792659082829719,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.2413777777777778e-05,
+      "loss": 0.4283,
+      "step": 17070
+    },
+    {
+      "epoch": 0.37948809100604336,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.2409333333333335e-05,
+      "loss": 0.4055,
+      "step": 17080
+    },
+    {
+      "epoch": 0.3797102737291148,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2404888888888892e-05,
+      "loss": 0.3983,
+      "step": 17090
+    },
+    {
+      "epoch": 0.37993245645218626,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.2400444444444446e-05,
+      "loss": 0.4344,
+      "step": 17100
+    },
+    {
+      "epoch": 0.3801546391752577,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2396000000000001e-05,
+      "loss": 0.4133,
+      "step": 17110
+    },
+    {
+      "epoch": 0.38037682189832916,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2391555555555556e-05,
+      "loss": 0.4089,
+      "step": 17120
+    },
+    {
+      "epoch": 0.3805990046214006,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2387111111111113e-05,
+      "loss": 0.3972,
+      "step": 17130
+    },
+    {
+      "epoch": 0.38082118734447207,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2382666666666667e-05,
+      "loss": 0.3929,
+      "step": 17140
+    },
+    {
+      "epoch": 0.3810433700675436,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.2378222222222222e-05,
+      "loss": 0.4264,
+      "step": 17150
+    },
+    {
+      "epoch": 0.381265552790615,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2373777777777777e-05,
+      "loss": 0.3825,
+      "step": 17160
+    },
+    {
+      "epoch": 0.3814877355136865,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.2369333333333335e-05,
+      "loss": 0.3898,
+      "step": 17170
+    },
+    {
+      "epoch": 0.38170991823675793,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.236488888888889e-05,
+      "loss": 0.3757,
+      "step": 17180
+    },
+    {
+      "epoch": 0.3819321009598294,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2360444444444447e-05,
+      "loss": 0.4031,
+      "step": 17190
+    },
+    {
+      "epoch": 0.38215428368290083,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.2356000000000002e-05,
+      "loss": 0.4101,
+      "step": 17200
+    },
+    {
+      "epoch": 0.3823764664059723,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2351555555555556e-05,
+      "loss": 0.4174,
+      "step": 17210
+    },
+    {
+      "epoch": 0.38259864912904373,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.2347111111111111e-05,
+      "loss": 0.3623,
+      "step": 17220
+    },
+    {
+      "epoch": 0.3828208318521152,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.2342666666666668e-05,
+      "loss": 0.3731,
+      "step": 17230
+    },
+    {
+      "epoch": 0.38304301457518664,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2338222222222223e-05,
+      "loss": 0.4104,
+      "step": 17240
+    },
+    {
+      "epoch": 0.3832651972982581,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2333777777777778e-05,
+      "loss": 0.4195,
+      "step": 17250
+    },
+    {
+      "epoch": 0.38348738002132954,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.2329333333333336e-05,
+      "loss": 0.3738,
+      "step": 17260
+    },
+    {
+      "epoch": 0.383709562744401,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.232488888888889e-05,
+      "loss": 0.3437,
+      "step": 17270
+    },
+    {
+      "epoch": 0.38393174546747244,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.2320444444444445e-05,
+      "loss": 0.3965,
+      "step": 17280
+    },
+    {
+      "epoch": 0.3841539281905439,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2316000000000002e-05,
+      "loss": 0.4006,
+      "step": 17290
+    },
+    {
+      "epoch": 0.38437611091361534,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.2311555555555557e-05,
+      "loss": 0.3767,
+      "step": 17300
+    },
+    {
+      "epoch": 0.3845982936366868,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.2307111111111112e-05,
+      "loss": 0.3993,
+      "step": 17310
+    },
+    {
+      "epoch": 0.38482047635975825,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.2302666666666667e-05,
+      "loss": 0.3828,
+      "step": 17320
+    },
+    {
+      "epoch": 0.3850426590828297,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.2298222222222223e-05,
+      "loss": 0.3965,
+      "step": 17330
+    },
+    {
+      "epoch": 0.38526484180590115,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2293777777777778e-05,
+      "loss": 0.3843,
+      "step": 17340
+    },
+    {
+      "epoch": 0.3854870245289726,
+      "grad_norm": 2.0,
+      "learning_rate": 1.2289333333333336e-05,
+      "loss": 0.4183,
+      "step": 17350
+    },
+    {
+      "epoch": 0.3857092072520441,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2284888888888891e-05,
+      "loss": 0.399,
+      "step": 17360
+    },
+    {
+      "epoch": 0.38593138997511556,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.2280444444444446e-05,
+      "loss": 0.3745,
+      "step": 17370
+    },
+    {
+      "epoch": 0.386153572698187,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.2276e-05,
+      "loss": 0.3713,
+      "step": 17380
+    },
+    {
+      "epoch": 0.38637575542125846,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.2271555555555557e-05,
+      "loss": 0.4097,
+      "step": 17390
+    },
+    {
+      "epoch": 0.3865979381443299,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2267111111111112e-05,
+      "loss": 0.4068,
+      "step": 17400
+    },
+    {
+      "epoch": 0.38682012086740136,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2262666666666667e-05,
+      "loss": 0.4166,
+      "step": 17410
+    },
+    {
+      "epoch": 0.3870423035904728,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2258222222222222e-05,
+      "loss": 0.4135,
+      "step": 17420
+    },
+    {
+      "epoch": 0.38726448631354426,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2253777777777778e-05,
+      "loss": 0.3857,
+      "step": 17430
+    },
+    {
+      "epoch": 0.3874866690366157,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2249333333333335e-05,
+      "loss": 0.3891,
+      "step": 17440
+    },
+    {
+      "epoch": 0.38770885175968717,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2244888888888891e-05,
+      "loss": 0.3675,
+      "step": 17450
+    },
+    {
+      "epoch": 0.3879310344827586,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.2240444444444446e-05,
+      "loss": 0.3991,
+      "step": 17460
+    },
+    {
+      "epoch": 0.38815321720583007,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.2236000000000001e-05,
+      "loss": 0.3881,
+      "step": 17470
+    },
+    {
+      "epoch": 0.3883753999289015,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2231555555555556e-05,
+      "loss": 0.4251,
+      "step": 17480
+    },
+    {
+      "epoch": 0.388597582651973,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2227111111111112e-05,
+      "loss": 0.4082,
+      "step": 17490
+    },
+    {
+      "epoch": 0.3888197653750444,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2222666666666667e-05,
+      "loss": 0.38,
+      "step": 17500
+    },
+    {
+      "epoch": 0.3890419480981159,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2218222222222222e-05,
+      "loss": 0.4046,
+      "step": 17510
+    },
+    {
+      "epoch": 0.3892641308211873,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.2213777777777777e-05,
+      "loss": 0.3826,
+      "step": 17520
+    },
+    {
+      "epoch": 0.3894863135442588,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.2209333333333335e-05,
+      "loss": 0.3907,
+      "step": 17530
+    },
+    {
+      "epoch": 0.38970849626733023,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.220488888888889e-05,
+      "loss": 0.371,
+      "step": 17540
+    },
+    {
+      "epoch": 0.3899306789904017,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2200444444444447e-05,
+      "loss": 0.3836,
+      "step": 17550
+    },
+    {
+      "epoch": 0.3901528617134732,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2196000000000001e-05,
+      "loss": 0.4352,
+      "step": 17560
+    },
+    {
+      "epoch": 0.39037504443654464,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.2191555555555556e-05,
+      "loss": 0.3747,
+      "step": 17570
+    },
+    {
+      "epoch": 0.3905972271596161,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.2187111111111111e-05,
+      "loss": 0.3985,
+      "step": 17580
+    },
+    {
+      "epoch": 0.39081940988268754,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.2182666666666668e-05,
+      "loss": 0.4214,
+      "step": 17590
+    },
+    {
+      "epoch": 0.391041592605759,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2178222222222223e-05,
+      "loss": 0.3523,
+      "step": 17600
+    },
+    {
+      "epoch": 0.39126377532883044,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2173777777777777e-05,
+      "loss": 0.389,
+      "step": 17610
+    },
+    {
+      "epoch": 0.3914859580519019,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.2169333333333336e-05,
+      "loss": 0.4053,
+      "step": 17620
+    },
+    {
+      "epoch": 0.39170814077497335,
+      "grad_norm": 2.25,
+      "learning_rate": 1.216488888888889e-05,
+      "loss": 0.3832,
+      "step": 17630
+    },
+    {
+      "epoch": 0.3919303234980448,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2160444444444445e-05,
+      "loss": 0.3999,
+      "step": 17640
+    },
+    {
+      "epoch": 0.39215250622111625,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2156000000000002e-05,
+      "loss": 0.4054,
+      "step": 17650
+    },
+    {
+      "epoch": 0.3923746889441877,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.2151555555555557e-05,
+      "loss": 0.4095,
+      "step": 17660
+    },
+    {
+      "epoch": 0.39259687166725915,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.2147111111111112e-05,
+      "loss": 0.4301,
+      "step": 17670
+    },
+    {
+      "epoch": 0.3928190543903306,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2142666666666666e-05,
+      "loss": 0.3823,
+      "step": 17680
+    },
+    {
+      "epoch": 0.39304123711340205,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.2138222222222223e-05,
+      "loss": 0.4237,
+      "step": 17690
+    },
+    {
+      "epoch": 0.3932634198364735,
+      "grad_norm": 2.25,
+      "learning_rate": 1.213377777777778e-05,
+      "loss": 0.4074,
+      "step": 17700
+    },
+    {
+      "epoch": 0.39348560255954496,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2129333333333336e-05,
+      "loss": 0.4239,
+      "step": 17710
+    },
+    {
+      "epoch": 0.3937077852826164,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2124888888888891e-05,
+      "loss": 0.3956,
+      "step": 17720
+    },
+    {
+      "epoch": 0.39392996800568786,
+      "grad_norm": 2.0,
+      "learning_rate": 1.2120444444444446e-05,
+      "loss": 0.3783,
+      "step": 17730
+    },
+    {
+      "epoch": 0.3941521507287593,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.2116e-05,
+      "loss": 0.3882,
+      "step": 17740
+    },
+    {
+      "epoch": 0.39437433345183076,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2111555555555557e-05,
+      "loss": 0.3935,
+      "step": 17750
+    },
+    {
+      "epoch": 0.3945965161749022,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2107111111111112e-05,
+      "loss": 0.4248,
+      "step": 17760
+    },
+    {
+      "epoch": 0.3948186988979737,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.2102666666666667e-05,
+      "loss": 0.4196,
+      "step": 17770
+    },
+    {
+      "epoch": 0.39504088162104517,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.2098222222222222e-05,
+      "loss": 0.4248,
+      "step": 17780
+    },
+    {
+      "epoch": 0.3952630643441166,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.209377777777778e-05,
+      "loss": 0.3692,
+      "step": 17790
+    },
+    {
+      "epoch": 0.3954852470671881,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2089333333333335e-05,
+      "loss": 0.3917,
+      "step": 17800
+    },
+    {
+      "epoch": 0.3957074297902595,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.2084888888888891e-05,
+      "loss": 0.3894,
+      "step": 17810
+    },
+    {
+      "epoch": 0.395929612513331,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.2080444444444446e-05,
+      "loss": 0.3956,
+      "step": 17820
+    },
+    {
+      "epoch": 0.3961517952364024,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2076000000000001e-05,
+      "loss": 0.3929,
+      "step": 17830
+    },
+    {
+      "epoch": 0.3963739779594739,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.2071555555555556e-05,
+      "loss": 0.406,
+      "step": 17840
+    },
+    {
+      "epoch": 0.39659616068254533,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.2067111111111112e-05,
+      "loss": 0.3919,
+      "step": 17850
+    },
+    {
+      "epoch": 0.3968183434056168,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.2062666666666667e-05,
+      "loss": 0.417,
+      "step": 17860
+    },
+    {
+      "epoch": 0.39704052612868823,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.2058222222222222e-05,
+      "loss": 0.3916,
+      "step": 17870
+    },
+    {
+      "epoch": 0.3972627088517597,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.205377777777778e-05,
+      "loss": 0.4051,
+      "step": 17880
+    },
+    {
+      "epoch": 0.39748489157483113,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2049333333333335e-05,
+      "loss": 0.3559,
+      "step": 17890
+    },
+    {
+      "epoch": 0.3977070742979026,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.204488888888889e-05,
+      "loss": 0.4128,
+      "step": 17900
+    },
+    {
+      "epoch": 0.39792925702097404,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.2040444444444446e-05,
+      "loss": 0.3935,
+      "step": 17910
+    },
+    {
+      "epoch": 0.3981514397440455,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.2036000000000001e-05,
+      "loss": 0.396,
+      "step": 17920
+    },
+    {
+      "epoch": 0.39837362246711694,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2031555555555556e-05,
+      "loss": 0.388,
+      "step": 17930
+    },
+    {
+      "epoch": 0.3985958051901884,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2027111111111111e-05,
+      "loss": 0.4001,
+      "step": 17940
+    },
+    {
+      "epoch": 0.39881798791325984,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2022666666666668e-05,
+      "loss": 0.3741,
+      "step": 17950
+    },
+    {
+      "epoch": 0.3990401706363313,
+      "grad_norm": 2.5,
+      "learning_rate": 1.2018222222222222e-05,
+      "loss": 0.4268,
+      "step": 17960
+    },
+    {
+      "epoch": 0.3992623533594028,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.2013777777777779e-05,
+      "loss": 0.4013,
+      "step": 17970
+    },
+    {
+      "epoch": 0.39948453608247425,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.2009333333333335e-05,
+      "loss": 0.3873,
+      "step": 17980
+    },
+    {
+      "epoch": 0.3997067188055457,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.200488888888889e-05,
+      "loss": 0.3654,
+      "step": 17990
+    },
+    {
+      "epoch": 0.39992890152861715,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.2000444444444445e-05,
+      "loss": 0.3867,
+      "step": 18000
+    },
+    {
+      "epoch": 0.4001510842516886,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.1996000000000002e-05,
+      "loss": 0.415,
+      "step": 18010
+    },
+    {
+      "epoch": 0.40037326697476006,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.1991555555555557e-05,
+      "loss": 0.3893,
+      "step": 18020
+    },
+    {
+      "epoch": 0.4005954496978315,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1987111111111111e-05,
+      "loss": 0.398,
+      "step": 18030
+    },
+    {
+      "epoch": 0.40081763242090296,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.1982666666666666e-05,
+      "loss": 0.3704,
+      "step": 18040
+    },
+    {
+      "epoch": 0.4010398151439744,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1978222222222223e-05,
+      "loss": 0.3863,
+      "step": 18050
+    },
+    {
+      "epoch": 0.40126199786704586,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.197377777777778e-05,
+      "loss": 0.3964,
+      "step": 18060
+    },
+    {
+      "epoch": 0.4014841805901173,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1969333333333336e-05,
+      "loss": 0.3743,
+      "step": 18070
+    },
+    {
+      "epoch": 0.40170636331318876,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.196488888888889e-05,
+      "loss": 0.3623,
+      "step": 18080
+    },
+    {
+      "epoch": 0.4019285460362602,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.1960444444444446e-05,
+      "loss": 0.358,
+      "step": 18090
+    },
+    {
+      "epoch": 0.40215072875933167,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.1956e-05,
+      "loss": 0.4099,
+      "step": 18100
+    },
+    {
+      "epoch": 0.4023729114824031,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.1951555555555557e-05,
+      "loss": 0.3787,
+      "step": 18110
+    },
+    {
+      "epoch": 0.40259509420547457,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.1947111111111112e-05,
+      "loss": 0.3866,
+      "step": 18120
+    },
+    {
+      "epoch": 0.402817276928546,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.1942666666666667e-05,
+      "loss": 0.3891,
+      "step": 18130
+    },
+    {
+      "epoch": 0.40303945965161747,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.1938222222222221e-05,
+      "loss": 0.3637,
+      "step": 18140
+    },
+    {
+      "epoch": 0.4032616423746889,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.193377777777778e-05,
+      "loss": 0.3642,
+      "step": 18150
+    },
+    {
+      "epoch": 0.4034838250977604,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1929333333333335e-05,
+      "loss": 0.3591,
+      "step": 18160
+    },
+    {
+      "epoch": 0.4037060078208318,
+      "grad_norm": 2.5,
+      "learning_rate": 1.1924888888888891e-05,
+      "loss": 0.3992,
+      "step": 18170
+    },
+    {
+      "epoch": 0.40392819054390333,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.1920444444444446e-05,
+      "loss": 0.3903,
+      "step": 18180
+    },
+    {
+      "epoch": 0.4041503732669748,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.1916e-05,
+      "loss": 0.4211,
+      "step": 18190
+    },
+    {
+      "epoch": 0.40437255599004623,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1911555555555556e-05,
+      "loss": 0.4168,
+      "step": 18200
+    },
+    {
+      "epoch": 0.4045947387131177,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.1907111111111112e-05,
+      "loss": 0.4052,
+      "step": 18210
+    },
+    {
+      "epoch": 0.40481692143618914,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1902666666666667e-05,
+      "loss": 0.404,
+      "step": 18220
+    },
+    {
+      "epoch": 0.4050391041592606,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.1898222222222222e-05,
+      "loss": 0.4127,
+      "step": 18230
+    },
+    {
+      "epoch": 0.40526128688233204,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.189377777777778e-05,
+      "loss": 0.3987,
+      "step": 18240
+    },
+    {
+      "epoch": 0.4054834696054035,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.1889333333333335e-05,
+      "loss": 0.3983,
+      "step": 18250
+    },
+    {
+      "epoch": 0.40570565232847494,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.188488888888889e-05,
+      "loss": 0.3852,
+      "step": 18260
+    },
+    {
+      "epoch": 0.4059278350515464,
+      "grad_norm": 2.625,
+      "learning_rate": 1.1880444444444446e-05,
+      "loss": 0.382,
+      "step": 18270
+    },
+    {
+      "epoch": 0.40615001777461784,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.1876000000000001e-05,
+      "loss": 0.3908,
+      "step": 18280
+    },
+    {
+      "epoch": 0.4063722004976893,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1871555555555556e-05,
+      "loss": 0.4179,
+      "step": 18290
+    },
+    {
+      "epoch": 0.40659438322076075,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.186711111111111e-05,
+      "loss": 0.4292,
+      "step": 18300
+    },
+    {
+      "epoch": 0.4068165659438322,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1862666666666667e-05,
+      "loss": 0.4112,
+      "step": 18310
+    },
+    {
+      "epoch": 0.40703874866690365,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.1858222222222224e-05,
+      "loss": 0.4137,
+      "step": 18320
+    },
+    {
+      "epoch": 0.4072609313899751,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.1853777777777779e-05,
+      "loss": 0.4142,
+      "step": 18330
+    },
+    {
+      "epoch": 0.40748311411304655,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1849333333333335e-05,
+      "loss": 0.3724,
+      "step": 18340
+    },
+    {
+      "epoch": 0.407705296836118,
+      "grad_norm": 2.125,
+      "learning_rate": 1.184488888888889e-05,
+      "loss": 0.3825,
+      "step": 18350
+    },
+    {
+      "epoch": 0.40792747955918945,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1840444444444445e-05,
+      "loss": 0.4147,
+      "step": 18360
+    },
+    {
+      "epoch": 0.4081496622822609,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1836000000000002e-05,
+      "loss": 0.38,
+      "step": 18370
+    },
+    {
+      "epoch": 0.4083718450053324,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1831555555555556e-05,
+      "loss": 0.3932,
+      "step": 18380
+    },
+    {
+      "epoch": 0.40859402772840386,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1827111111111111e-05,
+      "loss": 0.4194,
+      "step": 18390
+    },
+    {
+      "epoch": 0.4088162104514753,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1822666666666666e-05,
+      "loss": 0.4206,
+      "step": 18400
+    },
+    {
+      "epoch": 0.40903839317454677,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1818222222222224e-05,
+      "loss": 0.3969,
+      "step": 18410
+    },
+    {
+      "epoch": 0.4092605758976182,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.1813777777777779e-05,
+      "loss": 0.3743,
+      "step": 18420
+    },
+    {
+      "epoch": 0.40948275862068967,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.1809333333333336e-05,
+      "loss": 0.4294,
+      "step": 18430
+    },
+    {
+      "epoch": 0.4097049413437611,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.180488888888889e-05,
+      "loss": 0.3588,
+      "step": 18440
+    },
+    {
+      "epoch": 0.40992712406683257,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1800444444444445e-05,
+      "loss": 0.404,
+      "step": 18450
+    },
+    {
+      "epoch": 0.410149306789904,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1796e-05,
+      "loss": 0.422,
+      "step": 18460
+    },
+    {
+      "epoch": 0.4103714895129755,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1791555555555557e-05,
+      "loss": 0.3502,
+      "step": 18470
+    },
+    {
+      "epoch": 0.4105936722360469,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1787111111111112e-05,
+      "loss": 0.3651,
+      "step": 18480
+    },
+    {
+      "epoch": 0.4108158549591184,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.1782666666666666e-05,
+      "loss": 0.4026,
+      "step": 18490
+    },
+    {
+      "epoch": 0.4110380376821898,
+      "grad_norm": 2.125,
+      "learning_rate": 1.1778222222222225e-05,
+      "loss": 0.4164,
+      "step": 18500
+    },
+    {
+      "epoch": 0.4112602204052613,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.177377777777778e-05,
+      "loss": 0.3597,
+      "step": 18510
+    },
+    {
+      "epoch": 0.41148240312833273,
+      "grad_norm": 2.0,
+      "learning_rate": 1.1769333333333334e-05,
+      "loss": 0.3407,
+      "step": 18520
+    },
+    {
+      "epoch": 0.4117045858514042,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.1764888888888891e-05,
+      "loss": 0.4096,
+      "step": 18530
+    },
+    {
+      "epoch": 0.41192676857447563,
+      "grad_norm": 2.25,
+      "learning_rate": 1.1760444444444446e-05,
+      "loss": 0.3792,
+      "step": 18540
+    },
+    {
+      "epoch": 0.4121489512975471,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.1756e-05,
+      "loss": 0.391,
+      "step": 18550
+    },
+    {
+      "epoch": 0.41237113402061853,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.1751555555555555e-05,
+      "loss": 0.3822,
+      "step": 18560
+    },
+    {
+      "epoch": 0.41259331674369,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.1747111111111112e-05,
+      "loss": 0.3866,
+      "step": 18570
+    },
+    {
+      "epoch": 0.41281549946676144,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1742666666666667e-05,
+      "loss": 0.3711,
+      "step": 18580
+    },
+    {
+      "epoch": 0.41303768218983294,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1738222222222223e-05,
+      "loss": 0.4253,
+      "step": 18590
+    },
+    {
+      "epoch": 0.4132598649129044,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.173377777777778e-05,
+      "loss": 0.3965,
+      "step": 18600
+    },
+    {
+      "epoch": 0.41348204763597585,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1729333333333335e-05,
+      "loss": 0.4161,
+      "step": 18610
+    },
+    {
+      "epoch": 0.4137042303590473,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.172488888888889e-05,
+      "loss": 0.3879,
+      "step": 18620
+    },
+    {
+      "epoch": 0.41392641308211875,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.1720444444444446e-05,
+      "loss": 0.4,
+      "step": 18630
+    },
+    {
+      "epoch": 0.4141485958051902,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1716000000000001e-05,
+      "loss": 0.4214,
+      "step": 18640
+    },
+    {
+      "epoch": 0.41437077852826165,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.1711555555555556e-05,
+      "loss": 0.4266,
+      "step": 18650
+    },
+    {
+      "epoch": 0.4145929612513331,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.170711111111111e-05,
+      "loss": 0.4241,
+      "step": 18660
+    },
+    {
+      "epoch": 0.41481514397440455,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1702666666666667e-05,
+      "loss": 0.3736,
+      "step": 18670
+    },
+    {
+      "epoch": 0.415037326697476,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1698222222222224e-05,
+      "loss": 0.3792,
+      "step": 18680
+    },
+    {
+      "epoch": 0.41525950942054746,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.1693777777777779e-05,
+      "loss": 0.3791,
+      "step": 18690
+    },
+    {
+      "epoch": 0.4154816921436189,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.1689333333333335e-05,
+      "loss": 0.4218,
+      "step": 18700
+    },
+    {
+      "epoch": 0.41570387486669036,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.168488888888889e-05,
+      "loss": 0.3575,
+      "step": 18710
+    },
+    {
+      "epoch": 0.4159260575897618,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.1680444444444445e-05,
+      "loss": 0.4201,
+      "step": 18720
+    },
+    {
+      "epoch": 0.41614824031283326,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1676000000000001e-05,
+      "loss": 0.4063,
+      "step": 18730
+    },
+    {
+      "epoch": 0.4163704230359047,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1671555555555556e-05,
+      "loss": 0.4136,
+      "step": 18740
+    },
+    {
+      "epoch": 0.41659260575897616,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1667111111111111e-05,
+      "loss": 0.3875,
+      "step": 18750
+    },
+    {
+      "epoch": 0.4168147884820476,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1662666666666666e-05,
+      "loss": 0.4041,
+      "step": 18760
+    },
+    {
+      "epoch": 0.41703697120511907,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1658222222222224e-05,
+      "loss": 0.368,
+      "step": 18770
+    },
+    {
+      "epoch": 0.4172591539281905,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.1653777777777779e-05,
+      "loss": 0.3725,
+      "step": 18780
+    },
+    {
+      "epoch": 0.417481336651262,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1649333333333336e-05,
+      "loss": 0.3615,
+      "step": 18790
+    },
+    {
+      "epoch": 0.4177035193743335,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.164488888888889e-05,
+      "loss": 0.3657,
+      "step": 18800
+    },
+    {
+      "epoch": 0.4179257020974049,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1640444444444445e-05,
+      "loss": 0.4285,
+      "step": 18810
+    },
+    {
+      "epoch": 0.4181478848204764,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1636e-05,
+      "loss": 0.3817,
+      "step": 18820
+    },
+    {
+      "epoch": 0.41837006754354783,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.1631555555555557e-05,
+      "loss": 0.4134,
+      "step": 18830
+    },
+    {
+      "epoch": 0.4185922502666193,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.1627111111111111e-05,
+      "loss": 0.4261,
+      "step": 18840
+    },
+    {
+      "epoch": 0.41881443298969073,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1622666666666666e-05,
+      "loss": 0.3703,
+      "step": 18850
+    },
+    {
+      "epoch": 0.4190366157127622,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.1618222222222225e-05,
+      "loss": 0.3884,
+      "step": 18860
+    },
+    {
+      "epoch": 0.41925879843583364,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.161377777777778e-05,
+      "loss": 0.364,
+      "step": 18870
+    },
+    {
+      "epoch": 0.4194809811589051,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.1609333333333334e-05,
+      "loss": 0.3927,
+      "step": 18880
+    },
+    {
+      "epoch": 0.41970316388197654,
+      "grad_norm": 2.375,
+      "learning_rate": 1.160488888888889e-05,
+      "loss": 0.3564,
+      "step": 18890
+    },
+    {
+      "epoch": 0.419925346605048,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.1600444444444446e-05,
+      "loss": 0.3842,
+      "step": 18900
+    },
+    {
+      "epoch": 0.42014752932811944,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.1596e-05,
+      "loss": 0.4098,
+      "step": 18910
+    },
+    {
+      "epoch": 0.4203697120511909,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.1591555555555555e-05,
+      "loss": 0.3775,
+      "step": 18920
+    },
+    {
+      "epoch": 0.42059189477426234,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1587111111111112e-05,
+      "loss": 0.4061,
+      "step": 18930
+    },
+    {
+      "epoch": 0.4208140774973338,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1582666666666668e-05,
+      "loss": 0.4203,
+      "step": 18940
+    },
+    {
+      "epoch": 0.42103626022040525,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.1578222222222223e-05,
+      "loss": 0.4124,
+      "step": 18950
+    },
+    {
+      "epoch": 0.4212584429434767,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.157377777777778e-05,
+      "loss": 0.3668,
+      "step": 18960
+    },
+    {
+      "epoch": 0.42148062566654815,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1569333333333335e-05,
+      "loss": 0.3855,
+      "step": 18970
+    },
+    {
+      "epoch": 0.4217028083896196,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.156488888888889e-05,
+      "loss": 0.3755,
+      "step": 18980
+    },
+    {
+      "epoch": 0.42192499111269105,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1560444444444446e-05,
+      "loss": 0.4197,
+      "step": 18990
+    },
+    {
+      "epoch": 0.42214717383576256,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.1556e-05,
+      "loss": 0.383,
+      "step": 19000
+    },
+    {
+      "epoch": 0.422369356558834,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.1551555555555556e-05,
+      "loss": 0.4043,
+      "step": 19010
+    },
+    {
+      "epoch": 0.42259153928190546,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.154711111111111e-05,
+      "loss": 0.401,
+      "step": 19020
+    },
+    {
+      "epoch": 0.4228137220049769,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.1542666666666669e-05,
+      "loss": 0.3877,
+      "step": 19030
+    },
+    {
+      "epoch": 0.42303590472804836,
+      "grad_norm": 2.5,
+      "learning_rate": 1.1538222222222224e-05,
+      "loss": 0.4399,
+      "step": 19040
+    },
+    {
+      "epoch": 0.4232580874511198,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.153377777777778e-05,
+      "loss": 0.4048,
+      "step": 19050
+    },
+    {
+      "epoch": 0.42348027017419126,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.1529333333333335e-05,
+      "loss": 0.3851,
+      "step": 19060
+    },
+    {
+      "epoch": 0.4237024528972627,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.152488888888889e-05,
+      "loss": 0.4034,
+      "step": 19070
+    },
+    {
+      "epoch": 0.42392463562033417,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1520444444444445e-05,
+      "loss": 0.4198,
+      "step": 19080
+    },
+    {
+      "epoch": 0.4241468183434056,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1516000000000001e-05,
+      "loss": 0.4207,
+      "step": 19090
+    },
+    {
+      "epoch": 0.42436900106647707,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.1511555555555556e-05,
+      "loss": 0.4425,
+      "step": 19100
+    },
+    {
+      "epoch": 0.4245911837895485,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.1507111111111111e-05,
+      "loss": 0.3781,
+      "step": 19110
+    },
+    {
+      "epoch": 0.42481336651261997,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.1502666666666669e-05,
+      "loss": 0.4106,
+      "step": 19120
+    },
+    {
+      "epoch": 0.4250355492356914,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1498222222222224e-05,
+      "loss": 0.3809,
+      "step": 19130
+    },
+    {
+      "epoch": 0.4252577319587629,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1493777777777779e-05,
+      "loss": 0.3521,
+      "step": 19140
+    },
+    {
+      "epoch": 0.4254799146818343,
+      "grad_norm": 2.25,
+      "learning_rate": 1.1489333333333335e-05,
+      "loss": 0.388,
+      "step": 19150
+    },
+    {
+      "epoch": 0.4257020974049058,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.148488888888889e-05,
+      "loss": 0.4044,
+      "step": 19160
+    },
+    {
+      "epoch": 0.42592428012797723,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.1480444444444445e-05,
+      "loss": 0.4445,
+      "step": 19170
+    },
+    {
+      "epoch": 0.4261464628510487,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.1476e-05,
+      "loss": 0.405,
+      "step": 19180
+    },
+    {
+      "epoch": 0.42636864557412013,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.1471555555555556e-05,
+      "loss": 0.3559,
+      "step": 19190
+    },
+    {
+      "epoch": 0.42659082829719164,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1467111111111111e-05,
+      "loss": 0.393,
+      "step": 19200
+    },
+    {
+      "epoch": 0.4268130110202631,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1462666666666668e-05,
+      "loss": 0.3441,
+      "step": 19210
+    },
+    {
+      "epoch": 0.42703519374333454,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.1458222222222224e-05,
+      "loss": 0.3888,
+      "step": 19220
+    },
+    {
+      "epoch": 0.427257376466406,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.145377777777778e-05,
+      "loss": 0.399,
+      "step": 19230
+    },
+    {
+      "epoch": 0.42747955918947744,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.1449333333333334e-05,
+      "loss": 0.3745,
+      "step": 19240
+    },
+    {
+      "epoch": 0.4277017419125489,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.144488888888889e-05,
+      "loss": 0.3968,
+      "step": 19250
+    },
+    {
+      "epoch": 0.42792392463562035,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.1440444444444445e-05,
+      "loss": 0.4016,
+      "step": 19260
+    },
+    {
+      "epoch": 0.4281461073586918,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1436e-05,
+      "loss": 0.3813,
+      "step": 19270
+    },
+    {
+      "epoch": 0.42836829008176325,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.1431555555555555e-05,
+      "loss": 0.4106,
+      "step": 19280
+    },
+    {
+      "epoch": 0.4285904728048347,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.1427111111111112e-05,
+      "loss": 0.3633,
+      "step": 19290
+    },
+    {
+      "epoch": 0.42881265552790615,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1422666666666668e-05,
+      "loss": 0.3968,
+      "step": 19300
+    },
+    {
+      "epoch": 0.4290348382509776,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1418222222222223e-05,
+      "loss": 0.4016,
+      "step": 19310
+    },
+    {
+      "epoch": 0.42925702097404905,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.141377777777778e-05,
+      "loss": 0.3636,
+      "step": 19320
+    },
+    {
+      "epoch": 0.4294792036971205,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.1409333333333334e-05,
+      "loss": 0.3992,
+      "step": 19330
+    },
+    {
+      "epoch": 0.42970138642019196,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.140488888888889e-05,
+      "loss": 0.4122,
+      "step": 19340
+    },
+    {
+      "epoch": 0.4299235691432634,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.1400444444444446e-05,
+      "loss": 0.3773,
+      "step": 19350
+    },
+    {
+      "epoch": 0.43014575186633486,
+      "grad_norm": 2.625,
+      "learning_rate": 1.1396e-05,
+      "loss": 0.3969,
+      "step": 19360
+    },
+    {
+      "epoch": 0.4303679345894063,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.1391555555555556e-05,
+      "loss": 0.3679,
+      "step": 19370
+    },
+    {
+      "epoch": 0.43059011731247776,
+      "grad_norm": 2.625,
+      "learning_rate": 1.138711111111111e-05,
+      "loss": 0.3742,
+      "step": 19380
+    },
+    {
+      "epoch": 0.4308123000355492,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.1382666666666669e-05,
+      "loss": 0.3824,
+      "step": 19390
+    },
+    {
+      "epoch": 0.43103448275862066,
+      "grad_norm": 2.5,
+      "learning_rate": 1.1378222222222223e-05,
+      "loss": 0.4468,
+      "step": 19400
+    },
+    {
+      "epoch": 0.43125666548169217,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.137377777777778e-05,
+      "loss": 0.3628,
+      "step": 19410
+    },
+    {
+      "epoch": 0.4314788482047636,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.1369333333333335e-05,
+      "loss": 0.3756,
+      "step": 19420
+    },
+    {
+      "epoch": 0.43170103092783507,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.136488888888889e-05,
+      "loss": 0.3864,
+      "step": 19430
+    },
+    {
+      "epoch": 0.4319232136509065,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1360444444444445e-05,
+      "loss": 0.4071,
+      "step": 19440
+    },
+    {
+      "epoch": 0.432145396373978,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.1356000000000001e-05,
+      "loss": 0.4236,
+      "step": 19450
+    },
+    {
+      "epoch": 0.4323675790970494,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1351555555555556e-05,
+      "loss": 0.423,
+      "step": 19460
+    },
+    {
+      "epoch": 0.4325897618201209,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.134711111111111e-05,
+      "loss": 0.379,
+      "step": 19470
+    },
+    {
+      "epoch": 0.43281194454319233,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.1342666666666669e-05,
+      "loss": 0.4025,
+      "step": 19480
+    },
+    {
+      "epoch": 0.4330341272662638,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1338222222222224e-05,
+      "loss": 0.3735,
+      "step": 19490
+    },
+    {
+      "epoch": 0.43325630998933523,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1333777777777779e-05,
+      "loss": 0.397,
+      "step": 19500
+    },
+    {
+      "epoch": 0.4334784927124067,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1329333333333335e-05,
+      "loss": 0.4139,
+      "step": 19510
+    },
+    {
+      "epoch": 0.43370067543547813,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.132488888888889e-05,
+      "loss": 0.391,
+      "step": 19520
+    },
+    {
+      "epoch": 0.4339228581585496,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.1320444444444445e-05,
+      "loss": 0.3767,
+      "step": 19530
+    },
+    {
+      "epoch": 0.43414504088162104,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1316e-05,
+      "loss": 0.4362,
+      "step": 19540
+    },
+    {
+      "epoch": 0.4343672236046925,
+      "grad_norm": 3.0,
+      "learning_rate": 1.1311555555555556e-05,
+      "loss": 0.4181,
+      "step": 19550
+    },
+    {
+      "epoch": 0.43458940632776394,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.1307111111111111e-05,
+      "loss": 0.363,
+      "step": 19560
+    },
+    {
+      "epoch": 0.4348115890508354,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.1302666666666668e-05,
+      "loss": 0.4238,
+      "step": 19570
+    },
+    {
+      "epoch": 0.43503377177390684,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1298222222222224e-05,
+      "loss": 0.3864,
+      "step": 19580
+    },
+    {
+      "epoch": 0.4352559544969783,
+      "grad_norm": 2.0,
+      "learning_rate": 1.1293777777777779e-05,
+      "loss": 0.4356,
+      "step": 19590
+    },
+    {
+      "epoch": 0.43547813722004974,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1289333333333334e-05,
+      "loss": 0.4085,
+      "step": 19600
+    },
+    {
+      "epoch": 0.43570031994312125,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.128488888888889e-05,
+      "loss": 0.3822,
+      "step": 19610
+    },
+    {
+      "epoch": 0.4359225026661927,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.1280444444444445e-05,
+      "loss": 0.3907,
+      "step": 19620
+    },
+    {
+      "epoch": 0.43614468538926415,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1276e-05,
+      "loss": 0.4203,
+      "step": 19630
+    },
+    {
+      "epoch": 0.4363668681123356,
+      "grad_norm": 1.875,
+      "learning_rate": 1.1271555555555555e-05,
+      "loss": 0.4278,
+      "step": 19640
+    },
+    {
+      "epoch": 0.43658905083540706,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.1267111111111113e-05,
+      "loss": 0.3878,
+      "step": 19650
+    },
+    {
+      "epoch": 0.4368112335584785,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.1262666666666668e-05,
+      "loss": 0.4072,
+      "step": 19660
+    },
+    {
+      "epoch": 0.43703341628154996,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.1258222222222223e-05,
+      "loss": 0.3515,
+      "step": 19670
+    },
+    {
+      "epoch": 0.4372555990046214,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.125377777777778e-05,
+      "loss": 0.3994,
+      "step": 19680
+    },
+    {
+      "epoch": 0.43747778172769286,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.1249333333333334e-05,
+      "loss": 0.4353,
+      "step": 19690
+    },
+    {
+      "epoch": 0.4376999644507643,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1244888888888889e-05,
+      "loss": 0.42,
+      "step": 19700
+    },
+    {
+      "epoch": 0.43792214717383576,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.1240444444444446e-05,
+      "loss": 0.3881,
+      "step": 19710
+    },
+    {
+      "epoch": 0.4381443298969072,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.1236e-05,
+      "loss": 0.3918,
+      "step": 19720
+    },
+    {
+      "epoch": 0.43836651261997867,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1231555555555555e-05,
+      "loss": 0.3917,
+      "step": 19730
+    },
+    {
+      "epoch": 0.4385886953430501,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1227111111111114e-05,
+      "loss": 0.3976,
+      "step": 19740
+    },
+    {
+      "epoch": 0.43881087806612157,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1222666666666668e-05,
+      "loss": 0.3924,
+      "step": 19750
+    },
+    {
+      "epoch": 0.439033060789193,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.1218222222222223e-05,
+      "loss": 0.3909,
+      "step": 19760
+    },
+    {
+      "epoch": 0.43925524351226447,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.121377777777778e-05,
+      "loss": 0.42,
+      "step": 19770
+    },
+    {
+      "epoch": 0.4394774262353359,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.1209333333333335e-05,
+      "loss": 0.3655,
+      "step": 19780
+    },
+    {
+      "epoch": 0.4396996089584074,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.120488888888889e-05,
+      "loss": 0.3663,
+      "step": 19790
+    },
+    {
+      "epoch": 0.4399217916814788,
+      "grad_norm": 2.125,
+      "learning_rate": 1.1200444444444444e-05,
+      "loss": 0.3854,
+      "step": 19800
+    },
+    {
+      "epoch": 0.4401439744045503,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.1196000000000001e-05,
+      "loss": 0.374,
+      "step": 19810
+    },
+    {
+      "epoch": 0.4403661571276218,
+      "grad_norm": 1.875,
+      "learning_rate": 1.1191555555555556e-05,
+      "loss": 0.3725,
+      "step": 19820
+    },
+    {
+      "epoch": 0.44058833985069323,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.1187111111111112e-05,
+      "loss": 0.3875,
+      "step": 19830
+    },
+    {
+      "epoch": 0.4408105225737647,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.1182666666666669e-05,
+      "loss": 0.4269,
+      "step": 19840
+    },
+    {
+      "epoch": 0.44103270529683614,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.1178222222222224e-05,
+      "loss": 0.3947,
+      "step": 19850
+    },
+    {
+      "epoch": 0.4412548880199076,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.1173777777777779e-05,
+      "loss": 0.4126,
+      "step": 19860
+    },
+    {
+      "epoch": 0.44147707074297904,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.1169333333333335e-05,
+      "loss": 0.4062,
+      "step": 19870
+    },
+    {
+      "epoch": 0.4416992534660505,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.116488888888889e-05,
+      "loss": 0.3836,
+      "step": 19880
+    },
+    {
+      "epoch": 0.44192143618912194,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.1160444444444445e-05,
+      "loss": 0.3643,
+      "step": 19890
+    },
+    {
+      "epoch": 0.4421436189121934,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.1156e-05,
+      "loss": 0.4177,
+      "step": 19900
+    },
+    {
+      "epoch": 0.44236580163526484,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.1151555555555556e-05,
+      "loss": 0.3948,
+      "step": 19910
+    },
+    {
+      "epoch": 0.4425879843583363,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.1147111111111113e-05,
+      "loss": 0.3898,
+      "step": 19920
+    },
+    {
+      "epoch": 0.44281016708140775,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.1142666666666667e-05,
+      "loss": 0.3983,
+      "step": 19930
+    },
+    {
+      "epoch": 0.4430323498044792,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1138222222222224e-05,
+      "loss": 0.3946,
+      "step": 19940
+    },
+    {
+      "epoch": 0.44325453252755065,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1133777777777779e-05,
+      "loss": 0.3744,
+      "step": 19950
+    },
+    {
+      "epoch": 0.4434767152506221,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1129333333333334e-05,
+      "loss": 0.4106,
+      "step": 19960
+    },
+    {
+      "epoch": 0.44369889797369355,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.112488888888889e-05,
+      "loss": 0.3814,
+      "step": 19970
+    },
+    {
+      "epoch": 0.443921080696765,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.1120444444444445e-05,
+      "loss": 0.4158,
+      "step": 19980
+    },
+    {
+      "epoch": 0.44414326341983645,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.1116e-05,
+      "loss": 0.4096,
+      "step": 19990
+    },
+    {
+      "epoch": 0.4443654461429079,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.1111555555555555e-05,
+      "loss": 0.3942,
+      "step": 20000
+    },
+    {
+      "epoch": 0.44458762886597936,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1107111111111113e-05,
+      "loss": 0.3741,
+      "step": 20010
+    },
+    {
+      "epoch": 0.44480981158905086,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.1102666666666668e-05,
+      "loss": 0.3804,
+      "step": 20020
+    },
+    {
+      "epoch": 0.4450319943121223,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.1098222222222223e-05,
+      "loss": 0.3675,
+      "step": 20030
+    },
+    {
+      "epoch": 0.44525417703519377,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.109377777777778e-05,
+      "loss": 0.3957,
+      "step": 20040
+    },
+    {
+      "epoch": 0.4454763597582652,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.1089333333333334e-05,
+      "loss": 0.3769,
+      "step": 20050
+    },
+    {
+      "epoch": 0.44569854248133667,
+      "grad_norm": 2.0,
+      "learning_rate": 1.1084888888888889e-05,
+      "loss": 0.3605,
+      "step": 20060
+    },
+    {
+      "epoch": 0.4459207252044081,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1080444444444445e-05,
+      "loss": 0.4117,
+      "step": 20070
+    },
+    {
+      "epoch": 0.44614290792747957,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.1076e-05,
+      "loss": 0.4063,
+      "step": 20080
+    },
+    {
+      "epoch": 0.446365090650551,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1071555555555555e-05,
+      "loss": 0.4014,
+      "step": 20090
+    },
+    {
+      "epoch": 0.4465872733736225,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.1067111111111113e-05,
+      "loss": 0.3968,
+      "step": 20100
+    },
+    {
+      "epoch": 0.4468094560966939,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1062666666666668e-05,
+      "loss": 0.3791,
+      "step": 20110
+    },
+    {
+      "epoch": 0.4470316388197654,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1058222222222223e-05,
+      "loss": 0.4061,
+      "step": 20120
+    },
+    {
+      "epoch": 0.4472538215428368,
+      "grad_norm": 2.375,
+      "learning_rate": 1.105377777777778e-05,
+      "loss": 0.4039,
+      "step": 20130
+    },
+    {
+      "epoch": 0.4474760042659083,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.1049333333333334e-05,
+      "loss": 0.3768,
+      "step": 20140
+    },
+    {
+      "epoch": 0.44769818698897973,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.104488888888889e-05,
+      "loss": 0.388,
+      "step": 20150
+    },
+    {
+      "epoch": 0.4479203697120512,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1040444444444444e-05,
+      "loss": 0.4123,
+      "step": 20160
+    },
+    {
+      "epoch": 0.44814255243512263,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.1036e-05,
+      "loss": 0.3996,
+      "step": 20170
+    },
+    {
+      "epoch": 0.4483647351581941,
+      "grad_norm": 2.0,
+      "learning_rate": 1.1031555555555556e-05,
+      "loss": 0.3852,
+      "step": 20180
+    },
+    {
+      "epoch": 0.44858691788126553,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1027111111111112e-05,
+      "loss": 0.3636,
+      "step": 20190
+    },
+    {
+      "epoch": 0.448809100604337,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.1022666666666669e-05,
+      "loss": 0.4001,
+      "step": 20200
+    },
+    {
+      "epoch": 0.44903128332740844,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1018222222222223e-05,
+      "loss": 0.3724,
+      "step": 20210
+    },
+    {
+      "epoch": 0.4492534660504799,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1013777777777778e-05,
+      "loss": 0.3837,
+      "step": 20220
+    },
+    {
+      "epoch": 0.4494756487735514,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.1009333333333335e-05,
+      "loss": 0.3334,
+      "step": 20230
+    },
+    {
+      "epoch": 0.44969783149662285,
+      "grad_norm": 2.75,
+      "learning_rate": 1.100488888888889e-05,
+      "loss": 0.41,
+      "step": 20240
+    },
+    {
+      "epoch": 0.4499200142196943,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1000444444444445e-05,
+      "loss": 0.4168,
+      "step": 20250
+    },
+    {
+      "epoch": 0.45014219694276575,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0996e-05,
+      "loss": 0.4111,
+      "step": 20260
+    },
+    {
+      "epoch": 0.4503643796658372,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.0991555555555558e-05,
+      "loss": 0.3801,
+      "step": 20270
+    },
+    {
+      "epoch": 0.45058656238890865,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0987111111111112e-05,
+      "loss": 0.3671,
+      "step": 20280
+    },
+    {
+      "epoch": 0.4508087451119801,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.0982666666666667e-05,
+      "loss": 0.3702,
+      "step": 20290
+    },
+    {
+      "epoch": 0.45103092783505155,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0978222222222224e-05,
+      "loss": 0.4006,
+      "step": 20300
+    },
+    {
+      "epoch": 0.451253110558123,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0973777777777779e-05,
+      "loss": 0.4027,
+      "step": 20310
+    },
+    {
+      "epoch": 0.45147529328119446,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.0969333333333334e-05,
+      "loss": 0.3681,
+      "step": 20320
+    },
+    {
+      "epoch": 0.4516974760042659,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.096488888888889e-05,
+      "loss": 0.3681,
+      "step": 20330
+    },
+    {
+      "epoch": 0.45191965872733736,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.0960444444444445e-05,
+      "loss": 0.3897,
+      "step": 20340
+    },
+    {
+      "epoch": 0.4521418414504088,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.0956e-05,
+      "loss": 0.4075,
+      "step": 20350
+    },
+    {
+      "epoch": 0.45236402417348026,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.0951555555555558e-05,
+      "loss": 0.377,
+      "step": 20360
+    },
+    {
+      "epoch": 0.4525862068965517,
+      "grad_norm": 2.0,
+      "learning_rate": 1.0947111111111113e-05,
+      "loss": 0.372,
+      "step": 20370
+    },
+    {
+      "epoch": 0.45280838961962316,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0942666666666668e-05,
+      "loss": 0.389,
+      "step": 20380
+    },
+    {
+      "epoch": 0.4530305723426946,
+      "grad_norm": 2.125,
+      "learning_rate": 1.0938222222222223e-05,
+      "loss": 0.3631,
+      "step": 20390
+    },
+    {
+      "epoch": 0.45325275506576607,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0933777777777779e-05,
+      "loss": 0.3857,
+      "step": 20400
+    },
+    {
+      "epoch": 0.4534749377888375,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0929333333333334e-05,
+      "loss": 0.3991,
+      "step": 20410
+    },
+    {
+      "epoch": 0.45369712051190897,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0924888888888889e-05,
+      "loss": 0.3523,
+      "step": 20420
+    },
+    {
+      "epoch": 0.4539193032349804,
+      "grad_norm": 2.0,
+      "learning_rate": 1.0920444444444445e-05,
+      "loss": 0.4037,
+      "step": 20430
+    },
+    {
+      "epoch": 0.4541414859580519,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0916e-05,
+      "loss": 0.3844,
+      "step": 20440
+    },
+    {
+      "epoch": 0.4543636686811234,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.0911555555555557e-05,
+      "loss": 0.3743,
+      "step": 20450
+    },
+    {
+      "epoch": 0.45458585140419483,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.0907111111111113e-05,
+      "loss": 0.3594,
+      "step": 20460
+    },
+    {
+      "epoch": 0.4548080341272663,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.0902666666666668e-05,
+      "loss": 0.391,
+      "step": 20470
+    },
+    {
+      "epoch": 0.45503021685033773,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.0898222222222223e-05,
+      "loss": 0.3538,
+      "step": 20480
+    },
+    {
+      "epoch": 0.4552523995734092,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.089377777777778e-05,
+      "loss": 0.398,
+      "step": 20490
+    },
+    {
+      "epoch": 0.45547458229648063,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0889333333333334e-05,
+      "loss": 0.4254,
+      "step": 20500
+    },
+    {
+      "epoch": 0.4556967650195521,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.088488888888889e-05,
+      "loss": 0.4078,
+      "step": 20510
+    },
+    {
+      "epoch": 0.45591894774262354,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.0880444444444444e-05,
+      "loss": 0.3843,
+      "step": 20520
+    },
+    {
+      "epoch": 0.456141130465695,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0876e-05,
+      "loss": 0.4057,
+      "step": 20530
+    },
+    {
+      "epoch": 0.45636331318876644,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0871555555555557e-05,
+      "loss": 0.3866,
+      "step": 20540
+    },
+    {
+      "epoch": 0.4565854959118379,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0867111111111112e-05,
+      "loss": 0.3896,
+      "step": 20550
+    },
+    {
+      "epoch": 0.45680767863490934,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.0862666666666668e-05,
+      "loss": 0.4137,
+      "step": 20560
+    },
+    {
+      "epoch": 0.4570298613579808,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.0858222222222223e-05,
+      "loss": 0.3563,
+      "step": 20570
+    },
+    {
+      "epoch": 0.45725204408105224,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.0853777777777778e-05,
+      "loss": 0.4091,
+      "step": 20580
+    },
+    {
+      "epoch": 0.4574742268041237,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.0849333333333335e-05,
+      "loss": 0.3981,
+      "step": 20590
+    },
+    {
+      "epoch": 0.45769640952719515,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.084488888888889e-05,
+      "loss": 0.4079,
+      "step": 20600
+    },
+    {
+      "epoch": 0.4579185922502666,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.0840444444444444e-05,
+      "loss": 0.3801,
+      "step": 20610
+    },
+    {
+      "epoch": 0.45814077497333805,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.0836e-05,
+      "loss": 0.3811,
+      "step": 20620
+    },
+    {
+      "epoch": 0.4583629576964095,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0831555555555557e-05,
+      "loss": 0.3834,
+      "step": 20630
+    },
+    {
+      "epoch": 0.458585140419481,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.0827111111111112e-05,
+      "loss": 0.394,
+      "step": 20640
+    },
+    {
+      "epoch": 0.45880732314255246,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.0822666666666667e-05,
+      "loss": 0.412,
+      "step": 20650
+    },
+    {
+      "epoch": 0.4590295058656239,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.0818222222222224e-05,
+      "loss": 0.424,
+      "step": 20660
+    },
+    {
+      "epoch": 0.45925168858869536,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0813777777777779e-05,
+      "loss": 0.3918,
+      "step": 20670
+    },
+    {
+      "epoch": 0.4594738713117668,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0809333333333333e-05,
+      "loss": 0.3855,
+      "step": 20680
+    },
+    {
+      "epoch": 0.45969605403483826,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.080488888888889e-05,
+      "loss": 0.3949,
+      "step": 20690
+    },
+    {
+      "epoch": 0.4599182367579097,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.0800444444444445e-05,
+      "loss": 0.4035,
+      "step": 20700
+    },
+    {
+      "epoch": 0.46014041948098117,
+      "grad_norm": 2.125,
+      "learning_rate": 1.0796e-05,
+      "loss": 0.3553,
+      "step": 20710
+    },
+    {
+      "epoch": 0.4603626022040526,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0791555555555558e-05,
+      "loss": 0.4035,
+      "step": 20720
+    },
+    {
+      "epoch": 0.46058478492712407,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.0787111111111113e-05,
+      "loss": 0.3682,
+      "step": 20730
+    },
+    {
+      "epoch": 0.4608069676501955,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0782666666666668e-05,
+      "loss": 0.384,
+      "step": 20740
+    },
+    {
+      "epoch": 0.46102915037326697,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.0778222222222224e-05,
+      "loss": 0.3888,
+      "step": 20750
+    },
+    {
+      "epoch": 0.4612513330963384,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.0773777777777779e-05,
+      "loss": 0.4149,
+      "step": 20760
+    },
+    {
+      "epoch": 0.4614735158194099,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0769333333333334e-05,
+      "loss": 0.3992,
+      "step": 20770
+    },
+    {
+      "epoch": 0.4616956985424813,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.0764888888888889e-05,
+      "loss": 0.4211,
+      "step": 20780
+    },
+    {
+      "epoch": 0.4619178812655528,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.0760444444444445e-05,
+      "loss": 0.3719,
+      "step": 20790
+    },
+    {
+      "epoch": 0.46214006398862423,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.0756e-05,
+      "loss": 0.3809,
+      "step": 20800
+    },
+    {
+      "epoch": 0.4623622467116957,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0751555555555557e-05,
+      "loss": 0.3777,
+      "step": 20810
+    },
+    {
+      "epoch": 0.46258442943476713,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.0747111111111113e-05,
+      "loss": 0.3977,
+      "step": 20820
+    },
+    {
+      "epoch": 0.4628066121578386,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0742666666666668e-05,
+      "loss": 0.3746,
+      "step": 20830
+    },
+    {
+      "epoch": 0.46302879488091003,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.0738222222222223e-05,
+      "loss": 0.4218,
+      "step": 20840
+    },
+    {
+      "epoch": 0.46325097760398154,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.073377777777778e-05,
+      "loss": 0.4456,
+      "step": 20850
+    },
+    {
+      "epoch": 0.463473160327053,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0729333333333334e-05,
+      "loss": 0.3884,
+      "step": 20860
+    },
+    {
+      "epoch": 0.46369534305012444,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.0724888888888889e-05,
+      "loss": 0.417,
+      "step": 20870
+    },
+    {
+      "epoch": 0.4639175257731959,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0720444444444444e-05,
+      "loss": 0.3808,
+      "step": 20880
+    },
+    {
+      "epoch": 0.46413970849626734,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.0716000000000002e-05,
+      "loss": 0.3692,
+      "step": 20890
+    },
+    {
+      "epoch": 0.4643618912193388,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.0711555555555557e-05,
+      "loss": 0.4041,
+      "step": 20900
+    },
+    {
+      "epoch": 0.46458407394241025,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0707111111111112e-05,
+      "loss": 0.3721,
+      "step": 20910
+    },
+    {
+      "epoch": 0.4648062566654817,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0702666666666668e-05,
+      "loss": 0.4174,
+      "step": 20920
+    },
+    {
+      "epoch": 0.46502843938855315,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0698222222222223e-05,
+      "loss": 0.3944,
+      "step": 20930
+    },
+    {
+      "epoch": 0.4652506221116246,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.0693777777777778e-05,
+      "loss": 0.3833,
+      "step": 20940
+    },
+    {
+      "epoch": 0.46547280483469605,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0689333333333335e-05,
+      "loss": 0.4119,
+      "step": 20950
+    },
+    {
+      "epoch": 0.4656949875577675,
+      "grad_norm": 1.75,
+      "learning_rate": 1.068488888888889e-05,
+      "loss": 0.3927,
+      "step": 20960
+    },
+    {
+      "epoch": 0.46591717028083895,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0680444444444444e-05,
+      "loss": 0.394,
+      "step": 20970
+    },
+    {
+      "epoch": 0.4661393530039104,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.0676000000000002e-05,
+      "loss": 0.3588,
+      "step": 20980
+    },
+    {
+      "epoch": 0.46636153572698186,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.0671555555555557e-05,
+      "loss": 0.3979,
+      "step": 20990
+    },
+    {
+      "epoch": 0.4665837184500533,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.0667111111111112e-05,
+      "loss": 0.3983,
+      "step": 21000
+    },
+    {
+      "epoch": 0.46680590117312476,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.0662666666666667e-05,
+      "loss": 0.3962,
+      "step": 21010
+    },
+    {
+      "epoch": 0.4670280838961962,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0658222222222224e-05,
+      "loss": 0.4133,
+      "step": 21020
+    },
+    {
+      "epoch": 0.46725026661926766,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.0653777777777778e-05,
+      "loss": 0.3879,
+      "step": 21030
+    },
+    {
+      "epoch": 0.4674724493423391,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.0649333333333333e-05,
+      "loss": 0.4,
+      "step": 21040
+    },
+    {
+      "epoch": 0.4676946320654106,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.064488888888889e-05,
+      "loss": 0.3913,
+      "step": 21050
+    },
+    {
+      "epoch": 0.46791681478848207,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0640444444444445e-05,
+      "loss": 0.3635,
+      "step": 21060
+    },
+    {
+      "epoch": 0.4681389975115535,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.0636000000000001e-05,
+      "loss": 0.3726,
+      "step": 21070
+    },
+    {
+      "epoch": 0.468361180234625,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.0631555555555558e-05,
+      "loss": 0.3947,
+      "step": 21080
+    },
+    {
+      "epoch": 0.4685833629576964,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.0627111111111113e-05,
+      "loss": 0.3738,
+      "step": 21090
+    },
+    {
+      "epoch": 0.4688055456807679,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.0622666666666667e-05,
+      "loss": 0.4123,
+      "step": 21100
+    },
+    {
+      "epoch": 0.46902772840383933,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.0618222222222224e-05,
+      "loss": 0.4035,
+      "step": 21110
+    },
+    {
+      "epoch": 0.4692499111269108,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0613777777777779e-05,
+      "loss": 0.4067,
+      "step": 21120
+    },
+    {
+      "epoch": 0.46947209384998223,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0609333333333334e-05,
+      "loss": 0.3745,
+      "step": 21130
+    },
+    {
+      "epoch": 0.4696942765730537,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.0604888888888888e-05,
+      "loss": 0.3901,
+      "step": 21140
+    },
+    {
+      "epoch": 0.46991645929612513,
+      "grad_norm": 2.75,
+      "learning_rate": 1.0600444444444445e-05,
+      "loss": 0.4082,
+      "step": 21150
+    },
+    {
+      "epoch": 0.4701386420191966,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.0596000000000002e-05,
+      "loss": 0.3994,
+      "step": 21160
+    },
+    {
+      "epoch": 0.47036082474226804,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0591555555555556e-05,
+      "loss": 0.4238,
+      "step": 21170
+    },
+    {
+      "epoch": 0.4705830074653395,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.0587111111111113e-05,
+      "loss": 0.4065,
+      "step": 21180
+    },
+    {
+      "epoch": 0.47080519018841094,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.0582666666666668e-05,
+      "loss": 0.4038,
+      "step": 21190
+    },
+    {
+      "epoch": 0.4710273729114824,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.0578222222222223e-05,
+      "loss": 0.3967,
+      "step": 21200
+    },
+    {
+      "epoch": 0.47124955563455384,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.057377777777778e-05,
+      "loss": 0.3742,
+      "step": 21210
+    },
+    {
+      "epoch": 0.4714717383576253,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.0569333333333334e-05,
+      "loss": 0.4002,
+      "step": 21220
+    },
+    {
+      "epoch": 0.47169392108069674,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0564888888888889e-05,
+      "loss": 0.3993,
+      "step": 21230
+    },
+    {
+      "epoch": 0.4719161038037682,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.0560444444444444e-05,
+      "loss": 0.4094,
+      "step": 21240
+    },
+    {
+      "epoch": 0.47213828652683965,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0556000000000002e-05,
+      "loss": 0.3779,
+      "step": 21250
+    },
+    {
+      "epoch": 0.47236046924991115,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.0551555555555557e-05,
+      "loss": 0.3729,
+      "step": 21260
+    },
+    {
+      "epoch": 0.4725826519729826,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.0547111111111112e-05,
+      "loss": 0.3719,
+      "step": 21270
+    },
+    {
+      "epoch": 0.47280483469605405,
+      "grad_norm": 2.5,
+      "learning_rate": 1.0542666666666668e-05,
+      "loss": 0.385,
+      "step": 21280
+    },
+    {
+      "epoch": 0.4730270174191255,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.0538222222222223e-05,
+      "loss": 0.3988,
+      "step": 21290
+    },
+    {
+      "epoch": 0.47324920014219696,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.0533777777777778e-05,
+      "loss": 0.3811,
+      "step": 21300
+    },
+    {
+      "epoch": 0.4734713828652684,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.0529333333333334e-05,
+      "loss": 0.4309,
+      "step": 21310
+    },
+    {
+      "epoch": 0.47369356558833986,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.052488888888889e-05,
+      "loss": 0.3763,
+      "step": 21320
+    },
+    {
+      "epoch": 0.4739157483114113,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.0520444444444444e-05,
+      "loss": 0.3886,
+      "step": 21330
+    },
+    {
+      "epoch": 0.47413793103448276,
+      "grad_norm": 3.0,
+      "learning_rate": 1.0516000000000002e-05,
+      "loss": 0.3976,
+      "step": 21340
+    },
+    {
+      "epoch": 0.4743601137575542,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0511555555555557e-05,
+      "loss": 0.3736,
+      "step": 21350
+    },
+    {
+      "epoch": 0.47458229648062566,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0507111111111112e-05,
+      "loss": 0.4161,
+      "step": 21360
+    },
+    {
+      "epoch": 0.4748044792036971,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0502666666666667e-05,
+      "loss": 0.4237,
+      "step": 21370
+    },
+    {
+      "epoch": 0.47502666192676857,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0498222222222223e-05,
+      "loss": 0.3793,
+      "step": 21380
+    },
+    {
+      "epoch": 0.47524884464984,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0493777777777778e-05,
+      "loss": 0.4024,
+      "step": 21390
+    },
+    {
+      "epoch": 0.47547102737291147,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.0489333333333333e-05,
+      "loss": 0.3968,
+      "step": 21400
+    },
+    {
+      "epoch": 0.4756932100959829,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.048488888888889e-05,
+      "loss": 0.402,
+      "step": 21410
+    },
+    {
+      "epoch": 0.4759153928190544,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.0480444444444444e-05,
+      "loss": 0.3966,
+      "step": 21420
+    },
+    {
+      "epoch": 0.4761375755421258,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.0476000000000001e-05,
+      "loss": 0.4159,
+      "step": 21430
+    },
+    {
+      "epoch": 0.4763597582651973,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0471555555555558e-05,
+      "loss": 0.4008,
+      "step": 21440
+    },
+    {
+      "epoch": 0.4765819409882687,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0467111111111112e-05,
+      "loss": 0.3913,
+      "step": 21450
+    },
+    {
+      "epoch": 0.47680412371134023,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0462666666666667e-05,
+      "loss": 0.3975,
+      "step": 21460
+    },
+    {
+      "epoch": 0.4770263064344117,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0458222222222224e-05,
+      "loss": 0.3919,
+      "step": 21470
+    },
+    {
+      "epoch": 0.47724848915748314,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0453777777777779e-05,
+      "loss": 0.4096,
+      "step": 21480
+    },
+    {
+      "epoch": 0.4774706718805546,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0449333333333333e-05,
+      "loss": 0.3979,
+      "step": 21490
+    },
+    {
+      "epoch": 0.47769285460362604,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.0444888888888888e-05,
+      "loss": 0.388,
+      "step": 21500
+    },
+    {
+      "epoch": 0.4779150373266975,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.0440444444444447e-05,
+      "loss": 0.3575,
+      "step": 21510
+    },
+    {
+      "epoch": 0.47813722004976894,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.0436000000000001e-05,
+      "loss": 0.4031,
+      "step": 21520
+    },
+    {
+      "epoch": 0.4783594027728404,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.0431555555555556e-05,
+      "loss": 0.386,
+      "step": 21530
+    },
+    {
+      "epoch": 0.47858158549591184,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.0427111111111113e-05,
+      "loss": 0.3975,
+      "step": 21540
+    },
+    {
+      "epoch": 0.4788037682189833,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0422666666666668e-05,
+      "loss": 0.3538,
+      "step": 21550
+    },
+    {
+      "epoch": 0.47902595094205475,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.0418222222222222e-05,
+      "loss": 0.3753,
+      "step": 21560
+    },
+    {
+      "epoch": 0.4792481336651262,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.0413777777777779e-05,
+      "loss": 0.3741,
+      "step": 21570
+    },
+    {
+      "epoch": 0.47947031638819765,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0409333333333334e-05,
+      "loss": 0.4022,
+      "step": 21580
+    },
+    {
+      "epoch": 0.4796924991112691,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.0404888888888889e-05,
+      "loss": 0.3794,
+      "step": 21590
+    },
+    {
+      "epoch": 0.47991468183434055,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.0400444444444447e-05,
+      "loss": 0.3702,
+      "step": 21600
+    },
+    {
+      "epoch": 0.480136864557412,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0396000000000002e-05,
+      "loss": 0.435,
+      "step": 21610
+    },
+    {
+      "epoch": 0.48035904728048345,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.0391555555555557e-05,
+      "loss": 0.389,
+      "step": 21620
+    },
+    {
+      "epoch": 0.4805812300035549,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.0387111111111111e-05,
+      "loss": 0.3863,
+      "step": 21630
+    },
+    {
+      "epoch": 0.48080341272662636,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.0382666666666668e-05,
+      "loss": 0.3822,
+      "step": 21640
+    },
+    {
+      "epoch": 0.4810255954496978,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.0378222222222223e-05,
+      "loss": 0.4093,
+      "step": 21650
+    },
+    {
+      "epoch": 0.48124777817276926,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0373777777777778e-05,
+      "loss": 0.3749,
+      "step": 21660
+    },
+    {
+      "epoch": 0.48146996089584077,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.0369333333333334e-05,
+      "loss": 0.4006,
+      "step": 21670
+    },
+    {
+      "epoch": 0.4816921436189122,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0364888888888889e-05,
+      "loss": 0.4306,
+      "step": 21680
+    },
+    {
+      "epoch": 0.48191432634198367,
+      "grad_norm": 2.5,
+      "learning_rate": 1.0360444444444446e-05,
+      "loss": 0.38,
+      "step": 21690
+    },
+    {
+      "epoch": 0.4821365090650551,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0356000000000002e-05,
+      "loss": 0.414,
+      "step": 21700
+    },
+    {
+      "epoch": 0.48235869178812657,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.0351555555555557e-05,
+      "loss": 0.403,
+      "step": 21710
+    },
+    {
+      "epoch": 0.482580874511198,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0347111111111112e-05,
+      "loss": 0.3978,
+      "step": 21720
+    },
+    {
+      "epoch": 0.4828030572342695,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0342666666666667e-05,
+      "loss": 0.393,
+      "step": 21730
+    },
+    {
+      "epoch": 0.4830252399573409,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0338222222222223e-05,
+      "loss": 0.3547,
+      "step": 21740
+    },
+    {
+      "epoch": 0.4832474226804124,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.0333777777777778e-05,
+      "loss": 0.4171,
+      "step": 21750
+    },
+    {
+      "epoch": 0.4834696054034838,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.0329333333333333e-05,
+      "loss": 0.417,
+      "step": 21760
+    },
+    {
+      "epoch": 0.4836917881265553,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.032488888888889e-05,
+      "loss": 0.4144,
+      "step": 21770
+    },
+    {
+      "epoch": 0.48391397084962673,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.0320444444444446e-05,
+      "loss": 0.3907,
+      "step": 21780
+    },
+    {
+      "epoch": 0.4841361535726982,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0316e-05,
+      "loss": 0.3828,
+      "step": 21790
+    },
+    {
+      "epoch": 0.48435833629576963,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.0311555555555557e-05,
+      "loss": 0.3694,
+      "step": 21800
+    },
+    {
+      "epoch": 0.4845805190188411,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.0307111111111112e-05,
+      "loss": 0.3973,
+      "step": 21810
+    },
+    {
+      "epoch": 0.48480270174191253,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.0302666666666667e-05,
+      "loss": 0.3438,
+      "step": 21820
+    },
+    {
+      "epoch": 0.485024884464984,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0298222222222224e-05,
+      "loss": 0.3919,
+      "step": 21830
+    },
+    {
+      "epoch": 0.48524706718805544,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.0293777777777778e-05,
+      "loss": 0.3648,
+      "step": 21840
+    },
+    {
+      "epoch": 0.4854692499111269,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0289333333333333e-05,
+      "loss": 0.3749,
+      "step": 21850
+    },
+    {
+      "epoch": 0.48569143263419834,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.0284888888888888e-05,
+      "loss": 0.4041,
+      "step": 21860
+    },
+    {
+      "epoch": 0.48591361535726985,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.0280444444444446e-05,
+      "loss": 0.4086,
+      "step": 21870
+    },
+    {
+      "epoch": 0.4861357980803413,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.0276000000000001e-05,
+      "loss": 0.3983,
+      "step": 21880
+    },
+    {
+      "epoch": 0.48635798080341275,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.0271555555555556e-05,
+      "loss": 0.4025,
+      "step": 21890
+    },
+    {
+      "epoch": 0.4865801635264842,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.0267111111111113e-05,
+      "loss": 0.3899,
+      "step": 21900
+    },
+    {
+      "epoch": 0.48680234624955565,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0262666666666667e-05,
+      "loss": 0.3517,
+      "step": 21910
+    },
+    {
+      "epoch": 0.4870245289726271,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.0258222222222222e-05,
+      "loss": 0.3603,
+      "step": 21920
+    },
+    {
+      "epoch": 0.48724671169569855,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.0253777777777779e-05,
+      "loss": 0.3897,
+      "step": 21930
+    },
+    {
+      "epoch": 0.48746889441877,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0249333333333334e-05,
+      "loss": 0.3824,
+      "step": 21940
+    },
+    {
+      "epoch": 0.48769107714184146,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.0244888888888889e-05,
+      "loss": 0.4117,
+      "step": 21950
+    },
+    {
+      "epoch": 0.4879132598649129,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.0240444444444447e-05,
+      "loss": 0.3726,
+      "step": 21960
+    },
+    {
+      "epoch": 0.48813544258798436,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0236000000000002e-05,
+      "loss": 0.3954,
+      "step": 21970
+    },
+    {
+      "epoch": 0.4883576253110558,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.0231555555555556e-05,
+      "loss": 0.397,
+      "step": 21980
+    },
+    {
+      "epoch": 0.48857980803412726,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.0227111111111111e-05,
+      "loss": 0.4118,
+      "step": 21990
+    },
+    {
+      "epoch": 0.4888019907571987,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0222666666666668e-05,
+      "loss": 0.4002,
+      "step": 22000
+    },
+    {
+      "epoch": 0.48902417348027016,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.0218222222222223e-05,
+      "loss": 0.409,
+      "step": 22010
+    },
+    {
+      "epoch": 0.4892463562033416,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.0213777777777778e-05,
+      "loss": 0.3543,
+      "step": 22020
+    },
+    {
+      "epoch": 0.48946853892641307,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0209333333333334e-05,
+      "loss": 0.4016,
+      "step": 22030
+    },
+    {
+      "epoch": 0.4896907216494845,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.0204888888888889e-05,
+      "loss": 0.3772,
+      "step": 22040
+    },
+    {
+      "epoch": 0.48991290437255597,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0200444444444445e-05,
+      "loss": 0.3841,
+      "step": 22050
+    },
+    {
+      "epoch": 0.4901350870956274,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.0196000000000002e-05,
+      "loss": 0.3945,
+      "step": 22060
+    },
+    {
+      "epoch": 0.49035726981869887,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.0191555555555557e-05,
+      "loss": 0.357,
+      "step": 22070
+    },
+    {
+      "epoch": 0.4905794525417704,
+      "grad_norm": 2.625,
+      "learning_rate": 1.0187111111111112e-05,
+      "loss": 0.4068,
+      "step": 22080
+    },
+    {
+      "epoch": 0.49080163526484183,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0182666666666667e-05,
+      "loss": 0.4001,
+      "step": 22090
+    },
+    {
+      "epoch": 0.4910238179879133,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0178222222222223e-05,
+      "loss": 0.3655,
+      "step": 22100
+    },
+    {
+      "epoch": 0.49124600071098473,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0173777777777778e-05,
+      "loss": 0.3908,
+      "step": 22110
+    },
+    {
+      "epoch": 0.4914681834340562,
+      "grad_norm": 2.625,
+      "learning_rate": 1.0169333333333333e-05,
+      "loss": 0.4094,
+      "step": 22120
+    },
+    {
+      "epoch": 0.49169036615712763,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0164888888888891e-05,
+      "loss": 0.3689,
+      "step": 22130
+    },
+    {
+      "epoch": 0.4919125488801991,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.0160444444444446e-05,
+      "loss": 0.382,
+      "step": 22140
+    },
+    {
+      "epoch": 0.49213473160327054,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0156e-05,
+      "loss": 0.3878,
+      "step": 22150
+    },
+    {
+      "epoch": 0.492356914326342,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.0151555555555557e-05,
+      "loss": 0.4285,
+      "step": 22160
+    },
+    {
+      "epoch": 0.49257909704941344,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.0147111111111112e-05,
+      "loss": 0.356,
+      "step": 22170
+    },
+    {
+      "epoch": 0.4928012797724849,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0142666666666667e-05,
+      "loss": 0.3692,
+      "step": 22180
+    },
+    {
+      "epoch": 0.49302346249555634,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.0138222222222223e-05,
+      "loss": 0.3603,
+      "step": 22190
+    },
+    {
+      "epoch": 0.4932456452186278,
+      "grad_norm": 2.125,
+      "learning_rate": 1.0133777777777778e-05,
+      "loss": 0.409,
+      "step": 22200
+    },
+    {
+      "epoch": 0.49346782794169924,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0129333333333333e-05,
+      "loss": 0.3898,
+      "step": 22210
+    },
+    {
+      "epoch": 0.4936900106647707,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.0124888888888891e-05,
+      "loss": 0.4,
+      "step": 22220
+    },
+    {
+      "epoch": 0.49391219338784215,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.0120444444444446e-05,
+      "loss": 0.3752,
+      "step": 22230
+    },
+    {
+      "epoch": 0.4941343761109136,
+      "grad_norm": 2.625,
+      "learning_rate": 1.0116000000000001e-05,
+      "loss": 0.4245,
+      "step": 22240
+    },
+    {
+      "epoch": 0.49435655883398505,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.0111555555555556e-05,
+      "loss": 0.4112,
+      "step": 22250
+    },
+    {
+      "epoch": 0.4945787415570565,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.0107111111111112e-05,
+      "loss": 0.3718,
+      "step": 22260
+    },
+    {
+      "epoch": 0.49480092428012795,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0102666666666667e-05,
+      "loss": 0.3746,
+      "step": 22270
+    },
+    {
+      "epoch": 0.49502310700319946,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0098222222222222e-05,
+      "loss": 0.3785,
+      "step": 22280
+    },
+    {
+      "epoch": 0.4952452897262709,
+      "grad_norm": 2.0,
+      "learning_rate": 1.0093777777777779e-05,
+      "loss": 0.3779,
+      "step": 22290
+    },
+    {
+      "epoch": 0.49546747244934236,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0089333333333334e-05,
+      "loss": 0.3781,
+      "step": 22300
+    },
+    {
+      "epoch": 0.4956896551724138,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.008488888888889e-05,
+      "loss": 0.3549,
+      "step": 22310
+    },
+    {
+      "epoch": 0.49591183789548526,
+      "grad_norm": 3.125,
+      "learning_rate": 1.0080444444444447e-05,
+      "loss": 0.417,
+      "step": 22320
+    },
+    {
+      "epoch": 0.4961340206185567,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0076000000000001e-05,
+      "loss": 0.3762,
+      "step": 22330
+    },
+    {
+      "epoch": 0.49635620334162817,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0071555555555556e-05,
+      "loss": 0.4215,
+      "step": 22340
+    },
+    {
+      "epoch": 0.4965783860646996,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.0067111111111111e-05,
+      "loss": 0.4219,
+      "step": 22350
+    },
+    {
+      "epoch": 0.49680056878777107,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.0062666666666668e-05,
+      "loss": 0.3871,
+      "step": 22360
+    },
+    {
+      "epoch": 0.4970227515108425,
+      "grad_norm": 2.0,
+      "learning_rate": 1.0058222222222223e-05,
+      "loss": 0.3559,
+      "step": 22370
+    },
+    {
+      "epoch": 0.49724493423391397,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.0053777777777777e-05,
+      "loss": 0.3805,
+      "step": 22380
+    },
+    {
+      "epoch": 0.4974671169569854,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.0049333333333334e-05,
+      "loss": 0.4216,
+      "step": 22390
+    },
+    {
+      "epoch": 0.4976892996800569,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.004488888888889e-05,
+      "loss": 0.3915,
+      "step": 22400
+    },
+    {
+      "epoch": 0.4979114824031283,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.0040444444444445e-05,
+      "loss": 0.3722,
+      "step": 22410
+    },
+    {
+      "epoch": 0.4981336651261998,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.0036000000000002e-05,
+      "loss": 0.4061,
+      "step": 22420
+    },
+    {
+      "epoch": 0.4983558478492712,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.0031555555555557e-05,
+      "loss": 0.4162,
+      "step": 22430
+    },
+    {
+      "epoch": 0.4985780305723427,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0027111111111112e-05,
+      "loss": 0.4154,
+      "step": 22440
+    },
+    {
+      "epoch": 0.49880021329541413,
+      "grad_norm": 2.625,
+      "learning_rate": 1.0022666666666666e-05,
+      "loss": 0.3969,
+      "step": 22450
+    },
+    {
+      "epoch": 0.4990223960184856,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0018222222222223e-05,
+      "loss": 0.3987,
+      "step": 22460
+    },
+    {
+      "epoch": 0.49924457874155703,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.0013777777777778e-05,
+      "loss": 0.387,
+      "step": 22470
+    },
+    {
+      "epoch": 0.4994667614646285,
+      "grad_norm": 2.875,
+      "learning_rate": 1.0009333333333333e-05,
+      "loss": 0.3965,
+      "step": 22480
+    },
+    {
+      "epoch": 0.4996889441877,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.000488888888889e-05,
+      "loss": 0.4201,
+      "step": 22490
+    },
+    {
+      "epoch": 0.49991112691077144,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0000444444444446e-05,
+      "loss": 0.407,
+      "step": 22500
+    },
+    {
+      "epoch": 0.5001333096338428,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.996e-06,
+      "loss": 0.4298,
+      "step": 22510
+    },
+    {
+      "epoch": 0.5003554923569143,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.991555555555557e-06,
+      "loss": 0.3779,
+      "step": 22520
+    },
+    {
+      "epoch": 0.5005776750799857,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.987111111111112e-06,
+      "loss": 0.3997,
+      "step": 22530
+    },
+    {
+      "epoch": 0.5007998578030572,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.982666666666667e-06,
+      "loss": 0.3894,
+      "step": 22540
+    },
+    {
+      "epoch": 0.5010220405261286,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.978222222222223e-06,
+      "loss": 0.4136,
+      "step": 22550
+    },
+    {
+      "epoch": 0.5012442232492001,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.973777777777778e-06,
+      "loss": 0.3812,
+      "step": 22560
+    },
+    {
+      "epoch": 0.5014664059722715,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.969333333333335e-06,
+      "loss": 0.4132,
+      "step": 22570
+    },
+    {
+      "epoch": 0.501688588695343,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.96488888888889e-06,
+      "loss": 0.3851,
+      "step": 22580
+    },
+    {
+      "epoch": 0.5019107714184146,
+      "grad_norm": 2.125,
+      "learning_rate": 9.960444444444444e-06,
+      "loss": 0.3617,
+      "step": 22590
+    },
+    {
+      "epoch": 0.502132954141486,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.956000000000001e-06,
+      "loss": 0.3363,
+      "step": 22600
+    },
+    {
+      "epoch": 0.5023551368645575,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.951555555555556e-06,
+      "loss": 0.3396,
+      "step": 22610
+    },
+    {
+      "epoch": 0.5025773195876289,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.947111111111112e-06,
+      "loss": 0.3656,
+      "step": 22620
+    },
+    {
+      "epoch": 0.5027995023107004,
+      "grad_norm": 2.0,
+      "learning_rate": 9.942666666666667e-06,
+      "loss": 0.4024,
+      "step": 22630
+    },
+    {
+      "epoch": 0.5030216850337718,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.938222222222224e-06,
+      "loss": 0.4049,
+      "step": 22640
+    },
+    {
+      "epoch": 0.5032438677568433,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.933777777777779e-06,
+      "loss": 0.4076,
+      "step": 22650
+    },
+    {
+      "epoch": 0.5034660504799147,
+      "grad_norm": 2.5,
+      "learning_rate": 9.929333333333333e-06,
+      "loss": 0.3608,
+      "step": 22660
+    },
+    {
+      "epoch": 0.5036882332029862,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.92488888888889e-06,
+      "loss": 0.4079,
+      "step": 22670
+    },
+    {
+      "epoch": 0.5039104159260576,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.920444444444445e-06,
+      "loss": 0.4066,
+      "step": 22680
+    },
+    {
+      "epoch": 0.5041325986491291,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.916000000000001e-06,
+      "loss": 0.4185,
+      "step": 22690
+    },
+    {
+      "epoch": 0.5043547813722005,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.911555555555556e-06,
+      "loss": 0.3507,
+      "step": 22700
+    },
+    {
+      "epoch": 0.504576964095272,
+      "grad_norm": 2.125,
+      "learning_rate": 9.907111111111111e-06,
+      "loss": 0.3864,
+      "step": 22710
+    },
+    {
+      "epoch": 0.5047991468183434,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.902666666666668e-06,
+      "loss": 0.4213,
+      "step": 22720
+    },
+    {
+      "epoch": 0.5050213295414149,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.898222222222224e-06,
+      "loss": 0.3925,
+      "step": 22730
+    },
+    {
+      "epoch": 0.5052435122644863,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.893777777777779e-06,
+      "loss": 0.4311,
+      "step": 22740
+    },
+    {
+      "epoch": 0.5054656949875578,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.889333333333334e-06,
+      "loss": 0.3781,
+      "step": 22750
+    },
+    {
+      "epoch": 0.5056878777106292,
+      "grad_norm": 2.25,
+      "learning_rate": 9.884888888888889e-06,
+      "loss": 0.3717,
+      "step": 22760
+    },
+    {
+      "epoch": 0.5059100604337007,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.880444444444445e-06,
+      "loss": 0.3664,
+      "step": 22770
+    },
+    {
+      "epoch": 0.5061322431567721,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.876000000000002e-06,
+      "loss": 0.3805,
+      "step": 22780
+    },
+    {
+      "epoch": 0.5063544258798436,
+      "grad_norm": 2.75,
+      "learning_rate": 9.871555555555557e-06,
+      "loss": 0.4212,
+      "step": 22790
+    },
+    {
+      "epoch": 0.5065766086029151,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.867111111111111e-06,
+      "loss": 0.4189,
+      "step": 22800
+    },
+    {
+      "epoch": 0.5067987913259865,
+      "grad_norm": 2.0,
+      "learning_rate": 9.862666666666668e-06,
+      "loss": 0.3924,
+      "step": 22810
+    },
+    {
+      "epoch": 0.507020974049058,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.858222222222223e-06,
+      "loss": 0.3742,
+      "step": 22820
+    },
+    {
+      "epoch": 0.5072431567721294,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.85377777777778e-06,
+      "loss": 0.3877,
+      "step": 22830
+    },
+    {
+      "epoch": 0.5074653394952009,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.849333333333334e-06,
+      "loss": 0.4126,
+      "step": 22840
+    },
+    {
+      "epoch": 0.5076875222182723,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.844888888888889e-06,
+      "loss": 0.3659,
+      "step": 22850
+    },
+    {
+      "epoch": 0.5079097049413438,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.840444444444446e-06,
+      "loss": 0.4144,
+      "step": 22860
+    },
+    {
+      "epoch": 0.5081318876644152,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.836e-06,
+      "loss": 0.3972,
+      "step": 22870
+    },
+    {
+      "epoch": 0.5083540703874867,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.831555555555557e-06,
+      "loss": 0.364,
+      "step": 22880
+    },
+    {
+      "epoch": 0.5085762531105581,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.827111111111112e-06,
+      "loss": 0.4129,
+      "step": 22890
+    },
+    {
+      "epoch": 0.5087984358336296,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.822666666666667e-06,
+      "loss": 0.4082,
+      "step": 22900
+    },
+    {
+      "epoch": 0.509020618556701,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.818222222222223e-06,
+      "loss": 0.4159,
+      "step": 22910
+    },
+    {
+      "epoch": 0.5092428012797725,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.813777777777778e-06,
+      "loss": 0.3992,
+      "step": 22920
+    },
+    {
+      "epoch": 0.5094649840028439,
+      "grad_norm": 2.75,
+      "learning_rate": 9.809333333333335e-06,
+      "loss": 0.3495,
+      "step": 22930
+    },
+    {
+      "epoch": 0.5096871667259154,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.80488888888889e-06,
+      "loss": 0.376,
+      "step": 22940
+    },
+    {
+      "epoch": 0.5099093494489868,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.800444444444446e-06,
+      "loss": 0.3861,
+      "step": 22950
+    },
+    {
+      "epoch": 0.5101315321720583,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.796e-06,
+      "loss": 0.4001,
+      "step": 22960
+    },
+    {
+      "epoch": 0.5103537148951297,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.791555555555556e-06,
+      "loss": 0.4003,
+      "step": 22970
+    },
+    {
+      "epoch": 0.5105758976182012,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.787111111111112e-06,
+      "loss": 0.3659,
+      "step": 22980
+    },
+    {
+      "epoch": 0.5107980803412726,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.782666666666667e-06,
+      "loss": 0.3836,
+      "step": 22990
+    },
+    {
+      "epoch": 0.5110202630643441,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.778222222222224e-06,
+      "loss": 0.3947,
+      "step": 23000
+    },
+    {
+      "epoch": 0.5112424457874156,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.773777777777778e-06,
+      "loss": 0.3722,
+      "step": 23010
+    },
+    {
+      "epoch": 0.511464628510487,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.769333333333333e-06,
+      "loss": 0.384,
+      "step": 23020
+    },
+    {
+      "epoch": 0.5116868112335585,
+      "grad_norm": 2.375,
+      "learning_rate": 9.76488888888889e-06,
+      "loss": 0.3549,
+      "step": 23030
+    },
+    {
+      "epoch": 0.5119089939566299,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.760444444444446e-06,
+      "loss": 0.395,
+      "step": 23040
+    },
+    {
+      "epoch": 0.5121311766797014,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.756000000000001e-06,
+      "loss": 0.34,
+      "step": 23050
+    },
+    {
+      "epoch": 0.5123533594027728,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.751555555555556e-06,
+      "loss": 0.3952,
+      "step": 23060
+    },
+    {
+      "epoch": 0.5125755421258443,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.74711111111111e-06,
+      "loss": 0.4082,
+      "step": 23070
+    },
+    {
+      "epoch": 0.5127977248489157,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.742666666666667e-06,
+      "loss": 0.3826,
+      "step": 23080
+    },
+    {
+      "epoch": 0.5130199075719872,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.738222222222224e-06,
+      "loss": 0.3905,
+      "step": 23090
+    },
+    {
+      "epoch": 0.5132420902950586,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.733777777777779e-06,
+      "loss": 0.3996,
+      "step": 23100
+    },
+    {
+      "epoch": 0.5134642730181301,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.729333333333334e-06,
+      "loss": 0.3944,
+      "step": 23110
+    },
+    {
+      "epoch": 0.5136864557412015,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.724888888888888e-06,
+      "loss": 0.3675,
+      "step": 23120
+    },
+    {
+      "epoch": 0.513908638464273,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.720444444444445e-06,
+      "loss": 0.4111,
+      "step": 23130
+    },
+    {
+      "epoch": 0.5141308211873444,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.716000000000002e-06,
+      "loss": 0.4227,
+      "step": 23140
+    },
+    {
+      "epoch": 0.5143530039104159,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.711555555555556e-06,
+      "loss": 0.396,
+      "step": 23150
+    },
+    {
+      "epoch": 0.5145751866334873,
+      "grad_norm": 2.625,
+      "learning_rate": 9.707111111111111e-06,
+      "loss": 0.3875,
+      "step": 23160
+    },
+    {
+      "epoch": 0.5147973693565588,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.702666666666668e-06,
+      "loss": 0.387,
+      "step": 23170
+    },
+    {
+      "epoch": 0.5150195520796302,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.698222222222223e-06,
+      "loss": 0.3835,
+      "step": 23180
+    },
+    {
+      "epoch": 0.5152417348027017,
+      "grad_norm": 2.5,
+      "learning_rate": 9.693777777777779e-06,
+      "loss": 0.3879,
+      "step": 23190
+    },
+    {
+      "epoch": 0.5154639175257731,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.689333333333334e-06,
+      "loss": 0.3927,
+      "step": 23200
+    },
+    {
+      "epoch": 0.5156861002488446,
+      "grad_norm": 2.625,
+      "learning_rate": 9.684888888888889e-06,
+      "loss": 0.3956,
+      "step": 23210
+    },
+    {
+      "epoch": 0.5159082829719162,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.680444444444445e-06,
+      "loss": 0.3682,
+      "step": 23220
+    },
+    {
+      "epoch": 0.5161304656949876,
+      "grad_norm": 2.5,
+      "learning_rate": 9.676e-06,
+      "loss": 0.4165,
+      "step": 23230
+    },
+    {
+      "epoch": 0.5163526484180591,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.671555555555557e-06,
+      "loss": 0.3805,
+      "step": 23240
+    },
+    {
+      "epoch": 0.5165748311411305,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.667111111111112e-06,
+      "loss": 0.3986,
+      "step": 23250
+    },
+    {
+      "epoch": 0.516797013864202,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.662666666666668e-06,
+      "loss": 0.3789,
+      "step": 23260
+    },
+    {
+      "epoch": 0.5170191965872734,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.658222222222223e-06,
+      "loss": 0.4063,
+      "step": 23270
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.653777777777778e-06,
+      "loss": 0.4068,
+      "step": 23280
+    },
+    {
+      "epoch": 0.5174635620334163,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.649333333333334e-06,
+      "loss": 0.3881,
+      "step": 23290
+    },
+    {
+      "epoch": 0.5176857447564878,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.64488888888889e-06,
+      "loss": 0.403,
+      "step": 23300
+    },
+    {
+      "epoch": 0.5179079274795592,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.640444444444446e-06,
+      "loss": 0.411,
+      "step": 23310
+    },
+    {
+      "epoch": 0.5181301102026307,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.636e-06,
+      "loss": 0.411,
+      "step": 23320
+    },
+    {
+      "epoch": 0.5183522929257021,
+      "grad_norm": 2.875,
+      "learning_rate": 9.631555555555555e-06,
+      "loss": 0.3763,
+      "step": 23330
+    },
+    {
+      "epoch": 0.5185744756487736,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.627111111111112e-06,
+      "loss": 0.4007,
+      "step": 23340
+    },
+    {
+      "epoch": 0.518796658371845,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.622666666666668e-06,
+      "loss": 0.3983,
+      "step": 23350
+    },
+    {
+      "epoch": 0.5190188410949165,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.618222222222223e-06,
+      "loss": 0.4572,
+      "step": 23360
+    },
+    {
+      "epoch": 0.5192410238179879,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.613777777777778e-06,
+      "loss": 0.3783,
+      "step": 23370
+    },
+    {
+      "epoch": 0.5194632065410594,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.609333333333333e-06,
+      "loss": 0.3842,
+      "step": 23380
+    },
+    {
+      "epoch": 0.5196853892641308,
+      "grad_norm": 2.5,
+      "learning_rate": 9.60488888888889e-06,
+      "loss": 0.3901,
+      "step": 23390
+    },
+    {
+      "epoch": 0.5199075719872023,
+      "grad_norm": 2.5,
+      "learning_rate": 9.600444444444446e-06,
+      "loss": 0.4148,
+      "step": 23400
+    },
+    {
+      "epoch": 0.5201297547102738,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.596000000000001e-06,
+      "loss": 0.3899,
+      "step": 23410
+    },
+    {
+      "epoch": 0.5203519374333452,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.591555555555556e-06,
+      "loss": 0.4579,
+      "step": 23420
+    },
+    {
+      "epoch": 0.5205741201564167,
+      "grad_norm": 2.75,
+      "learning_rate": 9.58711111111111e-06,
+      "loss": 0.3828,
+      "step": 23430
+    },
+    {
+      "epoch": 0.5207963028794881,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.582666666666667e-06,
+      "loss": 0.3632,
+      "step": 23440
+    },
+    {
+      "epoch": 0.5210184856025596,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.578222222222224e-06,
+      "loss": 0.4083,
+      "step": 23450
+    },
+    {
+      "epoch": 0.521240668325631,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.573777777777779e-06,
+      "loss": 0.37,
+      "step": 23460
+    },
+    {
+      "epoch": 0.5214628510487025,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.569333333333333e-06,
+      "loss": 0.3641,
+      "step": 23470
+    },
+    {
+      "epoch": 0.5216850337717739,
+      "grad_norm": 1.9140625,
+      "learning_rate": 9.56488888888889e-06,
+      "loss": 0.3725,
+      "step": 23480
+    },
+    {
+      "epoch": 0.5219072164948454,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.560444444444445e-06,
+      "loss": 0.3971,
+      "step": 23490
+    },
+    {
+      "epoch": 0.5221293992179168,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.556000000000001e-06,
+      "loss": 0.3923,
+      "step": 23500
+    },
+    {
+      "epoch": 0.5223515819409883,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.551555555555556e-06,
+      "loss": 0.3926,
+      "step": 23510
+    },
+    {
+      "epoch": 0.5225737646640597,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.547111111111111e-06,
+      "loss": 0.3701,
+      "step": 23520
+    },
+    {
+      "epoch": 0.5227959473871312,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.542666666666668e-06,
+      "loss": 0.3761,
+      "step": 23530
+    },
+    {
+      "epoch": 0.5230181301102026,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.538222222222222e-06,
+      "loss": 0.3749,
+      "step": 23540
+    },
+    {
+      "epoch": 0.5232403128332741,
+      "grad_norm": 2.0,
+      "learning_rate": 9.533777777777779e-06,
+      "loss": 0.401,
+      "step": 23550
+    },
+    {
+      "epoch": 0.5234624955563455,
+      "grad_norm": 3.125,
+      "learning_rate": 9.529333333333334e-06,
+      "loss": 0.3932,
+      "step": 23560
+    },
+    {
+      "epoch": 0.523684678279417,
+      "grad_norm": 2.5,
+      "learning_rate": 9.52488888888889e-06,
+      "loss": 0.409,
+      "step": 23570
+    },
+    {
+      "epoch": 0.5239068610024884,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.520444444444445e-06,
+      "loss": 0.3663,
+      "step": 23580
+    },
+    {
+      "epoch": 0.5241290437255599,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.516e-06,
+      "loss": 0.4164,
+      "step": 23590
+    },
+    {
+      "epoch": 0.5243512264486313,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.511555555555557e-06,
+      "loss": 0.3705,
+      "step": 23600
+    },
+    {
+      "epoch": 0.5245734091717028,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.507111111111111e-06,
+      "loss": 0.3507,
+      "step": 23610
+    },
+    {
+      "epoch": 0.5247955918947743,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.502666666666668e-06,
+      "loss": 0.4186,
+      "step": 23620
+    },
+    {
+      "epoch": 0.5250177746178457,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.498222222222223e-06,
+      "loss": 0.4051,
+      "step": 23630
+    },
+    {
+      "epoch": 0.5252399573409172,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.493777777777778e-06,
+      "loss": 0.4121,
+      "step": 23640
+    },
+    {
+      "epoch": 0.5254621400639886,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.489333333333334e-06,
+      "loss": 0.372,
+      "step": 23650
+    },
+    {
+      "epoch": 0.5256843227870601,
+      "grad_norm": 2.875,
+      "learning_rate": 9.48488888888889e-06,
+      "loss": 0.4198,
+      "step": 23660
+    },
+    {
+      "epoch": 0.5259065055101315,
+      "grad_norm": 2.5,
+      "learning_rate": 9.480444444444446e-06,
+      "loss": 0.368,
+      "step": 23670
+    },
+    {
+      "epoch": 0.526128688233203,
+      "grad_norm": 2.625,
+      "learning_rate": 9.476e-06,
+      "loss": 0.3731,
+      "step": 23680
+    },
+    {
+      "epoch": 0.5263508709562744,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.471555555555555e-06,
+      "loss": 0.3682,
+      "step": 23690
+    },
+    {
+      "epoch": 0.5265730536793459,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.467111111111112e-06,
+      "loss": 0.3844,
+      "step": 23700
+    },
+    {
+      "epoch": 0.5267952364024173,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.462666666666668e-06,
+      "loss": 0.3779,
+      "step": 23710
+    },
+    {
+      "epoch": 0.5270174191254888,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.458222222222223e-06,
+      "loss": 0.3784,
+      "step": 23720
+    },
+    {
+      "epoch": 0.5272396018485602,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.453777777777778e-06,
+      "loss": 0.3681,
+      "step": 23730
+    },
+    {
+      "epoch": 0.5274617845716317,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.449333333333333e-06,
+      "loss": 0.3966,
+      "step": 23740
+    },
+    {
+      "epoch": 0.5276839672947031,
+      "grad_norm": 2.5,
+      "learning_rate": 9.44488888888889e-06,
+      "loss": 0.36,
+      "step": 23750
+    },
+    {
+      "epoch": 0.5279061500177746,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.440444444444446e-06,
+      "loss": 0.3675,
+      "step": 23760
+    },
+    {
+      "epoch": 0.528128332740846,
+      "grad_norm": 2.375,
+      "learning_rate": 9.436e-06,
+      "loss": 0.3679,
+      "step": 23770
+    },
+    {
+      "epoch": 0.5283505154639175,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.431555555555556e-06,
+      "loss": 0.3862,
+      "step": 23780
+    },
+    {
+      "epoch": 0.5285726981869889,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.427111111111112e-06,
+      "loss": 0.365,
+      "step": 23790
+    },
+    {
+      "epoch": 0.5287948809100604,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.422666666666667e-06,
+      "loss": 0.3974,
+      "step": 23800
+    },
+    {
+      "epoch": 0.5290170636331318,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.418222222222224e-06,
+      "loss": 0.35,
+      "step": 23810
+    },
+    {
+      "epoch": 0.5292392463562033,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.413777777777778e-06,
+      "loss": 0.423,
+      "step": 23820
+    },
+    {
+      "epoch": 0.5294614290792748,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.409333333333333e-06,
+      "loss": 0.3607,
+      "step": 23830
+    },
+    {
+      "epoch": 0.5296836118023462,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.40488888888889e-06,
+      "loss": 0.3973,
+      "step": 23840
+    },
+    {
+      "epoch": 0.5299057945254178,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.400444444444445e-06,
+      "loss": 0.4067,
+      "step": 23850
+    },
+    {
+      "epoch": 0.5301279772484891,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.396000000000001e-06,
+      "loss": 0.4117,
+      "step": 23860
+    },
+    {
+      "epoch": 0.5303501599715607,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.391555555555556e-06,
+      "loss": 0.3517,
+      "step": 23870
+    },
+    {
+      "epoch": 0.530572342694632,
+      "grad_norm": 2.375,
+      "learning_rate": 9.387111111111113e-06,
+      "loss": 0.3877,
+      "step": 23880
+    },
+    {
+      "epoch": 0.5307945254177036,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.382666666666667e-06,
+      "loss": 0.4231,
+      "step": 23890
+    },
+    {
+      "epoch": 0.531016708140775,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.378222222222222e-06,
+      "loss": 0.3845,
+      "step": 23900
+    },
+    {
+      "epoch": 0.5312388908638465,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.373777777777779e-06,
+      "loss": 0.436,
+      "step": 23910
+    },
+    {
+      "epoch": 0.5314610735869179,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.369333333333334e-06,
+      "loss": 0.3605,
+      "step": 23920
+    },
+    {
+      "epoch": 0.5316832563099894,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.36488888888889e-06,
+      "loss": 0.3881,
+      "step": 23930
+    },
+    {
+      "epoch": 0.5319054390330608,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.360444444444445e-06,
+      "loss": 0.4014,
+      "step": 23940
+    },
+    {
+      "epoch": 0.5321276217561323,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.356e-06,
+      "loss": 0.3885,
+      "step": 23950
+    },
+    {
+      "epoch": 0.5323498044792037,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.351555555555556e-06,
+      "loss": 0.3429,
+      "step": 23960
+    },
+    {
+      "epoch": 0.5325719872022752,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.347111111111113e-06,
+      "loss": 0.4116,
+      "step": 23970
+    },
+    {
+      "epoch": 0.5327941699253466,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.342666666666668e-06,
+      "loss": 0.4032,
+      "step": 23980
+    },
+    {
+      "epoch": 0.5330163526484181,
+      "grad_norm": 1.875,
+      "learning_rate": 9.338222222222223e-06,
+      "loss": 0.3314,
+      "step": 23990
+    },
+    {
+      "epoch": 0.5332385353714895,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.333777777777777e-06,
+      "loss": 0.3691,
+      "step": 24000
+    },
+    {
+      "epoch": 0.533460718094561,
+      "grad_norm": 2.125,
+      "learning_rate": 9.329333333333334e-06,
+      "loss": 0.3932,
+      "step": 24010
+    },
+    {
+      "epoch": 0.5336829008176324,
+      "grad_norm": 2.25,
+      "learning_rate": 9.32488888888889e-06,
+      "loss": 0.3472,
+      "step": 24020
+    },
+    {
+      "epoch": 0.5339050835407039,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.320444444444445e-06,
+      "loss": 0.4251,
+      "step": 24030
+    },
+    {
+      "epoch": 0.5341272662637754,
+      "grad_norm": 2.375,
+      "learning_rate": 9.316e-06,
+      "loss": 0.3884,
+      "step": 24040
+    },
+    {
+      "epoch": 0.5343494489868468,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.311555555555555e-06,
+      "loss": 0.3509,
+      "step": 24050
+    },
+    {
+      "epoch": 0.5345716317099183,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.307111111111112e-06,
+      "loss": 0.3587,
+      "step": 24060
+    },
+    {
+      "epoch": 0.5347938144329897,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.302666666666668e-06,
+      "loss": 0.4026,
+      "step": 24070
+    },
+    {
+      "epoch": 0.5350159971560612,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.298222222222223e-06,
+      "loss": 0.3798,
+      "step": 24080
+    },
+    {
+      "epoch": 0.5352381798791326,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.293777777777778e-06,
+      "loss": 0.3895,
+      "step": 24090
+    },
+    {
+      "epoch": 0.5354603626022041,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.289333333333334e-06,
+      "loss": 0.3804,
+      "step": 24100
+    },
+    {
+      "epoch": 0.5356825453252755,
+      "grad_norm": 2.125,
+      "learning_rate": 9.28488888888889e-06,
+      "loss": 0.3708,
+      "step": 24110
+    },
+    {
+      "epoch": 0.535904728048347,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.280444444444446e-06,
+      "loss": 0.3976,
+      "step": 24120
+    },
+    {
+      "epoch": 0.5361269107714184,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.276e-06,
+      "loss": 0.3709,
+      "step": 24130
+    },
+    {
+      "epoch": 0.5363490934944899,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.271555555555555e-06,
+      "loss": 0.3845,
+      "step": 24140
+    },
+    {
+      "epoch": 0.5365712762175613,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.267111111111112e-06,
+      "loss": 0.3443,
+      "step": 24150
+    },
+    {
+      "epoch": 0.5367934589406328,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.262666666666667e-06,
+      "loss": 0.3965,
+      "step": 24160
+    },
+    {
+      "epoch": 0.5370156416637042,
+      "grad_norm": 2.5,
+      "learning_rate": 9.258222222222223e-06,
+      "loss": 0.3619,
+      "step": 24170
+    },
+    {
+      "epoch": 0.5372378243867757,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.253777777777778e-06,
+      "loss": 0.4331,
+      "step": 24180
+    },
+    {
+      "epoch": 0.5374600071098471,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.249333333333335e-06,
+      "loss": 0.3977,
+      "step": 24190
+    },
+    {
+      "epoch": 0.5376821898329186,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.24488888888889e-06,
+      "loss": 0.404,
+      "step": 24200
+    },
+    {
+      "epoch": 0.53790437255599,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.240444444444444e-06,
+      "loss": 0.4076,
+      "step": 24210
+    },
+    {
+      "epoch": 0.5381265552790615,
+      "grad_norm": 2.5,
+      "learning_rate": 9.236000000000001e-06,
+      "loss": 0.4087,
+      "step": 24220
+    },
+    {
+      "epoch": 0.538348738002133,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.231555555555556e-06,
+      "loss": 0.4065,
+      "step": 24230
+    },
+    {
+      "epoch": 0.5385709207252044,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.227111111111112e-06,
+      "loss": 0.4035,
+      "step": 24240
+    },
+    {
+      "epoch": 0.5387931034482759,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.222666666666667e-06,
+      "loss": 0.3972,
+      "step": 24250
+    },
+    {
+      "epoch": 0.5390152861713473,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.218222222222222e-06,
+      "loss": 0.4168,
+      "step": 24260
+    },
+    {
+      "epoch": 0.5392374688944188,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.213777777777779e-06,
+      "loss": 0.3985,
+      "step": 24270
+    },
+    {
+      "epoch": 0.5394596516174902,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.209333333333335e-06,
+      "loss": 0.3861,
+      "step": 24280
+    },
+    {
+      "epoch": 0.5396818343405617,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.20488888888889e-06,
+      "loss": 0.3952,
+      "step": 24290
+    },
+    {
+      "epoch": 0.5399040170636331,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.200444444444445e-06,
+      "loss": 0.3793,
+      "step": 24300
+    },
+    {
+      "epoch": 0.5401261997867046,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.196e-06,
+      "loss": 0.3831,
+      "step": 24310
+    },
+    {
+      "epoch": 0.540348382509776,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.191555555555556e-06,
+      "loss": 0.3802,
+      "step": 24320
+    },
+    {
+      "epoch": 0.5405705652328475,
+      "grad_norm": 2.5,
+      "learning_rate": 9.187111111111113e-06,
+      "loss": 0.4176,
+      "step": 24330
+    },
+    {
+      "epoch": 0.5407927479559189,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.182666666666668e-06,
+      "loss": 0.342,
+      "step": 24340
+    },
+    {
+      "epoch": 0.5410149306789904,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.178222222222222e-06,
+      "loss": 0.3505,
+      "step": 24350
+    },
+    {
+      "epoch": 0.5412371134020618,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.173777777777777e-06,
+      "loss": 0.4308,
+      "step": 24360
+    },
+    {
+      "epoch": 0.5414592961251333,
+      "grad_norm": 2.375,
+      "learning_rate": 9.169333333333334e-06,
+      "loss": 0.3593,
+      "step": 24370
+    },
+    {
+      "epoch": 0.5416814788482047,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.16488888888889e-06,
+      "loss": 0.3972,
+      "step": 24380
+    },
+    {
+      "epoch": 0.5419036615712762,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.160444444444445e-06,
+      "loss": 0.3626,
+      "step": 24390
+    },
+    {
+      "epoch": 0.5421258442943476,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.156e-06,
+      "loss": 0.3998,
+      "step": 24400
+    },
+    {
+      "epoch": 0.5423480270174191,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.151555555555557e-06,
+      "loss": 0.4031,
+      "step": 24410
+    },
+    {
+      "epoch": 0.5425702097404905,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.147111111111111e-06,
+      "loss": 0.3568,
+      "step": 24420
+    },
+    {
+      "epoch": 0.542792392463562,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.142666666666668e-06,
+      "loss": 0.3548,
+      "step": 24430
+    },
+    {
+      "epoch": 0.5430145751866335,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.138222222222223e-06,
+      "loss": 0.3947,
+      "step": 24440
+    },
+    {
+      "epoch": 0.5432367579097049,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.133777777777778e-06,
+      "loss": 0.4046,
+      "step": 24450
+    },
+    {
+      "epoch": 0.5434589406327764,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.129333333333334e-06,
+      "loss": 0.4305,
+      "step": 24460
+    },
+    {
+      "epoch": 0.5436811233558478,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.124888888888889e-06,
+      "loss": 0.413,
+      "step": 24470
+    },
+    {
+      "epoch": 0.5439033060789193,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.120444444444446e-06,
+      "loss": 0.3882,
+      "step": 24480
+    },
+    {
+      "epoch": 0.5441254888019907,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.116e-06,
+      "loss": 0.4162,
+      "step": 24490
+    },
+    {
+      "epoch": 0.5443476715250622,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.111555555555557e-06,
+      "loss": 0.402,
+      "step": 24500
+    },
+    {
+      "epoch": 0.5445698542481336,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.107111111111112e-06,
+      "loss": 0.4079,
+      "step": 24510
+    },
+    {
+      "epoch": 0.5447920369712052,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.102666666666667e-06,
+      "loss": 0.3852,
+      "step": 24520
+    },
+    {
+      "epoch": 0.5450142196942765,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.098222222222223e-06,
+      "loss": 0.3792,
+      "step": 24530
+    },
+    {
+      "epoch": 0.545236402417348,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.093777777777778e-06,
+      "loss": 0.3756,
+      "step": 24540
+    },
+    {
+      "epoch": 0.5454585851404194,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.089333333333335e-06,
+      "loss": 0.379,
+      "step": 24550
+    },
+    {
+      "epoch": 0.545680767863491,
+      "grad_norm": 2.25,
+      "learning_rate": 9.08488888888889e-06,
+      "loss": 0.3756,
+      "step": 24560
+    },
+    {
+      "epoch": 0.5459029505865624,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.080444444444444e-06,
+      "loss": 0.3867,
+      "step": 24570
+    },
+    {
+      "epoch": 0.5461251333096339,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.076000000000001e-06,
+      "loss": 0.3852,
+      "step": 24580
+    },
+    {
+      "epoch": 0.5463473160327053,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.071555555555557e-06,
+      "loss": 0.3832,
+      "step": 24590
+    },
+    {
+      "epoch": 0.5465694987557768,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.067111111111112e-06,
+      "loss": 0.363,
+      "step": 24600
+    },
+    {
+      "epoch": 0.5467916814788482,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.062666666666667e-06,
+      "loss": 0.4217,
+      "step": 24610
+    },
+    {
+      "epoch": 0.5470138642019197,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.058222222222222e-06,
+      "loss": 0.3823,
+      "step": 24620
+    },
+    {
+      "epoch": 0.5472360469249911,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.053777777777778e-06,
+      "loss": 0.3898,
+      "step": 24630
+    },
+    {
+      "epoch": 0.5474582296480626,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.049333333333335e-06,
+      "loss": 0.3935,
+      "step": 24640
+    },
+    {
+      "epoch": 0.5476804123711341,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.04488888888889e-06,
+      "loss": 0.3809,
+      "step": 24650
+    },
+    {
+      "epoch": 0.5479025950942055,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.040444444444445e-06,
+      "loss": 0.3726,
+      "step": 24660
+    },
+    {
+      "epoch": 0.548124777817277,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.036e-06,
+      "loss": 0.3664,
+      "step": 24670
+    },
+    {
+      "epoch": 0.5483469605403484,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.031555555555556e-06,
+      "loss": 0.3716,
+      "step": 24680
+    },
+    {
+      "epoch": 0.5485691432634199,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.027111111111113e-06,
+      "loss": 0.4025,
+      "step": 24690
+    },
+    {
+      "epoch": 0.5487913259864913,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.022666666666667e-06,
+      "loss": 0.3298,
+      "step": 24700
+    },
+    {
+      "epoch": 0.5490135087095628,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.018222222222222e-06,
+      "loss": 0.3857,
+      "step": 24710
+    },
+    {
+      "epoch": 0.5492356914326342,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.013777777777779e-06,
+      "loss": 0.3528,
+      "step": 24720
+    },
+    {
+      "epoch": 0.5494578741557057,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.009333333333334e-06,
+      "loss": 0.3563,
+      "step": 24730
+    },
+    {
+      "epoch": 0.5496800568787771,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.00488888888889e-06,
+      "loss": 0.3696,
+      "step": 24740
+    },
+    {
+      "epoch": 0.5499022396018486,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.000444444444445e-06,
+      "loss": 0.3847,
+      "step": 24750
+    },
+    {
+      "epoch": 0.55012442232492,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.996e-06,
+      "loss": 0.3865,
+      "step": 24760
+    },
+    {
+      "epoch": 0.5503466050479915,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.991555555555556e-06,
+      "loss": 0.3664,
+      "step": 24770
+    },
+    {
+      "epoch": 0.5505687877710629,
+      "grad_norm": 2.625,
+      "learning_rate": 8.987111111111111e-06,
+      "loss": 0.4078,
+      "step": 24780
+    },
+    {
+      "epoch": 0.5507909704941344,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.982666666666668e-06,
+      "loss": 0.3581,
+      "step": 24790
+    },
+    {
+      "epoch": 0.5510131532172058,
+      "grad_norm": 2.25,
+      "learning_rate": 8.978222222222223e-06,
+      "loss": 0.3675,
+      "step": 24800
+    },
+    {
+      "epoch": 0.5512353359402773,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.97377777777778e-06,
+      "loss": 0.4063,
+      "step": 24810
+    },
+    {
+      "epoch": 0.5514575186633487,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.969333333333334e-06,
+      "loss": 0.385,
+      "step": 24820
+    },
+    {
+      "epoch": 0.5516797013864202,
+      "grad_norm": 2.625,
+      "learning_rate": 8.964888888888889e-06,
+      "loss": 0.4183,
+      "step": 24830
+    },
+    {
+      "epoch": 0.5519018841094916,
+      "grad_norm": 3.484375,
+      "learning_rate": 8.960444444444445e-06,
+      "loss": 0.4121,
+      "step": 24840
+    },
+    {
+      "epoch": 0.5521240668325631,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.956e-06,
+      "loss": 0.3987,
+      "step": 24850
+    },
+    {
+      "epoch": 0.5523462495556346,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.951555555555557e-06,
+      "loss": 0.3913,
+      "step": 24860
+    },
+    {
+      "epoch": 0.552568432278706,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.947111111111112e-06,
+      "loss": 0.4085,
+      "step": 24870
+    },
+    {
+      "epoch": 0.5527906150017775,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.942666666666667e-06,
+      "loss": 0.3699,
+      "step": 24880
+    },
+    {
+      "epoch": 0.5530127977248489,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.938222222222223e-06,
+      "loss": 0.3996,
+      "step": 24890
+    },
+    {
+      "epoch": 0.5532349804479204,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.93377777777778e-06,
+      "loss": 0.378,
+      "step": 24900
+    },
+    {
+      "epoch": 0.5534571631709918,
+      "grad_norm": 2.75,
+      "learning_rate": 8.929333333333334e-06,
+      "loss": 0.4091,
+      "step": 24910
+    },
+    {
+      "epoch": 0.5536793458940633,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.92488888888889e-06,
+      "loss": 0.3781,
+      "step": 24920
+    },
+    {
+      "epoch": 0.5539015286171347,
+      "grad_norm": 2.5,
+      "learning_rate": 8.920444444444444e-06,
+      "loss": 0.3566,
+      "step": 24930
+    },
+    {
+      "epoch": 0.5541237113402062,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.916e-06,
+      "loss": 0.3532,
+      "step": 24940
+    },
+    {
+      "epoch": 0.5543458940632776,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.911555555555557e-06,
+      "loss": 0.3517,
+      "step": 24950
+    },
+    {
+      "epoch": 0.5545680767863491,
+      "grad_norm": 2.625,
+      "learning_rate": 8.907111111111112e-06,
+      "loss": 0.3661,
+      "step": 24960
+    },
+    {
+      "epoch": 0.5547902595094205,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.902666666666667e-06,
+      "loss": 0.3667,
+      "step": 24970
+    },
+    {
+      "epoch": 0.555012442232492,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.898222222222222e-06,
+      "loss": 0.384,
+      "step": 24980
+    },
+    {
+      "epoch": 0.5552346249555634,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.893777777777778e-06,
+      "loss": 0.3959,
+      "step": 24990
+    },
+    {
+      "epoch": 0.5554568076786349,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.889333333333335e-06,
+      "loss": 0.3585,
+      "step": 25000
+    },
+    {
+      "epoch": 0.5556789904017063,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.88488888888889e-06,
+      "loss": 0.3841,
+      "step": 25010
+    },
+    {
+      "epoch": 0.5559011731247778,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.880444444444445e-06,
+      "loss": 0.3772,
+      "step": 25020
+    },
+    {
+      "epoch": 0.5561233558478492,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.876e-06,
+      "loss": 0.3723,
+      "step": 25030
+    },
+    {
+      "epoch": 0.5563455385709207,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.871555555555556e-06,
+      "loss": 0.4183,
+      "step": 25040
+    },
+    {
+      "epoch": 0.5565677212939921,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.867111111111112e-06,
+      "loss": 0.3735,
+      "step": 25050
+    },
+    {
+      "epoch": 0.5567899040170636,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.862666666666667e-06,
+      "loss": 0.3475,
+      "step": 25060
+    },
+    {
+      "epoch": 0.5570120867401351,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.858222222222222e-06,
+      "loss": 0.3926,
+      "step": 25070
+    },
+    {
+      "epoch": 0.5572342694632065,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.853777777777779e-06,
+      "loss": 0.4119,
+      "step": 25080
+    },
+    {
+      "epoch": 0.557456452186278,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.849333333333334e-06,
+      "loss": 0.3947,
+      "step": 25090
+    },
+    {
+      "epoch": 0.5576786349093494,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.84488888888889e-06,
+      "loss": 0.4107,
+      "step": 25100
+    },
+    {
+      "epoch": 0.5579008176324209,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.840444444444445e-06,
+      "loss": 0.4038,
+      "step": 25110
+    },
+    {
+      "epoch": 0.5581230003554923,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.836000000000001e-06,
+      "loss": 0.3674,
+      "step": 25120
+    },
+    {
+      "epoch": 0.5583451830785638,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.831555555555556e-06,
+      "loss": 0.428,
+      "step": 25130
+    },
+    {
+      "epoch": 0.5585673658016352,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.827111111111111e-06,
+      "loss": 0.3403,
+      "step": 25140
+    },
+    {
+      "epoch": 0.5587895485247067,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.822666666666668e-06,
+      "loss": 0.3861,
+      "step": 25150
+    },
+    {
+      "epoch": 0.5590117312477781,
+      "grad_norm": 2.625,
+      "learning_rate": 8.818222222222223e-06,
+      "loss": 0.3654,
+      "step": 25160
+    },
+    {
+      "epoch": 0.5592339139708496,
+      "grad_norm": 2.5,
+      "learning_rate": 8.813777777777779e-06,
+      "loss": 0.3821,
+      "step": 25170
+    },
+    {
+      "epoch": 0.559456096693921,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.809333333333334e-06,
+      "loss": 0.3817,
+      "step": 25180
+    },
+    {
+      "epoch": 0.5596782794169926,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.804888888888889e-06,
+      "loss": 0.3875,
+      "step": 25190
+    },
+    {
+      "epoch": 0.559900462140064,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.800444444444445e-06,
+      "loss": 0.4084,
+      "step": 25200
+    },
+    {
+      "epoch": 0.5601226448631355,
+      "grad_norm": 2.5,
+      "learning_rate": 8.796000000000002e-06,
+      "loss": 0.4038,
+      "step": 25210
+    },
+    {
+      "epoch": 0.5603448275862069,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.791555555555557e-06,
+      "loss": 0.3568,
+      "step": 25220
+    },
+    {
+      "epoch": 0.5605670103092784,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.787111111111112e-06,
+      "loss": 0.3638,
+      "step": 25230
+    },
+    {
+      "epoch": 0.5607891930323498,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.782666666666666e-06,
+      "loss": 0.4023,
+      "step": 25240
+    },
+    {
+      "epoch": 0.5610113757554213,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.778222222222223e-06,
+      "loss": 0.3715,
+      "step": 25250
+    },
+    {
+      "epoch": 0.5612335584784928,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.77377777777778e-06,
+      "loss": 0.3963,
+      "step": 25260
+    },
+    {
+      "epoch": 0.5614557412015642,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.769333333333334e-06,
+      "loss": 0.3837,
+      "step": 25270
+    },
+    {
+      "epoch": 0.5616779239246357,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.764888888888889e-06,
+      "loss": 0.4131,
+      "step": 25280
+    },
+    {
+      "epoch": 0.5619001066477071,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.760444444444444e-06,
+      "loss": 0.3561,
+      "step": 25290
+    },
+    {
+      "epoch": 0.5621222893707786,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.756e-06,
+      "loss": 0.3487,
+      "step": 25300
+    },
+    {
+      "epoch": 0.56234447209385,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.751555555555557e-06,
+      "loss": 0.388,
+      "step": 25310
+    },
+    {
+      "epoch": 0.5625666548169215,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.747111111111112e-06,
+      "loss": 0.3671,
+      "step": 25320
+    },
+    {
+      "epoch": 0.5627888375399929,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.742666666666667e-06,
+      "loss": 0.4174,
+      "step": 25330
+    },
+    {
+      "epoch": 0.5630110202630644,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.738222222222222e-06,
+      "loss": 0.405,
+      "step": 25340
+    },
+    {
+      "epoch": 0.5632332029861358,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.733777777777778e-06,
+      "loss": 0.3833,
+      "step": 25350
+    },
+    {
+      "epoch": 0.5634553857092073,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.729333333333335e-06,
+      "loss": 0.4073,
+      "step": 25360
+    },
+    {
+      "epoch": 0.5636775684322787,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.72488888888889e-06,
+      "loss": 0.3715,
+      "step": 25370
+    },
+    {
+      "epoch": 0.5638997511553502,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.720444444444444e-06,
+      "loss": 0.3855,
+      "step": 25380
+    },
+    {
+      "epoch": 0.5641219338784216,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.716000000000001e-06,
+      "loss": 0.3614,
+      "step": 25390
+    },
+    {
+      "epoch": 0.5643441166014931,
+      "grad_norm": 2.125,
+      "learning_rate": 8.711555555555556e-06,
+      "loss": 0.3656,
+      "step": 25400
+    },
+    {
+      "epoch": 0.5645662993245645,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.707111111111112e-06,
+      "loss": 0.3981,
+      "step": 25410
+    },
+    {
+      "epoch": 0.564788482047636,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.702666666666667e-06,
+      "loss": 0.4118,
+      "step": 25420
+    },
+    {
+      "epoch": 0.5650106647707074,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.698222222222224e-06,
+      "loss": 0.3816,
+      "step": 25430
+    },
+    {
+      "epoch": 0.5652328474937789,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.693777777777779e-06,
+      "loss": 0.3911,
+      "step": 25440
+    },
+    {
+      "epoch": 0.5654550302168503,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.689333333333333e-06,
+      "loss": 0.3567,
+      "step": 25450
+    },
+    {
+      "epoch": 0.5656772129399218,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.68488888888889e-06,
+      "loss": 0.3928,
+      "step": 25460
+    },
+    {
+      "epoch": 0.5658993956629933,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.680444444444445e-06,
+      "loss": 0.3439,
+      "step": 25470
+    },
+    {
+      "epoch": 0.5661215783860647,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.676000000000001e-06,
+      "loss": 0.3818,
+      "step": 25480
+    },
+    {
+      "epoch": 0.5663437611091362,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.671555555555556e-06,
+      "loss": 0.3726,
+      "step": 25490
+    },
+    {
+      "epoch": 0.5665659438322076,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.667111111111111e-06,
+      "loss": 0.3898,
+      "step": 25500
+    },
+    {
+      "epoch": 0.5667881265552791,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.662666666666668e-06,
+      "loss": 0.4124,
+      "step": 25510
+    },
+    {
+      "epoch": 0.5670103092783505,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.658222222222224e-06,
+      "loss": 0.3822,
+      "step": 25520
+    },
+    {
+      "epoch": 0.567232492001422,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.653777777777779e-06,
+      "loss": 0.3788,
+      "step": 25530
+    },
+    {
+      "epoch": 0.5674546747244934,
+      "grad_norm": 2.125,
+      "learning_rate": 8.649333333333334e-06,
+      "loss": 0.3916,
+      "step": 25540
+    },
+    {
+      "epoch": 0.5676768574475649,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.644888888888889e-06,
+      "loss": 0.3715,
+      "step": 25550
+    },
+    {
+      "epoch": 0.5678990401706363,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.640444444444445e-06,
+      "loss": 0.3836,
+      "step": 25560
+    },
+    {
+      "epoch": 0.5681212228937078,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.636000000000002e-06,
+      "loss": 0.4102,
+      "step": 25570
+    },
+    {
+      "epoch": 0.5683434056167792,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.631555555555557e-06,
+      "loss": 0.3541,
+      "step": 25580
+    },
+    {
+      "epoch": 0.5685655883398507,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.627111111111111e-06,
+      "loss": 0.3905,
+      "step": 25590
+    },
+    {
+      "epoch": 0.5687877710629221,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.622666666666666e-06,
+      "loss": 0.3719,
+      "step": 25600
+    },
+    {
+      "epoch": 0.5690099537859936,
+      "grad_norm": 2.5,
+      "learning_rate": 8.618222222222223e-06,
+      "loss": 0.3946,
+      "step": 25610
+    },
+    {
+      "epoch": 0.569232136509065,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.61377777777778e-06,
+      "loss": 0.4057,
+      "step": 25620
+    },
+    {
+      "epoch": 0.5694543192321365,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.609333333333334e-06,
+      "loss": 0.4055,
+      "step": 25630
+    },
+    {
+      "epoch": 0.5696765019552079,
+      "grad_norm": 2.25,
+      "learning_rate": 8.604888888888889e-06,
+      "loss": 0.4136,
+      "step": 25640
+    },
+    {
+      "epoch": 0.5698986846782794,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.600444444444444e-06,
+      "loss": 0.396,
+      "step": 25650
+    },
+    {
+      "epoch": 0.5701208674013508,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.596e-06,
+      "loss": 0.3647,
+      "step": 25660
+    },
+    {
+      "epoch": 0.5703430501244223,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.591555555555557e-06,
+      "loss": 0.3717,
+      "step": 25670
+    },
+    {
+      "epoch": 0.5705652328474938,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.587111111111112e-06,
+      "loss": 0.3827,
+      "step": 25680
+    },
+    {
+      "epoch": 0.5707874155705652,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.582666666666667e-06,
+      "loss": 0.4217,
+      "step": 25690
+    },
+    {
+      "epoch": 0.5710095982936367,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.578222222222223e-06,
+      "loss": 0.4349,
+      "step": 25700
+    },
+    {
+      "epoch": 0.5712317810167081,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.573777777777778e-06,
+      "loss": 0.3873,
+      "step": 25710
+    },
+    {
+      "epoch": 0.5714539637397796,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.569333333333335e-06,
+      "loss": 0.4171,
+      "step": 25720
+    },
+    {
+      "epoch": 0.571676146462851,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.56488888888889e-06,
+      "loss": 0.4147,
+      "step": 25730
+    },
+    {
+      "epoch": 0.5718983291859225,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.560444444444446e-06,
+      "loss": 0.4042,
+      "step": 25740
+    },
+    {
+      "epoch": 0.5721205119089939,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.556e-06,
+      "loss": 0.4124,
+      "step": 25750
+    },
+    {
+      "epoch": 0.5723426946320654,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.551555555555556e-06,
+      "loss": 0.3899,
+      "step": 25760
+    },
+    {
+      "epoch": 0.5725648773551368,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.547111111111112e-06,
+      "loss": 0.397,
+      "step": 25770
+    },
+    {
+      "epoch": 0.5727870600782083,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.542666666666667e-06,
+      "loss": 0.3999,
+      "step": 25780
+    },
+    {
+      "epoch": 0.5730092428012797,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.538222222222224e-06,
+      "loss": 0.3891,
+      "step": 25790
+    },
+    {
+      "epoch": 0.5732314255243512,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.533777777777778e-06,
+      "loss": 0.4335,
+      "step": 25800
+    },
+    {
+      "epoch": 0.5734536082474226,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.529333333333333e-06,
+      "loss": 0.4054,
+      "step": 25810
+    },
+    {
+      "epoch": 0.5736757909704941,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.52488888888889e-06,
+      "loss": 0.3576,
+      "step": 25820
+    },
+    {
+      "epoch": 0.5738979736935655,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.520444444444446e-06,
+      "loss": 0.428,
+      "step": 25830
+    },
+    {
+      "epoch": 0.574120156416637,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.516000000000001e-06,
+      "loss": 0.362,
+      "step": 25840
+    },
+    {
+      "epoch": 0.5743423391397084,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.511555555555556e-06,
+      "loss": 0.3951,
+      "step": 25850
+    },
+    {
+      "epoch": 0.57456452186278,
+      "grad_norm": 2.375,
+      "learning_rate": 8.50711111111111e-06,
+      "loss": 0.3886,
+      "step": 25860
+    },
+    {
+      "epoch": 0.5747867045858513,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.502666666666667e-06,
+      "loss": 0.4179,
+      "step": 25870
+    },
+    {
+      "epoch": 0.5750088873089229,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.498222222222224e-06,
+      "loss": 0.3508,
+      "step": 25880
+    },
+    {
+      "epoch": 0.5752310700319944,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.493777777777779e-06,
+      "loss": 0.3949,
+      "step": 25890
+    },
+    {
+      "epoch": 0.5754532527550658,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.489333333333334e-06,
+      "loss": 0.4156,
+      "step": 25900
+    },
+    {
+      "epoch": 0.5756754354781373,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.484888888888888e-06,
+      "loss": 0.3551,
+      "step": 25910
+    },
+    {
+      "epoch": 0.5758976182012087,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.480444444444445e-06,
+      "loss": 0.3424,
+      "step": 25920
+    },
+    {
+      "epoch": 0.5761198009242802,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.476000000000002e-06,
+      "loss": 0.4259,
+      "step": 25930
+    },
+    {
+      "epoch": 0.5763419836473516,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.471555555555556e-06,
+      "loss": 0.3698,
+      "step": 25940
+    },
+    {
+      "epoch": 0.5765641663704231,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.467111111111111e-06,
+      "loss": 0.3744,
+      "step": 25950
+    },
+    {
+      "epoch": 0.5767863490934945,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.462666666666666e-06,
+      "loss": 0.3816,
+      "step": 25960
+    },
+    {
+      "epoch": 0.577008531816566,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.458222222222223e-06,
+      "loss": 0.3592,
+      "step": 25970
+    },
+    {
+      "epoch": 0.5772307145396374,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.453777777777779e-06,
+      "loss": 0.361,
+      "step": 25980
+    },
+    {
+      "epoch": 0.5774528972627089,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.449333333333334e-06,
+      "loss": 0.3623,
+      "step": 25990
+    },
+    {
+      "epoch": 0.5776750799857803,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.444888888888889e-06,
+      "loss": 0.4267,
+      "step": 26000
+    },
+    {
+      "epoch": 0.5778972627088518,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.440444444444445e-06,
+      "loss": 0.4347,
+      "step": 26010
+    },
+    {
+      "epoch": 0.5781194454319232,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.436e-06,
+      "loss": 0.3921,
+      "step": 26020
+    },
+    {
+      "epoch": 0.5783416281549947,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.431555555555557e-06,
+      "loss": 0.3784,
+      "step": 26030
+    },
+    {
+      "epoch": 0.5785638108780661,
+      "grad_norm": 2.25,
+      "learning_rate": 8.427111111111112e-06,
+      "loss": 0.3643,
+      "step": 26040
+    },
+    {
+      "epoch": 0.5787859936011376,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.422666666666668e-06,
+      "loss": 0.3954,
+      "step": 26050
+    },
+    {
+      "epoch": 0.579008176324209,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.418222222222223e-06,
+      "loss": 0.3729,
+      "step": 26060
+    },
+    {
+      "epoch": 0.5792303590472805,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.413777777777778e-06,
+      "loss": 0.4098,
+      "step": 26070
+    },
+    {
+      "epoch": 0.579452541770352,
+      "grad_norm": 2.5,
+      "learning_rate": 8.409333333333334e-06,
+      "loss": 0.3831,
+      "step": 26080
+    },
+    {
+      "epoch": 0.5796747244934234,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.40488888888889e-06,
+      "loss": 0.3327,
+      "step": 26090
+    },
+    {
+      "epoch": 0.5798969072164949,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.400444444444446e-06,
+      "loss": 0.389,
+      "step": 26100
+    },
+    {
+      "epoch": 0.5801190899395663,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.396e-06,
+      "loss": 0.4266,
+      "step": 26110
+    },
+    {
+      "epoch": 0.5803412726626378,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.391555555555555e-06,
+      "loss": 0.3712,
+      "step": 26120
+    },
+    {
+      "epoch": 0.5805634553857092,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.387111111111112e-06,
+      "loss": 0.3867,
+      "step": 26130
+    },
+    {
+      "epoch": 0.5807856381087807,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.382666666666669e-06,
+      "loss": 0.4288,
+      "step": 26140
+    },
+    {
+      "epoch": 0.5810078208318521,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.378222222222223e-06,
+      "loss": 0.3929,
+      "step": 26150
+    },
+    {
+      "epoch": 0.5812300035549236,
+      "grad_norm": 2.375,
+      "learning_rate": 8.373777777777778e-06,
+      "loss": 0.3626,
+      "step": 26160
+    },
+    {
+      "epoch": 0.581452186277995,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.369333333333333e-06,
+      "loss": 0.3701,
+      "step": 26170
+    },
+    {
+      "epoch": 0.5816743690010665,
+      "grad_norm": 2.25,
+      "learning_rate": 8.36488888888889e-06,
+      "loss": 0.3522,
+      "step": 26180
+    },
+    {
+      "epoch": 0.5818965517241379,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.360444444444446e-06,
+      "loss": 0.3935,
+      "step": 26190
+    },
+    {
+      "epoch": 0.5821187344472094,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.356000000000001e-06,
+      "loss": 0.3453,
+      "step": 26200
+    },
+    {
+      "epoch": 0.5823409171702808,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.351555555555556e-06,
+      "loss": 0.3886,
+      "step": 26210
+    },
+    {
+      "epoch": 0.5825630998933523,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.34711111111111e-06,
+      "loss": 0.4037,
+      "step": 26220
+    },
+    {
+      "epoch": 0.5827852826164237,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.342666666666667e-06,
+      "loss": 0.3791,
+      "step": 26230
+    },
+    {
+      "epoch": 0.5830074653394952,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.338222222222224e-06,
+      "loss": 0.3771,
+      "step": 26240
+    },
+    {
+      "epoch": 0.5832296480625666,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.333777777777779e-06,
+      "loss": 0.3691,
+      "step": 26250
+    },
+    {
+      "epoch": 0.5834518307856381,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.329333333333333e-06,
+      "loss": 0.3587,
+      "step": 26260
+    },
+    {
+      "epoch": 0.5836740135087095,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.324888888888888e-06,
+      "loss": 0.3607,
+      "step": 26270
+    },
+    {
+      "epoch": 0.583896196231781,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.320444444444445e-06,
+      "loss": 0.3732,
+      "step": 26280
+    },
+    {
+      "epoch": 0.5841183789548525,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.316000000000001e-06,
+      "loss": 0.3848,
+      "step": 26290
+    },
+    {
+      "epoch": 0.5843405616779239,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.311555555555556e-06,
+      "loss": 0.3833,
+      "step": 26300
+    },
+    {
+      "epoch": 0.5845627444009954,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.307111111111111e-06,
+      "loss": 0.351,
+      "step": 26310
+    },
+    {
+      "epoch": 0.5847849271240668,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.302666666666668e-06,
+      "loss": 0.3716,
+      "step": 26320
+    },
+    {
+      "epoch": 0.5850071098471383,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.298222222222222e-06,
+      "loss": 0.3664,
+      "step": 26330
+    },
+    {
+      "epoch": 0.5852292925702097,
+      "grad_norm": 3.125,
+      "learning_rate": 8.293777777777779e-06,
+      "loss": 0.4228,
+      "step": 26340
+    },
+    {
+      "epoch": 0.5854514752932812,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.289333333333334e-06,
+      "loss": 0.3897,
+      "step": 26350
+    },
+    {
+      "epoch": 0.5856736580163526,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.28488888888889e-06,
+      "loss": 0.3806,
+      "step": 26360
+    },
+    {
+      "epoch": 0.5858958407394241,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.280444444444445e-06,
+      "loss": 0.378,
+      "step": 26370
+    },
+    {
+      "epoch": 0.5861180234624955,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.276e-06,
+      "loss": 0.3796,
+      "step": 26380
+    },
+    {
+      "epoch": 0.586340206185567,
+      "grad_norm": 2.125,
+      "learning_rate": 8.271555555555557e-06,
+      "loss": 0.3603,
+      "step": 26390
+    },
+    {
+      "epoch": 0.5865623889086384,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.267111111111111e-06,
+      "loss": 0.3759,
+      "step": 26400
+    },
+    {
+      "epoch": 0.5867845716317099,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.262666666666668e-06,
+      "loss": 0.3788,
+      "step": 26410
+    },
+    {
+      "epoch": 0.5870067543547813,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.258222222222223e-06,
+      "loss": 0.3692,
+      "step": 26420
+    },
+    {
+      "epoch": 0.5872289370778528,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.253777777777778e-06,
+      "loss": 0.4502,
+      "step": 26430
+    },
+    {
+      "epoch": 0.5874511198009242,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.249333333333334e-06,
+      "loss": 0.3761,
+      "step": 26440
+    },
+    {
+      "epoch": 0.5876733025239957,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.24488888888889e-06,
+      "loss": 0.3842,
+      "step": 26450
+    },
+    {
+      "epoch": 0.5878954852470671,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.240444444444446e-06,
+      "loss": 0.3799,
+      "step": 26460
+    },
+    {
+      "epoch": 0.5881176679701386,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.236e-06,
+      "loss": 0.3935,
+      "step": 26470
+    },
+    {
+      "epoch": 0.58833985069321,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.231555555555555e-06,
+      "loss": 0.4229,
+      "step": 26480
+    },
+    {
+      "epoch": 0.5885620334162815,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.227111111111112e-06,
+      "loss": 0.39,
+      "step": 26490
+    },
+    {
+      "epoch": 0.5887842161393531,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.222666666666668e-06,
+      "loss": 0.386,
+      "step": 26500
+    },
+    {
+      "epoch": 0.5890063988624245,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.218222222222223e-06,
+      "loss": 0.3853,
+      "step": 26510
+    },
+    {
+      "epoch": 0.589228581585496,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.213777777777778e-06,
+      "loss": 0.4081,
+      "step": 26520
+    },
+    {
+      "epoch": 0.5894507643085674,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.209333333333333e-06,
+      "loss": 0.3724,
+      "step": 26530
+    },
+    {
+      "epoch": 0.5896729470316389,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.20488888888889e-06,
+      "loss": 0.3663,
+      "step": 26540
+    },
+    {
+      "epoch": 0.5898951297547103,
+      "grad_norm": 2.25,
+      "learning_rate": 8.200444444444446e-06,
+      "loss": 0.402,
+      "step": 26550
+    },
+    {
+      "epoch": 0.5901173124777818,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.196e-06,
+      "loss": 0.3919,
+      "step": 26560
+    },
+    {
+      "epoch": 0.5903394952008532,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.191555555555556e-06,
+      "loss": 0.3896,
+      "step": 26570
+    },
+    {
+      "epoch": 0.5905616779239247,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.18711111111111e-06,
+      "loss": 0.4163,
+      "step": 26580
+    },
+    {
+      "epoch": 0.5907838606469961,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.182666666666667e-06,
+      "loss": 0.3443,
+      "step": 26590
+    },
+    {
+      "epoch": 0.5910060433700676,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.178222222222224e-06,
+      "loss": 0.3769,
+      "step": 26600
+    },
+    {
+      "epoch": 0.591228226093139,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.173777777777778e-06,
+      "loss": 0.3943,
+      "step": 26610
+    },
+    {
+      "epoch": 0.5914504088162105,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.169333333333333e-06,
+      "loss": 0.406,
+      "step": 26620
+    },
+    {
+      "epoch": 0.5916725915392819,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.16488888888889e-06,
+      "loss": 0.3781,
+      "step": 26630
+    },
+    {
+      "epoch": 0.5918947742623534,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.160444444444445e-06,
+      "loss": 0.3836,
+      "step": 26640
+    },
+    {
+      "epoch": 0.5921169569854248,
+      "grad_norm": 2.25,
+      "learning_rate": 8.156000000000001e-06,
+      "loss": 0.4014,
+      "step": 26650
+    },
+    {
+      "epoch": 0.5923391397084963,
+      "grad_norm": 2.375,
+      "learning_rate": 8.151555555555556e-06,
+      "loss": 0.3852,
+      "step": 26660
+    },
+    {
+      "epoch": 0.5925613224315677,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.147111111111113e-06,
+      "loss": 0.3652,
+      "step": 26670
+    },
+    {
+      "epoch": 0.5927835051546392,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.142666666666667e-06,
+      "loss": 0.3818,
+      "step": 26680
+    },
+    {
+      "epoch": 0.5930056878777106,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.138222222222222e-06,
+      "loss": 0.3772,
+      "step": 26690
+    },
+    {
+      "epoch": 0.5932278706007821,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.133777777777779e-06,
+      "loss": 0.4421,
+      "step": 26700
+    },
+    {
+      "epoch": 0.5934500533238536,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.129333333333334e-06,
+      "loss": 0.3549,
+      "step": 26710
+    },
+    {
+      "epoch": 0.593672236046925,
+      "grad_norm": 2.625,
+      "learning_rate": 8.12488888888889e-06,
+      "loss": 0.4144,
+      "step": 26720
+    },
+    {
+      "epoch": 0.5938944187699965,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.120444444444445e-06,
+      "loss": 0.3973,
+      "step": 26730
+    },
+    {
+      "epoch": 0.5941166014930679,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.116e-06,
+      "loss": 0.4206,
+      "step": 26740
+    },
+    {
+      "epoch": 0.5943387842161394,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.111555555555556e-06,
+      "loss": 0.3941,
+      "step": 26750
+    },
+    {
+      "epoch": 0.5945609669392108,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.107111111111113e-06,
+      "loss": 0.3912,
+      "step": 26760
+    },
+    {
+      "epoch": 0.5947831496622823,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.102666666666668e-06,
+      "loss": 0.4154,
+      "step": 26770
+    },
+    {
+      "epoch": 0.5950053323853537,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.098222222222223e-06,
+      "loss": 0.3884,
+      "step": 26780
+    },
+    {
+      "epoch": 0.5952275151084252,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.093777777777777e-06,
+      "loss": 0.3626,
+      "step": 26790
+    },
+    {
+      "epoch": 0.5954496978314966,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.089333333333334e-06,
+      "loss": 0.4196,
+      "step": 26800
+    },
+    {
+      "epoch": 0.5956718805545681,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.08488888888889e-06,
+      "loss": 0.365,
+      "step": 26810
+    },
+    {
+      "epoch": 0.5958940632776395,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.080444444444445e-06,
+      "loss": 0.36,
+      "step": 26820
+    },
+    {
+      "epoch": 0.596116246000711,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.076e-06,
+      "loss": 0.362,
+      "step": 26830
+    },
+    {
+      "epoch": 0.5963384287237824,
+      "grad_norm": 2.375,
+      "learning_rate": 8.071555555555555e-06,
+      "loss": 0.3744,
+      "step": 26840
+    },
+    {
+      "epoch": 0.5965606114468539,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.067111111111112e-06,
+      "loss": 0.3593,
+      "step": 26850
+    },
+    {
+      "epoch": 0.5967827941699253,
+      "grad_norm": 2.375,
+      "learning_rate": 8.062666666666668e-06,
+      "loss": 0.3708,
+      "step": 26860
+    },
+    {
+      "epoch": 0.5970049768929968,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.058222222222223e-06,
+      "loss": 0.372,
+      "step": 26870
+    },
+    {
+      "epoch": 0.5972271596160682,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.053777777777778e-06,
+      "loss": 0.3874,
+      "step": 26880
+    },
+    {
+      "epoch": 0.5974493423391397,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.049333333333333e-06,
+      "loss": 0.3587,
+      "step": 26890
+    },
+    {
+      "epoch": 0.5976715250622112,
+      "grad_norm": 2.25,
+      "learning_rate": 8.04488888888889e-06,
+      "loss": 0.3558,
+      "step": 26900
+    },
+    {
+      "epoch": 0.5978937077852826,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.040444444444446e-06,
+      "loss": 0.3847,
+      "step": 26910
+    },
+    {
+      "epoch": 0.5981158905083541,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.036e-06,
+      "loss": 0.4165,
+      "step": 26920
+    },
+    {
+      "epoch": 0.5983380732314255,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.031555555555555e-06,
+      "loss": 0.3804,
+      "step": 26930
+    },
+    {
+      "epoch": 0.598560255954497,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.027111111111112e-06,
+      "loss": 0.3987,
+      "step": 26940
+    },
+    {
+      "epoch": 0.5987824386775684,
+      "grad_norm": 3.0,
+      "learning_rate": 8.022666666666667e-06,
+      "loss": 0.4194,
+      "step": 26950
+    },
+    {
+      "epoch": 0.5990046214006399,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.018222222222223e-06,
+      "loss": 0.3575,
+      "step": 26960
+    },
+    {
+      "epoch": 0.5992268041237113,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.013777777777778e-06,
+      "loss": 0.3862,
+      "step": 26970
+    },
+    {
+      "epoch": 0.5994489868467828,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.009333333333335e-06,
+      "loss": 0.3968,
+      "step": 26980
+    },
+    {
+      "epoch": 0.5996711695698542,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.00488888888889e-06,
+      "loss": 0.3703,
+      "step": 26990
+    },
+    {
+      "epoch": 0.5998933522929257,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.000444444444444e-06,
+      "loss": 0.388,
+      "step": 27000
+    },
+    {
+      "epoch": 0.6001155350159971,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.996000000000001e-06,
+      "loss": 0.3826,
+      "step": 27010
+    },
+    {
+      "epoch": 0.6003377177390686,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.991555555555556e-06,
+      "loss": 0.382,
+      "step": 27020
+    },
+    {
+      "epoch": 0.60055990046214,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.987111111111112e-06,
+      "loss": 0.3621,
+      "step": 27030
+    },
+    {
+      "epoch": 0.6007820831852115,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.982666666666667e-06,
+      "loss": 0.4083,
+      "step": 27040
+    },
+    {
+      "epoch": 0.6010042659082829,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.978222222222222e-06,
+      "loss": 0.3844,
+      "step": 27050
+    },
+    {
+      "epoch": 0.6012264486313544,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.973777777777779e-06,
+      "loss": 0.3781,
+      "step": 27060
+    },
+    {
+      "epoch": 0.6014486313544258,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.969333333333335e-06,
+      "loss": 0.3586,
+      "step": 27070
+    },
+    {
+      "epoch": 0.6016708140774973,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.96488888888889e-06,
+      "loss": 0.3765,
+      "step": 27080
+    },
+    {
+      "epoch": 0.6018929968005687,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.960444444444445e-06,
+      "loss": 0.3895,
+      "step": 27090
+    },
+    {
+      "epoch": 0.6021151795236402,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.956e-06,
+      "loss": 0.3906,
+      "step": 27100
+    },
+    {
+      "epoch": 0.6023373622467117,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.951555555555556e-06,
+      "loss": 0.3882,
+      "step": 27110
+    },
+    {
+      "epoch": 0.6025595449697831,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.947111111111113e-06,
+      "loss": 0.3914,
+      "step": 27120
+    },
+    {
+      "epoch": 0.6027817276928547,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.942666666666668e-06,
+      "loss": 0.3816,
+      "step": 27130
+    },
+    {
+      "epoch": 0.603003910415926,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.938222222222222e-06,
+      "loss": 0.3587,
+      "step": 27140
+    },
+    {
+      "epoch": 0.6032260931389976,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.933777777777777e-06,
+      "loss": 0.4219,
+      "step": 27150
+    },
+    {
+      "epoch": 0.603448275862069,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.929333333333334e-06,
+      "loss": 0.3842,
+      "step": 27160
+    },
+    {
+      "epoch": 0.6036704585851405,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.92488888888889e-06,
+      "loss": 0.3762,
+      "step": 27170
+    },
+    {
+      "epoch": 0.6038926413082119,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.920444444444445e-06,
+      "loss": 0.3777,
+      "step": 27180
+    },
+    {
+      "epoch": 0.6041148240312834,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.916e-06,
+      "loss": 0.389,
+      "step": 27190
+    },
+    {
+      "epoch": 0.6043370067543548,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.911555555555555e-06,
+      "loss": 0.3915,
+      "step": 27200
+    },
+    {
+      "epoch": 0.6045591894774263,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.907111111111111e-06,
+      "loss": 0.3741,
+      "step": 27210
+    },
+    {
+      "epoch": 0.6047813722004977,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.902666666666668e-06,
+      "loss": 0.3503,
+      "step": 27220
+    },
+    {
+      "epoch": 0.6050035549235692,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.898222222222223e-06,
+      "loss": 0.3603,
+      "step": 27230
+    },
+    {
+      "epoch": 0.6052257376466406,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.893777777777778e-06,
+      "loss": 0.3596,
+      "step": 27240
+    },
+    {
+      "epoch": 0.6054479203697121,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.889333333333334e-06,
+      "loss": 0.4017,
+      "step": 27250
+    },
+    {
+      "epoch": 0.6056701030927835,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.884888888888889e-06,
+      "loss": 0.375,
+      "step": 27260
+    },
+    {
+      "epoch": 0.605892285815855,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.880444444444446e-06,
+      "loss": 0.3834,
+      "step": 27270
+    },
+    {
+      "epoch": 0.6061144685389264,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.876e-06,
+      "loss": 0.3794,
+      "step": 27280
+    },
+    {
+      "epoch": 0.6063366512619979,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.871555555555557e-06,
+      "loss": 0.3893,
+      "step": 27290
+    },
+    {
+      "epoch": 0.6065588339850693,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.867111111111112e-06,
+      "loss": 0.3511,
+      "step": 27300
+    },
+    {
+      "epoch": 0.6067810167081408,
+      "grad_norm": 2.5,
+      "learning_rate": 7.862666666666667e-06,
+      "loss": 0.3741,
+      "step": 27310
+    },
+    {
+      "epoch": 0.6070031994312123,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.858222222222223e-06,
+      "loss": 0.4052,
+      "step": 27320
+    },
+    {
+      "epoch": 0.6072253821542837,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.853777777777778e-06,
+      "loss": 0.4087,
+      "step": 27330
+    },
+    {
+      "epoch": 0.6074475648773552,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.849333333333335e-06,
+      "loss": 0.3986,
+      "step": 27340
+    },
+    {
+      "epoch": 0.6076697476004266,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.84488888888889e-06,
+      "loss": 0.3892,
+      "step": 27350
+    },
+    {
+      "epoch": 0.6078919303234981,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.840444444444444e-06,
+      "loss": 0.3862,
+      "step": 27360
+    },
+    {
+      "epoch": 0.6081141130465695,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.836000000000001e-06,
+      "loss": 0.3889,
+      "step": 27370
+    },
+    {
+      "epoch": 0.608336295769641,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.831555555555557e-06,
+      "loss": 0.3702,
+      "step": 27380
+    },
+    {
+      "epoch": 0.6085584784927124,
+      "grad_norm": 2.625,
+      "learning_rate": 7.827111111111112e-06,
+      "loss": 0.3859,
+      "step": 27390
+    },
+    {
+      "epoch": 0.6087806612157839,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.822666666666667e-06,
+      "loss": 0.4044,
+      "step": 27400
+    },
+    {
+      "epoch": 0.6090028439388553,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.818222222222222e-06,
+      "loss": 0.3996,
+      "step": 27410
+    },
+    {
+      "epoch": 0.6092250266619268,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.813777777777778e-06,
+      "loss": 0.3879,
+      "step": 27420
+    },
+    {
+      "epoch": 0.6094472093849982,
+      "grad_norm": 3.0,
+      "learning_rate": 7.809333333333335e-06,
+      "loss": 0.3552,
+      "step": 27430
+    },
+    {
+      "epoch": 0.6096693921080697,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.80488888888889e-06,
+      "loss": 0.3508,
+      "step": 27440
+    },
+    {
+      "epoch": 0.6098915748311411,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.800444444444445e-06,
+      "loss": 0.3573,
+      "step": 27450
+    },
+    {
+      "epoch": 0.6101137575542126,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.796e-06,
+      "loss": 0.385,
+      "step": 27460
+    },
+    {
+      "epoch": 0.610335940277284,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.791555555555556e-06,
+      "loss": 0.3949,
+      "step": 27470
+    },
+    {
+      "epoch": 0.6105581230003555,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.787111111111113e-06,
+      "loss": 0.393,
+      "step": 27480
+    },
+    {
+      "epoch": 0.6107803057234269,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.782666666666667e-06,
+      "loss": 0.3738,
+      "step": 27490
+    },
+    {
+      "epoch": 0.6110024884464984,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.778222222222222e-06,
+      "loss": 0.4097,
+      "step": 27500
+    },
+    {
+      "epoch": 0.6112246711695698,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.773777777777777e-06,
+      "loss": 0.3911,
+      "step": 27510
+    },
+    {
+      "epoch": 0.6114468538926413,
+      "grad_norm": 2.5,
+      "learning_rate": 7.769333333333334e-06,
+      "loss": 0.4197,
+      "step": 27520
+    },
+    {
+      "epoch": 0.6116690366157128,
+      "grad_norm": 2.375,
+      "learning_rate": 7.76488888888889e-06,
+      "loss": 0.3551,
+      "step": 27530
+    },
+    {
+      "epoch": 0.6118912193387842,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.760444444444445e-06,
+      "loss": 0.3914,
+      "step": 27540
+    },
+    {
+      "epoch": 0.6121134020618557,
+      "grad_norm": 3.125,
+      "learning_rate": 7.756e-06,
+      "loss": 0.371,
+      "step": 27550
+    },
+    {
+      "epoch": 0.6123355847849271,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.751555555555556e-06,
+      "loss": 0.3976,
+      "step": 27560
+    },
+    {
+      "epoch": 0.6125577675079986,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.747111111111111e-06,
+      "loss": 0.3953,
+      "step": 27570
+    },
+    {
+      "epoch": 0.61277995023107,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.742666666666668e-06,
+      "loss": 0.3821,
+      "step": 27580
+    },
+    {
+      "epoch": 0.6130021329541415,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.738222222222223e-06,
+      "loss": 0.3977,
+      "step": 27590
+    },
+    {
+      "epoch": 0.6132243156772129,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.73377777777778e-06,
+      "loss": 0.392,
+      "step": 27600
+    },
+    {
+      "epoch": 0.6134464984002844,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.729333333333334e-06,
+      "loss": 0.3432,
+      "step": 27610
+    },
+    {
+      "epoch": 0.6136686811233558,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.724888888888889e-06,
+      "loss": 0.381,
+      "step": 27620
+    },
+    {
+      "epoch": 0.6138908638464273,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.720444444444445e-06,
+      "loss": 0.3911,
+      "step": 27630
+    },
+    {
+      "epoch": 0.6141130465694987,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.716e-06,
+      "loss": 0.3796,
+      "step": 27640
+    },
+    {
+      "epoch": 0.6143352292925702,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.711555555555557e-06,
+      "loss": 0.3854,
+      "step": 27650
+    },
+    {
+      "epoch": 0.6145574120156416,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.707111111111112e-06,
+      "loss": 0.4328,
+      "step": 27660
+    },
+    {
+      "epoch": 0.6147795947387131,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.702666666666667e-06,
+      "loss": 0.3718,
+      "step": 27670
+    },
+    {
+      "epoch": 0.6150017774617845,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.698222222222223e-06,
+      "loss": 0.3647,
+      "step": 27680
+    },
+    {
+      "epoch": 0.615223960184856,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.69377777777778e-06,
+      "loss": 0.4028,
+      "step": 27690
+    },
+    {
+      "epoch": 0.6154461429079274,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.689333333333334e-06,
+      "loss": 0.399,
+      "step": 27700
+    },
+    {
+      "epoch": 0.6156683256309989,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.68488888888889e-06,
+      "loss": 0.3847,
+      "step": 27710
+    },
+    {
+      "epoch": 0.6158905083540704,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.680444444444444e-06,
+      "loss": 0.3966,
+      "step": 27720
+    },
+    {
+      "epoch": 0.6161126910771418,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.676e-06,
+      "loss": 0.4005,
+      "step": 27730
+    },
+    {
+      "epoch": 0.6163348738002133,
+      "grad_norm": 2.25,
+      "learning_rate": 7.671555555555557e-06,
+      "loss": 0.3959,
+      "step": 27740
+    },
+    {
+      "epoch": 0.6165570565232847,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.667111111111112e-06,
+      "loss": 0.3794,
+      "step": 27750
+    },
+    {
+      "epoch": 0.6167792392463562,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.662666666666667e-06,
+      "loss": 0.376,
+      "step": 27760
+    },
+    {
+      "epoch": 0.6170014219694276,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.658222222222222e-06,
+      "loss": 0.388,
+      "step": 27770
+    },
+    {
+      "epoch": 0.6172236046924992,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.653777777777778e-06,
+      "loss": 0.398,
+      "step": 27780
+    },
+    {
+      "epoch": 0.6174457874155705,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.649333333333335e-06,
+      "loss": 0.3529,
+      "step": 27790
+    },
+    {
+      "epoch": 0.617667970138642,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.64488888888889e-06,
+      "loss": 0.3881,
+      "step": 27800
+    },
+    {
+      "epoch": 0.6178901528617134,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.640444444444445e-06,
+      "loss": 0.357,
+      "step": 27810
+    },
+    {
+      "epoch": 0.618112335584785,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.636e-06,
+      "loss": 0.3694,
+      "step": 27820
+    },
+    {
+      "epoch": 0.6183345183078564,
+      "grad_norm": 2.25,
+      "learning_rate": 7.631555555555556e-06,
+      "loss": 0.396,
+      "step": 27830
+    },
+    {
+      "epoch": 0.6185567010309279,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.627111111111112e-06,
+      "loss": 0.4501,
+      "step": 27840
+    },
+    {
+      "epoch": 0.6187788837539993,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.622666666666667e-06,
+      "loss": 0.3959,
+      "step": 27850
+    },
+    {
+      "epoch": 0.6190010664770708,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.618222222222222e-06,
+      "loss": 0.3726,
+      "step": 27860
+    },
+    {
+      "epoch": 0.6192232492001422,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.613777777777779e-06,
+      "loss": 0.3592,
+      "step": 27870
+    },
+    {
+      "epoch": 0.6194454319232137,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.609333333333334e-06,
+      "loss": 0.3812,
+      "step": 27880
+    },
+    {
+      "epoch": 0.6196676146462851,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.604888888888889e-06,
+      "loss": 0.3947,
+      "step": 27890
+    },
+    {
+      "epoch": 0.6198897973693566,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.600444444444445e-06,
+      "loss": 0.3816,
+      "step": 27900
+    },
+    {
+      "epoch": 0.620111980092428,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.5960000000000015e-06,
+      "loss": 0.3579,
+      "step": 27910
+    },
+    {
+      "epoch": 0.6203341628154995,
+      "grad_norm": 2.5,
+      "learning_rate": 7.591555555555556e-06,
+      "loss": 0.3766,
+      "step": 27920
+    },
+    {
+      "epoch": 0.620556345538571,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.587111111111112e-06,
+      "loss": 0.4472,
+      "step": 27930
+    },
+    {
+      "epoch": 0.6207785282616424,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.582666666666667e-06,
+      "loss": 0.3622,
+      "step": 27940
+    },
+    {
+      "epoch": 0.6210007109847139,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.5782222222222225e-06,
+      "loss": 0.3767,
+      "step": 27950
+    },
+    {
+      "epoch": 0.6212228937077853,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.573777777777779e-06,
+      "loss": 0.3808,
+      "step": 27960
+    },
+    {
+      "epoch": 0.6214450764308568,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.569333333333334e-06,
+      "loss": 0.3676,
+      "step": 27970
+    },
+    {
+      "epoch": 0.6216672591539282,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.56488888888889e-06,
+      "loss": 0.3983,
+      "step": 27980
+    },
+    {
+      "epoch": 0.6218894418769997,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.5604444444444445e-06,
+      "loss": 0.4203,
+      "step": 27990
+    },
+    {
+      "epoch": 0.6221116246000711,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.556000000000001e-06,
+      "loss": 0.4059,
+      "step": 28000
+    },
+    {
+      "epoch": 0.6223338073231426,
+      "grad_norm": 2.25,
+      "learning_rate": 7.551555555555557e-06,
+      "loss": 0.3762,
+      "step": 28010
+    },
+    {
+      "epoch": 0.622555990046214,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.5471111111111115e-06,
+      "loss": 0.3727,
+      "step": 28020
+    },
+    {
+      "epoch": 0.6227781727692855,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.542666666666667e-06,
+      "loss": 0.3787,
+      "step": 28030
+    },
+    {
+      "epoch": 0.6230003554923569,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.538222222222222e-06,
+      "loss": 0.3643,
+      "step": 28040
+    },
+    {
+      "epoch": 0.6232225382154284,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.533777777777779e-06,
+      "loss": 0.3842,
+      "step": 28050
+    },
+    {
+      "epoch": 0.6234447209384998,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.529333333333334e-06,
+      "loss": 0.3752,
+      "step": 28060
+    },
+    {
+      "epoch": 0.6236669036615713,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.524888888888889e-06,
+      "loss": 0.3572,
+      "step": 28070
+    },
+    {
+      "epoch": 0.6238890863846427,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.520444444444445e-06,
+      "loss": 0.3834,
+      "step": 28080
+    },
+    {
+      "epoch": 0.6241112691077142,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.516000000000001e-06,
+      "loss": 0.3519,
+      "step": 28090
+    },
+    {
+      "epoch": 0.6243334518307856,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.511555555555556e-06,
+      "loss": 0.3824,
+      "step": 28100
+    },
+    {
+      "epoch": 0.6245556345538571,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.507111111111112e-06,
+      "loss": 0.3752,
+      "step": 28110
+    },
+    {
+      "epoch": 0.6247778172769285,
+      "grad_norm": 2.25,
+      "learning_rate": 7.502666666666667e-06,
+      "loss": 0.3523,
+      "step": 28120
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.4982222222222225e-06,
+      "loss": 0.3832,
+      "step": 28130
+    },
+    {
+      "epoch": 0.6252221827230715,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.493777777777779e-06,
+      "loss": 0.4148,
+      "step": 28140
+    },
+    {
+      "epoch": 0.6254443654461429,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.489333333333334e-06,
+      "loss": 0.366,
+      "step": 28150
+    },
+    {
+      "epoch": 0.6256665481692144,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.4848888888888895e-06,
+      "loss": 0.3789,
+      "step": 28160
+    },
+    {
+      "epoch": 0.6258887308922858,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.480444444444444e-06,
+      "loss": 0.3716,
+      "step": 28170
+    },
+    {
+      "epoch": 0.6261109136153573,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.476000000000001e-06,
+      "loss": 0.3479,
+      "step": 28180
+    },
+    {
+      "epoch": 0.6263330963384287,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.471555555555557e-06,
+      "loss": 0.3754,
+      "step": 28190
+    },
+    {
+      "epoch": 0.6265552790615002,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.4671111111111115e-06,
+      "loss": 0.3692,
+      "step": 28200
+    },
+    {
+      "epoch": 0.6267774617845716,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.462666666666667e-06,
+      "loss": 0.3499,
+      "step": 28210
+    },
+    {
+      "epoch": 0.6269996445076431,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.458222222222224e-06,
+      "loss": 0.3593,
+      "step": 28220
+    },
+    {
+      "epoch": 0.6272218272307145,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.4537777777777785e-06,
+      "loss": 0.3982,
+      "step": 28230
+    },
+    {
+      "epoch": 0.627444009953786,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.449333333333334e-06,
+      "loss": 0.3793,
+      "step": 28240
+    },
+    {
+      "epoch": 0.6276661926768574,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.444888888888889e-06,
+      "loss": 0.3641,
+      "step": 28250
+    },
+    {
+      "epoch": 0.6278883753999289,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.440444444444445e-06,
+      "loss": 0.3766,
+      "step": 28260
+    },
+    {
+      "epoch": 0.6281105581230003,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.436000000000001e-06,
+      "loss": 0.3702,
+      "step": 28270
+    },
+    {
+      "epoch": 0.6283327408460718,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.431555555555556e-06,
+      "loss": 0.4161,
+      "step": 28280
+    },
+    {
+      "epoch": 0.6285549235691432,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.427111111111112e-06,
+      "loss": 0.3946,
+      "step": 28290
+    },
+    {
+      "epoch": 0.6287771062922147,
+      "grad_norm": 2.375,
+      "learning_rate": 7.422666666666667e-06,
+      "loss": 0.4037,
+      "step": 28300
+    },
+    {
+      "epoch": 0.6289992890152861,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.418222222222223e-06,
+      "loss": 0.3448,
+      "step": 28310
+    },
+    {
+      "epoch": 0.6292214717383576,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.413777777777779e-06,
+      "loss": 0.3975,
+      "step": 28320
+    },
+    {
+      "epoch": 0.629443654461429,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.409333333333334e-06,
+      "loss": 0.3917,
+      "step": 28330
+    },
+    {
+      "epoch": 0.6296658371845005,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.4048888888888895e-06,
+      "loss": 0.3971,
+      "step": 28340
+    },
+    {
+      "epoch": 0.629888019907572,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.400444444444444e-06,
+      "loss": 0.3488,
+      "step": 28350
+    },
+    {
+      "epoch": 0.6301102026306434,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.396000000000001e-06,
+      "loss": 0.3703,
+      "step": 28360
+    },
+    {
+      "epoch": 0.6303323853537149,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.3915555555555565e-06,
+      "loss": 0.3957,
+      "step": 28370
+    },
+    {
+      "epoch": 0.6305545680767863,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.387111111111111e-06,
+      "loss": 0.3904,
+      "step": 28380
+    },
+    {
+      "epoch": 0.6307767507998578,
+      "grad_norm": 2.75,
+      "learning_rate": 7.382666666666667e-06,
+      "loss": 0.3894,
+      "step": 28390
+    },
+    {
+      "epoch": 0.6309989335229292,
+      "grad_norm": 2.25,
+      "learning_rate": 7.378222222222224e-06,
+      "loss": 0.3837,
+      "step": 28400
+    },
+    {
+      "epoch": 0.6312211162460007,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.3737777777777785e-06,
+      "loss": 0.3542,
+      "step": 28410
+    },
+    {
+      "epoch": 0.6314432989690721,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.369333333333334e-06,
+      "loss": 0.4,
+      "step": 28420
+    },
+    {
+      "epoch": 0.6316654816921436,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.364888888888889e-06,
+      "loss": 0.3896,
+      "step": 28430
+    },
+    {
+      "epoch": 0.631887664415215,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.360444444444445e-06,
+      "loss": 0.3741,
+      "step": 28440
+    },
+    {
+      "epoch": 0.6321098471382866,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.356000000000001e-06,
+      "loss": 0.386,
+      "step": 28450
+    },
+    {
+      "epoch": 0.632332029861358,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.351555555555556e-06,
+      "loss": 0.372,
+      "step": 28460
+    },
+    {
+      "epoch": 0.6325542125844295,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.347111111111112e-06,
+      "loss": 0.3576,
+      "step": 28470
+    },
+    {
+      "epoch": 0.6327763953075008,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.342666666666667e-06,
+      "loss": 0.3976,
+      "step": 28480
+    },
+    {
+      "epoch": 0.6329985780305724,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.338222222222223e-06,
+      "loss": 0.3562,
+      "step": 28490
+    },
+    {
+      "epoch": 0.6332207607536438,
+      "grad_norm": 2.5,
+      "learning_rate": 7.333777777777779e-06,
+      "loss": 0.3926,
+      "step": 28500
+    },
+    {
+      "epoch": 0.6334429434767153,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.329333333333334e-06,
+      "loss": 0.3995,
+      "step": 28510
+    },
+    {
+      "epoch": 0.6336651261997867,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.324888888888889e-06,
+      "loss": 0.3772,
+      "step": 28520
+    },
+    {
+      "epoch": 0.6338873089228582,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.320444444444446e-06,
+      "loss": 0.3619,
+      "step": 28530
+    },
+    {
+      "epoch": 0.6341094916459296,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.316000000000001e-06,
+      "loss": 0.3671,
+      "step": 28540
+    },
+    {
+      "epoch": 0.6343316743690011,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.3115555555555565e-06,
+      "loss": 0.3554,
+      "step": 28550
+    },
+    {
+      "epoch": 0.6345538570920726,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.307111111111111e-06,
+      "loss": 0.3984,
+      "step": 28560
+    },
+    {
+      "epoch": 0.634776039815144,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.302666666666667e-06,
+      "loss": 0.3698,
+      "step": 28570
+    },
+    {
+      "epoch": 0.6349982225382155,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.2982222222222235e-06,
+      "loss": 0.358,
+      "step": 28580
+    },
+    {
+      "epoch": 0.6352204052612869,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.293777777777778e-06,
+      "loss": 0.3577,
+      "step": 28590
+    },
+    {
+      "epoch": 0.6354425879843584,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.289333333333334e-06,
+      "loss": 0.4029,
+      "step": 28600
+    },
+    {
+      "epoch": 0.6356647707074298,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.284888888888889e-06,
+      "loss": 0.3996,
+      "step": 28610
+    },
+    {
+      "epoch": 0.6358869534305013,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.2804444444444455e-06,
+      "loss": 0.3708,
+      "step": 28620
+    },
+    {
+      "epoch": 0.6361091361535727,
+      "grad_norm": 2.125,
+      "learning_rate": 7.276000000000001e-06,
+      "loss": 0.3876,
+      "step": 28630
+    },
+    {
+      "epoch": 0.6363313188766442,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.271555555555556e-06,
+      "loss": 0.3658,
+      "step": 28640
+    },
+    {
+      "epoch": 0.6365535015997156,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.267111111111112e-06,
+      "loss": 0.3854,
+      "step": 28650
+    },
+    {
+      "epoch": 0.6367756843227871,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.2626666666666665e-06,
+      "loss": 0.3653,
+      "step": 28660
+    },
+    {
+      "epoch": 0.6369978670458585,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.258222222222223e-06,
+      "loss": 0.3889,
+      "step": 28670
+    },
+    {
+      "epoch": 0.63722004976893,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.253777777777779e-06,
+      "loss": 0.3712,
+      "step": 28680
+    },
+    {
+      "epoch": 0.6374422324920014,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.249333333333334e-06,
+      "loss": 0.3837,
+      "step": 28690
+    },
+    {
+      "epoch": 0.6376644152150729,
+      "grad_norm": 2.375,
+      "learning_rate": 7.244888888888889e-06,
+      "loss": 0.3379,
+      "step": 28700
+    },
+    {
+      "epoch": 0.6378865979381443,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.240444444444446e-06,
+      "loss": 0.3851,
+      "step": 28710
+    },
+    {
+      "epoch": 0.6381087806612158,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.236000000000001e-06,
+      "loss": 0.3968,
+      "step": 28720
+    },
+    {
+      "epoch": 0.6383309633842872,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.231555555555556e-06,
+      "loss": 0.3234,
+      "step": 28730
+    },
+    {
+      "epoch": 0.6385531461073587,
+      "grad_norm": 2.125,
+      "learning_rate": 7.227111111111111e-06,
+      "loss": 0.3903,
+      "step": 28740
+    },
+    {
+      "epoch": 0.6387753288304302,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.222666666666667e-06,
+      "loss": 0.359,
+      "step": 28750
+    },
+    {
+      "epoch": 0.6389975115535016,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.2182222222222235e-06,
+      "loss": 0.3954,
+      "step": 28760
+    },
+    {
+      "epoch": 0.6392196942765731,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.213777777777778e-06,
+      "loss": 0.4062,
+      "step": 28770
+    },
+    {
+      "epoch": 0.6394418769996445,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.209333333333334e-06,
+      "loss": 0.3806,
+      "step": 28780
+    },
+    {
+      "epoch": 0.639664059722716,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.204888888888889e-06,
+      "loss": 0.3964,
+      "step": 28790
+    },
+    {
+      "epoch": 0.6398862424457874,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.200444444444445e-06,
+      "loss": 0.3821,
+      "step": 28800
+    },
+    {
+      "epoch": 0.6401084251688589,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.196000000000001e-06,
+      "loss": 0.3998,
+      "step": 28810
+    },
+    {
+      "epoch": 0.6403306078919303,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.191555555555556e-06,
+      "loss": 0.3813,
+      "step": 28820
+    },
+    {
+      "epoch": 0.6405527906150018,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.187111111111112e-06,
+      "loss": 0.437,
+      "step": 28830
+    },
+    {
+      "epoch": 0.6407749733380732,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.182666666666668e-06,
+      "loss": 0.3915,
+      "step": 28840
+    },
+    {
+      "epoch": 0.6409971560611447,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.178222222222223e-06,
+      "loss": 0.372,
+      "step": 28850
+    },
+    {
+      "epoch": 0.6412193387842161,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.173777777777779e-06,
+      "loss": 0.3675,
+      "step": 28860
+    },
+    {
+      "epoch": 0.6414415215072876,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.1693333333333335e-06,
+      "loss": 0.3891,
+      "step": 28870
+    },
+    {
+      "epoch": 0.641663704230359,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.164888888888889e-06,
+      "loss": 0.3852,
+      "step": 28880
+    },
+    {
+      "epoch": 0.6418858869534305,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.160444444444446e-06,
+      "loss": 0.3704,
+      "step": 28890
+    },
+    {
+      "epoch": 0.6421080696765019,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.156000000000001e-06,
+      "loss": 0.4069,
+      "step": 28900
+    },
+    {
+      "epoch": 0.6423302523995734,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.151555555555556e-06,
+      "loss": 0.3886,
+      "step": 28910
+    },
+    {
+      "epoch": 0.6425524351226448,
+      "grad_norm": 2.25,
+      "learning_rate": 7.147111111111111e-06,
+      "loss": 0.3955,
+      "step": 28920
+    },
+    {
+      "epoch": 0.6427746178457163,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.142666666666668e-06,
+      "loss": 0.3679,
+      "step": 28930
+    },
+    {
+      "epoch": 0.6429968005687877,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.138222222222223e-06,
+      "loss": 0.3426,
+      "step": 28940
+    },
+    {
+      "epoch": 0.6432189832918592,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.133777777777778e-06,
+      "loss": 0.378,
+      "step": 28950
+    },
+    {
+      "epoch": 0.6434411660149307,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.129333333333334e-06,
+      "loss": 0.4086,
+      "step": 28960
+    },
+    {
+      "epoch": 0.6436633487380021,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.124888888888889e-06,
+      "loss": 0.3604,
+      "step": 28970
+    },
+    {
+      "epoch": 0.6438855314610736,
+      "grad_norm": 3.0,
+      "learning_rate": 7.120444444444445e-06,
+      "loss": 0.41,
+      "step": 28980
+    },
+    {
+      "epoch": 0.644107714184145,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.116000000000001e-06,
+      "loss": 0.3723,
+      "step": 28990
+    },
+    {
+      "epoch": 0.6443298969072165,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.111555555555556e-06,
+      "loss": 0.4027,
+      "step": 29000
+    },
+    {
+      "epoch": 0.6445520796302879,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.1071111111111115e-06,
+      "loss": 0.3697,
+      "step": 29010
+    },
+    {
+      "epoch": 0.6447742623533594,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.102666666666668e-06,
+      "loss": 0.3818,
+      "step": 29020
+    },
+    {
+      "epoch": 0.6449964450764308,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.098222222222223e-06,
+      "loss": 0.3822,
+      "step": 29030
+    },
+    {
+      "epoch": 0.6452186277995023,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.093777777777779e-06,
+      "loss": 0.358,
+      "step": 29040
+    },
+    {
+      "epoch": 0.6454408105225737,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.0893333333333334e-06,
+      "loss": 0.4125,
+      "step": 29050
+    },
+    {
+      "epoch": 0.6456629932456452,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.084888888888889e-06,
+      "loss": 0.3836,
+      "step": 29060
+    },
+    {
+      "epoch": 0.6458851759687166,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.080444444444446e-06,
+      "loss": 0.3953,
+      "step": 29070
+    },
+    {
+      "epoch": 0.6461073586917881,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.0760000000000005e-06,
+      "loss": 0.4048,
+      "step": 29080
+    },
+    {
+      "epoch": 0.6463295414148595,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.071555555555556e-06,
+      "loss": 0.3697,
+      "step": 29090
+    },
+    {
+      "epoch": 0.646551724137931,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.067111111111111e-06,
+      "loss": 0.3944,
+      "step": 29100
+    },
+    {
+      "epoch": 0.6467739068610024,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.062666666666668e-06,
+      "loss": 0.3524,
+      "step": 29110
+    },
+    {
+      "epoch": 0.646996089584074,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.058222222222223e-06,
+      "loss": 0.3895,
+      "step": 29120
+    },
+    {
+      "epoch": 0.6472182723071453,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.053777777777778e-06,
+      "loss": 0.4205,
+      "step": 29130
+    },
+    {
+      "epoch": 0.6474404550302169,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.049333333333334e-06,
+      "loss": 0.3825,
+      "step": 29140
+    },
+    {
+      "epoch": 0.6476626377532883,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.04488888888889e-06,
+      "loss": 0.4281,
+      "step": 29150
+    },
+    {
+      "epoch": 0.6478848204763598,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.040444444444445e-06,
+      "loss": 0.3488,
+      "step": 29160
+    },
+    {
+      "epoch": 0.6481070031994313,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.036000000000001e-06,
+      "loss": 0.3986,
+      "step": 29170
+    },
+    {
+      "epoch": 0.6483291859225027,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.031555555555556e-06,
+      "loss": 0.3739,
+      "step": 29180
+    },
+    {
+      "epoch": 0.6485513686455742,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.0271111111111114e-06,
+      "loss": 0.3763,
+      "step": 29190
+    },
+    {
+      "epoch": 0.6487735513686456,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.022666666666668e-06,
+      "loss": 0.3906,
+      "step": 29200
+    },
+    {
+      "epoch": 0.6489957340917171,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.018222222222223e-06,
+      "loss": 0.3285,
+      "step": 29210
+    },
+    {
+      "epoch": 0.6492179168147885,
+      "grad_norm": 2.5,
+      "learning_rate": 7.0137777777777785e-06,
+      "loss": 0.4033,
+      "step": 29220
+    },
+    {
+      "epoch": 0.64944009953786,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.009333333333333e-06,
+      "loss": 0.4007,
+      "step": 29230
+    },
+    {
+      "epoch": 0.6496622822609314,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.00488888888889e-06,
+      "loss": 0.4126,
+      "step": 29240
+    },
+    {
+      "epoch": 0.6498844649840029,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.000444444444446e-06,
+      "loss": 0.357,
+      "step": 29250
+    },
+    {
+      "epoch": 0.6501066477070743,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.9960000000000004e-06,
+      "loss": 0.3915,
+      "step": 29260
+    },
+    {
+      "epoch": 0.6503288304301458,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.991555555555556e-06,
+      "loss": 0.3613,
+      "step": 29270
+    },
+    {
+      "epoch": 0.6505510131532172,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.987111111111111e-06,
+      "loss": 0.3869,
+      "step": 29280
+    },
+    {
+      "epoch": 0.6507731958762887,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.9826666666666675e-06,
+      "loss": 0.3879,
+      "step": 29290
+    },
+    {
+      "epoch": 0.6509953785993601,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.978222222222223e-06,
+      "loss": 0.3714,
+      "step": 29300
+    },
+    {
+      "epoch": 0.6512175613224316,
+      "grad_norm": 2.375,
+      "learning_rate": 6.973777777777778e-06,
+      "loss": 0.426,
+      "step": 29310
+    },
+    {
+      "epoch": 0.651439744045503,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.969333333333334e-06,
+      "loss": 0.3995,
+      "step": 29320
+    },
+    {
+      "epoch": 0.6516619267685745,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.96488888888889e-06,
+      "loss": 0.3908,
+      "step": 29330
+    },
+    {
+      "epoch": 0.6518841094916459,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.960444444444445e-06,
+      "loss": 0.3951,
+      "step": 29340
+    },
+    {
+      "epoch": 0.6521062922147174,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.956000000000001e-06,
+      "loss": 0.3983,
+      "step": 29350
+    },
+    {
+      "epoch": 0.6523284749377888,
+      "grad_norm": 2.625,
+      "learning_rate": 6.951555555555556e-06,
+      "loss": 0.3974,
+      "step": 29360
+    },
+    {
+      "epoch": 0.6525506576608603,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.947111111111111e-06,
+      "loss": 0.3993,
+      "step": 29370
+    },
+    {
+      "epoch": 0.6527728403839318,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.942666666666668e-06,
+      "loss": 0.3831,
+      "step": 29380
+    },
+    {
+      "epoch": 0.6529950231070032,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.938222222222223e-06,
+      "loss": 0.412,
+      "step": 29390
+    },
+    {
+      "epoch": 0.6532172058300747,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.9337777777777784e-06,
+      "loss": 0.3932,
+      "step": 29400
+    },
+    {
+      "epoch": 0.6534393885531461,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.929333333333333e-06,
+      "loss": 0.3707,
+      "step": 29410
+    },
+    {
+      "epoch": 0.6536615712762176,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.92488888888889e-06,
+      "loss": 0.3733,
+      "step": 29420
+    },
+    {
+      "epoch": 0.653883753999289,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.9204444444444455e-06,
+      "loss": 0.3707,
+      "step": 29430
+    },
+    {
+      "epoch": 0.6541059367223605,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.916e-06,
+      "loss": 0.3863,
+      "step": 29440
+    },
+    {
+      "epoch": 0.6543281194454319,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.911555555555556e-06,
+      "loss": 0.3652,
+      "step": 29450
+    },
+    {
+      "epoch": 0.6545503021685034,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.907111111111113e-06,
+      "loss": 0.3804,
+      "step": 29460
+    },
+    {
+      "epoch": 0.6547724848915748,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.9026666666666674e-06,
+      "loss": 0.3855,
+      "step": 29470
+    },
+    {
+      "epoch": 0.6549946676146463,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.898222222222223e-06,
+      "loss": 0.3665,
+      "step": 29480
+    },
+    {
+      "epoch": 0.6552168503377177,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.893777777777778e-06,
+      "loss": 0.3597,
+      "step": 29490
+    },
+    {
+      "epoch": 0.6554390330607892,
+      "grad_norm": 3.25,
+      "learning_rate": 6.889333333333334e-06,
+      "loss": 0.358,
+      "step": 29500
+    },
+    {
+      "epoch": 0.6556612157838606,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.88488888888889e-06,
+      "loss": 0.3633,
+      "step": 29510
+    },
+    {
+      "epoch": 0.6558833985069321,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.880444444444445e-06,
+      "loss": 0.3802,
+      "step": 29520
+    },
+    {
+      "epoch": 0.6561055812300035,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.876000000000001e-06,
+      "loss": 0.3688,
+      "step": 29530
+    },
+    {
+      "epoch": 0.656327763953075,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.871555555555556e-06,
+      "loss": 0.3919,
+      "step": 29540
+    },
+    {
+      "epoch": 0.6565499466761464,
+      "grad_norm": 2.375,
+      "learning_rate": 6.867111111111112e-06,
+      "loss": 0.394,
+      "step": 29550
+    },
+    {
+      "epoch": 0.6567721293992179,
+      "grad_norm": 2.375,
+      "learning_rate": 6.862666666666668e-06,
+      "loss": 0.3783,
+      "step": 29560
+    },
+    {
+      "epoch": 0.6569943121222894,
+      "grad_norm": 2.375,
+      "learning_rate": 6.858222222222223e-06,
+      "loss": 0.3726,
+      "step": 29570
+    },
+    {
+      "epoch": 0.6572164948453608,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.853777777777778e-06,
+      "loss": 0.3872,
+      "step": 29580
+    },
+    {
+      "epoch": 0.6574386775684323,
+      "grad_norm": 2.5,
+      "learning_rate": 6.849333333333333e-06,
+      "loss": 0.3658,
+      "step": 29590
+    },
+    {
+      "epoch": 0.6576608602915037,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.84488888888889e-06,
+      "loss": 0.3468,
+      "step": 29600
+    },
+    {
+      "epoch": 0.6578830430145752,
+      "grad_norm": 2.625,
+      "learning_rate": 6.8404444444444454e-06,
+      "loss": 0.4079,
+      "step": 29610
+    },
+    {
+      "epoch": 0.6581052257376466,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.836e-06,
+      "loss": 0.3706,
+      "step": 29620
+    },
+    {
+      "epoch": 0.6583274084607181,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.831555555555556e-06,
+      "loss": 0.3939,
+      "step": 29630
+    },
+    {
+      "epoch": 0.6585495911837895,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.8271111111111125e-06,
+      "loss": 0.4351,
+      "step": 29640
+    },
+    {
+      "epoch": 0.658771773906861,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.822666666666667e-06,
+      "loss": 0.4096,
+      "step": 29650
+    },
+    {
+      "epoch": 0.6589939566299324,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.818222222222223e-06,
+      "loss": 0.3693,
+      "step": 29660
+    },
+    {
+      "epoch": 0.6592161393530039,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.813777777777778e-06,
+      "loss": 0.3968,
+      "step": 29670
+    },
+    {
+      "epoch": 0.6594383220760753,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.809333333333334e-06,
+      "loss": 0.3775,
+      "step": 29680
+    },
+    {
+      "epoch": 0.6596605047991468,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.80488888888889e-06,
+      "loss": 0.3893,
+      "step": 29690
+    },
+    {
+      "epoch": 0.6598826875222182,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.800444444444445e-06,
+      "loss": 0.3933,
+      "step": 29700
+    },
+    {
+      "epoch": 0.6601048702452897,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.796000000000001e-06,
+      "loss": 0.412,
+      "step": 29710
+    },
+    {
+      "epoch": 0.6603270529683611,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.7915555555555555e-06,
+      "loss": 0.3521,
+      "step": 29720
+    },
+    {
+      "epoch": 0.6605492356914326,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.787111111111112e-06,
+      "loss": 0.3412,
+      "step": 29730
+    },
+    {
+      "epoch": 0.660771418414504,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.782666666666668e-06,
+      "loss": 0.367,
+      "step": 29740
+    },
+    {
+      "epoch": 0.6609936011375755,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.778222222222223e-06,
+      "loss": 0.3539,
+      "step": 29750
+    },
+    {
+      "epoch": 0.6612157838606469,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.773777777777778e-06,
+      "loss": 0.3605,
+      "step": 29760
+    },
+    {
+      "epoch": 0.6614379665837185,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.769333333333335e-06,
+      "loss": 0.3604,
+      "step": 29770
+    },
+    {
+      "epoch": 0.66166014930679,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.76488888888889e-06,
+      "loss": 0.3803,
+      "step": 29780
+    },
+    {
+      "epoch": 0.6618823320298614,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.760444444444445e-06,
+      "loss": 0.4003,
+      "step": 29790
+    },
+    {
+      "epoch": 0.6621045147529329,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.756e-06,
+      "loss": 0.3609,
+      "step": 29800
+    },
+    {
+      "epoch": 0.6623266974760043,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.751555555555556e-06,
+      "loss": 0.3755,
+      "step": 29810
+    },
+    {
+      "epoch": 0.6625488801990758,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.7471111111111124e-06,
+      "loss": 0.3864,
+      "step": 29820
+    },
+    {
+      "epoch": 0.6627710629221472,
+      "grad_norm": 2.625,
+      "learning_rate": 6.742666666666667e-06,
+      "loss": 0.3745,
+      "step": 29830
+    },
+    {
+      "epoch": 0.6629932456452187,
+      "grad_norm": 2.125,
+      "learning_rate": 6.738222222222223e-06,
+      "loss": 0.3283,
+      "step": 29840
+    },
+    {
+      "epoch": 0.6632154283682901,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.733777777777778e-06,
+      "loss": 0.3893,
+      "step": 29850
+    },
+    {
+      "epoch": 0.6634376110913616,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.729333333333334e-06,
+      "loss": 0.3899,
+      "step": 29860
+    },
+    {
+      "epoch": 0.663659793814433,
+      "grad_norm": 3.125,
+      "learning_rate": 6.72488888888889e-06,
+      "loss": 0.3545,
+      "step": 29870
+    },
+    {
+      "epoch": 0.6638819765375045,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.720444444444445e-06,
+      "loss": 0.357,
+      "step": 29880
+    },
+    {
+      "epoch": 0.6641041592605759,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.716000000000001e-06,
+      "loss": 0.369,
+      "step": 29890
+    },
+    {
+      "epoch": 0.6643263419836474,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.7115555555555554e-06,
+      "loss": 0.4066,
+      "step": 29900
+    },
+    {
+      "epoch": 0.6645485247067188,
+      "grad_norm": 2.375,
+      "learning_rate": 6.707111111111112e-06,
+      "loss": 0.3894,
+      "step": 29910
+    },
+    {
+      "epoch": 0.6647707074297903,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.702666666666668e-06,
+      "loss": 0.3918,
+      "step": 29920
+    },
+    {
+      "epoch": 0.6649928901528617,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.6982222222222225e-06,
+      "loss": 0.402,
+      "step": 29930
+    },
+    {
+      "epoch": 0.6652150728759332,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.693777777777778e-06,
+      "loss": 0.3942,
+      "step": 29940
+    },
+    {
+      "epoch": 0.6654372555990046,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.689333333333335e-06,
+      "loss": 0.3799,
+      "step": 29950
+    },
+    {
+      "epoch": 0.6656594383220761,
+      "grad_norm": 2.375,
+      "learning_rate": 6.68488888888889e-06,
+      "loss": 0.4013,
+      "step": 29960
+    },
+    {
+      "epoch": 0.6658816210451475,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.680444444444445e-06,
+      "loss": 0.406,
+      "step": 29970
+    },
+    {
+      "epoch": 0.666103803768219,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.676e-06,
+      "loss": 0.3626,
+      "step": 29980
+    },
+    {
+      "epoch": 0.6663259864912905,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.671555555555556e-06,
+      "loss": 0.4231,
+      "step": 29990
+    },
+    {
+      "epoch": 0.6665481692143619,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.667111111111112e-06,
+      "loss": 0.4099,
+      "step": 30000
+    },
+    {
+      "epoch": 0.6667703519374334,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.662666666666667e-06,
+      "loss": 0.3699,
+      "step": 30010
+    },
+    {
+      "epoch": 0.6669925346605048,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.658222222222223e-06,
+      "loss": 0.3677,
+      "step": 30020
+    },
+    {
+      "epoch": 0.6672147173835763,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.653777777777778e-06,
+      "loss": 0.3877,
+      "step": 30030
+    },
+    {
+      "epoch": 0.6674369001066477,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.649333333333334e-06,
+      "loss": 0.3941,
+      "step": 30040
+    },
+    {
+      "epoch": 0.6676590828297192,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.64488888888889e-06,
+      "loss": 0.3895,
+      "step": 30050
+    },
+    {
+      "epoch": 0.6678812655527906,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.640444444444445e-06,
+      "loss": 0.4168,
+      "step": 30060
+    },
+    {
+      "epoch": 0.6681034482758621,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.6360000000000005e-06,
+      "loss": 0.3704,
+      "step": 30070
+    },
+    {
+      "epoch": 0.6683256309989335,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.631555555555557e-06,
+      "loss": 0.3571,
+      "step": 30080
+    },
+    {
+      "epoch": 0.668547813722005,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.627111111111112e-06,
+      "loss": 0.3695,
+      "step": 30090
+    },
+    {
+      "epoch": 0.6687699964450764,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.622666666666668e-06,
+      "loss": 0.3893,
+      "step": 30100
+    },
+    {
+      "epoch": 0.6689921791681479,
+      "grad_norm": 2.5,
+      "learning_rate": 6.618222222222222e-06,
+      "loss": 0.3676,
+      "step": 30110
+    },
+    {
+      "epoch": 0.6692143618912193,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.613777777777778e-06,
+      "loss": 0.3457,
+      "step": 30120
+    },
+    {
+      "epoch": 0.6694365446142908,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.609333333333335e-06,
+      "loss": 0.388,
+      "step": 30130
+    },
+    {
+      "epoch": 0.6696587273373622,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.6048888888888895e-06,
+      "loss": 0.3966,
+      "step": 30140
+    },
+    {
+      "epoch": 0.6698809100604337,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.600444444444445e-06,
+      "loss": 0.3633,
+      "step": 30150
+    },
+    {
+      "epoch": 0.6701030927835051,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.596e-06,
+      "loss": 0.3536,
+      "step": 30160
+    },
+    {
+      "epoch": 0.6703252755065766,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.5915555555555566e-06,
+      "loss": 0.3891,
+      "step": 30170
+    },
+    {
+      "epoch": 0.670547458229648,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.587111111111112e-06,
+      "loss": 0.379,
+      "step": 30180
+    },
+    {
+      "epoch": 0.6707696409527195,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.582666666666667e-06,
+      "loss": 0.379,
+      "step": 30190
+    },
+    {
+      "epoch": 0.670991823675791,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.578222222222223e-06,
+      "loss": 0.3898,
+      "step": 30200
+    },
+    {
+      "epoch": 0.6712140063988624,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.573777777777778e-06,
+      "loss": 0.3695,
+      "step": 30210
+    },
+    {
+      "epoch": 0.6714361891219339,
+      "grad_norm": 2.875,
+      "learning_rate": 6.569333333333334e-06,
+      "loss": 0.3822,
+      "step": 30220
+    },
+    {
+      "epoch": 0.6716583718450053,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.56488888888889e-06,
+      "loss": 0.3455,
+      "step": 30230
+    },
+    {
+      "epoch": 0.6718805545680768,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.560444444444445e-06,
+      "loss": 0.3591,
+      "step": 30240
+    },
+    {
+      "epoch": 0.6721027372911482,
+      "grad_norm": 2.75,
+      "learning_rate": 6.556e-06,
+      "loss": 0.3789,
+      "step": 30250
+    },
+    {
+      "epoch": 0.6723249200142197,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.551555555555557e-06,
+      "loss": 0.3421,
+      "step": 30260
+    },
+    {
+      "epoch": 0.6725471027372911,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.547111111111112e-06,
+      "loss": 0.3677,
+      "step": 30270
+    },
+    {
+      "epoch": 0.6727692854603626,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.5426666666666675e-06,
+      "loss": 0.3675,
+      "step": 30280
+    },
+    {
+      "epoch": 0.672991468183434,
+      "grad_norm": 2.375,
+      "learning_rate": 6.538222222222222e-06,
+      "loss": 0.3698,
+      "step": 30290
+    },
+    {
+      "epoch": 0.6732136509065055,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.533777777777778e-06,
+      "loss": 0.3691,
+      "step": 30300
+    },
+    {
+      "epoch": 0.6734358336295769,
+      "grad_norm": 2.375,
+      "learning_rate": 6.5293333333333346e-06,
+      "loss": 0.3903,
+      "step": 30310
+    },
+    {
+      "epoch": 0.6736580163526484,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.524888888888889e-06,
+      "loss": 0.3781,
+      "step": 30320
+    },
+    {
+      "epoch": 0.6738801990757198,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.520444444444445e-06,
+      "loss": 0.3968,
+      "step": 30330
+    },
+    {
+      "epoch": 0.6741023817987913,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.516e-06,
+      "loss": 0.3784,
+      "step": 30340
+    },
+    {
+      "epoch": 0.6743245645218627,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.5115555555555565e-06,
+      "loss": 0.3678,
+      "step": 30350
+    },
+    {
+      "epoch": 0.6745467472449342,
+      "grad_norm": 2.375,
+      "learning_rate": 6.507111111111112e-06,
+      "loss": 0.3828,
+      "step": 30360
+    },
+    {
+      "epoch": 0.6747689299680056,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.502666666666667e-06,
+      "loss": 0.3999,
+      "step": 30370
+    },
+    {
+      "epoch": 0.6749911126910771,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.498222222222223e-06,
+      "loss": 0.3861,
+      "step": 30380
+    },
+    {
+      "epoch": 0.6752132954141487,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.493777777777779e-06,
+      "loss": 0.3735,
+      "step": 30390
+    },
+    {
+      "epoch": 0.67543547813722,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.489333333333334e-06,
+      "loss": 0.3889,
+      "step": 30400
+    },
+    {
+      "epoch": 0.6756576608602916,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.48488888888889e-06,
+      "loss": 0.3923,
+      "step": 30410
+    },
+    {
+      "epoch": 0.675879843583363,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.480444444444445e-06,
+      "loss": 0.3722,
+      "step": 30420
+    },
+    {
+      "epoch": 0.6761020263064345,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.476e-06,
+      "loss": 0.4002,
+      "step": 30430
+    },
+    {
+      "epoch": 0.6763242090295059,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.471555555555557e-06,
+      "loss": 0.4155,
+      "step": 30440
+    },
+    {
+      "epoch": 0.6765463917525774,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.467111111111112e-06,
+      "loss": 0.4074,
+      "step": 30450
+    },
+    {
+      "epoch": 0.6767685744756488,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.462666666666667e-06,
+      "loss": 0.3554,
+      "step": 30460
+    },
+    {
+      "epoch": 0.6769907571987203,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.458222222222222e-06,
+      "loss": 0.3737,
+      "step": 30470
+    },
+    {
+      "epoch": 0.6772129399217917,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.453777777777779e-06,
+      "loss": 0.3717,
+      "step": 30480
+    },
+    {
+      "epoch": 0.6774351226448632,
+      "grad_norm": 2.21875,
+      "learning_rate": 6.4493333333333345e-06,
+      "loss": 0.3333,
+      "step": 30490
+    },
+    {
+      "epoch": 0.6776573053679346,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.444888888888889e-06,
+      "loss": 0.405,
+      "step": 30500
+    },
+    {
+      "epoch": 0.6778794880910061,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.440444444444445e-06,
+      "loss": 0.3492,
+      "step": 30510
+    },
+    {
+      "epoch": 0.6781016708140775,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.436e-06,
+      "loss": 0.4077,
+      "step": 30520
+    },
+    {
+      "epoch": 0.678323853537149,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.431555555555556e-06,
+      "loss": 0.3826,
+      "step": 30530
+    },
+    {
+      "epoch": 0.6785460362602204,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.427111111111112e-06,
+      "loss": 0.3546,
+      "step": 30540
+    },
+    {
+      "epoch": 0.6787682189832919,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.422666666666667e-06,
+      "loss": 0.3804,
+      "step": 30550
+    },
+    {
+      "epoch": 0.6789904017063633,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.418222222222223e-06,
+      "loss": 0.4225,
+      "step": 30560
+    },
+    {
+      "epoch": 0.6792125844294348,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.413777777777779e-06,
+      "loss": 0.3724,
+      "step": 30570
+    },
+    {
+      "epoch": 0.6794347671525062,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.409333333333334e-06,
+      "loss": 0.3994,
+      "step": 30580
+    },
+    {
+      "epoch": 0.6796569498755777,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.40488888888889e-06,
+      "loss": 0.3607,
+      "step": 30590
+    },
+    {
+      "epoch": 0.6798791325986492,
+      "grad_norm": 1.921875,
+      "learning_rate": 6.4004444444444446e-06,
+      "loss": 0.393,
+      "step": 30600
+    },
+    {
+      "epoch": 0.6801013153217206,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.396e-06,
+      "loss": 0.3754,
+      "step": 30610
+    },
+    {
+      "epoch": 0.6803234980447921,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.391555555555557e-06,
+      "loss": 0.4055,
+      "step": 30620
+    },
+    {
+      "epoch": 0.6805456807678635,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.387111111111112e-06,
+      "loss": 0.3717,
+      "step": 30630
+    },
+    {
+      "epoch": 0.680767863490935,
+      "grad_norm": 2.375,
+      "learning_rate": 6.382666666666667e-06,
+      "loss": 0.3224,
+      "step": 30640
+    },
+    {
+      "epoch": 0.6809900462140064,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.378222222222222e-06,
+      "loss": 0.3843,
+      "step": 30650
+    },
+    {
+      "epoch": 0.6812122289370779,
+      "grad_norm": 2.0,
+      "learning_rate": 6.373777777777779e-06,
+      "loss": 0.4126,
+      "step": 30660
+    },
+    {
+      "epoch": 0.6814344116601493,
+      "grad_norm": 2.375,
+      "learning_rate": 6.369333333333334e-06,
+      "loss": 0.3901,
+      "step": 30670
+    },
+    {
+      "epoch": 0.6816565943832208,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.364888888888889e-06,
+      "loss": 0.3607,
+      "step": 30680
+    },
+    {
+      "epoch": 0.6818787771062922,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.360444444444445e-06,
+      "loss": 0.3835,
+      "step": 30690
+    },
+    {
+      "epoch": 0.6821009598293637,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.356000000000001e-06,
+      "loss": 0.3999,
+      "step": 30700
+    },
+    {
+      "epoch": 0.6823231425524351,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.351555555555556e-06,
+      "loss": 0.3683,
+      "step": 30710
+    },
+    {
+      "epoch": 0.6825453252755066,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.347111111111112e-06,
+      "loss": 0.3753,
+      "step": 30720
+    },
+    {
+      "epoch": 0.682767507998578,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.342666666666667e-06,
+      "loss": 0.375,
+      "step": 30730
+    },
+    {
+      "epoch": 0.6829896907216495,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.3382222222222226e-06,
+      "loss": 0.3567,
+      "step": 30740
+    },
+    {
+      "epoch": 0.6832118734447209,
+      "grad_norm": 2.625,
+      "learning_rate": 6.333777777777779e-06,
+      "loss": 0.3455,
+      "step": 30750
+    },
+    {
+      "epoch": 0.6834340561677924,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.329333333333334e-06,
+      "loss": 0.3536,
+      "step": 30760
+    },
+    {
+      "epoch": 0.6836562388908638,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.32488888888889e-06,
+      "loss": 0.3903,
+      "step": 30770
+    },
+    {
+      "epoch": 0.6838784216139353,
+      "grad_norm": 2.625,
+      "learning_rate": 6.3204444444444445e-06,
+      "loss": 0.402,
+      "step": 30780
+    },
+    {
+      "epoch": 0.6841006043370067,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.316000000000001e-06,
+      "loss": 0.3951,
+      "step": 30790
+    },
+    {
+      "epoch": 0.6843227870600782,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.311555555555557e-06,
+      "loss": 0.3408,
+      "step": 30800
+    },
+    {
+      "epoch": 0.6845449697831497,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.3071111111111116e-06,
+      "loss": 0.3605,
+      "step": 30810
+    },
+    {
+      "epoch": 0.6847671525062211,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.302666666666667e-06,
+      "loss": 0.4112,
+      "step": 30820
+    },
+    {
+      "epoch": 0.6849893352292926,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.298222222222222e-06,
+      "loss": 0.379,
+      "step": 30830
+    },
+    {
+      "epoch": 0.685211517952364,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.293777777777779e-06,
+      "loss": 0.4127,
+      "step": 30840
+    },
+    {
+      "epoch": 0.6854337006754355,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.289333333333334e-06,
+      "loss": 0.4008,
+      "step": 30850
+    },
+    {
+      "epoch": 0.6856558833985069,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.284888888888889e-06,
+      "loss": 0.4076,
+      "step": 30860
+    },
+    {
+      "epoch": 0.6858780661215784,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.280444444444445e-06,
+      "loss": 0.3614,
+      "step": 30870
+    },
+    {
+      "epoch": 0.6861002488446498,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.2760000000000006e-06,
+      "loss": 0.357,
+      "step": 30880
+    },
+    {
+      "epoch": 0.6863224315677213,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.271555555555556e-06,
+      "loss": 0.3925,
+      "step": 30890
+    },
+    {
+      "epoch": 0.6865446142907927,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.267111111111112e-06,
+      "loss": 0.393,
+      "step": 30900
+    },
+    {
+      "epoch": 0.6867667970138642,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.262666666666667e-06,
+      "loss": 0.4081,
+      "step": 30910
+    },
+    {
+      "epoch": 0.6869889797369356,
+      "grad_norm": 2.125,
+      "learning_rate": 6.2582222222222225e-06,
+      "loss": 0.4138,
+      "step": 30920
+    },
+    {
+      "epoch": 0.6872111624600071,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.253777777777779e-06,
+      "loss": 0.38,
+      "step": 30930
+    },
+    {
+      "epoch": 0.6874333451830785,
+      "grad_norm": 3.125,
+      "learning_rate": 6.249333333333334e-06,
+      "loss": 0.397,
+      "step": 30940
+    },
+    {
+      "epoch": 0.68765552790615,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.2448888888888896e-06,
+      "loss": 0.4009,
+      "step": 30950
+    },
+    {
+      "epoch": 0.6878777106292214,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.240444444444444e-06,
+      "loss": 0.3613,
+      "step": 30960
+    },
+    {
+      "epoch": 0.6880998933522929,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.236000000000001e-06,
+      "loss": 0.4026,
+      "step": 30970
+    },
+    {
+      "epoch": 0.6883220760753643,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.231555555555557e-06,
+      "loss": 0.3755,
+      "step": 30980
+    },
+    {
+      "epoch": 0.6885442587984358,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.2271111111111115e-06,
+      "loss": 0.4011,
+      "step": 30990
+    },
+    {
+      "epoch": 0.6887664415215072,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.222666666666667e-06,
+      "loss": 0.3851,
+      "step": 31000
+    },
+    {
+      "epoch": 0.6889886242445787,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.218222222222223e-06,
+      "loss": 0.3872,
+      "step": 31010
+    },
+    {
+      "epoch": 0.6892108069676502,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.2137777777777786e-06,
+      "loss": 0.3654,
+      "step": 31020
+    },
+    {
+      "epoch": 0.6894329896907216,
+      "grad_norm": 2.25,
+      "learning_rate": 6.209333333333334e-06,
+      "loss": 0.3944,
+      "step": 31030
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.204888888888889e-06,
+      "loss": 0.3752,
+      "step": 31040
+    },
+    {
+      "epoch": 0.6898773551368645,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.200444444444445e-06,
+      "loss": 0.3726,
+      "step": 31050
+    },
+    {
+      "epoch": 0.690099537859936,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.196000000000001e-06,
+      "loss": 0.3663,
+      "step": 31060
+    },
+    {
+      "epoch": 0.6903217205830074,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.191555555555556e-06,
+      "loss": 0.3583,
+      "step": 31070
+    },
+    {
+      "epoch": 0.690543903306079,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.187111111111112e-06,
+      "loss": 0.3539,
+      "step": 31080
+    },
+    {
+      "epoch": 0.6907660860291504,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.182666666666667e-06,
+      "loss": 0.4004,
+      "step": 31090
+    },
+    {
+      "epoch": 0.6909882687522219,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.178222222222223e-06,
+      "loss": 0.3762,
+      "step": 31100
+    },
+    {
+      "epoch": 0.6912104514752933,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.173777777777779e-06,
+      "loss": 0.3915,
+      "step": 31110
+    },
+    {
+      "epoch": 0.6914326341983648,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.169333333333334e-06,
+      "loss": 0.4161,
+      "step": 31120
+    },
+    {
+      "epoch": 0.6916548169214362,
+      "grad_norm": 1.9453125,
+      "learning_rate": 6.1648888888888895e-06,
+      "loss": 0.333,
+      "step": 31130
+    },
+    {
+      "epoch": 0.6918769996445077,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.160444444444444e-06,
+      "loss": 0.3612,
+      "step": 31140
+    },
+    {
+      "epoch": 0.6920991823675791,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.156000000000001e-06,
+      "loss": 0.3741,
+      "step": 31150
+    },
+    {
+      "epoch": 0.6923213650906506,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.1515555555555566e-06,
+      "loss": 0.3687,
+      "step": 31160
+    },
+    {
+      "epoch": 0.692543547813722,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.147111111111111e-06,
+      "loss": 0.4184,
+      "step": 31170
+    },
+    {
+      "epoch": 0.6927657305367935,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.142666666666667e-06,
+      "loss": 0.3746,
+      "step": 31180
+    },
+    {
+      "epoch": 0.6929879132598649,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.138222222222223e-06,
+      "loss": 0.4053,
+      "step": 31190
+    },
+    {
+      "epoch": 0.6932100959829364,
+      "grad_norm": 1.984375,
+      "learning_rate": 6.1337777777777785e-06,
+      "loss": 0.3382,
+      "step": 31200
+    },
+    {
+      "epoch": 0.6934322787060079,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.129333333333334e-06,
+      "loss": 0.4028,
+      "step": 31210
+    },
+    {
+      "epoch": 0.6936544614290793,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.124888888888889e-06,
+      "loss": 0.4148,
+      "step": 31220
+    },
+    {
+      "epoch": 0.6938766441521508,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.120444444444445e-06,
+      "loss": 0.3841,
+      "step": 31230
+    },
+    {
+      "epoch": 0.6940988268752222,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.116000000000001e-06,
+      "loss": 0.4052,
+      "step": 31240
+    },
+    {
+      "epoch": 0.6943210095982937,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.111555555555556e-06,
+      "loss": 0.3667,
+      "step": 31250
+    },
+    {
+      "epoch": 0.6945431923213651,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.107111111111112e-06,
+      "loss": 0.3724,
+      "step": 31260
+    },
+    {
+      "epoch": 0.6947653750444366,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.102666666666667e-06,
+      "loss": 0.3665,
+      "step": 31270
+    },
+    {
+      "epoch": 0.694987557767508,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.098222222222223e-06,
+      "loss": 0.3945,
+      "step": 31280
+    },
+    {
+      "epoch": 0.6952097404905795,
+      "grad_norm": 2.75,
+      "learning_rate": 6.093777777777779e-06,
+      "loss": 0.3324,
+      "step": 31290
+    },
+    {
+      "epoch": 0.6954319232136509,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.089333333333334e-06,
+      "loss": 0.3721,
+      "step": 31300
+    },
+    {
+      "epoch": 0.6956541059367224,
+      "grad_norm": 2.375,
+      "learning_rate": 6.084888888888889e-06,
+      "loss": 0.3802,
+      "step": 31310
+    },
+    {
+      "epoch": 0.6958762886597938,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.080444444444445e-06,
+      "loss": 0.3598,
+      "step": 31320
+    },
+    {
+      "epoch": 0.6960984713828653,
+      "grad_norm": 2.375,
+      "learning_rate": 6.076000000000001e-06,
+      "loss": 0.3883,
+      "step": 31330
+    },
+    {
+      "epoch": 0.6963206541059367,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.0715555555555565e-06,
+      "loss": 0.3485,
+      "step": 31340
+    },
+    {
+      "epoch": 0.6965428368290082,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.067111111111111e-06,
+      "loss": 0.4005,
+      "step": 31350
+    },
+    {
+      "epoch": 0.6967650195520796,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.062666666666667e-06,
+      "loss": 0.3968,
+      "step": 31360
+    },
+    {
+      "epoch": 0.6969872022751511,
+      "grad_norm": 2.25,
+      "learning_rate": 6.058222222222223e-06,
+      "loss": 0.3712,
+      "step": 31370
+    },
+    {
+      "epoch": 0.6972093849982225,
+      "grad_norm": 2.5,
+      "learning_rate": 6.053777777777778e-06,
+      "loss": 0.4079,
+      "step": 31380
+    },
+    {
+      "epoch": 0.697431567721294,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.049333333333334e-06,
+      "loss": 0.369,
+      "step": 31390
+    },
+    {
+      "epoch": 0.6976537504443654,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.044888888888889e-06,
+      "loss": 0.3667,
+      "step": 31400
+    },
+    {
+      "epoch": 0.6978759331674369,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.0404444444444455e-06,
+      "loss": 0.3936,
+      "step": 31410
+    },
+    {
+      "epoch": 0.6980981158905084,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.036000000000001e-06,
+      "loss": 0.415,
+      "step": 31420
+    },
+    {
+      "epoch": 0.6983202986135798,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.031555555555556e-06,
+      "loss": 0.3716,
+      "step": 31430
+    },
+    {
+      "epoch": 0.6985424813366513,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.027111111111112e-06,
+      "loss": 0.3485,
+      "step": 31440
+    },
+    {
+      "epoch": 0.6987646640597227,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.0226666666666665e-06,
+      "loss": 0.3572,
+      "step": 31450
+    },
+    {
+      "epoch": 0.6989868467827942,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.018222222222223e-06,
+      "loss": 0.3455,
+      "step": 31460
+    },
+    {
+      "epoch": 0.6992090295058656,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.013777777777779e-06,
+      "loss": 0.38,
+      "step": 31470
+    },
+    {
+      "epoch": 0.6994312122289371,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.009333333333334e-06,
+      "loss": 0.3581,
+      "step": 31480
+    },
+    {
+      "epoch": 0.6996533949520085,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.004888888888889e-06,
+      "loss": 0.4038,
+      "step": 31490
+    },
+    {
+      "epoch": 0.69987557767508,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.000444444444445e-06,
+      "loss": 0.3667,
+      "step": 31500
+    },
+    {
+      "epoch": 0.7000977603981514,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.996000000000001e-06,
+      "loss": 0.3834,
+      "step": 31510
+    },
+    {
+      "epoch": 0.7003199431212229,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.991555555555556e-06,
+      "loss": 0.364,
+      "step": 31520
+    },
+    {
+      "epoch": 0.7005421258442943,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.987111111111111e-06,
+      "loss": 0.4001,
+      "step": 31530
+    },
+    {
+      "epoch": 0.7007643085673658,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.982666666666667e-06,
+      "loss": 0.3787,
+      "step": 31540
+    },
+    {
+      "epoch": 0.7009864912904372,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.978222222222223e-06,
+      "loss": 0.3874,
+      "step": 31550
+    },
+    {
+      "epoch": 0.7012086740135087,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.973777777777778e-06,
+      "loss": 0.3811,
+      "step": 31560
+    },
+    {
+      "epoch": 0.7014308567365801,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.969333333333334e-06,
+      "loss": 0.4122,
+      "step": 31570
+    },
+    {
+      "epoch": 0.7016530394596516,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.964888888888889e-06,
+      "loss": 0.4028,
+      "step": 31580
+    },
+    {
+      "epoch": 0.701875222182723,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.960444444444445e-06,
+      "loss": 0.4028,
+      "step": 31590
+    },
+    {
+      "epoch": 0.7020974049057945,
+      "grad_norm": 2.25,
+      "learning_rate": 5.956000000000001e-06,
+      "loss": 0.3656,
+      "step": 31600
+    },
+    {
+      "epoch": 0.7023195876288659,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.951555555555556e-06,
+      "loss": 0.3823,
+      "step": 31610
+    },
+    {
+      "epoch": 0.7025417703519374,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.947111111111112e-06,
+      "loss": 0.3883,
+      "step": 31620
+    },
+    {
+      "epoch": 0.7027639530750089,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.942666666666667e-06,
+      "loss": 0.3654,
+      "step": 31630
+    },
+    {
+      "epoch": 0.7029861357980803,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.938222222222223e-06,
+      "loss": 0.4223,
+      "step": 31640
+    },
+    {
+      "epoch": 0.7032083185211518,
+      "grad_norm": 2.375,
+      "learning_rate": 5.933777777777779e-06,
+      "loss": 0.3847,
+      "step": 31650
+    },
+    {
+      "epoch": 0.7034305012442232,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.9293333333333335e-06,
+      "loss": 0.3774,
+      "step": 31660
+    },
+    {
+      "epoch": 0.7036526839672947,
+      "grad_norm": 2.625,
+      "learning_rate": 5.924888888888889e-06,
+      "loss": 0.3562,
+      "step": 31670
+    },
+    {
+      "epoch": 0.7038748666903661,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.920444444444445e-06,
+      "loss": 0.3559,
+      "step": 31680
+    },
+    {
+      "epoch": 0.7040970494134376,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.916000000000001e-06,
+      "loss": 0.4212,
+      "step": 31690
+    },
+    {
+      "epoch": 0.704319232136509,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.911555555555556e-06,
+      "loss": 0.3567,
+      "step": 31700
+    },
+    {
+      "epoch": 0.7045414148595806,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.907111111111111e-06,
+      "loss": 0.3667,
+      "step": 31710
+    },
+    {
+      "epoch": 0.704763597582652,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.902666666666668e-06,
+      "loss": 0.3793,
+      "step": 31720
+    },
+    {
+      "epoch": 0.7049857803057235,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.8982222222222225e-06,
+      "loss": 0.3814,
+      "step": 31730
+    },
+    {
+      "epoch": 0.7052079630287948,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.893777777777778e-06,
+      "loss": 0.3853,
+      "step": 31740
+    },
+    {
+      "epoch": 0.7054301457518664,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.889333333333334e-06,
+      "loss": 0.4039,
+      "step": 31750
+    },
+    {
+      "epoch": 0.7056523284749378,
+      "grad_norm": 2.625,
+      "learning_rate": 5.884888888888889e-06,
+      "loss": 0.3638,
+      "step": 31760
+    },
+    {
+      "epoch": 0.7058745111980093,
+      "grad_norm": 2.75,
+      "learning_rate": 5.880444444444445e-06,
+      "loss": 0.433,
+      "step": 31770
+    },
+    {
+      "epoch": 0.7060966939210807,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.876000000000001e-06,
+      "loss": 0.434,
+      "step": 31780
+    },
+    {
+      "epoch": 0.7063188766441522,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.871555555555556e-06,
+      "loss": 0.384,
+      "step": 31790
+    },
+    {
+      "epoch": 0.7065410593672236,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.8671111111111115e-06,
+      "loss": 0.3686,
+      "step": 31800
+    },
+    {
+      "epoch": 0.7067632420902951,
+      "grad_norm": 2.375,
+      "learning_rate": 5.862666666666667e-06,
+      "loss": 0.3816,
+      "step": 31810
+    },
+    {
+      "epoch": 0.7069854248133665,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.858222222222223e-06,
+      "loss": 0.3428,
+      "step": 31820
+    },
+    {
+      "epoch": 0.707207607536438,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.853777777777779e-06,
+      "loss": 0.3623,
+      "step": 31830
+    },
+    {
+      "epoch": 0.7074297902595095,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.8493333333333335e-06,
+      "loss": 0.4127,
+      "step": 31840
+    },
+    {
+      "epoch": 0.7076519729825809,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.844888888888889e-06,
+      "loss": 0.4007,
+      "step": 31850
+    },
+    {
+      "epoch": 0.7078741557056524,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.840444444444445e-06,
+      "loss": 0.4189,
+      "step": 31860
+    },
+    {
+      "epoch": 0.7080963384287238,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.8360000000000005e-06,
+      "loss": 0.3812,
+      "step": 31870
+    },
+    {
+      "epoch": 0.7083185211517953,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.831555555555556e-06,
+      "loss": 0.3905,
+      "step": 31880
+    },
+    {
+      "epoch": 0.7085407038748667,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.827111111111111e-06,
+      "loss": 0.3401,
+      "step": 31890
+    },
+    {
+      "epoch": 0.7087628865979382,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.822666666666668e-06,
+      "loss": 0.3723,
+      "step": 31900
+    },
+    {
+      "epoch": 0.7089850693210096,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.818222222222223e-06,
+      "loss": 0.3983,
+      "step": 31910
+    },
+    {
+      "epoch": 0.7092072520440811,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.813777777777778e-06,
+      "loss": 0.3972,
+      "step": 31920
+    },
+    {
+      "epoch": 0.7094294347671525,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.809333333333334e-06,
+      "loss": 0.3765,
+      "step": 31930
+    },
+    {
+      "epoch": 0.709651617490224,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.8048888888888895e-06,
+      "loss": 0.4079,
+      "step": 31940
+    },
+    {
+      "epoch": 0.7098738002132954,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.800444444444445e-06,
+      "loss": 0.3835,
+      "step": 31950
+    },
+    {
+      "epoch": 0.7100959829363669,
+      "grad_norm": 2.109375,
+      "learning_rate": 5.796000000000001e-06,
+      "loss": 0.3712,
+      "step": 31960
+    },
+    {
+      "epoch": 0.7103181656594383,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.791555555555556e-06,
+      "loss": 0.3595,
+      "step": 31970
+    },
+    {
+      "epoch": 0.7105403483825098,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.7871111111111115e-06,
+      "loss": 0.4171,
+      "step": 31980
+    },
+    {
+      "epoch": 0.7107625311055812,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.782666666666667e-06,
+      "loss": 0.375,
+      "step": 31990
+    },
+    {
+      "epoch": 0.7109847138286527,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.778222222222223e-06,
+      "loss": 0.3762,
+      "step": 32000
+    },
+    {
+      "epoch": 0.7112068965517241,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.7737777777777785e-06,
+      "loss": 0.3487,
+      "step": 32010
+    },
+    {
+      "epoch": 0.7114290792747956,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.769333333333333e-06,
+      "loss": 0.387,
+      "step": 32020
+    },
+    {
+      "epoch": 0.711651261997867,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.76488888888889e-06,
+      "loss": 0.3527,
+      "step": 32030
+    },
+    {
+      "epoch": 0.7118734447209385,
+      "grad_norm": 2.5,
+      "learning_rate": 5.760444444444445e-06,
+      "loss": 0.3559,
+      "step": 32040
+    },
+    {
+      "epoch": 0.71209562744401,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.7560000000000005e-06,
+      "loss": 0.3764,
+      "step": 32050
+    },
+    {
+      "epoch": 0.7123178101670814,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.751555555555556e-06,
+      "loss": 0.4182,
+      "step": 32060
+    },
+    {
+      "epoch": 0.7125399928901529,
+      "grad_norm": 2.625,
+      "learning_rate": 5.747111111111111e-06,
+      "loss": 0.4145,
+      "step": 32070
+    },
+    {
+      "epoch": 0.7127621756132243,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.7426666666666675e-06,
+      "loss": 0.3925,
+      "step": 32080
+    },
+    {
+      "epoch": 0.7129843583362958,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.738222222222223e-06,
+      "loss": 0.3647,
+      "step": 32090
+    },
+    {
+      "epoch": 0.7132065410593672,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.733777777777778e-06,
+      "loss": 0.4259,
+      "step": 32100
+    },
+    {
+      "epoch": 0.7134287237824387,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.729333333333334e-06,
+      "loss": 0.363,
+      "step": 32110
+    },
+    {
+      "epoch": 0.7136509065055101,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.7248888888888895e-06,
+      "loss": 0.385,
+      "step": 32120
+    },
+    {
+      "epoch": 0.7138730892285816,
+      "grad_norm": 3.0,
+      "learning_rate": 5.720444444444445e-06,
+      "loss": 0.3991,
+      "step": 32130
+    },
+    {
+      "epoch": 0.714095271951653,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.716000000000001e-06,
+      "loss": 0.374,
+      "step": 32140
+    },
+    {
+      "epoch": 0.7143174546747245,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.711555555555556e-06,
+      "loss": 0.3876,
+      "step": 32150
+    },
+    {
+      "epoch": 0.7145396373977959,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.707111111111111e-06,
+      "loss": 0.4159,
+      "step": 32160
+    },
+    {
+      "epoch": 0.7147618201208674,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.702666666666667e-06,
+      "loss": 0.3928,
+      "step": 32170
+    },
+    {
+      "epoch": 0.7149840028439388,
+      "grad_norm": 2.875,
+      "learning_rate": 5.698222222222223e-06,
+      "loss": 0.3942,
+      "step": 32180
+    },
+    {
+      "epoch": 0.7152061855670103,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.6937777777777785e-06,
+      "loss": 0.3635,
+      "step": 32190
+    },
+    {
+      "epoch": 0.7154283682900817,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.689333333333333e-06,
+      "loss": 0.3683,
+      "step": 32200
+    },
+    {
+      "epoch": 0.7156505510131532,
+      "grad_norm": 2.125,
+      "learning_rate": 5.68488888888889e-06,
+      "loss": 0.3795,
+      "step": 32210
+    },
+    {
+      "epoch": 0.7158727337362246,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.680444444444445e-06,
+      "loss": 0.3697,
+      "step": 32220
+    },
+    {
+      "epoch": 0.7160949164592961,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.676e-06,
+      "loss": 0.3657,
+      "step": 32230
+    },
+    {
+      "epoch": 0.7163170991823676,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.671555555555556e-06,
+      "loss": 0.3692,
+      "step": 32240
+    },
+    {
+      "epoch": 0.716539281905439,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.667111111111112e-06,
+      "loss": 0.3662,
+      "step": 32250
+    },
+    {
+      "epoch": 0.7167614646285105,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.6626666666666675e-06,
+      "loss": 0.3842,
+      "step": 32260
+    },
+    {
+      "epoch": 0.7169836473515819,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.658222222222223e-06,
+      "loss": 0.3638,
+      "step": 32270
+    },
+    {
+      "epoch": 0.7172058300746534,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.653777777777778e-06,
+      "loss": 0.4022,
+      "step": 32280
+    },
+    {
+      "epoch": 0.7174280127977248,
+      "grad_norm": 3.25,
+      "learning_rate": 5.649333333333334e-06,
+      "loss": 0.3911,
+      "step": 32290
+    },
+    {
+      "epoch": 0.7176501955207963,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.644888888888889e-06,
+      "loss": 0.3572,
+      "step": 32300
+    },
+    {
+      "epoch": 0.7178723782438677,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.640444444444445e-06,
+      "loss": 0.3877,
+      "step": 32310
+    },
+    {
+      "epoch": 0.7180945609669392,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.636000000000001e-06,
+      "loss": 0.4065,
+      "step": 32320
+    },
+    {
+      "epoch": 0.7183167436900106,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.631555555555556e-06,
+      "loss": 0.3379,
+      "step": 32330
+    },
+    {
+      "epoch": 0.7185389264130821,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.627111111111112e-06,
+      "loss": 0.3465,
+      "step": 32340
+    },
+    {
+      "epoch": 0.7187611091361535,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.622666666666667e-06,
+      "loss": 0.4159,
+      "step": 32350
+    },
+    {
+      "epoch": 0.718983291859225,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.618222222222223e-06,
+      "loss": 0.3727,
+      "step": 32360
+    },
+    {
+      "epoch": 0.7192054745822964,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.613777777777778e-06,
+      "loss": 0.3809,
+      "step": 32370
+    },
+    {
+      "epoch": 0.719427657305368,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.609333333333333e-06,
+      "loss": 0.35,
+      "step": 32380
+    },
+    {
+      "epoch": 0.7196498400284393,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.60488888888889e-06,
+      "loss": 0.4121,
+      "step": 32390
+    },
+    {
+      "epoch": 0.7198720227515109,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.600444444444445e-06,
+      "loss": 0.435,
+      "step": 32400
+    },
+    {
+      "epoch": 0.7200942054745822,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.596e-06,
+      "loss": 0.3537,
+      "step": 32410
+    },
+    {
+      "epoch": 0.7203163881976538,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.591555555555556e-06,
+      "loss": 0.3979,
+      "step": 32420
+    },
+    {
+      "epoch": 0.7205385709207252,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.587111111111112e-06,
+      "loss": 0.3518,
+      "step": 32430
+    },
+    {
+      "epoch": 0.7207607536437967,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.582666666666667e-06,
+      "loss": 0.3972,
+      "step": 32440
+    },
+    {
+      "epoch": 0.7209829363668682,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.578222222222223e-06,
+      "loss": 0.399,
+      "step": 32450
+    },
+    {
+      "epoch": 0.7212051190899396,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.573777777777778e-06,
+      "loss": 0.4075,
+      "step": 32460
+    },
+    {
+      "epoch": 0.7214273018130111,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.569333333333334e-06,
+      "loss": 0.383,
+      "step": 32470
+    },
+    {
+      "epoch": 0.7216494845360825,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.564888888888889e-06,
+      "loss": 0.4107,
+      "step": 32480
+    },
+    {
+      "epoch": 0.721871667259154,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.560444444444445e-06,
+      "loss": 0.3835,
+      "step": 32490
+    },
+    {
+      "epoch": 0.7220938499822254,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.556000000000001e-06,
+      "loss": 0.4052,
+      "step": 32500
+    },
+    {
+      "epoch": 0.7223160327052969,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.5515555555555555e-06,
+      "loss": 0.3796,
+      "step": 32510
+    },
+    {
+      "epoch": 0.7225382154283683,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.547111111111112e-06,
+      "loss": 0.386,
+      "step": 32520
+    },
+    {
+      "epoch": 0.7227603981514398,
+      "grad_norm": 2.5,
+      "learning_rate": 5.542666666666667e-06,
+      "loss": 0.3719,
+      "step": 32530
+    },
+    {
+      "epoch": 0.7229825808745112,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.538222222222223e-06,
+      "loss": 0.3423,
+      "step": 32540
+    },
+    {
+      "epoch": 0.7232047635975827,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.533777777777778e-06,
+      "loss": 0.378,
+      "step": 32550
+    },
+    {
+      "epoch": 0.7234269463206541,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.529333333333334e-06,
+      "loss": 0.3324,
+      "step": 32560
+    },
+    {
+      "epoch": 0.7236491290437256,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.52488888888889e-06,
+      "loss": 0.3906,
+      "step": 32570
+    },
+    {
+      "epoch": 0.723871311766797,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.5204444444444445e-06,
+      "loss": 0.3532,
+      "step": 32580
+    },
+    {
+      "epoch": 0.7240934944898685,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.516e-06,
+      "loss": 0.4058,
+      "step": 32590
+    },
+    {
+      "epoch": 0.7243156772129399,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.511555555555556e-06,
+      "loss": 0.3908,
+      "step": 32600
+    },
+    {
+      "epoch": 0.7245378599360114,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.507111111111112e-06,
+      "loss": 0.3788,
+      "step": 32610
+    },
+    {
+      "epoch": 0.7247600426590828,
+      "grad_norm": 2.75,
+      "learning_rate": 5.502666666666667e-06,
+      "loss": 0.4041,
+      "step": 32620
+    },
+    {
+      "epoch": 0.7249822253821543,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.498222222222223e-06,
+      "loss": 0.3679,
+      "step": 32630
+    },
+    {
+      "epoch": 0.7252044081052257,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.493777777777778e-06,
+      "loss": 0.4047,
+      "step": 32640
+    },
+    {
+      "epoch": 0.7254265908282972,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.489333333333334e-06,
+      "loss": 0.3897,
+      "step": 32650
+    },
+    {
+      "epoch": 0.7256487735513687,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.484888888888889e-06,
+      "loss": 0.3811,
+      "step": 32660
+    },
+    {
+      "epoch": 0.7258709562744401,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.480444444444445e-06,
+      "loss": 0.3456,
+      "step": 32670
+    },
+    {
+      "epoch": 0.7260931389975116,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.476000000000001e-06,
+      "loss": 0.3679,
+      "step": 32680
+    },
+    {
+      "epoch": 0.726315321720583,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.4715555555555554e-06,
+      "loss": 0.3522,
+      "step": 32690
+    },
+    {
+      "epoch": 0.7265375044436545,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.467111111111112e-06,
+      "loss": 0.4007,
+      "step": 32700
+    },
+    {
+      "epoch": 0.7267596871667259,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.462666666666667e-06,
+      "loss": 0.3458,
+      "step": 32710
+    },
+    {
+      "epoch": 0.7269818698897974,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.4582222222222225e-06,
+      "loss": 0.3631,
+      "step": 32720
+    },
+    {
+      "epoch": 0.7272040526128688,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.453777777777778e-06,
+      "loss": 0.379,
+      "step": 32730
+    },
+    {
+      "epoch": 0.7274262353359403,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.449333333333334e-06,
+      "loss": 0.3544,
+      "step": 32740
+    },
+    {
+      "epoch": 0.7276484180590117,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.44488888888889e-06,
+      "loss": 0.3485,
+      "step": 32750
+    },
+    {
+      "epoch": 0.7278706007820832,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.4404444444444444e-06,
+      "loss": 0.4027,
+      "step": 32760
+    },
+    {
+      "epoch": 0.7280927835051546,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.436e-06,
+      "loss": 0.4076,
+      "step": 32770
+    },
+    {
+      "epoch": 0.7283149662282261,
+      "grad_norm": 2.203125,
+      "learning_rate": 5.431555555555556e-06,
+      "loss": 0.3965,
+      "step": 32780
+    },
+    {
+      "epoch": 0.7285371489512975,
+      "grad_norm": 2.25,
+      "learning_rate": 5.4271111111111115e-06,
+      "loss": 0.3601,
+      "step": 32790
+    },
+    {
+      "epoch": 0.728759331674369,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.422666666666667e-06,
+      "loss": 0.3642,
+      "step": 32800
+    },
+    {
+      "epoch": 0.7289815143974404,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.418222222222223e-06,
+      "loss": 0.3831,
+      "step": 32810
+    },
+    {
+      "epoch": 0.7292036971205119,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.413777777777778e-06,
+      "loss": 0.3745,
+      "step": 32820
+    },
+    {
+      "epoch": 0.7294258798435833,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.409333333333334e-06,
+      "loss": 0.3593,
+      "step": 32830
+    },
+    {
+      "epoch": 0.7296480625666548,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.404888888888889e-06,
+      "loss": 0.3802,
+      "step": 32840
+    },
+    {
+      "epoch": 0.7298702452897262,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.400444444444445e-06,
+      "loss": 0.3895,
+      "step": 32850
+    },
+    {
+      "epoch": 0.7300924280127977,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.3960000000000005e-06,
+      "loss": 0.3854,
+      "step": 32860
+    },
+    {
+      "epoch": 0.7303146107358692,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.391555555555556e-06,
+      "loss": 0.389,
+      "step": 32870
+    },
+    {
+      "epoch": 0.7305367934589406,
+      "grad_norm": 3.140625,
+      "learning_rate": 5.387111111111112e-06,
+      "loss": 0.3785,
+      "step": 32880
+    },
+    {
+      "epoch": 0.7307589761820121,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.382666666666667e-06,
+      "loss": 0.4139,
+      "step": 32890
+    },
+    {
+      "epoch": 0.7309811589050835,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.3782222222222224e-06,
+      "loss": 0.3552,
+      "step": 32900
+    },
+    {
+      "epoch": 0.731203341628155,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.373777777777778e-06,
+      "loss": 0.3713,
+      "step": 32910
+    },
+    {
+      "epoch": 0.7314255243512264,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.369333333333334e-06,
+      "loss": 0.3602,
+      "step": 32920
+    },
+    {
+      "epoch": 0.7316477070742979,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.3648888888888895e-06,
+      "loss": 0.3437,
+      "step": 32930
+    },
+    {
+      "epoch": 0.7318698897973693,
+      "grad_norm": 2.25,
+      "learning_rate": 5.360444444444445e-06,
+      "loss": 0.3611,
+      "step": 32940
+    },
+    {
+      "epoch": 0.7320920725204408,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.356e-06,
+      "loss": 0.3704,
+      "step": 32950
+    },
+    {
+      "epoch": 0.7323142552435122,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.351555555555557e-06,
+      "loss": 0.4018,
+      "step": 32960
+    },
+    {
+      "epoch": 0.7325364379665837,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.3471111111111114e-06,
+      "loss": 0.4128,
+      "step": 32970
+    },
+    {
+      "epoch": 0.7327586206896551,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.342666666666667e-06,
+      "loss": 0.3732,
+      "step": 32980
+    },
+    {
+      "epoch": 0.7329808034127266,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.338222222222223e-06,
+      "loss": 0.3742,
+      "step": 32990
+    },
+    {
+      "epoch": 0.733202986135798,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.333777777777778e-06,
+      "loss": 0.3759,
+      "step": 33000
+    },
+    {
+      "epoch": 0.7334251688588695,
+      "grad_norm": 3.125,
+      "learning_rate": 5.329333333333334e-06,
+      "loss": 0.3769,
+      "step": 33010
+    },
+    {
+      "epoch": 0.7336473515819409,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.324888888888889e-06,
+      "loss": 0.3619,
+      "step": 33020
+    },
+    {
+      "epoch": 0.7338695343050124,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.320444444444445e-06,
+      "loss": 0.4092,
+      "step": 33030
+    },
+    {
+      "epoch": 0.7340917170280838,
+      "grad_norm": 2.625,
+      "learning_rate": 5.3160000000000004e-06,
+      "loss": 0.4058,
+      "step": 33040
+    },
+    {
+      "epoch": 0.7343138997511554,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.311555555555556e-06,
+      "loss": 0.4174,
+      "step": 33050
+    },
+    {
+      "epoch": 0.7345360824742269,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.307111111111112e-06,
+      "loss": 0.4063,
+      "step": 33060
+    },
+    {
+      "epoch": 0.7347582651972983,
+      "grad_norm": 2.75,
+      "learning_rate": 5.302666666666667e-06,
+      "loss": 0.3657,
+      "step": 33070
+    },
+    {
+      "epoch": 0.7349804479203698,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.298222222222222e-06,
+      "loss": 0.3567,
+      "step": 33080
+    },
+    {
+      "epoch": 0.7352026306434412,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.293777777777778e-06,
+      "loss": 0.395,
+      "step": 33090
+    },
+    {
+      "epoch": 0.7354248133665127,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.289333333333334e-06,
+      "loss": 0.3675,
+      "step": 33100
+    },
+    {
+      "epoch": 0.7356469960895841,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.2848888888888894e-06,
+      "loss": 0.4294,
+      "step": 33110
+    },
+    {
+      "epoch": 0.7358691788126556,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.280444444444445e-06,
+      "loss": 0.3893,
+      "step": 33120
+    },
+    {
+      "epoch": 0.736091361535727,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.276e-06,
+      "loss": 0.3707,
+      "step": 33130
+    },
+    {
+      "epoch": 0.7363135442587985,
+      "grad_norm": 2.375,
+      "learning_rate": 5.2715555555555565e-06,
+      "loss": 0.4016,
+      "step": 33140
+    },
+    {
+      "epoch": 0.7365357269818699,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.267111111111111e-06,
+      "loss": 0.3467,
+      "step": 33150
+    },
+    {
+      "epoch": 0.7367579097049414,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.262666666666667e-06,
+      "loss": 0.3979,
+      "step": 33160
+    },
+    {
+      "epoch": 0.7369800924280128,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.258222222222223e-06,
+      "loss": 0.4023,
+      "step": 33170
+    },
+    {
+      "epoch": 0.7372022751510843,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.2537777777777784e-06,
+      "loss": 0.3649,
+      "step": 33180
+    },
+    {
+      "epoch": 0.7374244578741557,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.249333333333334e-06,
+      "loss": 0.3974,
+      "step": 33190
+    },
+    {
+      "epoch": 0.7376466405972272,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.244888888888889e-06,
+      "loss": 0.3581,
+      "step": 33200
+    },
+    {
+      "epoch": 0.7378688233202986,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.240444444444445e-06,
+      "loss": 0.3588,
+      "step": 33210
+    },
+    {
+      "epoch": 0.7380910060433701,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.236e-06,
+      "loss": 0.367,
+      "step": 33220
+    },
+    {
+      "epoch": 0.7383131887664415,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.231555555555556e-06,
+      "loss": 0.4256,
+      "step": 33230
+    },
+    {
+      "epoch": 0.738535371489513,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.227111111111112e-06,
+      "loss": 0.3773,
+      "step": 33240
+    },
+    {
+      "epoch": 0.7387575542125844,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.222666666666667e-06,
+      "loss": 0.3831,
+      "step": 33250
+    },
+    {
+      "epoch": 0.7389797369356559,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.218222222222222e-06,
+      "loss": 0.3948,
+      "step": 33260
+    },
+    {
+      "epoch": 0.7392019196587274,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.213777777777779e-06,
+      "loss": 0.3287,
+      "step": 33270
+    },
+    {
+      "epoch": 0.7394241023817988,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.209333333333334e-06,
+      "loss": 0.3519,
+      "step": 33280
+    },
+    {
+      "epoch": 0.7396462851048703,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.204888888888889e-06,
+      "loss": 0.3808,
+      "step": 33290
+    },
+    {
+      "epoch": 0.7398684678279417,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.200444444444445e-06,
+      "loss": 0.3694,
+      "step": 33300
+    },
+    {
+      "epoch": 0.7400906505510132,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.196e-06,
+      "loss": 0.3577,
+      "step": 33310
+    },
+    {
+      "epoch": 0.7403128332740846,
+      "grad_norm": 2.875,
+      "learning_rate": 5.1915555555555564e-06,
+      "loss": 0.4061,
+      "step": 33320
+    },
+    {
+      "epoch": 0.7405350159971561,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.187111111111111e-06,
+      "loss": 0.3947,
+      "step": 33330
+    },
+    {
+      "epoch": 0.7407571987202275,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.182666666666667e-06,
+      "loss": 0.3818,
+      "step": 33340
+    },
+    {
+      "epoch": 0.740979381443299,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.178222222222223e-06,
+      "loss": 0.3778,
+      "step": 33350
+    },
+    {
+      "epoch": 0.7412015641663704,
+      "grad_norm": 2.0,
+      "learning_rate": 5.173777777777778e-06,
+      "loss": 0.3661,
+      "step": 33360
+    },
+    {
+      "epoch": 0.7414237468894419,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.169333333333334e-06,
+      "loss": 0.364,
+      "step": 33370
+    },
+    {
+      "epoch": 0.7416459296125133,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.164888888888889e-06,
+      "loss": 0.3494,
+      "step": 33380
+    },
+    {
+      "epoch": 0.7418681123355848,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.160444444444445e-06,
+      "loss": 0.3959,
+      "step": 33390
+    },
+    {
+      "epoch": 0.7420902950586562,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.156e-06,
+      "loss": 0.3567,
+      "step": 33400
+    },
+    {
+      "epoch": 0.7423124777817277,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.151555555555556e-06,
+      "loss": 0.3257,
+      "step": 33410
+    },
+    {
+      "epoch": 0.7425346605047991,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.147111111111112e-06,
+      "loss": 0.3331,
+      "step": 33420
+    },
+    {
+      "epoch": 0.7427568432278706,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.1426666666666665e-06,
+      "loss": 0.3841,
+      "step": 33430
+    },
+    {
+      "epoch": 0.742979025950942,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.138222222222222e-06,
+      "loss": 0.3922,
+      "step": 33440
+    },
+    {
+      "epoch": 0.7432012086740135,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.133777777777779e-06,
+      "loss": 0.3382,
+      "step": 33450
+    },
+    {
+      "epoch": 0.7434233913970849,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.129333333333334e-06,
+      "loss": 0.3916,
+      "step": 33460
+    },
+    {
+      "epoch": 0.7436455741201564,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.124888888888889e-06,
+      "loss": 0.3768,
+      "step": 33470
+    },
+    {
+      "epoch": 0.7438677568432279,
+      "grad_norm": 2.5,
+      "learning_rate": 5.120444444444445e-06,
+      "loss": 0.35,
+      "step": 33480
+    },
+    {
+      "epoch": 0.7440899395662993,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.116000000000001e-06,
+      "loss": 0.3907,
+      "step": 33490
+    },
+    {
+      "epoch": 0.7443121222893708,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.111555555555556e-06,
+      "loss": 0.4041,
+      "step": 33500
+    },
+    {
+      "epoch": 0.7445343050124422,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.107111111111111e-06,
+      "loss": 0.3545,
+      "step": 33510
+    },
+    {
+      "epoch": 0.7447564877355137,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.102666666666667e-06,
+      "loss": 0.4054,
+      "step": 33520
+    },
+    {
+      "epoch": 0.7449786704585851,
+      "grad_norm": 2.75,
+      "learning_rate": 5.0982222222222226e-06,
+      "loss": 0.3445,
+      "step": 33530
+    },
+    {
+      "epoch": 0.7452008531816566,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.093777777777778e-06,
+      "loss": 0.377,
+      "step": 33540
+    },
+    {
+      "epoch": 0.745423035904728,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.089333333333334e-06,
+      "loss": 0.3659,
+      "step": 33550
+    },
+    {
+      "epoch": 0.7456452186277995,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.084888888888889e-06,
+      "loss": 0.3905,
+      "step": 33560
+    },
+    {
+      "epoch": 0.7458674013508709,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.0804444444444445e-06,
+      "loss": 0.4162,
+      "step": 33570
+    },
+    {
+      "epoch": 0.7460895840739424,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.076000000000001e-06,
+      "loss": 0.4015,
+      "step": 33580
+    },
+    {
+      "epoch": 0.7463117667970138,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.071555555555556e-06,
+      "loss": 0.3701,
+      "step": 33590
+    },
+    {
+      "epoch": 0.7465339495200853,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.0671111111111116e-06,
+      "loss": 0.3764,
+      "step": 33600
+    },
+    {
+      "epoch": 0.7467561322431567,
+      "grad_norm": 1.828125,
+      "learning_rate": 5.062666666666666e-06,
+      "loss": 0.3787,
+      "step": 33610
+    },
+    {
+      "epoch": 0.7469783149662282,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.058222222222222e-06,
+      "loss": 0.3511,
+      "step": 33620
+    },
+    {
+      "epoch": 0.7472004976892996,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.053777777777779e-06,
+      "loss": 0.386,
+      "step": 33630
+    },
+    {
+      "epoch": 0.7474226804123711,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.0493333333333335e-06,
+      "loss": 0.4056,
+      "step": 33640
+    },
+    {
+      "epoch": 0.7476448631354425,
+      "grad_norm": 2.625,
+      "learning_rate": 5.044888888888889e-06,
+      "loss": 0.4146,
+      "step": 33650
+    },
+    {
+      "epoch": 0.747867045858514,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.040444444444445e-06,
+      "loss": 0.3825,
+      "step": 33660
+    },
+    {
+      "epoch": 0.7480892285815854,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.0360000000000006e-06,
+      "loss": 0.3307,
+      "step": 33670
+    },
+    {
+      "epoch": 0.748311411304657,
+      "grad_norm": 2.5,
+      "learning_rate": 5.031555555555556e-06,
+      "loss": 0.3789,
+      "step": 33680
+    },
+    {
+      "epoch": 0.7485335940277285,
+      "grad_norm": 2.625,
+      "learning_rate": 5.027111111111111e-06,
+      "loss": 0.3696,
+      "step": 33690
+    },
+    {
+      "epoch": 0.7487557767507999,
+      "grad_norm": 2.5,
+      "learning_rate": 5.022666666666667e-06,
+      "loss": 0.4013,
+      "step": 33700
+    },
+    {
+      "epoch": 0.7489779594738714,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.0182222222222225e-06,
+      "loss": 0.3663,
+      "step": 33710
+    },
+    {
+      "epoch": 0.7492001421969428,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.013777777777778e-06,
+      "loss": 0.3976,
+      "step": 33720
+    },
+    {
+      "epoch": 0.7494223249200143,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.009333333333334e-06,
+      "loss": 0.3801,
+      "step": 33730
+    },
+    {
+      "epoch": 0.7496445076430857,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.004888888888889e-06,
+      "loss": 0.3946,
+      "step": 33740
+    },
+    {
+      "epoch": 0.7498666903661572,
+      "grad_norm": 2.1875,
+      "learning_rate": 5.000444444444444e-06,
+      "loss": 0.3812,
+      "step": 33750
+    },
+    {
+      "epoch": 0.7500888730892286,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.996e-06,
+      "loss": 0.4265,
+      "step": 33760
+    },
+    {
+      "epoch": 0.7503110558123001,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.991555555555556e-06,
+      "loss": 0.3479,
+      "step": 33770
+    },
+    {
+      "epoch": 0.7505332385353715,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.9871111111111115e-06,
+      "loss": 0.3982,
+      "step": 33780
+    },
+    {
+      "epoch": 0.750755421258443,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.982666666666667e-06,
+      "loss": 0.4098,
+      "step": 33790
+    },
+    {
+      "epoch": 0.7509776039815144,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.978222222222223e-06,
+      "loss": 0.4078,
+      "step": 33800
+    },
+    {
+      "epoch": 0.7511997867045859,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.973777777777778e-06,
+      "loss": 0.3804,
+      "step": 33810
+    },
+    {
+      "epoch": 0.7514219694276573,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.969333333333333e-06,
+      "loss": 0.3774,
+      "step": 33820
+    },
+    {
+      "epoch": 0.7516441521507288,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.964888888888889e-06,
+      "loss": 0.3881,
+      "step": 33830
+    },
+    {
+      "epoch": 0.7518663348738002,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.960444444444445e-06,
+      "loss": 0.3709,
+      "step": 33840
+    },
+    {
+      "epoch": 0.7520885175968717,
+      "grad_norm": 2.5,
+      "learning_rate": 4.9560000000000005e-06,
+      "loss": 0.3514,
+      "step": 33850
+    },
+    {
+      "epoch": 0.7523107003199431,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.951555555555556e-06,
+      "loss": 0.3587,
+      "step": 33860
+    },
+    {
+      "epoch": 0.7525328830430146,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.947111111111111e-06,
+      "loss": 0.3717,
+      "step": 33870
+    },
+    {
+      "epoch": 0.7527550657660861,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.9426666666666676e-06,
+      "loss": 0.3629,
+      "step": 33880
+    },
+    {
+      "epoch": 0.7529772484891575,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.938222222222222e-06,
+      "loss": 0.3995,
+      "step": 33890
+    },
+    {
+      "epoch": 0.753199431212229,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.933777777777778e-06,
+      "loss": 0.3856,
+      "step": 33900
+    },
+    {
+      "epoch": 0.7534216139353004,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.929333333333334e-06,
+      "loss": 0.382,
+      "step": 33910
+    },
+    {
+      "epoch": 0.7536437966583719,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.924888888888889e-06,
+      "loss": 0.3794,
+      "step": 33920
+    },
+    {
+      "epoch": 0.7538659793814433,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.920444444444445e-06,
+      "loss": 0.3611,
+      "step": 33930
+    },
+    {
+      "epoch": 0.7540881621045148,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.916e-06,
+      "loss": 0.3722,
+      "step": 33940
+    },
+    {
+      "epoch": 0.7543103448275862,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.911555555555556e-06,
+      "loss": 0.366,
+      "step": 33950
+    },
+    {
+      "epoch": 0.7545325275506577,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.907111111111111e-06,
+      "loss": 0.3687,
+      "step": 33960
+    },
+    {
+      "epoch": 0.7547547102737291,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.902666666666667e-06,
+      "loss": 0.3423,
+      "step": 33970
+    },
+    {
+      "epoch": 0.7549768929968006,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.898222222222223e-06,
+      "loss": 0.4181,
+      "step": 33980
+    },
+    {
+      "epoch": 0.755199075719872,
+      "grad_norm": 3.125,
+      "learning_rate": 4.8937777777777785e-06,
+      "loss": 0.3907,
+      "step": 33990
+    },
+    {
+      "epoch": 0.7554212584429435,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.889333333333333e-06,
+      "loss": 0.3739,
+      "step": 34000
+    },
+    {
+      "epoch": 0.7556434411660149,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.884888888888889e-06,
+      "loss": 0.3817,
+      "step": 34010
+    },
+    {
+      "epoch": 0.7558656238890864,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.880444444444445e-06,
+      "loss": 0.3424,
+      "step": 34020
+    },
+    {
+      "epoch": 0.7560878066121578,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.876e-06,
+      "loss": 0.3504,
+      "step": 34030
+    },
+    {
+      "epoch": 0.7563099893352293,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.871555555555556e-06,
+      "loss": 0.4067,
+      "step": 34040
+    },
+    {
+      "epoch": 0.7565321720583007,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.867111111111111e-06,
+      "loss": 0.3779,
+      "step": 34050
+    },
+    {
+      "epoch": 0.7567543547813722,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.8626666666666675e-06,
+      "loss": 0.3708,
+      "step": 34060
+    },
+    {
+      "epoch": 0.7569765375044436,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.858222222222222e-06,
+      "loss": 0.3694,
+      "step": 34070
+    },
+    {
+      "epoch": 0.7571987202275151,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.853777777777778e-06,
+      "loss": 0.3966,
+      "step": 34080
+    },
+    {
+      "epoch": 0.7574209029505866,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.849333333333334e-06,
+      "loss": 0.3978,
+      "step": 34090
+    },
+    {
+      "epoch": 0.757643085673658,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.844888888888889e-06,
+      "loss": 0.3731,
+      "step": 34100
+    },
+    {
+      "epoch": 0.7578652683967295,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.840444444444445e-06,
+      "loss": 0.352,
+      "step": 34110
+    },
+    {
+      "epoch": 0.7580874511198009,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.836e-06,
+      "loss": 0.367,
+      "step": 34120
+    },
+    {
+      "epoch": 0.7583096338428724,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.831555555555556e-06,
+      "loss": 0.3731,
+      "step": 34130
+    },
+    {
+      "epoch": 0.7585318165659438,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.827111111111111e-06,
+      "loss": 0.3773,
+      "step": 34140
+    },
+    {
+      "epoch": 0.7587539992890153,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.822666666666667e-06,
+      "loss": 0.4028,
+      "step": 34150
+    },
+    {
+      "epoch": 0.7589761820120867,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.818222222222223e-06,
+      "loss": 0.3595,
+      "step": 34160
+    },
+    {
+      "epoch": 0.7591983647351582,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.813777777777778e-06,
+      "loss": 0.3652,
+      "step": 34170
+    },
+    {
+      "epoch": 0.7594205474582296,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.809333333333333e-06,
+      "loss": 0.3695,
+      "step": 34180
+    },
+    {
+      "epoch": 0.7596427301813011,
+      "grad_norm": 2.375,
+      "learning_rate": 4.80488888888889e-06,
+      "loss": 0.3856,
+      "step": 34190
+    },
+    {
+      "epoch": 0.7598649129043725,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.800444444444445e-06,
+      "loss": 0.3393,
+      "step": 34200
+    },
+    {
+      "epoch": 0.760087095627444,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.796e-06,
+      "loss": 0.3476,
+      "step": 34210
+    },
+    {
+      "epoch": 0.7603092783505154,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.791555555555556e-06,
+      "loss": 0.3445,
+      "step": 34220
+    },
+    {
+      "epoch": 0.7605314610735869,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.787111111111111e-06,
+      "loss": 0.3893,
+      "step": 34230
+    },
+    {
+      "epoch": 0.7607536437966583,
+      "grad_norm": 2.5,
+      "learning_rate": 4.782666666666667e-06,
+      "loss": 0.4079,
+      "step": 34240
+    },
+    {
+      "epoch": 0.7609758265197298,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.778222222222222e-06,
+      "loss": 0.3897,
+      "step": 34250
+    },
+    {
+      "epoch": 0.7611980092428012,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.773777777777778e-06,
+      "loss": 0.3997,
+      "step": 34260
+    },
+    {
+      "epoch": 0.7614201919658727,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.769333333333334e-06,
+      "loss": 0.3641,
+      "step": 34270
+    },
+    {
+      "epoch": 0.7616423746889441,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.764888888888889e-06,
+      "loss": 0.3564,
+      "step": 34280
+    },
+    {
+      "epoch": 0.7618645574120156,
+      "grad_norm": 2.625,
+      "learning_rate": 4.760444444444445e-06,
+      "loss": 0.3899,
+      "step": 34290
+    },
+    {
+      "epoch": 0.7620867401350871,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.756000000000001e-06,
+      "loss": 0.3494,
+      "step": 34300
+    },
+    {
+      "epoch": 0.7623089228581585,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.7515555555555556e-06,
+      "loss": 0.4105,
+      "step": 34310
+    },
+    {
+      "epoch": 0.76253110558123,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.747111111111111e-06,
+      "loss": 0.3266,
+      "step": 34320
+    },
+    {
+      "epoch": 0.7627532883043014,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.742666666666667e-06,
+      "loss": 0.3259,
+      "step": 34330
+    },
+    {
+      "epoch": 0.762975471027373,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.738222222222223e-06,
+      "loss": 0.3529,
+      "step": 34340
+    },
+    {
+      "epoch": 0.7631976537504443,
+      "grad_norm": 2.375,
+      "learning_rate": 4.733777777777778e-06,
+      "loss": 0.4033,
+      "step": 34350
+    },
+    {
+      "epoch": 0.7634198364735159,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.729333333333333e-06,
+      "loss": 0.3745,
+      "step": 34360
+    },
+    {
+      "epoch": 0.7636420191965873,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.72488888888889e-06,
+      "loss": 0.3941,
+      "step": 34370
+    },
+    {
+      "epoch": 0.7638642019196588,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.7204444444444446e-06,
+      "loss": 0.3911,
+      "step": 34380
+    },
+    {
+      "epoch": 0.7640863846427302,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.716e-06,
+      "loss": 0.3599,
+      "step": 34390
+    },
+    {
+      "epoch": 0.7643085673658017,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.711555555555556e-06,
+      "loss": 0.4085,
+      "step": 34400
+    },
+    {
+      "epoch": 0.7645307500888731,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.707111111111112e-06,
+      "loss": 0.3546,
+      "step": 34410
+    },
+    {
+      "epoch": 0.7647529328119446,
+      "grad_norm": 3.125,
+      "learning_rate": 4.702666666666667e-06,
+      "loss": 0.4001,
+      "step": 34420
+    },
+    {
+      "epoch": 0.764975115535016,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.698222222222222e-06,
+      "loss": 0.4021,
+      "step": 34430
+    },
+    {
+      "epoch": 0.7651972982580875,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.693777777777778e-06,
+      "loss": 0.3948,
+      "step": 34440
+    },
+    {
+      "epoch": 0.7654194809811589,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.6893333333333336e-06,
+      "loss": 0.3804,
+      "step": 34450
+    },
+    {
+      "epoch": 0.7656416637042304,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.684888888888889e-06,
+      "loss": 0.3928,
+      "step": 34460
+    },
+    {
+      "epoch": 0.7658638464273018,
+      "grad_norm": 2.75,
+      "learning_rate": 4.680444444444445e-06,
+      "loss": 0.3607,
+      "step": 34470
+    },
+    {
+      "epoch": 0.7660860291503733,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.676000000000001e-06,
+      "loss": 0.4113,
+      "step": 34480
+    },
+    {
+      "epoch": 0.7663082118734447,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.6715555555555555e-06,
+      "loss": 0.3627,
+      "step": 34490
+    },
+    {
+      "epoch": 0.7665303945965162,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.667111111111112e-06,
+      "loss": 0.3499,
+      "step": 34500
+    },
+    {
+      "epoch": 0.7667525773195877,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.662666666666667e-06,
+      "loss": 0.3732,
+      "step": 34510
+    },
+    {
+      "epoch": 0.7669747600426591,
+      "grad_norm": 1.953125,
+      "learning_rate": 4.6582222222222226e-06,
+      "loss": 0.3323,
+      "step": 34520
+    },
+    {
+      "epoch": 0.7671969427657306,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.653777777777778e-06,
+      "loss": 0.3621,
+      "step": 34530
+    },
+    {
+      "epoch": 0.767419125488802,
+      "grad_norm": 2.625,
+      "learning_rate": 4.649333333333333e-06,
+      "loss": 0.4036,
+      "step": 34540
+    },
+    {
+      "epoch": 0.7676413082118735,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.64488888888889e-06,
+      "loss": 0.3702,
+      "step": 34550
+    },
+    {
+      "epoch": 0.7678634909349449,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.6404444444444445e-06,
+      "loss": 0.3808,
+      "step": 34560
+    },
+    {
+      "epoch": 0.7680856736580164,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.636e-06,
+      "loss": 0.3658,
+      "step": 34570
+    },
+    {
+      "epoch": 0.7683078563810878,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.631555555555556e-06,
+      "loss": 0.3817,
+      "step": 34580
+    },
+    {
+      "epoch": 0.7685300391041593,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6271111111111116e-06,
+      "loss": 0.3879,
+      "step": 34590
+    },
+    {
+      "epoch": 0.7687522218272307,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.622666666666667e-06,
+      "loss": 0.3624,
+      "step": 34600
+    },
+    {
+      "epoch": 0.7689744045503022,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.618222222222223e-06,
+      "loss": 0.3752,
+      "step": 34610
+    },
+    {
+      "epoch": 0.7691965872733736,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.613777777777778e-06,
+      "loss": 0.3627,
+      "step": 34620
+    },
+    {
+      "epoch": 0.7694187699964451,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.6093333333333335e-06,
+      "loss": 0.389,
+      "step": 34630
+    },
+    {
+      "epoch": 0.7696409527195165,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.604888888888889e-06,
+      "loss": 0.3609,
+      "step": 34640
+    },
+    {
+      "epoch": 0.769863135442588,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.600444444444445e-06,
+      "loss": 0.3974,
+      "step": 34650
+    },
+    {
+      "epoch": 0.7700853181656594,
+      "grad_norm": 2.5,
+      "learning_rate": 4.5960000000000006e-06,
+      "loss": 0.4181,
+      "step": 34660
+    },
+    {
+      "epoch": 0.7703075008887309,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.591555555555555e-06,
+      "loss": 0.3805,
+      "step": 34670
+    },
+    {
+      "epoch": 0.7705296836118023,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.587111111111112e-06,
+      "loss": 0.374,
+      "step": 34680
+    },
+    {
+      "epoch": 0.7707518663348738,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.582666666666667e-06,
+      "loss": 0.3311,
+      "step": 34690
+    },
+    {
+      "epoch": 0.7709740490579452,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.5782222222222225e-06,
+      "loss": 0.3876,
+      "step": 34700
+    },
+    {
+      "epoch": 0.7711962317810167,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.573777777777778e-06,
+      "loss": 0.3584,
+      "step": 34710
+    },
+    {
+      "epoch": 0.7714184145040882,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.569333333333334e-06,
+      "loss": 0.4144,
+      "step": 34720
+    },
+    {
+      "epoch": 0.7716405972271596,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.5648888888888895e-06,
+      "loss": 0.3428,
+      "step": 34730
+    },
+    {
+      "epoch": 0.7718627799502311,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.560444444444444e-06,
+      "loss": 0.3809,
+      "step": 34740
+    },
+    {
+      "epoch": 0.7720849626733025,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.556e-06,
+      "loss": 0.4017,
+      "step": 34750
+    },
+    {
+      "epoch": 0.772307145396374,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.551555555555556e-06,
+      "loss": 0.3841,
+      "step": 34760
+    },
+    {
+      "epoch": 0.7725293281194454,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.5471111111111115e-06,
+      "loss": 0.3991,
+      "step": 34770
+    },
+    {
+      "epoch": 0.7727515108425169,
+      "grad_norm": 2.625,
+      "learning_rate": 4.542666666666667e-06,
+      "loss": 0.4064,
+      "step": 34780
+    },
+    {
+      "epoch": 0.7729736935655883,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.538222222222223e-06,
+      "loss": 0.3668,
+      "step": 34790
+    },
+    {
+      "epoch": 0.7731958762886598,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.533777777777778e-06,
+      "loss": 0.4388,
+      "step": 34800
+    },
+    {
+      "epoch": 0.7734180590117312,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.529333333333334e-06,
+      "loss": 0.3686,
+      "step": 34810
+    },
+    {
+      "epoch": 0.7736402417348027,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.524888888888889e-06,
+      "loss": 0.3894,
+      "step": 34820
+    },
+    {
+      "epoch": 0.7738624244578741,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.520444444444445e-06,
+      "loss": 0.3523,
+      "step": 34830
+    },
+    {
+      "epoch": 0.7740846071809456,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.5160000000000005e-06,
+      "loss": 0.3411,
+      "step": 34840
+    },
+    {
+      "epoch": 0.774306789904017,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.511555555555555e-06,
+      "loss": 0.376,
+      "step": 34850
+    },
+    {
+      "epoch": 0.7745289726270885,
+      "grad_norm": 2.875,
+      "learning_rate": 4.507111111111112e-06,
+      "loss": 0.3893,
+      "step": 34860
+    },
+    {
+      "epoch": 0.7747511553501599,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.502666666666667e-06,
+      "loss": 0.3377,
+      "step": 34870
+    },
+    {
+      "epoch": 0.7749733380732314,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.498222222222222e-06,
+      "loss": 0.396,
+      "step": 34880
+    },
+    {
+      "epoch": 0.7751955207963028,
+      "grad_norm": 2.3125,
+      "learning_rate": 4.493777777777778e-06,
+      "loss": 0.3816,
+      "step": 34890
+    },
+    {
+      "epoch": 0.7754177035193743,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.489333333333334e-06,
+      "loss": 0.3772,
+      "step": 34900
+    },
+    {
+      "epoch": 0.7756398862424458,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.4848888888888895e-06,
+      "loss": 0.3348,
+      "step": 34910
+    },
+    {
+      "epoch": 0.7758620689655172,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.480444444444445e-06,
+      "loss": 0.4088,
+      "step": 34920
+    },
+    {
+      "epoch": 0.7760842516885887,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.476e-06,
+      "loss": 0.4077,
+      "step": 34930
+    },
+    {
+      "epoch": 0.7763064344116601,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.471555555555556e-06,
+      "loss": 0.3465,
+      "step": 34940
+    },
+    {
+      "epoch": 0.7765286171347316,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.467111111111111e-06,
+      "loss": 0.3538,
+      "step": 34950
+    },
+    {
+      "epoch": 0.776750799857803,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.462666666666667e-06,
+      "loss": 0.3795,
+      "step": 34960
+    },
+    {
+      "epoch": 0.7769729825808745,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.458222222222223e-06,
+      "loss": 0.3513,
+      "step": 34970
+    },
+    {
+      "epoch": 0.777195165303946,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.453777777777778e-06,
+      "loss": 0.3852,
+      "step": 34980
+    },
+    {
+      "epoch": 0.7774173480270175,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.449333333333334e-06,
+      "loss": 0.3594,
+      "step": 34990
+    },
+    {
+      "epoch": 0.7776395307500888,
+      "grad_norm": 2.125,
+      "learning_rate": 4.444888888888889e-06,
+      "loss": 0.3742,
+      "step": 35000
+    },
+    {
+      "epoch": 0.7778617134731604,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.440444444444445e-06,
+      "loss": 0.4009,
+      "step": 35010
+    },
+    {
+      "epoch": 0.7780838961962318,
+      "grad_norm": 2.375,
+      "learning_rate": 4.436e-06,
+      "loss": 0.3819,
+      "step": 35020
+    },
+    {
+      "epoch": 0.7783060789193033,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.431555555555556e-06,
+      "loss": 0.3715,
+      "step": 35030
+    },
+    {
+      "epoch": 0.7785282616423747,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.427111111111112e-06,
+      "loss": 0.3911,
+      "step": 35040
+    },
+    {
+      "epoch": 0.7787504443654462,
+      "grad_norm": 2.5,
+      "learning_rate": 4.422666666666667e-06,
+      "loss": 0.4007,
+      "step": 35050
+    },
+    {
+      "epoch": 0.7789726270885176,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.418222222222222e-06,
+      "loss": 0.3836,
+      "step": 35060
+    },
+    {
+      "epoch": 0.7791948098115891,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.413777777777778e-06,
+      "loss": 0.3945,
+      "step": 35070
+    },
+    {
+      "epoch": 0.7794169925346605,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.409333333333334e-06,
+      "loss": 0.3924,
+      "step": 35080
+    },
+    {
+      "epoch": 0.779639175257732,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.404888888888889e-06,
+      "loss": 0.3972,
+      "step": 35090
+    },
+    {
+      "epoch": 0.7798613579808034,
+      "grad_norm": 2.75,
+      "learning_rate": 4.400444444444445e-06,
+      "loss": 0.4165,
+      "step": 35100
+    },
+    {
+      "epoch": 0.7800835407038749,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.396e-06,
+      "loss": 0.3743,
+      "step": 35110
+    },
+    {
+      "epoch": 0.7803057234269464,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.3915555555555565e-06,
+      "loss": 0.3532,
+      "step": 35120
+    },
+    {
+      "epoch": 0.7805279061500178,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.387111111111111e-06,
+      "loss": 0.373,
+      "step": 35130
+    },
+    {
+      "epoch": 0.7807500888730893,
+      "grad_norm": 2.75,
+      "learning_rate": 4.382666666666667e-06,
+      "loss": 0.3815,
+      "step": 35140
+    },
+    {
+      "epoch": 0.7809722715961607,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.378222222222223e-06,
+      "loss": 0.3535,
+      "step": 35150
+    },
+    {
+      "epoch": 0.7811944543192322,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.3737777777777775e-06,
+      "loss": 0.371,
+      "step": 35160
+    },
+    {
+      "epoch": 0.7814166370423036,
+      "grad_norm": 2.25,
+      "learning_rate": 4.369333333333334e-06,
+      "loss": 0.3683,
+      "step": 35170
+    },
+    {
+      "epoch": 0.7816388197653751,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.364888888888889e-06,
+      "loss": 0.3765,
+      "step": 35180
+    },
+    {
+      "epoch": 0.7818610024884465,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.360444444444445e-06,
+      "loss": 0.3667,
+      "step": 35190
+    },
+    {
+      "epoch": 0.782083185211518,
+      "grad_norm": 2.125,
+      "learning_rate": 4.356e-06,
+      "loss": 0.3965,
+      "step": 35200
+    },
+    {
+      "epoch": 0.7823053679345894,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.351555555555556e-06,
+      "loss": 0.3767,
+      "step": 35210
+    },
+    {
+      "epoch": 0.7825275506576609,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.347111111111112e-06,
+      "loss": 0.3617,
+      "step": 35220
+    },
+    {
+      "epoch": 0.7827497333807323,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.342666666666667e-06,
+      "loss": 0.3547,
+      "step": 35230
+    },
+    {
+      "epoch": 0.7829719161038038,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.338222222222222e-06,
+      "loss": 0.3716,
+      "step": 35240
+    },
+    {
+      "epoch": 0.7831940988268752,
+      "grad_norm": 2.375,
+      "learning_rate": 4.333777777777778e-06,
+      "loss": 0.378,
+      "step": 35250
+    },
+    {
+      "epoch": 0.7834162815499467,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.329333333333334e-06,
+      "loss": 0.3583,
+      "step": 35260
+    },
+    {
+      "epoch": 0.7836384642730181,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.324888888888889e-06,
+      "loss": 0.3724,
+      "step": 35270
+    },
+    {
+      "epoch": 0.7838606469960896,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.320444444444445e-06,
+      "loss": 0.3697,
+      "step": 35280
+    },
+    {
+      "epoch": 0.784082829719161,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.316e-06,
+      "loss": 0.3935,
+      "step": 35290
+    },
+    {
+      "epoch": 0.7843050124422325,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.311555555555556e-06,
+      "loss": 0.3499,
+      "step": 35300
+    },
+    {
+      "epoch": 0.7845271951653039,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.307111111111111e-06,
+      "loss": 0.4539,
+      "step": 35310
+    },
+    {
+      "epoch": 0.7847493778883754,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.302666666666667e-06,
+      "loss": 0.3557,
+      "step": 35320
+    },
+    {
+      "epoch": 0.7849715606114469,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.298222222222223e-06,
+      "loss": 0.3916,
+      "step": 35330
+    },
+    {
+      "epoch": 0.7851937433345183,
+      "grad_norm": 2.5,
+      "learning_rate": 4.293777777777778e-06,
+      "loss": 0.3796,
+      "step": 35340
+    },
+    {
+      "epoch": 0.7854159260575898,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.289333333333334e-06,
+      "loss": 0.4244,
+      "step": 35350
+    },
+    {
+      "epoch": 0.7856381087806612,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.284888888888889e-06,
+      "loss": 0.3966,
+      "step": 35360
+    },
+    {
+      "epoch": 0.7858602915037327,
+      "grad_norm": 1.7421875,
+      "learning_rate": 4.2804444444444445e-06,
+      "loss": 0.3677,
+      "step": 35370
+    },
+    {
+      "epoch": 0.7860824742268041,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.276e-06,
+      "loss": 0.3921,
+      "step": 35380
+    },
+    {
+      "epoch": 0.7863046569498756,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.271555555555556e-06,
+      "loss": 0.4123,
+      "step": 35390
+    },
+    {
+      "epoch": 0.786526839672947,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.267111111111112e-06,
+      "loss": 0.3939,
+      "step": 35400
+    },
+    {
+      "epoch": 0.7867490223960185,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.262666666666667e-06,
+      "loss": 0.3451,
+      "step": 35410
+    },
+    {
+      "epoch": 0.7869712051190899,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.258222222222222e-06,
+      "loss": 0.4128,
+      "step": 35420
+    },
+    {
+      "epoch": 0.7871933878421614,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.253777777777779e-06,
+      "loss": 0.3621,
+      "step": 35430
+    },
+    {
+      "epoch": 0.7874155705652328,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.2493333333333335e-06,
+      "loss": 0.3936,
+      "step": 35440
+    },
+    {
+      "epoch": 0.7876377532883043,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.244888888888889e-06,
+      "loss": 0.3349,
+      "step": 35450
+    },
+    {
+      "epoch": 0.7878599360113757,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.240444444444445e-06,
+      "loss": 0.3618,
+      "step": 35460
+    },
+    {
+      "epoch": 0.7880821187344472,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.236e-06,
+      "loss": 0.3727,
+      "step": 35470
+    },
+    {
+      "epoch": 0.7883043014575186,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.231555555555556e-06,
+      "loss": 0.36,
+      "step": 35480
+    },
+    {
+      "epoch": 0.7885264841805901,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.227111111111111e-06,
+      "loss": 0.3867,
+      "step": 35490
+    },
+    {
+      "epoch": 0.7887486669036615,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.222666666666667e-06,
+      "loss": 0.3428,
+      "step": 35500
+    },
+    {
+      "epoch": 0.788970849626733,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.2182222222222225e-06,
+      "loss": 0.3838,
+      "step": 35510
+    },
+    {
+      "epoch": 0.7891930323498044,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.213777777777778e-06,
+      "loss": 0.3552,
+      "step": 35520
+    },
+    {
+      "epoch": 0.7894152150728759,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.209333333333334e-06,
+      "loss": 0.3469,
+      "step": 35530
+    },
+    {
+      "epoch": 0.7896373977959474,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.20488888888889e-06,
+      "loss": 0.3799,
+      "step": 35540
+    },
+    {
+      "epoch": 0.7898595805190188,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.2004444444444445e-06,
+      "loss": 0.351,
+      "step": 35550
+    },
+    {
+      "epoch": 0.7900817632420903,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.196e-06,
+      "loss": 0.334,
+      "step": 35560
+    },
+    {
+      "epoch": 0.7903039459651617,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.191555555555556e-06,
+      "loss": 0.3548,
+      "step": 35570
+    },
+    {
+      "epoch": 0.7905261286882332,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.1871111111111115e-06,
+      "loss": 0.3785,
+      "step": 35580
+    },
+    {
+      "epoch": 0.7907483114113046,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.182666666666667e-06,
+      "loss": 0.3791,
+      "step": 35590
+    },
+    {
+      "epoch": 0.7909704941343761,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.178222222222222e-06,
+      "loss": 0.3997,
+      "step": 35600
+    },
+    {
+      "epoch": 0.7911926768574475,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.173777777777779e-06,
+      "loss": 0.3568,
+      "step": 35610
+    },
+    {
+      "epoch": 0.791414859580519,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.1693333333333335e-06,
+      "loss": 0.3632,
+      "step": 35620
+    },
+    {
+      "epoch": 0.7916370423035904,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.164888888888889e-06,
+      "loss": 0.4123,
+      "step": 35630
+    },
+    {
+      "epoch": 0.791859225026662,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.160444444444445e-06,
+      "loss": 0.3667,
+      "step": 35640
+    },
+    {
+      "epoch": 0.7920814077497333,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.1560000000000005e-06,
+      "loss": 0.3872,
+      "step": 35650
+    },
+    {
+      "epoch": 0.7923035904728049,
+      "grad_norm": 2.5,
+      "learning_rate": 4.151555555555556e-06,
+      "loss": 0.3686,
+      "step": 35660
+    },
+    {
+      "epoch": 0.7925257731958762,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.147111111111111e-06,
+      "loss": 0.3643,
+      "step": 35670
+    },
+    {
+      "epoch": 0.7927479559189478,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.142666666666667e-06,
+      "loss": 0.3725,
+      "step": 35680
+    },
+    {
+      "epoch": 0.7929701386420192,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.1382222222222224e-06,
+      "loss": 0.4178,
+      "step": 35690
+    },
+    {
+      "epoch": 0.7931923213650907,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.133777777777778e-06,
+      "loss": 0.3527,
+      "step": 35700
+    },
+    {
+      "epoch": 0.793414504088162,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.129333333333334e-06,
+      "loss": 0.4105,
+      "step": 35710
+    },
+    {
+      "epoch": 0.7936366868112336,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.1248888888888895e-06,
+      "loss": 0.3827,
+      "step": 35720
+    },
+    {
+      "epoch": 0.7938588695343051,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.120444444444444e-06,
+      "loss": 0.3481,
+      "step": 35730
+    },
+    {
+      "epoch": 0.7940810522573765,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.116000000000001e-06,
+      "loss": 0.378,
+      "step": 35740
+    },
+    {
+      "epoch": 0.794303234980448,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.111555555555556e-06,
+      "loss": 0.3797,
+      "step": 35750
+    },
+    {
+      "epoch": 0.7945254177035194,
+      "grad_norm": 3.125,
+      "learning_rate": 4.1071111111111114e-06,
+      "loss": 0.4157,
+      "step": 35760
+    },
+    {
+      "epoch": 0.7947476004265909,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.102666666666667e-06,
+      "loss": 0.3935,
+      "step": 35770
+    },
+    {
+      "epoch": 0.7949697831496623,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.098222222222222e-06,
+      "loss": 0.395,
+      "step": 35780
+    },
+    {
+      "epoch": 0.7951919658727338,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.0937777777777785e-06,
+      "loss": 0.3708,
+      "step": 35790
+    },
+    {
+      "epoch": 0.7954141485958052,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.089333333333333e-06,
+      "loss": 0.3731,
+      "step": 35800
+    },
+    {
+      "epoch": 0.7956363313188767,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.084888888888889e-06,
+      "loss": 0.3528,
+      "step": 35810
+    },
+    {
+      "epoch": 0.7958585140419481,
+      "grad_norm": 2.625,
+      "learning_rate": 4.080444444444445e-06,
+      "loss": 0.3945,
+      "step": 35820
+    },
+    {
+      "epoch": 0.7960806967650196,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.0760000000000004e-06,
+      "loss": 0.3837,
+      "step": 35830
+    },
+    {
+      "epoch": 0.796302879488091,
+      "grad_norm": 1.953125,
+      "learning_rate": 4.071555555555556e-06,
+      "loss": 0.3751,
+      "step": 35840
+    },
+    {
+      "epoch": 0.7965250622111625,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.067111111111112e-06,
+      "loss": 0.3691,
+      "step": 35850
+    },
+    {
+      "epoch": 0.7967472449342339,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.062666666666667e-06,
+      "loss": 0.3627,
+      "step": 35860
+    },
+    {
+      "epoch": 0.7969694276573054,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.058222222222222e-06,
+      "loss": 0.3748,
+      "step": 35870
+    },
+    {
+      "epoch": 0.7971916103803768,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.053777777777778e-06,
+      "loss": 0.3858,
+      "step": 35880
+    },
+    {
+      "epoch": 0.7974137931034483,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.049333333333334e-06,
+      "loss": 0.3567,
+      "step": 35890
+    },
+    {
+      "epoch": 0.7976359758265197,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.0448888888888894e-06,
+      "loss": 0.3555,
+      "step": 35900
+    },
+    {
+      "epoch": 0.7978581585495912,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.040444444444444e-06,
+      "loss": 0.3841,
+      "step": 35910
+    },
+    {
+      "epoch": 0.7980803412726626,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.036000000000001e-06,
+      "loss": 0.3513,
+      "step": 35920
+    },
+    {
+      "epoch": 0.7983025239957341,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.031555555555556e-06,
+      "loss": 0.3697,
+      "step": 35930
+    },
+    {
+      "epoch": 0.7985247067188056,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.027111111111111e-06,
+      "loss": 0.3785,
+      "step": 35940
+    },
+    {
+      "epoch": 0.798746889441877,
+      "grad_norm": 2.625,
+      "learning_rate": 4.022666666666667e-06,
+      "loss": 0.3818,
+      "step": 35950
+    },
+    {
+      "epoch": 0.7989690721649485,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.018222222222223e-06,
+      "loss": 0.3765,
+      "step": 35960
+    },
+    {
+      "epoch": 0.7991912548880199,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.0137777777777784e-06,
+      "loss": 0.3733,
+      "step": 35970
+    },
+    {
+      "epoch": 0.7994134376110914,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.009333333333333e-06,
+      "loss": 0.3615,
+      "step": 35980
+    },
+    {
+      "epoch": 0.7996356203341628,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.004888888888889e-06,
+      "loss": 0.3991,
+      "step": 35990
+    },
+    {
+      "epoch": 0.7998578030572343,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.000444444444445e-06,
+      "loss": 0.4218,
+      "step": 36000
+    },
+    {
+      "epoch": 0.8000799857803057,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.996e-06,
+      "loss": 0.4153,
+      "step": 36010
+    },
+    {
+      "epoch": 0.8003021685033772,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.991555555555556e-06,
+      "loss": 0.416,
+      "step": 36020
+    },
+    {
+      "epoch": 0.8005243512264486,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.987111111111112e-06,
+      "loss": 0.3752,
+      "step": 36030
+    },
+    {
+      "epoch": 0.8007465339495201,
+      "grad_norm": 2.25,
+      "learning_rate": 3.982666666666667e-06,
+      "loss": 0.4093,
+      "step": 36040
+    },
+    {
+      "epoch": 0.8009687166725915,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.978222222222223e-06,
+      "loss": 0.3814,
+      "step": 36050
+    },
+    {
+      "epoch": 0.801190899395663,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.973777777777778e-06,
+      "loss": 0.3947,
+      "step": 36060
+    },
+    {
+      "epoch": 0.8014130821187344,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.969333333333334e-06,
+      "loss": 0.4031,
+      "step": 36070
+    },
+    {
+      "epoch": 0.8016352648418059,
+      "grad_norm": 2.375,
+      "learning_rate": 3.964888888888889e-06,
+      "loss": 0.3883,
+      "step": 36080
+    },
+    {
+      "epoch": 0.8018574475648773,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.960444444444444e-06,
+      "loss": 0.3801,
+      "step": 36090
+    },
+    {
+      "epoch": 0.8020796302879488,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.956000000000001e-06,
+      "loss": 0.4001,
+      "step": 36100
+    },
+    {
+      "epoch": 0.8023018130110202,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.951555555555556e-06,
+      "loss": 0.3976,
+      "step": 36110
+    },
+    {
+      "epoch": 0.8025239957340917,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.947111111111111e-06,
+      "loss": 0.4109,
+      "step": 36120
+    },
+    {
+      "epoch": 0.8027461784571631,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.942666666666667e-06,
+      "loss": 0.3503,
+      "step": 36130
+    },
+    {
+      "epoch": 0.8029683611802346,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.938222222222223e-06,
+      "loss": 0.3853,
+      "step": 36140
+    },
+    {
+      "epoch": 0.8031905439033061,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.933777777777778e-06,
+      "loss": 0.4025,
+      "step": 36150
+    },
+    {
+      "epoch": 0.8034127266263775,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.929333333333334e-06,
+      "loss": 0.3681,
+      "step": 36160
+    },
+    {
+      "epoch": 0.803634909349449,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.924888888888889e-06,
+      "loss": 0.3452,
+      "step": 36170
+    },
+    {
+      "epoch": 0.8038570920725204,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.920444444444445e-06,
+      "loss": 0.3523,
+      "step": 36180
+    },
+    {
+      "epoch": 0.8040792747955919,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.916e-06,
+      "loss": 0.3614,
+      "step": 36190
+    },
+    {
+      "epoch": 0.8043014575186633,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.911555555555556e-06,
+      "loss": 0.4246,
+      "step": 36200
+    },
+    {
+      "epoch": 0.8045236402417348,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.907111111111112e-06,
+      "loss": 0.3876,
+      "step": 36210
+    },
+    {
+      "epoch": 0.8047458229648062,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.9026666666666665e-06,
+      "loss": 0.3782,
+      "step": 36220
+    },
+    {
+      "epoch": 0.8049680056878777,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.898222222222223e-06,
+      "loss": 0.4159,
+      "step": 36230
+    },
+    {
+      "epoch": 0.8051901884109491,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.893777777777778e-06,
+      "loss": 0.3788,
+      "step": 36240
+    },
+    {
+      "epoch": 0.8054123711340206,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.889333333333334e-06,
+      "loss": 0.3817,
+      "step": 36250
+    },
+    {
+      "epoch": 0.805634553857092,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.884888888888889e-06,
+      "loss": 0.4005,
+      "step": 36260
+    },
+    {
+      "epoch": 0.8058567365801635,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.880444444444445e-06,
+      "loss": 0.3908,
+      "step": 36270
+    },
+    {
+      "epoch": 0.8060789193032349,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.876000000000001e-06,
+      "loss": 0.3803,
+      "step": 36280
+    },
+    {
+      "epoch": 0.8063011020263064,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.8715555555555555e-06,
+      "loss": 0.3769,
+      "step": 36290
+    },
+    {
+      "epoch": 0.8065232847493778,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.867111111111111e-06,
+      "loss": 0.3546,
+      "step": 36300
+    },
+    {
+      "epoch": 0.8067454674724494,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.862666666666667e-06,
+      "loss": 0.4036,
+      "step": 36310
+    },
+    {
+      "epoch": 0.8069676501955207,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.858222222222223e-06,
+      "loss": 0.3939,
+      "step": 36320
+    },
+    {
+      "epoch": 0.8071898329185923,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.853777777777778e-06,
+      "loss": 0.4026,
+      "step": 36330
+    },
+    {
+      "epoch": 0.8074120156416636,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.849333333333334e-06,
+      "loss": 0.4102,
+      "step": 36340
+    },
+    {
+      "epoch": 0.8076341983647352,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.844888888888889e-06,
+      "loss": 0.3906,
+      "step": 36350
+    },
+    {
+      "epoch": 0.8078563810878067,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.840444444444445e-06,
+      "loss": 0.328,
+      "step": 36360
+    },
+    {
+      "epoch": 0.8080785638108781,
+      "grad_norm": 2.875,
+      "learning_rate": 3.836e-06,
+      "loss": 0.409,
+      "step": 36370
+    },
+    {
+      "epoch": 0.8083007465339496,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.831555555555556e-06,
+      "loss": 0.3531,
+      "step": 36380
+    },
+    {
+      "epoch": 0.808522929257021,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.827111111111112e-06,
+      "loss": 0.3441,
+      "step": 36390
+    },
+    {
+      "epoch": 0.8087451119800925,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.8226666666666664e-06,
+      "loss": 0.3811,
+      "step": 36400
+    },
+    {
+      "epoch": 0.8089672947031639,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.818222222222223e-06,
+      "loss": 0.3645,
+      "step": 36410
+    },
+    {
+      "epoch": 0.8091894774262354,
+      "grad_norm": 2.625,
+      "learning_rate": 3.813777777777778e-06,
+      "loss": 0.3758,
+      "step": 36420
+    },
+    {
+      "epoch": 0.8094116601493068,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.809333333333334e-06,
+      "loss": 0.3884,
+      "step": 36430
+    },
+    {
+      "epoch": 0.8096338428723783,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.804888888888889e-06,
+      "loss": 0.399,
+      "step": 36440
+    },
+    {
+      "epoch": 0.8098560255954497,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.800444444444445e-06,
+      "loss": 0.3821,
+      "step": 36450
+    },
+    {
+      "epoch": 0.8100782083185212,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.796e-06,
+      "loss": 0.3441,
+      "step": 36460
+    },
+    {
+      "epoch": 0.8103003910415926,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.7915555555555563e-06,
+      "loss": 0.3482,
+      "step": 36470
+    },
+    {
+      "epoch": 0.8105225737646641,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.7871111111111115e-06,
+      "loss": 0.3525,
+      "step": 36480
+    },
+    {
+      "epoch": 0.8107447564877355,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.782666666666667e-06,
+      "loss": 0.3821,
+      "step": 36490
+    },
+    {
+      "epoch": 0.810966939210807,
+      "grad_norm": 2.75,
+      "learning_rate": 3.7782222222222225e-06,
+      "loss": 0.364,
+      "step": 36500
+    },
+    {
+      "epoch": 0.8111891219338784,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.7737777777777778e-06,
+      "loss": 0.3863,
+      "step": 36510
+    },
+    {
+      "epoch": 0.8114113046569499,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.769333333333334e-06,
+      "loss": 0.3838,
+      "step": 36520
+    },
+    {
+      "epoch": 0.8116334873800213,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.764888888888889e-06,
+      "loss": 0.3697,
+      "step": 36530
+    },
+    {
+      "epoch": 0.8118556701030928,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.760444444444445e-06,
+      "loss": 0.3424,
+      "step": 36540
+    },
+    {
+      "epoch": 0.8120778528261643,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.756e-06,
+      "loss": 0.3625,
+      "step": 36550
+    },
+    {
+      "epoch": 0.8123000355492357,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.7515555555555562e-06,
+      "loss": 0.3735,
+      "step": 36560
+    },
+    {
+      "epoch": 0.8125222182723072,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.7471111111111115e-06,
+      "loss": 0.4031,
+      "step": 36570
+    },
+    {
+      "epoch": 0.8127444009953786,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.742666666666667e-06,
+      "loss": 0.3842,
+      "step": 36580
+    },
+    {
+      "epoch": 0.8129665837184501,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.7382222222222225e-06,
+      "loss": 0.36,
+      "step": 36590
+    },
+    {
+      "epoch": 0.8131887664415215,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.7337777777777777e-06,
+      "loss": 0.3619,
+      "step": 36600
+    },
+    {
+      "epoch": 0.813410949164593,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.729333333333334e-06,
+      "loss": 0.3943,
+      "step": 36610
+    },
+    {
+      "epoch": 0.8136331318876644,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.724888888888889e-06,
+      "loss": 0.3786,
+      "step": 36620
+    },
+    {
+      "epoch": 0.8138553146107359,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.720444444444445e-06,
+      "loss": 0.3729,
+      "step": 36630
+    },
+    {
+      "epoch": 0.8140774973338073,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.716e-06,
+      "loss": 0.3617,
+      "step": 36640
+    },
+    {
+      "epoch": 0.8142996800568788,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.711555555555556e-06,
+      "loss": 0.4053,
+      "step": 36650
+    },
+    {
+      "epoch": 0.8145218627799502,
+      "grad_norm": 1.9453125,
+      "learning_rate": 3.7071111111111115e-06,
+      "loss": 0.3896,
+      "step": 36660
+    },
+    {
+      "epoch": 0.8147440455030217,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.702666666666667e-06,
+      "loss": 0.4024,
+      "step": 36670
+    },
+    {
+      "epoch": 0.8149662282260931,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.6982222222222224e-06,
+      "loss": 0.3581,
+      "step": 36680
+    },
+    {
+      "epoch": 0.8151884109491646,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.6937777777777785e-06,
+      "loss": 0.3698,
+      "step": 36690
+    },
+    {
+      "epoch": 0.815410593672236,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.689333333333334e-06,
+      "loss": 0.3626,
+      "step": 36700
+    },
+    {
+      "epoch": 0.8156327763953075,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.684888888888889e-06,
+      "loss": 0.4152,
+      "step": 36710
+    },
+    {
+      "epoch": 0.8158549591183789,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.6804444444444448e-06,
+      "loss": 0.3904,
+      "step": 36720
+    },
+    {
+      "epoch": 0.8160771418414504,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.676e-06,
+      "loss": 0.3647,
+      "step": 36730
+    },
+    {
+      "epoch": 0.8162993245645218,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.671555555555556e-06,
+      "loss": 0.3711,
+      "step": 36740
+    },
+    {
+      "epoch": 0.8165215072875933,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.6671111111111114e-06,
+      "loss": 0.3562,
+      "step": 36750
+    },
+    {
+      "epoch": 0.8167436900106648,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.662666666666667e-06,
+      "loss": 0.3905,
+      "step": 36760
+    },
+    {
+      "epoch": 0.8169658727337362,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.6582222222222224e-06,
+      "loss": 0.3713,
+      "step": 36770
+    },
+    {
+      "epoch": 0.8171880554568077,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.6537777777777785e-06,
+      "loss": 0.3718,
+      "step": 36780
+    },
+    {
+      "epoch": 0.8174102381798791,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.6493333333333338e-06,
+      "loss": 0.3669,
+      "step": 36790
+    },
+    {
+      "epoch": 0.8176324209029506,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.644888888888889e-06,
+      "loss": 0.3661,
+      "step": 36800
+    },
+    {
+      "epoch": 0.817854603626022,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.6404444444444447e-06,
+      "loss": 0.4021,
+      "step": 36810
+    },
+    {
+      "epoch": 0.8180767863490935,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.636e-06,
+      "loss": 0.3581,
+      "step": 36820
+    },
+    {
+      "epoch": 0.8182989690721649,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.631555555555556e-06,
+      "loss": 0.3789,
+      "step": 36830
+    },
+    {
+      "epoch": 0.8185211517952364,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.6271111111111114e-06,
+      "loss": 0.3591,
+      "step": 36840
+    },
+    {
+      "epoch": 0.8187433345183078,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.622666666666667e-06,
+      "loss": 0.3703,
+      "step": 36850
+    },
+    {
+      "epoch": 0.8189655172413793,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.6182222222222223e-06,
+      "loss": 0.4259,
+      "step": 36860
+    },
+    {
+      "epoch": 0.8191876999644507,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.6137777777777785e-06,
+      "loss": 0.3824,
+      "step": 36870
+    },
+    {
+      "epoch": 0.8194098826875222,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.6093333333333337e-06,
+      "loss": 0.3941,
+      "step": 36880
+    },
+    {
+      "epoch": 0.8196320654105936,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.6048888888888894e-06,
+      "loss": 0.3845,
+      "step": 36890
+    },
+    {
+      "epoch": 0.8198542481336651,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.6004444444444447e-06,
+      "loss": 0.3932,
+      "step": 36900
+    },
+    {
+      "epoch": 0.8200764308567365,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.596e-06,
+      "loss": 0.3741,
+      "step": 36910
+    },
+    {
+      "epoch": 0.820298613579808,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.591555555555556e-06,
+      "loss": 0.3793,
+      "step": 36920
+    },
+    {
+      "epoch": 0.8205207963028794,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.5871111111111113e-06,
+      "loss": 0.3621,
+      "step": 36930
+    },
+    {
+      "epoch": 0.820742979025951,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.582666666666667e-06,
+      "loss": 0.3599,
+      "step": 36940
+    },
+    {
+      "epoch": 0.8209651617490223,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.5782222222222223e-06,
+      "loss": 0.4144,
+      "step": 36950
+    },
+    {
+      "epoch": 0.8211873444720938,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.5737777777777784e-06,
+      "loss": 0.3619,
+      "step": 36960
+    },
+    {
+      "epoch": 0.8214095271951654,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.5693333333333337e-06,
+      "loss": 0.3865,
+      "step": 36970
+    },
+    {
+      "epoch": 0.8216317099182368,
+      "grad_norm": 2.375,
+      "learning_rate": 3.5648888888888894e-06,
+      "loss": 0.3761,
+      "step": 36980
+    },
+    {
+      "epoch": 0.8218538926413083,
+      "grad_norm": 2.5,
+      "learning_rate": 3.5604444444444447e-06,
+      "loss": 0.3877,
+      "step": 36990
+    },
+    {
+      "epoch": 0.8220760753643797,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.5560000000000008e-06,
+      "loss": 0.3583,
+      "step": 37000
+    },
+    {
+      "epoch": 0.8222982580874512,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.551555555555556e-06,
+      "loss": 0.3937,
+      "step": 37010
+    },
+    {
+      "epoch": 0.8225204408105226,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.5471111111111113e-06,
+      "loss": 0.4002,
+      "step": 37020
+    },
+    {
+      "epoch": 0.8227426235335941,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.542666666666667e-06,
+      "loss": 0.4044,
+      "step": 37030
+    },
+    {
+      "epoch": 0.8229648062566655,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.5382222222222223e-06,
+      "loss": 0.3772,
+      "step": 37040
+    },
+    {
+      "epoch": 0.823186988979737,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.5337777777777784e-06,
+      "loss": 0.3827,
+      "step": 37050
+    },
+    {
+      "epoch": 0.8234091717028084,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.5293333333333336e-06,
+      "loss": 0.3963,
+      "step": 37060
+    },
+    {
+      "epoch": 0.8236313544258799,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.5248888888888893e-06,
+      "loss": 0.4422,
+      "step": 37070
+    },
+    {
+      "epoch": 0.8238535371489513,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.5204444444444446e-06,
+      "loss": 0.4013,
+      "step": 37080
+    },
+    {
+      "epoch": 0.8240757198720228,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.5160000000000007e-06,
+      "loss": 0.3957,
+      "step": 37090
+    },
+    {
+      "epoch": 0.8242979025950942,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.511555555555556e-06,
+      "loss": 0.3686,
+      "step": 37100
+    },
+    {
+      "epoch": 0.8245200853181657,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.5071111111111113e-06,
+      "loss": 0.3786,
+      "step": 37110
+    },
+    {
+      "epoch": 0.8247422680412371,
+      "grad_norm": 2.203125,
+      "learning_rate": 3.502666666666667e-06,
+      "loss": 0.3427,
+      "step": 37120
+    },
+    {
+      "epoch": 0.8249644507643086,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.4982222222222222e-06,
+      "loss": 0.3927,
+      "step": 37130
+    },
+    {
+      "epoch": 0.82518663348738,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.4937777777777783e-06,
+      "loss": 0.395,
+      "step": 37140
+    },
+    {
+      "epoch": 0.8254088162104515,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.4893333333333336e-06,
+      "loss": 0.4044,
+      "step": 37150
+    },
+    {
+      "epoch": 0.8256309989335229,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.4848888888888893e-06,
+      "loss": 0.3699,
+      "step": 37160
+    },
+    {
+      "epoch": 0.8258531816565944,
+      "grad_norm": 2.375,
+      "learning_rate": 3.4804444444444446e-06,
+      "loss": 0.3468,
+      "step": 37170
+    },
+    {
+      "epoch": 0.8260753643796659,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.4760000000000007e-06,
+      "loss": 0.3921,
+      "step": 37180
+    },
+    {
+      "epoch": 0.8262975471027373,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.471555555555556e-06,
+      "loss": 0.3611,
+      "step": 37190
+    },
+    {
+      "epoch": 0.8265197298258088,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.4671111111111116e-06,
+      "loss": 0.361,
+      "step": 37200
+    },
+    {
+      "epoch": 0.8267419125488802,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.462666666666667e-06,
+      "loss": 0.3554,
+      "step": 37210
+    },
+    {
+      "epoch": 0.8269640952719517,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.458222222222222e-06,
+      "loss": 0.3574,
+      "step": 37220
+    },
+    {
+      "epoch": 0.8271862779950231,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.4537777777777783e-06,
+      "loss": 0.3913,
+      "step": 37230
+    },
+    {
+      "epoch": 0.8274084607180946,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.4493333333333336e-06,
+      "loss": 0.3534,
+      "step": 37240
+    },
+    {
+      "epoch": 0.827630643441166,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.4448888888888893e-06,
+      "loss": 0.3659,
+      "step": 37250
+    },
+    {
+      "epoch": 0.8278528261642375,
+      "grad_norm": 2.875,
+      "learning_rate": 3.4404444444444445e-06,
+      "loss": 0.3847,
+      "step": 37260
+    },
+    {
+      "epoch": 0.8280750088873089,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.4360000000000006e-06,
+      "loss": 0.3676,
+      "step": 37270
+    },
+    {
+      "epoch": 0.8282971916103804,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.431555555555556e-06,
+      "loss": 0.3727,
+      "step": 37280
+    },
+    {
+      "epoch": 0.8285193743334518,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.4271111111111116e-06,
+      "loss": 0.36,
+      "step": 37290
+    },
+    {
+      "epoch": 0.8287415570565233,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.422666666666667e-06,
+      "loss": 0.3995,
+      "step": 37300
+    },
+    {
+      "epoch": 0.8289637397795947,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.4182222222222226e-06,
+      "loss": 0.3648,
+      "step": 37310
+    },
+    {
+      "epoch": 0.8291859225026662,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.4137777777777783e-06,
+      "loss": 0.4034,
+      "step": 37320
+    },
+    {
+      "epoch": 0.8294081052257376,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.4093333333333335e-06,
+      "loss": 0.3502,
+      "step": 37330
+    },
+    {
+      "epoch": 0.8296302879488091,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.4048888888888892e-06,
+      "loss": 0.3505,
+      "step": 37340
+    },
+    {
+      "epoch": 0.8298524706718805,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.4004444444444445e-06,
+      "loss": 0.3957,
+      "step": 37350
+    },
+    {
+      "epoch": 0.830074653394952,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.3960000000000006e-06,
+      "loss": 0.3776,
+      "step": 37360
+    },
+    {
+      "epoch": 0.8302968361180235,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.391555555555556e-06,
+      "loss": 0.3782,
+      "step": 37370
+    },
+    {
+      "epoch": 0.8305190188410949,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.3871111111111116e-06,
+      "loss": 0.3891,
+      "step": 37380
+    },
+    {
+      "epoch": 0.8307412015641664,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.382666666666667e-06,
+      "loss": 0.3471,
+      "step": 37390
+    },
+    {
+      "epoch": 0.8309633842872378,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.3782222222222225e-06,
+      "loss": 0.4007,
+      "step": 37400
+    },
+    {
+      "epoch": 0.8311855670103093,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.3737777777777782e-06,
+      "loss": 0.3866,
+      "step": 37410
+    },
+    {
+      "epoch": 0.8314077497333807,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.3693333333333335e-06,
+      "loss": 0.3708,
+      "step": 37420
+    },
+    {
+      "epoch": 0.8316299324564522,
+      "grad_norm": 2.375,
+      "learning_rate": 3.364888888888889e-06,
+      "loss": 0.3908,
+      "step": 37430
+    },
+    {
+      "epoch": 0.8318521151795236,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.3604444444444444e-06,
+      "loss": 0.3849,
+      "step": 37440
+    },
+    {
+      "epoch": 0.8320742979025951,
+      "grad_norm": 2.75,
+      "learning_rate": 3.3560000000000006e-06,
+      "loss": 0.3863,
+      "step": 37450
+    },
+    {
+      "epoch": 0.8322964806256665,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.351555555555556e-06,
+      "loss": 0.3868,
+      "step": 37460
+    },
+    {
+      "epoch": 0.832518663348738,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.3471111111111115e-06,
+      "loss": 0.3625,
+      "step": 37470
+    },
+    {
+      "epoch": 0.8327408460718094,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.342666666666667e-06,
+      "loss": 0.3774,
+      "step": 37480
+    },
+    {
+      "epoch": 0.8329630287948809,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.3382222222222225e-06,
+      "loss": 0.4237,
+      "step": 37490
+    },
+    {
+      "epoch": 0.8331852115179523,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.333777777777778e-06,
+      "loss": 0.3394,
+      "step": 37500
+    },
+    {
+      "epoch": 0.8334073942410238,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.329333333333334e-06,
+      "loss": 0.4392,
+      "step": 37510
+    },
+    {
+      "epoch": 0.8336295769640952,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.324888888888889e-06,
+      "loss": 0.3995,
+      "step": 37520
+    },
+    {
+      "epoch": 0.8338517596871667,
+      "grad_norm": 2.171875,
+      "learning_rate": 3.3204444444444444e-06,
+      "loss": 0.3664,
+      "step": 37530
+    },
+    {
+      "epoch": 0.8340739424102381,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.3160000000000005e-06,
+      "loss": 0.3388,
+      "step": 37540
+    },
+    {
+      "epoch": 0.8342961251333096,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.311555555555556e-06,
+      "loss": 0.3804,
+      "step": 37550
+    },
+    {
+      "epoch": 0.834518307856381,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.3071111111111115e-06,
+      "loss": 0.3699,
+      "step": 37560
+    },
+    {
+      "epoch": 0.8347404905794525,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.3026666666666668e-06,
+      "loss": 0.3796,
+      "step": 37570
+    },
+    {
+      "epoch": 0.834962673302524,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.298222222222223e-06,
+      "loss": 0.3949,
+      "step": 37580
+    },
+    {
+      "epoch": 0.8351848560255954,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.293777777777778e-06,
+      "loss": 0.3359,
+      "step": 37590
+    },
+    {
+      "epoch": 0.835407038748667,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.289333333333334e-06,
+      "loss": 0.3812,
+      "step": 37600
+    },
+    {
+      "epoch": 0.8356292214717383,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.284888888888889e-06,
+      "loss": 0.3936,
+      "step": 37610
+    },
+    {
+      "epoch": 0.8358514041948099,
+      "grad_norm": 3.0,
+      "learning_rate": 3.280444444444445e-06,
+      "loss": 0.396,
+      "step": 37620
+    },
+    {
+      "epoch": 0.8360735869178813,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.2760000000000005e-06,
+      "loss": 0.4006,
+      "step": 37630
+    },
+    {
+      "epoch": 0.8362957696409528,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.2715555555555558e-06,
+      "loss": 0.4128,
+      "step": 37640
+    },
+    {
+      "epoch": 0.8365179523640242,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.2671111111111114e-06,
+      "loss": 0.3945,
+      "step": 37650
+    },
+    {
+      "epoch": 0.8367401350870957,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.2626666666666667e-06,
+      "loss": 0.3853,
+      "step": 37660
+    },
+    {
+      "epoch": 0.8369623178101671,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.258222222222223e-06,
+      "loss": 0.3778,
+      "step": 37670
+    },
+    {
+      "epoch": 0.8371845005332386,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.253777777777778e-06,
+      "loss": 0.3505,
+      "step": 37680
+    },
+    {
+      "epoch": 0.83740668325631,
+      "grad_norm": 2.109375,
+      "learning_rate": 3.249333333333334e-06,
+      "loss": 0.3252,
+      "step": 37690
+    },
+    {
+      "epoch": 0.8376288659793815,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.244888888888889e-06,
+      "loss": 0.3809,
+      "step": 37700
+    },
+    {
+      "epoch": 0.8378510487024529,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.2404444444444448e-06,
+      "loss": 0.3932,
+      "step": 37710
+    },
+    {
+      "epoch": 0.8380732314255244,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.2360000000000004e-06,
+      "loss": 0.3969,
+      "step": 37720
+    },
+    {
+      "epoch": 0.8382954141485958,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.2315555555555557e-06,
+      "loss": 0.3881,
+      "step": 37730
+    },
+    {
+      "epoch": 0.8385175968716673,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.2271111111111114e-06,
+      "loss": 0.402,
+      "step": 37740
+    },
+    {
+      "epoch": 0.8387397795947387,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.2226666666666667e-06,
+      "loss": 0.3891,
+      "step": 37750
+    },
+    {
+      "epoch": 0.8389619623178102,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.2182222222222228e-06,
+      "loss": 0.3798,
+      "step": 37760
+    },
+    {
+      "epoch": 0.8391841450408816,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.213777777777778e-06,
+      "loss": 0.381,
+      "step": 37770
+    },
+    {
+      "epoch": 0.8394063277639531,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.2093333333333337e-06,
+      "loss": 0.3803,
+      "step": 37780
+    },
+    {
+      "epoch": 0.8396285104870246,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.204888888888889e-06,
+      "loss": 0.3757,
+      "step": 37790
+    },
+    {
+      "epoch": 0.839850693210096,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.2004444444444447e-06,
+      "loss": 0.3806,
+      "step": 37800
+    },
+    {
+      "epoch": 0.8400728759331675,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.1960000000000004e-06,
+      "loss": 0.4007,
+      "step": 37810
+    },
+    {
+      "epoch": 0.8402950586562389,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.191555555555556e-06,
+      "loss": 0.4154,
+      "step": 37820
+    },
+    {
+      "epoch": 0.8405172413793104,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.1871111111111114e-06,
+      "loss": 0.4186,
+      "step": 37830
+    },
+    {
+      "epoch": 0.8407394241023818,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.1826666666666666e-06,
+      "loss": 0.4009,
+      "step": 37840
+    },
+    {
+      "epoch": 0.8409616068254533,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.1782222222222227e-06,
+      "loss": 0.4162,
+      "step": 37850
+    },
+    {
+      "epoch": 0.8411837895485247,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.173777777777778e-06,
+      "loss": 0.3911,
+      "step": 37860
+    },
+    {
+      "epoch": 0.8414059722715962,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.1693333333333337e-06,
+      "loss": 0.3891,
+      "step": 37870
+    },
+    {
+      "epoch": 0.8416281549946676,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.164888888888889e-06,
+      "loss": 0.3632,
+      "step": 37880
+    },
+    {
+      "epoch": 0.8418503377177391,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.1604444444444447e-06,
+      "loss": 0.3702,
+      "step": 37890
+    },
+    {
+      "epoch": 0.8420725204408105,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.1560000000000004e-06,
+      "loss": 0.3673,
+      "step": 37900
+    },
+    {
+      "epoch": 0.842294703163882,
+      "grad_norm": 2.375,
+      "learning_rate": 3.151555555555556e-06,
+      "loss": 0.3518,
+      "step": 37910
+    },
+    {
+      "epoch": 0.8425168858869534,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.1471111111111113e-06,
+      "loss": 0.382,
+      "step": 37920
+    },
+    {
+      "epoch": 0.8427390686100249,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.142666666666667e-06,
+      "loss": 0.377,
+      "step": 37930
+    },
+    {
+      "epoch": 0.8429612513330963,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.1382222222222227e-06,
+      "loss": 0.4045,
+      "step": 37940
+    },
+    {
+      "epoch": 0.8431834340561678,
+      "grad_norm": 3.25,
+      "learning_rate": 3.133777777777778e-06,
+      "loss": 0.3554,
+      "step": 37950
+    },
+    {
+      "epoch": 0.8434056167792392,
+      "grad_norm": 2.75,
+      "learning_rate": 3.1293333333333337e-06,
+      "loss": 0.3976,
+      "step": 37960
+    },
+    {
+      "epoch": 0.8436277995023107,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.124888888888889e-06,
+      "loss": 0.4088,
+      "step": 37970
+    },
+    {
+      "epoch": 0.8438499822253821,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.1204444444444446e-06,
+      "loss": 0.3825,
+      "step": 37980
+    },
+    {
+      "epoch": 0.8440721649484536,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.1160000000000003e-06,
+      "loss": 0.3749,
+      "step": 37990
+    },
+    {
+      "epoch": 0.8442943476715251,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.111555555555556e-06,
+      "loss": 0.3941,
+      "step": 38000
+    },
+    {
+      "epoch": 0.8445165303945965,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.1071111111111113e-06,
+      "loss": 0.3638,
+      "step": 38010
+    },
+    {
+      "epoch": 0.844738713117668,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.102666666666667e-06,
+      "loss": 0.3978,
+      "step": 38020
+    },
+    {
+      "epoch": 0.8449608958407394,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.0982222222222227e-06,
+      "loss": 0.3633,
+      "step": 38030
+    },
+    {
+      "epoch": 0.8451830785638109,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.093777777777778e-06,
+      "loss": 0.3862,
+      "step": 38040
+    },
+    {
+      "epoch": 0.8454052612868823,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.0893333333333336e-06,
+      "loss": 0.3785,
+      "step": 38050
+    },
+    {
+      "epoch": 0.8456274440099538,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.084888888888889e-06,
+      "loss": 0.389,
+      "step": 38060
+    },
+    {
+      "epoch": 0.8458496267330252,
+      "grad_norm": 2.875,
+      "learning_rate": 3.0804444444444446e-06,
+      "loss": 0.3706,
+      "step": 38070
+    },
+    {
+      "epoch": 0.8460718094560967,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.0760000000000003e-06,
+      "loss": 0.3671,
+      "step": 38080
+    },
+    {
+      "epoch": 0.8462939921791681,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.071555555555556e-06,
+      "loss": 0.4127,
+      "step": 38090
+    },
+    {
+      "epoch": 0.8465161749022396,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.0671111111111112e-06,
+      "loss": 0.3619,
+      "step": 38100
+    },
+    {
+      "epoch": 0.846738357625311,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.062666666666667e-06,
+      "loss": 0.3844,
+      "step": 38110
+    },
+    {
+      "epoch": 0.8469605403483825,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.0582222222222226e-06,
+      "loss": 0.4027,
+      "step": 38120
+    },
+    {
+      "epoch": 0.8471827230714539,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.0537777777777783e-06,
+      "loss": 0.379,
+      "step": 38130
+    },
+    {
+      "epoch": 0.8474049057945254,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.0493333333333336e-06,
+      "loss": 0.4034,
+      "step": 38140
+    },
+    {
+      "epoch": 0.8476270885175968,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.044888888888889e-06,
+      "loss": 0.4175,
+      "step": 38150
+    },
+    {
+      "epoch": 0.8478492712406683,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.0404444444444445e-06,
+      "loss": 0.3888,
+      "step": 38160
+    },
+    {
+      "epoch": 0.8480714539637397,
+      "grad_norm": 2.390625,
+      "learning_rate": 3.0360000000000002e-06,
+      "loss": 0.349,
+      "step": 38170
+    },
+    {
+      "epoch": 0.8482936366868112,
+      "grad_norm": 2.375,
+      "learning_rate": 3.031555555555556e-06,
+      "loss": 0.3918,
+      "step": 38180
+    },
+    {
+      "epoch": 0.8485158194098826,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.027111111111111e-06,
+      "loss": 0.3855,
+      "step": 38190
+    },
+    {
+      "epoch": 0.8487380021329541,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.022666666666667e-06,
+      "loss": 0.3281,
+      "step": 38200
+    },
+    {
+      "epoch": 0.8489601848560256,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.0182222222222226e-06,
+      "loss": 0.3725,
+      "step": 38210
+    },
+    {
+      "epoch": 0.849182367579097,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.0137777777777783e-06,
+      "loss": 0.3511,
+      "step": 38220
+    },
+    {
+      "epoch": 0.8494045503021685,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.0093333333333335e-06,
+      "loss": 0.3693,
+      "step": 38230
+    },
+    {
+      "epoch": 0.8496267330252399,
+      "grad_norm": 2.234375,
+      "learning_rate": 3.0048888888888892e-06,
+      "loss": 0.3991,
+      "step": 38240
+    },
+    {
+      "epoch": 0.8498489157483115,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.0004444444444445e-06,
+      "loss": 0.3783,
+      "step": 38250
+    },
+    {
+      "epoch": 0.8500710984713828,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.996e-06,
+      "loss": 0.3797,
+      "step": 38260
+    },
+    {
+      "epoch": 0.8502932811944544,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.991555555555556e-06,
+      "loss": 0.3732,
+      "step": 38270
+    },
+    {
+      "epoch": 0.8505154639175257,
+      "grad_norm": 2.125,
+      "learning_rate": 2.987111111111111e-06,
+      "loss": 0.3435,
+      "step": 38280
+    },
+    {
+      "epoch": 0.8507376466405973,
+      "grad_norm": 1.96875,
+      "learning_rate": 2.982666666666667e-06,
+      "loss": 0.3585,
+      "step": 38290
+    },
+    {
+      "epoch": 0.8509598293636687,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9782222222222225e-06,
+      "loss": 0.3673,
+      "step": 38300
+    },
+    {
+      "epoch": 0.8511820120867402,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.9737777777777782e-06,
+      "loss": 0.3812,
+      "step": 38310
+    },
+    {
+      "epoch": 0.8514041948098116,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.9693333333333335e-06,
+      "loss": 0.3847,
+      "step": 38320
+    },
+    {
+      "epoch": 0.8516263775328831,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.964888888888889e-06,
+      "loss": 0.379,
+      "step": 38330
+    },
+    {
+      "epoch": 0.8518485602559545,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9604444444444445e-06,
+      "loss": 0.3578,
+      "step": 38340
+    },
+    {
+      "epoch": 0.852070742979026,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.956e-06,
+      "loss": 0.3734,
+      "step": 38350
+    },
+    {
+      "epoch": 0.8522929257020974,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.951555555555556e-06,
+      "loss": 0.3857,
+      "step": 38360
+    },
+    {
+      "epoch": 0.8525151084251689,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.947111111111111e-06,
+      "loss": 0.3981,
+      "step": 38370
+    },
+    {
+      "epoch": 0.8527372911482403,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.942666666666667e-06,
+      "loss": 0.3664,
+      "step": 38380
+    },
+    {
+      "epoch": 0.8529594738713118,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.9382222222222225e-06,
+      "loss": 0.3549,
+      "step": 38390
+    },
+    {
+      "epoch": 0.8531816565943833,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.933777777777778e-06,
+      "loss": 0.3706,
+      "step": 38400
+    },
+    {
+      "epoch": 0.8534038393174547,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.9293333333333335e-06,
+      "loss": 0.3724,
+      "step": 38410
+    },
+    {
+      "epoch": 0.8536260220405262,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.924888888888889e-06,
+      "loss": 0.411,
+      "step": 38420
+    },
+    {
+      "epoch": 0.8538482047635976,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.9204444444444444e-06,
+      "loss": 0.3301,
+      "step": 38430
+    },
+    {
+      "epoch": 0.8540703874866691,
+      "grad_norm": 2.75,
+      "learning_rate": 2.9160000000000005e-06,
+      "loss": 0.3766,
+      "step": 38440
+    },
+    {
+      "epoch": 0.8542925702097405,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.911555555555556e-06,
+      "loss": 0.3619,
+      "step": 38450
+    },
+    {
+      "epoch": 0.854514752932812,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.907111111111111e-06,
+      "loss": 0.3743,
+      "step": 38460
+    },
+    {
+      "epoch": 0.8547369356558834,
+      "grad_norm": 2.25,
+      "learning_rate": 2.9026666666666668e-06,
+      "loss": 0.3742,
+      "step": 38470
+    },
+    {
+      "epoch": 0.8549591183789549,
+      "grad_norm": 2.25,
+      "learning_rate": 2.8982222222222225e-06,
+      "loss": 0.3755,
+      "step": 38480
+    },
+    {
+      "epoch": 0.8551813011020263,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.893777777777778e-06,
+      "loss": 0.3777,
+      "step": 38490
+    },
+    {
+      "epoch": 0.8554034838250978,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8893333333333334e-06,
+      "loss": 0.3395,
+      "step": 38500
+    },
+    {
+      "epoch": 0.8556256665481692,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.884888888888889e-06,
+      "loss": 0.3713,
+      "step": 38510
+    },
+    {
+      "epoch": 0.8558478492712407,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.880444444444445e-06,
+      "loss": 0.3555,
+      "step": 38520
+    },
+    {
+      "epoch": 0.8560700319943121,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.8760000000000005e-06,
+      "loss": 0.3666,
+      "step": 38530
+    },
+    {
+      "epoch": 0.8562922147173836,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.8715555555555558e-06,
+      "loss": 0.3908,
+      "step": 38540
+    },
+    {
+      "epoch": 0.856514397440455,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.8671111111111115e-06,
+      "loss": 0.3939,
+      "step": 38550
+    },
+    {
+      "epoch": 0.8567365801635265,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.8626666666666667e-06,
+      "loss": 0.402,
+      "step": 38560
+    },
+    {
+      "epoch": 0.8569587628865979,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.8582222222222224e-06,
+      "loss": 0.3823,
+      "step": 38570
+    },
+    {
+      "epoch": 0.8571809456096694,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.853777777777778e-06,
+      "loss": 0.3862,
+      "step": 38580
+    },
+    {
+      "epoch": 0.8574031283327408,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8493333333333334e-06,
+      "loss": 0.3953,
+      "step": 38590
+    },
+    {
+      "epoch": 0.8576253110558123,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.844888888888889e-06,
+      "loss": 0.392,
+      "step": 38600
+    },
+    {
+      "epoch": 0.8578474937788838,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.8404444444444448e-06,
+      "loss": 0.3912,
+      "step": 38610
+    },
+    {
+      "epoch": 0.8580696765019552,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.8360000000000005e-06,
+      "loss": 0.3748,
+      "step": 38620
+    },
+    {
+      "epoch": 0.8582918592250267,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.8315555555555557e-06,
+      "loss": 0.3653,
+      "step": 38630
+    },
+    {
+      "epoch": 0.8585140419480981,
+      "grad_norm": 2.375,
+      "learning_rate": 2.8271111111111114e-06,
+      "loss": 0.3418,
+      "step": 38640
+    },
+    {
+      "epoch": 0.8587362246711696,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8226666666666667e-06,
+      "loss": 0.3835,
+      "step": 38650
+    },
+    {
+      "epoch": 0.858958407394241,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.8182222222222224e-06,
+      "loss": 0.3827,
+      "step": 38660
+    },
+    {
+      "epoch": 0.8591805901173125,
+      "grad_norm": 2.375,
+      "learning_rate": 2.813777777777778e-06,
+      "loss": 0.3809,
+      "step": 38670
+    },
+    {
+      "epoch": 0.8594027728403839,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8093333333333333e-06,
+      "loss": 0.3883,
+      "step": 38680
+    },
+    {
+      "epoch": 0.8596249555634554,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.804888888888889e-06,
+      "loss": 0.3659,
+      "step": 38690
+    },
+    {
+      "epoch": 0.8598471382865268,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8004444444444447e-06,
+      "loss": 0.3739,
+      "step": 38700
+    },
+    {
+      "epoch": 0.8600693210095983,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.7960000000000004e-06,
+      "loss": 0.354,
+      "step": 38710
+    },
+    {
+      "epoch": 0.8602915037326697,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.7915555555555557e-06,
+      "loss": 0.3551,
+      "step": 38720
+    },
+    {
+      "epoch": 0.8605136864557412,
+      "grad_norm": 2.5,
+      "learning_rate": 2.7871111111111114e-06,
+      "loss": 0.3688,
+      "step": 38730
+    },
+    {
+      "epoch": 0.8607358691788126,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.7826666666666666e-06,
+      "loss": 0.3777,
+      "step": 38740
+    },
+    {
+      "epoch": 0.8609580519018841,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.7782222222222228e-06,
+      "loss": 0.4131,
+      "step": 38750
+    },
+    {
+      "epoch": 0.8611802346249555,
+      "grad_norm": 3.0,
+      "learning_rate": 2.773777777777778e-06,
+      "loss": 0.3677,
+      "step": 38760
+    },
+    {
+      "epoch": 0.861402417348027,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.7693333333333333e-06,
+      "loss": 0.3922,
+      "step": 38770
+    },
+    {
+      "epoch": 0.8616246000710984,
+      "grad_norm": 2.875,
+      "learning_rate": 2.764888888888889e-06,
+      "loss": 0.378,
+      "step": 38780
+    },
+    {
+      "epoch": 0.8618467827941699,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7604444444444447e-06,
+      "loss": 0.3579,
+      "step": 38790
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.7560000000000004e-06,
+      "loss": 0.3672,
+      "step": 38800
+    },
+    {
+      "epoch": 0.8622911482403128,
+      "grad_norm": 2.5,
+      "learning_rate": 2.7515555555555556e-06,
+      "loss": 0.3729,
+      "step": 38810
+    },
+    {
+      "epoch": 0.8625133309633843,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.7471111111111113e-06,
+      "loss": 0.3626,
+      "step": 38820
+    },
+    {
+      "epoch": 0.8627355136864557,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.7426666666666666e-06,
+      "loss": 0.4237,
+      "step": 38830
+    },
+    {
+      "epoch": 0.8629576964095272,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.7382222222222227e-06,
+      "loss": 0.4017,
+      "step": 38840
+    },
+    {
+      "epoch": 0.8631798791325986,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.733777777777778e-06,
+      "loss": 0.3437,
+      "step": 38850
+    },
+    {
+      "epoch": 0.8634020618556701,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.7293333333333333e-06,
+      "loss": 0.3578,
+      "step": 38860
+    },
+    {
+      "epoch": 0.8636242445787415,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.724888888888889e-06,
+      "loss": 0.3849,
+      "step": 38870
+    },
+    {
+      "epoch": 0.863846427301813,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.7204444444444446e-06,
+      "loss": 0.3795,
+      "step": 38880
+    },
+    {
+      "epoch": 0.8640686100248844,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.7160000000000003e-06,
+      "loss": 0.3713,
+      "step": 38890
+    },
+    {
+      "epoch": 0.864290792747956,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.7115555555555556e-06,
+      "loss": 0.3595,
+      "step": 38900
+    },
+    {
+      "epoch": 0.8645129754710273,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.7071111111111113e-06,
+      "loss": 0.3469,
+      "step": 38910
+    },
+    {
+      "epoch": 0.8647351581940989,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.7026666666666666e-06,
+      "loss": 0.3666,
+      "step": 38920
+    },
+    {
+      "epoch": 0.8649573409171702,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.6982222222222227e-06,
+      "loss": 0.3731,
+      "step": 38930
+    },
+    {
+      "epoch": 0.8651795236402418,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.693777777777778e-06,
+      "loss": 0.385,
+      "step": 38940
+    },
+    {
+      "epoch": 0.8654017063633132,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.6893333333333336e-06,
+      "loss": 0.3696,
+      "step": 38950
+    },
+    {
+      "epoch": 0.8656238890863847,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.684888888888889e-06,
+      "loss": 0.4151,
+      "step": 38960
+    },
+    {
+      "epoch": 0.865846071809456,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.6804444444444446e-06,
+      "loss": 0.3815,
+      "step": 38970
+    },
+    {
+      "epoch": 0.8660682545325276,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.6760000000000003e-06,
+      "loss": 0.3737,
+      "step": 38980
+    },
+    {
+      "epoch": 0.866290437255599,
+      "grad_norm": 2.75,
+      "learning_rate": 2.6715555555555556e-06,
+      "loss": 0.378,
+      "step": 38990
+    },
+    {
+      "epoch": 0.8665126199786705,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.6671111111111113e-06,
+      "loss": 0.3694,
+      "step": 39000
+    },
+    {
+      "epoch": 0.8667348027017419,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.6626666666666665e-06,
+      "loss": 0.3726,
+      "step": 39010
+    },
+    {
+      "epoch": 0.8669569854248134,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.6582222222222226e-06,
+      "loss": 0.3751,
+      "step": 39020
+    },
+    {
+      "epoch": 0.8671791681478849,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.653777777777778e-06,
+      "loss": 0.3833,
+      "step": 39030
+    },
+    {
+      "epoch": 0.8674013508709563,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.6493333333333336e-06,
+      "loss": 0.4114,
+      "step": 39040
+    },
+    {
+      "epoch": 0.8676235335940278,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.644888888888889e-06,
+      "loss": 0.3759,
+      "step": 39050
+    },
+    {
+      "epoch": 0.8678457163170992,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.640444444444445e-06,
+      "loss": 0.3894,
+      "step": 39060
+    },
+    {
+      "epoch": 0.8680678990401707,
+      "grad_norm": 2.625,
+      "learning_rate": 2.6360000000000003e-06,
+      "loss": 0.3688,
+      "step": 39070
+    },
+    {
+      "epoch": 0.8682900817632421,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.6315555555555555e-06,
+      "loss": 0.3668,
+      "step": 39080
+    },
+    {
+      "epoch": 0.8685122644863136,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.6271111111111112e-06,
+      "loss": 0.3603,
+      "step": 39090
+    },
+    {
+      "epoch": 0.868734447209385,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.6226666666666665e-06,
+      "loss": 0.3859,
+      "step": 39100
+    },
+    {
+      "epoch": 0.8689566299324565,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.6182222222222226e-06,
+      "loss": 0.3455,
+      "step": 39110
+    },
+    {
+      "epoch": 0.8691788126555279,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.613777777777778e-06,
+      "loss": 0.3751,
+      "step": 39120
+    },
+    {
+      "epoch": 0.8694009953785994,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.6093333333333336e-06,
+      "loss": 0.3494,
+      "step": 39130
+    },
+    {
+      "epoch": 0.8696231781016708,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.604888888888889e-06,
+      "loss": 0.4063,
+      "step": 39140
+    },
+    {
+      "epoch": 0.8698453608247423,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.600444444444445e-06,
+      "loss": 0.4022,
+      "step": 39150
+    },
+    {
+      "epoch": 0.8700675435478137,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.5960000000000002e-06,
+      "loss": 0.3903,
+      "step": 39160
+    },
+    {
+      "epoch": 0.8702897262708852,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.5915555555555555e-06,
+      "loss": 0.4057,
+      "step": 39170
+    },
+    {
+      "epoch": 0.8705119089939566,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.587111111111111e-06,
+      "loss": 0.3593,
+      "step": 39180
+    },
+    {
+      "epoch": 0.8707340917170281,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.5826666666666664e-06,
+      "loss": 0.3396,
+      "step": 39190
+    },
+    {
+      "epoch": 0.8709562744400995,
+      "grad_norm": 2.375,
+      "learning_rate": 2.5782222222222226e-06,
+      "loss": 0.3696,
+      "step": 39200
+    },
+    {
+      "epoch": 0.871178457163171,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.573777777777778e-06,
+      "loss": 0.3804,
+      "step": 39210
+    },
+    {
+      "epoch": 0.8714006398862425,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.5693333333333335e-06,
+      "loss": 0.4056,
+      "step": 39220
+    },
+    {
+      "epoch": 0.8716228226093139,
+      "grad_norm": 2.375,
+      "learning_rate": 2.564888888888889e-06,
+      "loss": 0.3712,
+      "step": 39230
+    },
+    {
+      "epoch": 0.8718450053323854,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.560444444444445e-06,
+      "loss": 0.3899,
+      "step": 39240
+    },
+    {
+      "epoch": 0.8720671880554568,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.556e-06,
+      "loss": 0.3869,
+      "step": 39250
+    },
+    {
+      "epoch": 0.8722893707785283,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.551555555555556e-06,
+      "loss": 0.3569,
+      "step": 39260
+    },
+    {
+      "epoch": 0.8725115535015997,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.547111111111111e-06,
+      "loss": 0.3602,
+      "step": 39270
+    },
+    {
+      "epoch": 0.8727337362246712,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.5426666666666664e-06,
+      "loss": 0.3651,
+      "step": 39280
+    },
+    {
+      "epoch": 0.8729559189477426,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.5382222222222225e-06,
+      "loss": 0.4032,
+      "step": 39290
+    },
+    {
+      "epoch": 0.8731781016708141,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.533777777777778e-06,
+      "loss": 0.4008,
+      "step": 39300
+    },
+    {
+      "epoch": 0.8734002843938855,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.5293333333333335e-06,
+      "loss": 0.3781,
+      "step": 39310
+    },
+    {
+      "epoch": 0.873622467116957,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.5248888888888888e-06,
+      "loss": 0.3624,
+      "step": 39320
+    },
+    {
+      "epoch": 0.8738446498400284,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.520444444444445e-06,
+      "loss": 0.3744,
+      "step": 39330
+    },
+    {
+      "epoch": 0.8740668325630999,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.516e-06,
+      "loss": 0.3678,
+      "step": 39340
+    },
+    {
+      "epoch": 0.8742890152861713,
+      "grad_norm": 2.078125,
+      "learning_rate": 2.511555555555556e-06,
+      "loss": 0.3681,
+      "step": 39350
+    },
+    {
+      "epoch": 0.8745111980092428,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.507111111111111e-06,
+      "loss": 0.3679,
+      "step": 39360
+    },
+    {
+      "epoch": 0.8747333807323142,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.5026666666666672e-06,
+      "loss": 0.3866,
+      "step": 39370
+    },
+    {
+      "epoch": 0.8749555634553857,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.4982222222222225e-06,
+      "loss": 0.3862,
+      "step": 39380
+    },
+    {
+      "epoch": 0.8751777461784571,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.493777777777778e-06,
+      "loss": 0.3422,
+      "step": 39390
+    },
+    {
+      "epoch": 0.8753999289015286,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.4893333333333334e-06,
+      "loss": 0.3805,
+      "step": 39400
+    },
+    {
+      "epoch": 0.8756221116246,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.484888888888889e-06,
+      "loss": 0.4086,
+      "step": 39410
+    },
+    {
+      "epoch": 0.8758442943476715,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.480444444444445e-06,
+      "loss": 0.3776,
+      "step": 39420
+    },
+    {
+      "epoch": 0.876066477070743,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.476e-06,
+      "loss": 0.3834,
+      "step": 39430
+    },
+    {
+      "epoch": 0.8762886597938144,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.4715555555555558e-06,
+      "loss": 0.3931,
+      "step": 39440
+    },
+    {
+      "epoch": 0.8765108425168859,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.467111111111111e-06,
+      "loss": 0.3288,
+      "step": 39450
+    },
+    {
+      "epoch": 0.8767330252399573,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.4626666666666667e-06,
+      "loss": 0.4021,
+      "step": 39460
+    },
+    {
+      "epoch": 0.8769552079630288,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.4582222222222224e-06,
+      "loss": 0.3973,
+      "step": 39470
+    },
+    {
+      "epoch": 0.8771773906861002,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.453777777777778e-06,
+      "loss": 0.3773,
+      "step": 39480
+    },
+    {
+      "epoch": 0.8773995734091717,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.4493333333333334e-06,
+      "loss": 0.3593,
+      "step": 39490
+    },
+    {
+      "epoch": 0.8776217561322431,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.444888888888889e-06,
+      "loss": 0.4039,
+      "step": 39500
+    },
+    {
+      "epoch": 0.8778439388553146,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.4404444444444448e-06,
+      "loss": 0.3548,
+      "step": 39510
+    },
+    {
+      "epoch": 0.878066121578386,
+      "grad_norm": 2.625,
+      "learning_rate": 2.4360000000000005e-06,
+      "loss": 0.3488,
+      "step": 39520
+    },
+    {
+      "epoch": 0.8782883043014575,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.4315555555555557e-06,
+      "loss": 0.3653,
+      "step": 39530
+    },
+    {
+      "epoch": 0.8785104870245289,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.427111111111111e-06,
+      "loss": 0.3435,
+      "step": 39540
+    },
+    {
+      "epoch": 0.8787326697476004,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.4226666666666667e-06,
+      "loss": 0.4013,
+      "step": 39550
+    },
+    {
+      "epoch": 0.8789548524706718,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.4182222222222224e-06,
+      "loss": 0.379,
+      "step": 39560
+    },
+    {
+      "epoch": 0.8791770351937434,
+      "grad_norm": 2.140625,
+      "learning_rate": 2.413777777777778e-06,
+      "loss": 0.334,
+      "step": 39570
+    },
+    {
+      "epoch": 0.8793992179168147,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.4093333333333334e-06,
+      "loss": 0.3664,
+      "step": 39580
+    },
+    {
+      "epoch": 0.8796214006398863,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.404888888888889e-06,
+      "loss": 0.332,
+      "step": 39590
+    },
+    {
+      "epoch": 0.8798435833629576,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.4004444444444447e-06,
+      "loss": 0.3774,
+      "step": 39600
+    },
+    {
+      "epoch": 0.8800657660860292,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.3960000000000004e-06,
+      "loss": 0.4064,
+      "step": 39610
+    },
+    {
+      "epoch": 0.8802879488091006,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.3915555555555557e-06,
+      "loss": 0.3968,
+      "step": 39620
+    },
+    {
+      "epoch": 0.8805101315321721,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.3871111111111114e-06,
+      "loss": 0.366,
+      "step": 39630
+    },
+    {
+      "epoch": 0.8807323142552436,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.3826666666666667e-06,
+      "loss": 0.3739,
+      "step": 39640
+    },
+    {
+      "epoch": 0.880954496978315,
+      "grad_norm": 2.875,
+      "learning_rate": 2.3782222222222224e-06,
+      "loss": 0.3934,
+      "step": 39650
+    },
+    {
+      "epoch": 0.8811766797013865,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.373777777777778e-06,
+      "loss": 0.3518,
+      "step": 39660
+    },
+    {
+      "epoch": 0.8813988624244579,
+      "grad_norm": 2.15625,
+      "learning_rate": 2.3693333333333333e-06,
+      "loss": 0.3619,
+      "step": 39670
+    },
+    {
+      "epoch": 0.8816210451475294,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.364888888888889e-06,
+      "loss": 0.3641,
+      "step": 39680
+    },
+    {
+      "epoch": 0.8818432278706008,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.3604444444444447e-06,
+      "loss": 0.3966,
+      "step": 39690
+    },
+    {
+      "epoch": 0.8820654105936723,
+      "grad_norm": 2.625,
+      "learning_rate": 2.3560000000000004e-06,
+      "loss": 0.3988,
+      "step": 39700
+    },
+    {
+      "epoch": 0.8822875933167437,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.3515555555555557e-06,
+      "loss": 0.3729,
+      "step": 39710
+    },
+    {
+      "epoch": 0.8825097760398152,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.3471111111111114e-06,
+      "loss": 0.3538,
+      "step": 39720
+    },
+    {
+      "epoch": 0.8827319587628866,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.342666666666667e-06,
+      "loss": 0.362,
+      "step": 39730
+    },
+    {
+      "epoch": 0.8829541414859581,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.3382222222222223e-06,
+      "loss": 0.4084,
+      "step": 39740
+    },
+    {
+      "epoch": 0.8831763242090295,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.333777777777778e-06,
+      "loss": 0.3524,
+      "step": 39750
+    },
+    {
+      "epoch": 0.883398506932101,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.3293333333333333e-06,
+      "loss": 0.3715,
+      "step": 39760
+    },
+    {
+      "epoch": 0.8836206896551724,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.324888888888889e-06,
+      "loss": 0.3663,
+      "step": 39770
+    },
+    {
+      "epoch": 0.8838428723782439,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.3204444444444447e-06,
+      "loss": 0.3583,
+      "step": 39780
+    },
+    {
+      "epoch": 0.8840650551013153,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.3160000000000004e-06,
+      "loss": 0.3297,
+      "step": 39790
+    },
+    {
+      "epoch": 0.8842872378243868,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.3115555555555556e-06,
+      "loss": 0.3392,
+      "step": 39800
+    },
+    {
+      "epoch": 0.8845094205474582,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.3071111111111113e-06,
+      "loss": 0.3786,
+      "step": 39810
+    },
+    {
+      "epoch": 0.8847316032705297,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.302666666666667e-06,
+      "loss": 0.4139,
+      "step": 39820
+    },
+    {
+      "epoch": 0.8849537859936011,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.2982222222222227e-06,
+      "loss": 0.4058,
+      "step": 39830
+    },
+    {
+      "epoch": 0.8851759687166726,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.293777777777778e-06,
+      "loss": 0.3488,
+      "step": 39840
+    },
+    {
+      "epoch": 0.8853981514397441,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.2893333333333332e-06,
+      "loss": 0.3846,
+      "step": 39850
+    },
+    {
+      "epoch": 0.8856203341628155,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.284888888888889e-06,
+      "loss": 0.3565,
+      "step": 39860
+    },
+    {
+      "epoch": 0.885842516885887,
+      "grad_norm": 2.234375,
+      "learning_rate": 2.2804444444444446e-06,
+      "loss": 0.3832,
+      "step": 39870
+    },
+    {
+      "epoch": 0.8860646996089584,
+      "grad_norm": 2.0,
+      "learning_rate": 2.2760000000000003e-06,
+      "loss": 0.389,
+      "step": 39880
+    },
+    {
+      "epoch": 0.8862868823320299,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.2715555555555556e-06,
+      "loss": 0.3651,
+      "step": 39890
+    },
+    {
+      "epoch": 0.8865090650551013,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.2671111111111113e-06,
+      "loss": 0.3626,
+      "step": 39900
+    },
+    {
+      "epoch": 0.8867312477781728,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.262666666666667e-06,
+      "loss": 0.4105,
+      "step": 39910
+    },
+    {
+      "epoch": 0.8869534305012442,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.2582222222222227e-06,
+      "loss": 0.4118,
+      "step": 39920
+    },
+    {
+      "epoch": 0.8871756132243157,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.253777777777778e-06,
+      "loss": 0.3644,
+      "step": 39930
+    },
+    {
+      "epoch": 0.8873977959473871,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.2493333333333336e-06,
+      "loss": 0.4027,
+      "step": 39940
+    },
+    {
+      "epoch": 0.8876199786704586,
+      "grad_norm": 2.5,
+      "learning_rate": 2.244888888888889e-06,
+      "loss": 0.3823,
+      "step": 39950
+    },
+    {
+      "epoch": 0.88784216139353,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.2404444444444446e-06,
+      "loss": 0.3794,
+      "step": 39960
+    },
+    {
+      "epoch": 0.8880643441166015,
+      "grad_norm": 2.25,
+      "learning_rate": 2.2360000000000003e-06,
+      "loss": 0.4008,
+      "step": 39970
+    },
+    {
+      "epoch": 0.8882865268396729,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.2315555555555555e-06,
+      "loss": 0.3852,
+      "step": 39980
+    },
+    {
+      "epoch": 0.8885087095627444,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.2271111111111112e-06,
+      "loss": 0.4021,
+      "step": 39990
+    },
+    {
+      "epoch": 0.8887308922858158,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.222666666666667e-06,
+      "loss": 0.3567,
+      "step": 40000
+    },
+    {
+      "epoch": 0.8889530750088873,
+      "grad_norm": 2.375,
+      "learning_rate": 2.2182222222222226e-06,
+      "loss": 0.4155,
+      "step": 40010
+    },
+    {
+      "epoch": 0.8891752577319587,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.213777777777778e-06,
+      "loss": 0.4003,
+      "step": 40020
+    },
+    {
+      "epoch": 0.8893974404550302,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.2093333333333336e-06,
+      "loss": 0.3741,
+      "step": 40030
+    },
+    {
+      "epoch": 0.8896196231781017,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.2048888888888893e-06,
+      "loss": 0.3505,
+      "step": 40040
+    },
+    {
+      "epoch": 0.8898418059011731,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.2004444444444445e-06,
+      "loss": 0.384,
+      "step": 40050
+    },
+    {
+      "epoch": 0.8900639886242446,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.1960000000000002e-06,
+      "loss": 0.3613,
+      "step": 40060
+    },
+    {
+      "epoch": 0.890286171347316,
+      "grad_norm": 2.375,
+      "learning_rate": 2.1915555555555555e-06,
+      "loss": 0.369,
+      "step": 40070
+    },
+    {
+      "epoch": 0.8905083540703875,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.187111111111111e-06,
+      "loss": 0.3752,
+      "step": 40080
+    },
+    {
+      "epoch": 0.8907305367934589,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.182666666666667e-06,
+      "loss": 0.3766,
+      "step": 40090
+    },
+    {
+      "epoch": 0.8909527195165304,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.1782222222222226e-06,
+      "loss": 0.3497,
+      "step": 40100
+    },
+    {
+      "epoch": 0.8911749022396018,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.173777777777778e-06,
+      "loss": 0.3781,
+      "step": 40110
+    },
+    {
+      "epoch": 0.8913970849626733,
+      "grad_norm": 2.75,
+      "learning_rate": 2.1693333333333335e-06,
+      "loss": 0.3893,
+      "step": 40120
+    },
+    {
+      "epoch": 0.8916192676857447,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.1648888888888892e-06,
+      "loss": 0.3513,
+      "step": 40130
+    },
+    {
+      "epoch": 0.8918414504088162,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.160444444444445e-06,
+      "loss": 0.3694,
+      "step": 40140
+    },
+    {
+      "epoch": 0.8920636331318876,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.156e-06,
+      "loss": 0.4148,
+      "step": 40150
+    },
+    {
+      "epoch": 0.8922858158549591,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.1515555555555555e-06,
+      "loss": 0.3905,
+      "step": 40160
+    },
+    {
+      "epoch": 0.8925079985780305,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.147111111111111e-06,
+      "loss": 0.4334,
+      "step": 40170
+    },
+    {
+      "epoch": 0.892730181301102,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.142666666666667e-06,
+      "loss": 0.4006,
+      "step": 40180
+    },
+    {
+      "epoch": 0.8929523640241734,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.1382222222222225e-06,
+      "loss": 0.3621,
+      "step": 40190
+    },
+    {
+      "epoch": 0.893174546747245,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.133777777777778e-06,
+      "loss": 0.4158,
+      "step": 40200
+    },
+    {
+      "epoch": 0.8933967294703163,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.1293333333333335e-06,
+      "loss": 0.3865,
+      "step": 40210
+    },
+    {
+      "epoch": 0.8936189121933878,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.124888888888889e-06,
+      "loss": 0.3357,
+      "step": 40220
+    },
+    {
+      "epoch": 0.8938410949164592,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.120444444444445e-06,
+      "loss": 0.3716,
+      "step": 40230
+    },
+    {
+      "epoch": 0.8940632776395308,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.116e-06,
+      "loss": 0.3947,
+      "step": 40240
+    },
+    {
+      "epoch": 0.8942854603626023,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.111555555555556e-06,
+      "loss": 0.3916,
+      "step": 40250
+    },
+    {
+      "epoch": 0.8945076430856737,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.107111111111111e-06,
+      "loss": 0.3989,
+      "step": 40260
+    },
+    {
+      "epoch": 0.8947298258087452,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.102666666666667e-06,
+      "loss": 0.38,
+      "step": 40270
+    },
+    {
+      "epoch": 0.8949520085318166,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.0982222222222225e-06,
+      "loss": 0.3718,
+      "step": 40280
+    },
+    {
+      "epoch": 0.8951741912548881,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.0937777777777778e-06,
+      "loss": 0.4015,
+      "step": 40290
+    },
+    {
+      "epoch": 0.8953963739779595,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.0893333333333335e-06,
+      "loss": 0.3837,
+      "step": 40300
+    },
+    {
+      "epoch": 0.895618556701031,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.084888888888889e-06,
+      "loss": 0.3657,
+      "step": 40310
+    },
+    {
+      "epoch": 0.8958407394241024,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.080444444444445e-06,
+      "loss": 0.3588,
+      "step": 40320
+    },
+    {
+      "epoch": 0.8960629221471739,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.076e-06,
+      "loss": 0.3911,
+      "step": 40330
+    },
+    {
+      "epoch": 0.8962851048702453,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.071555555555556e-06,
+      "loss": 0.4072,
+      "step": 40340
+    },
+    {
+      "epoch": 0.8965072875933168,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.0671111111111115e-06,
+      "loss": 0.3658,
+      "step": 40350
+    },
+    {
+      "epoch": 0.8967294703163882,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.0626666666666668e-06,
+      "loss": 0.4061,
+      "step": 40360
+    },
+    {
+      "epoch": 0.8969516530394597,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.0582222222222225e-06,
+      "loss": 0.3724,
+      "step": 40370
+    },
+    {
+      "epoch": 0.8971738357625311,
+      "grad_norm": 2.375,
+      "learning_rate": 2.0537777777777777e-06,
+      "loss": 0.3883,
+      "step": 40380
+    },
+    {
+      "epoch": 0.8973960184856026,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.0493333333333334e-06,
+      "loss": 0.3974,
+      "step": 40390
+    },
+    {
+      "epoch": 0.897618201208674,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.044888888888889e-06,
+      "loss": 0.3535,
+      "step": 40400
+    },
+    {
+      "epoch": 0.8978403839317455,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.040444444444445e-06,
+      "loss": 0.3663,
+      "step": 40410
+    },
+    {
+      "epoch": 0.8980625666548169,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.036e-06,
+      "loss": 0.3851,
+      "step": 40420
+    },
+    {
+      "epoch": 0.8982847493778884,
+      "grad_norm": 2.75,
+      "learning_rate": 2.0315555555555558e-06,
+      "loss": 0.3804,
+      "step": 40430
+    },
+    {
+      "epoch": 0.8985069321009598,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.0271111111111115e-06,
+      "loss": 0.3553,
+      "step": 40440
+    },
+    {
+      "epoch": 0.8987291148240313,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.022666666666667e-06,
+      "loss": 0.3897,
+      "step": 40450
+    },
+    {
+      "epoch": 0.8989512975471028,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.0182222222222224e-06,
+      "loss": 0.3835,
+      "step": 40460
+    },
+    {
+      "epoch": 0.8991734802701742,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.0137777777777777e-06,
+      "loss": 0.3883,
+      "step": 40470
+    },
+    {
+      "epoch": 0.8993956629932457,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.0093333333333334e-06,
+      "loss": 0.3752,
+      "step": 40480
+    },
+    {
+      "epoch": 0.8996178457163171,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.004888888888889e-06,
+      "loss": 0.3865,
+      "step": 40490
+    },
+    {
+      "epoch": 0.8998400284393886,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.0004444444444448e-06,
+      "loss": 0.373,
+      "step": 40500
+    },
+    {
+      "epoch": 0.90006221116246,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.996e-06,
+      "loss": 0.3439,
+      "step": 40510
+    },
+    {
+      "epoch": 0.9002843938855315,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.9915555555555557e-06,
+      "loss": 0.3622,
+      "step": 40520
+    },
+    {
+      "epoch": 0.9005065766086029,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.9871111111111114e-06,
+      "loss": 0.3596,
+      "step": 40530
+    },
+    {
+      "epoch": 0.9007287593316744,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.982666666666667e-06,
+      "loss": 0.3888,
+      "step": 40540
+    },
+    {
+      "epoch": 0.9009509420547458,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9782222222222224e-06,
+      "loss": 0.3853,
+      "step": 40550
+    },
+    {
+      "epoch": 0.9011731247778173,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.973777777777778e-06,
+      "loss": 0.3725,
+      "step": 40560
+    },
+    {
+      "epoch": 0.9013953075008887,
+      "grad_norm": 2.375,
+      "learning_rate": 1.9693333333333333e-06,
+      "loss": 0.3684,
+      "step": 40570
+    },
+    {
+      "epoch": 0.9016174902239602,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.964888888888889e-06,
+      "loss": 0.4225,
+      "step": 40580
+    },
+    {
+      "epoch": 0.9018396729470316,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.9604444444444447e-06,
+      "loss": 0.3596,
+      "step": 40590
+    },
+    {
+      "epoch": 0.9020618556701031,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.956e-06,
+      "loss": 0.3713,
+      "step": 40600
+    },
+    {
+      "epoch": 0.9022840383931745,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.9515555555555557e-06,
+      "loss": 0.3995,
+      "step": 40610
+    },
+    {
+      "epoch": 0.902506221116246,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9471111111111114e-06,
+      "loss": 0.3526,
+      "step": 40620
+    },
+    {
+      "epoch": 0.9027284038393174,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.942666666666667e-06,
+      "loss": 0.3605,
+      "step": 40630
+    },
+    {
+      "epoch": 0.9029505865623889,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.9382222222222223e-06,
+      "loss": 0.3665,
+      "step": 40640
+    },
+    {
+      "epoch": 0.9031727692854603,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.933777777777778e-06,
+      "loss": 0.3558,
+      "step": 40650
+    },
+    {
+      "epoch": 0.9033949520085318,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9293333333333337e-06,
+      "loss": 0.3958,
+      "step": 40660
+    },
+    {
+      "epoch": 0.9036171347316033,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.924888888888889e-06,
+      "loss": 0.38,
+      "step": 40670
+    },
+    {
+      "epoch": 0.9038393174546747,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.9204444444444447e-06,
+      "loss": 0.3761,
+      "step": 40680
+    },
+    {
+      "epoch": 0.9040615001777462,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.916e-06,
+      "loss": 0.3434,
+      "step": 40690
+    },
+    {
+      "epoch": 0.9042836829008176,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.9115555555555556e-06,
+      "loss": 0.4032,
+      "step": 40700
+    },
+    {
+      "epoch": 0.9045058656238891,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.9071111111111113e-06,
+      "loss": 0.4014,
+      "step": 40710
+    },
+    {
+      "epoch": 0.9047280483469605,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.9026666666666668e-06,
+      "loss": 0.3608,
+      "step": 40720
+    },
+    {
+      "epoch": 0.904950231070032,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.8982222222222225e-06,
+      "loss": 0.3829,
+      "step": 40730
+    },
+    {
+      "epoch": 0.9051724137931034,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.893777777777778e-06,
+      "loss": 0.3839,
+      "step": 40740
+    },
+    {
+      "epoch": 0.9053945965161749,
+      "grad_norm": 2.625,
+      "learning_rate": 1.8893333333333335e-06,
+      "loss": 0.4213,
+      "step": 40750
+    },
+    {
+      "epoch": 0.9056167792392463,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.8848888888888892e-06,
+      "loss": 0.411,
+      "step": 40760
+    },
+    {
+      "epoch": 0.9058389619623178,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.8804444444444444e-06,
+      "loss": 0.3776,
+      "step": 40770
+    },
+    {
+      "epoch": 0.9060611446853892,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.8760000000000001e-06,
+      "loss": 0.3827,
+      "step": 40780
+    },
+    {
+      "epoch": 0.9062833274084607,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.8715555555555556e-06,
+      "loss": 0.329,
+      "step": 40790
+    },
+    {
+      "epoch": 0.9065055101315321,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.8671111111111113e-06,
+      "loss": 0.3922,
+      "step": 40800
+    },
+    {
+      "epoch": 0.9067276928546036,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.8626666666666668e-06,
+      "loss": 0.3798,
+      "step": 40810
+    },
+    {
+      "epoch": 0.906949875577675,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.8582222222222225e-06,
+      "loss": 0.3537,
+      "step": 40820
+    },
+    {
+      "epoch": 0.9071720583007465,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.853777777777778e-06,
+      "loss": 0.3708,
+      "step": 40830
+    },
+    {
+      "epoch": 0.9073942410238179,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.8493333333333336e-06,
+      "loss": 0.3599,
+      "step": 40840
+    },
+    {
+      "epoch": 0.9076164237468894,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.8448888888888891e-06,
+      "loss": 0.3792,
+      "step": 40850
+    },
+    {
+      "epoch": 0.9078386064699608,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.8404444444444446e-06,
+      "loss": 0.4045,
+      "step": 40860
+    },
+    {
+      "epoch": 0.9080607891930323,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.8360000000000003e-06,
+      "loss": 0.3799,
+      "step": 40870
+    },
+    {
+      "epoch": 0.9082829719161039,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8315555555555556e-06,
+      "loss": 0.3598,
+      "step": 40880
+    },
+    {
+      "epoch": 0.9085051546391752,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.8271111111111113e-06,
+      "loss": 0.3637,
+      "step": 40890
+    },
+    {
+      "epoch": 0.9087273373622468,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.8226666666666667e-06,
+      "loss": 0.3867,
+      "step": 40900
+    },
+    {
+      "epoch": 0.9089495200853182,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8182222222222224e-06,
+      "loss": 0.3726,
+      "step": 40910
+    },
+    {
+      "epoch": 0.9091717028083897,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.813777777777778e-06,
+      "loss": 0.3596,
+      "step": 40920
+    },
+    {
+      "epoch": 0.909393885531461,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.8093333333333336e-06,
+      "loss": 0.368,
+      "step": 40930
+    },
+    {
+      "epoch": 0.9096160682545326,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.804888888888889e-06,
+      "loss": 0.3872,
+      "step": 40940
+    },
+    {
+      "epoch": 0.909838250977604,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.8004444444444446e-06,
+      "loss": 0.4331,
+      "step": 40950
+    },
+    {
+      "epoch": 0.9100604337006755,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7960000000000003e-06,
+      "loss": 0.396,
+      "step": 40960
+    },
+    {
+      "epoch": 0.9102826164237469,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.7915555555555557e-06,
+      "loss": 0.4018,
+      "step": 40970
+    },
+    {
+      "epoch": 0.9105047991468184,
+      "grad_norm": 3.0,
+      "learning_rate": 1.7871111111111112e-06,
+      "loss": 0.3579,
+      "step": 40980
+    },
+    {
+      "epoch": 0.9107269818698898,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.7826666666666667e-06,
+      "loss": 0.3482,
+      "step": 40990
+    },
+    {
+      "epoch": 0.9109491645929613,
+      "grad_norm": 2.625,
+      "learning_rate": 1.7782222222222224e-06,
+      "loss": 0.3721,
+      "step": 41000
+    },
+    {
+      "epoch": 0.9111713473160327,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.7737777777777779e-06,
+      "loss": 0.3953,
+      "step": 41010
+    },
+    {
+      "epoch": 0.9113935300391042,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.7693333333333336e-06,
+      "loss": 0.3973,
+      "step": 41020
+    },
+    {
+      "epoch": 0.9116157127621756,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.764888888888889e-06,
+      "loss": 0.4075,
+      "step": 41030
+    },
+    {
+      "epoch": 0.9118378954852471,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.7604444444444445e-06,
+      "loss": 0.3856,
+      "step": 41040
+    },
+    {
+      "epoch": 0.9120600782083185,
+      "grad_norm": 2.75,
+      "learning_rate": 1.7560000000000002e-06,
+      "loss": 0.4153,
+      "step": 41050
+    },
+    {
+      "epoch": 0.91228226093139,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.7515555555555557e-06,
+      "loss": 0.3414,
+      "step": 41060
+    },
+    {
+      "epoch": 0.9125044436544615,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.7471111111111114e-06,
+      "loss": 0.3496,
+      "step": 41070
+    },
+    {
+      "epoch": 0.9127266263775329,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.7426666666666667e-06,
+      "loss": 0.3977,
+      "step": 41080
+    },
+    {
+      "epoch": 0.9129488091006044,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.7382222222222223e-06,
+      "loss": 0.3732,
+      "step": 41090
+    },
+    {
+      "epoch": 0.9131709918236758,
+      "grad_norm": 2.5,
+      "learning_rate": 1.7337777777777778e-06,
+      "loss": 0.4004,
+      "step": 41100
+    },
+    {
+      "epoch": 0.9133931745467473,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.7293333333333335e-06,
+      "loss": 0.3833,
+      "step": 41110
+    },
+    {
+      "epoch": 0.9136153572698187,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.724888888888889e-06,
+      "loss": 0.3491,
+      "step": 41120
+    },
+    {
+      "epoch": 0.9138375399928902,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.7204444444444445e-06,
+      "loss": 0.3881,
+      "step": 41130
+    },
+    {
+      "epoch": 0.9140597227159616,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.7160000000000002e-06,
+      "loss": 0.3619,
+      "step": 41140
+    },
+    {
+      "epoch": 0.9142819054390331,
+      "grad_norm": 2.5,
+      "learning_rate": 1.7115555555555557e-06,
+      "loss": 0.3969,
+      "step": 41150
+    },
+    {
+      "epoch": 0.9145040881621045,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.7071111111111113e-06,
+      "loss": 0.3802,
+      "step": 41160
+    },
+    {
+      "epoch": 0.914726270885176,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.7026666666666668e-06,
+      "loss": 0.4202,
+      "step": 41170
+    },
+    {
+      "epoch": 0.9149484536082474,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.6982222222222225e-06,
+      "loss": 0.354,
+      "step": 41180
+    },
+    {
+      "epoch": 0.9151706363313189,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.6937777777777778e-06,
+      "loss": 0.371,
+      "step": 41190
+    },
+    {
+      "epoch": 0.9153928190543903,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.6893333333333335e-06,
+      "loss": 0.3855,
+      "step": 41200
+    },
+    {
+      "epoch": 0.9156150017774618,
+      "grad_norm": 2.625,
+      "learning_rate": 1.684888888888889e-06,
+      "loss": 0.4043,
+      "step": 41210
+    },
+    {
+      "epoch": 0.9158371845005332,
+      "grad_norm": 2.875,
+      "learning_rate": 1.6804444444444444e-06,
+      "loss": 0.3803,
+      "step": 41220
+    },
+    {
+      "epoch": 0.9160593672236047,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.6760000000000001e-06,
+      "loss": 0.4139,
+      "step": 41230
+    },
+    {
+      "epoch": 0.9162815499466761,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.6715555555555556e-06,
+      "loss": 0.3653,
+      "step": 41240
+    },
+    {
+      "epoch": 0.9165037326697476,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.6671111111111113e-06,
+      "loss": 0.3962,
+      "step": 41250
+    },
+    {
+      "epoch": 0.916725915392819,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.6626666666666668e-06,
+      "loss": 0.3863,
+      "step": 41260
+    },
+    {
+      "epoch": 0.9169480981158905,
+      "grad_norm": 2.75,
+      "learning_rate": 1.6582222222222225e-06,
+      "loss": 0.3965,
+      "step": 41270
+    },
+    {
+      "epoch": 0.917170280838962,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.653777777777778e-06,
+      "loss": 0.3669,
+      "step": 41280
+    },
+    {
+      "epoch": 0.9173924635620334,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.6493333333333334e-06,
+      "loss": 0.4242,
+      "step": 41290
+    },
+    {
+      "epoch": 0.9176146462851049,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.644888888888889e-06,
+      "loss": 0.3731,
+      "step": 41300
+    },
+    {
+      "epoch": 0.9178368290081763,
+      "grad_norm": 2.5,
+      "learning_rate": 1.6404444444444446e-06,
+      "loss": 0.3651,
+      "step": 41310
+    },
+    {
+      "epoch": 0.9180590117312478,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.636e-06,
+      "loss": 0.3503,
+      "step": 41320
+    },
+    {
+      "epoch": 0.9182811944543192,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.6315555555555556e-06,
+      "loss": 0.3702,
+      "step": 41330
+    },
+    {
+      "epoch": 0.9185033771773907,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.6271111111111113e-06,
+      "loss": 0.3517,
+      "step": 41340
+    },
+    {
+      "epoch": 0.9187255599004621,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.6226666666666667e-06,
+      "loss": 0.3837,
+      "step": 41350
+    },
+    {
+      "epoch": 0.9189477426235336,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.6182222222222224e-06,
+      "loss": 0.3561,
+      "step": 41360
+    },
+    {
+      "epoch": 0.919169925346605,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.613777777777778e-06,
+      "loss": 0.3496,
+      "step": 41370
+    },
+    {
+      "epoch": 0.9193921080696765,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.6093333333333336e-06,
+      "loss": 0.4501,
+      "step": 41380
+    },
+    {
+      "epoch": 0.9196142907927479,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.6048888888888889e-06,
+      "loss": 0.3776,
+      "step": 41390
+    },
+    {
+      "epoch": 0.9198364735158194,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.6004444444444446e-06,
+      "loss": 0.3998,
+      "step": 41400
+    },
+    {
+      "epoch": 0.9200586562388908,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.596e-06,
+      "loss": 0.3667,
+      "step": 41410
+    },
+    {
+      "epoch": 0.9202808389619623,
+      "grad_norm": 2.5,
+      "learning_rate": 1.5915555555555555e-06,
+      "loss": 0.4079,
+      "step": 41420
+    },
+    {
+      "epoch": 0.9205030216850337,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.5871111111111112e-06,
+      "loss": 0.3854,
+      "step": 41430
+    },
+    {
+      "epoch": 0.9207252044081052,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.5826666666666667e-06,
+      "loss": 0.383,
+      "step": 41440
+    },
+    {
+      "epoch": 0.9209473871311766,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.5782222222222224e-06,
+      "loss": 0.3772,
+      "step": 41450
+    },
+    {
+      "epoch": 0.9211695698542481,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.5737777777777779e-06,
+      "loss": 0.3157,
+      "step": 41460
+    },
+    {
+      "epoch": 0.9213917525773195,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.5693333333333336e-06,
+      "loss": 0.3859,
+      "step": 41470
+    },
+    {
+      "epoch": 0.921613935300391,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.564888888888889e-06,
+      "loss": 0.3857,
+      "step": 41480
+    },
+    {
+      "epoch": 0.9218361180234625,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.5604444444444447e-06,
+      "loss": 0.3758,
+      "step": 41490
+    },
+    {
+      "epoch": 0.9220583007465339,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.556e-06,
+      "loss": 0.361,
+      "step": 41500
+    },
+    {
+      "epoch": 0.9222804834696054,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.5515555555555555e-06,
+      "loss": 0.4005,
+      "step": 41510
+    },
+    {
+      "epoch": 0.9225026661926768,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5471111111111112e-06,
+      "loss": 0.3538,
+      "step": 41520
+    },
+    {
+      "epoch": 0.9227248489157484,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.5426666666666667e-06,
+      "loss": 0.4144,
+      "step": 41530
+    },
+    {
+      "epoch": 0.9229470316388197,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.5382222222222224e-06,
+      "loss": 0.3523,
+      "step": 41540
+    },
+    {
+      "epoch": 0.9231692143618913,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.5337777777777778e-06,
+      "loss": 0.3618,
+      "step": 41550
+    },
+    {
+      "epoch": 0.9233913970849627,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.5293333333333335e-06,
+      "loss": 0.3997,
+      "step": 41560
+    },
+    {
+      "epoch": 0.9236135798080342,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.524888888888889e-06,
+      "loss": 0.366,
+      "step": 41570
+    },
+    {
+      "epoch": 0.9238357625311056,
+      "grad_norm": 2.625,
+      "learning_rate": 1.5204444444444447e-06,
+      "loss": 0.4361,
+      "step": 41580
+    },
+    {
+      "epoch": 0.9240579452541771,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.5160000000000002e-06,
+      "loss": 0.392,
+      "step": 41590
+    },
+    {
+      "epoch": 0.9242801279772485,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.5115555555555554e-06,
+      "loss": 0.4082,
+      "step": 41600
+    },
+    {
+      "epoch": 0.92450231070032,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.5071111111111111e-06,
+      "loss": 0.4094,
+      "step": 41610
+    },
+    {
+      "epoch": 0.9247244934233914,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.5026666666666666e-06,
+      "loss": 0.3583,
+      "step": 41620
+    },
+    {
+      "epoch": 0.9249466761464629,
+      "grad_norm": 2.875,
+      "learning_rate": 1.4982222222222223e-06,
+      "loss": 0.3857,
+      "step": 41630
+    },
+    {
+      "epoch": 0.9251688588695343,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.4937777777777778e-06,
+      "loss": 0.3817,
+      "step": 41640
+    },
+    {
+      "epoch": 0.9253910415926058,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4893333333333335e-06,
+      "loss": 0.3696,
+      "step": 41650
+    },
+    {
+      "epoch": 0.9256132243156772,
+      "grad_norm": 3.0,
+      "learning_rate": 1.484888888888889e-06,
+      "loss": 0.3652,
+      "step": 41660
+    },
+    {
+      "epoch": 0.9258354070387487,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.4804444444444447e-06,
+      "loss": 0.4169,
+      "step": 41670
+    },
+    {
+      "epoch": 0.9260575897618201,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.4760000000000001e-06,
+      "loss": 0.369,
+      "step": 41680
+    },
+    {
+      "epoch": 0.9262797724848916,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.4715555555555558e-06,
+      "loss": 0.3739,
+      "step": 41690
+    },
+    {
+      "epoch": 0.9265019552079631,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.467111111111111e-06,
+      "loss": 0.3888,
+      "step": 41700
+    },
+    {
+      "epoch": 0.9267241379310345,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.4626666666666666e-06,
+      "loss": 0.3685,
+      "step": 41710
+    },
+    {
+      "epoch": 0.926946320654106,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.4582222222222223e-06,
+      "loss": 0.4017,
+      "step": 41720
+    },
+    {
+      "epoch": 0.9271685033771774,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.4537777777777778e-06,
+      "loss": 0.358,
+      "step": 41730
+    },
+    {
+      "epoch": 0.9273906861002489,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4493333333333334e-06,
+      "loss": 0.3853,
+      "step": 41740
+    },
+    {
+      "epoch": 0.9276128688233203,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.444888888888889e-06,
+      "loss": 0.3745,
+      "step": 41750
+    },
+    {
+      "epoch": 0.9278350515463918,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.4404444444444446e-06,
+      "loss": 0.3652,
+      "step": 41760
+    },
+    {
+      "epoch": 0.9280572342694632,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.436e-06,
+      "loss": 0.3897,
+      "step": 41770
+    },
+    {
+      "epoch": 0.9282794169925347,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.4315555555555558e-06,
+      "loss": 0.3989,
+      "step": 41780
+    },
+    {
+      "epoch": 0.9285015997156061,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.4271111111111113e-06,
+      "loss": 0.3566,
+      "step": 41790
+    },
+    {
+      "epoch": 0.9287237824386776,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.422666666666667e-06,
+      "loss": 0.4067,
+      "step": 41800
+    },
+    {
+      "epoch": 0.928945965161749,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.4182222222222222e-06,
+      "loss": 0.3799,
+      "step": 41810
+    },
+    {
+      "epoch": 0.9291681478848205,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.4137777777777777e-06,
+      "loss": 0.3472,
+      "step": 41820
+    },
+    {
+      "epoch": 0.9293903306078919,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.4093333333333334e-06,
+      "loss": 0.3585,
+      "step": 41830
+    },
+    {
+      "epoch": 0.9296125133309634,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.4048888888888889e-06,
+      "loss": 0.3858,
+      "step": 41840
+    },
+    {
+      "epoch": 0.9298346960540348,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.4004444444444446e-06,
+      "loss": 0.3849,
+      "step": 41850
+    },
+    {
+      "epoch": 0.9300568787771063,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.396e-06,
+      "loss": 0.3503,
+      "step": 41860
+    },
+    {
+      "epoch": 0.9302790615001777,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3915555555555558e-06,
+      "loss": 0.3858,
+      "step": 41870
+    },
+    {
+      "epoch": 0.9305012442232492,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.3871111111111112e-06,
+      "loss": 0.3878,
+      "step": 41880
+    },
+    {
+      "epoch": 0.9307234269463207,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.382666666666667e-06,
+      "loss": 0.3741,
+      "step": 41890
+    },
+    {
+      "epoch": 0.9309456096693921,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.3782222222222224e-06,
+      "loss": 0.3998,
+      "step": 41900
+    },
+    {
+      "epoch": 0.9311677923924636,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.3737777777777777e-06,
+      "loss": 0.3897,
+      "step": 41910
+    },
+    {
+      "epoch": 0.931389975115535,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.3693333333333334e-06,
+      "loss": 0.3934,
+      "step": 41920
+    },
+    {
+      "epoch": 0.9316121578386065,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.3648888888888888e-06,
+      "loss": 0.3852,
+      "step": 41930
+    },
+    {
+      "epoch": 0.9318343405616779,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.3604444444444445e-06,
+      "loss": 0.4274,
+      "step": 41940
+    },
+    {
+      "epoch": 0.9320565232847494,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.356e-06,
+      "loss": 0.3154,
+      "step": 41950
+    },
+    {
+      "epoch": 0.9322787060078208,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.3515555555555557e-06,
+      "loss": 0.4125,
+      "step": 41960
+    },
+    {
+      "epoch": 0.9325008887308923,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.3471111111111112e-06,
+      "loss": 0.3975,
+      "step": 41970
+    },
+    {
+      "epoch": 0.9327230714539637,
+      "grad_norm": 2.75,
+      "learning_rate": 1.3426666666666669e-06,
+      "loss": 0.3655,
+      "step": 41980
+    },
+    {
+      "epoch": 0.9329452541770352,
+      "grad_norm": 2.625,
+      "learning_rate": 1.3382222222222224e-06,
+      "loss": 0.4096,
+      "step": 41990
+    },
+    {
+      "epoch": 0.9331674369001066,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.333777777777778e-06,
+      "loss": 0.3845,
+      "step": 42000
+    },
+    {
+      "epoch": 0.9333896196231781,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.3293333333333333e-06,
+      "loss": 0.3792,
+      "step": 42010
+    },
+    {
+      "epoch": 0.9336118023462495,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.3248888888888888e-06,
+      "loss": 0.3942,
+      "step": 42020
+    },
+    {
+      "epoch": 0.933833985069321,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.3204444444444445e-06,
+      "loss": 0.39,
+      "step": 42030
+    },
+    {
+      "epoch": 0.9340561677923924,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.316e-06,
+      "loss": 0.3978,
+      "step": 42040
+    },
+    {
+      "epoch": 0.9342783505154639,
+      "grad_norm": 2.5,
+      "learning_rate": 1.3115555555555557e-06,
+      "loss": 0.3792,
+      "step": 42050
+    },
+    {
+      "epoch": 0.9345005332385353,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.3071111111111112e-06,
+      "loss": 0.3787,
+      "step": 42060
+    },
+    {
+      "epoch": 0.9347227159616068,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.3026666666666668e-06,
+      "loss": 0.3818,
+      "step": 42070
+    },
+    {
+      "epoch": 0.9349448986846782,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.2982222222222223e-06,
+      "loss": 0.3714,
+      "step": 42080
+    },
+    {
+      "epoch": 0.9351670814077497,
+      "grad_norm": 2.5,
+      "learning_rate": 1.293777777777778e-06,
+      "loss": 0.3789,
+      "step": 42090
+    },
+    {
+      "epoch": 0.9353892641308212,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.2893333333333335e-06,
+      "loss": 0.358,
+      "step": 42100
+    },
+    {
+      "epoch": 0.9356114468538926,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.2848888888888892e-06,
+      "loss": 0.3733,
+      "step": 42110
+    },
+    {
+      "epoch": 0.9358336295769641,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.2804444444444445e-06,
+      "loss": 0.3934,
+      "step": 42120
+    },
+    {
+      "epoch": 0.9360558123000355,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.276e-06,
+      "loss": 0.3869,
+      "step": 42130
+    },
+    {
+      "epoch": 0.936277995023107,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2715555555555556e-06,
+      "loss": 0.4351,
+      "step": 42140
+    },
+    {
+      "epoch": 0.9365001777461784,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2671111111111111e-06,
+      "loss": 0.3776,
+      "step": 42150
+    },
+    {
+      "epoch": 0.93672236046925,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.2626666666666668e-06,
+      "loss": 0.3752,
+      "step": 42160
+    },
+    {
+      "epoch": 0.9369445431923213,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.2582222222222223e-06,
+      "loss": 0.3555,
+      "step": 42170
+    },
+    {
+      "epoch": 0.9371667259153929,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.253777777777778e-06,
+      "loss": 0.394,
+      "step": 42180
+    },
+    {
+      "epoch": 0.9373889086384642,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.2493333333333335e-06,
+      "loss": 0.4158,
+      "step": 42190
+    },
+    {
+      "epoch": 0.9376110913615358,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.244888888888889e-06,
+      "loss": 0.3608,
+      "step": 42200
+    },
+    {
+      "epoch": 0.9378332740846071,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.2404444444444446e-06,
+      "loss": 0.3904,
+      "step": 42210
+    },
+    {
+      "epoch": 0.9380554568076787,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.2360000000000001e-06,
+      "loss": 0.4005,
+      "step": 42220
+    },
+    {
+      "epoch": 0.93827763953075,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.2315555555555558e-06,
+      "loss": 0.3674,
+      "step": 42230
+    },
+    {
+      "epoch": 0.9384998222538216,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.2271111111111113e-06,
+      "loss": 0.3818,
+      "step": 42240
+    },
+    {
+      "epoch": 0.938722004976893,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.2226666666666668e-06,
+      "loss": 0.3794,
+      "step": 42250
+    },
+    {
+      "epoch": 0.9389441876999645,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.2182222222222222e-06,
+      "loss": 0.411,
+      "step": 42260
+    },
+    {
+      "epoch": 0.9391663704230359,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.213777777777778e-06,
+      "loss": 0.3834,
+      "step": 42270
+    },
+    {
+      "epoch": 0.9393885531461074,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.2093333333333334e-06,
+      "loss": 0.3438,
+      "step": 42280
+    },
+    {
+      "epoch": 0.9396107358691788,
+      "grad_norm": 2.25,
+      "learning_rate": 1.2048888888888891e-06,
+      "loss": 0.4101,
+      "step": 42290
+    },
+    {
+      "epoch": 0.9398329185922503,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.2004444444444446e-06,
+      "loss": 0.4113,
+      "step": 42300
+    },
+    {
+      "epoch": 0.9400551013153218,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.196e-06,
+      "loss": 0.3796,
+      "step": 42310
+    },
+    {
+      "epoch": 0.9402772840383932,
+      "grad_norm": 2.625,
+      "learning_rate": 1.1915555555555558e-06,
+      "loss": 0.3958,
+      "step": 42320
+    },
+    {
+      "epoch": 0.9404994667614647,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.1871111111111112e-06,
+      "loss": 0.4032,
+      "step": 42330
+    },
+    {
+      "epoch": 0.9407216494845361,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.1826666666666667e-06,
+      "loss": 0.3915,
+      "step": 42340
+    },
+    {
+      "epoch": 0.9409438322076076,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1782222222222222e-06,
+      "loss": 0.4121,
+      "step": 42350
+    },
+    {
+      "epoch": 0.941166014930679,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.173777777777778e-06,
+      "loss": 0.3798,
+      "step": 42360
+    },
+    {
+      "epoch": 0.9413881976537505,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.1693333333333334e-06,
+      "loss": 0.3774,
+      "step": 42370
+    },
+    {
+      "epoch": 0.9416103803768219,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.164888888888889e-06,
+      "loss": 0.3736,
+      "step": 42380
+    },
+    {
+      "epoch": 0.9418325630998934,
+      "grad_norm": 2.625,
+      "learning_rate": 1.1604444444444445e-06,
+      "loss": 0.3654,
+      "step": 42390
+    },
+    {
+      "epoch": 0.9420547458229648,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.156e-06,
+      "loss": 0.3836,
+      "step": 42400
+    },
+    {
+      "epoch": 0.9422769285460363,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1515555555555557e-06,
+      "loss": 0.3899,
+      "step": 42410
+    },
+    {
+      "epoch": 0.9424991112691077,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.1471111111111112e-06,
+      "loss": 0.3751,
+      "step": 42420
+    },
+    {
+      "epoch": 0.9427212939921792,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.1426666666666667e-06,
+      "loss": 0.3956,
+      "step": 42430
+    },
+    {
+      "epoch": 0.9429434767152506,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.1382222222222224e-06,
+      "loss": 0.3709,
+      "step": 42440
+    },
+    {
+      "epoch": 0.9431656594383221,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.1337777777777779e-06,
+      "loss": 0.3564,
+      "step": 42450
+    },
+    {
+      "epoch": 0.9433878421613935,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1293333333333333e-06,
+      "loss": 0.3594,
+      "step": 42460
+    },
+    {
+      "epoch": 0.943610024884465,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.124888888888889e-06,
+      "loss": 0.3577,
+      "step": 42470
+    },
+    {
+      "epoch": 0.9438322076075364,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.1204444444444445e-06,
+      "loss": 0.3817,
+      "step": 42480
+    },
+    {
+      "epoch": 0.9440543903306079,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.1160000000000002e-06,
+      "loss": 0.3739,
+      "step": 42490
+    },
+    {
+      "epoch": 0.9442765730536793,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.1115555555555557e-06,
+      "loss": 0.376,
+      "step": 42500
+    },
+    {
+      "epoch": 0.9444987557767508,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.1071111111111112e-06,
+      "loss": 0.4159,
+      "step": 42510
+    },
+    {
+      "epoch": 0.9447209384998223,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.1026666666666666e-06,
+      "loss": 0.3683,
+      "step": 42520
+    },
+    {
+      "epoch": 0.9449431212228937,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.0982222222222223e-06,
+      "loss": 0.3693,
+      "step": 42530
+    },
+    {
+      "epoch": 0.9451653039459652,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.0937777777777778e-06,
+      "loss": 0.4035,
+      "step": 42540
+    },
+    {
+      "epoch": 0.9453874866690366,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0893333333333333e-06,
+      "loss": 0.3564,
+      "step": 42550
+    },
+    {
+      "epoch": 0.9456096693921081,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.084888888888889e-06,
+      "loss": 0.3892,
+      "step": 42560
+    },
+    {
+      "epoch": 0.9458318521151795,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.0804444444444445e-06,
+      "loss": 0.4029,
+      "step": 42570
+    },
+    {
+      "epoch": 0.946054034838251,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.0760000000000002e-06,
+      "loss": 0.3823,
+      "step": 42580
+    },
+    {
+      "epoch": 0.9462762175613224,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.0715555555555556e-06,
+      "loss": 0.3705,
+      "step": 42590
+    },
+    {
+      "epoch": 0.9464984002843939,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.0671111111111113e-06,
+      "loss": 0.3629,
+      "step": 42600
+    },
+    {
+      "epoch": 0.9467205830074653,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0626666666666668e-06,
+      "loss": 0.3675,
+      "step": 42610
+    },
+    {
+      "epoch": 0.9469427657305368,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0582222222222223e-06,
+      "loss": 0.39,
+      "step": 42620
+    },
+    {
+      "epoch": 0.9471649484536082,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.0537777777777778e-06,
+      "loss": 0.3632,
+      "step": 42630
+    },
+    {
+      "epoch": 0.9473871311766797,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.0493333333333335e-06,
+      "loss": 0.3646,
+      "step": 42640
+    },
+    {
+      "epoch": 0.9476093138997511,
+      "grad_norm": 2.375,
+      "learning_rate": 1.044888888888889e-06,
+      "loss": 0.4065,
+      "step": 42650
+    },
+    {
+      "epoch": 0.9478314966228226,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.0404444444444444e-06,
+      "loss": 0.3758,
+      "step": 42660
+    },
+    {
+      "epoch": 0.948053679345894,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.0360000000000001e-06,
+      "loss": 0.3788,
+      "step": 42670
+    },
+    {
+      "epoch": 0.9482758620689655,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.0315555555555556e-06,
+      "loss": 0.3928,
+      "step": 42680
+    },
+    {
+      "epoch": 0.9484980447920369,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.0271111111111113e-06,
+      "loss": 0.3944,
+      "step": 42690
+    },
+    {
+      "epoch": 0.9487202275151084,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.0226666666666668e-06,
+      "loss": 0.3966,
+      "step": 42700
+    },
+    {
+      "epoch": 0.9489424102381799,
+      "grad_norm": 2.125,
+      "learning_rate": 1.0182222222222223e-06,
+      "loss": 0.3592,
+      "step": 42710
+    },
+    {
+      "epoch": 0.9491645929612513,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.0137777777777777e-06,
+      "loss": 0.3896,
+      "step": 42720
+    },
+    {
+      "epoch": 0.9493867756843228,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.0093333333333334e-06,
+      "loss": 0.3657,
+      "step": 42730
+    },
+    {
+      "epoch": 0.9496089584073942,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.004888888888889e-06,
+      "loss": 0.3815,
+      "step": 42740
+    },
+    {
+      "epoch": 0.9498311411304657,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.0004444444444446e-06,
+      "loss": 0.3694,
+      "step": 42750
+    },
+    {
+      "epoch": 0.9500533238535371,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.96e-07,
+      "loss": 0.3747,
+      "step": 42760
+    },
+    {
+      "epoch": 0.9502755065766086,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.915555555555556e-07,
+      "loss": 0.3493,
+      "step": 42770
+    },
+    {
+      "epoch": 0.95049768929968,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.871111111111113e-07,
+      "loss": 0.419,
+      "step": 42780
+    },
+    {
+      "epoch": 0.9507198720227515,
+      "grad_norm": 2.75,
+      "learning_rate": 9.826666666666667e-07,
+      "loss": 0.3941,
+      "step": 42790
+    },
+    {
+      "epoch": 0.9509420547458229,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.782222222222224e-07,
+      "loss": 0.3739,
+      "step": 42800
+    },
+    {
+      "epoch": 0.9511642374688944,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.737777777777777e-07,
+      "loss": 0.3549,
+      "step": 42810
+    },
+    {
+      "epoch": 0.9513864201919658,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.693333333333334e-07,
+      "loss": 0.403,
+      "step": 42820
+    },
+    {
+      "epoch": 0.9516086029150373,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.648888888888889e-07,
+      "loss": 0.3846,
+      "step": 42830
+    },
+    {
+      "epoch": 0.9518307856381087,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.604444444444446e-07,
+      "loss": 0.3849,
+      "step": 42840
+    },
+    {
+      "epoch": 0.9520529683611803,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.56e-07,
+      "loss": 0.3603,
+      "step": 42850
+    },
+    {
+      "epoch": 0.9522751510842516,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.515555555555555e-07,
+      "loss": 0.3794,
+      "step": 42860
+    },
+    {
+      "epoch": 0.9524973338073232,
+      "grad_norm": 2.875,
+      "learning_rate": 9.471111111111111e-07,
+      "loss": 0.4247,
+      "step": 42870
+    },
+    {
+      "epoch": 0.9527195165303946,
+      "grad_norm": 2.375,
+      "learning_rate": 9.426666666666667e-07,
+      "loss": 0.3788,
+      "step": 42880
+    },
+    {
+      "epoch": 0.9529416992534661,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.382222222222223e-07,
+      "loss": 0.3558,
+      "step": 42890
+    },
+    {
+      "epoch": 0.9531638819765375,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.337777777777779e-07,
+      "loss": 0.3491,
+      "step": 42900
+    },
+    {
+      "epoch": 0.953386064699609,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.293333333333334e-07,
+      "loss": 0.388,
+      "step": 42910
+    },
+    {
+      "epoch": 0.9536082474226805,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.248888888888889e-07,
+      "loss": 0.3641,
+      "step": 42920
+    },
+    {
+      "epoch": 0.9538304301457519,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.204444444444445e-07,
+      "loss": 0.3852,
+      "step": 42930
+    },
+    {
+      "epoch": 0.9540526128688234,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.160000000000001e-07,
+      "loss": 0.3894,
+      "step": 42940
+    },
+    {
+      "epoch": 0.9542747955918948,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.115555555555557e-07,
+      "loss": 0.39,
+      "step": 42950
+    },
+    {
+      "epoch": 0.9544969783149663,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.071111111111113e-07,
+      "loss": 0.328,
+      "step": 42960
+    },
+    {
+      "epoch": 0.9547191610380377,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.026666666666666e-07,
+      "loss": 0.3202,
+      "step": 42970
+    },
+    {
+      "epoch": 0.9549413437611092,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.982222222222222e-07,
+      "loss": 0.3493,
+      "step": 42980
+    },
+    {
+      "epoch": 0.9551635264841806,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.937777777777778e-07,
+      "loss": 0.3455,
+      "step": 42990
+    },
+    {
+      "epoch": 0.9553857092072521,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.893333333333334e-07,
+      "loss": 0.3665,
+      "step": 43000
+    },
+    {
+      "epoch": 0.9556078919303235,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.84888888888889e-07,
+      "loss": 0.3596,
+      "step": 43010
+    },
+    {
+      "epoch": 0.955830074653395,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.804444444444445e-07,
+      "loss": 0.4035,
+      "step": 43020
+    },
+    {
+      "epoch": 0.9560522573764664,
+      "grad_norm": 2.875,
+      "learning_rate": 8.760000000000001e-07,
+      "loss": 0.4075,
+      "step": 43030
+    },
+    {
+      "epoch": 0.9562744400995379,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.715555555555556e-07,
+      "loss": 0.3831,
+      "step": 43040
+    },
+    {
+      "epoch": 0.9564966228226093,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.671111111111112e-07,
+      "loss": 0.3732,
+      "step": 43050
+    },
+    {
+      "epoch": 0.9567188055456808,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.626666666666668e-07,
+      "loss": 0.3609,
+      "step": 43060
+    },
+    {
+      "epoch": 0.9569409882687522,
+      "grad_norm": 3.0,
+      "learning_rate": 8.582222222222222e-07,
+      "loss": 0.4022,
+      "step": 43070
+    },
+    {
+      "epoch": 0.9571631709918237,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.537777777777778e-07,
+      "loss": 0.3679,
+      "step": 43080
+    },
+    {
+      "epoch": 0.9573853537148951,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.493333333333334e-07,
+      "loss": 0.3745,
+      "step": 43090
+    },
+    {
+      "epoch": 0.9576075364379666,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.44888888888889e-07,
+      "loss": 0.4094,
+      "step": 43100
+    },
+    {
+      "epoch": 0.957829719161038,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.404444444444445e-07,
+      "loss": 0.3889,
+      "step": 43110
+    },
+    {
+      "epoch": 0.9580519018841095,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.36e-07,
+      "loss": 0.3631,
+      "step": 43120
+    },
+    {
+      "epoch": 0.958274084607181,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.315555555555556e-07,
+      "loss": 0.3851,
+      "step": 43130
+    },
+    {
+      "epoch": 0.9584962673302524,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.271111111111112e-07,
+      "loss": 0.3862,
+      "step": 43140
+    },
+    {
+      "epoch": 0.9587184500533239,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.226666666666668e-07,
+      "loss": 0.3828,
+      "step": 43150
+    },
+    {
+      "epoch": 0.9589406327763953,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.182222222222224e-07,
+      "loss": 0.3638,
+      "step": 43160
+    },
+    {
+      "epoch": 0.9591628154994668,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.137777777777777e-07,
+      "loss": 0.4227,
+      "step": 43170
+    },
+    {
+      "epoch": 0.9593849982225382,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.093333333333333e-07,
+      "loss": 0.3688,
+      "step": 43180
+    },
+    {
+      "epoch": 0.9596071809456097,
+      "grad_norm": 2.625,
+      "learning_rate": 8.048888888888889e-07,
+      "loss": 0.3755,
+      "step": 43190
+    },
+    {
+      "epoch": 0.9598293636686811,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.004444444444445e-07,
+      "loss": 0.3626,
+      "step": 43200
+    },
+    {
+      "epoch": 0.9600515463917526,
+      "grad_norm": 2.375,
+      "learning_rate": 7.960000000000001e-07,
+      "loss": 0.3469,
+      "step": 43210
+    },
+    {
+      "epoch": 0.960273729114824,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.915555555555557e-07,
+      "loss": 0.3896,
+      "step": 43220
+    },
+    {
+      "epoch": 0.9604959118378955,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.871111111111112e-07,
+      "loss": 0.3818,
+      "step": 43230
+    },
+    {
+      "epoch": 0.9607180945609669,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.826666666666667e-07,
+      "loss": 0.3521,
+      "step": 43240
+    },
+    {
+      "epoch": 0.9609402772840384,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.782222222222223e-07,
+      "loss": 0.3824,
+      "step": 43250
+    },
+    {
+      "epoch": 0.9611624600071098,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.737777777777779e-07,
+      "loss": 0.3383,
+      "step": 43260
+    },
+    {
+      "epoch": 0.9613846427301813,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.693333333333335e-07,
+      "loss": 0.3442,
+      "step": 43270
+    },
+    {
+      "epoch": 0.9616068254532527,
+      "grad_norm": 2.375,
+      "learning_rate": 7.648888888888889e-07,
+      "loss": 0.3612,
+      "step": 43280
+    },
+    {
+      "epoch": 0.9618290081763242,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.604444444444445e-07,
+      "loss": 0.3635,
+      "step": 43290
+    },
+    {
+      "epoch": 0.9620511908993956,
+      "grad_norm": 3.0,
+      "learning_rate": 7.56e-07,
+      "loss": 0.3861,
+      "step": 43300
+    },
+    {
+      "epoch": 0.9622733736224671,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.515555555555556e-07,
+      "loss": 0.3739,
+      "step": 43310
+    },
+    {
+      "epoch": 0.9624955563455385,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.471111111111112e-07,
+      "loss": 0.336,
+      "step": 43320
+    },
+    {
+      "epoch": 0.96271773906861,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.426666666666667e-07,
+      "loss": 0.3691,
+      "step": 43330
+    },
+    {
+      "epoch": 0.9629399217916815,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.382222222222223e-07,
+      "loss": 0.3891,
+      "step": 43340
+    },
+    {
+      "epoch": 0.9631621045147529,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.337777777777779e-07,
+      "loss": 0.4046,
+      "step": 43350
+    },
+    {
+      "epoch": 0.9633842872378244,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.293333333333335e-07,
+      "loss": 0.3876,
+      "step": 43360
+    },
+    {
+      "epoch": 0.9636064699608958,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.24888888888889e-07,
+      "loss": 0.3588,
+      "step": 43370
+    },
+    {
+      "epoch": 0.9638286526839673,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.204444444444444e-07,
+      "loss": 0.3977,
+      "step": 43380
+    },
+    {
+      "epoch": 0.9640508354070387,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.16e-07,
+      "loss": 0.4028,
+      "step": 43390
+    },
+    {
+      "epoch": 0.9642730181301102,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.115555555555556e-07,
+      "loss": 0.373,
+      "step": 43400
+    },
+    {
+      "epoch": 0.9644952008531816,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.071111111111112e-07,
+      "loss": 0.3863,
+      "step": 43410
+    },
+    {
+      "epoch": 0.9647173835762531,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.026666666666668e-07,
+      "loss": 0.3439,
+      "step": 43420
+    },
+    {
+      "epoch": 0.9649395662993245,
+      "grad_norm": 2.125,
+      "learning_rate": 6.982222222222222e-07,
+      "loss": 0.3628,
+      "step": 43430
+    },
+    {
+      "epoch": 0.965161749022396,
+      "grad_norm": 2.125,
+      "learning_rate": 6.937777777777778e-07,
+      "loss": 0.3671,
+      "step": 43440
+    },
+    {
+      "epoch": 0.9653839317454674,
+      "grad_norm": 2.25,
+      "learning_rate": 6.893333333333334e-07,
+      "loss": 0.3832,
+      "step": 43450
+    },
+    {
+      "epoch": 0.965606114468539,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.84888888888889e-07,
+      "loss": 0.3556,
+      "step": 43460
+    },
+    {
+      "epoch": 0.9658282971916103,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.804444444444446e-07,
+      "loss": 0.3729,
+      "step": 43470
+    },
+    {
+      "epoch": 0.9660504799146818,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.76e-07,
+      "loss": 0.4146,
+      "step": 43480
+    },
+    {
+      "epoch": 0.9662726626377532,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.715555555555556e-07,
+      "loss": 0.3773,
+      "step": 43490
+    },
+    {
+      "epoch": 0.9664948453608248,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.671111111111111e-07,
+      "loss": 0.3963,
+      "step": 43500
+    },
+    {
+      "epoch": 0.9667170280838961,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.626666666666667e-07,
+      "loss": 0.3538,
+      "step": 43510
+    },
+    {
+      "epoch": 0.9669392108069677,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.582222222222223e-07,
+      "loss": 0.3399,
+      "step": 43520
+    },
+    {
+      "epoch": 0.9671613935300392,
+      "grad_norm": 2.5,
+      "learning_rate": 6.537777777777779e-07,
+      "loss": 0.3567,
+      "step": 43530
+    },
+    {
+      "epoch": 0.9673835762531106,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.493333333333334e-07,
+      "loss": 0.4067,
+      "step": 43540
+    },
+    {
+      "epoch": 0.9676057589761821,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.44888888888889e-07,
+      "loss": 0.3565,
+      "step": 43550
+    },
+    {
+      "epoch": 0.9678279416992535,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.404444444444446e-07,
+      "loss": 0.3918,
+      "step": 43560
+    },
+    {
+      "epoch": 0.968050124422325,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.360000000000001e-07,
+      "loss": 0.3324,
+      "step": 43570
+    },
+    {
+      "epoch": 0.9682723071453964,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.315555555555557e-07,
+      "loss": 0.3777,
+      "step": 43580
+    },
+    {
+      "epoch": 0.9684944898684679,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.271111111111111e-07,
+      "loss": 0.3734,
+      "step": 43590
+    },
+    {
+      "epoch": 0.9687166725915393,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.226666666666667e-07,
+      "loss": 0.3634,
+      "step": 43600
+    },
+    {
+      "epoch": 0.9689388553146108,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.182222222222223e-07,
+      "loss": 0.3954,
+      "step": 43610
+    },
+    {
+      "epoch": 0.9691610380376822,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.137777777777779e-07,
+      "loss": 0.4005,
+      "step": 43620
+    },
+    {
+      "epoch": 0.9693832207607537,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.093333333333333e-07,
+      "loss": 0.3846,
+      "step": 43630
+    },
+    {
+      "epoch": 0.9696054034838251,
+      "grad_norm": 2.375,
+      "learning_rate": 6.048888888888889e-07,
+      "loss": 0.3847,
+      "step": 43640
+    },
+    {
+      "epoch": 0.9698275862068966,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.004444444444445e-07,
+      "loss": 0.3993,
+      "step": 43650
+    },
+    {
+      "epoch": 0.970049768929968,
+      "grad_norm": 2.75,
+      "learning_rate": 5.960000000000001e-07,
+      "loss": 0.3835,
+      "step": 43660
+    },
+    {
+      "epoch": 0.9702719516530395,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.915555555555557e-07,
+      "loss": 0.3449,
+      "step": 43670
+    },
+    {
+      "epoch": 0.9704941343761109,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.871111111111112e-07,
+      "loss": 0.3962,
+      "step": 43680
+    },
+    {
+      "epoch": 0.9707163170991824,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.826666666666667e-07,
+      "loss": 0.3536,
+      "step": 43690
+    },
+    {
+      "epoch": 0.9709384998222538,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.782222222222222e-07,
+      "loss": 0.3503,
+      "step": 43700
+    },
+    {
+      "epoch": 0.9711606825453253,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.737777777777778e-07,
+      "loss": 0.3566,
+      "step": 43710
+    },
+    {
+      "epoch": 0.9713828652683967,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.693333333333334e-07,
+      "loss": 0.4052,
+      "step": 43720
+    },
+    {
+      "epoch": 0.9716050479914682,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.648888888888889e-07,
+      "loss": 0.3622,
+      "step": 43730
+    },
+    {
+      "epoch": 0.9718272307145397,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.604444444444445e-07,
+      "loss": 0.3668,
+      "step": 43740
+    },
+    {
+      "epoch": 0.9720494134376111,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.560000000000001e-07,
+      "loss": 0.3619,
+      "step": 43750
+    },
+    {
+      "epoch": 0.9722715961606826,
+      "grad_norm": 2.875,
+      "learning_rate": 5.515555555555556e-07,
+      "loss": 0.4131,
+      "step": 43760
+    },
+    {
+      "epoch": 0.972493778883754,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.471111111111112e-07,
+      "loss": 0.3642,
+      "step": 43770
+    },
+    {
+      "epoch": 0.9727159616068255,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.426666666666667e-07,
+      "loss": 0.3549,
+      "step": 43780
+    },
+    {
+      "epoch": 0.9729381443298969,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.382222222222223e-07,
+      "loss": 0.3783,
+      "step": 43790
+    },
+    {
+      "epoch": 0.9731603270529684,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.337777777777779e-07,
+      "loss": 0.3919,
+      "step": 43800
+    },
+    {
+      "epoch": 0.9733825097760398,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.293333333333334e-07,
+      "loss": 0.3956,
+      "step": 43810
+    },
+    {
+      "epoch": 0.9736046924991113,
+      "grad_norm": 3.125,
+      "learning_rate": 5.24888888888889e-07,
+      "loss": 0.4097,
+      "step": 43820
+    },
+    {
+      "epoch": 0.9738268752221827,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.204444444444444e-07,
+      "loss": 0.3635,
+      "step": 43830
+    },
+    {
+      "epoch": 0.9740490579452542,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.16e-07,
+      "loss": 0.4085,
+      "step": 43840
+    },
+    {
+      "epoch": 0.9742712406683256,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.115555555555556e-07,
+      "loss": 0.3866,
+      "step": 43850
+    },
+    {
+      "epoch": 0.9744934233913971,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.071111111111112e-07,
+      "loss": 0.3987,
+      "step": 43860
+    },
+    {
+      "epoch": 0.9747156061144685,
+      "grad_norm": 2.046875,
+      "learning_rate": 5.026666666666667e-07,
+      "loss": 0.3365,
+      "step": 43870
+    },
+    {
+      "epoch": 0.97493778883754,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.982222222222223e-07,
+      "loss": 0.3901,
+      "step": 43880
+    },
+    {
+      "epoch": 0.9751599715606114,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.937777777777778e-07,
+      "loss": 0.388,
+      "step": 43890
+    },
+    {
+      "epoch": 0.9753821542836829,
+      "grad_norm": 2.5,
+      "learning_rate": 4.893333333333334e-07,
+      "loss": 0.3724,
+      "step": 43900
+    },
+    {
+      "epoch": 0.9756043370067543,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.848888888888889e-07,
+      "loss": 0.3528,
+      "step": 43910
+    },
+    {
+      "epoch": 0.9758265197298258,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.804444444444445e-07,
+      "loss": 0.351,
+      "step": 43920
+    },
+    {
+      "epoch": 0.9760487024528972,
+      "grad_norm": 2.375,
+      "learning_rate": 4.760000000000001e-07,
+      "loss": 0.3827,
+      "step": 43930
+    },
+    {
+      "epoch": 0.9762708851759687,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.7155555555555556e-07,
+      "loss": 0.3802,
+      "step": 43940
+    },
+    {
+      "epoch": 0.9764930678990402,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.6711111111111115e-07,
+      "loss": 0.3979,
+      "step": 43950
+    },
+    {
+      "epoch": 0.9767152506221116,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.626666666666667e-07,
+      "loss": 0.3912,
+      "step": 43960
+    },
+    {
+      "epoch": 0.9769374333451831,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.5822222222222227e-07,
+      "loss": 0.42,
+      "step": 43970
+    },
+    {
+      "epoch": 0.9771596160682545,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.5377777777777785e-07,
+      "loss": 0.3759,
+      "step": 43980
+    },
+    {
+      "epoch": 0.977381798791326,
+      "grad_norm": 2.5,
+      "learning_rate": 4.4933333333333333e-07,
+      "loss": 0.3879,
+      "step": 43990
+    },
+    {
+      "epoch": 0.9776039815143974,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.448888888888889e-07,
+      "loss": 0.3706,
+      "step": 44000
+    },
+    {
+      "epoch": 0.9778261642374689,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.4044444444444445e-07,
+      "loss": 0.3588,
+      "step": 44010
+    },
+    {
+      "epoch": 0.9780483469605403,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.3600000000000004e-07,
+      "loss": 0.3787,
+      "step": 44020
+    },
+    {
+      "epoch": 0.9782705296836118,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.315555555555556e-07,
+      "loss": 0.3515,
+      "step": 44030
+    },
+    {
+      "epoch": 0.9784927124066832,
+      "grad_norm": 2.75,
+      "learning_rate": 4.271111111111111e-07,
+      "loss": 0.3578,
+      "step": 44040
+    },
+    {
+      "epoch": 0.9787148951297547,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.226666666666667e-07,
+      "loss": 0.3804,
+      "step": 44050
+    },
+    {
+      "epoch": 0.9789370778528261,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.182222222222222e-07,
+      "loss": 0.4112,
+      "step": 44060
+    },
+    {
+      "epoch": 0.9791592605758976,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.137777777777778e-07,
+      "loss": 0.3923,
+      "step": 44070
+    },
+    {
+      "epoch": 0.979381443298969,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.093333333333334e-07,
+      "loss": 0.3167,
+      "step": 44080
+    },
+    {
+      "epoch": 0.9796036260220405,
+      "grad_norm": 2.125,
+      "learning_rate": 4.048888888888889e-07,
+      "loss": 0.4019,
+      "step": 44090
+    },
+    {
+      "epoch": 0.9798258087451119,
+      "grad_norm": 2.75,
+      "learning_rate": 4.0044444444444447e-07,
+      "loss": 0.3911,
+      "step": 44100
+    },
+    {
+      "epoch": 0.9800479914681834,
+      "grad_norm": 2.875,
+      "learning_rate": 3.9600000000000005e-07,
+      "loss": 0.356,
+      "step": 44110
+    },
+    {
+      "epoch": 0.9802701741912548,
+      "grad_norm": 2.25,
+      "learning_rate": 3.915555555555556e-07,
+      "loss": 0.3712,
+      "step": 44120
+    },
+    {
+      "epoch": 0.9804923569143263,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.8711111111111117e-07,
+      "loss": 0.4247,
+      "step": 44130
+    },
+    {
+      "epoch": 0.9807145396373977,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.8266666666666665e-07,
+      "loss": 0.4134,
+      "step": 44140
+    },
+    {
+      "epoch": 0.9809367223604692,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.7822222222222224e-07,
+      "loss": 0.3876,
+      "step": 44150
+    },
+    {
+      "epoch": 0.9811589050835408,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.737777777777778e-07,
+      "loss": 0.3674,
+      "step": 44160
+    },
+    {
+      "epoch": 0.9813810878066122,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.6933333333333336e-07,
+      "loss": 0.3346,
+      "step": 44170
+    },
+    {
+      "epoch": 0.9816032705296837,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.6488888888888894e-07,
+      "loss": 0.3563,
+      "step": 44180
+    },
+    {
+      "epoch": 0.981825453252755,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.604444444444444e-07,
+      "loss": 0.4051,
+      "step": 44190
+    },
+    {
+      "epoch": 0.9820476359758266,
+      "grad_norm": 2.5,
+      "learning_rate": 3.56e-07,
+      "loss": 0.3901,
+      "step": 44200
+    },
+    {
+      "epoch": 0.982269818698898,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.515555555555556e-07,
+      "loss": 0.4045,
+      "step": 44210
+    },
+    {
+      "epoch": 0.9824920014219695,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.4711111111111113e-07,
+      "loss": 0.3926,
+      "step": 44220
+    },
+    {
+      "epoch": 0.9827141841450409,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.426666666666667e-07,
+      "loss": 0.3516,
+      "step": 44230
+    },
+    {
+      "epoch": 0.9829363668681124,
+      "grad_norm": 2.75,
+      "learning_rate": 3.382222222222222e-07,
+      "loss": 0.393,
+      "step": 44240
+    },
+    {
+      "epoch": 0.9831585495911838,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.337777777777778e-07,
+      "loss": 0.3764,
+      "step": 44250
+    },
+    {
+      "epoch": 0.9833807323142553,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.2933333333333337e-07,
+      "loss": 0.3908,
+      "step": 44260
+    },
+    {
+      "epoch": 0.9836029150373267,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.248888888888889e-07,
+      "loss": 0.4109,
+      "step": 44270
+    },
+    {
+      "epoch": 0.9838250977603982,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.204444444444445e-07,
+      "loss": 0.3827,
+      "step": 44280
+    },
+    {
+      "epoch": 0.9840472804834696,
+      "grad_norm": 2.3125,
+      "learning_rate": 3.160000000000001e-07,
+      "loss": 0.386,
+      "step": 44290
+    },
+    {
+      "epoch": 0.9842694632065411,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.1155555555555556e-07,
+      "loss": 0.3903,
+      "step": 44300
+    },
+    {
+      "epoch": 0.9844916459296125,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.0711111111111114e-07,
+      "loss": 0.4183,
+      "step": 44310
+    },
+    {
+      "epoch": 0.984713828652684,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.026666666666667e-07,
+      "loss": 0.39,
+      "step": 44320
+    },
+    {
+      "epoch": 0.9849360113757554,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.9822222222222226e-07,
+      "loss": 0.4059,
+      "step": 44330
+    },
+    {
+      "epoch": 0.9851581940988269,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.937777777777778e-07,
+      "loss": 0.3491,
+      "step": 44340
+    },
+    {
+      "epoch": 0.9853803768218983,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.8933333333333333e-07,
+      "loss": 0.3557,
+      "step": 44350
+    },
+    {
+      "epoch": 0.9856025595449698,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.848888888888889e-07,
+      "loss": 0.4059,
+      "step": 44360
+    },
+    {
+      "epoch": 0.9858247422680413,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.8044444444444445e-07,
+      "loss": 0.3636,
+      "step": 44370
+    },
+    {
+      "epoch": 0.9860469249911127,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.7600000000000004e-07,
+      "loss": 0.3925,
+      "step": 44380
+    },
+    {
+      "epoch": 0.9862691077141842,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.7155555555555557e-07,
+      "loss": 0.358,
+      "step": 44390
+    },
+    {
+      "epoch": 0.9864912904372556,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.6711111111111116e-07,
+      "loss": 0.3771,
+      "step": 44400
+    },
+    {
+      "epoch": 0.9867134731603271,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.626666666666667e-07,
+      "loss": 0.4057,
+      "step": 44410
+    },
+    {
+      "epoch": 0.9869356558833985,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.582222222222222e-07,
+      "loss": 0.3639,
+      "step": 44420
+    },
+    {
+      "epoch": 0.98715783860647,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.537777777777778e-07,
+      "loss": 0.397,
+      "step": 44430
+    },
+    {
+      "epoch": 0.9873800213295414,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.4933333333333334e-07,
+      "loss": 0.3519,
+      "step": 44440
+    },
+    {
+      "epoch": 0.9876022040526129,
+      "grad_norm": 1.90625,
+      "learning_rate": 2.4488888888888893e-07,
+      "loss": 0.3831,
+      "step": 44450
+    },
+    {
+      "epoch": 0.9878243867756843,
+      "grad_norm": 2.3125,
+      "learning_rate": 2.4044444444444446e-07,
+      "loss": 0.3542,
+      "step": 44460
+    },
+    {
+      "epoch": 0.9880465694987558,
+      "grad_norm": 2.296875,
+      "learning_rate": 2.3600000000000002e-07,
+      "loss": 0.3645,
+      "step": 44470
+    },
+    {
+      "epoch": 0.9882687522218272,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.3155555555555556e-07,
+      "loss": 0.3595,
+      "step": 44480
+    },
+    {
+      "epoch": 0.9884909349448987,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.2711111111111114e-07,
+      "loss": 0.3782,
+      "step": 44490
+    },
+    {
+      "epoch": 0.9887131176679701,
+      "grad_norm": 3.0,
+      "learning_rate": 2.226666666666667e-07,
+      "loss": 0.4173,
+      "step": 44500
+    },
+    {
+      "epoch": 0.9889353003910416,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.1822222222222224e-07,
+      "loss": 0.4032,
+      "step": 44510
+    },
+    {
+      "epoch": 0.989157483114113,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.137777777777778e-07,
+      "loss": 0.3444,
+      "step": 44520
+    },
+    {
+      "epoch": 0.9893796658371845,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.0933333333333335e-07,
+      "loss": 0.3868,
+      "step": 44530
+    },
+    {
+      "epoch": 0.9896018485602559,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.0488888888888891e-07,
+      "loss": 0.3836,
+      "step": 44540
+    },
+    {
+      "epoch": 0.9898240312833274,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.0044444444444447e-07,
+      "loss": 0.3686,
+      "step": 44550
+    },
+    {
+      "epoch": 0.9900462140063989,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.96e-07,
+      "loss": 0.3866,
+      "step": 44560
+    },
+    {
+      "epoch": 0.9902683967294703,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.9155555555555557e-07,
+      "loss": 0.3696,
+      "step": 44570
+    },
+    {
+      "epoch": 0.9904905794525418,
+      "grad_norm": 2.125,
+      "learning_rate": 1.8711111111111113e-07,
+      "loss": 0.376,
+      "step": 44580
+    },
+    {
+      "epoch": 0.9907127621756132,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.826666666666667e-07,
+      "loss": 0.3585,
+      "step": 44590
+    },
+    {
+      "epoch": 0.9909349448986847,
+      "grad_norm": 3.0,
+      "learning_rate": 1.7822222222222222e-07,
+      "loss": 0.3533,
+      "step": 44600
+    },
+    {
+      "epoch": 0.9911571276217561,
+      "grad_norm": 2.75,
+      "learning_rate": 1.7377777777777778e-07,
+      "loss": 0.3631,
+      "step": 44610
+    },
+    {
+      "epoch": 0.9913793103448276,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.6933333333333337e-07,
+      "loss": 0.3594,
+      "step": 44620
+    },
+    {
+      "epoch": 0.991601493067899,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.648888888888889e-07,
+      "loss": 0.3878,
+      "step": 44630
+    },
+    {
+      "epoch": 0.9918236757909705,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.6044444444444446e-07,
+      "loss": 0.3635,
+      "step": 44640
+    },
+    {
+      "epoch": 0.9920458585140419,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.56e-07,
+      "loss": 0.3724,
+      "step": 44650
+    },
+    {
+      "epoch": 0.9922680412371134,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.5155555555555558e-07,
+      "loss": 0.3796,
+      "step": 44660
+    },
+    {
+      "epoch": 0.9924902239601848,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.4711111111111111e-07,
+      "loss": 0.3429,
+      "step": 44670
+    },
+    {
+      "epoch": 0.9927124066832563,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.4266666666666667e-07,
+      "loss": 0.3699,
+      "step": 44680
+    },
+    {
+      "epoch": 0.9929345894063277,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.3822222222222223e-07,
+      "loss": 0.3655,
+      "step": 44690
+    },
+    {
+      "epoch": 0.9931567721293992,
+      "grad_norm": 2.125,
+      "learning_rate": 1.337777777777778e-07,
+      "loss": 0.392,
+      "step": 44700
+    },
+    {
+      "epoch": 0.9933789548524706,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.2933333333333335e-07,
+      "loss": 0.4055,
+      "step": 44710
+    },
+    {
+      "epoch": 0.9936011375755421,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.248888888888889e-07,
+      "loss": 0.4049,
+      "step": 44720
+    },
+    {
+      "epoch": 0.9938233202986135,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.2044444444444445e-07,
+      "loss": 0.3453,
+      "step": 44730
+    },
+    {
+      "epoch": 0.994045503021685,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.16e-07,
+      "loss": 0.3585,
+      "step": 44740
+    },
+    {
+      "epoch": 0.9942676857447564,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.1155555555555557e-07,
+      "loss": 0.3736,
+      "step": 44750
+    },
+    {
+      "epoch": 0.9944898684678279,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.0711111111111111e-07,
+      "loss": 0.3338,
+      "step": 44760
+    },
+    {
+      "epoch": 0.9947120511908994,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.0266666666666667e-07,
+      "loss": 0.3935,
+      "step": 44770
+    },
+    {
+      "epoch": 0.9949342339139708,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.822222222222222e-08,
+      "loss": 0.3645,
+      "step": 44780
+    },
+    {
+      "epoch": 0.9951564166370424,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.377777777777779e-08,
+      "loss": 0.4132,
+      "step": 44790
+    },
+    {
+      "epoch": 0.9953785993601137,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.933333333333334e-08,
+      "loss": 0.3645,
+      "step": 44800
+    },
+    {
+      "epoch": 0.9956007820831853,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.48888888888889e-08,
+      "loss": 0.3802,
+      "step": 44810
+    },
+    {
+      "epoch": 0.9958229648062566,
+      "grad_norm": 2.375,
+      "learning_rate": 8.044444444444445e-08,
+      "loss": 0.4019,
+      "step": 44820
+    },
+    {
+      "epoch": 0.9960451475293282,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.6e-08,
+      "loss": 0.3698,
+      "step": 44830
+    },
+    {
+      "epoch": 0.9962673302523996,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.155555555555557e-08,
+      "loss": 0.3613,
+      "step": 44840
+    },
+    {
+      "epoch": 0.9964895129754711,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.711111111111111e-08,
+      "loss": 0.334,
+      "step": 44850
+    },
+    {
+      "epoch": 0.9967116956985425,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.266666666666667e-08,
+      "loss": 0.3732,
+      "step": 44860
+    },
+    {
+      "epoch": 0.996933878421614,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.822222222222223e-08,
+      "loss": 0.3744,
+      "step": 44870
+    },
+    {
+      "epoch": 0.9971560611446854,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.3777777777777785e-08,
+      "loss": 0.3746,
+      "step": 44880
+    },
+    {
+      "epoch": 0.9973782438677569,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.933333333333333e-08,
+      "loss": 0.3952,
+      "step": 44890
+    },
+    {
+      "epoch": 0.9976004265908283,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.488888888888889e-08,
+      "loss": 0.3463,
+      "step": 44900
+    },
+    {
+      "epoch": 0.9978226093138998,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.0444444444444445e-08,
+      "loss": 0.3189,
+      "step": 44910
+    },
+    {
+      "epoch": 0.9980447920369712,
+      "grad_norm": 3.0,
+      "learning_rate": 3.6000000000000005e-08,
+      "loss": 0.4051,
+      "step": 44920
+    },
+    {
+      "epoch": 0.9982669747600427,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.155555555555556e-08,
+      "loss": 0.3898,
+      "step": 44930
+    },
+    {
+      "epoch": 0.9984891574831141,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.7111111111111115e-08,
+      "loss": 0.4055,
+      "step": 44940
+    },
+    {
+      "epoch": 0.9987113402061856,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.266666666666667e-08,
+      "loss": 0.3668,
+      "step": 44950
+    },
+    {
+      "epoch": 0.998933522929257,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.8222222222222224e-08,
+      "loss": 0.3609,
+      "step": 44960
+    },
+    {
+      "epoch": 0.9991557056523285,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.3777777777777778e-08,
+      "loss": 0.3752,
+      "step": 44970
+    },
+    {
+      "epoch": 0.9993778883754,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.333333333333334e-09,
+      "loss": 0.3807,
+      "step": 44980
+    },
+    {
+      "epoch": 0.9996000710984714,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.888888888888889e-09,
+      "loss": 0.3794,
+      "step": 44990
+    },
+    {
+      "epoch": 0.9998222538215429,
+      "grad_norm": 2.75,
+      "learning_rate": 4.444444444444445e-10,
+      "loss": 0.3628,
+      "step": 45000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 45000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3115342813852418e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}