diff --git "a/checkpoint-13500/trainer_state.json" "b/checkpoint-13500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-13500/trainer_state.json"
@@ -0,0 +1,9484 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.957123699757167,
+  "eval_steps": 500,
+  "global_step": 13500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014497481062665362,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.9987142857142857e-05,
+      "loss": 1.3618,
+      "step": 10
+    },
+    {
+      "epoch": 0.0028994962125330724,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9972857142857144e-05,
+      "loss": 0.8381,
+      "step": 20
+    },
+    {
+      "epoch": 0.004349244318799609,
+      "grad_norm": 1.75,
+      "learning_rate": 1.995857142857143e-05,
+      "loss": 0.7389,
+      "step": 30
+    },
+    {
+      "epoch": 0.005798992425066145,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9944285714285717e-05,
+      "loss": 0.667,
+      "step": 40
+    },
+    {
+      "epoch": 0.007248740531332681,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9930000000000004e-05,
+      "loss": 0.6777,
+      "step": 50
+    },
+    {
+      "epoch": 0.008698488637599217,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9915714285714287e-05,
+      "loss": 0.6787,
+      "step": 60
+    },
+    {
+      "epoch": 0.010148236743865753,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9901428571428574e-05,
+      "loss": 0.7067,
+      "step": 70
+    },
+    {
+      "epoch": 0.01159798485013229,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9887142857142858e-05,
+      "loss": 0.6614,
+      "step": 80
+    },
+    {
+      "epoch": 0.013047732956398826,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9872857142857145e-05,
+      "loss": 0.6237,
+      "step": 90
+    },
+    {
+      "epoch": 0.014497481062665362,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9858571428571428e-05,
+      "loss": 0.629,
+      "step": 100
+    },
+    {
+      "epoch": 0.015947229168931897,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9844285714285715e-05,
+      "loss": 0.6437,
+      "step": 110
+    },
+    {
+      "epoch": 0.017396977275198434,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.983e-05,
+      "loss": 0.637,
+      "step": 120
+    },
+    {
+      "epoch": 0.01884672538146497,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.981571428571429e-05,
+      "loss": 0.641,
+      "step": 130
+    },
+    {
+      "epoch": 0.020296473487731507,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9801428571428575e-05,
+      "loss": 0.6419,
+      "step": 140
+    },
+    {
+      "epoch": 0.02174622159399804,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.978714285714286e-05,
+      "loss": 0.6563,
+      "step": 150
+    },
+    {
+      "epoch": 0.02319596970026458,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9772857142857145e-05,
+      "loss": 0.6336,
+      "step": 160
+    },
+    {
+      "epoch": 0.024645717806531114,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.975857142857143e-05,
+      "loss": 0.6273,
+      "step": 170
+    },
+    {
+      "epoch": 0.02609546591279765,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9744285714285716e-05,
+      "loss": 0.5986,
+      "step": 180
+    },
+    {
+      "epoch": 0.027545214019064186,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9730000000000003e-05,
+      "loss": 0.5938,
+      "step": 190
+    },
+    {
+      "epoch": 0.028994962125330724,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9715714285714286e-05,
+      "loss": 0.6233,
+      "step": 200
+    },
+    {
+      "epoch": 0.03044471023159726,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9701428571428573e-05,
+      "loss": 0.598,
+      "step": 210
+    },
+    {
+      "epoch": 0.03189445833786379,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.968714285714286e-05,
+      "loss": 0.5952,
+      "step": 220
+    },
+    {
+      "epoch": 0.03334420644413033,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9672857142857143e-05,
+      "loss": 0.6146,
+      "step": 230
+    },
+    {
+      "epoch": 0.03479395455039687,
+      "grad_norm": 1.875,
+      "learning_rate": 1.965857142857143e-05,
+      "loss": 0.6049,
+      "step": 240
+    },
+    {
+      "epoch": 0.03624370265666341,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9644285714285717e-05,
+      "loss": 0.5934,
+      "step": 250
+    },
+    {
+      "epoch": 0.03769345076292994,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9630000000000003e-05,
+      "loss": 0.6216,
+      "step": 260
+    },
+    {
+      "epoch": 0.039143198869196476,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9615714285714287e-05,
+      "loss": 0.6243,
+      "step": 270
+    },
+    {
+      "epoch": 0.040592946975463014,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9601428571428574e-05,
+      "loss": 0.5642,
+      "step": 280
+    },
+    {
+      "epoch": 0.04204269508172955,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9587142857142857e-05,
+      "loss": 0.5497,
+      "step": 290
+    },
+    {
+      "epoch": 0.04349244318799608,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9572857142857144e-05,
+      "loss": 0.5838,
+      "step": 300
+    },
+    {
+      "epoch": 0.04494219129426262,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.955857142857143e-05,
+      "loss": 0.5611,
+      "step": 310
+    },
+    {
+      "epoch": 0.04639193940052916,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9544285714285714e-05,
+      "loss": 0.5583,
+      "step": 320
+    },
+    {
+      "epoch": 0.047841687506795696,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.953e-05,
+      "loss": 0.6231,
+      "step": 330
+    },
+    {
+      "epoch": 0.04929143561306223,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9515714285714288e-05,
+      "loss": 0.5866,
+      "step": 340
+    },
+    {
+      "epoch": 0.050741183719328765,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9501428571428575e-05,
+      "loss": 0.6276,
+      "step": 350
+    },
+    {
+      "epoch": 0.0521909318255953,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9487142857142858e-05,
+      "loss": 0.5807,
+      "step": 360
+    },
+    {
+      "epoch": 0.05364067993186184,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9472857142857145e-05,
+      "loss": 0.5872,
+      "step": 370
+    },
+    {
+      "epoch": 0.05509042803812837,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9458571428571432e-05,
+      "loss": 0.5537,
+      "step": 380
+    },
+    {
+      "epoch": 0.05654017614439491,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9444285714285715e-05,
+      "loss": 0.5759,
+      "step": 390
+    },
+    {
+      "epoch": 0.05798992425066145,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9430000000000002e-05,
+      "loss": 0.5865,
+      "step": 400
+    },
+    {
+      "epoch": 0.059439672356927986,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9415714285714285e-05,
+      "loss": 0.5987,
+      "step": 410
+    },
+    {
+      "epoch": 0.06088942046319452,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9401428571428572e-05,
+      "loss": 0.5719,
+      "step": 420
+    },
+    {
+      "epoch": 0.062339168569461055,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.938714285714286e-05,
+      "loss": 0.4918,
+      "step": 430
+    },
+    {
+      "epoch": 0.06378891667572759,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9372857142857146e-05,
+      "loss": 0.5742,
+      "step": 440
+    },
+    {
+      "epoch": 0.06523866478199412,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.935857142857143e-05,
+      "loss": 0.5629,
+      "step": 450
+    },
+    {
+      "epoch": 0.06668841288826066,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9344285714285716e-05,
+      "loss": 0.5642,
+      "step": 460
+    },
+    {
+      "epoch": 0.0681381609945272,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9330000000000003e-05,
+      "loss": 0.5652,
+      "step": 470
+    },
+    {
+      "epoch": 0.06958790910079374,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9315714285714286e-05,
+      "loss": 0.5984,
+      "step": 480
+    },
+    {
+      "epoch": 0.07103765720706028,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9301428571428573e-05,
+      "loss": 0.5607,
+      "step": 490
+    },
+    {
+      "epoch": 0.07248740531332681,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9287142857142857e-05,
+      "loss": 0.5492,
+      "step": 500
+    },
+    {
+      "epoch": 0.07393715341959335,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9272857142857143e-05,
+      "loss": 0.5905,
+      "step": 510
+    },
+    {
+      "epoch": 0.07538690152585988,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.925857142857143e-05,
+      "loss": 0.5439,
+      "step": 520
+    },
+    {
+      "epoch": 0.07683664963212641,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9244285714285714e-05,
+      "loss": 0.5865,
+      "step": 530
+    },
+    {
+      "epoch": 0.07828639773839295,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.923e-05,
+      "loss": 0.542,
+      "step": 540
+    },
+    {
+      "epoch": 0.07973614584465949,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9215714285714287e-05,
+      "loss": 0.543,
+      "step": 550
+    },
+    {
+      "epoch": 0.08118589395092603,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9201428571428574e-05,
+      "loss": 0.502,
+      "step": 560
+    },
+    {
+      "epoch": 0.08263564205719257,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.918714285714286e-05,
+      "loss": 0.5264,
+      "step": 570
+    },
+    {
+      "epoch": 0.0840853901634591,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9172857142857144e-05,
+      "loss": 0.515,
+      "step": 580
+    },
+    {
+      "epoch": 0.08553513826972564,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.915857142857143e-05,
+      "loss": 0.5219,
+      "step": 590
+    },
+    {
+      "epoch": 0.08698488637599217,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9144285714285715e-05,
+      "loss": 0.5348,
+      "step": 600
+    },
+    {
+      "epoch": 0.0884346344822587,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.913e-05,
+      "loss": 0.532,
+      "step": 610
+    },
+    {
+      "epoch": 0.08988438258852524,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9115714285714285e-05,
+      "loss": 0.5263,
+      "step": 620
+    },
+    {
+      "epoch": 0.09133413069479178,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9101428571428572e-05,
+      "loss": 0.5698,
+      "step": 630
+    },
+    {
+      "epoch": 0.09278387880105832,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.908714285714286e-05,
+      "loss": 0.4835,
+      "step": 640
+    },
+    {
+      "epoch": 0.09423362690732486,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9072857142857145e-05,
+      "loss": 0.5404,
+      "step": 650
+    },
+    {
+      "epoch": 0.09568337501359139,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9058571428571432e-05,
+      "loss": 0.5061,
+      "step": 660
+    },
+    {
+      "epoch": 0.09713312311985793,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9044285714285716e-05,
+      "loss": 0.5508,
+      "step": 670
+    },
+    {
+      "epoch": 0.09858287122612445,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.9030000000000002e-05,
+      "loss": 0.5165,
+      "step": 680
+    },
+    {
+      "epoch": 0.10003261933239099,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9015714285714286e-05,
+      "loss": 0.5336,
+      "step": 690
+    },
+    {
+      "epoch": 0.10148236743865753,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9001428571428573e-05,
+      "loss": 0.5299,
+      "step": 700
+    },
+    {
+      "epoch": 0.10293211554492407,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.898714285714286e-05,
+      "loss": 0.5151,
+      "step": 710
+    },
+    {
+      "epoch": 0.1043818636511906,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8972857142857143e-05,
+      "loss": 0.5621,
+      "step": 720
+    },
+    {
+      "epoch": 0.10583161175745714,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.895857142857143e-05,
+      "loss": 0.5075,
+      "step": 730
+    },
+    {
+      "epoch": 0.10728135986372368,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.8944285714285717e-05,
+      "loss": 0.4995,
+      "step": 740
+    },
+    {
+      "epoch": 0.10873110796999022,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.893e-05,
+      "loss": 0.5096,
+      "step": 750
+    },
+    {
+      "epoch": 0.11018085607625674,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8915714285714287e-05,
+      "loss": 0.4965,
+      "step": 760
+    },
+    {
+      "epoch": 0.11163060418252328,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.8901428571428574e-05,
+      "loss": 0.5114,
+      "step": 770
+    },
+    {
+      "epoch": 0.11308035228878982,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.888714285714286e-05,
+      "loss": 0.5009,
+      "step": 780
+    },
+    {
+      "epoch": 0.11453010039505636,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.8872857142857144e-05,
+      "loss": 0.521,
+      "step": 790
+    },
+    {
+      "epoch": 0.1159798485013229,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.885857142857143e-05,
+      "loss": 0.4932,
+      "step": 800
+    },
+    {
+      "epoch": 0.11742959660758943,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8844285714285714e-05,
+      "loss": 0.4661,
+      "step": 810
+    },
+    {
+      "epoch": 0.11887934471385597,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.883e-05,
+      "loss": 0.4992,
+      "step": 820
+    },
+    {
+      "epoch": 0.12032909282012251,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8815714285714288e-05,
+      "loss": 0.4953,
+      "step": 830
+    },
+    {
+      "epoch": 0.12177884092638903,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.880142857142857e-05,
+      "loss": 0.4806,
+      "step": 840
+    },
+    {
+      "epoch": 0.12322858903265557,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8787142857142858e-05,
+      "loss": 0.5048,
+      "step": 850
+    },
+    {
+      "epoch": 0.12467833713892211,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.8772857142857145e-05,
+      "loss": 0.4715,
+      "step": 860
+    },
+    {
+      "epoch": 0.12612808524518865,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.875857142857143e-05,
+      "loss": 0.5255,
+      "step": 870
+    },
+    {
+      "epoch": 0.12757783335145517,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.8744285714285715e-05,
+      "loss": 0.4847,
+      "step": 880
+    },
+    {
+      "epoch": 0.12902758145772172,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.8730000000000002e-05,
+      "loss": 0.4672,
+      "step": 890
+    },
+    {
+      "epoch": 0.13047732956398825,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.871571428571429e-05,
+      "loss": 0.4915,
+      "step": 900
+    },
+    {
+      "epoch": 0.1319270776702548,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8701428571428572e-05,
+      "loss": 0.4913,
+      "step": 910
+    },
+    {
+      "epoch": 0.13337682577652132,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.868714285714286e-05,
+      "loss": 0.4836,
+      "step": 920
+    },
+    {
+      "epoch": 0.13482657388278788,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.8672857142857142e-05,
+      "loss": 0.4809,
+      "step": 930
+    },
+    {
+      "epoch": 0.1362763219890544,
+      "grad_norm": 2.125,
+      "learning_rate": 1.865857142857143e-05,
+      "loss": 0.4398,
+      "step": 940
+    },
+    {
+      "epoch": 0.13772607009532095,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8644285714285716e-05,
+      "loss": 0.4701,
+      "step": 950
+    },
+    {
+      "epoch": 0.13917581820158748,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.8630000000000003e-05,
+      "loss": 0.4475,
+      "step": 960
+    },
+    {
+      "epoch": 0.140625566307854,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.861571428571429e-05,
+      "loss": 0.4357,
+      "step": 970
+    },
+    {
+      "epoch": 0.14207531441412055,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.8601428571428573e-05,
+      "loss": 0.4504,
+      "step": 980
+    },
+    {
+      "epoch": 0.14352506252038708,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.858714285714286e-05,
+      "loss": 0.4868,
+      "step": 990
+    },
+    {
+      "epoch": 0.14497481062665363,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.8572857142857143e-05,
+      "loss": 0.4788,
+      "step": 1000
+    },
+    {
+      "epoch": 0.14642455873292015,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.855857142857143e-05,
+      "loss": 0.4693,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1478743068391867,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8544285714285717e-05,
+      "loss": 0.4538,
+      "step": 1020
+    },
+    {
+      "epoch": 0.14932405494545323,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.853e-05,
+      "loss": 0.4613,
+      "step": 1030
+    },
+    {
+      "epoch": 0.15077380305171975,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.8515714285714287e-05,
+      "loss": 0.4627,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1522235511579863,
+      "grad_norm": 2.0,
+      "learning_rate": 1.850142857142857e-05,
+      "loss": 0.4583,
+      "step": 1050
+    },
+    {
+      "epoch": 0.15367329926425283,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8487142857142858e-05,
+      "loss": 0.45,
+      "step": 1060
+    },
+    {
+      "epoch": 0.15512304737051938,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.8472857142857144e-05,
+      "loss": 0.4408,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1565727954767859,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.845857142857143e-05,
+      "loss": 0.4414,
+      "step": 1080
+    },
+    {
+      "epoch": 0.15802254358305245,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.8444285714285718e-05,
+      "loss": 0.4629,
+      "step": 1090
+    },
+    {
+      "epoch": 0.15947229168931898,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.843e-05,
+      "loss": 0.4622,
+      "step": 1100
+    },
+    {
+      "epoch": 0.16092203979558553,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.8415714285714288e-05,
+      "loss": 0.4543,
+      "step": 1110
+    },
+    {
+      "epoch": 0.16237178790185205,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.840142857142857e-05,
+      "loss": 0.4618,
+      "step": 1120
+    },
+    {
+      "epoch": 0.16382153600811858,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.838714285714286e-05,
+      "loss": 0.4301,
+      "step": 1130
+    },
+    {
+      "epoch": 0.16527128411438513,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.8372857142857142e-05,
+      "loss": 0.453,
+      "step": 1140
+    },
+    {
+      "epoch": 0.16672103222065165,
+      "grad_norm": 2.0,
+      "learning_rate": 1.835857142857143e-05,
+      "loss": 0.4387,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1681707803269182,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8344285714285716e-05,
+      "loss": 0.4143,
+      "step": 1160
+    },
+    {
+      "epoch": 0.16962052843318473,
+      "grad_norm": 1.875,
+      "learning_rate": 1.8330000000000002e-05,
+      "loss": 0.4672,
+      "step": 1170
+    },
+    {
+      "epoch": 0.17107027653945128,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.831571428571429e-05,
+      "loss": 0.456,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1725200246457178,
+      "grad_norm": 2.125,
+      "learning_rate": 1.8301428571428573e-05,
+      "loss": 0.4561,
+      "step": 1190
+    },
+    {
+      "epoch": 0.17396977275198433,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.828714285714286e-05,
+      "loss": 0.4441,
+      "step": 1200
+    },
+    {
+      "epoch": 0.17541952085825088,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8272857142857143e-05,
+      "loss": 0.453,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1768692689645174,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.825857142857143e-05,
+      "loss": 0.4231,
+      "step": 1220
+    },
+    {
+      "epoch": 0.17831901707078396,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.8244285714285717e-05,
+      "loss": 0.4145,
+      "step": 1230
+    },
+    {
+      "epoch": 0.17976876517705048,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.823e-05,
+      "loss": 0.4516,
+      "step": 1240
+    },
+    {
+      "epoch": 0.18121851328331703,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.8215714285714287e-05,
+      "loss": 0.4489,
+      "step": 1250
+    },
+    {
+      "epoch": 0.18266826138958356,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.8201428571428574e-05,
+      "loss": 0.4523,
+      "step": 1260
+    },
+    {
+      "epoch": 0.18411800949585008,
+      "grad_norm": 1.875,
+      "learning_rate": 1.818714285714286e-05,
+      "loss": 0.4579,
+      "step": 1270
+    },
+    {
+      "epoch": 0.18556775760211663,
+      "grad_norm": 2.125,
+      "learning_rate": 1.8172857142857144e-05,
+      "loss": 0.4335,
+      "step": 1280
+    },
+    {
+      "epoch": 0.18701750570838316,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.815857142857143e-05,
+      "loss": 0.4244,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1884672538146497,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.8144285714285717e-05,
+      "loss": 0.3976,
+      "step": 1300
+    },
+    {
+      "epoch": 0.18991700192091623,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.813e-05,
+      "loss": 0.4345,
+      "step": 1310
+    },
+    {
+      "epoch": 0.19136675002718279,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8115714285714288e-05,
+      "loss": 0.4204,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1928164981334493,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.810142857142857e-05,
+      "loss": 0.4014,
+      "step": 1330
+    },
+    {
+      "epoch": 0.19426624623971586,
+      "grad_norm": 2.0,
+      "learning_rate": 1.8087142857142858e-05,
+      "loss": 0.4295,
+      "step": 1340
+    },
+    {
+      "epoch": 0.19571599434598239,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8072857142857145e-05,
+      "loss": 0.4046,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1971657424522489,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.8058571428571428e-05,
+      "loss": 0.4213,
+      "step": 1360
+    },
+    {
+      "epoch": 0.19861549055851546,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.8044285714285715e-05,
+      "loss": 0.4186,
+      "step": 1370
+    },
+    {
+      "epoch": 0.20006523866478199,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.8030000000000002e-05,
+      "loss": 0.4438,
+      "step": 1380
+    },
+    {
+      "epoch": 0.20151498677104854,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.801571428571429e-05,
+      "loss": 0.4184,
+      "step": 1390
+    },
+    {
+      "epoch": 0.20296473487731506,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.8001428571428572e-05,
+      "loss": 0.4228,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2044144829835816,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.798714285714286e-05,
+      "loss": 0.4018,
+      "step": 1410
+    },
+    {
+      "epoch": 0.20586423108984814,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.7972857142857146e-05,
+      "loss": 0.4109,
+      "step": 1420
+    },
+    {
+      "epoch": 0.20731397919611466,
+      "grad_norm": 2.25,
+      "learning_rate": 1.795857142857143e-05,
+      "loss": 0.3866,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2087637273023812,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.7944285714285716e-05,
+      "loss": 0.4344,
+      "step": 1440
+    },
+    {
+      "epoch": 0.21021347540864774,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.793e-05,
+      "loss": 0.4,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2116632235149143,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.7915714285714286e-05,
+      "loss": 0.4229,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2131129716211808,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.7901428571428573e-05,
+      "loss": 0.386,
+      "step": 1470
+    },
+    {
+      "epoch": 0.21456271972744737,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.788714285714286e-05,
+      "loss": 0.396,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2160124678337139,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.7872857142857147e-05,
+      "loss": 0.3972,
+      "step": 1490
+    },
+    {
+      "epoch": 0.21746221593998044,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.785857142857143e-05,
+      "loss": 0.359,
+      "step": 1500
+    },
+    {
+      "epoch": 0.21891196404624697,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7844285714285717e-05,
+      "loss": 0.3996,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2203617121525135,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.783e-05,
+      "loss": 0.4019,
+      "step": 1520
+    },
+    {
+      "epoch": 0.22181146025878004,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.7815714285714287e-05,
+      "loss": 0.4037,
+      "step": 1530
+    },
+    {
+      "epoch": 0.22326120836504657,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.7801428571428574e-05,
+      "loss": 0.4173,
+      "step": 1540
+    },
+    {
+      "epoch": 0.22471095647131312,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.7787142857142857e-05,
+      "loss": 0.4191,
+      "step": 1550
+    },
+    {
+      "epoch": 0.22616070457757964,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.7772857142857144e-05,
+      "loss": 0.3939,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2276104526838462,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.775857142857143e-05,
+      "loss": 0.4173,
+      "step": 1570
+    },
+    {
+      "epoch": 0.22906020079011272,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.7744285714285715e-05,
+      "loss": 0.4093,
+      "step": 1580
+    },
+    {
+      "epoch": 0.23050994889637924,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.773e-05,
+      "loss": 0.3745,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2319596970026458,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.7715714285714288e-05,
+      "loss": 0.3717,
+      "step": 1600
+    },
+    {
+      "epoch": 0.23340944510891232,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.7701428571428575e-05,
+      "loss": 0.3746,
+      "step": 1610
+    },
+    {
+      "epoch": 0.23485919321517887,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.768714285714286e-05,
+      "loss": 0.4211,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2363089413214454,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.7672857142857145e-05,
+      "loss": 0.405,
+      "step": 1630
+    },
+    {
+      "epoch": 0.23775868942771194,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.765857142857143e-05,
+      "loss": 0.3834,
+      "step": 1640
+    },
+    {
+      "epoch": 0.23920843753397847,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7644285714285715e-05,
+      "loss": 0.4133,
+      "step": 1650
+    },
+    {
+      "epoch": 0.24065818564024502,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.763e-05,
+      "loss": 0.4348,
+      "step": 1660
+    },
+    {
+      "epoch": 0.24210793374651154,
+      "grad_norm": 2.75,
+      "learning_rate": 1.7615714285714286e-05,
+      "loss": 0.3676,
+      "step": 1670
+    },
+    {
+      "epoch": 0.24355768185277807,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.7601428571428573e-05,
+      "loss": 0.385,
+      "step": 1680
+    },
+    {
+      "epoch": 0.24500742995904462,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.758714285714286e-05,
+      "loss": 0.3871,
+      "step": 1690
+    },
+    {
+      "epoch": 0.24645717806531114,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.7572857142857146e-05,
+      "loss": 0.3553,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2479069261715777,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.755857142857143e-05,
+      "loss": 0.3638,
+      "step": 1710
+    },
+    {
+      "epoch": 0.24935667427784422,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.7544285714285716e-05,
+      "loss": 0.392,
+      "step": 1720
+    },
+    {
+      "epoch": 0.25080642238411077,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.753e-05,
+      "loss": 0.3908,
+      "step": 1730
+    },
+    {
+      "epoch": 0.2522561704903773,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.7515714285714287e-05,
+      "loss": 0.367,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2537059185966438,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.7501428571428574e-05,
+      "loss": 0.3688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.25515566670291034,
+      "grad_norm": 2.125,
+      "learning_rate": 1.7487142857142857e-05,
+      "loss": 0.3938,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2566054148091769,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.7472857142857144e-05,
+      "loss": 0.3925,
+      "step": 1770
+    },
+    {
+      "epoch": 0.25805516291544345,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.745857142857143e-05,
+      "loss": 0.3619,
+      "step": 1780
+    },
+    {
+      "epoch": 0.25950491102170997,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7444285714285717e-05,
+      "loss": 0.4038,
+      "step": 1790
+    },
+    {
+      "epoch": 0.2609546591279765,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.743e-05,
+      "loss": 0.3533,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2624044072342431,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.7415714285714288e-05,
+      "loss": 0.3591,
+      "step": 1810
+    },
+    {
+      "epoch": 0.2638541553405096,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.7401428571428574e-05,
+      "loss": 0.3641,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2653039034467761,
+      "grad_norm": 1.875,
+      "learning_rate": 1.7387142857142858e-05,
+      "loss": 0.3802,
+      "step": 1830
+    },
+    {
+      "epoch": 0.26675365155304265,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.7372857142857145e-05,
+      "loss": 0.3516,
+      "step": 1840
+    },
+    {
+      "epoch": 0.26820339965930917,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.7358571428571428e-05,
+      "loss": 0.3574,
+      "step": 1850
+    },
+    {
+      "epoch": 0.26965314776557575,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7344285714285715e-05,
+      "loss": 0.372,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2711028958718423,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.7330000000000002e-05,
+      "loss": 0.3802,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2725526439781088,
+      "grad_norm": 2.25,
+      "learning_rate": 1.7315714285714285e-05,
+      "loss": 0.3818,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2740023920843753,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.7301428571428572e-05,
+      "loss": 0.3603,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2754521401906419,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.728714285714286e-05,
+      "loss": 0.3635,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2769018882969084,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.7272857142857146e-05,
+      "loss": 0.3677,
+      "step": 1910
+    },
+    {
+      "epoch": 0.27835163640317495,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.725857142857143e-05,
+      "loss": 0.3388,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2798013845094415,
+      "grad_norm": 1.875,
+      "learning_rate": 1.7244285714285716e-05,
+      "loss": 0.3538,
+      "step": 1930
+    },
+    {
+      "epoch": 0.281251132615708,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.7230000000000003e-05,
+      "loss": 0.3496,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2827008807219746,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7215714285714286e-05,
+      "loss": 0.3487,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2841506288282411,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.7201428571428573e-05,
+      "loss": 0.3574,
+      "step": 1960
+    },
+    {
+      "epoch": 0.2856003769345076,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.7187142857142856e-05,
+      "loss": 0.3586,
+      "step": 1970
+    },
+    {
+      "epoch": 0.28705012504077415,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.7172857142857143e-05,
+      "loss": 0.351,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2884998731470407,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.715857142857143e-05,
+      "loss": 0.3543,
+      "step": 1990
+    },
+    {
+      "epoch": 0.28994962125330725,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.7144285714285717e-05,
+      "loss": 0.361,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2913993693595738,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7130000000000004e-05,
+      "loss": 0.3361,
+      "step": 2010
+    },
+    {
+      "epoch": 0.2928491174658403,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.7115714285714287e-05,
+      "loss": 0.3812,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2942988655721068,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.7101428571428574e-05,
+      "loss": 0.368,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2957486136783734,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.7087142857142857e-05,
+      "loss": 0.3782,
+      "step": 2040
+    },
+    {
+      "epoch": 0.29719836178463993,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.7072857142857144e-05,
+      "loss": 0.3515,
+      "step": 2050
+    },
+    {
+      "epoch": 0.29864810989090645,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.705857142857143e-05,
+      "loss": 0.3584,
+      "step": 2060
+    },
+    {
+      "epoch": 0.300097857997173,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.7044285714285714e-05,
+      "loss": 0.3164,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3015476061034395,
+      "grad_norm": 2.375,
+      "learning_rate": 1.703e-05,
+      "loss": 0.3491,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3029973542097061,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.7015714285714288e-05,
+      "loss": 0.3395,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3044471023159726,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.700142857142857e-05,
+      "loss": 0.3581,
+      "step": 2100
+    },
+    {
+      "epoch": 0.30589685042223913,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.698714285714286e-05,
+      "loss": 0.3542,
+      "step": 2110
+    },
+    {
+      "epoch": 0.30734659852850565,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.6972857142857145e-05,
+      "loss": 0.3562,
+      "step": 2120
+    },
+    {
+      "epoch": 0.30879634663477223,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6958571428571432e-05,
+      "loss": 0.3519,
+      "step": 2130
+    },
+    {
+      "epoch": 0.31024609474103876,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.6944285714285715e-05,
+      "loss": 0.3325,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3116958428473053,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.6930000000000002e-05,
+      "loss": 0.3392,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3131455909535718,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6915714285714286e-05,
+      "loss": 0.3659,
+      "step": 2160
+    },
+    {
+      "epoch": 0.31459533905983833,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.6901428571428572e-05,
+      "loss": 0.327,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3160450871661049,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.6887142857142856e-05,
+      "loss": 0.3399,
+      "step": 2180
+    },
+    {
+      "epoch": 0.31749483527237143,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6872857142857143e-05,
+      "loss": 0.3356,
+      "step": 2190
+    },
+    {
+      "epoch": 0.31894458337863796,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.685857142857143e-05,
+      "loss": 0.3242,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3203943314849045,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.6844285714285716e-05,
+      "loss": 0.3531,
+      "step": 2210
+    },
+    {
+      "epoch": 0.32184407959117106,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.6830000000000003e-05,
+      "loss": 0.3553,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3232938276974376,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6815714285714287e-05,
+      "loss": 0.3118,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3247435758037041,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.6801428571428573e-05,
+      "loss": 0.3222,
+      "step": 2240
+    },
+    {
+      "epoch": 0.32619332390997063,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.6787142857142857e-05,
+      "loss": 0.3336,
+      "step": 2250
+    },
+    {
+      "epoch": 0.32764307201623716,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.6772857142857144e-05,
+      "loss": 0.3117,
+      "step": 2260
+    },
+    {
+      "epoch": 0.32909282012250374,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.675857142857143e-05,
+      "loss": 0.3404,
+      "step": 2270
+    },
+    {
+      "epoch": 0.33054256822877026,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.6744285714285714e-05,
+      "loss": 0.3373,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3319923163350368,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.673e-05,
+      "loss": 0.3411,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3334420644413033,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.6715714285714288e-05,
+      "loss": 0.3402,
+      "step": 2300
+    },
+    {
+      "epoch": 0.33489181254756983,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6701428571428574e-05,
+      "loss": 0.3405,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3363415606538364,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.668714285714286e-05,
+      "loss": 0.3274,
+      "step": 2320
+    },
+    {
+      "epoch": 0.33779130876010294,
+      "grad_norm": 2.625,
+      "learning_rate": 1.6672857142857145e-05,
+      "loss": 0.352,
+      "step": 2330
+    },
+    {
+      "epoch": 0.33924105686636946,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.665857142857143e-05,
+      "loss": 0.311,
+      "step": 2340
+    },
+    {
+      "epoch": 0.340690804972636,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.6644285714285715e-05,
+      "loss": 0.3359,
+      "step": 2350
+    },
+    {
+      "epoch": 0.34214055307890257,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.6630000000000002e-05,
+      "loss": 0.3263,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3435903011851691,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.6615714285714285e-05,
+      "loss": 0.3077,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3450400492914356,
+      "grad_norm": 2.0,
+      "learning_rate": 1.6601428571428572e-05,
+      "loss": 0.3289,
+      "step": 2380
+    },
+    {
+      "epoch": 0.34648979739770214,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.658714285714286e-05,
+      "loss": 0.3385,
+      "step": 2390
+    },
+    {
+      "epoch": 0.34793954550396866,
+      "grad_norm": 2.375,
+      "learning_rate": 1.6572857142857142e-05,
+      "loss": 0.3184,
+      "step": 2400
+    },
+    {
+      "epoch": 0.34938929361023524,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.655857142857143e-05,
+      "loss": 0.3194,
+      "step": 2410
+    },
+    {
+      "epoch": 0.35083904171650176,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.6544285714285716e-05,
+      "loss": 0.3162,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3522887898227683,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.6530000000000003e-05,
+      "loss": 0.3462,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3537385379290348,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.6515714285714286e-05,
+      "loss": 0.3053,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3551882860353014,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.6501428571428573e-05,
+      "loss": 0.3382,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3566380341415679,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.648714285714286e-05,
+      "loss": 0.3137,
+      "step": 2460
+    },
+    {
+      "epoch": 0.35808778224783444,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.6472857142857143e-05,
+      "loss": 0.341,
+      "step": 2470
+    },
+    {
+      "epoch": 0.35953753035410096,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.645857142857143e-05,
+      "loss": 0.3196,
+      "step": 2480
+    },
+    {
+      "epoch": 0.3609872784603675,
+      "grad_norm": 2.5,
+      "learning_rate": 1.6444285714285713e-05,
+      "loss": 0.3201,
+      "step": 2490
+    },
+    {
+      "epoch": 0.36243702656663407,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.643e-05,
+      "loss": 0.3077,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3638867746729006,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.6415714285714287e-05,
+      "loss": 0.3095,
+      "step": 2510
+    },
+    {
+      "epoch": 0.3653365227791671,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.6401428571428574e-05,
+      "loss": 0.3421,
+      "step": 2520
+    },
+    {
+      "epoch": 0.36678627088543364,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.638714285714286e-05,
+      "loss": 0.3385,
+      "step": 2530
+    },
+    {
+      "epoch": 0.36823601899170016,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.6372857142857144e-05,
+      "loss": 0.3242,
+      "step": 2540
+    },
+    {
+      "epoch": 0.36968576709796674,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.635857142857143e-05,
+      "loss": 0.3367,
+      "step": 2550
+    },
+    {
+      "epoch": 0.37113551520423327,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.6344285714285714e-05,
+      "loss": 0.3121,
+      "step": 2560
+    },
+    {
+      "epoch": 0.3725852633104998,
+      "grad_norm": 2.25,
+      "learning_rate": 1.633e-05,
+      "loss": 0.3084,
+      "step": 2570
+    },
+    {
+      "epoch": 0.3740350114167663,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.6315714285714288e-05,
+      "loss": 0.2978,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3754847595230329,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.630142857142857e-05,
+      "loss": 0.2917,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3769345076292994,
+      "grad_norm": 2.125,
+      "learning_rate": 1.6287142857142858e-05,
+      "loss": 0.303,
+      "step": 2600
+    },
+    {
+      "epoch": 0.37838425573556594,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.6272857142857145e-05,
+      "loss": 0.3118,
+      "step": 2610
+    },
+    {
+      "epoch": 0.37983400384183247,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.6258571428571432e-05,
+      "loss": 0.3313,
+      "step": 2620
+    },
+    {
+      "epoch": 0.381283751948099,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.6244285714285715e-05,
+      "loss": 0.3187,
+      "step": 2630
+    },
+    {
+      "epoch": 0.38273350005436557,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.6230000000000002e-05,
+      "loss": 0.303,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3841832481606321,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.621571428571429e-05,
+      "loss": 0.3225,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3856329962668986,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.6201428571428572e-05,
+      "loss": 0.2918,
+      "step": 2660
+    },
+    {
+      "epoch": 0.38708274437316514,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.618714285714286e-05,
+      "loss": 0.331,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3885324924794317,
+      "grad_norm": 2.375,
+      "learning_rate": 1.6172857142857143e-05,
+      "loss": 0.3187,
+      "step": 2680
+    },
+    {
+      "epoch": 0.38998224058569825,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.615857142857143e-05,
+      "loss": 0.3219,
+      "step": 2690
+    },
+    {
+      "epoch": 0.39143198869196477,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.6144285714285713e-05,
+      "loss": 0.3005,
+      "step": 2700
+    },
+    {
+      "epoch": 0.3928817367982313,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.613e-05,
+      "loss": 0.3147,
+      "step": 2710
+    },
+    {
+      "epoch": 0.3943314849044978,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.6115714285714287e-05,
+      "loss": 0.3102,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3957812330107644,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.6101428571428573e-05,
+      "loss": 0.3051,
+      "step": 2730
+    },
+    {
+      "epoch": 0.3972309811170309,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.608714285714286e-05,
+      "loss": 0.314,
+      "step": 2740
+    },
+    {
+      "epoch": 0.39868072922329745,
+      "grad_norm": 2.5,
+      "learning_rate": 1.6072857142857144e-05,
+      "loss": 0.3123,
+      "step": 2750
+    },
+    {
+      "epoch": 0.40013047732956397,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.605857142857143e-05,
+      "loss": 0.3068,
+      "step": 2760
+    },
+    {
+      "epoch": 0.40158022543583055,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.6044285714285717e-05,
+      "loss": 0.3041,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4030299735420971,
+      "grad_norm": 2.0,
+      "learning_rate": 1.603e-05,
+      "loss": 0.3288,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4044797216483636,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.6015714285714288e-05,
+      "loss": 0.3056,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4059294697546301,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.600142857142857e-05,
+      "loss": 0.296,
+      "step": 2800
+    },
+    {
+      "epoch": 0.40737921786089665,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5987142857142858e-05,
+      "loss": 0.2872,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4088289659671632,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5972857142857145e-05,
+      "loss": 0.3164,
+      "step": 2820
+    },
+    {
+      "epoch": 0.41027871407342975,
+      "grad_norm": 2.25,
+      "learning_rate": 1.595857142857143e-05,
+      "loss": 0.2898,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4117284621796963,
+      "grad_norm": 2.0,
+      "learning_rate": 1.5944285714285718e-05,
+      "loss": 0.3266,
+      "step": 2840
+    },
+    {
+      "epoch": 0.4131782102859628,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.593e-05,
+      "loss": 0.3057,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4146279583922293,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.591571428571429e-05,
+      "loss": 0.2856,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4160777064984959,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.5901428571428572e-05,
+      "loss": 0.32,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4175274546047624,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.588714285714286e-05,
+      "loss": 0.3132,
+      "step": 2880
+    },
+    {
+      "epoch": 0.41897720271102895,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5872857142857142e-05,
+      "loss": 0.3149,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4204269508172955,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.585857142857143e-05,
+      "loss": 0.2923,
+      "step": 2900
+    },
+    {
+      "epoch": 0.42187669892356205,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.5844285714285716e-05,
+      "loss": 0.3025,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4233264470298286,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5830000000000003e-05,
+      "loss": 0.3102,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4247761951360951,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5815714285714286e-05,
+      "loss": 0.3039,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4262259432423616,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.5801428571428573e-05,
+      "loss": 0.321,
+      "step": 2940
+    },
+    {
+      "epoch": 0.42767569134862815,
+      "grad_norm": 1.875,
+      "learning_rate": 1.578714285714286e-05,
+      "loss": 0.3115,
+      "step": 2950
+    },
+    {
+      "epoch": 0.42912543945489473,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.5772857142857143e-05,
+      "loss": 0.2833,
+      "step": 2960
+    },
+    {
+      "epoch": 0.43057518756116125,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.575857142857143e-05,
+      "loss": 0.3012,
+      "step": 2970
+    },
+    {
+      "epoch": 0.4320249356674278,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.5744285714285717e-05,
+      "loss": 0.2979,
+      "step": 2980
+    },
+    {
+      "epoch": 0.4334746837736943,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.573e-05,
+      "loss": 0.2834,
+      "step": 2990
+    },
+    {
+      "epoch": 0.4349244318799609,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.5715714285714287e-05,
+      "loss": 0.2936,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4363741799862274,
+      "grad_norm": 2.375,
+      "learning_rate": 1.570142857142857e-05,
+      "loss": 0.3027,
+      "step": 3010
+    },
+    {
+      "epoch": 0.43782392809249393,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5687142857142857e-05,
+      "loss": 0.2792,
+      "step": 3020
+    },
+    {
+      "epoch": 0.43927367619876045,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.5672857142857144e-05,
+      "loss": 0.2759,
+      "step": 3030
+    },
+    {
+      "epoch": 0.440723424305027,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.565857142857143e-05,
+      "loss": 0.2947,
+      "step": 3040
+    },
+    {
+      "epoch": 0.44217317241129356,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.5644285714285718e-05,
+      "loss": 0.2883,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4436229205175601,
+      "grad_norm": 2.25,
+      "learning_rate": 1.563e-05,
+      "loss": 0.2886,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4450726686238266,
+      "grad_norm": 2.25,
+      "learning_rate": 1.5615714285714288e-05,
+      "loss": 0.3036,
+      "step": 3070
+    },
+    {
+      "epoch": 0.44652241673009313,
+      "grad_norm": 1.875,
+      "learning_rate": 1.560142857142857e-05,
+      "loss": 0.3059,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4479721648363597,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5587142857142858e-05,
+      "loss": 0.3031,
+      "step": 3090
+    },
+    {
+      "epoch": 0.44942191294262623,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.5572857142857145e-05,
+      "loss": 0.315,
+      "step": 3100
+    },
+    {
+      "epoch": 0.45087166104889276,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.555857142857143e-05,
+      "loss": 0.2855,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4523214091551593,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.5544285714285715e-05,
+      "loss": 0.2836,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4537711572614258,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.5530000000000002e-05,
+      "loss": 0.2882,
+      "step": 3130
+    },
+    {
+      "epoch": 0.4552209053676924,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.551571428571429e-05,
+      "loss": 0.299,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4566706534739589,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.5501428571428572e-05,
+      "loss": 0.2849,
+      "step": 3150
+    },
+    {
+      "epoch": 0.45812040158022543,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.548714285714286e-05,
+      "loss": 0.277,
+      "step": 3160
+    },
+    {
+      "epoch": 0.45957014968649196,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.5472857142857146e-05,
+      "loss": 0.2912,
+      "step": 3170
+    },
+    {
+      "epoch": 0.4610198977927585,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.545857142857143e-05,
+      "loss": 0.2952,
+      "step": 3180
+    },
+    {
+      "epoch": 0.46246964589902506,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.5444285714285716e-05,
+      "loss": 0.2902,
+      "step": 3190
+    },
+    {
+      "epoch": 0.4639193940052916,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.543e-05,
+      "loss": 0.2905,
+      "step": 3200
+    },
+    {
+      "epoch": 0.4653691421115581,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.5415714285714286e-05,
+      "loss": 0.3043,
+      "step": 3210
+    },
+    {
+      "epoch": 0.46681889021782463,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.5401428571428573e-05,
+      "loss": 0.2888,
+      "step": 3220
+    },
+    {
+      "epoch": 0.4682686383240912,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.5387142857142857e-05,
+      "loss": 0.2735,
+      "step": 3230
+    },
+    {
+      "epoch": 0.46971838643035774,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.5372857142857144e-05,
+      "loss": 0.2663,
+      "step": 3240
+    },
+    {
+      "epoch": 0.47116813453662426,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.535857142857143e-05,
+      "loss": 0.2902,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4726178826428908,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5344285714285717e-05,
+      "loss": 0.2794,
+      "step": 3260
+    },
+    {
+      "epoch": 0.4740676307491573,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.533e-05,
+      "loss": 0.299,
+      "step": 3270
+    },
+    {
+      "epoch": 0.4755173788554239,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.5315714285714287e-05,
+      "loss": 0.2939,
+      "step": 3280
+    },
+    {
+      "epoch": 0.4769671269616904,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.5301428571428574e-05,
+      "loss": 0.2835,
+      "step": 3290
+    },
+    {
+      "epoch": 0.47841687506795694,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.5287142857142858e-05,
+      "loss": 0.3007,
+      "step": 3300
+    },
+    {
+      "epoch": 0.47986662317422346,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.5272857142857144e-05,
+      "loss": 0.2717,
+      "step": 3310
+    },
+    {
+      "epoch": 0.48131637128049004,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.525857142857143e-05,
+      "loss": 0.2917,
+      "step": 3320
+    },
+    {
+      "epoch": 0.48276611938675656,
+      "grad_norm": 2.75,
+      "learning_rate": 1.5244285714285716e-05,
+      "loss": 0.2864,
+      "step": 3330
+    },
+    {
+      "epoch": 0.4842158674930231,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.523e-05,
+      "loss": 0.2842,
+      "step": 3340
+    },
+    {
+      "epoch": 0.4856656155992896,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.5215714285714287e-05,
+      "loss": 0.2899,
+      "step": 3350
+    },
+    {
+      "epoch": 0.48711536370555614,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.5201428571428574e-05,
+      "loss": 0.2687,
+      "step": 3360
+    },
+    {
+      "epoch": 0.4885651118118227,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.5187142857142859e-05,
+      "loss": 0.3019,
+      "step": 3370
+    },
+    {
+      "epoch": 0.49001485991808924,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.5172857142857145e-05,
+      "loss": 0.2697,
+      "step": 3380
+    },
+    {
+      "epoch": 0.49146460802435576,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.5158571428571429e-05,
+      "loss": 0.2888,
+      "step": 3390
+    },
+    {
+      "epoch": 0.4929143561306223,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.5144285714285716e-05,
+      "loss": 0.2858,
+      "step": 3400
+    },
+    {
+      "epoch": 0.49436410423688887,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.513e-05,
+      "loss": 0.28,
+      "step": 3410
+    },
+    {
+      "epoch": 0.4958138523431554,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5115714285714288e-05,
+      "loss": 0.2578,
+      "step": 3420
+    },
+    {
+      "epoch": 0.4972636004494219,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5101428571428573e-05,
+      "loss": 0.2728,
+      "step": 3430
+    },
+    {
+      "epoch": 0.49871334855568844,
+      "grad_norm": 2.125,
+      "learning_rate": 1.5087142857142858e-05,
+      "loss": 0.2803,
+      "step": 3440
+    },
+    {
+      "epoch": 0.500163096661955,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.5072857142857145e-05,
+      "loss": 0.2832,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5016128447682215,
+      "grad_norm": 2.0,
+      "learning_rate": 1.505857142857143e-05,
+      "loss": 0.2721,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5030625928744881,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.5044285714285715e-05,
+      "loss": 0.2835,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5045123409807546,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.503e-05,
+      "loss": 0.2819,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5059620890870211,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.5015714285714287e-05,
+      "loss": 0.2686,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5074118371932876,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.5001428571428574e-05,
+      "loss": 0.251,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5088615852995542,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.4987142857142857e-05,
+      "loss": 0.2782,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5103113334058207,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.4972857142857144e-05,
+      "loss": 0.2734,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5117610815120873,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.4958571428571429e-05,
+      "loss": 0.2836,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5132108296183538,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.4944285714285716e-05,
+      "loss": 0.2934,
+      "step": 3540
+    },
+    {
+      "epoch": 0.5146605777246204,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.4930000000000003e-05,
+      "loss": 0.2827,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5161103258308869,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4915714285714286e-05,
+      "loss": 0.2682,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5175600739371534,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.4901428571428573e-05,
+      "loss": 0.2787,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5190098220434199,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4887142857142858e-05,
+      "loss": 0.2711,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5204595701496865,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4872857142857145e-05,
+      "loss": 0.2658,
+      "step": 3590
+    },
+    {
+      "epoch": 0.521909318255953,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4858571428571428e-05,
+      "loss": 0.2826,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5233590663622195,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.4844285714285715e-05,
+      "loss": 0.2708,
+      "step": 3610
+    },
+    {
+      "epoch": 0.5248088144684862,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.4830000000000002e-05,
+      "loss": 0.2821,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5262585625747527,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.4815714285714287e-05,
+      "loss": 0.2758,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5277083106810192,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.4801428571428574e-05,
+      "loss": 0.2743,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5291580587872857,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4787142857142857e-05,
+      "loss": 0.2716,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5306078068935522,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.4772857142857144e-05,
+      "loss": 0.2688,
+      "step": 3660
+    },
+    {
+      "epoch": 0.5320575549998188,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.475857142857143e-05,
+      "loss": 0.2865,
+      "step": 3670
+    },
+    {
+      "epoch": 0.5335073031060853,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4744285714285716e-05,
+      "loss": 0.2751,
+      "step": 3680
+    },
+    {
+      "epoch": 0.5349570512123518,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.4730000000000001e-05,
+      "loss": 0.2789,
+      "step": 3690
+    },
+    {
+      "epoch": 0.5364067993186183,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.4715714285714286e-05,
+      "loss": 0.277,
+      "step": 3700
+    },
+    {
+      "epoch": 0.537856547424885,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.4701428571428573e-05,
+      "loss": 0.2576,
+      "step": 3710
+    },
+    {
+      "epoch": 0.5393062955311515,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.4687142857142858e-05,
+      "loss": 0.303,
+      "step": 3720
+    },
+    {
+      "epoch": 0.540756043637418,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4672857142857143e-05,
+      "loss": 0.2587,
+      "step": 3730
+    },
+    {
+      "epoch": 0.5422057917436846,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.4658571428571429e-05,
+      "loss": 0.2569,
+      "step": 3740
+    },
+    {
+      "epoch": 0.5436555398499511,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4644285714285715e-05,
+      "loss": 0.2586,
+      "step": 3750
+    },
+    {
+      "epoch": 0.5451052879562176,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.4630000000000002e-05,
+      "loss": 0.2802,
+      "step": 3760
+    },
+    {
+      "epoch": 0.5465550360624841,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.4615714285714286e-05,
+      "loss": 0.2653,
+      "step": 3770
+    },
+    {
+      "epoch": 0.5480047841687506,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.4601428571428572e-05,
+      "loss": 0.2735,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5494545322750172,
+      "grad_norm": 2.375,
+      "learning_rate": 1.4587142857142858e-05,
+      "loss": 0.2717,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5509042803812838,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.4572857142857144e-05,
+      "loss": 0.2718,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5523540284875503,
+      "grad_norm": 2.5,
+      "learning_rate": 1.4558571428571431e-05,
+      "loss": 0.2712,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5538037765938169,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.4544285714285715e-05,
+      "loss": 0.2479,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5552535247000834,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.4530000000000001e-05,
+      "loss": 0.2633,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5567032728063499,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4515714285714287e-05,
+      "loss": 0.259,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5581530209126164,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.4501428571428573e-05,
+      "loss": 0.2672,
+      "step": 3850
+    },
+    {
+      "epoch": 0.559602769018883,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4487142857142857e-05,
+      "loss": 0.2474,
+      "step": 3860
+    },
+    {
+      "epoch": 0.5610525171251495,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.4472857142857144e-05,
+      "loss": 0.2642,
+      "step": 3870
+    },
+    {
+      "epoch": 0.562502265231416,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.445857142857143e-05,
+      "loss": 0.2695,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5639520133376825,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4444285714285716e-05,
+      "loss": 0.2683,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5654017614439492,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4430000000000002e-05,
+      "loss": 0.2835,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5668515095502157,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.4415714285714286e-05,
+      "loss": 0.2713,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5683012576564822,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.4401428571428573e-05,
+      "loss": 0.2733,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5697510057627487,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.4387142857142858e-05,
+      "loss": 0.2761,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5712007538690153,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4372857142857145e-05,
+      "loss": 0.2791,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5726505019752818,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.435857142857143e-05,
+      "loss": 0.2525,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5741002500815483,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.4344285714285715e-05,
+      "loss": 0.2772,
+      "step": 3960
+    },
+    {
+      "epoch": 0.5755499981878148,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.4330000000000002e-05,
+      "loss": 0.2647,
+      "step": 3970
+    },
+    {
+      "epoch": 0.5769997462940814,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.4315714285714287e-05,
+      "loss": 0.2661,
+      "step": 3980
+    },
+    {
+      "epoch": 0.578449494400348,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4301428571428572e-05,
+      "loss": 0.2564,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5798992425066145,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4287142857142857e-05,
+      "loss": 0.282,
+      "step": 4000
+    },
+    {
+      "epoch": 0.581348990612881,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4272857142857144e-05,
+      "loss": 0.2573,
+      "step": 4010
+    },
+    {
+      "epoch": 0.5827987387191476,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.425857142857143e-05,
+      "loss": 0.2453,
+      "step": 4020
+    },
+    {
+      "epoch": 0.5842484868254141,
+      "grad_norm": 2.125,
+      "learning_rate": 1.4244285714285714e-05,
+      "loss": 0.2638,
+      "step": 4030
+    },
+    {
+      "epoch": 0.5856982349316806,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.4230000000000001e-05,
+      "loss": 0.2532,
+      "step": 4040
+    },
+    {
+      "epoch": 0.5871479830379471,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4215714285714286e-05,
+      "loss": 0.2663,
+      "step": 4050
+    },
+    {
+      "epoch": 0.5885977311442137,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4201428571428573e-05,
+      "loss": 0.2553,
+      "step": 4060
+    },
+    {
+      "epoch": 0.5900474792504802,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.418714285714286e-05,
+      "loss": 0.2657,
+      "step": 4070
+    },
+    {
+      "epoch": 0.5914972273567468,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.4172857142857143e-05,
+      "loss": 0.2366,
+      "step": 4080
+    },
+    {
+      "epoch": 0.5929469754630133,
+      "grad_norm": 2.5,
+      "learning_rate": 1.415857142857143e-05,
+      "loss": 0.2643,
+      "step": 4090
+    },
+    {
+      "epoch": 0.5943967235692799,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.4144285714285715e-05,
+      "loss": 0.2516,
+      "step": 4100
+    },
+    {
+      "epoch": 0.5958464716755464,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4130000000000002e-05,
+      "loss": 0.2542,
+      "step": 4110
+    },
+    {
+      "epoch": 0.5972962197818129,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4115714285714285e-05,
+      "loss": 0.2525,
+      "step": 4120
+    },
+    {
+      "epoch": 0.5987459678880794,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4101428571428572e-05,
+      "loss": 0.2588,
+      "step": 4130
+    },
+    {
+      "epoch": 0.600195715994346,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4087142857142859e-05,
+      "loss": 0.2559,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6016454641006125,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4072857142857144e-05,
+      "loss": 0.2766,
+      "step": 4150
+    },
+    {
+      "epoch": 0.603095212206879,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.4058571428571431e-05,
+      "loss": 0.2717,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6045449603131456,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.4044285714285714e-05,
+      "loss": 0.2735,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6059947084194122,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.4030000000000001e-05,
+      "loss": 0.2671,
+      "step": 4180
+    },
+    {
+      "epoch": 0.6074444565256787,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.4015714285714286e-05,
+      "loss": 0.2483,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6088942046319452,
+      "grad_norm": 2.0,
+      "learning_rate": 1.4001428571428573e-05,
+      "loss": 0.259,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6103439527382117,
+      "grad_norm": 2.5,
+      "learning_rate": 1.398714285714286e-05,
+      "loss": 0.2583,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6117937008444783,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.3972857142857143e-05,
+      "loss": 0.2518,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6132434489507448,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.395857142857143e-05,
+      "loss": 0.2706,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6146931970570113,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.3944285714285715e-05,
+      "loss": 0.2578,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6161429451632778,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.393e-05,
+      "loss": 0.241,
+      "step": 4250
+    },
+    {
+      "epoch": 0.6175926932695445,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.3915714285714287e-05,
+      "loss": 0.2616,
+      "step": 4260
+    },
+    {
+      "epoch": 0.619042441375811,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3901428571428572e-05,
+      "loss": 0.2599,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6204921894820775,
+      "grad_norm": 2.5,
+      "learning_rate": 1.388714285714286e-05,
+      "loss": 0.2738,
+      "step": 4280
+    },
+    {
+      "epoch": 0.621941937588344,
+      "grad_norm": 2.0,
+      "learning_rate": 1.3872857142857143e-05,
+      "loss": 0.2345,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6233916856946106,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.385857142857143e-05,
+      "loss": 0.265,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6248414338008771,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.3844285714285715e-05,
+      "loss": 0.2606,
+      "step": 4310
+    },
+    {
+      "epoch": 0.6262911819071436,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.3830000000000001e-05,
+      "loss": 0.2313,
+      "step": 4320
+    },
+    {
+      "epoch": 0.6277409300134101,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.3815714285714288e-05,
+      "loss": 0.2438,
+      "step": 4330
+    },
+    {
+      "epoch": 0.6291906781196767,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3801428571428572e-05,
+      "loss": 0.2429,
+      "step": 4340
+    },
+    {
+      "epoch": 0.6306404262259433,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.3787142857142858e-05,
+      "loss": 0.2514,
+      "step": 4350
+    },
+    {
+      "epoch": 0.6320901743322098,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3772857142857144e-05,
+      "loss": 0.2519,
+      "step": 4360
+    },
+    {
+      "epoch": 0.6335399224384763,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.375857142857143e-05,
+      "loss": 0.2508,
+      "step": 4370
+    },
+    {
+      "epoch": 0.6349896705447429,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.3744285714285714e-05,
+      "loss": 0.2836,
+      "step": 4380
+    },
+    {
+      "epoch": 0.6364394186510094,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.373e-05,
+      "loss": 0.2424,
+      "step": 4390
+    },
+    {
+      "epoch": 0.6378891667572759,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3715714285714288e-05,
+      "loss": 0.2614,
+      "step": 4400
+    },
+    {
+      "epoch": 0.6393389148635424,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3701428571428573e-05,
+      "loss": 0.2678,
+      "step": 4410
+    },
+    {
+      "epoch": 0.640788662969809,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.368714285714286e-05,
+      "loss": 0.2631,
+      "step": 4420
+    },
+    {
+      "epoch": 0.6422384110760755,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3672857142857143e-05,
+      "loss": 0.252,
+      "step": 4430
+    },
+    {
+      "epoch": 0.6436881591823421,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.365857142857143e-05,
+      "loss": 0.2636,
+      "step": 4440
+    },
+    {
+      "epoch": 0.6451379072886086,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3644285714285715e-05,
+      "loss": 0.2577,
+      "step": 4450
+    },
+    {
+      "epoch": 0.6465876553948752,
+      "grad_norm": 2.25,
+      "learning_rate": 1.3630000000000002e-05,
+      "loss": 0.2728,
+      "step": 4460
+    },
+    {
+      "epoch": 0.6480374035011417,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3615714285714288e-05,
+      "loss": 0.2503,
+      "step": 4470
+    },
+    {
+      "epoch": 0.6494871516074082,
+      "grad_norm": 2.0,
+      "learning_rate": 1.3601428571428572e-05,
+      "loss": 0.2608,
+      "step": 4480
+    },
+    {
+      "epoch": 0.6509368997136747,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3587142857142859e-05,
+      "loss": 0.2579,
+      "step": 4490
+    },
+    {
+      "epoch": 0.6523866478199413,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3572857142857144e-05,
+      "loss": 0.2589,
+      "step": 4500
+    },
+    {
+      "epoch": 0.6538363959262078,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.355857142857143e-05,
+      "loss": 0.2733,
+      "step": 4510
+    },
+    {
+      "epoch": 0.6552861440324743,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.3544285714285716e-05,
+      "loss": 0.2622,
+      "step": 4520
+    },
+    {
+      "epoch": 0.6567358921387408,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3530000000000001e-05,
+      "loss": 0.2731,
+      "step": 4530
+    },
+    {
+      "epoch": 0.6581856402450075,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3515714285714288e-05,
+      "loss": 0.2487,
+      "step": 4540
+    },
+    {
+      "epoch": 0.659635388351274,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.3501428571428571e-05,
+      "loss": 0.2474,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6610851364575405,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.3487142857142858e-05,
+      "loss": 0.2536,
+      "step": 4560
+    },
+    {
+      "epoch": 0.662534884563807,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.3472857142857143e-05,
+      "loss": 0.2512,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6639846326700736,
+      "grad_norm": 1.875,
+      "learning_rate": 1.345857142857143e-05,
+      "loss": 0.246,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6654343807763401,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.3444285714285717e-05,
+      "loss": 0.236,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6668841288826066,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.343e-05,
+      "loss": 0.2535,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6683338769888731,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3415714285714287e-05,
+      "loss": 0.2511,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6697836250951397,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3401428571428572e-05,
+      "loss": 0.2386,
+      "step": 4620
+    },
+    {
+      "epoch": 0.6712333732014063,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.3387142857142859e-05,
+      "loss": 0.2424,
+      "step": 4630
+    },
+    {
+      "epoch": 0.6726831213076728,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.3372857142857142e-05,
+      "loss": 0.2433,
+      "step": 4640
+    },
+    {
+      "epoch": 0.6741328694139394,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.335857142857143e-05,
+      "loss": 0.2464,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6755826175202059,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.3344285714285716e-05,
+      "loss": 0.2581,
+      "step": 4660
+    },
+    {
+      "epoch": 0.6770323656264724,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3330000000000001e-05,
+      "loss": 0.2482,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6784821137327389,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.3315714285714288e-05,
+      "loss": 0.2425,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6799318618390054,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.3301428571428571e-05,
+      "loss": 0.2454,
+      "step": 4690
+    },
+    {
+      "epoch": 0.681381609945272,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.3287142857142858e-05,
+      "loss": 0.2524,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6828313580515385,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3272857142857143e-05,
+      "loss": 0.262,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6842811061578051,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.325857142857143e-05,
+      "loss": 0.2397,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6857308542640717,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3244285714285717e-05,
+      "loss": 0.2548,
+      "step": 4730
+    },
+    {
+      "epoch": 0.6871806023703382,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.323e-05,
+      "loss": 0.2472,
+      "step": 4740
+    },
+    {
+      "epoch": 0.6886303504766047,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.3215714285714287e-05,
+      "loss": 0.2483,
+      "step": 4750
+    },
+    {
+      "epoch": 0.6900800985828712,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.3201428571428572e-05,
+      "loss": 0.2646,
+      "step": 4760
+    },
+    {
+      "epoch": 0.6915298466891378,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.3187142857142859e-05,
+      "loss": 0.2526,
+      "step": 4770
+    },
+    {
+      "epoch": 0.6929795947954043,
+      "grad_norm": 2.375,
+      "learning_rate": 1.3172857142857144e-05,
+      "loss": 0.2601,
+      "step": 4780
+    },
+    {
+      "epoch": 0.6944293429016708,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.315857142857143e-05,
+      "loss": 0.2511,
+      "step": 4790
+    },
+    {
+      "epoch": 0.6958790910079373,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.3144285714285716e-05,
+      "loss": 0.2519,
+      "step": 4800
+    },
+    {
+      "epoch": 0.697328839114204,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.3130000000000001e-05,
+      "loss": 0.2613,
+      "step": 4810
+    },
+    {
+      "epoch": 0.6987785872204705,
+      "grad_norm": 2.5,
+      "learning_rate": 1.3115714285714286e-05,
+      "loss": 0.2627,
+      "step": 4820
+    },
+    {
+      "epoch": 0.700228335326737,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3101428571428572e-05,
+      "loss": 0.2426,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7016780834330035,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3087142857142858e-05,
+      "loss": 0.2546,
+      "step": 4840
+    },
+    {
+      "epoch": 0.70312783153927,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.3072857142857145e-05,
+      "loss": 0.242,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7045775796455366,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3058571428571429e-05,
+      "loss": 0.2535,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7060273277518031,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.3044285714285715e-05,
+      "loss": 0.2435,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7074770758580696,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.303e-05,
+      "loss": 0.2729,
+      "step": 4880
+    },
+    {
+      "epoch": 0.7089268239643362,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.3015714285714287e-05,
+      "loss": 0.24,
+      "step": 4890
+    },
+    {
+      "epoch": 0.7103765720706028,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3001428571428571e-05,
+      "loss": 0.251,
+      "step": 4900
+    },
+    {
+      "epoch": 0.7118263201768693,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2987142857142858e-05,
+      "loss": 0.2587,
+      "step": 4910
+    },
+    {
+      "epoch": 0.7132760682831358,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2972857142857144e-05,
+      "loss": 0.2539,
+      "step": 4920
+    },
+    {
+      "epoch": 0.7147258163894024,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.295857142857143e-05,
+      "loss": 0.2347,
+      "step": 4930
+    },
+    {
+      "epoch": 0.7161755644956689,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2944285714285716e-05,
+      "loss": 0.2503,
+      "step": 4940
+    },
+    {
+      "epoch": 0.7176253126019354,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.293e-05,
+      "loss": 0.2427,
+      "step": 4950
+    },
+    {
+      "epoch": 0.7190750607082019,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.2915714285714287e-05,
+      "loss": 0.2596,
+      "step": 4960
+    },
+    {
+      "epoch": 0.7205248088144685,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.2901428571428572e-05,
+      "loss": 0.2199,
+      "step": 4970
+    },
+    {
+      "epoch": 0.721974556920735,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.2887142857142859e-05,
+      "loss": 0.2315,
+      "step": 4980
+    },
+    {
+      "epoch": 0.7234243050270016,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.2872857142857145e-05,
+      "loss": 0.2456,
+      "step": 4990
+    },
+    {
+      "epoch": 0.7248740531332681,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.2858571428571429e-05,
+      "loss": 0.2416,
+      "step": 5000
+    },
+    {
+      "epoch": 0.7263238012395347,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2844285714285716e-05,
+      "loss": 0.2485,
+      "step": 5010
+    },
+    {
+      "epoch": 0.7277735493458012,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.283e-05,
+      "loss": 0.2491,
+      "step": 5020
+    },
+    {
+      "epoch": 0.7292232974520677,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2815714285714288e-05,
+      "loss": 0.2484,
+      "step": 5030
+    },
+    {
+      "epoch": 0.7306730455583342,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2801428571428573e-05,
+      "loss": 0.2442,
+      "step": 5040
+    },
+    {
+      "epoch": 0.7321227936646008,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2787142857142858e-05,
+      "loss": 0.2451,
+      "step": 5050
+    },
+    {
+      "epoch": 0.7335725417708673,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.2772857142857145e-05,
+      "loss": 0.2259,
+      "step": 5060
+    },
+    {
+      "epoch": 0.7350222898771338,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.275857142857143e-05,
+      "loss": 0.2544,
+      "step": 5070
+    },
+    {
+      "epoch": 0.7364720379834003,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.2744285714285715e-05,
+      "loss": 0.2424,
+      "step": 5080
+    },
+    {
+      "epoch": 0.737921786089667,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.273e-05,
+      "loss": 0.2501,
+      "step": 5090
+    },
+    {
+      "epoch": 0.7393715341959335,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.2715714285714287e-05,
+      "loss": 0.2494,
+      "step": 5100
+    },
+    {
+      "epoch": 0.7408212823022,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2701428571428574e-05,
+      "loss": 0.2511,
+      "step": 5110
+    },
+    {
+      "epoch": 0.7422710304084665,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2687142857142857e-05,
+      "loss": 0.2355,
+      "step": 5120
+    },
+    {
+      "epoch": 0.7437207785147331,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2672857142857144e-05,
+      "loss": 0.2419,
+      "step": 5130
+    },
+    {
+      "epoch": 0.7451705266209996,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.2658571428571429e-05,
+      "loss": 0.2248,
+      "step": 5140
+    },
+    {
+      "epoch": 0.7466202747272661,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.2644285714285716e-05,
+      "loss": 0.2349,
+      "step": 5150
+    },
+    {
+      "epoch": 0.7480700228335326,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.263e-05,
+      "loss": 0.2268,
+      "step": 5160
+    },
+    {
+      "epoch": 0.7495197709397992,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.2615714285714286e-05,
+      "loss": 0.2327,
+      "step": 5170
+    },
+    {
+      "epoch": 0.7509695190460658,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2601428571428573e-05,
+      "loss": 0.2688,
+      "step": 5180
+    },
+    {
+      "epoch": 0.7524192671523323,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.2587142857142858e-05,
+      "loss": 0.2685,
+      "step": 5190
+    },
+    {
+      "epoch": 0.7538690152585988,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2572857142857145e-05,
+      "loss": 0.24,
+      "step": 5200
+    },
+    {
+      "epoch": 0.7553187633648654,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.2558571428571428e-05,
+      "loss": 0.24,
+      "step": 5210
+    },
+    {
+      "epoch": 0.7567685114711319,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.2544285714285715e-05,
+      "loss": 0.2236,
+      "step": 5220
+    },
+    {
+      "epoch": 0.7582182595773984,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.253e-05,
+      "loss": 0.2413,
+      "step": 5230
+    },
+    {
+      "epoch": 0.7596680076836649,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2515714285714287e-05,
+      "loss": 0.2329,
+      "step": 5240
+    },
+    {
+      "epoch": 0.7611177557899315,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.2501428571428574e-05,
+      "loss": 0.2191,
+      "step": 5250
+    },
+    {
+      "epoch": 0.762567503896198,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.2487142857142857e-05,
+      "loss": 0.2389,
+      "step": 5260
+    },
+    {
+      "epoch": 0.7640172520024646,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.2472857142857144e-05,
+      "loss": 0.2311,
+      "step": 5270
+    },
+    {
+      "epoch": 0.7654670001087311,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.245857142857143e-05,
+      "loss": 0.2252,
+      "step": 5280
+    },
+    {
+      "epoch": 0.7669167482149977,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2444285714285716e-05,
+      "loss": 0.2366,
+      "step": 5290
+    },
+    {
+      "epoch": 0.7683664963212642,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2430000000000001e-05,
+      "loss": 0.2541,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7698162444275307,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2415714285714286e-05,
+      "loss": 0.2465,
+      "step": 5310
+    },
+    {
+      "epoch": 0.7712659925337972,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2401428571428573e-05,
+      "loss": 0.2374,
+      "step": 5320
+    },
+    {
+      "epoch": 0.7727157406400638,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.2387142857142858e-05,
+      "loss": 0.2395,
+      "step": 5330
+    },
+    {
+      "epoch": 0.7741654887463303,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2372857142857143e-05,
+      "loss": 0.2485,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7756152368525968,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2358571428571429e-05,
+      "loss": 0.2449,
+      "step": 5350
+    },
+    {
+      "epoch": 0.7770649849588634,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.2344285714285715e-05,
+      "loss": 0.235,
+      "step": 5360
+    },
+    {
+      "epoch": 0.77851473306513,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2330000000000002e-05,
+      "loss": 0.2176,
+      "step": 5370
+    },
+    {
+      "epoch": 0.7799644811713965,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.2315714285714286e-05,
+      "loss": 0.2314,
+      "step": 5380
+    },
+    {
+      "epoch": 0.781414229277663,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.2301428571428572e-05,
+      "loss": 0.2392,
+      "step": 5390
+    },
+    {
+      "epoch": 0.7828639773839295,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.2287142857142858e-05,
+      "loss": 0.2534,
+      "step": 5400
+    },
+    {
+      "epoch": 0.7843137254901961,
+      "grad_norm": 2.125,
+      "learning_rate": 1.2272857142857144e-05,
+      "loss": 0.2487,
+      "step": 5410
+    },
+    {
+      "epoch": 0.7857634735964626,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.2258571428571428e-05,
+      "loss": 0.2216,
+      "step": 5420
+    },
+    {
+      "epoch": 0.7872132217027291,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.2244285714285715e-05,
+      "loss": 0.2408,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7886629698089956,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.2230000000000001e-05,
+      "loss": 0.2486,
+      "step": 5440
+    },
+    {
+      "epoch": 0.7901127179152623,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2215714285714287e-05,
+      "loss": 0.2363,
+      "step": 5450
+    },
+    {
+      "epoch": 0.7915624660215288,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.2201428571428573e-05,
+      "loss": 0.226,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7930122141277953,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2187142857142857e-05,
+      "loss": 0.2442,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7944619622340618,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.2172857142857144e-05,
+      "loss": 0.2334,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7959117103403284,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.2158571428571429e-05,
+      "loss": 0.2252,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7973614584465949,
+      "grad_norm": 2.625,
+      "learning_rate": 1.2144285714285716e-05,
+      "loss": 0.2322,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7988112065528614,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.2130000000000002e-05,
+      "loss": 0.2444,
+      "step": 5510
+    },
+    {
+      "epoch": 0.8002609546591279,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2115714285714286e-05,
+      "loss": 0.2459,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8017107027653945,
+      "grad_norm": 2.625,
+      "learning_rate": 1.2101428571428573e-05,
+      "loss": 0.2396,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8031604508716611,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2087142857142858e-05,
+      "loss": 0.2258,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8046101989779276,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.2072857142857145e-05,
+      "loss": 0.2436,
+      "step": 5550
+    },
+    {
+      "epoch": 0.8060599470841942,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.205857142857143e-05,
+      "loss": 0.2405,
+      "step": 5560
+    },
+    {
+      "epoch": 0.8075096951904607,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.2044285714285715e-05,
+      "loss": 0.2411,
+      "step": 5570
+    },
+    {
+      "epoch": 0.8089594432967272,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.2030000000000002e-05,
+      "loss": 0.2436,
+      "step": 5580
+    },
+    {
+      "epoch": 0.8104091914029937,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.2015714285714287e-05,
+      "loss": 0.2359,
+      "step": 5590
+    },
+    {
+      "epoch": 0.8118589395092602,
+      "grad_norm": 2.5,
+      "learning_rate": 1.2001428571428572e-05,
+      "loss": 0.2577,
+      "step": 5600
+    },
+    {
+      "epoch": 0.8133086876155268,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.1987142857142857e-05,
+      "loss": 0.2298,
+      "step": 5610
+    },
+    {
+      "epoch": 0.8147584357217933,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1972857142857144e-05,
+      "loss": 0.2433,
+      "step": 5620
+    },
+    {
+      "epoch": 0.8162081838280599,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.195857142857143e-05,
+      "loss": 0.2321,
+      "step": 5630
+    },
+    {
+      "epoch": 0.8176579319343265,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.1944285714285714e-05,
+      "loss": 0.2446,
+      "step": 5640
+    },
+    {
+      "epoch": 0.819107680040593,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.1930000000000001e-05,
+      "loss": 0.2452,
+      "step": 5650
+    },
+    {
+      "epoch": 0.8205574281468595,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1915714285714286e-05,
+      "loss": 0.2335,
+      "step": 5660
+    },
+    {
+      "epoch": 0.822007176253126,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.1901428571428573e-05,
+      "loss": 0.2238,
+      "step": 5670
+    },
+    {
+      "epoch": 0.8234569243593926,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1887142857142856e-05,
+      "loss": 0.2184,
+      "step": 5680
+    },
+    {
+      "epoch": 0.8249066724656591,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.1872857142857143e-05,
+      "loss": 0.225,
+      "step": 5690
+    },
+    {
+      "epoch": 0.8263564205719256,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.185857142857143e-05,
+      "loss": 0.2363,
+      "step": 5700
+    },
+    {
+      "epoch": 0.8278061686781921,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1844285714285715e-05,
+      "loss": 0.229,
+      "step": 5710
+    },
+    {
+      "epoch": 0.8292559167844586,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.1830000000000002e-05,
+      "loss": 0.2484,
+      "step": 5720
+    },
+    {
+      "epoch": 0.8307056648907253,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.1815714285714285e-05,
+      "loss": 0.2324,
+      "step": 5730
+    },
+    {
+      "epoch": 0.8321554129969918,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.1801428571428572e-05,
+      "loss": 0.2499,
+      "step": 5740
+    },
+    {
+      "epoch": 0.8336051611032583,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1787142857142857e-05,
+      "loss": 0.2431,
+      "step": 5750
+    },
+    {
+      "epoch": 0.8350549092095249,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1772857142857144e-05,
+      "loss": 0.2378,
+      "step": 5760
+    },
+    {
+      "epoch": 0.8365046573157914,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.1758571428571431e-05,
+      "loss": 0.2479,
+      "step": 5770
+    },
+    {
+      "epoch": 0.8379544054220579,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1744285714285714e-05,
+      "loss": 0.2325,
+      "step": 5780
+    },
+    {
+      "epoch": 0.8394041535283244,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1730000000000001e-05,
+      "loss": 0.237,
+      "step": 5790
+    },
+    {
+      "epoch": 0.840853901634591,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1715714285714286e-05,
+      "loss": 0.2322,
+      "step": 5800
+    },
+    {
+      "epoch": 0.8423036497408575,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1701428571428573e-05,
+      "loss": 0.2351,
+      "step": 5810
+    },
+    {
+      "epoch": 0.8437533978471241,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.168714285714286e-05,
+      "loss": 0.2309,
+      "step": 5820
+    },
+    {
+      "epoch": 0.8452031459533906,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1672857142857143e-05,
+      "loss": 0.2402,
+      "step": 5830
+    },
+    {
+      "epoch": 0.8466528940596572,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.165857142857143e-05,
+      "loss": 0.2453,
+      "step": 5840
+    },
+    {
+      "epoch": 0.8481026421659237,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.1644285714285715e-05,
+      "loss": 0.2349,
+      "step": 5850
+    },
+    {
+      "epoch": 0.8495523902721902,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.163e-05,
+      "loss": 0.214,
+      "step": 5860
+    },
+    {
+      "epoch": 0.8510021383784567,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1615714285714286e-05,
+      "loss": 0.2413,
+      "step": 5870
+    },
+    {
+      "epoch": 0.8524518864847233,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.1601428571428572e-05,
+      "loss": 0.2252,
+      "step": 5880
+    },
+    {
+      "epoch": 0.8539016345909898,
+      "grad_norm": 1.875,
+      "learning_rate": 1.158714285714286e-05,
+      "loss": 0.2311,
+      "step": 5890
+    },
+    {
+      "epoch": 0.8553513826972563,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.1572857142857143e-05,
+      "loss": 0.2316,
+      "step": 5900
+    },
+    {
+      "epoch": 0.8568011308035229,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.155857142857143e-05,
+      "loss": 0.2315,
+      "step": 5910
+    },
+    {
+      "epoch": 0.8582508789097895,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.1544285714285715e-05,
+      "loss": 0.2475,
+      "step": 5920
+    },
+    {
+      "epoch": 0.859700627016056,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.1530000000000001e-05,
+      "loss": 0.2251,
+      "step": 5930
+    },
+    {
+      "epoch": 0.8611503751223225,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1515714285714285e-05,
+      "loss": 0.2405,
+      "step": 5940
+    },
+    {
+      "epoch": 0.862600123228589,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1501428571428572e-05,
+      "loss": 0.2346,
+      "step": 5950
+    },
+    {
+      "epoch": 0.8640498713348556,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.1487142857142858e-05,
+      "loss": 0.212,
+      "step": 5960
+    },
+    {
+      "epoch": 0.8654996194411221,
+      "grad_norm": 2.125,
+      "learning_rate": 1.1472857142857144e-05,
+      "loss": 0.2262,
+      "step": 5970
+    },
+    {
+      "epoch": 0.8669493675473886,
+      "grad_norm": 2.25,
+      "learning_rate": 1.145857142857143e-05,
+      "loss": 0.2199,
+      "step": 5980
+    },
+    {
+      "epoch": 0.8683991156536551,
+      "grad_norm": 2.25,
+      "learning_rate": 1.1444285714285714e-05,
+      "loss": 0.2357,
+      "step": 5990
+    },
+    {
+      "epoch": 0.8698488637599218,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.143e-05,
+      "loss": 0.2367,
+      "step": 6000
+    },
+    {
+      "epoch": 0.8712986118661883,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.1415714285714287e-05,
+      "loss": 0.2368,
+      "step": 6010
+    },
+    {
+      "epoch": 0.8727483599724548,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1401428571428573e-05,
+      "loss": 0.2264,
+      "step": 6020
+    },
+    {
+      "epoch": 0.8741981080787213,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.138714285714286e-05,
+      "loss": 0.2286,
+      "step": 6030
+    },
+    {
+      "epoch": 0.8756478561849879,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.1372857142857143e-05,
+      "loss": 0.2475,
+      "step": 6040
+    },
+    {
+      "epoch": 0.8770976042912544,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.135857142857143e-05,
+      "loss": 0.2328,
+      "step": 6050
+    },
+    {
+      "epoch": 0.8785473523975209,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.1344285714285715e-05,
+      "loss": 0.2405,
+      "step": 6060
+    },
+    {
+      "epoch": 0.8799971005037874,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.1330000000000002e-05,
+      "loss": 0.2412,
+      "step": 6070
+    },
+    {
+      "epoch": 0.881446848610054,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.1315714285714288e-05,
+      "loss": 0.2476,
+      "step": 6080
+    },
+    {
+      "epoch": 0.8828965967163206,
+      "grad_norm": 1.875,
+      "learning_rate": 1.1301428571428572e-05,
+      "loss": 0.2305,
+      "step": 6090
+    },
+    {
+      "epoch": 0.8843463448225871,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.1287142857142859e-05,
+      "loss": 0.2292,
+      "step": 6100
+    },
+    {
+      "epoch": 0.8857960929288536,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.1272857142857144e-05,
+      "loss": 0.223,
+      "step": 6110
+    },
+    {
+      "epoch": 0.8872458410351202,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.125857142857143e-05,
+      "loss": 0.2352,
+      "step": 6120
+    },
+    {
+      "epoch": 0.8886955891413867,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.1244285714285714e-05,
+      "loss": 0.2373,
+      "step": 6130
+    },
+    {
+      "epoch": 0.8901453372476532,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1230000000000001e-05,
+      "loss": 0.2304,
+      "step": 6140
+    },
+    {
+      "epoch": 0.8915950853539197,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1215714285714288e-05,
+      "loss": 0.232,
+      "step": 6150
+    },
+    {
+      "epoch": 0.8930448334601863,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.1201428571428571e-05,
+      "loss": 0.2523,
+      "step": 6160
+    },
+    {
+      "epoch": 0.8944945815664528,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.1187142857142858e-05,
+      "loss": 0.2271,
+      "step": 6170
+    },
+    {
+      "epoch": 0.8959443296727194,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1172857142857143e-05,
+      "loss": 0.2295,
+      "step": 6180
+    },
+    {
+      "epoch": 0.8973940777789859,
+      "grad_norm": 2.125,
+      "learning_rate": 1.115857142857143e-05,
+      "loss": 0.261,
+      "step": 6190
+    },
+    {
+      "epoch": 0.8988438258852525,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1144285714285713e-05,
+      "loss": 0.2184,
+      "step": 6200
+    },
+    {
+      "epoch": 0.900293573991519,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.113e-05,
+      "loss": 0.2322,
+      "step": 6210
+    },
+    {
+      "epoch": 0.9017433220977855,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.1115714285714287e-05,
+      "loss": 0.2497,
+      "step": 6220
+    },
+    {
+      "epoch": 0.903193070204052,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.1101428571428572e-05,
+      "loss": 0.2368,
+      "step": 6230
+    },
+    {
+      "epoch": 0.9046428183103186,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.1087142857142859e-05,
+      "loss": 0.2301,
+      "step": 6240
+    },
+    {
+      "epoch": 0.9060925664165851,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.1072857142857142e-05,
+      "loss": 0.2162,
+      "step": 6250
+    },
+    {
+      "epoch": 0.9075423145228516,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.105857142857143e-05,
+      "loss": 0.2289,
+      "step": 6260
+    },
+    {
+      "epoch": 0.9089920626291181,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1044285714285716e-05,
+      "loss": 0.21,
+      "step": 6270
+    },
+    {
+      "epoch": 0.9104418107353848,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.1030000000000001e-05,
+      "loss": 0.2345,
+      "step": 6280
+    },
+    {
+      "epoch": 0.9118915588416513,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1015714285714288e-05,
+      "loss": 0.2229,
+      "step": 6290
+    },
+    {
+      "epoch": 0.9133413069479178,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.1001428571428571e-05,
+      "loss": 0.2281,
+      "step": 6300
+    },
+    {
+      "epoch": 0.9147910550541843,
+      "grad_norm": 2.25,
+      "learning_rate": 1.0987142857142858e-05,
+      "loss": 0.2288,
+      "step": 6310
+    },
+    {
+      "epoch": 0.9162408031604509,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.0972857142857143e-05,
+      "loss": 0.237,
+      "step": 6320
+    },
+    {
+      "epoch": 0.9176905512667174,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.095857142857143e-05,
+      "loss": 0.2377,
+      "step": 6330
+    },
+    {
+      "epoch": 0.9191402993729839,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.0944285714285717e-05,
+      "loss": 0.2284,
+      "step": 6340
+    },
+    {
+      "epoch": 0.9205900474792504,
+      "grad_norm": 2.125,
+      "learning_rate": 1.093e-05,
+      "loss": 0.2203,
+      "step": 6350
+    },
+    {
+      "epoch": 0.922039795585517,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.0915714285714287e-05,
+      "loss": 0.2127,
+      "step": 6360
+    },
+    {
+      "epoch": 0.9234895436917836,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.0901428571428572e-05,
+      "loss": 0.2265,
+      "step": 6370
+    },
+    {
+      "epoch": 0.9249392917980501,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0887142857142859e-05,
+      "loss": 0.211,
+      "step": 6380
+    },
+    {
+      "epoch": 0.9263890399043166,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.0872857142857143e-05,
+      "loss": 0.2274,
+      "step": 6390
+    },
+    {
+      "epoch": 0.9278387880105832,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.085857142857143e-05,
+      "loss": 0.2389,
+      "step": 6400
+    },
+    {
+      "epoch": 0.9292885361168497,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.0844285714285716e-05,
+      "loss": 0.2426,
+      "step": 6410
+    },
+    {
+      "epoch": 0.9307382842231162,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0830000000000001e-05,
+      "loss": 0.1981,
+      "step": 6420
+    },
+    {
+      "epoch": 0.9321880323293827,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.0815714285714286e-05,
+      "loss": 0.2419,
+      "step": 6430
+    },
+    {
+      "epoch": 0.9336377804356493,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0801428571428572e-05,
+      "loss": 0.2388,
+      "step": 6440
+    },
+    {
+      "epoch": 0.9350875285419158,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.0787142857142858e-05,
+      "loss": 0.2282,
+      "step": 6450
+    },
+    {
+      "epoch": 0.9365372766481824,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.0772857142857142e-05,
+      "loss": 0.2325,
+      "step": 6460
+    },
+    {
+      "epoch": 0.937987024754449,
+      "grad_norm": 2.0,
+      "learning_rate": 1.0758571428571429e-05,
+      "loss": 0.2354,
+      "step": 6470
+    },
+    {
+      "epoch": 0.9394367728607155,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0744285714285715e-05,
+      "loss": 0.2426,
+      "step": 6480
+    },
+    {
+      "epoch": 0.940886520966982,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.073e-05,
+      "loss": 0.2254,
+      "step": 6490
+    },
+    {
+      "epoch": 0.9423362690732485,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.0715714285714287e-05,
+      "loss": 0.2166,
+      "step": 6500
+    },
+    {
+      "epoch": 0.943786017179515,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.0701428571428571e-05,
+      "loss": 0.2233,
+      "step": 6510
+    },
+    {
+      "epoch": 0.9452357652857816,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.0687142857142858e-05,
+      "loss": 0.2191,
+      "step": 6520
+    },
+    {
+      "epoch": 0.9466855133920481,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0672857142857144e-05,
+      "loss": 0.2301,
+      "step": 6530
+    },
+    {
+      "epoch": 0.9481352614983146,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.065857142857143e-05,
+      "loss": 0.2332,
+      "step": 6540
+    },
+    {
+      "epoch": 0.9495850096045813,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0644285714285716e-05,
+      "loss": 0.2412,
+      "step": 6550
+    },
+    {
+      "epoch": 0.9510347577108478,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.063e-05,
+      "loss": 0.2372,
+      "step": 6560
+    },
+    {
+      "epoch": 0.9524845058171143,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.0615714285714287e-05,
+      "loss": 0.2264,
+      "step": 6570
+    },
+    {
+      "epoch": 0.9539342539233808,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.0601428571428572e-05,
+      "loss": 0.2215,
+      "step": 6580
+    },
+    {
+      "epoch": 0.9553840020296474,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.0587142857142859e-05,
+      "loss": 0.2174,
+      "step": 6590
+    },
+    {
+      "epoch": 0.9568337501359139,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0572857142857145e-05,
+      "loss": 0.2182,
+      "step": 6600
+    },
+    {
+      "epoch": 0.9582834982421804,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0558571428571429e-05,
+      "loss": 0.2322,
+      "step": 6610
+    },
+    {
+      "epoch": 0.9597332463484469,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.0544285714285716e-05,
+      "loss": 0.1999,
+      "step": 6620
+    },
+    {
+      "epoch": 0.9611829944547134,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.053e-05,
+      "loss": 0.2347,
+      "step": 6630
+    },
+    {
+      "epoch": 0.9626327425609801,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.0515714285714288e-05,
+      "loss": 0.2089,
+      "step": 6640
+    },
+    {
+      "epoch": 0.9640824906672466,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.0501428571428571e-05,
+      "loss": 0.232,
+      "step": 6650
+    },
+    {
+      "epoch": 0.9655322387735131,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.0487142857142858e-05,
+      "loss": 0.2343,
+      "step": 6660
+    },
+    {
+      "epoch": 0.9669819868797797,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.0472857142857145e-05,
+      "loss": 0.2332,
+      "step": 6670
+    },
+    {
+      "epoch": 0.9684317349860462,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.045857142857143e-05,
+      "loss": 0.2228,
+      "step": 6680
+    },
+    {
+      "epoch": 0.9698814830923127,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.0444285714285715e-05,
+      "loss": 0.2189,
+      "step": 6690
+    },
+    {
+      "epoch": 0.9713312311985792,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.043e-05,
+      "loss": 0.2158,
+      "step": 6700
+    },
+    {
+      "epoch": 0.9727809793048457,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0415714285714287e-05,
+      "loss": 0.2389,
+      "step": 6710
+    },
+    {
+      "epoch": 0.9742307274111123,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.0401428571428572e-05,
+      "loss": 0.2135,
+      "step": 6720
+    },
+    {
+      "epoch": 0.9756804755173789,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.0387142857142857e-05,
+      "loss": 0.2358,
+      "step": 6730
+    },
+    {
+      "epoch": 0.9771302236236454,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.0372857142857144e-05,
+      "loss": 0.2145,
+      "step": 6740
+    },
+    {
+      "epoch": 0.978579971729912,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0358571428571429e-05,
+      "loss": 0.2219,
+      "step": 6750
+    },
+    {
+      "epoch": 0.9800297198361785,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.0344285714285716e-05,
+      "loss": 0.2246,
+      "step": 6760
+    },
+    {
+      "epoch": 0.981479467942445,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.033e-05,
+      "loss": 0.2188,
+      "step": 6770
+    },
+    {
+      "epoch": 0.9829292160487115,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.0315714285714286e-05,
+      "loss": 0.2222,
+      "step": 6780
+    },
+    {
+      "epoch": 0.984378964154978,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.0301428571428573e-05,
+      "loss": 0.2267,
+      "step": 6790
+    },
+    {
+      "epoch": 0.9858287122612446,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.0287142857142858e-05,
+      "loss": 0.237,
+      "step": 6800
+    },
+    {
+      "epoch": 0.9872784603675111,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.0272857142857145e-05,
+      "loss": 0.2388,
+      "step": 6810
+    },
+    {
+      "epoch": 0.9887282084737777,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.0258571428571428e-05,
+      "loss": 0.2232,
+      "step": 6820
+    },
+    {
+      "epoch": 0.9901779565800443,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.0244285714285715e-05,
+      "loss": 0.2187,
+      "step": 6830
+    },
+    {
+      "epoch": 0.9916277046863108,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.023e-05,
+      "loss": 0.218,
+      "step": 6840
+    },
+    {
+      "epoch": 0.9930774527925773,
+      "grad_norm": 1.875,
+      "learning_rate": 1.0215714285714287e-05,
+      "loss": 0.2227,
+      "step": 6850
+    },
+    {
+      "epoch": 0.9945272008988438,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0201428571428574e-05,
+      "loss": 0.2394,
+      "step": 6860
+    },
+    {
+      "epoch": 0.9959769490051104,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.0187142857142857e-05,
+      "loss": 0.2394,
+      "step": 6870
+    },
+    {
+      "epoch": 0.9974266971113769,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.0172857142857144e-05,
+      "loss": 0.2354,
+      "step": 6880
+    },
+    {
+      "epoch": 0.9988764452176434,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.015857142857143e-05,
+      "loss": 0.2096,
+      "step": 6890
+    },
+    {
+      "epoch": 1.0002899496212534,
+      "grad_norm": 1.875,
+      "learning_rate": 1.0144285714285716e-05,
+      "loss": 0.2204,
+      "step": 6900
+    },
+    {
+      "epoch": 1.00173969772752,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.013e-05,
+      "loss": 0.201,
+      "step": 6910
+    },
+    {
+      "epoch": 1.0031894458337864,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.0115714285714286e-05,
+      "loss": 0.2083,
+      "step": 6920
+    },
+    {
+      "epoch": 1.004639193940053,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.0101428571428573e-05,
+      "loss": 0.2113,
+      "step": 6930
+    },
+    {
+      "epoch": 1.0060889420463195,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.0087142857142858e-05,
+      "loss": 0.1838,
+      "step": 6940
+    },
+    {
+      "epoch": 1.007538690152586,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0072857142857143e-05,
+      "loss": 0.216,
+      "step": 6950
+    },
+    {
+      "epoch": 1.0089884382588525,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.0058571428571429e-05,
+      "loss": 0.2188,
+      "step": 6960
+    },
+    {
+      "epoch": 1.010438186365119,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0044285714285715e-05,
+      "loss": 0.1987,
+      "step": 6970
+    },
+    {
+      "epoch": 1.0118879344713856,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.003e-05,
+      "loss": 0.2046,
+      "step": 6980
+    },
+    {
+      "epoch": 1.013337682577652,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0015714285714286e-05,
+      "loss": 0.1939,
+      "step": 6990
+    },
+    {
+      "epoch": 1.0147874306839186,
+      "grad_norm": 2.0,
+      "learning_rate": 1.0001428571428572e-05,
+      "loss": 0.2157,
+      "step": 7000
+    },
+    {
+      "epoch": 1.0162371787901852,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.987142857142858e-06,
+      "loss": 0.2023,
+      "step": 7010
+    },
+    {
+      "epoch": 1.0176869268964517,
+      "grad_norm": 1.8359375,
+      "learning_rate": 9.972857142857144e-06,
+      "loss": 0.2168,
+      "step": 7020
+    },
+    {
+      "epoch": 1.0191366750027182,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.95857142857143e-06,
+      "loss": 0.2152,
+      "step": 7030
+    },
+    {
+      "epoch": 1.0205864231089847,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.944285714285715e-06,
+      "loss": 0.1925,
+      "step": 7040
+    },
+    {
+      "epoch": 1.0220361712152513,
+      "grad_norm": 1.375,
+      "learning_rate": 9.930000000000001e-06,
+      "loss": 0.212,
+      "step": 7050
+    },
+    {
+      "epoch": 1.0234859193215178,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.915714285714287e-06,
+      "loss": 0.2087,
+      "step": 7060
+    },
+    {
+      "epoch": 1.0249356674277845,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.901428571428572e-06,
+      "loss": 0.187,
+      "step": 7070
+    },
+    {
+      "epoch": 1.026385415534051,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.887142857142859e-06,
+      "loss": 0.2219,
+      "step": 7080
+    },
+    {
+      "epoch": 1.0278351636403176,
+      "grad_norm": 1.921875,
+      "learning_rate": 9.872857142857144e-06,
+      "loss": 0.1973,
+      "step": 7090
+    },
+    {
+      "epoch": 1.029284911746584,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.858571428571429e-06,
+      "loss": 0.2048,
+      "step": 7100
+    },
+    {
+      "epoch": 1.0307346598528506,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.844285714285716e-06,
+      "loss": 0.1819,
+      "step": 7110
+    },
+    {
+      "epoch": 1.0321844079591171,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.83e-06,
+      "loss": 0.2033,
+      "step": 7120
+    },
+    {
+      "epoch": 1.0336341560653837,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.815714285714286e-06,
+      "loss": 0.2093,
+      "step": 7130
+    },
+    {
+      "epoch": 1.0350839041716502,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.801428571428571e-06,
+      "loss": 0.2198,
+      "step": 7140
+    },
+    {
+      "epoch": 1.0365336522779167,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.787142857142858e-06,
+      "loss": 0.1839,
+      "step": 7150
+    },
+    {
+      "epoch": 1.0379834003841832,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.772857142857145e-06,
+      "loss": 0.1911,
+      "step": 7160
+    },
+    {
+      "epoch": 1.0394331484904498,
+      "grad_norm": 1.515625,
+      "learning_rate": 9.75857142857143e-06,
+      "loss": 0.2161,
+      "step": 7170
+    },
+    {
+      "epoch": 1.0408828965967163,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.744285714285715e-06,
+      "loss": 0.2004,
+      "step": 7180
+    },
+    {
+      "epoch": 1.0423326447029828,
+      "grad_norm": 1.84375,
+      "learning_rate": 9.73e-06,
+      "loss": 0.2118,
+      "step": 7190
+    },
+    {
+      "epoch": 1.0437823928092493,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.715714285714287e-06,
+      "loss": 0.206,
+      "step": 7200
+    },
+    {
+      "epoch": 1.0452321409155159,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.701428571428572e-06,
+      "loss": 0.2187,
+      "step": 7210
+    },
+    {
+      "epoch": 1.0466818890217824,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.687142857142859e-06,
+      "loss": 0.1944,
+      "step": 7220
+    },
+    {
+      "epoch": 1.048131637128049,
+      "grad_norm": 1.5546875,
+      "learning_rate": 9.672857142857144e-06,
+      "loss": 0.1908,
+      "step": 7230
+    },
+    {
+      "epoch": 1.0495813852343154,
+      "grad_norm": 1.90625,
+      "learning_rate": 9.658571428571429e-06,
+      "loss": 0.1946,
+      "step": 7240
+    },
+    {
+      "epoch": 1.0510311333405822,
+      "grad_norm": 2.0,
+      "learning_rate": 9.644285714285714e-06,
+      "loss": 0.2026,
+      "step": 7250
+    },
+    {
+      "epoch": 1.0524808814468487,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.630000000000001e-06,
+      "loss": 0.1932,
+      "step": 7260
+    },
+    {
+      "epoch": 1.0539306295531152,
+      "grad_norm": 2.0,
+      "learning_rate": 9.615714285714286e-06,
+      "loss": 0.2008,
+      "step": 7270
+    },
+    {
+      "epoch": 1.0553803776593818,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.601428571428573e-06,
+      "loss": 0.2057,
+      "step": 7280
+    },
+    {
+      "epoch": 1.0568301257656483,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.587142857142858e-06,
+      "loss": 0.1897,
+      "step": 7290
+    },
+    {
+      "epoch": 1.0582798738719148,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.572857142857143e-06,
+      "loss": 0.2056,
+      "step": 7300
+    },
+    {
+      "epoch": 1.0597296219781813,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.55857142857143e-06,
+      "loss": 0.1976,
+      "step": 7310
+    },
+    {
+      "epoch": 1.0611793700844478,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.544285714285715e-06,
+      "loss": 0.2027,
+      "step": 7320
+    },
+    {
+      "epoch": 1.0626291181907144,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.53e-06,
+      "loss": 0.1972,
+      "step": 7330
+    },
+    {
+      "epoch": 1.064078866296981,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.515714285714287e-06,
+      "loss": 0.201,
+      "step": 7340
+    },
+    {
+      "epoch": 1.0655286144032474,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.501428571428572e-06,
+      "loss": 0.1929,
+      "step": 7350
+    },
+    {
+      "epoch": 1.066978362509514,
+      "grad_norm": 1.4609375,
+      "learning_rate": 9.487142857142857e-06,
+      "loss": 0.1933,
+      "step": 7360
+    },
+    {
+      "epoch": 1.0684281106157805,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.472857142857144e-06,
+      "loss": 0.2084,
+      "step": 7370
+    },
+    {
+      "epoch": 1.069877858722047,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.45857142857143e-06,
+      "loss": 0.2003,
+      "step": 7380
+    },
+    {
+      "epoch": 1.0713276068283135,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.444285714285714e-06,
+      "loss": 0.2057,
+      "step": 7390
+    },
+    {
+      "epoch": 1.07277735493458,
+      "grad_norm": 1.921875,
+      "learning_rate": 9.43e-06,
+      "loss": 0.2041,
+      "step": 7400
+    },
+    {
+      "epoch": 1.0742271030408466,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.415714285714286e-06,
+      "loss": 0.196,
+      "step": 7410
+    },
+    {
+      "epoch": 1.075676851147113,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.401428571428573e-06,
+      "loss": 0.2179,
+      "step": 7420
+    },
+    {
+      "epoch": 1.0771265992533796,
+      "grad_norm": 1.453125,
+      "learning_rate": 9.387142857142858e-06,
+      "loss": 0.2012,
+      "step": 7430
+    },
+    {
+      "epoch": 1.0785763473596464,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.372857142857143e-06,
+      "loss": 0.1966,
+      "step": 7440
+    },
+    {
+      "epoch": 1.0800260954659129,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.358571428571429e-06,
+      "loss": 0.2109,
+      "step": 7450
+    },
+    {
+      "epoch": 1.0814758435721794,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.344285714285715e-06,
+      "loss": 0.2008,
+      "step": 7460
+    },
+    {
+      "epoch": 1.082925591678446,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.33e-06,
+      "loss": 0.2022,
+      "step": 7470
+    },
+    {
+      "epoch": 1.0843753397847125,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.315714285714287e-06,
+      "loss": 0.1981,
+      "step": 7480
+    },
+    {
+      "epoch": 1.085825087890979,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.301428571428572e-06,
+      "loss": 0.2026,
+      "step": 7490
+    },
+    {
+      "epoch": 1.0872748359972455,
+      "grad_norm": 1.890625,
+      "learning_rate": 9.287142857142858e-06,
+      "loss": 0.1985,
+      "step": 7500
+    },
+    {
+      "epoch": 1.088724584103512,
+      "grad_norm": 2.5,
+      "learning_rate": 9.272857142857143e-06,
+      "loss": 0.2012,
+      "step": 7510
+    },
+    {
+      "epoch": 1.0901743322097786,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.25857142857143e-06,
+      "loss": 0.2057,
+      "step": 7520
+    },
+    {
+      "epoch": 1.091624080316045,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.244285714285715e-06,
+      "loss": 0.1991,
+      "step": 7530
+    },
+    {
+      "epoch": 1.0930738284223116,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.230000000000001e-06,
+      "loss": 0.1941,
+      "step": 7540
+    },
+    {
+      "epoch": 1.0945235765285781,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.215714285714287e-06,
+      "loss": 0.1992,
+      "step": 7550
+    },
+    {
+      "epoch": 1.0959733246348446,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.201428571428572e-06,
+      "loss": 0.1924,
+      "step": 7560
+    },
+    {
+      "epoch": 1.0974230727411112,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.187142857142858e-06,
+      "loss": 0.1968,
+      "step": 7570
+    },
+    {
+      "epoch": 1.0988728208473777,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.172857142857144e-06,
+      "loss": 0.1838,
+      "step": 7580
+    },
+    {
+      "epoch": 1.1003225689536442,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.158571428571429e-06,
+      "loss": 0.2,
+      "step": 7590
+    },
+    {
+      "epoch": 1.1017723170599107,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.144285714285716e-06,
+      "loss": 0.2008,
+      "step": 7600
+    },
+    {
+      "epoch": 1.1032220651661775,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.13e-06,
+      "loss": 0.2118,
+      "step": 7610
+    },
+    {
+      "epoch": 1.104671813272444,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.115714285714286e-06,
+      "loss": 0.2059,
+      "step": 7620
+    },
+    {
+      "epoch": 1.1061215613787105,
+      "grad_norm": 1.84375,
+      "learning_rate": 9.101428571428573e-06,
+      "loss": 0.1865,
+      "step": 7630
+    },
+    {
+      "epoch": 1.107571309484977,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.087142857142858e-06,
+      "loss": 0.1966,
+      "step": 7640
+    },
+    {
+      "epoch": 1.1090210575912436,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.072857142857143e-06,
+      "loss": 0.1961,
+      "step": 7650
+    },
+    {
+      "epoch": 1.11047080569751,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.058571428571428e-06,
+      "loss": 0.2088,
+      "step": 7660
+    },
+    {
+      "epoch": 1.1119205538037766,
+      "grad_norm": 1.921875,
+      "learning_rate": 9.044285714285715e-06,
+      "loss": 0.2034,
+      "step": 7670
+    },
+    {
+      "epoch": 1.1133703019100432,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.030000000000002e-06,
+      "loss": 0.1816,
+      "step": 7680
+    },
+    {
+      "epoch": 1.1148200500163097,
+      "grad_norm": 2.0,
+      "learning_rate": 9.015714285714287e-06,
+      "loss": 0.2106,
+      "step": 7690
+    },
+    {
+      "epoch": 1.1162697981225762,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.001428571428572e-06,
+      "loss": 0.1921,
+      "step": 7700
+    },
+    {
+      "epoch": 1.1177195462288427,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.987142857142857e-06,
+      "loss": 0.203,
+      "step": 7710
+    },
+    {
+      "epoch": 1.1191692943351093,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.972857142857144e-06,
+      "loss": 0.2016,
+      "step": 7720
+    },
+    {
+      "epoch": 1.1206190424413758,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.95857142857143e-06,
+      "loss": 0.2001,
+      "step": 7730
+    },
+    {
+      "epoch": 1.1220687905476423,
+      "grad_norm": 2.0,
+      "learning_rate": 8.944285714285716e-06,
+      "loss": 0.2063,
+      "step": 7740
+    },
+    {
+      "epoch": 1.1235185386539088,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.930000000000001e-06,
+      "loss": 0.2004,
+      "step": 7750
+    },
+    {
+      "epoch": 1.1249682867601754,
+      "grad_norm": 2.25,
+      "learning_rate": 8.915714285714286e-06,
+      "loss": 0.2026,
+      "step": 7760
+    },
+    {
+      "epoch": 1.1264180348664419,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.901428571428571e-06,
+      "loss": 0.1966,
+      "step": 7770
+    },
+    {
+      "epoch": 1.1278677829727084,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.887142857142858e-06,
+      "loss": 0.2082,
+      "step": 7780
+    },
+    {
+      "epoch": 1.129317531078975,
+      "grad_norm": 1.890625,
+      "learning_rate": 8.872857142857143e-06,
+      "loss": 0.1965,
+      "step": 7790
+    },
+    {
+      "epoch": 1.1307672791852417,
+      "grad_norm": 2.0,
+      "learning_rate": 8.85857142857143e-06,
+      "loss": 0.2007,
+      "step": 7800
+    },
+    {
+      "epoch": 1.1322170272915082,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.844285714285715e-06,
+      "loss": 0.1984,
+      "step": 7810
+    },
+    {
+      "epoch": 1.1336667753977747,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.83e-06,
+      "loss": 0.1969,
+      "step": 7820
+    },
+    {
+      "epoch": 1.1351165235040412,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.815714285714287e-06,
+      "loss": 0.2157,
+      "step": 7830
+    },
+    {
+      "epoch": 1.1365662716103078,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.801428571428572e-06,
+      "loss": 0.2063,
+      "step": 7840
+    },
+    {
+      "epoch": 1.1380160197165743,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.787142857142857e-06,
+      "loss": 0.1991,
+      "step": 7850
+    },
+    {
+      "epoch": 1.1394657678228408,
+      "grad_norm": 1.75,
+      "learning_rate": 8.772857142857144e-06,
+      "loss": 0.1918,
+      "step": 7860
+    },
+    {
+      "epoch": 1.1409155159291073,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.75857142857143e-06,
+      "loss": 0.198,
+      "step": 7870
+    },
+    {
+      "epoch": 1.1423652640353739,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.744285714285716e-06,
+      "loss": 0.2008,
+      "step": 7880
+    },
+    {
+      "epoch": 1.1438150121416404,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.730000000000001e-06,
+      "loss": 0.2081,
+      "step": 7890
+    },
+    {
+      "epoch": 1.145264760247907,
+      "grad_norm": 1.4609375,
+      "learning_rate": 8.715714285714286e-06,
+      "loss": 0.1879,
+      "step": 7900
+    },
+    {
+      "epoch": 1.1467145083541734,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.701428571428571e-06,
+      "loss": 0.1816,
+      "step": 7910
+    },
+    {
+      "epoch": 1.14816425646044,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.687142857142856e-06,
+      "loss": 0.2008,
+      "step": 7920
+    },
+    {
+      "epoch": 1.1496140045667065,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.672857142857143e-06,
+      "loss": 0.1943,
+      "step": 7930
+    },
+    {
+      "epoch": 1.151063752672973,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.65857142857143e-06,
+      "loss": 0.1975,
+      "step": 7940
+    },
+    {
+      "epoch": 1.1525135007792395,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.644285714285715e-06,
+      "loss": 0.1713,
+      "step": 7950
+    },
+    {
+      "epoch": 1.153963248885506,
+      "grad_norm": 1.625,
+      "learning_rate": 8.63e-06,
+      "loss": 0.197,
+      "step": 7960
+    },
+    {
+      "epoch": 1.1554129969917728,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.615714285714286e-06,
+      "loss": 0.2022,
+      "step": 7970
+    },
+    {
+      "epoch": 1.156862745098039,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.601428571428572e-06,
+      "loss": 0.1997,
+      "step": 7980
+    },
+    {
+      "epoch": 1.1583124932043058,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.587142857142859e-06,
+      "loss": 0.1888,
+      "step": 7990
+    },
+    {
+      "epoch": 1.1597622413105724,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.572857142857144e-06,
+      "loss": 0.2019,
+      "step": 8000
+    },
+    {
+      "epoch": 1.161211989416839,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.55857142857143e-06,
+      "loss": 0.2099,
+      "step": 8010
+    },
+    {
+      "epoch": 1.1626617375231054,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.544285714285715e-06,
+      "loss": 0.1978,
+      "step": 8020
+    },
+    {
+      "epoch": 1.164111485629372,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.530000000000001e-06,
+      "loss": 0.1958,
+      "step": 8030
+    },
+    {
+      "epoch": 1.1655612337356385,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.515714285714286e-06,
+      "loss": 0.2026,
+      "step": 8040
+    },
+    {
+      "epoch": 1.167010981841905,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.501428571428572e-06,
+      "loss": 0.2019,
+      "step": 8050
+    },
+    {
+      "epoch": 1.1684607299481715,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.487142857142858e-06,
+      "loss": 0.2039,
+      "step": 8060
+    },
+    {
+      "epoch": 1.169910478054438,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.472857142857144e-06,
+      "loss": 0.1953,
+      "step": 8070
+    },
+    {
+      "epoch": 1.1713602261607046,
+      "grad_norm": 1.4140625,
+      "learning_rate": 8.458571428571429e-06,
+      "loss": 0.1924,
+      "step": 8080
+    },
+    {
+      "epoch": 1.172809974266971,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.444285714285715e-06,
+      "loss": 0.2241,
+      "step": 8090
+    },
+    {
+      "epoch": 1.1742597223732376,
+      "grad_norm": 1.75,
+      "learning_rate": 8.43e-06,
+      "loss": 0.1985,
+      "step": 8100
+    },
+    {
+      "epoch": 1.1757094704795041,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.415714285714286e-06,
+      "loss": 0.2064,
+      "step": 8110
+    },
+    {
+      "epoch": 1.1771592185857707,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.401428571428573e-06,
+      "loss": 0.2137,
+      "step": 8120
+    },
+    {
+      "epoch": 1.1786089666920372,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.387142857142858e-06,
+      "loss": 0.214,
+      "step": 8130
+    },
+    {
+      "epoch": 1.1800587147983037,
+      "grad_norm": 1.4609375,
+      "learning_rate": 8.372857142857144e-06,
+      "loss": 0.2006,
+      "step": 8140
+    },
+    {
+      "epoch": 1.1815084629045702,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.35857142857143e-06,
+      "loss": 0.212,
+      "step": 8150
+    },
+    {
+      "epoch": 1.182958211010837,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.344285714285715e-06,
+      "loss": 0.202,
+      "step": 8160
+    },
+    {
+      "epoch": 1.1844079591171033,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.33e-06,
+      "loss": 0.2044,
+      "step": 8170
+    },
+    {
+      "epoch": 1.18585770722337,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.315714285714287e-06,
+      "loss": 0.1901,
+      "step": 8180
+    },
+    {
+      "epoch": 1.1873074553296366,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.301428571428572e-06,
+      "loss": 0.2212,
+      "step": 8190
+    },
+    {
+      "epoch": 1.188757203435903,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.287142857142859e-06,
+      "loss": 0.1885,
+      "step": 8200
+    },
+    {
+      "epoch": 1.1902069515421696,
+      "grad_norm": 1.625,
+      "learning_rate": 8.272857142857144e-06,
+      "loss": 0.1944,
+      "step": 8210
+    },
+    {
+      "epoch": 1.1916566996484361,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.258571428571429e-06,
+      "loss": 0.2136,
+      "step": 8220
+    },
+    {
+      "epoch": 1.1931064477547026,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.244285714285714e-06,
+      "loss": 0.1903,
+      "step": 8230
+    },
+    {
+      "epoch": 1.1945561958609692,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.23e-06,
+      "loss": 0.1863,
+      "step": 8240
+    },
+    {
+      "epoch": 1.1960059439672357,
+      "grad_norm": 1.8046875,
+      "learning_rate": 8.215714285714288e-06,
+      "loss": 0.2084,
+      "step": 8250
+    },
+    {
+      "epoch": 1.1974556920735022,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.201428571428573e-06,
+      "loss": 0.1898,
+      "step": 8260
+    },
+    {
+      "epoch": 1.1989054401797687,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.187142857142858e-06,
+      "loss": 0.2051,
+      "step": 8270
+    },
+    {
+      "epoch": 1.2003551882860353,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.172857142857143e-06,
+      "loss": 0.1956,
+      "step": 8280
+    },
+    {
+      "epoch": 1.2018049363923018,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.15857142857143e-06,
+      "loss": 0.2029,
+      "step": 8290
+    },
+    {
+      "epoch": 1.2032546844985683,
+      "grad_norm": 1.4609375,
+      "learning_rate": 8.144285714285715e-06,
+      "loss": 0.1989,
+      "step": 8300
+    },
+    {
+      "epoch": 1.2047044326048348,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.13e-06,
+      "loss": 0.1992,
+      "step": 8310
+    },
+    {
+      "epoch": 1.2061541807111014,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.115714285714287e-06,
+      "loss": 0.1984,
+      "step": 8320
+    },
+    {
+      "epoch": 1.2076039288173679,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.101428571428572e-06,
+      "loss": 0.1959,
+      "step": 8330
+    },
+    {
+      "epoch": 1.2090536769236344,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.087142857142857e-06,
+      "loss": 0.1953,
+      "step": 8340
+    },
+    {
+      "epoch": 1.2105034250299012,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.072857142857144e-06,
+      "loss": 0.2063,
+      "step": 8350
+    },
+    {
+      "epoch": 1.2119531731361677,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.058571428571429e-06,
+      "loss": 0.2213,
+      "step": 8360
+    },
+    {
+      "epoch": 1.2134029212424342,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.044285714285714e-06,
+      "loss": 0.1945,
+      "step": 8370
+    },
+    {
+      "epoch": 1.2148526693487007,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.030000000000001e-06,
+      "loss": 0.1903,
+      "step": 8380
+    },
+    {
+      "epoch": 1.2163024174549673,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.015714285714286e-06,
+      "loss": 0.195,
+      "step": 8390
+    },
+    {
+      "epoch": 1.2177521655612338,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.001428571428573e-06,
+      "loss": 0.2023,
+      "step": 8400
+    },
+    {
+      "epoch": 1.2192019136675003,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.987142857142858e-06,
+      "loss": 0.2029,
+      "step": 8410
+    },
+    {
+      "epoch": 1.2206516617737668,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.972857142857143e-06,
+      "loss": 0.1894,
+      "step": 8420
+    },
+    {
+      "epoch": 1.2221014098800334,
+      "grad_norm": 1.484375,
+      "learning_rate": 7.958571428571428e-06,
+      "loss": 0.2058,
+      "step": 8430
+    },
+    {
+      "epoch": 1.2235511579862999,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.944285714285715e-06,
+      "loss": 0.1899,
+      "step": 8440
+    },
+    {
+      "epoch": 1.2250009060925664,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.93e-06,
+      "loss": 0.2022,
+      "step": 8450
+    },
+    {
+      "epoch": 1.226450654198833,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.915714285714287e-06,
+      "loss": 0.1996,
+      "step": 8460
+    },
+    {
+      "epoch": 1.2279004023050994,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.901428571428572e-06,
+      "loss": 0.2094,
+      "step": 8470
+    },
+    {
+      "epoch": 1.229350150411366,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.887142857142857e-06,
+      "loss": 0.2069,
+      "step": 8480
+    },
+    {
+      "epoch": 1.2307998985176325,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.872857142857143e-06,
+      "loss": 0.2014,
+      "step": 8490
+    },
+    {
+      "epoch": 1.232249646623899,
+      "grad_norm": 1.4140625,
+      "learning_rate": 7.85857142857143e-06,
+      "loss": 0.1849,
+      "step": 8500
+    },
+    {
+      "epoch": 1.2336993947301655,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.844285714285716e-06,
+      "loss": 0.2006,
+      "step": 8510
+    },
+    {
+      "epoch": 1.2351491428364323,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.830000000000001e-06,
+      "loss": 0.187,
+      "step": 8520
+    },
+    {
+      "epoch": 1.2365988909426986,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.815714285714286e-06,
+      "loss": 0.1911,
+      "step": 8530
+    },
+    {
+      "epoch": 1.2380486390489653,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.801428571428572e-06,
+      "loss": 0.2149,
+      "step": 8540
+    },
+    {
+      "epoch": 1.2394983871552319,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.787142857142858e-06,
+      "loss": 0.1859,
+      "step": 8550
+    },
+    {
+      "epoch": 1.2409481352614984,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.772857142857143e-06,
+      "loss": 0.2096,
+      "step": 8560
+    },
+    {
+      "epoch": 1.242397883367765,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.75857142857143e-06,
+      "loss": 0.1981,
+      "step": 8570
+    },
+    {
+      "epoch": 1.2438476314740314,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.744285714285715e-06,
+      "loss": 0.1973,
+      "step": 8580
+    },
+    {
+      "epoch": 1.245297379580298,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.73e-06,
+      "loss": 0.1964,
+      "step": 8590
+    },
+    {
+      "epoch": 1.2467471276865645,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.715714285714286e-06,
+      "loss": 0.1909,
+      "step": 8600
+    },
+    {
+      "epoch": 1.248196875792831,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.701428571428572e-06,
+      "loss": 0.1812,
+      "step": 8610
+    },
+    {
+      "epoch": 1.2496466238990975,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.687142857142858e-06,
+      "loss": 0.1878,
+      "step": 8620
+    },
+    {
+      "epoch": 1.251096372005364,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.672857142857143e-06,
+      "loss": 0.1981,
+      "step": 8630
+    },
+    {
+      "epoch": 1.2525461201116306,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.65857142857143e-06,
+      "loss": 0.2017,
+      "step": 8640
+    },
+    {
+      "epoch": 1.253995868217897,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.644285714285715e-06,
+      "loss": 0.209,
+      "step": 8650
+    },
+    {
+      "epoch": 1.2554456163241636,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.630000000000001e-06,
+      "loss": 0.1866,
+      "step": 8660
+    },
+    {
+      "epoch": 1.2568953644304302,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.615714285714287e-06,
+      "loss": 0.1977,
+      "step": 8670
+    },
+    {
+      "epoch": 1.2583451125366967,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.601428571428572e-06,
+      "loss": 0.2159,
+      "step": 8680
+    },
+    {
+      "epoch": 1.2597948606429634,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.587142857142858e-06,
+      "loss": 0.1984,
+      "step": 8690
+    },
+    {
+      "epoch": 1.2612446087492297,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.572857142857144e-06,
+      "loss": 0.1922,
+      "step": 8700
+    },
+    {
+      "epoch": 1.2626943568554965,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.55857142857143e-06,
+      "loss": 0.1891,
+      "step": 8710
+    },
+    {
+      "epoch": 1.2641441049617628,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.544285714285715e-06,
+      "loss": 0.1944,
+      "step": 8720
+    },
+    {
+      "epoch": 1.2655938530680295,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.530000000000001e-06,
+      "loss": 0.1977,
+      "step": 8730
+    },
+    {
+      "epoch": 1.267043601174296,
+      "grad_norm": 1.84375,
+      "learning_rate": 7.515714285714286e-06,
+      "loss": 0.1962,
+      "step": 8740
+    },
+    {
+      "epoch": 1.2684933492805626,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.501428571428572e-06,
+      "loss": 0.2018,
+      "step": 8750
+    },
+    {
+      "epoch": 1.269943097386829,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.487142857142857e-06,
+      "loss": 0.1981,
+      "step": 8760
+    },
+    {
+      "epoch": 1.2713928454930956,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.472857142857144e-06,
+      "loss": 0.1927,
+      "step": 8770
+    },
+    {
+      "epoch": 1.2728425935993621,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.45857142857143e-06,
+      "loss": 0.1915,
+      "step": 8780
+    },
+    {
+      "epoch": 1.2742923417056287,
+      "grad_norm": 2.0,
+      "learning_rate": 7.444285714285715e-06,
+      "loss": 0.1878,
+      "step": 8790
+    },
+    {
+      "epoch": 1.2757420898118952,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.430000000000001e-06,
+      "loss": 0.1994,
+      "step": 8800
+    },
+    {
+      "epoch": 1.2771918379181617,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.415714285714286e-06,
+      "loss": 0.2047,
+      "step": 8810
+    },
+    {
+      "epoch": 1.2786415860244282,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.401428571428572e-06,
+      "loss": 0.2087,
+      "step": 8820
+    },
+    {
+      "epoch": 1.2800913341306948,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.387142857142858e-06,
+      "loss": 0.1947,
+      "step": 8830
+    },
+    {
+      "epoch": 1.2815410822369613,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.372857142857144e-06,
+      "loss": 0.1992,
+      "step": 8840
+    },
+    {
+      "epoch": 1.2829908303432278,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.358571428571429e-06,
+      "loss": 0.1911,
+      "step": 8850
+    },
+    {
+      "epoch": 1.2844405784494943,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.344285714285715e-06,
+      "loss": 0.1914,
+      "step": 8860
+    },
+    {
+      "epoch": 1.2858903265557609,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.33e-06,
+      "loss": 0.1975,
+      "step": 8870
+    },
+    {
+      "epoch": 1.2873400746620276,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.315714285714286e-06,
+      "loss": 0.2005,
+      "step": 8880
+    },
+    {
+      "epoch": 1.288789822768294,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.301428571428571e-06,
+      "loss": 0.1938,
+      "step": 8890
+    },
+    {
+      "epoch": 1.2902395708745606,
+      "grad_norm": 1.5703125,
+      "learning_rate": 7.287142857142858e-06,
+      "loss": 0.188,
+      "step": 8900
+    },
+    {
+      "epoch": 1.291689318980827,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.272857142857144e-06,
+      "loss": 0.1951,
+      "step": 8910
+    },
+    {
+      "epoch": 1.2931390670870937,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.258571428571429e-06,
+      "loss": 0.1998,
+      "step": 8920
+    },
+    {
+      "epoch": 1.2945888151933602,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.244285714285715e-06,
+      "loss": 0.2077,
+      "step": 8930
+    },
+    {
+      "epoch": 1.2960385632996267,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.23e-06,
+      "loss": 0.1929,
+      "step": 8940
+    },
+    {
+      "epoch": 1.2974883114058933,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.215714285714286e-06,
+      "loss": 0.2127,
+      "step": 8950
+    },
+    {
+      "epoch": 1.2989380595121598,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.201428571428572e-06,
+      "loss": 0.2068,
+      "step": 8960
+    },
+    {
+      "epoch": 1.3003878076184263,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.187142857142858e-06,
+      "loss": 0.1905,
+      "step": 8970
+    },
+    {
+      "epoch": 1.3018375557246928,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.172857142857143e-06,
+      "loss": 0.1879,
+      "step": 8980
+    },
+    {
+      "epoch": 1.3032873038309594,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.158571428571429e-06,
+      "loss": 0.1911,
+      "step": 8990
+    },
+    {
+      "epoch": 1.304737051937226,
+      "grad_norm": 1.75,
+      "learning_rate": 7.144285714285714e-06,
+      "loss": 0.1967,
+      "step": 9000
+    },
+    {
+      "epoch": 1.3061868000434924,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.13e-06,
+      "loss": 0.1873,
+      "step": 9010
+    },
+    {
+      "epoch": 1.307636548149759,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.1157142857142855e-06,
+      "loss": 0.1971,
+      "step": 9020
+    },
+    {
+      "epoch": 1.3090862962560255,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.101428571428572e-06,
+      "loss": 0.2125,
+      "step": 9030
+    },
+    {
+      "epoch": 1.310536044362292,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.087142857142858e-06,
+      "loss": 0.1959,
+      "step": 9040
+    },
+    {
+      "epoch": 1.3119857924685585,
+      "grad_norm": 1.75,
+      "learning_rate": 7.072857142857143e-06,
+      "loss": 0.212,
+      "step": 9050
+    },
+    {
+      "epoch": 1.313435540574825,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.058571428571429e-06,
+      "loss": 0.2028,
+      "step": 9060
+    },
+    {
+      "epoch": 1.3148852886810918,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.0442857142857145e-06,
+      "loss": 0.199,
+      "step": 9070
+    },
+    {
+      "epoch": 1.316335036787358,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.0300000000000005e-06,
+      "loss": 0.1944,
+      "step": 9080
+    },
+    {
+      "epoch": 1.3177847848936248,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.0157142857142864e-06,
+      "loss": 0.2057,
+      "step": 9090
+    },
+    {
+      "epoch": 1.3192345329998914,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.001428571428572e-06,
+      "loss": 0.2031,
+      "step": 9100
+    },
+    {
+      "epoch": 1.3206842811061579,
+      "grad_norm": 1.953125,
+      "learning_rate": 6.9871428571428575e-06,
+      "loss": 0.1974,
+      "step": 9110
+    },
+    {
+      "epoch": 1.3221340292124244,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.9728571428571435e-06,
+      "loss": 0.1987,
+      "step": 9120
+    },
+    {
+      "epoch": 1.323583777318691,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.958571428571429e-06,
+      "loss": 0.1904,
+      "step": 9130
+    },
+    {
+      "epoch": 1.3250335254249574,
+      "grad_norm": 1.96875,
+      "learning_rate": 6.944285714285715e-06,
+      "loss": 0.2172,
+      "step": 9140
+    },
+    {
+      "epoch": 1.326483273531224,
+      "grad_norm": 1.875,
+      "learning_rate": 6.93e-06,
+      "loss": 0.1912,
+      "step": 9150
+    },
+    {
+      "epoch": 1.3279330216374905,
+      "grad_norm": 1.375,
+      "learning_rate": 6.9157142857142865e-06,
+      "loss": 0.1815,
+      "step": 9160
+    },
+    {
+      "epoch": 1.329382769743757,
+      "grad_norm": 1.75,
+      "learning_rate": 6.9014285714285725e-06,
+      "loss": 0.2022,
+      "step": 9170
+    },
+    {
+      "epoch": 1.3308325178500235,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.887142857142858e-06,
+      "loss": 0.1815,
+      "step": 9180
+    },
+    {
+      "epoch": 1.33228226595629,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.872857142857144e-06,
+      "loss": 0.1919,
+      "step": 9190
+    },
+    {
+      "epoch": 1.3337320140625566,
+      "grad_norm": 1.9453125,
+      "learning_rate": 6.858571428571429e-06,
+      "loss": 0.2074,
+      "step": 9200
+    },
+    {
+      "epoch": 1.3351817621688231,
+      "grad_norm": 2.25,
+      "learning_rate": 6.844285714285715e-06,
+      "loss": 0.2182,
+      "step": 9210
+    },
+    {
+      "epoch": 1.3366315102750896,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.830000000000001e-06,
+      "loss": 0.1913,
+      "step": 9220
+    },
+    {
+      "epoch": 1.3380812583813562,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.815714285714287e-06,
+      "loss": 0.1937,
+      "step": 9230
+    },
+    {
+      "epoch": 1.339531006487623,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.801428571428572e-06,
+      "loss": 0.2092,
+      "step": 9240
+    },
+    {
+      "epoch": 1.3409807545938892,
+      "grad_norm": 1.828125,
+      "learning_rate": 6.787142857142858e-06,
+      "loss": 0.2057,
+      "step": 9250
+    },
+    {
+      "epoch": 1.342430502700156,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.772857142857143e-06,
+      "loss": 0.1917,
+      "step": 9260
+    },
+    {
+      "epoch": 1.3438802508064223,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.758571428571429e-06,
+      "loss": 0.1931,
+      "step": 9270
+    },
+    {
+      "epoch": 1.345329998912689,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.744285714285714e-06,
+      "loss": 0.2087,
+      "step": 9280
+    },
+    {
+      "epoch": 1.3467797470189555,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.730000000000001e-06,
+      "loss": 0.1963,
+      "step": 9290
+    },
+    {
+      "epoch": 1.348229495125222,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.715714285714287e-06,
+      "loss": 0.2073,
+      "step": 9300
+    },
+    {
+      "epoch": 1.3496792432314886,
+      "grad_norm": 1.5390625,
+      "learning_rate": 6.701428571428572e-06,
+      "loss": 0.196,
+      "step": 9310
+    },
+    {
+      "epoch": 1.351128991337755,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.687142857142858e-06,
+      "loss": 0.1795,
+      "step": 9320
+    },
+    {
+      "epoch": 1.3525787394440216,
+      "grad_norm": 2.0,
+      "learning_rate": 6.672857142857143e-06,
+      "loss": 0.1746,
+      "step": 9330
+    },
+    {
+      "epoch": 1.3540284875502882,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.658571428571429e-06,
+      "loss": 0.1972,
+      "step": 9340
+    },
+    {
+      "epoch": 1.3554782356565547,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.644285714285715e-06,
+      "loss": 0.1974,
+      "step": 9350
+    },
+    {
+      "epoch": 1.3569279837628212,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.630000000000001e-06,
+      "loss": 0.2013,
+      "step": 9360
+    },
+    {
+      "epoch": 1.3583777318690877,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.615714285714286e-06,
+      "loss": 0.1947,
+      "step": 9370
+    },
+    {
+      "epoch": 1.3598274799753542,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.601428571428572e-06,
+      "loss": 0.1809,
+      "step": 9380
+    },
+    {
+      "epoch": 1.3612772280816208,
+      "grad_norm": 1.8359375,
+      "learning_rate": 6.587142857142857e-06,
+      "loss": 0.2116,
+      "step": 9390
+    },
+    {
+      "epoch": 1.3627269761878873,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.572857142857143e-06,
+      "loss": 0.1874,
+      "step": 9400
+    },
+    {
+      "epoch": 1.3641767242941538,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.558571428571428e-06,
+      "loss": 0.2119,
+      "step": 9410
+    },
+    {
+      "epoch": 1.3656264724004203,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.544285714285715e-06,
+      "loss": 0.198,
+      "step": 9420
+    },
+    {
+      "epoch": 1.367076220506687,
+      "grad_norm": 1.453125,
+      "learning_rate": 6.530000000000001e-06,
+      "loss": 0.1865,
+      "step": 9430
+    },
+    {
+      "epoch": 1.3685259686129534,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.515714285714286e-06,
+      "loss": 0.2077,
+      "step": 9440
+    },
+    {
+      "epoch": 1.3699757167192201,
+      "grad_norm": 2.109375,
+      "learning_rate": 6.501428571428572e-06,
+      "loss": 0.2024,
+      "step": 9450
+    },
+    {
+      "epoch": 1.3714254648254864,
+      "grad_norm": 1.8671875,
+      "learning_rate": 6.487142857142857e-06,
+      "loss": 0.1963,
+      "step": 9460
+    },
+    {
+      "epoch": 1.3728752129317532,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.472857142857143e-06,
+      "loss": 0.1884,
+      "step": 9470
+    },
+    {
+      "epoch": 1.3743249610380197,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.45857142857143e-06,
+      "loss": 0.1947,
+      "step": 9480
+    },
+    {
+      "epoch": 1.3757747091442862,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.444285714285715e-06,
+      "loss": 0.197,
+      "step": 9490
+    },
+    {
+      "epoch": 1.3772244572505528,
+      "grad_norm": 1.5703125,
+      "learning_rate": 6.43e-06,
+      "loss": 0.2169,
+      "step": 9500
+    },
+    {
+      "epoch": 1.3786742053568193,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.415714285714286e-06,
+      "loss": 0.1946,
+      "step": 9510
+    },
+    {
+      "epoch": 1.3801239534630858,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.401428571428571e-06,
+      "loss": 0.195,
+      "step": 9520
+    },
+    {
+      "epoch": 1.3815737015693523,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.387142857142857e-06,
+      "loss": 0.1993,
+      "step": 9530
+    },
+    {
+      "epoch": 1.3830234496756189,
+      "grad_norm": 2.0625,
+      "learning_rate": 6.3728571428571425e-06,
+      "loss": 0.1851,
+      "step": 9540
+    },
+    {
+      "epoch": 1.3844731977818854,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.358571428571429e-06,
+      "loss": 0.1973,
+      "step": 9550
+    },
+    {
+      "epoch": 1.385922945888152,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.344285714285715e-06,
+      "loss": 0.197,
+      "step": 9560
+    },
+    {
+      "epoch": 1.3873726939944184,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.33e-06,
+      "loss": 0.1849,
+      "step": 9570
+    },
+    {
+      "epoch": 1.388822442100685,
+      "grad_norm": 1.8984375,
+      "learning_rate": 6.315714285714286e-06,
+      "loss": 0.2005,
+      "step": 9580
+    },
+    {
+      "epoch": 1.3902721902069515,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.3014285714285715e-06,
+      "loss": 0.2064,
+      "step": 9590
+    },
+    {
+      "epoch": 1.391721938313218,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.2871428571428574e-06,
+      "loss": 0.1922,
+      "step": 9600
+    },
+    {
+      "epoch": 1.3931716864194845,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.272857142857144e-06,
+      "loss": 0.1783,
+      "step": 9610
+    },
+    {
+      "epoch": 1.3946214345257513,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.258571428571429e-06,
+      "loss": 0.2021,
+      "step": 9620
+    },
+    {
+      "epoch": 1.3960711826320176,
+      "grad_norm": 1.5625,
+      "learning_rate": 6.244285714285715e-06,
+      "loss": 0.1798,
+      "step": 9630
+    },
+    {
+      "epoch": 1.3975209307382843,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.2300000000000005e-06,
+      "loss": 0.2034,
+      "step": 9640
+    },
+    {
+      "epoch": 1.3989706788445508,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.215714285714286e-06,
+      "loss": 0.1998,
+      "step": 9650
+    },
+    {
+      "epoch": 1.4004204269508174,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.201428571428572e-06,
+      "loss": 0.1818,
+      "step": 9660
+    },
+    {
+      "epoch": 1.401870175057084,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.187142857142857e-06,
+      "loss": 0.1831,
+      "step": 9670
+    },
+    {
+      "epoch": 1.4033199231633504,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.1728571428571435e-06,
+      "loss": 0.1889,
+      "step": 9680
+    },
+    {
+      "epoch": 1.404769671269617,
+      "grad_norm": 1.375,
+      "learning_rate": 6.1585714285714295e-06,
+      "loss": 0.2006,
+      "step": 9690
+    },
+    {
+      "epoch": 1.4062194193758835,
+      "grad_norm": 2.25,
+      "learning_rate": 6.144285714285715e-06,
+      "loss": 0.2139,
+      "step": 9700
+    },
+    {
+      "epoch": 1.40766916748215,
+      "grad_norm": 1.9453125,
+      "learning_rate": 6.130000000000001e-06,
+      "loss": 0.1858,
+      "step": 9710
+    },
+    {
+      "epoch": 1.4091189155884165,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.115714285714286e-06,
+      "loss": 0.1954,
+      "step": 9720
+    },
+    {
+      "epoch": 1.410568663694683,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.101428571428572e-06,
+      "loss": 0.2009,
+      "step": 9730
+    },
+    {
+      "epoch": 1.4120184118009496,
+      "grad_norm": 2.15625,
+      "learning_rate": 6.0871428571428585e-06,
+      "loss": 0.2043,
+      "step": 9740
+    },
+    {
+      "epoch": 1.413468159907216,
+      "grad_norm": 2.046875,
+      "learning_rate": 6.072857142857144e-06,
+      "loss": 0.1927,
+      "step": 9750
+    },
+    {
+      "epoch": 1.4149179080134826,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.05857142857143e-06,
+      "loss": 0.2042,
+      "step": 9760
+    },
+    {
+      "epoch": 1.4163676561197491,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.044285714285715e-06,
+      "loss": 0.1887,
+      "step": 9770
+    },
+    {
+      "epoch": 1.4178174042260157,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.030000000000001e-06,
+      "loss": 0.2028,
+      "step": 9780
+    },
+    {
+      "epoch": 1.4192671523322824,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.015714285714286e-06,
+      "loss": 0.1751,
+      "step": 9790
+    },
+    {
+      "epoch": 1.4207169004385487,
+      "grad_norm": 1.8984375,
+      "learning_rate": 6.001428571428571e-06,
+      "loss": 0.2099,
+      "step": 9800
+    },
+    {
+      "epoch": 1.4221666485448154,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.987142857142858e-06,
+      "loss": 0.2096,
+      "step": 9810
+    },
+    {
+      "epoch": 1.4236163966510818,
+      "grad_norm": 1.3359375,
+      "learning_rate": 5.972857142857144e-06,
+      "loss": 0.1966,
+      "step": 9820
+    },
+    {
+      "epoch": 1.4250661447573485,
+      "grad_norm": 1.625,
+      "learning_rate": 5.958571428571429e-06,
+      "loss": 0.1912,
+      "step": 9830
+    },
+    {
+      "epoch": 1.426515892863615,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.944285714285715e-06,
+      "loss": 0.1948,
+      "step": 9840
+    },
+    {
+      "epoch": 1.4279656409698815,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.93e-06,
+      "loss": 0.1898,
+      "step": 9850
+    },
+    {
+      "epoch": 1.429415389076148,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.915714285714286e-06,
+      "loss": 0.1866,
+      "step": 9860
+    },
+    {
+      "epoch": 1.4308651371824146,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.901428571428573e-06,
+      "loss": 0.185,
+      "step": 9870
+    },
+    {
+      "epoch": 1.4323148852886811,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.887142857142858e-06,
+      "loss": 0.1928,
+      "step": 9880
+    },
+    {
+      "epoch": 1.4337646333949476,
+      "grad_norm": 1.859375,
+      "learning_rate": 5.872857142857144e-06,
+      "loss": 0.1883,
+      "step": 9890
+    },
+    {
+      "epoch": 1.4352143815012142,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.858571428571429e-06,
+      "loss": 0.1909,
+      "step": 9900
+    },
+    {
+      "epoch": 1.4366641296074807,
+      "grad_norm": 1.53125,
+      "learning_rate": 5.844285714285715e-06,
+      "loss": 0.1939,
+      "step": 9910
+    },
+    {
+      "epoch": 1.4381138777137472,
+      "grad_norm": 1.9609375,
+      "learning_rate": 5.83e-06,
+      "loss": 0.2059,
+      "step": 9920
+    },
+    {
+      "epoch": 1.4395636258200137,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.815714285714286e-06,
+      "loss": 0.1977,
+      "step": 9930
+    },
+    {
+      "epoch": 1.4410133739262803,
+      "grad_norm": 1.8515625,
+      "learning_rate": 5.801428571428572e-06,
+      "loss": 0.189,
+      "step": 9940
+    },
+    {
+      "epoch": 1.4424631220325468,
+      "grad_norm": 1.9921875,
+      "learning_rate": 5.787142857142858e-06,
+      "loss": 0.189,
+      "step": 9950
+    },
+    {
+      "epoch": 1.4439128701388133,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.772857142857143e-06,
+      "loss": 0.1938,
+      "step": 9960
+    },
+    {
+      "epoch": 1.4453626182450798,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.758571428571429e-06,
+      "loss": 0.1961,
+      "step": 9970
+    },
+    {
+      "epoch": 1.4468123663513466,
+      "grad_norm": 1.8671875,
+      "learning_rate": 5.744285714285714e-06,
+      "loss": 0.1976,
+      "step": 9980
+    },
+    {
+      "epoch": 1.4482621144576129,
+      "grad_norm": 1.921875,
+      "learning_rate": 5.73e-06,
+      "loss": 0.1851,
+      "step": 9990
+    },
+    {
+      "epoch": 1.4497118625638796,
+      "grad_norm": 1.625,
+      "learning_rate": 5.715714285714287e-06,
+      "loss": 0.1888,
+      "step": 10000
+    },
+    {
+      "epoch": 1.451161610670146,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.701428571428572e-06,
+      "loss": 0.1965,
+      "step": 10010
+    },
+    {
+      "epoch": 1.4526113587764127,
+      "grad_norm": 1.8359375,
+      "learning_rate": 5.687142857142858e-06,
+      "loss": 0.1991,
+      "step": 10020
+    },
+    {
+      "epoch": 1.4540611068826792,
+      "grad_norm": 1.4609375,
+      "learning_rate": 5.672857142857143e-06,
+      "loss": 0.2002,
+      "step": 10030
+    },
+    {
+      "epoch": 1.4555108549889457,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.658571428571429e-06,
+      "loss": 0.2017,
+      "step": 10040
+    },
+    {
+      "epoch": 1.4569606030952122,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.644285714285714e-06,
+      "loss": 0.2051,
+      "step": 10050
+    },
+    {
+      "epoch": 1.4584103512014788,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.63e-06,
+      "loss": 0.1849,
+      "step": 10060
+    },
+    {
+      "epoch": 1.4598600993077453,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.615714285714286e-06,
+      "loss": 0.1869,
+      "step": 10070
+    },
+    {
+      "epoch": 1.4613098474140118,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.601428571428572e-06,
+      "loss": 0.2027,
+      "step": 10080
+    },
+    {
+      "epoch": 1.4627595955202783,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.587142857142857e-06,
+      "loss": 0.1982,
+      "step": 10090
+    },
+    {
+      "epoch": 1.4642093436265449,
+      "grad_norm": 1.8828125,
+      "learning_rate": 5.572857142857143e-06,
+      "loss": 0.1786,
+      "step": 10100
+    },
+    {
+      "epoch": 1.4656590917328114,
+      "grad_norm": 2.0,
+      "learning_rate": 5.5585714285714285e-06,
+      "loss": 0.1912,
+      "step": 10110
+    },
+    {
+      "epoch": 1.467108839839078,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.5442857142857144e-06,
+      "loss": 0.1842,
+      "step": 10120
+    },
+    {
+      "epoch": 1.4685585879453444,
+      "grad_norm": 1.8515625,
+      "learning_rate": 5.530000000000001e-06,
+      "loss": 0.2038,
+      "step": 10130
+    },
+    {
+      "epoch": 1.470008336051611,
+      "grad_norm": 1.4921875,
+      "learning_rate": 5.515714285714286e-06,
+      "loss": 0.196,
+      "step": 10140
+    },
+    {
+      "epoch": 1.4714580841578775,
+      "grad_norm": 1.75,
+      "learning_rate": 5.501428571428572e-06,
+      "loss": 0.1826,
+      "step": 10150
+    },
+    {
+      "epoch": 1.472907832264144,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.4871428571428575e-06,
+      "loss": 0.1929,
+      "step": 10160
+    },
+    {
+      "epoch": 1.4743575803704108,
+      "grad_norm": 1.8828125,
+      "learning_rate": 5.4728571428571435e-06,
+      "loss": 0.1973,
+      "step": 10170
+    },
+    {
+      "epoch": 1.475807328476677,
+      "grad_norm": 1.875,
+      "learning_rate": 5.458571428571429e-06,
+      "loss": 0.1733,
+      "step": 10180
+    },
+    {
+      "epoch": 1.4772570765829438,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.444285714285715e-06,
+      "loss": 0.1881,
+      "step": 10190
+    },
+    {
+      "epoch": 1.4787068246892103,
+      "grad_norm": 2.078125,
+      "learning_rate": 5.4300000000000005e-06,
+      "loss": 0.1863,
+      "step": 10200
+    },
+    {
+      "epoch": 1.4801565727954769,
+      "grad_norm": 1.90625,
+      "learning_rate": 5.4157142857142865e-06,
+      "loss": 0.2044,
+      "step": 10210
+    },
+    {
+      "epoch": 1.4816063209017434,
+      "grad_norm": 1.9453125,
+      "learning_rate": 5.401428571428572e-06,
+      "loss": 0.1976,
+      "step": 10220
+    },
+    {
+      "epoch": 1.48305606900801,
+      "grad_norm": 1.828125,
+      "learning_rate": 5.387142857142858e-06,
+      "loss": 0.1965,
+      "step": 10230
+    },
+    {
+      "epoch": 1.4845058171142764,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.372857142857143e-06,
+      "loss": 0.1809,
+      "step": 10240
+    },
+    {
+      "epoch": 1.485955565220543,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.358571428571429e-06,
+      "loss": 0.1933,
+      "step": 10250
+    },
+    {
+      "epoch": 1.4874053133268095,
+      "grad_norm": 1.5546875,
+      "learning_rate": 5.3442857142857155e-06,
+      "loss": 0.2007,
+      "step": 10260
+    },
+    {
+      "epoch": 1.488855061433076,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.330000000000001e-06,
+      "loss": 0.1895,
+      "step": 10270
+    },
+    {
+      "epoch": 1.4903048095393425,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.315714285714287e-06,
+      "loss": 0.2075,
+      "step": 10280
+    },
+    {
+      "epoch": 1.491754557645609,
+      "grad_norm": 1.8515625,
+      "learning_rate": 5.301428571428572e-06,
+      "loss": 0.1971,
+      "step": 10290
+    },
+    {
+      "epoch": 1.4932043057518756,
+      "grad_norm": 2.0625,
+      "learning_rate": 5.287142857142858e-06,
+      "loss": 0.2043,
+      "step": 10300
+    },
+    {
+      "epoch": 1.494654053858142,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.272857142857143e-06,
+      "loss": 0.2001,
+      "step": 10310
+    },
+    {
+      "epoch": 1.4961038019644086,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.25857142857143e-06,
+      "loss": 0.1905,
+      "step": 10320
+    },
+    {
+      "epoch": 1.4975535500706751,
+      "grad_norm": 1.625,
+      "learning_rate": 5.244285714285715e-06,
+      "loss": 0.2038,
+      "step": 10330
+    },
+    {
+      "epoch": 1.499003298176942,
+      "grad_norm": 2.15625,
+      "learning_rate": 5.230000000000001e-06,
+      "loss": 0.1861,
+      "step": 10340
+    },
+    {
+      "epoch": 1.5004530462832082,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.215714285714286e-06,
+      "loss": 0.1979,
+      "step": 10350
+    },
+    {
+      "epoch": 1.501902794389475,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.201428571428572e-06,
+      "loss": 0.1993,
+      "step": 10360
+    },
+    {
+      "epoch": 1.5033525424957412,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.187142857142857e-06,
+      "loss": 0.1899,
+      "step": 10370
+    },
+    {
+      "epoch": 1.504802290602008,
+      "grad_norm": 1.875,
+      "learning_rate": 5.172857142857143e-06,
+      "loss": 0.1841,
+      "step": 10380
+    },
+    {
+      "epoch": 1.5062520387082743,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.15857142857143e-06,
+      "loss": 0.2045,
+      "step": 10390
+    },
+    {
+      "epoch": 1.507701786814541,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.144285714285715e-06,
+      "loss": 0.1876,
+      "step": 10400
+    },
+    {
+      "epoch": 1.5091515349208076,
+      "grad_norm": 1.9921875,
+      "learning_rate": 5.130000000000001e-06,
+      "loss": 0.1798,
+      "step": 10410
+    },
+    {
+      "epoch": 1.510601283027074,
+      "grad_norm": 2.125,
+      "learning_rate": 5.115714285714286e-06,
+      "loss": 0.2032,
+      "step": 10420
+    },
+    {
+      "epoch": 1.5120510311333406,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.101428571428572e-06,
+      "loss": 0.2015,
+      "step": 10430
+    },
+    {
+      "epoch": 1.5135007792396071,
+      "grad_norm": 1.953125,
+      "learning_rate": 5.087142857142857e-06,
+      "loss": 0.2089,
+      "step": 10440
+    },
+    {
+      "epoch": 1.5149505273458737,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.072857142857144e-06,
+      "loss": 0.1944,
+      "step": 10450
+    },
+    {
+      "epoch": 1.5164002754521402,
+      "grad_norm": 1.296875,
+      "learning_rate": 5.058571428571429e-06,
+      "loss": 0.1787,
+      "step": 10460
+    },
+    {
+      "epoch": 1.5178500235584067,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.044285714285715e-06,
+      "loss": 0.1786,
+      "step": 10470
+    },
+    {
+      "epoch": 1.5192997716646732,
+      "grad_norm": 1.9453125,
+      "learning_rate": 5.03e-06,
+      "loss": 0.1985,
+      "step": 10480
+    },
+    {
+      "epoch": 1.5207495197709398,
+      "grad_norm": 1.5234375,
+      "learning_rate": 5.015714285714286e-06,
+      "loss": 0.1981,
+      "step": 10490
+    },
+    {
+      "epoch": 1.5221992678772063,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.001428571428571e-06,
+      "loss": 0.1997,
+      "step": 10500
+    },
+    {
+      "epoch": 1.523649015983473,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.987142857142858e-06,
+      "loss": 0.1991,
+      "step": 10510
+    },
+    {
+      "epoch": 1.5250987640897393,
+      "grad_norm": 2.234375,
+      "learning_rate": 4.972857142857143e-06,
+      "loss": 0.1967,
+      "step": 10520
+    },
+    {
+      "epoch": 1.526548512196006,
+      "grad_norm": 1.7578125,
+      "learning_rate": 4.958571428571428e-06,
+      "loss": 0.2023,
+      "step": 10530
+    },
+    {
+      "epoch": 1.5279982603022724,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.944285714285715e-06,
+      "loss": 0.1885,
+      "step": 10540
+    },
+    {
+      "epoch": 1.5294480084085391,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.93e-06,
+      "loss": 0.2025,
+      "step": 10550
+    },
+    {
+      "epoch": 1.5308977565148054,
+      "grad_norm": 1.953125,
+      "learning_rate": 4.915714285714286e-06,
+      "loss": 0.1975,
+      "step": 10560
+    },
+    {
+      "epoch": 1.5323475046210722,
+      "grad_norm": 1.7109375,
+      "learning_rate": 4.901428571428572e-06,
+      "loss": 0.205,
+      "step": 10570
+    },
+    {
+      "epoch": 1.5337972527273385,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.887142857142857e-06,
+      "loss": 0.1963,
+      "step": 10580
+    },
+    {
+      "epoch": 1.5352470008336052,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.872857142857143e-06,
+      "loss": 0.2013,
+      "step": 10590
+    },
+    {
+      "epoch": 1.5366967489398717,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.858571428571429e-06,
+      "loss": 0.2025,
+      "step": 10600
+    },
+    {
+      "epoch": 1.5381464970461383,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.844285714285714e-06,
+      "loss": 0.1949,
+      "step": 10610
+    },
+    {
+      "epoch": 1.5395962451524048,
+      "grad_norm": 1.703125,
+      "learning_rate": 4.83e-06,
+      "loss": 0.1804,
+      "step": 10620
+    },
+    {
+      "epoch": 1.5410459932586713,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.815714285714286e-06,
+      "loss": 0.1937,
+      "step": 10630
+    },
+    {
+      "epoch": 1.5424957413649378,
+      "grad_norm": 1.75,
+      "learning_rate": 4.801428571428572e-06,
+      "loss": 0.193,
+      "step": 10640
+    },
+    {
+      "epoch": 1.5439454894712044,
+      "grad_norm": 2.03125,
+      "learning_rate": 4.787142857142857e-06,
+      "loss": 0.185,
+      "step": 10650
+    },
+    {
+      "epoch": 1.5453952375774709,
+      "grad_norm": 2.265625,
+      "learning_rate": 4.772857142857143e-06,
+      "loss": 0.1887,
+      "step": 10660
+    },
+    {
+      "epoch": 1.5468449856837374,
+      "grad_norm": 1.8125,
+      "learning_rate": 4.758571428571429e-06,
+      "loss": 0.1984,
+      "step": 10670
+    },
+    {
+      "epoch": 1.5482947337900042,
+      "grad_norm": 1.625,
+      "learning_rate": 4.7442857142857145e-06,
+      "loss": 0.1971,
+      "step": 10680
+    },
+    {
+      "epoch": 1.5497444818962705,
+      "grad_norm": 1.765625,
+      "learning_rate": 4.7300000000000005e-06,
+      "loss": 0.1785,
+      "step": 10690
+    },
+    {
+      "epoch": 1.5511942300025372,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.7157142857142864e-06,
+      "loss": 0.1983,
+      "step": 10700
+    },
+    {
+      "epoch": 1.5526439781088035,
+      "grad_norm": 1.765625,
+      "learning_rate": 4.7014285714285716e-06,
+      "loss": 0.1954,
+      "step": 10710
+    },
+    {
+      "epoch": 1.5540937262150702,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.6871428571428575e-06,
+      "loss": 0.1869,
+      "step": 10720
+    },
+    {
+      "epoch": 1.5555434743213366,
+      "grad_norm": 1.953125,
+      "learning_rate": 4.6728571428571435e-06,
+      "loss": 0.2108,
+      "step": 10730
+    },
+    {
+      "epoch": 1.5569932224276033,
+      "grad_norm": 1.625,
+      "learning_rate": 4.658571428571429e-06,
+      "loss": 0.1925,
+      "step": 10740
+    },
+    {
+      "epoch": 1.5584429705338696,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.644285714285715e-06,
+      "loss": 0.1821,
+      "step": 10750
+    },
+    {
+      "epoch": 1.5598927186401363,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.6300000000000006e-06,
+      "loss": 0.1672,
+      "step": 10760
+    },
+    {
+      "epoch": 1.5613424667464026,
+      "grad_norm": 1.7734375,
+      "learning_rate": 4.6157142857142865e-06,
+      "loss": 0.2089,
+      "step": 10770
+    },
+    {
+      "epoch": 1.5627922148526694,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.601428571428572e-06,
+      "loss": 0.1892,
+      "step": 10780
+    },
+    {
+      "epoch": 1.564241962958936,
+      "grad_norm": 1.6484375,
+      "learning_rate": 4.587142857142858e-06,
+      "loss": 0.194,
+      "step": 10790
+    },
+    {
+      "epoch": 1.5656917110652024,
+      "grad_norm": 1.875,
+      "learning_rate": 4.572857142857144e-06,
+      "loss": 0.229,
+      "step": 10800
+    },
+    {
+      "epoch": 1.567141459171469,
+      "grad_norm": 1.75,
+      "learning_rate": 4.558571428571429e-06,
+      "loss": 0.1851,
+      "step": 10810
+    },
+    {
+      "epoch": 1.5685912072777355,
+      "grad_norm": 1.7890625,
+      "learning_rate": 4.544285714285715e-06,
+      "loss": 0.2125,
+      "step": 10820
+    },
+    {
+      "epoch": 1.570040955384002,
+      "grad_norm": 1.5625,
+      "learning_rate": 4.530000000000001e-06,
+      "loss": 0.18,
+      "step": 10830
+    },
+    {
+      "epoch": 1.5714907034902685,
+      "grad_norm": 2.1875,
+      "learning_rate": 4.515714285714286e-06,
+      "loss": 0.1781,
+      "step": 10840
+    },
+    {
+      "epoch": 1.572940451596535,
+      "grad_norm": 1.953125,
+      "learning_rate": 4.501428571428572e-06,
+      "loss": 0.1838,
+      "step": 10850
+    },
+    {
+      "epoch": 1.5743901997028016,
+      "grad_norm": 1.7734375,
+      "learning_rate": 4.487142857142858e-06,
+      "loss": 0.1987,
+      "step": 10860
+    },
+    {
+      "epoch": 1.5758399478090683,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.472857142857143e-06,
+      "loss": 0.1897,
+      "step": 10870
+    },
+    {
+      "epoch": 1.5772896959153346,
+      "grad_norm": 1.8671875,
+      "learning_rate": 4.458571428571429e-06,
+      "loss": 0.1774,
+      "step": 10880
+    },
+    {
+      "epoch": 1.5787394440216014,
+      "grad_norm": 1.78125,
+      "learning_rate": 4.444285714285715e-06,
+      "loss": 0.193,
+      "step": 10890
+    },
+    {
+      "epoch": 1.5801891921278677,
+      "grad_norm": 1.5703125,
+      "learning_rate": 4.430000000000001e-06,
+      "loss": 0.1844,
+      "step": 10900
+    },
+    {
+      "epoch": 1.5816389402341344,
+      "grad_norm": 1.78125,
+      "learning_rate": 4.415714285714286e-06,
+      "loss": 0.1907,
+      "step": 10910
+    },
+    {
+      "epoch": 1.5830886883404007,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.401428571428572e-06,
+      "loss": 0.1881,
+      "step": 10920
+    },
+    {
+      "epoch": 1.5845384364466675,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4.387142857142858e-06,
+      "loss": 0.1909,
+      "step": 10930
+    },
+    {
+      "epoch": 1.5859881845529338,
+      "grad_norm": 1.875,
+      "learning_rate": 4.372857142857143e-06,
+      "loss": 0.1869,
+      "step": 10940
+    },
+    {
+      "epoch": 1.5874379326592005,
+      "grad_norm": 1.7890625,
+      "learning_rate": 4.358571428571429e-06,
+      "loss": 0.2021,
+      "step": 10950
+    },
+    {
+      "epoch": 1.588887680765467,
+      "grad_norm": 1.734375,
+      "learning_rate": 4.344285714285715e-06,
+      "loss": 0.1966,
+      "step": 10960
+    },
+    {
+      "epoch": 1.5903374288717336,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.33e-06,
+      "loss": 0.1806,
+      "step": 10970
+    },
+    {
+      "epoch": 1.591787176978,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.315714285714286e-06,
+      "loss": 0.1946,
+      "step": 10980
+    },
+    {
+      "epoch": 1.5932369250842666,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.301428571428572e-06,
+      "loss": 0.2043,
+      "step": 10990
+    },
+    {
+      "epoch": 1.5946866731905331,
+      "grad_norm": 1.71875,
+      "learning_rate": 4.287142857142857e-06,
+      "loss": 0.1878,
+      "step": 11000
+    },
+    {
+      "epoch": 1.5961364212967997,
+      "grad_norm": 1.65625,
+      "learning_rate": 4.272857142857143e-06,
+      "loss": 0.2184,
+      "step": 11010
+    },
+    {
+      "epoch": 1.5975861694030662,
+      "grad_norm": 1.9453125,
+      "learning_rate": 4.258571428571429e-06,
+      "loss": 0.1944,
+      "step": 11020
+    },
+    {
+      "epoch": 1.5990359175093327,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.244285714285715e-06,
+      "loss": 0.204,
+      "step": 11030
+    },
+    {
+      "epoch": 1.6004856656155992,
+      "grad_norm": 1.6484375,
+      "learning_rate": 4.23e-06,
+      "loss": 0.1937,
+      "step": 11040
+    },
+    {
+      "epoch": 1.6019354137218658,
+      "grad_norm": 1.90625,
+      "learning_rate": 4.215714285714286e-06,
+      "loss": 0.1957,
+      "step": 11050
+    },
+    {
+      "epoch": 1.6033851618281325,
+      "grad_norm": 1.7109375,
+      "learning_rate": 4.201428571428572e-06,
+      "loss": 0.1958,
+      "step": 11060
+    },
+    {
+      "epoch": 1.6048349099343988,
+      "grad_norm": 1.640625,
+      "learning_rate": 4.187142857142857e-06,
+      "loss": 0.185,
+      "step": 11070
+    },
+    {
+      "epoch": 1.6062846580406656,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.172857142857143e-06,
+      "loss": 0.1994,
+      "step": 11080
+    },
+    {
+      "epoch": 1.6077344061469319,
+      "grad_norm": 2.125,
+      "learning_rate": 4.158571428571429e-06,
+      "loss": 0.1804,
+      "step": 11090
+    },
+    {
+      "epoch": 1.6091841542531986,
+      "grad_norm": 1.703125,
+      "learning_rate": 4.144285714285714e-06,
+      "loss": 0.1928,
+      "step": 11100
+    },
+    {
+      "epoch": 1.610633902359465,
+      "grad_norm": 1.6875,
+      "learning_rate": 4.13e-06,
+      "loss": 0.1947,
+      "step": 11110
+    },
+    {
+      "epoch": 1.6120836504657317,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.115714285714286e-06,
+      "loss": 0.1921,
+      "step": 11120
+    },
+    {
+      "epoch": 1.613533398571998,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.101428571428571e-06,
+      "loss": 0.1902,
+      "step": 11130
+    },
+    {
+      "epoch": 1.6149831466782647,
+      "grad_norm": 1.7421875,
+      "learning_rate": 4.087142857142857e-06,
+      "loss": 0.1951,
+      "step": 11140
+    },
+    {
+      "epoch": 1.6164328947845312,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.072857142857143e-06,
+      "loss": 0.1891,
+      "step": 11150
+    },
+    {
+      "epoch": 1.6178826428907978,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.058571428571429e-06,
+      "loss": 0.1913,
+      "step": 11160
+    },
+    {
+      "epoch": 1.6193323909970643,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.044285714285714e-06,
+      "loss": 0.18,
+      "step": 11170
+    },
+    {
+      "epoch": 1.6207821391033308,
+      "grad_norm": 2.125,
+      "learning_rate": 4.03e-06,
+      "loss": 0.1828,
+      "step": 11180
+    },
+    {
+      "epoch": 1.6222318872095973,
+      "grad_norm": 1.6015625,
+      "learning_rate": 4.015714285714286e-06,
+      "loss": 0.1816,
+      "step": 11190
+    },
+    {
+      "epoch": 1.6236816353158638,
+      "grad_norm": 1.8515625,
+      "learning_rate": 4.0014285714285715e-06,
+      "loss": 0.1877,
+      "step": 11200
+    },
+    {
+      "epoch": 1.6251313834221304,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.9871428571428574e-06,
+      "loss": 0.1887,
+      "step": 11210
+    },
+    {
+      "epoch": 1.626581131528397,
+      "grad_norm": 1.8203125,
+      "learning_rate": 3.972857142857143e-06,
+      "loss": 0.2062,
+      "step": 11220
+    },
+    {
+      "epoch": 1.6280308796346636,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.9585714285714285e-06,
+      "loss": 0.184,
+      "step": 11230
+    },
+    {
+      "epoch": 1.62948062774093,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.9442857142857145e-06,
+      "loss": 0.1899,
+      "step": 11240
+    },
+    {
+      "epoch": 1.6309303758471967,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.9300000000000005e-06,
+      "loss": 0.1977,
+      "step": 11250
+    },
+    {
+      "epoch": 1.632380123953463,
+      "grad_norm": 1.7734375,
+      "learning_rate": 3.915714285714286e-06,
+      "loss": 0.19,
+      "step": 11260
+    },
+    {
+      "epoch": 1.6338298720597297,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.901428571428572e-06,
+      "loss": 0.1975,
+      "step": 11270
+    },
+    {
+      "epoch": 1.635279620165996,
+      "grad_norm": 1.4296875,
+      "learning_rate": 3.8871428571428576e-06,
+      "loss": 0.2074,
+      "step": 11280
+    },
+    {
+      "epoch": 1.6367293682722628,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.8728571428571435e-06,
+      "loss": 0.1913,
+      "step": 11290
+    },
+    {
+      "epoch": 1.638179116378529,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.858571428571429e-06,
+      "loss": 0.1897,
+      "step": 11300
+    },
+    {
+      "epoch": 1.6396288644847958,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.844285714285715e-06,
+      "loss": 0.1863,
+      "step": 11310
+    },
+    {
+      "epoch": 1.6410786125910624,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.830000000000001e-06,
+      "loss": 0.1946,
+      "step": 11320
+    },
+    {
+      "epoch": 1.6425283606973289,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.815714285714286e-06,
+      "loss": 0.2067,
+      "step": 11330
+    },
+    {
+      "epoch": 1.6439781088035954,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.8014285714285713e-06,
+      "loss": 0.1972,
+      "step": 11340
+    },
+    {
+      "epoch": 1.645427856909862,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.7871428571428577e-06,
+      "loss": 0.1893,
+      "step": 11350
+    },
+    {
+      "epoch": 1.6468776050161285,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.7728571428571432e-06,
+      "loss": 0.1916,
+      "step": 11360
+    },
+    {
+      "epoch": 1.648327353122395,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.7585714285714288e-06,
+      "loss": 0.1944,
+      "step": 11370
+    },
+    {
+      "epoch": 1.6497771012286615,
+      "grad_norm": 1.5390625,
+      "learning_rate": 3.7442857142857147e-06,
+      "loss": 0.1918,
+      "step": 11380
+    },
+    {
+      "epoch": 1.651226849334928,
+      "grad_norm": 2.125,
+      "learning_rate": 3.7300000000000003e-06,
+      "loss": 0.186,
+      "step": 11390
+    },
+    {
+      "epoch": 1.6526765974411946,
+      "grad_norm": 1.8046875,
+      "learning_rate": 3.715714285714286e-06,
+      "loss": 0.1984,
+      "step": 11400
+    },
+    {
+      "epoch": 1.654126345547461,
+      "grad_norm": 1.6953125,
+      "learning_rate": 3.701428571428572e-06,
+      "loss": 0.1865,
+      "step": 11410
+    },
+    {
+      "epoch": 1.6555760936537278,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.6871428571428574e-06,
+      "loss": 0.1878,
+      "step": 11420
+    },
+    {
+      "epoch": 1.6570258417599941,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.672857142857143e-06,
+      "loss": 0.1947,
+      "step": 11430
+    },
+    {
+      "epoch": 1.6584755898662609,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.6585714285714293e-06,
+      "loss": 0.1914,
+      "step": 11440
+    },
+    {
+      "epoch": 1.6599253379725272,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.644285714285715e-06,
+      "loss": 0.2029,
+      "step": 11450
+    },
+    {
+      "epoch": 1.661375086078794,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.6300000000000004e-06,
+      "loss": 0.2035,
+      "step": 11460
+    },
+    {
+      "epoch": 1.6628248341850602,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.6157142857142864e-06,
+      "loss": 0.1953,
+      "step": 11470
+    },
+    {
+      "epoch": 1.664274582291327,
+      "grad_norm": 1.765625,
+      "learning_rate": 3.601428571428572e-06,
+      "loss": 0.1908,
+      "step": 11480
+    },
+    {
+      "epoch": 1.6657243303975933,
+      "grad_norm": 2.125,
+      "learning_rate": 3.5871428571428575e-06,
+      "loss": 0.1782,
+      "step": 11490
+    },
+    {
+      "epoch": 1.66717407850386,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.572857142857143e-06,
+      "loss": 0.1854,
+      "step": 11500
+    },
+    {
+      "epoch": 1.6686238266101265,
+      "grad_norm": 1.875,
+      "learning_rate": 3.558571428571429e-06,
+      "loss": 0.1912,
+      "step": 11510
+    },
+    {
+      "epoch": 1.670073574716393,
+      "grad_norm": 1.7265625,
+      "learning_rate": 3.5442857142857145e-06,
+      "loss": 0.2005,
+      "step": 11520
+    },
+    {
+      "epoch": 1.6715233228226596,
+      "grad_norm": 2.078125,
+      "learning_rate": 3.53e-06,
+      "loss": 0.2009,
+      "step": 11530
+    },
+    {
+      "epoch": 1.672973070928926,
+      "grad_norm": 1.8203125,
+      "learning_rate": 3.515714285714286e-06,
+      "loss": 0.1834,
+      "step": 11540
+    },
+    {
+      "epoch": 1.6744228190351926,
+      "grad_norm": 2.046875,
+      "learning_rate": 3.5014285714285716e-06,
+      "loss": 0.1953,
+      "step": 11550
+    },
+    {
+      "epoch": 1.6758725671414592,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.487142857142857e-06,
+      "loss": 0.1964,
+      "step": 11560
+    },
+    {
+      "epoch": 1.6773223152477257,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.4728571428571435e-06,
+      "loss": 0.1983,
+      "step": 11570
+    },
+    {
+      "epoch": 1.6787720633539922,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.458571428571429e-06,
+      "loss": 0.1865,
+      "step": 11580
+    },
+    {
+      "epoch": 1.6802218114602587,
+      "grad_norm": 1.515625,
+      "learning_rate": 3.4442857142857146e-06,
+      "loss": 0.176,
+      "step": 11590
+    },
+    {
+      "epoch": 1.6816715595665253,
+      "grad_norm": 2.0,
+      "learning_rate": 3.4300000000000006e-06,
+      "loss": 0.1747,
+      "step": 11600
+    },
+    {
+      "epoch": 1.683121307672792,
+      "grad_norm": 1.765625,
+      "learning_rate": 3.415714285714286e-06,
+      "loss": 0.2012,
+      "step": 11610
+    },
+    {
+      "epoch": 1.6845710557790583,
+      "grad_norm": 2.21875,
+      "learning_rate": 3.4014285714285717e-06,
+      "loss": 0.1767,
+      "step": 11620
+    },
+    {
+      "epoch": 1.686020803885325,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.3871428571428573e-06,
+      "loss": 0.2058,
+      "step": 11630
+    },
+    {
+      "epoch": 1.6874705519915913,
+      "grad_norm": 1.6328125,
+      "learning_rate": 3.3728571428571432e-06,
+      "loss": 0.182,
+      "step": 11640
+    },
+    {
+      "epoch": 1.688920300097858,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.3585714285714288e-06,
+      "loss": 0.1874,
+      "step": 11650
+    },
+    {
+      "epoch": 1.6903700482041244,
+      "grad_norm": 1.9765625,
+      "learning_rate": 3.3442857142857143e-06,
+      "loss": 0.1963,
+      "step": 11660
+    },
+    {
+      "epoch": 1.6918197963103911,
+      "grad_norm": 1.9609375,
+      "learning_rate": 3.3300000000000003e-06,
+      "loss": 0.1982,
+      "step": 11670
+    },
+    {
+      "epoch": 1.6932695444166574,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.315714285714286e-06,
+      "loss": 0.1945,
+      "step": 11680
+    },
+    {
+      "epoch": 1.6947192925229242,
+      "grad_norm": 1.703125,
+      "learning_rate": 3.3014285714285714e-06,
+      "loss": 0.1898,
+      "step": 11690
+    },
+    {
+      "epoch": 1.6961690406291907,
+      "grad_norm": 1.6796875,
+      "learning_rate": 3.287142857142858e-06,
+      "loss": 0.1837,
+      "step": 11700
+    },
+    {
+      "epoch": 1.6976187887354572,
+      "grad_norm": 1.765625,
+      "learning_rate": 3.2728571428571433e-06,
+      "loss": 0.19,
+      "step": 11710
+    },
+    {
+      "epoch": 1.6990685368417238,
+      "grad_norm": 1.84375,
+      "learning_rate": 3.258571428571429e-06,
+      "loss": 0.1964,
+      "step": 11720
+    },
+    {
+      "epoch": 1.7005182849479903,
+      "grad_norm": 1.6015625,
+      "learning_rate": 3.244285714285715e-06,
+      "loss": 0.2012,
+      "step": 11730
+    },
+    {
+      "epoch": 1.7019680330542568,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.2300000000000004e-06,
+      "loss": 0.1843,
+      "step": 11740
+    },
+    {
+      "epoch": 1.7034177811605233,
+      "grad_norm": 1.921875,
+      "learning_rate": 3.215714285714286e-06,
+      "loss": 0.1966,
+      "step": 11750
+    },
+    {
+      "epoch": 1.7048675292667899,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3.2014285714285715e-06,
+      "loss": 0.1888,
+      "step": 11760
+    },
+    {
+      "epoch": 1.7063172773730564,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.1871428571428575e-06,
+      "loss": 0.1979,
+      "step": 11770
+    },
+    {
+      "epoch": 1.7077670254793231,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.172857142857143e-06,
+      "loss": 0.1699,
+      "step": 11780
+    },
+    {
+      "epoch": 1.7092167735855894,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.1585714285714286e-06,
+      "loss": 0.188,
+      "step": 11790
+    },
+    {
+      "epoch": 1.7106665216918562,
+      "grad_norm": 1.71875,
+      "learning_rate": 3.1442857142857146e-06,
+      "loss": 0.1806,
+      "step": 11800
+    },
+    {
+      "epoch": 1.7121162697981225,
+      "grad_norm": 1.5,
+      "learning_rate": 3.13e-06,
+      "loss": 0.1776,
+      "step": 11810
+    },
+    {
+      "epoch": 1.7135660179043892,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.1157142857142857e-06,
+      "loss": 0.168,
+      "step": 11820
+    },
+    {
+      "epoch": 1.7150157660106555,
+      "grad_norm": 1.984375,
+      "learning_rate": 3.101428571428572e-06,
+      "loss": 0.1908,
+      "step": 11830
+    },
+    {
+      "epoch": 1.7164655141169223,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.0871428571428576e-06,
+      "loss": 0.1821,
+      "step": 11840
+    },
+    {
+      "epoch": 1.7179152622231886,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.072857142857143e-06,
+      "loss": 0.1978,
+      "step": 11850
+    },
+    {
+      "epoch": 1.7193650103294553,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.058571428571429e-06,
+      "loss": 0.1953,
+      "step": 11860
+    },
+    {
+      "epoch": 1.7208147584357218,
+      "grad_norm": 1.8515625,
+      "learning_rate": 3.0442857142857147e-06,
+      "loss": 0.2031,
+      "step": 11870
+    },
+    {
+      "epoch": 1.7222645065419884,
+      "grad_norm": 1.484375,
+      "learning_rate": 3.0300000000000002e-06,
+      "loss": 0.1895,
+      "step": 11880
+    },
+    {
+      "epoch": 1.723714254648255,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.0157142857142858e-06,
+      "loss": 0.1923,
+      "step": 11890
+    },
+    {
+      "epoch": 1.7251640027545214,
+      "grad_norm": 2.09375,
+      "learning_rate": 3.0014285714285717e-06,
+      "loss": 0.1818,
+      "step": 11900
+    },
+    {
+      "epoch": 1.726613750860788,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.9871428571428573e-06,
+      "loss": 0.2045,
+      "step": 11910
+    },
+    {
+      "epoch": 1.7280634989670545,
+      "grad_norm": 2.125,
+      "learning_rate": 2.972857142857143e-06,
+      "loss": 0.1908,
+      "step": 11920
+    },
+    {
+      "epoch": 1.729513247073321,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.958571428571429e-06,
+      "loss": 0.1882,
+      "step": 11930
+    },
+    {
+      "epoch": 1.7309629951795875,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.9442857142857144e-06,
+      "loss": 0.2009,
+      "step": 11940
+    },
+    {
+      "epoch": 1.732412743285854,
+      "grad_norm": 1.6640625,
+      "learning_rate": 2.93e-06,
+      "loss": 0.1882,
+      "step": 11950
+    },
+    {
+      "epoch": 1.7338624913921206,
+      "grad_norm": 1.84375,
+      "learning_rate": 2.9157142857142863e-06,
+      "loss": 0.1852,
+      "step": 11960
+    },
+    {
+      "epoch": 1.7353122394983873,
+      "grad_norm": 1.6015625,
+      "learning_rate": 2.901428571428572e-06,
+      "loss": 0.1827,
+      "step": 11970
+    },
+    {
+      "epoch": 1.7367619876046536,
+      "grad_norm": 1.8984375,
+      "learning_rate": 2.8871428571428574e-06,
+      "loss": 0.2016,
+      "step": 11980
+    },
+    {
+      "epoch": 1.7382117357109204,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.8728571428571434e-06,
+      "loss": 0.1882,
+      "step": 11990
+    },
+    {
+      "epoch": 1.7396614838171867,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.858571428571429e-06,
+      "loss": 0.1929,
+      "step": 12000
+    },
+    {
+      "epoch": 1.7411112319234534,
+      "grad_norm": 1.8359375,
+      "learning_rate": 2.8442857142857145e-06,
+      "loss": 0.1897,
+      "step": 12010
+    },
+    {
+      "epoch": 1.7425609800297197,
+      "grad_norm": 1.625,
+      "learning_rate": 2.83e-06,
+      "loss": 0.1858,
+      "step": 12020
+    },
+    {
+      "epoch": 1.7440107281359865,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.815714285714286e-06,
+      "loss": 0.1687,
+      "step": 12030
+    },
+    {
+      "epoch": 1.7454604762422528,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.8014285714285715e-06,
+      "loss": 0.1865,
+      "step": 12040
+    },
+    {
+      "epoch": 1.7469102243485195,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.787142857142857e-06,
+      "loss": 0.1901,
+      "step": 12050
+    },
+    {
+      "epoch": 1.748359972454786,
+      "grad_norm": 1.5703125,
+      "learning_rate": 2.772857142857143e-06,
+      "loss": 0.2004,
+      "step": 12060
+    },
+    {
+      "epoch": 1.7498097205610526,
+      "grad_norm": 1.8203125,
+      "learning_rate": 2.7585714285714286e-06,
+      "loss": 0.2008,
+      "step": 12070
+    },
+    {
+      "epoch": 1.751259468667319,
+      "grad_norm": 1.765625,
+      "learning_rate": 2.744285714285714e-06,
+      "loss": 0.1941,
+      "step": 12080
+    },
+    {
+      "epoch": 1.7527092167735856,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.7300000000000005e-06,
+      "loss": 0.1904,
+      "step": 12090
+    },
+    {
+      "epoch": 1.7541589648798521,
+      "grad_norm": 1.8515625,
+      "learning_rate": 2.715714285714286e-06,
+      "loss": 0.2061,
+      "step": 12100
+    },
+    {
+      "epoch": 1.7556087129861186,
+      "grad_norm": 1.875,
+      "learning_rate": 2.7014285714285716e-06,
+      "loss": 0.2014,
+      "step": 12110
+    },
+    {
+      "epoch": 1.7570584610923852,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.6871428571428576e-06,
+      "loss": 0.201,
+      "step": 12120
+    },
+    {
+      "epoch": 1.7585082091986517,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.672857142857143e-06,
+      "loss": 0.1832,
+      "step": 12130
+    },
+    {
+      "epoch": 1.7599579573049182,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.6585714285714287e-06,
+      "loss": 0.1938,
+      "step": 12140
+    },
+    {
+      "epoch": 1.7614077054111847,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.6442857142857143e-06,
+      "loss": 0.1944,
+      "step": 12150
+    },
+    {
+      "epoch": 1.7628574535174515,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.6300000000000002e-06,
+      "loss": 0.1876,
+      "step": 12160
+    },
+    {
+      "epoch": 1.7643072016237178,
+      "grad_norm": 1.59375,
+      "learning_rate": 2.6157142857142858e-06,
+      "loss": 0.2107,
+      "step": 12170
+    },
+    {
+      "epoch": 1.7657569497299845,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.6014285714285713e-06,
+      "loss": 0.2021,
+      "step": 12180
+    },
+    {
+      "epoch": 1.7672066978362508,
+      "grad_norm": 1.8125,
+      "learning_rate": 2.5871428571428573e-06,
+      "loss": 0.1837,
+      "step": 12190
+    },
+    {
+      "epoch": 1.7686564459425176,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.572857142857143e-06,
+      "loss": 0.1755,
+      "step": 12200
+    },
+    {
+      "epoch": 1.7701061940487839,
+      "grad_norm": 1.78125,
+      "learning_rate": 2.5585714285714284e-06,
+      "loss": 0.2064,
+      "step": 12210
+    },
+    {
+      "epoch": 1.7715559421550506,
+      "grad_norm": 1.5546875,
+      "learning_rate": 2.544285714285715e-06,
+      "loss": 0.1967,
+      "step": 12220
+    },
+    {
+      "epoch": 1.773005690261317,
+      "grad_norm": 2.0,
+      "learning_rate": 2.5300000000000003e-06,
+      "loss": 0.1826,
+      "step": 12230
+    },
+    {
+      "epoch": 1.7744554383675837,
+      "grad_norm": 2.09375,
+      "learning_rate": 2.515714285714286e-06,
+      "loss": 0.1891,
+      "step": 12240
+    },
+    {
+      "epoch": 1.7759051864738502,
+      "grad_norm": 1.734375,
+      "learning_rate": 2.501428571428572e-06,
+      "loss": 0.1817,
+      "step": 12250
+    },
+    {
+      "epoch": 1.7773549345801167,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.4871428571428574e-06,
+      "loss": 0.2038,
+      "step": 12260
+    },
+    {
+      "epoch": 1.7788046826863833,
+      "grad_norm": 1.5,
+      "learning_rate": 2.472857142857143e-06,
+      "loss": 0.1887,
+      "step": 12270
+    },
+    {
+      "epoch": 1.7802544307926498,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.458571428571429e-06,
+      "loss": 0.1987,
+      "step": 12280
+    },
+    {
+      "epoch": 1.7817041788989163,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.4442857142857145e-06,
+      "loss": 0.1859,
+      "step": 12290
+    },
+    {
+      "epoch": 1.7831539270051828,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.43e-06,
+      "loss": 0.1944,
+      "step": 12300
+    },
+    {
+      "epoch": 1.7846036751114494,
+      "grad_norm": 1.4296875,
+      "learning_rate": 2.415714285714286e-06,
+      "loss": 0.2096,
+      "step": 12310
+    },
+    {
+      "epoch": 1.7860534232177159,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.4014285714285715e-06,
+      "loss": 0.1839,
+      "step": 12320
+    },
+    {
+      "epoch": 1.7875031713239826,
+      "grad_norm": 1.984375,
+      "learning_rate": 2.3871428571428575e-06,
+      "loss": 0.1788,
+      "step": 12330
+    },
+    {
+      "epoch": 1.788952919430249,
+      "grad_norm": 1.9453125,
+      "learning_rate": 2.372857142857143e-06,
+      "loss": 0.1948,
+      "step": 12340
+    },
+    {
+      "epoch": 1.7904026675365157,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.3585714285714286e-06,
+      "loss": 0.1905,
+      "step": 12350
+    },
+    {
+      "epoch": 1.791852415642782,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.3442857142857146e-06,
+      "loss": 0.171,
+      "step": 12360
+    },
+    {
+      "epoch": 1.7933021637490487,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.33e-06,
+      "loss": 0.2056,
+      "step": 12370
+    },
+    {
+      "epoch": 1.794751911855315,
+      "grad_norm": 1.671875,
+      "learning_rate": 2.3157142857142857e-06,
+      "loss": 0.207,
+      "step": 12380
+    },
+    {
+      "epoch": 1.7962016599615818,
+      "grad_norm": 1.8828125,
+      "learning_rate": 2.3014285714285717e-06,
+      "loss": 0.1978,
+      "step": 12390
+    },
+    {
+      "epoch": 1.797651408067848,
+      "grad_norm": 2.0,
+      "learning_rate": 2.287142857142857e-06,
+      "loss": 0.1988,
+      "step": 12400
+    },
+    {
+      "epoch": 1.7991011561741148,
+      "grad_norm": 1.796875,
+      "learning_rate": 2.272857142857143e-06,
+      "loss": 0.1873,
+      "step": 12410
+    },
+    {
+      "epoch": 1.8005509042803813,
+      "grad_norm": 1.53125,
+      "learning_rate": 2.2585714285714287e-06,
+      "loss": 0.1894,
+      "step": 12420
+    },
+    {
+      "epoch": 1.8020006523866479,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.2442857142857143e-06,
+      "loss": 0.1772,
+      "step": 12430
+    },
+    {
+      "epoch": 1.8034504004929144,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.2300000000000002e-06,
+      "loss": 0.1901,
+      "step": 12440
+    },
+    {
+      "epoch": 1.804900148599181,
+      "grad_norm": 1.484375,
+      "learning_rate": 2.215714285714286e-06,
+      "loss": 0.1843,
+      "step": 12450
+    },
+    {
+      "epoch": 1.8063498967054474,
+      "grad_norm": 1.75,
+      "learning_rate": 2.2014285714285718e-06,
+      "loss": 0.2003,
+      "step": 12460
+    },
+    {
+      "epoch": 1.807799644811714,
+      "grad_norm": 1.7578125,
+      "learning_rate": 2.1871428571428573e-06,
+      "loss": 0.2003,
+      "step": 12470
+    },
+    {
+      "epoch": 1.8092493929179805,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.172857142857143e-06,
+      "loss": 0.1999,
+      "step": 12480
+    },
+    {
+      "epoch": 1.810699141024247,
+      "grad_norm": 1.453125,
+      "learning_rate": 2.158571428571429e-06,
+      "loss": 0.1912,
+      "step": 12490
+    },
+    {
+      "epoch": 1.8121488891305135,
+      "grad_norm": 1.6875,
+      "learning_rate": 2.1442857142857144e-06,
+      "loss": 0.2037,
+      "step": 12500
+    },
+    {
+      "epoch": 1.81359863723678,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.13e-06,
+      "loss": 0.1777,
+      "step": 12510
+    },
+    {
+      "epoch": 1.8150483853430468,
+      "grad_norm": 1.7734375,
+      "learning_rate": 2.115714285714286e-06,
+      "loss": 0.1949,
+      "step": 12520
+    },
+    {
+      "epoch": 1.816498133449313,
+      "grad_norm": 1.71875,
+      "learning_rate": 2.1014285714285715e-06,
+      "loss": 0.1922,
+      "step": 12530
+    },
+    {
+      "epoch": 1.8179478815555798,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.0871428571428574e-06,
+      "loss": 0.1858,
+      "step": 12540
+    },
+    {
+      "epoch": 1.8193976296618461,
+      "grad_norm": 1.59375,
+      "learning_rate": 2.072857142857143e-06,
+      "loss": 0.1938,
+      "step": 12550
+    },
+    {
+      "epoch": 1.820847377768113,
+      "grad_norm": 2.015625,
+      "learning_rate": 2.0585714285714285e-06,
+      "loss": 0.2006,
+      "step": 12560
+    },
+    {
+      "epoch": 1.8222971258743792,
+      "grad_norm": 1.7421875,
+      "learning_rate": 2.0442857142857145e-06,
+      "loss": 0.195,
+      "step": 12570
+    },
+    {
+      "epoch": 1.823746873980646,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.0300000000000005e-06,
+      "loss": 0.2044,
+      "step": 12580
+    },
+    {
+      "epoch": 1.8251966220869122,
+      "grad_norm": 2.390625,
+      "learning_rate": 2.015714285714286e-06,
+      "loss": 0.1762,
+      "step": 12590
+    },
+    {
+      "epoch": 1.826646370193179,
+      "grad_norm": 1.8671875,
+      "learning_rate": 2.0014285714285716e-06,
+      "loss": 0.1925,
+      "step": 12600
+    },
+    {
+      "epoch": 1.8280961182994455,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.987142857142857e-06,
+      "loss": 0.2019,
+      "step": 12610
+    },
+    {
+      "epoch": 1.829545866405712,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.972857142857143e-06,
+      "loss": 0.1934,
+      "step": 12620
+    },
+    {
+      "epoch": 1.8309956145119786,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9585714285714286e-06,
+      "loss": 0.173,
+      "step": 12630
+    },
+    {
+      "epoch": 1.832445362618245,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.944285714285714e-06,
+      "loss": 0.1892,
+      "step": 12640
+    },
+    {
+      "epoch": 1.8338951107245116,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.93e-06,
+      "loss": 0.2048,
+      "step": 12650
+    },
+    {
+      "epoch": 1.8353448588307781,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9157142857142857e-06,
+      "loss": 0.1973,
+      "step": 12660
+    },
+    {
+      "epoch": 1.8367946069370447,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9014285714285715e-06,
+      "loss": 0.1988,
+      "step": 12670
+    },
+    {
+      "epoch": 1.8382443550433112,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8871428571428574e-06,
+      "loss": 0.1812,
+      "step": 12680
+    },
+    {
+      "epoch": 1.8396941031495777,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.872857142857143e-06,
+      "loss": 0.1932,
+      "step": 12690
+    },
+    {
+      "epoch": 1.8411438512558442,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.8585714285714287e-06,
+      "loss": 0.2022,
+      "step": 12700
+    },
+    {
+      "epoch": 1.842593599362111,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.8442857142857145e-06,
+      "loss": 0.1923,
+      "step": 12710
+    },
+    {
+      "epoch": 1.8440433474683773,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.83e-06,
+      "loss": 0.1989,
+      "step": 12720
+    },
+    {
+      "epoch": 1.845493095574644,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.815714285714286e-06,
+      "loss": 0.1917,
+      "step": 12730
+    },
+    {
+      "epoch": 1.8469428436809103,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.8014285714285716e-06,
+      "loss": 0.1887,
+      "step": 12740
+    },
+    {
+      "epoch": 1.848392591787177,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.7871428571428573e-06,
+      "loss": 0.2134,
+      "step": 12750
+    },
+    {
+      "epoch": 1.8498423398934434,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.772857142857143e-06,
+      "loss": 0.2094,
+      "step": 12760
+    },
+    {
+      "epoch": 1.8512920879997101,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.7585714285714286e-06,
+      "loss": 0.1875,
+      "step": 12770
+    },
+    {
+      "epoch": 1.8527418361059764,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.7442857142857144e-06,
+      "loss": 0.1763,
+      "step": 12780
+    },
+    {
+      "epoch": 1.8541915842122432,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.73e-06,
+      "loss": 0.1874,
+      "step": 12790
+    },
+    {
+      "epoch": 1.8556413323185097,
+      "grad_norm": 2.0,
+      "learning_rate": 1.715714285714286e-06,
+      "loss": 0.1897,
+      "step": 12800
+    },
+    {
+      "epoch": 1.8570910804247762,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7014285714285717e-06,
+      "loss": 0.1848,
+      "step": 12810
+    },
+    {
+      "epoch": 1.8585408285310427,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.6871428571428572e-06,
+      "loss": 0.1862,
+      "step": 12820
+    },
+    {
+      "epoch": 1.8599905766373093,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.672857142857143e-06,
+      "loss": 0.2066,
+      "step": 12830
+    },
+    {
+      "epoch": 1.8614403247435758,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.6585714285714288e-06,
+      "loss": 0.1945,
+      "step": 12840
+    },
+    {
+      "epoch": 1.8628900728498423,
+      "grad_norm": 2.0,
+      "learning_rate": 1.6442857142857143e-06,
+      "loss": 0.1799,
+      "step": 12850
+    },
+    {
+      "epoch": 1.8643398209561088,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.6300000000000003e-06,
+      "loss": 0.195,
+      "step": 12860
+    },
+    {
+      "epoch": 1.8657895690623754,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.6157142857142858e-06,
+      "loss": 0.1953,
+      "step": 12870
+    },
+    {
+      "epoch": 1.867239317168642,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6014285714285716e-06,
+      "loss": 0.1981,
+      "step": 12880
+    },
+    {
+      "epoch": 1.8686890652749084,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5871428571428573e-06,
+      "loss": 0.1951,
+      "step": 12890
+    },
+    {
+      "epoch": 1.8701388133811752,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5728571428571429e-06,
+      "loss": 0.1862,
+      "step": 12900
+    },
+    {
+      "epoch": 1.8715885614874415,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5585714285714287e-06,
+      "loss": 0.1772,
+      "step": 12910
+    },
+    {
+      "epoch": 1.8730383095937082,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5442857142857142e-06,
+      "loss": 0.1843,
+      "step": 12920
+    },
+    {
+      "epoch": 1.8744880576999745,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5300000000000002e-06,
+      "loss": 0.1861,
+      "step": 12930
+    },
+    {
+      "epoch": 1.8759378058062413,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.515714285714286e-06,
+      "loss": 0.2098,
+      "step": 12940
+    },
+    {
+      "epoch": 1.8773875539125076,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5014285714285715e-06,
+      "loss": 0.1949,
+      "step": 12950
+    },
+    {
+      "epoch": 1.8788373020187743,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4871428571428572e-06,
+      "loss": 0.1905,
+      "step": 12960
+    },
+    {
+      "epoch": 1.8802870501250408,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.472857142857143e-06,
+      "loss": 0.1877,
+      "step": 12970
+    },
+    {
+      "epoch": 1.8817367982313074,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4585714285714286e-06,
+      "loss": 0.2009,
+      "step": 12980
+    },
+    {
+      "epoch": 1.8831865463375739,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.4442857142857145e-06,
+      "loss": 0.2006,
+      "step": 12990
+    },
+    {
+      "epoch": 1.8846362944438404,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.43e-06,
+      "loss": 0.2025,
+      "step": 13000
+    },
+    {
+      "epoch": 1.886086042550107,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.4157142857142858e-06,
+      "loss": 0.18,
+      "step": 13010
+    },
+    {
+      "epoch": 1.8875357906563734,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.4014285714285716e-06,
+      "loss": 0.191,
+      "step": 13020
+    },
+    {
+      "epoch": 1.88898553876264,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3871428571428571e-06,
+      "loss": 0.1994,
+      "step": 13030
+    },
+    {
+      "epoch": 1.8904352868689065,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.372857142857143e-06,
+      "loss": 0.1888,
+      "step": 13040
+    },
+    {
+      "epoch": 1.891885034975173,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.3585714285714289e-06,
+      "loss": 0.2014,
+      "step": 13050
+    },
+    {
+      "epoch": 1.8933347830814395,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.3442857142857144e-06,
+      "loss": 0.1922,
+      "step": 13060
+    },
+    {
+      "epoch": 1.8947845311877063,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3300000000000002e-06,
+      "loss": 0.1857,
+      "step": 13070
+    },
+    {
+      "epoch": 1.8962342792939726,
+      "grad_norm": 1.875,
+      "learning_rate": 1.3157142857142857e-06,
+      "loss": 0.2022,
+      "step": 13080
+    },
+    {
+      "epoch": 1.8976840274002393,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.3014285714285715e-06,
+      "loss": 0.1887,
+      "step": 13090
+    },
+    {
+      "epoch": 1.8991337755065056,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2871428571428575e-06,
+      "loss": 0.1821,
+      "step": 13100
+    },
+    {
+      "epoch": 1.9005835236127724,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.2728571428571428e-06,
+      "loss": 0.1925,
+      "step": 13110
+    },
+    {
+      "epoch": 1.9020332717190387,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.2585714285714288e-06,
+      "loss": 0.1944,
+      "step": 13120
+    },
+    {
+      "epoch": 1.9034830198253054,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2442857142857143e-06,
+      "loss": 0.188,
+      "step": 13130
+    },
+    {
+      "epoch": 1.9049327679315717,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.23e-06,
+      "loss": 0.1921,
+      "step": 13140
+    },
+    {
+      "epoch": 1.9063825160378385,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.2157142857142858e-06,
+      "loss": 0.193,
+      "step": 13150
+    },
+    {
+      "epoch": 1.907832264144105,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2014285714285716e-06,
+      "loss": 0.2015,
+      "step": 13160
+    },
+    {
+      "epoch": 1.9092820122503715,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.1871428571428574e-06,
+      "loss": 0.1873,
+      "step": 13170
+    },
+    {
+      "epoch": 1.910731760356638,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.172857142857143e-06,
+      "loss": 0.1958,
+      "step": 13180
+    },
+    {
+      "epoch": 1.9121815084629046,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.1585714285714287e-06,
+      "loss": 0.1951,
+      "step": 13190
+    },
+    {
+      "epoch": 1.913631256569171,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1442857142857144e-06,
+      "loss": 0.1943,
+      "step": 13200
+    },
+    {
+      "epoch": 1.9150810046754376,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1300000000000002e-06,
+      "loss": 0.1912,
+      "step": 13210
+    },
+    {
+      "epoch": 1.9165307527817041,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1157142857142857e-06,
+      "loss": 0.1891,
+      "step": 13220
+    },
+    {
+      "epoch": 1.9179805008879707,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1014285714285715e-06,
+      "loss": 0.1972,
+      "step": 13230
+    },
+    {
+      "epoch": 1.9194302489942372,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0871428571428573e-06,
+      "loss": 0.1796,
+      "step": 13240
+    },
+    {
+      "epoch": 1.9208799971005037,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.072857142857143e-06,
+      "loss": 0.1904,
+      "step": 13250
+    },
+    {
+      "epoch": 1.9223297452067705,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0585714285714286e-06,
+      "loss": 0.1746,
+      "step": 13260
+    },
+    {
+      "epoch": 1.9237794933130368,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0442857142857143e-06,
+      "loss": 0.191,
+      "step": 13270
+    },
+    {
+      "epoch": 1.9252292414193035,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.03e-06,
+      "loss": 0.187,
+      "step": 13280
+    },
+    {
+      "epoch": 1.9266789895255698,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.0157142857142859e-06,
+      "loss": 0.192,
+      "step": 13290
+    },
+    {
+      "epoch": 1.9281287376318366,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0014285714285716e-06,
+      "loss": 0.2085,
+      "step": 13300
+    },
+    {
+      "epoch": 1.9295784857381029,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.871428571428572e-07,
+      "loss": 0.1922,
+      "step": 13310
+    },
+    {
+      "epoch": 1.9310282338443696,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.72857142857143e-07,
+      "loss": 0.1878,
+      "step": 13320
+    },
+    {
+      "epoch": 1.932477981950636,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.585714285714287e-07,
+      "loss": 0.2004,
+      "step": 13330
+    },
+    {
+      "epoch": 1.9339277300569027,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.442857142857144e-07,
+      "loss": 0.1911,
+      "step": 13340
+    },
+    {
+      "epoch": 1.9353774781631692,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.300000000000001e-07,
+      "loss": 0.2187,
+      "step": 13350
+    },
+    {
+      "epoch": 1.9368272262694357,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.157142857142857e-07,
+      "loss": 0.1891,
+      "step": 13360
+    },
+    {
+      "epoch": 1.9382769743757022,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.014285714285716e-07,
+      "loss": 0.1928,
+      "step": 13370
+    },
+    {
+      "epoch": 1.9397267224819688,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.871428571428573e-07,
+      "loss": 0.1971,
+      "step": 13380
+    },
+    {
+      "epoch": 1.9411764705882353,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.728571428571429e-07,
+      "loss": 0.1912,
+      "step": 13390
+    },
+    {
+      "epoch": 1.9426262186945018,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.585714285714286e-07,
+      "loss": 0.1761,
+      "step": 13400
+    },
+    {
+      "epoch": 1.9440759668007683,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.442857142857143e-07,
+      "loss": 0.186,
+      "step": 13410
+    },
+    {
+      "epoch": 1.9455257149070349,
+      "grad_norm": 1.5,
+      "learning_rate": 8.300000000000001e-07,
+      "loss": 0.1931,
+      "step": 13420
+    },
+    {
+      "epoch": 1.9469754630133016,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.157142857142858e-07,
+      "loss": 0.1922,
+      "step": 13430
+    },
+    {
+      "epoch": 1.948425211119568,
+      "grad_norm": 1.5390625,
+      "learning_rate": 8.014285714285715e-07,
+      "loss": 0.2032,
+      "step": 13440
+    },
+    {
+      "epoch": 1.9498749592258346,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.871428571428572e-07,
+      "loss": 0.1868,
+      "step": 13450
+    },
+    {
+      "epoch": 1.951324707332101,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.728571428571429e-07,
+      "loss": 0.188,
+      "step": 13460
+    },
+    {
+      "epoch": 1.9527744554383677,
+      "grad_norm": 1.4765625,
+      "learning_rate": 7.585714285714287e-07,
+      "loss": 0.1853,
+      "step": 13470
+    },
+    {
+      "epoch": 1.954224203544634,
+      "grad_norm": 2.125,
+      "learning_rate": 7.442857142857143e-07,
+      "loss": 0.1865,
+      "step": 13480
+    },
+    {
+      "epoch": 1.9556739516509007,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.3e-07,
+      "loss": 0.1879,
+      "step": 13490
+    },
+    {
+      "epoch": 1.957123699757167,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.157142857142859e-07,
+      "loss": 0.2061,
+      "step": 13500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 14000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 9.603473818292724e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}