diff --git "a/checkpoint-11500/trainer_state.json" "b/checkpoint-11500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-11500/trainer_state.json"
@@ -0,0 +1,8084 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.964795452877083,
+  "eval_steps": 500,
+  "global_step": 11500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008389525677192026,
+      "grad_norm": 290.0,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 6.6219,
+      "step": 10
+    },
+    {
+      "epoch": 0.0016779051354384052,
+      "grad_norm": 4.15625,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.5478,
+      "step": 20
+    },
+    {
+      "epoch": 0.0025168577031576076,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.16e-05,
+      "loss": 0.9445,
+      "step": 30
+    },
+    {
+      "epoch": 0.0033558102708768104,
+      "grad_norm": 1.3125,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 0.693,
+      "step": 40
+    },
+    {
+      "epoch": 0.004194762838596013,
+      "grad_norm": 1.3671875,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.6403,
+      "step": 50
+    },
+    {
+      "epoch": 0.005033715406315215,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.36e-05,
+      "loss": 0.6304,
+      "step": 60
+    },
+    {
+      "epoch": 0.005872667974034418,
+      "grad_norm": 1.515625,
+      "learning_rate": 2.76e-05,
+      "loss": 0.599,
+      "step": 70
+    },
+    {
+      "epoch": 0.006711620541753621,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.16e-05,
+      "loss": 0.591,
+      "step": 80
+    },
+    {
+      "epoch": 0.007550573109472823,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.5600000000000005e-05,
+      "loss": 0.6058,
+      "step": 90
+    },
+    {
+      "epoch": 0.008389525677192027,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.96e-05,
+      "loss": 0.6065,
+      "step": 100
+    },
+    {
+      "epoch": 0.009228478244911228,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.360000000000001e-05,
+      "loss": 0.6054,
+      "step": 110
+    },
+    {
+      "epoch": 0.01006743081263043,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.7600000000000005e-05,
+      "loss": 0.6284,
+      "step": 120
+    },
+    {
+      "epoch": 0.010906383380349634,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.160000000000001e-05,
+      "loss": 0.6296,
+      "step": 130
+    },
+    {
+      "epoch": 0.011745335948068836,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.56e-05,
+      "loss": 0.534,
+      "step": 140
+    },
+    {
+      "epoch": 0.012584288515788038,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.9600000000000005e-05,
+      "loss": 0.5969,
+      "step": 150
+    },
+    {
+      "epoch": 0.013423241083507242,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.360000000000001e-05,
+      "loss": 0.6417,
+      "step": 160
+    },
+    {
+      "epoch": 0.014262193651226444,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.76e-05,
+      "loss": 0.6363,
+      "step": 170
+    },
+    {
+      "epoch": 0.015101146218945646,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.16e-05,
+      "loss": 0.6482,
+      "step": 180
+    },
+    {
+      "epoch": 0.015940098786664848,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.560000000000001e-05,
+      "loss": 0.6197,
+      "step": 190
+    },
+    {
+      "epoch": 0.016779051354384053,
+      "grad_norm": 0.83984375,
+      "learning_rate": 7.960000000000001e-05,
+      "loss": 0.599,
+      "step": 200
+    },
+    {
+      "epoch": 0.017618003922103255,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.993856655290104e-05,
+      "loss": 0.6171,
+      "step": 210
+    },
+    {
+      "epoch": 0.018456956489822457,
+      "grad_norm": 0.81640625,
+      "learning_rate": 7.98703071672355e-05,
+      "loss": 0.621,
+      "step": 220
+    },
+    {
+      "epoch": 0.01929590905754166,
+      "grad_norm": 0.84765625,
+      "learning_rate": 7.980204778156997e-05,
+      "loss": 0.5833,
+      "step": 230
+    },
+    {
+      "epoch": 0.02013486162526086,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.973378839590445e-05,
+      "loss": 0.62,
+      "step": 240
+    },
+    {
+      "epoch": 0.020973814192980063,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.966552901023891e-05,
+      "loss": 0.6168,
+      "step": 250
+    },
+    {
+      "epoch": 0.021812766760699268,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.959726962457338e-05,
+      "loss": 0.596,
+      "step": 260
+    },
+    {
+      "epoch": 0.02265171932841847,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.952901023890786e-05,
+      "loss": 0.5832,
+      "step": 270
+    },
+    {
+      "epoch": 0.023490671896137672,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.946075085324232e-05,
+      "loss": 0.6071,
+      "step": 280
+    },
+    {
+      "epoch": 0.024329624463856874,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.93924914675768e-05,
+      "loss": 0.6342,
+      "step": 290
+    },
+    {
+      "epoch": 0.025168577031576076,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.932423208191127e-05,
+      "loss": 0.6141,
+      "step": 300
+    },
+    {
+      "epoch": 0.02600752959929528,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.925597269624573e-05,
+      "loss": 0.6573,
+      "step": 310
+    },
+    {
+      "epoch": 0.026846482167014483,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.918771331058021e-05,
+      "loss": 0.5759,
+      "step": 320
+    },
+    {
+      "epoch": 0.027685434734733685,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.911945392491469e-05,
+      "loss": 0.5954,
+      "step": 330
+    },
+    {
+      "epoch": 0.028524387302452887,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.905119453924916e-05,
+      "loss": 0.6044,
+      "step": 340
+    },
+    {
+      "epoch": 0.02936333987017209,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.898293515358362e-05,
+      "loss": 0.6609,
+      "step": 350
+    },
+    {
+      "epoch": 0.03020229243789129,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.891467576791809e-05,
+      "loss": 0.6294,
+      "step": 360
+    },
+    {
+      "epoch": 0.031041245005610497,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.884641638225257e-05,
+      "loss": 0.6168,
+      "step": 370
+    },
+    {
+      "epoch": 0.031880197573329695,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.877815699658705e-05,
+      "loss": 0.5928,
+      "step": 380
+    },
+    {
+      "epoch": 0.0327191501410489,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.870989761092151e-05,
+      "loss": 0.6392,
+      "step": 390
+    },
+    {
+      "epoch": 0.033558102708768106,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.864163822525598e-05,
+      "loss": 0.595,
+      "step": 400
+    },
+    {
+      "epoch": 0.03439705527648731,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.857337883959044e-05,
+      "loss": 0.5803,
+      "step": 410
+    },
+    {
+      "epoch": 0.03523600784420651,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.850511945392492e-05,
+      "loss": 0.5735,
+      "step": 420
+    },
+    {
+      "epoch": 0.03607496041192571,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.843686006825939e-05,
+      "loss": 0.5848,
+      "step": 430
+    },
+    {
+      "epoch": 0.036913912979644914,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.836860068259385e-05,
+      "loss": 0.6076,
+      "step": 440
+    },
+    {
+      "epoch": 0.037752865547364116,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.830034129692833e-05,
+      "loss": 0.6163,
+      "step": 450
+    },
+    {
+      "epoch": 0.03859181811508332,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.823208191126281e-05,
+      "loss": 0.598,
+      "step": 460
+    },
+    {
+      "epoch": 0.03943077068280252,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.816382252559728e-05,
+      "loss": 0.5121,
+      "step": 470
+    },
+    {
+      "epoch": 0.04026972325052172,
+      "grad_norm": 0.7265625,
+      "learning_rate": 7.809556313993174e-05,
+      "loss": 0.5521,
+      "step": 480
+    },
+    {
+      "epoch": 0.041108675818240924,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.802730375426621e-05,
+      "loss": 0.62,
+      "step": 490
+    },
+    {
+      "epoch": 0.041947628385960126,
+      "grad_norm": 0.625,
+      "learning_rate": 7.795904436860069e-05,
+      "loss": 0.5939,
+      "step": 500
+    },
+    {
+      "epoch": 0.042786580953679335,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.789078498293517e-05,
+      "loss": 0.5646,
+      "step": 510
+    },
+    {
+      "epoch": 0.043625533521398537,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.782252559726963e-05,
+      "loss": 0.595,
+      "step": 520
+    },
+    {
+      "epoch": 0.04446448608911774,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.77542662116041e-05,
+      "loss": 0.5661,
+      "step": 530
+    },
+    {
+      "epoch": 0.04530343865683694,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.768600682593857e-05,
+      "loss": 0.5619,
+      "step": 540
+    },
+    {
+      "epoch": 0.04614239122455614,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.761774744027304e-05,
+      "loss": 0.5654,
+      "step": 550
+    },
+    {
+      "epoch": 0.046981343792275344,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.754948805460752e-05,
+      "loss": 0.5833,
+      "step": 560
+    },
+    {
+      "epoch": 0.047820296359994546,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.748122866894199e-05,
+      "loss": 0.5263,
+      "step": 570
+    },
+    {
+      "epoch": 0.04865924892771375,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.741296928327645e-05,
+      "loss": 0.582,
+      "step": 580
+    },
+    {
+      "epoch": 0.04949820149543295,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.734470989761092e-05,
+      "loss": 0.5396,
+      "step": 590
+    },
+    {
+      "epoch": 0.05033715406315215,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.72764505119454e-05,
+      "loss": 0.5699,
+      "step": 600
+    },
+    {
+      "epoch": 0.051176106630871354,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.720819112627987e-05,
+      "loss": 0.5638,
+      "step": 610
+    },
+    {
+      "epoch": 0.05201505919859056,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.713993174061433e-05,
+      "loss": 0.5641,
+      "step": 620
+    },
+    {
+      "epoch": 0.052854011766309765,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.707167235494881e-05,
+      "loss": 0.5358,
+      "step": 630
+    },
+    {
+      "epoch": 0.05369296433402897,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.700341296928329e-05,
+      "loss": 0.565,
+      "step": 640
+    },
+    {
+      "epoch": 0.05453191690174817,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.693515358361776e-05,
+      "loss": 0.5771,
+      "step": 650
+    },
+    {
+      "epoch": 0.05537086946946737,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.686689419795222e-05,
+      "loss": 0.5628,
+      "step": 660
+    },
+    {
+      "epoch": 0.05620982203718657,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.679863481228669e-05,
+      "loss": 0.589,
+      "step": 670
+    },
+    {
+      "epoch": 0.057048774604905775,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.673037542662117e-05,
+      "loss": 0.5531,
+      "step": 680
+    },
+    {
+      "epoch": 0.05788772717262498,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.666211604095565e-05,
+      "loss": 0.5415,
+      "step": 690
+    },
+    {
+      "epoch": 0.05872667974034418,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.659385665529011e-05,
+      "loss": 0.5729,
+      "step": 700
+    },
+    {
+      "epoch": 0.05956563230806338,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.652559726962458e-05,
+      "loss": 0.5536,
+      "step": 710
+    },
+    {
+      "epoch": 0.06040458487578258,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.645733788395904e-05,
+      "loss": 0.5431,
+      "step": 720
+    },
+    {
+      "epoch": 0.061243537443501785,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.638907849829352e-05,
+      "loss": 0.5555,
+      "step": 730
+    },
+    {
+      "epoch": 0.062082490011220993,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.6320819112628e-05,
+      "loss": 0.5444,
+      "step": 740
+    },
+    {
+      "epoch": 0.0629214425789402,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.625255972696247e-05,
+      "loss": 0.5285,
+      "step": 750
+    },
+    {
+      "epoch": 0.06376039514665939,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.618430034129693e-05,
+      "loss": 0.5676,
+      "step": 760
+    },
+    {
+      "epoch": 0.0645993477143786,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.611604095563141e-05,
+      "loss": 0.5007,
+      "step": 770
+    },
+    {
+      "epoch": 0.0654383002820978,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.604778156996588e-05,
+      "loss": 0.5397,
+      "step": 780
+    },
+    {
+      "epoch": 0.066277252849817,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.597952218430036e-05,
+      "loss": 0.5538,
+      "step": 790
+    },
+    {
+      "epoch": 0.06711620541753621,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.591126279863481e-05,
+      "loss": 0.5183,
+      "step": 800
+    },
+    {
+      "epoch": 0.06795515798525541,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.584300341296929e-05,
+      "loss": 0.5385,
+      "step": 810
+    },
+    {
+      "epoch": 0.06879411055297462,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.577474402730377e-05,
+      "loss": 0.5702,
+      "step": 820
+    },
+    {
+      "epoch": 0.06963306312069381,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.570648464163823e-05,
+      "loss": 0.5602,
+      "step": 830
+    },
+    {
+      "epoch": 0.07047201568841302,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.56382252559727e-05,
+      "loss": 0.5444,
+      "step": 840
+    },
+    {
+      "epoch": 0.07131096825613222,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.556996587030716e-05,
+      "loss": 0.5568,
+      "step": 850
+    },
+    {
+      "epoch": 0.07214992082385142,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.550170648464164e-05,
+      "loss": 0.5868,
+      "step": 860
+    },
+    {
+      "epoch": 0.07298887339157062,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.543344709897612e-05,
+      "loss": 0.5531,
+      "step": 870
+    },
+    {
+      "epoch": 0.07382782595928983,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.536518771331059e-05,
+      "loss": 0.5235,
+      "step": 880
+    },
+    {
+      "epoch": 0.07466677852700902,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.529692832764505e-05,
+      "loss": 0.5234,
+      "step": 890
+    },
+    {
+      "epoch": 0.07550573109472823,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.522866894197953e-05,
+      "loss": 0.4985,
+      "step": 900
+    },
+    {
+      "epoch": 0.07634468366244744,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.5160409556314e-05,
+      "loss": 0.5425,
+      "step": 910
+    },
+    {
+      "epoch": 0.07718363623016664,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.509215017064848e-05,
+      "loss": 0.5343,
+      "step": 920
+    },
+    {
+      "epoch": 0.07802258879788584,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.502389078498294e-05,
+      "loss": 0.5259,
+      "step": 930
+    },
+    {
+      "epoch": 0.07886154136560504,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.495563139931741e-05,
+      "loss": 0.5176,
+      "step": 940
+    },
+    {
+      "epoch": 0.07970049393332425,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.488737201365189e-05,
+      "loss": 0.5314,
+      "step": 950
+    },
+    {
+      "epoch": 0.08053944650104344,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.481911262798635e-05,
+      "loss": 0.5583,
+      "step": 960
+    },
+    {
+      "epoch": 0.08137839906876265,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.475085324232083e-05,
+      "loss": 0.5162,
+      "step": 970
+    },
+    {
+      "epoch": 0.08221735163648185,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.46825938566553e-05,
+      "loss": 0.5017,
+      "step": 980
+    },
+    {
+      "epoch": 0.08305630420420106,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.461433447098976e-05,
+      "loss": 0.5186,
+      "step": 990
+    },
+    {
+      "epoch": 0.08389525677192025,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.454607508532424e-05,
+      "loss": 0.5654,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08473420933963946,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.447781569965871e-05,
+      "loss": 0.5267,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08557316190735867,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.440955631399318e-05,
+      "loss": 0.5202,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08641211447507786,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.434129692832765e-05,
+      "loss": 0.5497,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08725106704279707,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.427303754266212e-05,
+      "loss": 0.517,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08809001961051627,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.42047781569966e-05,
+      "loss": 0.5921,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08892897217823548,
+      "grad_norm": 0.474609375,
+      "learning_rate": 7.413651877133106e-05,
+      "loss": 0.5061,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08976792474595467,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.406825938566553e-05,
+      "loss": 0.5261,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09060687731367388,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.400000000000001e-05,
+      "loss": 0.5014,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09144582988139308,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.393174061433448e-05,
+      "loss": 0.5191,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09228478244911228,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.386348122866895e-05,
+      "loss": 0.5717,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09312373501683148,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.379522184300342e-05,
+      "loss": 0.5369,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09396268758455069,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.372696245733789e-05,
+      "loss": 0.5232,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0948016401522699,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.365870307167237e-05,
+      "loss": 0.5449,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09564059271998909,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.359044368600683e-05,
+      "loss": 0.537,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0964795452877083,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.352218430034131e-05,
+      "loss": 0.5266,
+      "step": 1150
+    },
+    {
+      "epoch": 0.0973184978554275,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.345392491467578e-05,
+      "loss": 0.5448,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0981574504231467,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.338566552901024e-05,
+      "loss": 0.5589,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0989964029908659,
+      "grad_norm": 0.48828125,
+      "learning_rate": 7.331740614334472e-05,
+      "loss": 0.5174,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09983535555858511,
+      "grad_norm": 0.5,
+      "learning_rate": 7.324914675767919e-05,
+      "loss": 0.5537,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1006743081263043,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.318088737201365e-05,
+      "loss": 0.5181,
+      "step": 1200
+    },
+    {
+      "epoch": 0.10151326069402351,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.311262798634813e-05,
+      "loss": 0.5371,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10235221326174271,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.30443686006826e-05,
+      "loss": 0.502,
+      "step": 1220
+    },
+    {
+      "epoch": 0.10319116582946192,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.297610921501708e-05,
+      "loss": 0.5532,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10403011839718113,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.290784982935154e-05,
+      "loss": 0.51,
+      "step": 1240
+    },
+    {
+      "epoch": 0.10486907096490032,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.283959044368601e-05,
+      "loss": 0.5173,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10570802353261953,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.277133105802049e-05,
+      "loss": 0.5247,
+      "step": 1260
+    },
+    {
+      "epoch": 0.10654697610033873,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.270307167235495e-05,
+      "loss": 0.5094,
+      "step": 1270
+    },
+    {
+      "epoch": 0.10738592866805793,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.263481228668943e-05,
+      "loss": 0.4913,
+      "step": 1280
+    },
+    {
+      "epoch": 0.10822488123577713,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.25665529010239e-05,
+      "loss": 0.517,
+      "step": 1290
+    },
+    {
+      "epoch": 0.10906383380349634,
+      "grad_norm": 0.5,
+      "learning_rate": 7.249829351535836e-05,
+      "loss": 0.5303,
+      "step": 1300
+    },
+    {
+      "epoch": 0.10990278637121553,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.243003412969284e-05,
+      "loss": 0.5111,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11074173893893474,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.236177474402731e-05,
+      "loss": 0.5216,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11158069150665394,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.229351535836179e-05,
+      "loss": 0.5426,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11241964407437315,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.222525597269625e-05,
+      "loss": 0.4916,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11325859664209235,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.215699658703072e-05,
+      "loss": 0.4829,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11409754920981155,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.20887372013652e-05,
+      "loss": 0.5266,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11493650177753076,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.202047781569966e-05,
+      "loss": 0.5466,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11577545434524995,
+      "grad_norm": 0.474609375,
+      "learning_rate": 7.195221843003413e-05,
+      "loss": 0.5389,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11661440691296916,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.188395904436861e-05,
+      "loss": 0.5168,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11745335948068836,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.181569965870307e-05,
+      "loss": 0.5077,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11829231204840757,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.174744027303755e-05,
+      "loss": 0.4983,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11913126461612676,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.167918088737202e-05,
+      "loss": 0.4972,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11997021718384597,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.161092150170648e-05,
+      "loss": 0.5226,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12080916975156517,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.154266211604096e-05,
+      "loss": 0.503,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12164812231928437,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.147440273037543e-05,
+      "loss": 0.5348,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12248707488700357,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.140614334470991e-05,
+      "loss": 0.4928,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12332602745472278,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.133788395904437e-05,
+      "loss": 0.5024,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12416498002244199,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.126962457337884e-05,
+      "loss": 0.4768,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12500393259016118,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.120136518771332e-05,
+      "loss": 0.4918,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1258428851578804,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.113310580204779e-05,
+      "loss": 0.4975,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1266818377255996,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.106484641638226e-05,
+      "loss": 0.4952,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12752079029331878,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.099658703071673e-05,
+      "loss": 0.5037,
+      "step": 1520
+    },
+    {
+      "epoch": 0.128359742861038,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.09283276450512e-05,
+      "loss": 0.5083,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1291986954287572,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.086006825938567e-05,
+      "loss": 0.4955,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1300376479964764,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.079180887372014e-05,
+      "loss": 0.5227,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1308766005641956,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.07235494880546e-05,
+      "loss": 0.5049,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1317155531319148,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.065529010238909e-05,
+      "loss": 0.5296,
+      "step": 1570
+    },
+    {
+      "epoch": 0.132554505699634,
+      "grad_norm": 0.5,
+      "learning_rate": 7.058703071672355e-05,
+      "loss": 0.5114,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13339345826735322,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.051877133105803e-05,
+      "loss": 0.4872,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13423241083507242,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.04505119453925e-05,
+      "loss": 0.5246,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1350713634027916,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.038225255972696e-05,
+      "loss": 0.5132,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13591031597051081,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.031399317406144e-05,
+      "loss": 0.5115,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13674926853823002,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.02457337883959e-05,
+      "loss": 0.5134,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13758822110594923,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.017747440273039e-05,
+      "loss": 0.5062,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1384271736736684,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.010921501706485e-05,
+      "loss": 0.4901,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13926612624138762,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.004095563139932e-05,
+      "loss": 0.5184,
+      "step": 1660
+    },
+    {
+      "epoch": 0.14010507880910683,
+      "grad_norm": 0.5,
+      "learning_rate": 6.99726962457338e-05,
+      "loss": 0.5085,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14094403137682604,
+      "grad_norm": 0.5,
+      "learning_rate": 6.990443686006826e-05,
+      "loss": 0.4884,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14178298394454525,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.983617747440274e-05,
+      "loss": 0.5136,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14262193651226443,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.976791808873721e-05,
+      "loss": 0.4814,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14346088907998364,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.969965870307167e-05,
+      "loss": 0.5241,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14429984164770285,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.963139931740615e-05,
+      "loss": 0.4941,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14513879421542206,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.956313993174062e-05,
+      "loss": 0.4865,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14597774678314124,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.949488054607508e-05,
+      "loss": 0.4577,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14681669935086045,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.942662116040956e-05,
+      "loss": 0.4727,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14765565191857966,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.935836177474403e-05,
+      "loss": 0.4624,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14849460448629886,
+      "grad_norm": 0.474609375,
+      "learning_rate": 6.929010238907851e-05,
+      "loss": 0.477,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14933355705401805,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.922184300341297e-05,
+      "loss": 0.4987,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15017250962173725,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.915358361774744e-05,
+      "loss": 0.5018,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15101146218945646,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.908532423208192e-05,
+      "loss": 0.506,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15185041475717567,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.901706484641638e-05,
+      "loss": 0.494,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15268936732489488,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.894880546075086e-05,
+      "loss": 0.5037,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15352831989261406,
+      "grad_norm": 0.81640625,
+      "learning_rate": 6.888054607508533e-05,
+      "loss": 0.4933,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15436727246033327,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.88122866894198e-05,
+      "loss": 0.4891,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15520622502805248,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.874402730375427e-05,
+      "loss": 0.5019,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1560451775957717,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.867576791808875e-05,
+      "loss": 0.4845,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15688413016349087,
+      "grad_norm": 0.462890625,
+      "learning_rate": 6.860750853242322e-05,
+      "loss": 0.464,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15772308273121008,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.853924914675768e-05,
+      "loss": 0.4677,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1585620352989293,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.847098976109215e-05,
+      "loss": 0.462,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1594009878666485,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.840273037542663e-05,
+      "loss": 0.4755,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16023994043436768,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.83344709897611e-05,
+      "loss": 0.5176,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1610788930020869,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.826621160409556e-05,
+      "loss": 0.452,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1619178455698061,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.819795221843004e-05,
+      "loss": 0.5008,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1627567981375253,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.81296928327645e-05,
+      "loss": 0.4664,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1635957507052445,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.806143344709898e-05,
+      "loss": 0.4695,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1644347032729637,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.799317406143345e-05,
+      "loss": 0.4591,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1652736558406829,
+      "grad_norm": 0.47265625,
+      "learning_rate": 6.792491467576792e-05,
+      "loss": 0.4595,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1661126084084021,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.78566552901024e-05,
+      "loss": 0.4739,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16695156097612132,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.778839590443687e-05,
+      "loss": 0.4793,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1677905135438405,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.772013651877134e-05,
+      "loss": 0.4749,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1686294661115597,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.76518771331058e-05,
+      "loss": 0.5148,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16946841867927892,
+      "grad_norm": 0.49609375,
+      "learning_rate": 6.758361774744027e-05,
+      "loss": 0.5101,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17030737124699813,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.751535836177475e-05,
+      "loss": 0.4781,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17114632381471734,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.744709897610923e-05,
+      "loss": 0.4708,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17198527638243652,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.73788395904437e-05,
+      "loss": 0.4739,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17282422895015573,
+      "grad_norm": 0.390625,
+      "learning_rate": 6.731058020477816e-05,
+      "loss": 0.458,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17366318151787494,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.724232081911263e-05,
+      "loss": 0.4762,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17450213408559415,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.71740614334471e-05,
+      "loss": 0.4539,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17534108665331333,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.710580204778159e-05,
+      "loss": 0.4771,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17618003922103254,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.703754266211605e-05,
+      "loss": 0.478,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17701899178875175,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.696928327645052e-05,
+      "loss": 0.4498,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17785794435647095,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.690102389078498e-05,
+      "loss": 0.4836,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17869689692419014,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.683276450511946e-05,
+      "loss": 0.4617,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17953584949190934,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.676450511945393e-05,
+      "loss": 0.5011,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18037480205962855,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.669624573378839e-05,
+      "loss": 0.4577,
+      "step": 2150
+    },
+    {
+      "epoch": 0.18121375462734776,
+      "grad_norm": 0.5,
+      "learning_rate": 6.662798634812287e-05,
+      "loss": 0.4857,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18205270719506697,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.655972696245735e-05,
+      "loss": 0.4846,
+      "step": 2170
+    },
+    {
+      "epoch": 0.18289165976278615,
+      "grad_norm": 0.482421875,
+      "learning_rate": 6.649146757679182e-05,
+      "loss": 0.4992,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18373061233050536,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.642320819112628e-05,
+      "loss": 0.452,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18456956489822457,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.635494880546075e-05,
+      "loss": 0.4598,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18540851746594378,
+      "grad_norm": 0.482421875,
+      "learning_rate": 6.628668941979523e-05,
+      "loss": 0.4405,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18624747003366296,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.62184300341297e-05,
+      "loss": 0.4606,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18708642260138217,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.615017064846417e-05,
+      "loss": 0.4765,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18792537516910138,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.608191126279864e-05,
+      "loss": 0.4773,
+      "step": 2240
+    },
+    {
+      "epoch": 0.1887643277368206,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.60136518771331e-05,
+      "loss": 0.4657,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1896032803045398,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.594539249146758e-05,
+      "loss": 0.448,
+      "step": 2260
+    },
+    {
+      "epoch": 0.19044223287225898,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.587713310580206e-05,
+      "loss": 0.4578,
+      "step": 2270
+    },
+    {
+      "epoch": 0.19128118543997819,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.580887372013653e-05,
+      "loss": 0.455,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1921201380076974,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.5740614334471e-05,
+      "loss": 0.4657,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1929590905754166,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.567235494880547e-05,
+      "loss": 0.4754,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19379804314313578,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.560409556313994e-05,
+      "loss": 0.4405,
+      "step": 2310
+    },
+    {
+      "epoch": 0.194636995710855,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.55358361774744e-05,
+      "loss": 0.4605,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1954759482785742,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.546757679180887e-05,
+      "loss": 0.5133,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1963149008462934,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.539931740614335e-05,
+      "loss": 0.4318,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1971538534140126,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.533105802047783e-05,
+      "loss": 0.4743,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1979928059817318,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.52627986348123e-05,
+      "loss": 0.4667,
+      "step": 2360
+    },
+    {
+      "epoch": 0.198831758549451,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.519453924914676e-05,
+      "loss": 0.4878,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19967071111717022,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.512627986348123e-05,
+      "loss": 0.4279,
+      "step": 2380
+    },
+    {
+      "epoch": 0.20050966368488943,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.50580204778157e-05,
+      "loss": 0.4737,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2013486162526086,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.498976109215018e-05,
+      "loss": 0.4355,
+      "step": 2400
+    },
+    {
+      "epoch": 0.20218756882032782,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.492150170648465e-05,
+      "loss": 0.4458,
+      "step": 2410
+    },
+    {
+      "epoch": 0.20302652138804703,
+      "grad_norm": 0.474609375,
+      "learning_rate": 6.485324232081912e-05,
+      "loss": 0.4723,
+      "step": 2420
+    },
+    {
+      "epoch": 0.20386547395576624,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.47849829351536e-05,
+      "loss": 0.4331,
+      "step": 2430
+    },
+    {
+      "epoch": 0.20470442652348542,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.471672354948806e-05,
+      "loss": 0.4613,
+      "step": 2440
+    },
+    {
+      "epoch": 0.20554337909120463,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.464846416382254e-05,
+      "loss": 0.4682,
+      "step": 2450
+    },
+    {
+      "epoch": 0.20638233165892383,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.4580204778157e-05,
+      "loss": 0.4343,
+      "step": 2460
+    },
+    {
+      "epoch": 0.20722128422664304,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.451194539249147e-05,
+      "loss": 0.4582,
+      "step": 2470
+    },
+    {
+      "epoch": 0.20806023679436225,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.444368600682595e-05,
+      "loss": 0.4496,
+      "step": 2480
+    },
+    {
+      "epoch": 0.20889918936208143,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.437542662116042e-05,
+      "loss": 0.4507,
+      "step": 2490
+    },
+    {
+      "epoch": 0.20973814192980064,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.430716723549488e-05,
+      "loss": 0.4319,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21057709449751985,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.423890784982935e-05,
+      "loss": 0.4843,
+      "step": 2510
+    },
+    {
+      "epoch": 0.21141604706523906,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.417064846416383e-05,
+      "loss": 0.4452,
+      "step": 2520
+    },
+    {
+      "epoch": 0.21225499963295824,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.41023890784983e-05,
+      "loss": 0.4459,
+      "step": 2530
+    },
+    {
+      "epoch": 0.21309395220067745,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.403412969283277e-05,
+      "loss": 0.4298,
+      "step": 2540
+    },
+    {
+      "epoch": 0.21393290476839666,
+      "grad_norm": 0.47265625,
+      "learning_rate": 6.396587030716724e-05,
+      "loss": 0.4787,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21477185733611587,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.389761092150172e-05,
+      "loss": 0.4694,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21561080990383505,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.382935153583618e-05,
+      "loss": 0.4112,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21644976247155426,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.376109215017066e-05,
+      "loss": 0.483,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21728871503927347,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.369283276450513e-05,
+      "loss": 0.4553,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21812766760699268,
+      "grad_norm": 0.5,
+      "learning_rate": 6.362457337883959e-05,
+      "loss": 0.4915,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21896662017471188,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.355631399317407e-05,
+      "loss": 0.4574,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21980557274243107,
+      "grad_norm": 0.5,
+      "learning_rate": 6.348805460750854e-05,
+      "loss": 0.4874,
+      "step": 2620
+    },
+    {
+      "epoch": 0.22064452531015027,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.341979522184302e-05,
+      "loss": 0.4514,
+      "step": 2630
+    },
+    {
+      "epoch": 0.22148347787786948,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.335153583617748e-05,
+      "loss": 0.4272,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2223224304455887,
+      "grad_norm": 0.41015625,
+      "learning_rate": 6.328327645051195e-05,
+      "loss": 0.451,
+      "step": 2650
+    },
+    {
+      "epoch": 0.22316138301330787,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.321501706484643e-05,
+      "loss": 0.4915,
+      "step": 2660
+    },
+    {
+      "epoch": 0.22400033558102708,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.314675767918089e-05,
+      "loss": 0.4305,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2248392881487463,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.307849829351536e-05,
+      "loss": 0.4472,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2256782407164655,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.301023890784984e-05,
+      "loss": 0.4676,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2265171932841847,
+      "grad_norm": 0.43359375,
+      "learning_rate": 6.29419795221843e-05,
+      "loss": 0.4966,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2273561458519039,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.287372013651878e-05,
+      "loss": 0.4278,
+      "step": 2710
+    },
+    {
+      "epoch": 0.2281950984196231,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.280546075085325e-05,
+      "loss": 0.4478,
+      "step": 2720
+    },
+    {
+      "epoch": 0.2290340509873423,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.273720136518771e-05,
+      "loss": 0.4217,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22987300355506152,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.266894197952219e-05,
+      "loss": 0.4293,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2307119561227807,
+      "grad_norm": 0.484375,
+      "learning_rate": 6.260068259385666e-05,
+      "loss": 0.46,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2315509086904999,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.253242320819114e-05,
+      "loss": 0.4623,
+      "step": 2760
+    },
+    {
+      "epoch": 0.23238986125821912,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.24641638225256e-05,
+      "loss": 0.4705,
+      "step": 2770
+    },
+    {
+      "epoch": 0.23322881382593832,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.239590443686007e-05,
+      "loss": 0.4416,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2340677663936575,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.232764505119455e-05,
+      "loss": 0.4106,
+      "step": 2790
+    },
+    {
+      "epoch": 0.23490671896137671,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.225938566552901e-05,
+      "loss": 0.4642,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23574567152909592,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.21911262798635e-05,
+      "loss": 0.4476,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23658462409681513,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.212286689419796e-05,
+      "loss": 0.455,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23742357666453434,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.205460750853242e-05,
+      "loss": 0.4555,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23826252923225352,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.19863481228669e-05,
+      "loss": 0.495,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23910148179997273,
+      "grad_norm": 0.435546875,
+      "learning_rate": 6.191808873720137e-05,
+      "loss": 0.4076,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23994043436769194,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.184982935153584e-05,
+      "loss": 0.4398,
+      "step": 2860
+    },
+    {
+      "epoch": 0.24077938693541115,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.178156996587031e-05,
+      "loss": 0.4383,
+      "step": 2870
+    },
+    {
+      "epoch": 0.24161833950313033,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.171331058020478e-05,
+      "loss": 0.4432,
+      "step": 2880
+    },
+    {
+      "epoch": 0.24245729207084954,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.164505119453926e-05,
+      "loss": 0.4463,
+      "step": 2890
+    },
+    {
+      "epoch": 0.24329624463856875,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.157679180887373e-05,
+      "loss": 0.4596,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24413519720628796,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.150853242320819e-05,
+      "loss": 0.5046,
+      "step": 2910
+    },
+    {
+      "epoch": 0.24497414977400714,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.144027303754267e-05,
+      "loss": 0.4507,
+      "step": 2920
+    },
+    {
+      "epoch": 0.24581310234172635,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.137201365187714e-05,
+      "loss": 0.4575,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24665205490944556,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.130375426621161e-05,
+      "loss": 0.4668,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24749100747716477,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.123549488054608e-05,
+      "loss": 0.4656,
+      "step": 2950
+    },
+    {
+      "epoch": 0.24832996004488397,
+      "grad_norm": 0.37890625,
+      "learning_rate": 6.116723549488055e-05,
+      "loss": 0.4224,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24916891261260316,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.109897610921503e-05,
+      "loss": 0.4412,
+      "step": 2970
+    },
+    {
+      "epoch": 0.25000786518032236,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.103071672354949e-05,
+      "loss": 0.4512,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2508468177480416,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.0962457337883964e-05,
+      "loss": 0.4302,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2516857703157608,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.0894197952218436e-05,
+      "loss": 0.4406,
+      "step": 3000
+    },
+    {
+      "epoch": 0.25252472288348,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.08259385665529e-05,
+      "loss": 0.4241,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2533636754511992,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.075767918088738e-05,
+      "loss": 0.4216,
+      "step": 3020
+    },
+    {
+      "epoch": 0.25420262801891835,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.068941979522185e-05,
+      "loss": 0.4415,
+      "step": 3030
+    },
+    {
+      "epoch": 0.25504158058663756,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.062116040955632e-05,
+      "loss": 0.4302,
+      "step": 3040
+    },
+    {
+      "epoch": 0.25588053315435677,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.055290102389079e-05,
+      "loss": 0.4468,
+      "step": 3050
+    },
+    {
+      "epoch": 0.256719485722076,
+      "grad_norm": 0.48828125,
+      "learning_rate": 6.048464163822526e-05,
+      "loss": 0.4397,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2575584382897952,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.0416382252559736e-05,
+      "loss": 0.4542,
+      "step": 3070
+    },
+    {
+      "epoch": 0.2583973908575144,
+      "grad_norm": 0.462890625,
+      "learning_rate": 6.0348122866894195e-05,
+      "loss": 0.4199,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2592363434252336,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.0279863481228675e-05,
+      "loss": 0.4216,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2600752959929528,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.021160409556315e-05,
+      "loss": 0.4309,
+      "step": 3100
+    },
+    {
+      "epoch": 0.260914248560672,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.014334470989761e-05,
+      "loss": 0.4567,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2617532011283912,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.0075085324232085e-05,
+      "loss": 0.4273,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2625921536961104,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.0006825938566564e-05,
+      "loss": 0.4112,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2634311062638296,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.993856655290103e-05,
+      "loss": 0.4498,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2642700588315488,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.98703071672355e-05,
+      "loss": 0.427,
+      "step": 3150
+    },
+    {
+      "epoch": 0.265109011399268,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.980204778156997e-05,
+      "loss": 0.4167,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2659479639669872,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.973378839590444e-05,
+      "loss": 0.3787,
+      "step": 3170
+    },
+    {
+      "epoch": 0.26678691653470643,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.966552901023891e-05,
+      "loss": 0.4317,
+      "step": 3180
+    },
+    {
+      "epoch": 0.26762586910242564,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.959726962457338e-05,
+      "loss": 0.4098,
+      "step": 3190
+    },
+    {
+      "epoch": 0.26846482167014485,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.952901023890786e-05,
+      "loss": 0.4357,
+      "step": 3200
+    },
+    {
+      "epoch": 0.269303774237864,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.9460750853242324e-05,
+      "loss": 0.4277,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2701427268055832,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.9392491467576796e-05,
+      "loss": 0.4205,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2709816793733024,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.932423208191127e-05,
+      "loss": 0.4259,
+      "step": 3230
+    },
+    {
+      "epoch": 0.27182063194102163,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.9255972696245734e-05,
+      "loss": 0.4296,
+      "step": 3240
+    },
+    {
+      "epoch": 0.27265958450874084,
+      "grad_norm": 0.3828125,
+      "learning_rate": 5.9187713310580214e-05,
+      "loss": 0.4092,
+      "step": 3250
+    },
+    {
+      "epoch": 0.27349853707646005,
+      "grad_norm": 0.46875,
+      "learning_rate": 5.9119453924914686e-05,
+      "loss": 0.4199,
+      "step": 3260
+    },
+    {
+      "epoch": 0.27433748964417926,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.905119453924915e-05,
+      "loss": 0.3933,
+      "step": 3270
+    },
+    {
+      "epoch": 0.27517644221189846,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.8982935153583624e-05,
+      "loss": 0.4492,
+      "step": 3280
+    },
+    {
+      "epoch": 0.2760153947796177,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.891467576791809e-05,
+      "loss": 0.4505,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2768543473473368,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.884641638225256e-05,
+      "loss": 0.4073,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27769329991505604,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.877815699658704e-05,
+      "loss": 0.4388,
+      "step": 3310
+    },
+    {
+      "epoch": 0.27853225248277524,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.870989761092151e-05,
+      "loss": 0.4237,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27937120505049445,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.864163822525598e-05,
+      "loss": 0.4001,
+      "step": 3330
+    },
+    {
+      "epoch": 0.28021015761821366,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.8573378839590445e-05,
+      "loss": 0.4324,
+      "step": 3340
+    },
+    {
+      "epoch": 0.28104911018593287,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.850511945392492e-05,
+      "loss": 0.4404,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2818880627536521,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.843686006825939e-05,
+      "loss": 0.4181,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2827270153213713,
+      "grad_norm": 0.470703125,
+      "learning_rate": 5.8368600682593856e-05,
+      "loss": 0.4123,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2835659678890905,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.8300341296928335e-05,
+      "loss": 0.4224,
+      "step": 3380
+    },
+    {
+      "epoch": 0.28440492045680965,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.823208191126281e-05,
+      "loss": 0.4059,
+      "step": 3390
+    },
+    {
+      "epoch": 0.28524387302452886,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.816382252559727e-05,
+      "loss": 0.4295,
+      "step": 3400
+    },
+    {
+      "epoch": 0.28608282559224807,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.8095563139931746e-05,
+      "loss": 0.4251,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2869217781599673,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.802730375426621e-05,
+      "loss": 0.4448,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2877607307276865,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.795904436860069e-05,
+      "loss": 0.4223,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2885996832954057,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.789078498293516e-05,
+      "loss": 0.4372,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2894386358631249,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.782252559726963e-05,
+      "loss": 0.3961,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2902775884308441,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.77542662116041e-05,
+      "loss": 0.4195,
+      "step": 3460
+    },
+    {
+      "epoch": 0.29111654099856327,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.768600682593857e-05,
+      "loss": 0.4226,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2919554935662825,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.761774744027304e-05,
+      "loss": 0.4318,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2927944461340017,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.754948805460752e-05,
+      "loss": 0.4511,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2936333987017209,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.7481228668941984e-05,
+      "loss": 0.4321,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2944723512694401,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.741296928327646e-05,
+      "loss": 0.4016,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2953113038371593,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.734470989761092e-05,
+      "loss": 0.3936,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2961502564048785,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.7276450511945395e-05,
+      "loss": 0.4152,
+      "step": 3530
+    },
+    {
+      "epoch": 0.29698920897259773,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.720819112627987e-05,
+      "loss": 0.4021,
+      "step": 3540
+    },
+    {
+      "epoch": 0.29782816154031694,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.713993174061433e-05,
+      "loss": 0.4409,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2986671141080361,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.707167235494881e-05,
+      "loss": 0.397,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2995060666757553,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.7003412969283285e-05,
+      "loss": 0.4422,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3003450192434745,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.693515358361775e-05,
+      "loss": 0.4104,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3011839718111937,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.686689419795222e-05,
+      "loss": 0.3982,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3020229243789129,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.679863481228669e-05,
+      "loss": 0.4193,
+      "step": 3600
+    },
+    {
+      "epoch": 0.30286187694663214,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.673037542662117e-05,
+      "loss": 0.4249,
+      "step": 3610
+    },
+    {
+      "epoch": 0.30370082951435134,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.666211604095564e-05,
+      "loss": 0.4162,
+      "step": 3620
+    },
+    {
+      "epoch": 0.30453978208207055,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.6593856655290106e-05,
+      "loss": 0.3931,
+      "step": 3630
+    },
+    {
+      "epoch": 0.30537873464978976,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.652559726962458e-05,
+      "loss": 0.3981,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3062176872175089,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.6457337883959044e-05,
+      "loss": 0.4286,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3070566397852281,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.6389078498293516e-05,
+      "loss": 0.431,
+      "step": 3660
+    },
+    {
+      "epoch": 0.30789559235294733,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.6320819112627996e-05,
+      "loss": 0.402,
+      "step": 3670
+    },
+    {
+      "epoch": 0.30873454492066654,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.625255972696246e-05,
+      "loss": 0.4212,
+      "step": 3680
+    },
+    {
+      "epoch": 0.30957349748838575,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.6184300341296934e-05,
+      "loss": 0.4355,
+      "step": 3690
+    },
+    {
+      "epoch": 0.31041245005610496,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.6116040955631406e-05,
+      "loss": 0.4358,
+      "step": 3700
+    },
+    {
+      "epoch": 0.31125140262382417,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.604778156996587e-05,
+      "loss": 0.4846,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3120903551915434,
+      "grad_norm": 0.470703125,
+      "learning_rate": 5.597952218430035e-05,
+      "loss": 0.3781,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3129293077592626,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.591126279863481e-05,
+      "loss": 0.4198,
+      "step": 3730
+    },
+    {
+      "epoch": 0.31376826032698174,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.584300341296929e-05,
+      "loss": 0.4387,
+      "step": 3740
+    },
+    {
+      "epoch": 0.31460721289470095,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.577474402730376e-05,
+      "loss": 0.4313,
+      "step": 3750
+    },
+    {
+      "epoch": 0.31544616546242016,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.570648464163823e-05,
+      "loss": 0.4184,
+      "step": 3760
+    },
+    {
+      "epoch": 0.31628511803013937,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.56382252559727e-05,
+      "loss": 0.4007,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3171240705978586,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.5569965870307165e-05,
+      "loss": 0.421,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3179630231655778,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.5501706484641645e-05,
+      "loss": 0.4016,
+      "step": 3790
+    },
+    {
+      "epoch": 0.318801975733297,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.543344709897612e-05,
+      "loss": 0.3737,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3196409283010162,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.536518771331058e-05,
+      "loss": 0.422,
+      "step": 3810
+    },
+    {
+      "epoch": 0.32047988086873536,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.5296928327645055e-05,
+      "loss": 0.42,
+      "step": 3820
+    },
+    {
+      "epoch": 0.32131883343645457,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.522866894197953e-05,
+      "loss": 0.4118,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3221577860041738,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.5160409556313993e-05,
+      "loss": 0.3828,
+      "step": 3840
+    },
+    {
+      "epoch": 0.322996738571893,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.509215017064847e-05,
+      "loss": 0.4322,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3238356911396122,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.502389078498294e-05,
+      "loss": 0.3888,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3246746437073314,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.495563139931741e-05,
+      "loss": 0.3914,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3255135962750506,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.488737201365188e-05,
+      "loss": 0.3864,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3263525488427698,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.481911262798635e-05,
+      "loss": 0.3853,
+      "step": 3890
+    },
+    {
+      "epoch": 0.327191501410489,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.475085324232083e-05,
+      "loss": 0.4229,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3280304539782082,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.4682593856655294e-05,
+      "loss": 0.432,
+      "step": 3910
+    },
+    {
+      "epoch": 0.3288694065459274,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.4614334470989766e-05,
+      "loss": 0.3766,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3297083591136466,
+      "grad_norm": 0.46484375,
+      "learning_rate": 5.454607508532424e-05,
+      "loss": 0.3977,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3305473116813658,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.4477815699658704e-05,
+      "loss": 0.3929,
+      "step": 3940
+    },
+    {
+      "epoch": 0.331386264249085,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.440955631399318e-05,
+      "loss": 0.4273,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3322252168168042,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.4341296928327656e-05,
+      "loss": 0.3961,
+      "step": 3960
+    },
+    {
+      "epoch": 0.33306416938452343,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.427303754266212e-05,
+      "loss": 0.3848,
+      "step": 3970
+    },
+    {
+      "epoch": 0.33390312195224264,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.4204778156996594e-05,
+      "loss": 0.3883,
+      "step": 3980
+    },
+    {
+      "epoch": 0.33474207451996185,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.413651877133106e-05,
+      "loss": 0.3954,
+      "step": 3990
+    },
+    {
+      "epoch": 0.335581027087681,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.406825938566553e-05,
+      "loss": 0.4048,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3364199796554002,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 0.3781,
+      "step": 4010
+    },
+    {
+      "epoch": 0.3372589322231194,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.393174061433447e-05,
+      "loss": 0.3827,
+      "step": 4020
+    },
+    {
+      "epoch": 0.33809788479083863,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.386348122866895e-05,
+      "loss": 0.402,
+      "step": 4030
+    },
+    {
+      "epoch": 0.33893683735855784,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.3795221843003415e-05,
+      "loss": 0.403,
+      "step": 4040
+    },
+    {
+      "epoch": 0.33977578992627705,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.372696245733789e-05,
+      "loss": 0.4247,
+      "step": 4050
+    },
+    {
+      "epoch": 0.34061474249399626,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.365870307167236e-05,
+      "loss": 0.3697,
+      "step": 4060
+    },
+    {
+      "epoch": 0.34145369506171547,
+      "grad_norm": 0.376953125,
+      "learning_rate": 5.3590443686006826e-05,
+      "loss": 0.4241,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3422926476294347,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.3522184300341305e-05,
+      "loss": 0.4229,
+      "step": 4080
+    },
+    {
+      "epoch": 0.34313160019715383,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.345392491467578e-05,
+      "loss": 0.3971,
+      "step": 4090
+    },
+    {
+      "epoch": 0.34397055276487304,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.338566552901024e-05,
+      "loss": 0.412,
+      "step": 4100
+    },
+    {
+      "epoch": 0.34480950533259225,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.3317406143344716e-05,
+      "loss": 0.3692,
+      "step": 4110
+    },
+    {
+      "epoch": 0.34564845790031146,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.324914675767918e-05,
+      "loss": 0.4232,
+      "step": 4120
+    },
+    {
+      "epoch": 0.34648741046803067,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.3180887372013654e-05,
+      "loss": 0.4087,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3473263630357499,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.311262798634813e-05,
+      "loss": 0.4235,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3481653156034691,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.30443686006826e-05,
+      "loss": 0.4036,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3490042681711883,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.297610921501707e-05,
+      "loss": 0.3665,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3498432207389075,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.290784982935154e-05,
+      "loss": 0.3914,
+      "step": 4170
+    },
+    {
+      "epoch": 0.35068217330662665,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.283959044368601e-05,
+      "loss": 0.3714,
+      "step": 4180
+    },
+    {
+      "epoch": 0.35152112587434586,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.277133105802048e-05,
+      "loss": 0.4048,
+      "step": 4190
+    },
+    {
+      "epoch": 0.35236007844206507,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.270307167235495e-05,
+      "loss": 0.3795,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3531990310097843,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.263481228668943e-05,
+      "loss": 0.4002,
+      "step": 4210
+    },
+    {
+      "epoch": 0.3540379835775035,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.25665529010239e-05,
+      "loss": 0.4126,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3548769361452227,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.2498293515358365e-05,
+      "loss": 0.3944,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3557158887129419,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.243003412969284e-05,
+      "loss": 0.4179,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3565548412806611,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.23617747440273e-05,
+      "loss": 0.3878,
+      "step": 4250
+    },
+    {
+      "epoch": 0.35739379384838027,
+      "grad_norm": 0.365234375,
+      "learning_rate": 5.229351535836178e-05,
+      "loss": 0.378,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3582327464160995,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.2225255972696255e-05,
+      "loss": 0.3705,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3590716989838187,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.215699658703072e-05,
+      "loss": 0.4008,
+      "step": 4280
+    },
+    {
+      "epoch": 0.3599106515515379,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.208873720136519e-05,
+      "loss": 0.3957,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3607496041192571,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.202047781569966e-05,
+      "loss": 0.4047,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3615885566869763,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.195221843003413e-05,
+      "loss": 0.4013,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3624275092546955,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.188395904436861e-05,
+      "loss": 0.4057,
+      "step": 4320
+    },
+    {
+      "epoch": 0.36326646182241473,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.1815699658703076e-05,
+      "loss": 0.3759,
+      "step": 4330
+    },
+    {
+      "epoch": 0.36410541439013394,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.174744027303755e-05,
+      "loss": 0.4095,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3649443669578531,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.1679180887372014e-05,
+      "loss": 0.3885,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3657833195255723,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.1610921501706486e-05,
+      "loss": 0.4258,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3666222720932915,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.1542662116040966e-05,
+      "loss": 0.396,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3674612246610107,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.1474402730375425e-05,
+      "loss": 0.3749,
+      "step": 4380
+    },
+    {
+      "epoch": 0.36830017722872993,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.1406143344709904e-05,
+      "loss": 0.3873,
+      "step": 4390
+    },
+    {
+      "epoch": 0.36913912979644914,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.1337883959044376e-05,
+      "loss": 0.3725,
+      "step": 4400
+    },
+    {
+      "epoch": 0.36997808236416835,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.126962457337884e-05,
+      "loss": 0.3794,
+      "step": 4410
+    },
+    {
+      "epoch": 0.37081703493188756,
+      "grad_norm": 0.46484375,
+      "learning_rate": 5.1201365187713314e-05,
+      "loss": 0.3908,
+      "step": 4420
+    },
+    {
+      "epoch": 0.37165598749960677,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.113310580204778e-05,
+      "loss": 0.3812,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3724949400673259,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.106484641638226e-05,
+      "loss": 0.4344,
+      "step": 4440
+    },
+    {
+      "epoch": 0.37333389263504513,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.099658703071673e-05,
+      "loss": 0.3884,
+      "step": 4450
+    },
+    {
+      "epoch": 0.37417284520276434,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.09283276450512e-05,
+      "loss": 0.381,
+      "step": 4460
+    },
+    {
+      "epoch": 0.37501179777048355,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.086006825938567e-05,
+      "loss": 0.3803,
+      "step": 4470
+    },
+    {
+      "epoch": 0.37585075033820275,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.0791808873720135e-05,
+      "loss": 0.4266,
+      "step": 4480
+    },
+    {
+      "epoch": 0.37668970290592196,
+      "grad_norm": 0.486328125,
+      "learning_rate": 5.072354948805461e-05,
+      "loss": 0.377,
+      "step": 4490
+    },
+    {
+      "epoch": 0.3775286554736412,
+      "grad_norm": 0.3671875,
+      "learning_rate": 5.065529010238909e-05,
+      "loss": 0.3635,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3783676080413604,
+      "grad_norm": 0.373046875,
+      "learning_rate": 5.058703071672355e-05,
+      "loss": 0.4067,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3792065606090796,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.0518771331058025e-05,
+      "loss": 0.3815,
+      "step": 4520
+    },
+    {
+      "epoch": 0.38004551317679874,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.04505119453925e-05,
+      "loss": 0.356,
+      "step": 4530
+    },
+    {
+      "epoch": 0.38088446574451795,
+      "grad_norm": 0.369140625,
+      "learning_rate": 5.0382252559726963e-05,
+      "loss": 0.394,
+      "step": 4540
+    },
+    {
+      "epoch": 0.38172341831223716,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.031399317406144e-05,
+      "loss": 0.3757,
+      "step": 4550
+    },
+    {
+      "epoch": 0.38256237087995637,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.024573378839591e-05,
+      "loss": 0.4115,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3834013234476756,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.017747440273038e-05,
+      "loss": 0.3735,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3842402760153948,
+      "grad_norm": 0.3828125,
+      "learning_rate": 5.010921501706485e-05,
+      "loss": 0.3696,
+      "step": 4580
+    },
+    {
+      "epoch": 0.385079228583114,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.004095563139932e-05,
+      "loss": 0.4015,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3859181811508332,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.997269624573379e-05,
+      "loss": 0.388,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3867571337185524,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.990443686006826e-05,
+      "loss": 0.3666,
+      "step": 4610
+    },
+    {
+      "epoch": 0.38759608628627157,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.9836177474402736e-05,
+      "loss": 0.3942,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3884350388539908,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.976791808873721e-05,
+      "loss": 0.3514,
+      "step": 4630
+    },
+    {
+      "epoch": 0.38927399142171,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.9699658703071674e-05,
+      "loss": 0.3885,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3901129439894292,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.963139931740615e-05,
+      "loss": 0.3653,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3909518965571484,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.956313993174062e-05,
+      "loss": 0.4304,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3917908491248676,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.9494880546075085e-05,
+      "loss": 0.4087,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3926298016925868,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.9426621160409564e-05,
+      "loss": 0.3567,
+      "step": 4680
+    },
+    {
+      "epoch": 0.39346875426030603,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.935836177474403e-05,
+      "loss": 0.3968,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3943077068280252,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.92901023890785e-05,
+      "loss": 0.4048,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3951466593957444,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.9221843003412975e-05,
+      "loss": 0.4115,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3959856119634636,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.915358361774744e-05,
+      "loss": 0.3706,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3968245645311828,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.908532423208192e-05,
+      "loss": 0.3831,
+      "step": 4730
+    },
+    {
+      "epoch": 0.397663517098902,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.9017064846416385e-05,
+      "loss": 0.394,
+      "step": 4740
+    },
+    {
+      "epoch": 0.39850246966662123,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.894880546075086e-05,
+      "loss": 0.3928,
+      "step": 4750
+    },
+    {
+      "epoch": 0.39934142223434044,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.888054607508533e-05,
+      "loss": 0.3838,
+      "step": 4760
+    },
+    {
+      "epoch": 0.40018037480205965,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.8812286689419796e-05,
+      "loss": 0.4014,
+      "step": 4770
+    },
+    {
+      "epoch": 0.40101932736977886,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.874402730375427e-05,
+      "loss": 0.3832,
+      "step": 4780
+    },
+    {
+      "epoch": 0.401858279937498,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.867576791808875e-05,
+      "loss": 0.3779,
+      "step": 4790
+    },
+    {
+      "epoch": 0.4026972325052172,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.860750853242321e-05,
+      "loss": 0.3528,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4035361850729364,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.8539249146757686e-05,
+      "loss": 0.3457,
+      "step": 4810
+    },
+    {
+      "epoch": 0.40437513764065564,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.847098976109215e-05,
+      "loss": 0.3936,
+      "step": 4820
+    },
+    {
+      "epoch": 0.40521409020837484,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.8402730375426624e-05,
+      "loss": 0.3731,
+      "step": 4830
+    },
+    {
+      "epoch": 0.40605304277609405,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.8334470989761096e-05,
+      "loss": 0.3713,
+      "step": 4840
+    },
+    {
+      "epoch": 0.40689199534381326,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.826621160409556e-05,
+      "loss": 0.3551,
+      "step": 4850
+    },
+    {
+      "epoch": 0.40773094791153247,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.819795221843004e-05,
+      "loss": 0.3949,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4085699004792517,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.812969283276451e-05,
+      "loss": 0.4025,
+      "step": 4870
+    },
+    {
+      "epoch": 0.40940885304697083,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.806143344709898e-05,
+      "loss": 0.4235,
+      "step": 4880
+    },
+    {
+      "epoch": 0.41024780561469004,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.799317406143345e-05,
+      "loss": 0.3775,
+      "step": 4890
+    },
+    {
+      "epoch": 0.41108675818240925,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.792491467576792e-05,
+      "loss": 0.3746,
+      "step": 4900
+    },
+    {
+      "epoch": 0.41192571075012846,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.78566552901024e-05,
+      "loss": 0.3956,
+      "step": 4910
+    },
+    {
+      "epoch": 0.41276466331784767,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.778839590443687e-05,
+      "loss": 0.3876,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4136036158855669,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.7720136518771335e-05,
+      "loss": 0.387,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4144425684532861,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.765187713310581e-05,
+      "loss": 0.4316,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4152815210210053,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.758361774744027e-05,
+      "loss": 0.3687,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4161204735887245,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.7515358361774745e-05,
+      "loss": 0.3934,
+      "step": 4960
+    },
+    {
+      "epoch": 0.41695942615644366,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.7447098976109225e-05,
+      "loss": 0.3847,
+      "step": 4970
+    },
+    {
+      "epoch": 0.41779837872416287,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.737883959044369e-05,
+      "loss": 0.3557,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4186373312918821,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.731058020477816e-05,
+      "loss": 0.4218,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4194762838596013,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.724232081911263e-05,
+      "loss": 0.4181,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4203152364273205,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.71740614334471e-05,
+      "loss": 0.3795,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4211541889950397,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.710580204778158e-05,
+      "loss": 0.4028,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4219931415627589,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.7037542662116046e-05,
+      "loss": 0.3825,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4228320941304781,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.696928327645052e-05,
+      "loss": 0.3888,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4236710466981973,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.6901023890784984e-05,
+      "loss": 0.4167,
+      "step": 5050
+    },
+    {
+      "epoch": 0.4245099992659165,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.6832764505119456e-05,
+      "loss": 0.3944,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4253489518336357,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.676450511945393e-05,
+      "loss": 0.3643,
+      "step": 5070
+    },
+    {
+      "epoch": 0.4261879044013549,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.6696245733788395e-05,
+      "loss": 0.3612,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4270268569690741,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.6627986348122874e-05,
+      "loss": 0.3632,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4278658095367933,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.6559726962457346e-05,
+      "loss": 0.3677,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4287047621045125,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.649146757679181e-05,
+      "loss": 0.3721,
+      "step": 5110
+    },
+    {
+      "epoch": 0.42954371467223174,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.6423208191126284e-05,
+      "loss": 0.3836,
+      "step": 5120
+    },
+    {
+      "epoch": 0.43038266723995094,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.635494880546075e-05,
+      "loss": 0.3635,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4312216198076701,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.628668941979522e-05,
+      "loss": 0.4082,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4320605723753893,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.62184300341297e-05,
+      "loss": 0.3845,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4328995249431085,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.615017064846417e-05,
+      "loss": 0.356,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4337384775108277,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.608191126279864e-05,
+      "loss": 0.3948,
+      "step": 5170
+    },
+    {
+      "epoch": 0.43457743007854693,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.6013651877133106e-05,
+      "loss": 0.3754,
+      "step": 5180
+    },
+    {
+      "epoch": 0.43541638264626614,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.594539249146758e-05,
+      "loss": 0.3837,
+      "step": 5190
+    },
+    {
+      "epoch": 0.43625533521398535,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.587713310580206e-05,
+      "loss": 0.3595,
+      "step": 5200
+    },
+    {
+      "epoch": 0.43709428778170456,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.580887372013652e-05,
+      "loss": 0.3998,
+      "step": 5210
+    },
+    {
+      "epoch": 0.43793324034942377,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.5740614334470995e-05,
+      "loss": 0.365,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4387721929171429,
+      "grad_norm": 0.36328125,
+      "learning_rate": 4.567235494880547e-05,
+      "loss": 0.3782,
+      "step": 5230
+    },
+    {
+      "epoch": 0.43961114548486213,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.5604095563139933e-05,
+      "loss": 0.362,
+      "step": 5240
+    },
+    {
+      "epoch": 0.44045009805258134,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.5535836177474406e-05,
+      "loss": 0.3825,
+      "step": 5250
+    },
+    {
+      "epoch": 0.44128905062030055,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.546757679180887e-05,
+      "loss": 0.363,
+      "step": 5260
+    },
+    {
+      "epoch": 0.44212800318801976,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.539931740614335e-05,
+      "loss": 0.3932,
+      "step": 5270
+    },
+    {
+      "epoch": 0.44296695575573897,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.533105802047782e-05,
+      "loss": 0.3776,
+      "step": 5280
+    },
+    {
+      "epoch": 0.4438059083234582,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.526279863481229e-05,
+      "loss": 0.3596,
+      "step": 5290
+    },
+    {
+      "epoch": 0.4446448608911774,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.519453924914676e-05,
+      "loss": 0.3707,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4454838134588966,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.512627986348123e-05,
+      "loss": 0.3773,
+      "step": 5310
+    },
+    {
+      "epoch": 0.44632276602661575,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.50580204778157e-05,
+      "loss": 0.3686,
+      "step": 5320
+    },
+    {
+      "epoch": 0.44716171859433496,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.498976109215018e-05,
+      "loss": 0.3864,
+      "step": 5330
+    },
+    {
+      "epoch": 0.44800067116205416,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.4921501706484644e-05,
+      "loss": 0.3432,
+      "step": 5340
+    },
+    {
+      "epoch": 0.4488396237297734,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.485324232081912e-05,
+      "loss": 0.3807,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4496785762974926,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.478498293515359e-05,
+      "loss": 0.4029,
+      "step": 5360
+    },
+    {
+      "epoch": 0.4505175288652118,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.4716723549488055e-05,
+      "loss": 0.3901,
+      "step": 5370
+    },
+    {
+      "epoch": 0.451356481432931,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.4648464163822534e-05,
+      "loss": 0.3755,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4521954340006502,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.4580204778157e-05,
+      "loss": 0.3339,
+      "step": 5390
+    },
+    {
+      "epoch": 0.4530343865683694,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.451194539249147e-05,
+      "loss": 0.3822,
+      "step": 5400
+    },
+    {
+      "epoch": 0.45387333913608857,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.4443686006825945e-05,
+      "loss": 0.3514,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4547122917038078,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.437542662116041e-05,
+      "loss": 0.3611,
+      "step": 5420
+    },
+    {
+      "epoch": 0.455551244271527,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.430716723549488e-05,
+      "loss": 0.3644,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4563901968392462,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.423890784982935e-05,
+      "loss": 0.3892,
+      "step": 5440
+    },
+    {
+      "epoch": 0.4572291494069654,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.417064846416383e-05,
+      "loss": 0.3801,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4580681019746846,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.41023890784983e-05,
+      "loss": 0.3483,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4589070545424038,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.4034129692832766e-05,
+      "loss": 0.3817,
+      "step": 5470
+    },
+    {
+      "epoch": 0.45974600711012303,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.396587030716724e-05,
+      "loss": 0.3588,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4605849596778422,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.389761092150172e-05,
+      "loss": 0.3589,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4614239122455614,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.3829351535836177e-05,
+      "loss": 0.393,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4622628648132806,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.3761092150170656e-05,
+      "loss": 0.3636,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4631018173809998,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.369283276450512e-05,
+      "loss": 0.3836,
+      "step": 5520
+    },
+    {
+      "epoch": 0.463940769948719,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.3624573378839594e-05,
+      "loss": 0.3607,
+      "step": 5530
+    },
+    {
+      "epoch": 0.46477972251643823,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.3556313993174066e-05,
+      "loss": 0.3756,
+      "step": 5540
+    },
+    {
+      "epoch": 0.46561867508415744,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.348805460750853e-05,
+      "loss": 0.3677,
+      "step": 5550
+    },
+    {
+      "epoch": 0.46645762765187665,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.341979522184301e-05,
+      "loss": 0.3466,
+      "step": 5560
+    },
+    {
+      "epoch": 0.46729658021959586,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.335153583617748e-05,
+      "loss": 0.3478,
+      "step": 5570
+    },
+    {
+      "epoch": 0.468135532787315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.328327645051195e-05,
+      "loss": 0.3491,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4689744853550342,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.321501706484642e-05,
+      "loss": 0.3583,
+      "step": 5590
+    },
+    {
+      "epoch": 0.46981343792275343,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.314675767918089e-05,
+      "loss": 0.4017,
+      "step": 5600
+    },
+    {
+      "epoch": 0.47065239049047264,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.307849829351536e-05,
+      "loss": 0.355,
+      "step": 5610
+    },
+    {
+      "epoch": 0.47149134305819185,
+      "grad_norm": 0.380859375,
+      "learning_rate": 4.301023890784984e-05,
+      "loss": 0.3713,
+      "step": 5620
+    },
+    {
+      "epoch": 0.47233029562591106,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.2941979522184305e-05,
+      "loss": 0.3893,
+      "step": 5630
+    },
+    {
+      "epoch": 0.47316924819363027,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.287372013651878e-05,
+      "loss": 0.3784,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4740082007613495,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.280546075085324e-05,
+      "loss": 0.3578,
+      "step": 5650
+    },
+    {
+      "epoch": 0.4748471533290687,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.2737201365187716e-05,
+      "loss": 0.3649,
+      "step": 5660
+    },
+    {
+      "epoch": 0.47568610589678784,
+      "grad_norm": 0.375,
+      "learning_rate": 4.2668941979522195e-05,
+      "loss": 0.3454,
+      "step": 5670
+    },
+    {
+      "epoch": 0.47652505846450705,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.260068259385666e-05,
+      "loss": 0.3565,
+      "step": 5680
+    },
+    {
+      "epoch": 0.47736401103222625,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.253242320819113e-05,
+      "loss": 0.3402,
+      "step": 5690
+    },
+    {
+      "epoch": 0.47820296359994546,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.24641638225256e-05,
+      "loss": 0.359,
+      "step": 5700
+    },
+    {
+      "epoch": 0.47904191616766467,
+      "grad_norm": 0.337890625,
+      "learning_rate": 4.239590443686007e-05,
+      "loss": 0.3324,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4798808687353839,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.2327645051194543e-05,
+      "loss": 0.376,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4807198213031031,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.225938566552901e-05,
+      "loss": 0.3775,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4815587738708223,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.219112627986349e-05,
+      "loss": 0.3478,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4823977264385415,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.2122866894197954e-05,
+      "loss": 0.3594,
+      "step": 5750
+    },
+    {
+      "epoch": 0.48323667900626066,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.2054607508532426e-05,
+      "loss": 0.3592,
+      "step": 5760
+    },
+    {
+      "epoch": 0.48407563157397987,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.19863481228669e-05,
+      "loss": 0.355,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4849145841416991,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.1918088737201365e-05,
+      "loss": 0.3213,
+      "step": 5780
+    },
+    {
+      "epoch": 0.4857535367094183,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.184982935153584e-05,
+      "loss": 0.3554,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4865924892771375,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.1781569965870316e-05,
+      "loss": 0.4197,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4874314418448567,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.171331058020478e-05,
+      "loss": 0.3765,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4882703944125759,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.1645051194539254e-05,
+      "loss": 0.3412,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4891093469802951,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.157679180887372e-05,
+      "loss": 0.3748,
+      "step": 5830
+    },
+    {
+      "epoch": 0.4899482995480143,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.150853242320819e-05,
+      "loss": 0.3282,
+      "step": 5840
+    },
+    {
+      "epoch": 0.4907872521157335,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.144027303754267e-05,
+      "loss": 0.3495,
+      "step": 5850
+    },
+    {
+      "epoch": 0.4916262046834527,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.137201365187714e-05,
+      "loss": 0.3633,
+      "step": 5860
+    },
+    {
+      "epoch": 0.4924651572511719,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.130375426621161e-05,
+      "loss": 0.3537,
+      "step": 5870
+    },
+    {
+      "epoch": 0.4933041098188911,
+      "grad_norm": 0.376953125,
+      "learning_rate": 4.1235494880546076e-05,
+      "loss": 0.3498,
+      "step": 5880
+    },
+    {
+      "epoch": 0.4941430623866103,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.116723549488055e-05,
+      "loss": 0.3607,
+      "step": 5890
+    },
+    {
+      "epoch": 0.49498201495432953,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.109897610921502e-05,
+      "loss": 0.3749,
+      "step": 5900
+    },
+    {
+      "epoch": 0.49582096752204874,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.1030716723549486e-05,
+      "loss": 0.3455,
+      "step": 5910
+    },
+    {
+      "epoch": 0.49665992008976795,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.0962457337883965e-05,
+      "loss": 0.3392,
+      "step": 5920
+    },
+    {
+      "epoch": 0.4974988726574871,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.089419795221844e-05,
+      "loss": 0.3759,
+      "step": 5930
+    },
+    {
+      "epoch": 0.4983378252252063,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.0825938566552904e-05,
+      "loss": 0.3544,
+      "step": 5940
+    },
+    {
+      "epoch": 0.4991767777929255,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.0757679180887376e-05,
+      "loss": 0.3617,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5000157303606447,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.068941979522184e-05,
+      "loss": 0.3591,
+      "step": 5960
+    },
+    {
+      "epoch": 0.500854682928364,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.0621160409556314e-05,
+      "loss": 0.3589,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5016936354960831,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.055290102389079e-05,
+      "loss": 0.3423,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5025325880638023,
+      "grad_norm": 0.375,
+      "learning_rate": 4.048464163822526e-05,
+      "loss": 0.3565,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5033715406315216,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.041638225255973e-05,
+      "loss": 0.3573,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5042104931992407,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.03481228668942e-05,
+      "loss": 0.3922,
+      "step": 6010
+    },
+    {
+      "epoch": 0.50504944576696,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.027986348122867e-05,
+      "loss": 0.378,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5058883983346791,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.021160409556315e-05,
+      "loss": 0.3494,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5067273509023984,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.0143344709897615e-05,
+      "loss": 0.3469,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5075663034701176,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.007508532423209e-05,
+      "loss": 0.337,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5084052560378367,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.000682593856656e-05,
+      "loss": 0.3258,
+      "step": 6060
+    },
+    {
+      "epoch": 0.509244208605556,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.9938566552901025e-05,
+      "loss": 0.3762,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5100831611732751,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.98703071672355e-05,
+      "loss": 0.3797,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5109221137409944,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.980204778156997e-05,
+      "loss": 0.3369,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5117610663087135,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.973378839590444e-05,
+      "loss": 0.3597,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5126000188764328,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.966552901023891e-05,
+      "loss": 0.3399,
+      "step": 6110
+    },
+    {
+      "epoch": 0.513438971444152,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.959726962457338e-05,
+      "loss": 0.3617,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5142779240118712,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.952901023890785e-05,
+      "loss": 0.3236,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5151168765795904,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.9460750853242325e-05,
+      "loss": 0.3636,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5159558291473095,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.939249146757679e-05,
+      "loss": 0.3515,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5167947817150288,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.9324232081911264e-05,
+      "loss": 0.374,
+      "step": 6160
+    },
+    {
+      "epoch": 0.517633734282748,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.9255972696245736e-05,
+      "loss": 0.335,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5184726868504672,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.918771331058021e-05,
+      "loss": 0.368,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5193116394181864,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.911945392491468e-05,
+      "loss": 0.3317,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5201505919859056,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.905119453924915e-05,
+      "loss": 0.3271,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5209895445536248,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.8982935153583626e-05,
+      "loss": 0.3437,
+      "step": 6210
+    },
+    {
+      "epoch": 0.521828497121344,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.891467576791809e-05,
+      "loss": 0.3447,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5226674496890632,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.8846416382252564e-05,
+      "loss": 0.3611,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5235064022567824,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.877815699658703e-05,
+      "loss": 0.376,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5243453548245016,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.87098976109215e-05,
+      "loss": 0.3582,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5251843073922208,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.8641638225255975e-05,
+      "loss": 0.3765,
+      "step": 6260
+    },
+    {
+      "epoch": 0.52602325995994,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.857337883959045e-05,
+      "loss": 0.3421,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5268622125276592,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.850511945392492e-05,
+      "loss": 0.3694,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5277011650953785,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.8436860068259385e-05,
+      "loss": 0.3709,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5285401176630976,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.8368600682593864e-05,
+      "loss": 0.3469,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5293790702308169,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.830034129692833e-05,
+      "loss": 0.3826,
+      "step": 6310
+    },
+    {
+      "epoch": 0.530218022798536,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.82320819112628e-05,
+      "loss": 0.3406,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5310569753662552,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.8163822525597275e-05,
+      "loss": 0.3424,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5318959279339744,
+      "grad_norm": 0.353515625,
+      "learning_rate": 3.809556313993174e-05,
+      "loss": 0.3457,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5327348805016936,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.802730375426621e-05,
+      "loss": 0.3666,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5335738330694129,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.7959044368600686e-05,
+      "loss": 0.3582,
+      "step": 6360
+    },
+    {
+      "epoch": 0.534412785637132,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.789078498293516e-05,
+      "loss": 0.3327,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5352517382048513,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.7822525597269624e-05,
+      "loss": 0.3599,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5360906907725704,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.77542662116041e-05,
+      "loss": 0.3429,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5369296433402897,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.768600682593857e-05,
+      "loss": 0.3605,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5377685959080089,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.761774744027304e-05,
+      "loss": 0.3699,
+      "step": 6410
+    },
+    {
+      "epoch": 0.538607548475728,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.7549488054607514e-05,
+      "loss": 0.3136,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5394465010434473,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.7481228668941986e-05,
+      "loss": 0.3854,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5402854536111664,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.741296928327645e-05,
+      "loss": 0.3788,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5411244061788857,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.7344709897610924e-05,
+      "loss": 0.3655,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5419633587466048,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.7276450511945397e-05,
+      "loss": 0.3541,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5428023113143241,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.720819112627986e-05,
+      "loss": 0.3621,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5436412638820433,
+      "grad_norm": 0.375,
+      "learning_rate": 3.713993174061434e-05,
+      "loss": 0.3473,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5444802164497625,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.707167235494881e-05,
+      "loss": 0.3243,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5453191690174817,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.700341296928328e-05,
+      "loss": 0.3351,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5461581215852008,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.693515358361775e-05,
+      "loss": 0.3426,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5469970741529201,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.6866894197952224e-05,
+      "loss": 0.3556,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5478360267206392,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.679863481228669e-05,
+      "loss": 0.3233,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5486749792883585,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.673037542662116e-05,
+      "loss": 0.3248,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5495139318560777,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.6662116040955635e-05,
+      "loss": 0.3561,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5503528844237969,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.659385665529011e-05,
+      "loss": 0.3514,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5511918369915161,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.652559726962458e-05,
+      "loss": 0.3569,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5520307895592353,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.6457337883959046e-05,
+      "loss": 0.3809,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5528697421269545,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.638907849829352e-05,
+      "loss": 0.3447,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5537086946946737,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.632081911262799e-05,
+      "loss": 0.3543,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5545476472623929,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.625255972696246e-05,
+      "loss": 0.3384,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5553865998301121,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.618430034129693e-05,
+      "loss": 0.354,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5562255523978313,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.61160409556314e-05,
+      "loss": 0.3303,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5570645049655505,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.6047781569965874e-05,
+      "loss": 0.356,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5579034575332698,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.5979522184300346e-05,
+      "loss": 0.3523,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5587424101009889,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.591126279863482e-05,
+      "loss": 0.3382,
+      "step": 6660
+    },
+    {
+      "epoch": 0.5595813626687082,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.5843003412969284e-05,
+      "loss": 0.3311,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5604203152364273,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.577474402730376e-05,
+      "loss": 0.3281,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5612592678041465,
+      "grad_norm": 0.375,
+      "learning_rate": 3.570648464163823e-05,
+      "loss": 0.3499,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5620982203718657,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.56382252559727e-05,
+      "loss": 0.3492,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5629371729395849,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.556996587030717e-05,
+      "loss": 0.3639,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5637761255073042,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.550170648464164e-05,
+      "loss": 0.3119,
+      "step": 6720
+    },
+    {
+      "epoch": 0.5646150780750233,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.543344709897611e-05,
+      "loss": 0.3589,
+      "step": 6730
+    },
+    {
+      "epoch": 0.5654540306427426,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.5365187713310585e-05,
+      "loss": 0.3621,
+      "step": 6740
+    },
+    {
+      "epoch": 0.5662929832104617,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.529692832764506e-05,
+      "loss": 0.3264,
+      "step": 6750
+    },
+    {
+      "epoch": 0.567131935778181,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.522866894197952e-05,
+      "loss": 0.3505,
+      "step": 6760
+    },
+    {
+      "epoch": 0.5679708883459001,
+      "grad_norm": 0.375,
+      "learning_rate": 3.5160409556313995e-05,
+      "loss": 0.3302,
+      "step": 6770
+    },
+    {
+      "epoch": 0.5688098409136193,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.509215017064847e-05,
+      "loss": 0.3317,
+      "step": 6780
+    },
+    {
+      "epoch": 0.5696487934813386,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.502389078498294e-05,
+      "loss": 0.3364,
+      "step": 6790
+    },
+    {
+      "epoch": 0.5704877460490577,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.4955631399317406e-05,
+      "loss": 0.3216,
+      "step": 6800
+    },
+    {
+      "epoch": 0.571326698616777,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.488737201365188e-05,
+      "loss": 0.3494,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5721656511844961,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.481911262798635e-05,
+      "loss": 0.3391,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5730046037522154,
+      "grad_norm": 0.375,
+      "learning_rate": 3.475085324232082e-05,
+      "loss": 0.3256,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5738435563199346,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.4682593856655296e-05,
+      "loss": 0.3261,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5746825088876537,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.461433447098976e-05,
+      "loss": 0.3597,
+      "step": 6850
+    },
+    {
+      "epoch": 0.575521461455373,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.4546075085324234e-05,
+      "loss": 0.3247,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5763604140230921,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.4477815699658706e-05,
+      "loss": 0.3373,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5771993665908114,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.440955631399318e-05,
+      "loss": 0.3392,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5780383191585305,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.4341296928327644e-05,
+      "loss": 0.3333,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5788772717262498,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.427303754266212e-05,
+      "loss": 0.3332,
+      "step": 6900
+    },
+    {
+      "epoch": 0.579716224293969,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.420477815699659e-05,
+      "loss": 0.3262,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5805551768616882,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.413651877133106e-05,
+      "loss": 0.3367,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5813941294294074,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.4068259385665534e-05,
+      "loss": 0.3586,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5822330819971265,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.4e-05,
+      "loss": 0.3445,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5830720345648458,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.393174061433447e-05,
+      "loss": 0.3563,
+      "step": 6950
+    },
+    {
+      "epoch": 0.583910987132565,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.3863481228668945e-05,
+      "loss": 0.3265,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5847499397002842,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.379522184300342e-05,
+      "loss": 0.3438,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5855888922680034,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.372696245733789e-05,
+      "loss": 0.3318,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5864278448357226,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.3658703071672355e-05,
+      "loss": 0.3124,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5872667974034418,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.359044368600683e-05,
+      "loss": 0.3352,
+      "step": 7000
+    },
+    {
+      "epoch": 0.588105749971161,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.35221843003413e-05,
+      "loss": 0.3296,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5889447025388802,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.345392491467577e-05,
+      "loss": 0.3408,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5897836551065994,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.338566552901024e-05,
+      "loss": 0.381,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5906226076743186,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.331740614334471e-05,
+      "loss": 0.3506,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5914615602420378,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.324914675767918e-05,
+      "loss": 0.3229,
+      "step": 7050
+    },
+    {
+      "epoch": 0.592300512809757,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.3180887372013656e-05,
+      "loss": 0.3506,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5931394653774762,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.311262798634813e-05,
+      "loss": 0.3208,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5939784179451955,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.3044368600682594e-05,
+      "loss": 0.3423,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5948173705129146,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.2976109215017066e-05,
+      "loss": 0.3651,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5956563230806339,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.290784982935154e-05,
+      "loss": 0.3303,
+      "step": 7100
+    },
+    {
+      "epoch": 0.596495275648353,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.283959044368601e-05,
+      "loss": 0.3473,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5973342282160722,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.277133105802048e-05,
+      "loss": 0.3125,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5981731807837914,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.2703071672354956e-05,
+      "loss": 0.3107,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5990121333515106,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.263481228668942e-05,
+      "loss": 0.3162,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5998510859192299,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.2566552901023894e-05,
+      "loss": 0.3519,
+      "step": 7150
+    },
+    {
+      "epoch": 0.600690038486949,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.2498293515358367e-05,
+      "loss": 0.3237,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6015289910546683,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.243003412969283e-05,
+      "loss": 0.3399,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6023679436223874,
+      "grad_norm": 0.36328125,
+      "learning_rate": 3.2361774744027305e-05,
+      "loss": 0.3439,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6032068961901067,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.229351535836178e-05,
+      "loss": 0.3279,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6040458487578259,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.222525597269625e-05,
+      "loss": 0.343,
+      "step": 7200
+    },
+    {
+      "epoch": 0.604884801325545,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.2156996587030715e-05,
+      "loss": 0.3484,
+      "step": 7210
+    },
+    {
+      "epoch": 0.6057237538932643,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.2088737201365195e-05,
+      "loss": 0.333,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6065627064609834,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.202047781569966e-05,
+      "loss": 0.3551,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6074016590287027,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.195221843003413e-05,
+      "loss": 0.3482,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6082406115964218,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.1883959044368605e-05,
+      "loss": 0.3644,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6090795641641411,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.181569965870308e-05,
+      "loss": 0.3326,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6099185167318603,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.174744027303754e-05,
+      "loss": 0.3247,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6107574692995795,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.1679180887372016e-05,
+      "loss": 0.3696,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6115964218672987,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.161092150170649e-05,
+      "loss": 0.3406,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6124353744350178,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.1542662116040954e-05,
+      "loss": 0.3316,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6132743270027371,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.147440273037543e-05,
+      "loss": 0.317,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6141132795704562,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.14061433447099e-05,
+      "loss": 0.3533,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6149522321381755,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.133788395904437e-05,
+      "loss": 0.3467,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6157911847058947,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.1269624573378844e-05,
+      "loss": 0.3187,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6166301372736139,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.1201365187713316e-05,
+      "loss": 0.3329,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6174690898413331,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.113310580204778e-05,
+      "loss": 0.344,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6183080424090523,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.1064846416382254e-05,
+      "loss": 0.3199,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6191469949767715,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.099658703071673e-05,
+      "loss": 0.3271,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6199859475444907,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.092832764505119e-05,
+      "loss": 0.3497,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6208249001122099,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.086006825938567e-05,
+      "loss": 0.3234,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6216638526799291,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.079180887372014e-05,
+      "loss": 0.3571,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6225028052476483,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.072354948805461e-05,
+      "loss": 0.3376,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6233417578153675,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.065529010238908e-05,
+      "loss": 0.3077,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6241807103830868,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.0587030716723555e-05,
+      "loss": 0.336,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6250196629508059,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.051877133105802e-05,
+      "loss": 0.3266,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6258586155185252,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.0450511945392493e-05,
+      "loss": 0.3642,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6266975680862443,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.0382252559726965e-05,
+      "loss": 0.3294,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6275365206539635,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.0313993174061438e-05,
+      "loss": 0.3365,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6283754732216827,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.0245733788395907e-05,
+      "loss": 0.3473,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6292144257894019,
+      "grad_norm": 0.375,
+      "learning_rate": 3.017747440273038e-05,
+      "loss": 0.3268,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6300533783571212,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.0109215017064848e-05,
+      "loss": 0.3146,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6308923309248403,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.0040955631399317e-05,
+      "loss": 0.3401,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6317312834925596,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.9972696245733793e-05,
+      "loss": 0.3098,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6325702360602787,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.9904436860068262e-05,
+      "loss": 0.306,
+      "step": 7540
+    },
+    {
+      "epoch": 0.633409188627998,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.983617747440273e-05,
+      "loss": 0.3441,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6342481411957172,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.9767918088737204e-05,
+      "loss": 0.3614,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6350870937634363,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.9699658703071676e-05,
+      "loss": 0.3234,
+      "step": 7570
+    },
+    {
+      "epoch": 0.6359260463311556,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.9631399317406145e-05,
+      "loss": 0.2922,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6367649988988747,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.9563139931740618e-05,
+      "loss": 0.3327,
+      "step": 7590
+    },
+    {
+      "epoch": 0.637603951466594,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.9494880546075087e-05,
+      "loss": 0.3221,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6384429040343131,
+      "grad_norm": 0.375,
+      "learning_rate": 2.942662116040956e-05,
+      "loss": 0.3255,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6392818566020324,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.935836177474403e-05,
+      "loss": 0.2936,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6401208091697516,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.92901023890785e-05,
+      "loss": 0.3064,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6409597617374707,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.922184300341297e-05,
+      "loss": 0.3213,
+      "step": 7640
+    },
+    {
+      "epoch": 0.64179871430519,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.9153583617747442e-05,
+      "loss": 0.3455,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6426376668729091,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.9085324232081915e-05,
+      "loss": 0.3068,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6434766194406284,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.9017064846416384e-05,
+      "loss": 0.3188,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6443155720083475,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.8948805460750856e-05,
+      "loss": 0.2951,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6451545245760668,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.8880546075085325e-05,
+      "loss": 0.3078,
+      "step": 7690
+    },
+    {
+      "epoch": 0.645993477143786,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.88122866894198e-05,
+      "loss": 0.3267,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6468324297115052,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.874402730375427e-05,
+      "loss": 0.3431,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6476713822792244,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.867576791808874e-05,
+      "loss": 0.3016,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6485103348469435,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.8607508532423208e-05,
+      "loss": 0.3218,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6493492874146628,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.853924914675768e-05,
+      "loss": 0.3162,
+      "step": 7740
+    },
+    {
+      "epoch": 0.650188239982382,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.8470989761092153e-05,
+      "loss": 0.2939,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6510271925501012,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.8402730375426622e-05,
+      "loss": 0.3235,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6518661451178204,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.8334470989761095e-05,
+      "loss": 0.3244,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6527050976855396,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.8266211604095564e-05,
+      "loss": 0.3491,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6535440502532588,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.819795221843004e-05,
+      "loss": 0.3061,
+      "step": 7790
+    },
+    {
+      "epoch": 0.654383002820978,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.812969283276451e-05,
+      "loss": 0.3574,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6552219553886972,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.8061433447098978e-05,
+      "loss": 0.3341,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6560609079564164,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.7993174061433447e-05,
+      "loss": 0.333,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6568998605241356,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.7924914675767923e-05,
+      "loss": 0.3102,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6577388130918548,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.7856655290102392e-05,
+      "loss": 0.3315,
+      "step": 7840
+    },
+    {
+      "epoch": 0.658577765659574,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.778839590443686e-05,
+      "loss": 0.3534,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6594167182272932,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.7720136518771333e-05,
+      "loss": 0.3407,
+      "step": 7860
+    },
+    {
+      "epoch": 0.6602556707950125,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.7651877133105802e-05,
+      "loss": 0.2997,
+      "step": 7870
+    },
+    {
+      "epoch": 0.6610946233627316,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.7583617747440278e-05,
+      "loss": 0.3279,
+      "step": 7880
+    },
+    {
+      "epoch": 0.6619335759304509,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.7515358361774747e-05,
+      "loss": 0.3335,
+      "step": 7890
+    },
+    {
+      "epoch": 0.66277252849817,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.7447098976109216e-05,
+      "loss": 0.3275,
+      "step": 7900
+    },
+    {
+      "epoch": 0.6636114810658892,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.7378839590443685e-05,
+      "loss": 0.3384,
+      "step": 7910
+    },
+    {
+      "epoch": 0.6644504336336085,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.731058020477816e-05,
+      "loss": 0.3416,
+      "step": 7920
+    },
+    {
+      "epoch": 0.6652893862013276,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.724232081911263e-05,
+      "loss": 0.3141,
+      "step": 7930
+    },
+    {
+      "epoch": 0.6661283387690469,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.71740614334471e-05,
+      "loss": 0.2982,
+      "step": 7940
+    },
+    {
+      "epoch": 0.666967291336766,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.7105802047781572e-05,
+      "loss": 0.2999,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6678062439044853,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.7037542662116044e-05,
+      "loss": 0.3125,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6686451964722044,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.6969283276450517e-05,
+      "loss": 0.3205,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6694841490399237,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.6901023890784986e-05,
+      "loss": 0.3469,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6703231016076429,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.6832764505119455e-05,
+      "loss": 0.3269,
+      "step": 7990
+    },
+    {
+      "epoch": 0.671162054175362,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.6764505119453924e-05,
+      "loss": 0.3178,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6720010067430813,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.66962457337884e-05,
+      "loss": 0.3161,
+      "step": 8010
+    },
+    {
+      "epoch": 0.6728399593108004,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.662798634812287e-05,
+      "loss": 0.3221,
+      "step": 8020
+    },
+    {
+      "epoch": 0.6736789118785197,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.6559726962457338e-05,
+      "loss": 0.3135,
+      "step": 8030
+    },
+    {
+      "epoch": 0.6745178644462388,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.649146757679181e-05,
+      "loss": 0.3181,
+      "step": 8040
+    },
+    {
+      "epoch": 0.6753568170139581,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.6423208191126283e-05,
+      "loss": 0.3273,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6761957695816773,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.6354948805460755e-05,
+      "loss": 0.3148,
+      "step": 8060
+    },
+    {
+      "epoch": 0.6770347221493965,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.6286689419795224e-05,
+      "loss": 0.3343,
+      "step": 8070
+    },
+    {
+      "epoch": 0.6778736747171157,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.6218430034129693e-05,
+      "loss": 0.3092,
+      "step": 8080
+    },
+    {
+      "epoch": 0.6787126272848348,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.6150170648464162e-05,
+      "loss": 0.3048,
+      "step": 8090
+    },
+    {
+      "epoch": 0.6795515798525541,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.6081911262798638e-05,
+      "loss": 0.2975,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6803905324202733,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.6013651877133107e-05,
+      "loss": 0.3113,
+      "step": 8110
+    },
+    {
+      "epoch": 0.6812294849879925,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.594539249146758e-05,
+      "loss": 0.3479,
+      "step": 8120
+    },
+    {
+      "epoch": 0.6820684375557117,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.587713310580205e-05,
+      "loss": 0.3433,
+      "step": 8130
+    },
+    {
+      "epoch": 0.6829073901234309,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.580887372013652e-05,
+      "loss": 0.3153,
+      "step": 8140
+    },
+    {
+      "epoch": 0.6837463426911501,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.5740614334470994e-05,
+      "loss": 0.3238,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6845852952588694,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.5672354948805463e-05,
+      "loss": 0.3233,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6854242478265885,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.5604095563139932e-05,
+      "loss": 0.3207,
+      "step": 8170
+    },
+    {
+      "epoch": 0.6862632003943077,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.5535836177474408e-05,
+      "loss": 0.3189,
+      "step": 8180
+    },
+    {
+      "epoch": 0.6871021529620269,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.5467576791808877e-05,
+      "loss": 0.312,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6879411055297461,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.5399317406143346e-05,
+      "loss": 0.3189,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6887800580974653,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.5331058020477818e-05,
+      "loss": 0.3152,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6896190106651845,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.5262798634812287e-05,
+      "loss": 0.3393,
+      "step": 8220
+    },
+    {
+      "epoch": 0.6904579632329038,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.519453924914676e-05,
+      "loss": 0.3043,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6912969158006229,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.5126279863481232e-05,
+      "loss": 0.3183,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6921358683683422,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.50580204778157e-05,
+      "loss": 0.312,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6929748209360613,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.498976109215017e-05,
+      "loss": 0.3176,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6938137735037805,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.4921501706484646e-05,
+      "loss": 0.3067,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6946527260714997,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.4853242320819115e-05,
+      "loss": 0.3491,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6954916786392189,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.4784982935153584e-05,
+      "loss": 0.3276,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6963306312069382,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.4716723549488057e-05,
+      "loss": 0.3044,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6971695837746573,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.464846416382253e-05,
+      "loss": 0.3063,
+      "step": 8310
+    },
+    {
+      "epoch": 0.6980085363423766,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.4580204778157e-05,
+      "loss": 0.3146,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6988474889100957,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.451194539249147e-05,
+      "loss": 0.3358,
+      "step": 8330
+    },
+    {
+      "epoch": 0.699686441477815,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.444368600682594e-05,
+      "loss": 0.3146,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7005253940455342,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.437542662116041e-05,
+      "loss": 0.3078,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7013643466132533,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.4307167235494885e-05,
+      "loss": 0.3059,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7022032991809726,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.4238907849829354e-05,
+      "loss": 0.3047,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7030422517486917,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.4170648464163823e-05,
+      "loss": 0.3344,
+      "step": 8380
+    },
+    {
+      "epoch": 0.703881204316411,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.4102389078498295e-05,
+      "loss": 0.3081,
+      "step": 8390
+    },
+    {
+      "epoch": 0.7047201568841301,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.4034129692832768e-05,
+      "loss": 0.2912,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7055591094518494,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.3965870307167237e-05,
+      "loss": 0.3128,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7063980620195686,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.389761092150171e-05,
+      "loss": 0.3087,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7072370145872878,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.382935153583618e-05,
+      "loss": 0.3116,
+      "step": 8430
+    },
+    {
+      "epoch": 0.708075967155007,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3761092150170654e-05,
+      "loss": 0.321,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7089149197227261,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.3692832764505123e-05,
+      "loss": 0.2924,
+      "step": 8450
+    },
+    {
+      "epoch": 0.7097538722904454,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.3624573378839592e-05,
+      "loss": 0.3071,
+      "step": 8460
+    },
+    {
+      "epoch": 0.7105928248581646,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.355631399317406e-05,
+      "loss": 0.3217,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7114317774258838,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.3488054607508534e-05,
+      "loss": 0.3306,
+      "step": 8480
+    },
+    {
+      "epoch": 0.712270729993603,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.3419795221843006e-05,
+      "loss": 0.3178,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7131096825613222,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.3351535836177475e-05,
+      "loss": 0.3169,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7139486351290414,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.3283276450511948e-05,
+      "loss": 0.2919,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7147875876967605,
+      "grad_norm": 0.376953125,
+      "learning_rate": 2.3215017064846417e-05,
+      "loss": 0.3017,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7156265402644798,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.3146757679180893e-05,
+      "loss": 0.3285,
+      "step": 8530
+    },
+    {
+      "epoch": 0.716465492832199,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.3078498293515362e-05,
+      "loss": 0.302,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7173044453999182,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.301023890784983e-05,
+      "loss": 0.3347,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7181433979676374,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.29419795221843e-05,
+      "loss": 0.3461,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7189823505353566,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.2873720136518772e-05,
+      "loss": 0.3167,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7198213031030758,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.2805460750853245e-05,
+      "loss": 0.3068,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7206602556707951,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.2737201365187714e-05,
+      "loss": 0.3143,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7214992082385142,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.2668941979522186e-05,
+      "loss": 0.3136,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7223381608062334,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.2600682593856655e-05,
+      "loss": 0.298,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7231771133739526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.253242320819113e-05,
+      "loss": 0.3019,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7240160659416718,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.24641638225256e-05,
+      "loss": 0.3116,
+      "step": 8630
+    },
+    {
+      "epoch": 0.724855018509391,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.239590443686007e-05,
+      "loss": 0.2842,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7256939710771102,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.232764505119454e-05,
+      "loss": 0.3347,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7265329236448295,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.2259385665529014e-05,
+      "loss": 0.2836,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7273718762125486,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.2191126279863483e-05,
+      "loss": 0.3001,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7282108287802679,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.2122866894197952e-05,
+      "loss": 0.3101,
+      "step": 8680
+    },
+    {
+      "epoch": 0.729049781347987,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.2054607508532425e-05,
+      "loss": 0.3106,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7298887339157062,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.1986348122866894e-05,
+      "loss": 0.3473,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7307276864834255,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.191808873720137e-05,
+      "loss": 0.3223,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7315666390511446,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.184982935153584e-05,
+      "loss": 0.3213,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7324055916188639,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.1781569965870308e-05,
+      "loss": 0.3069,
+      "step": 8730
+    },
+    {
+      "epoch": 0.733244544186583,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.1713310580204777e-05,
+      "loss": 0.3324,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7340834967543023,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.1645051194539253e-05,
+      "loss": 0.3394,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7349224493220214,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.1576791808873722e-05,
+      "loss": 0.3097,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7357614018897407,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.1508532423208194e-05,
+      "loss": 0.3354,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7366003544574599,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.1440273037542663e-05,
+      "loss": 0.2888,
+      "step": 8780
+    },
+    {
+      "epoch": 0.737439307025179,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1372013651877136e-05,
+      "loss": 0.3024,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7382782595928983,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.1303754266211608e-05,
+      "loss": 0.3279,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7391172121606174,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.1235494880546077e-05,
+      "loss": 0.3129,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7399561647283367,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.1167235494880546e-05,
+      "loss": 0.291,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7407951172960558,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.1098976109215015e-05,
+      "loss": 0.285,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7416340698637751,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.103071672354949e-05,
+      "loss": 0.332,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7424730224314943,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.096245733788396e-05,
+      "loss": 0.3272,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7433119749992135,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.0894197952218433e-05,
+      "loss": 0.2977,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7441509275669327,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.0825938566552902e-05,
+      "loss": 0.3118,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7449898801346518,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.0757679180887374e-05,
+      "loss": 0.3112,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7458288327023711,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.0689419795221847e-05,
+      "loss": 0.3137,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7466677852700903,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.0621160409556316e-05,
+      "loss": 0.3162,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7475067378378095,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.0552901023890785e-05,
+      "loss": 0.3101,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7483456904055287,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.0484641638225254e-05,
+      "loss": 0.3113,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7491846429732479,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.041638225255973e-05,
+      "loss": 0.2909,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7500235955409671,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.03481228668942e-05,
+      "loss": 0.3233,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7508625481086864,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.027986348122867e-05,
+      "loss": 0.2976,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7517015006764055,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.021160409556314e-05,
+      "loss": 0.3359,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7525404532441247,
+      "grad_norm": 0.376953125,
+      "learning_rate": 2.0143344709897613e-05,
+      "loss": 0.308,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7533794058118439,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.0075085324232085e-05,
+      "loss": 0.3131,
+      "step": 8980
+    },
+    {
+      "epoch": 0.7542183583795631,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.0006825938566554e-05,
+      "loss": 0.3192,
+      "step": 8990
+    },
+    {
+      "epoch": 0.7550573109472823,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.9938566552901027e-05,
+      "loss": 0.3064,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7558962635150015,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.9870307167235496e-05,
+      "loss": 0.2888,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7567352160827208,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.980204778156997e-05,
+      "loss": 0.3259,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7575741686504399,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9733788395904437e-05,
+      "loss": 0.3083,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7584131212181592,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.966552901023891e-05,
+      "loss": 0.3019,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7592520737858783,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9597269624573382e-05,
+      "loss": 0.2958,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7600910263535975,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.952901023890785e-05,
+      "loss": 0.292,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7609299789213168,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9460750853242324e-05,
+      "loss": 0.3166,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7617689314890359,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9392491467576793e-05,
+      "loss": 0.3226,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7626078840567552,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9324232081911265e-05,
+      "loss": 0.3234,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7634468366244743,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9255972696245734e-05,
+      "loss": 0.3238,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7642857891921936,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.9187713310580207e-05,
+      "loss": 0.3389,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7651247417599127,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.9119453924914676e-05,
+      "loss": 0.3019,
+      "step": 9120
+    },
+    {
+      "epoch": 0.765963694327632,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.905119453924915e-05,
+      "loss": 0.3517,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7668026468953512,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.898293515358362e-05,
+      "loss": 0.2821,
+      "step": 9140
+    },
+    {
+      "epoch": 0.7676415994630703,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.891467576791809e-05,
+      "loss": 0.318,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7684805520307896,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8846416382252562e-05,
+      "loss": 0.3093,
+      "step": 9160
+    },
+    {
+      "epoch": 0.7693195045985087,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.877815699658703e-05,
+      "loss": 0.3299,
+      "step": 9170
+    },
+    {
+      "epoch": 0.770158457166228,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8709897610921504e-05,
+      "loss": 0.3146,
+      "step": 9180
+    },
+    {
+      "epoch": 0.7709974097339471,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.8641638225255973e-05,
+      "loss": 0.2951,
+      "step": 9190
+    },
+    {
+      "epoch": 0.7718363623016664,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.8573378839590445e-05,
+      "loss": 0.3226,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7726753148693856,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8505119453924914e-05,
+      "loss": 0.3018,
+      "step": 9210
+    },
+    {
+      "epoch": 0.7735142674371048,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.8436860068259387e-05,
+      "loss": 0.3199,
+      "step": 9220
+    },
+    {
+      "epoch": 0.774353220004824,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.836860068259386e-05,
+      "loss": 0.3067,
+      "step": 9230
+    },
+    {
+      "epoch": 0.7751921725725431,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.830034129692833e-05,
+      "loss": 0.2861,
+      "step": 9240
+    },
+    {
+      "epoch": 0.7760311251402624,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.82320819112628e-05,
+      "loss": 0.3188,
+      "step": 9250
+    },
+    {
+      "epoch": 0.7768700777079816,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8163822525597273e-05,
+      "loss": 0.3236,
+      "step": 9260
+    },
+    {
+      "epoch": 0.7777090302757008,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8095563139931742e-05,
+      "loss": 0.3075,
+      "step": 9270
+    },
+    {
+      "epoch": 0.77854798284342,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.802730375426621e-05,
+      "loss": 0.308,
+      "step": 9280
+    },
+    {
+      "epoch": 0.7793869354111392,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7959044368600684e-05,
+      "loss": 0.3105,
+      "step": 9290
+    },
+    {
+      "epoch": 0.7802258879788584,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7890784982935153e-05,
+      "loss": 0.291,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7810648405465775,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7822525597269625e-05,
+      "loss": 0.2903,
+      "step": 9310
+    },
+    {
+      "epoch": 0.7819037931142968,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.7754266211604098e-05,
+      "loss": 0.3183,
+      "step": 9320
+    },
+    {
+      "epoch": 0.782742745682016,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.768600682593857e-05,
+      "loss": 0.3169,
+      "step": 9330
+    },
+    {
+      "epoch": 0.7835816982497352,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.761774744027304e-05,
+      "loss": 0.2839,
+      "step": 9340
+    },
+    {
+      "epoch": 0.7844206508174544,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7549488054607512e-05,
+      "loss": 0.2799,
+      "step": 9350
+    },
+    {
+      "epoch": 0.7852596033851736,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.748122866894198e-05,
+      "loss": 0.3072,
+      "step": 9360
+    },
+    {
+      "epoch": 0.7860985559528928,
+      "grad_norm": 0.375,
+      "learning_rate": 1.7412969283276453e-05,
+      "loss": 0.322,
+      "step": 9370
+    },
+    {
+      "epoch": 0.7869375085206121,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7344709897610922e-05,
+      "loss": 0.3228,
+      "step": 9380
+    },
+    {
+      "epoch": 0.7877764610883312,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.727645051194539e-05,
+      "loss": 0.3282,
+      "step": 9390
+    },
+    {
+      "epoch": 0.7886154136560504,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7208191126279864e-05,
+      "loss": 0.3361,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7894543662237696,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.7139931740614336e-05,
+      "loss": 0.3411,
+      "step": 9410
+    },
+    {
+      "epoch": 0.7902933187914888,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.707167235494881e-05,
+      "loss": 0.2957,
+      "step": 9420
+    },
+    {
+      "epoch": 0.791132271359208,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.7003412969283278e-05,
+      "loss": 0.2747,
+      "step": 9430
+    },
+    {
+      "epoch": 0.7919712239269272,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.693515358361775e-05,
+      "loss": 0.301,
+      "step": 9440
+    },
+    {
+      "epoch": 0.7928101764946465,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.686689419795222e-05,
+      "loss": 0.3174,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7936491290623656,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.6798634812286692e-05,
+      "loss": 0.3069,
+      "step": 9460
+    },
+    {
+      "epoch": 0.7944880816300849,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.673037542662116e-05,
+      "loss": 0.2877,
+      "step": 9470
+    },
+    {
+      "epoch": 0.795327034197804,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.6662116040955633e-05,
+      "loss": 0.3435,
+      "step": 9480
+    },
+    {
+      "epoch": 0.7961659867655232,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.6593856655290102e-05,
+      "loss": 0.2862,
+      "step": 9490
+    },
+    {
+      "epoch": 0.7970049393332425,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.6525597269624575e-05,
+      "loss": 0.3139,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7978438919009616,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6457337883959047e-05,
+      "loss": 0.3095,
+      "step": 9510
+    },
+    {
+      "epoch": 0.7986828444686809,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.6389078498293516e-05,
+      "loss": 0.3284,
+      "step": 9520
+    },
+    {
+      "epoch": 0.7995217970364,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.632081911262799e-05,
+      "loss": 0.3056,
+      "step": 9530
+    },
+    {
+      "epoch": 0.8003607496041193,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.6252559726962458e-05,
+      "loss": 0.3237,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8011997021718384,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.618430034129693e-05,
+      "loss": 0.2981,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8020386547395577,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.61160409556314e-05,
+      "loss": 0.3133,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8028776073072769,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.6047781569965872e-05,
+      "loss": 0.3057,
+      "step": 9570
+    },
+    {
+      "epoch": 0.803716559874996,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.597952218430034e-05,
+      "loss": 0.3068,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8045555124427153,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.5911262798634813e-05,
+      "loss": 0.3063,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8053944650104344,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5843003412969286e-05,
+      "loss": 0.3034,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8062334175781537,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.5774744027303755e-05,
+      "loss": 0.3159,
+      "step": 9610
+    },
+    {
+      "epoch": 0.8070723701458729,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.5706484641638227e-05,
+      "loss": 0.3064,
+      "step": 9620
+    },
+    {
+      "epoch": 0.8079113227135921,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.5638225255972696e-05,
+      "loss": 0.307,
+      "step": 9630
+    },
+    {
+      "epoch": 0.8087502752813113,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.556996587030717e-05,
+      "loss": 0.2971,
+      "step": 9640
+    },
+    {
+      "epoch": 0.8095892278490305,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5501706484641638e-05,
+      "loss": 0.2962,
+      "step": 9650
+    },
+    {
+      "epoch": 0.8104281804167497,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.543344709897611e-05,
+      "loss": 0.312,
+      "step": 9660
+    },
+    {
+      "epoch": 0.8112671329844688,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.536518771331058e-05,
+      "loss": 0.3082,
+      "step": 9670
+    },
+    {
+      "epoch": 0.8121060855521881,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.5296928327645052e-05,
+      "loss": 0.312,
+      "step": 9680
+    },
+    {
+      "epoch": 0.8129450381199073,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5228668941979523e-05,
+      "loss": 0.3184,
+      "step": 9690
+    },
+    {
+      "epoch": 0.8137839906876265,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.5160409556313995e-05,
+      "loss": 0.2945,
+      "step": 9700
+    },
+    {
+      "epoch": 0.8146229432553457,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5092150170648466e-05,
+      "loss": 0.3275,
+      "step": 9710
+    },
+    {
+      "epoch": 0.8154618958230649,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.5023890784982937e-05,
+      "loss": 0.3295,
+      "step": 9720
+    },
+    {
+      "epoch": 0.8163008483907841,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.4955631399317407e-05,
+      "loss": 0.2932,
+      "step": 9730
+    },
+    {
+      "epoch": 0.8171398009585034,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.4887372013651877e-05,
+      "loss": 0.3053,
+      "step": 9740
+    },
+    {
+      "epoch": 0.8179787535262225,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4819112627986349e-05,
+      "loss": 0.3193,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8188177060939417,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.475085324232082e-05,
+      "loss": 0.3288,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8196566586616609,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.4682593856655292e-05,
+      "loss": 0.2928,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8204956112293801,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4614334470989761e-05,
+      "loss": 0.3261,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8213345637970993,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4546075085324234e-05,
+      "loss": 0.2812,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8221735163648185,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.4477815699658704e-05,
+      "loss": 0.2891,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8230124689325378,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.4409556313993175e-05,
+      "loss": 0.3091,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8238514215002569,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.4341296928327646e-05,
+      "loss": 0.3233,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8246903740679762,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4273037542662118e-05,
+      "loss": 0.2931,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8255293266356953,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.4204778156996588e-05,
+      "loss": 0.327,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8263682792034145,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4136518771331058e-05,
+      "loss": 0.3128,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8272072317711338,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.406825938566553e-05,
+      "loss": 0.3158,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8280461843388529,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4e-05,
+      "loss": 0.3069,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8288851369065722,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3931740614334472e-05,
+      "loss": 0.2917,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8297240894742913,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3863481228668943e-05,
+      "loss": 0.3108,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8305630420420106,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3795221843003414e-05,
+      "loss": 0.3268,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8314019946097297,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3726962457337885e-05,
+      "loss": 0.3301,
+      "step": 9910
+    },
+    {
+      "epoch": 0.832240947177449,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3658703071672357e-05,
+      "loss": 0.2925,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8330798997451682,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3590443686006826e-05,
+      "loss": 0.307,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8339188523128873,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.3522184300341298e-05,
+      "loss": 0.3352,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8347578048806066,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.345392491467577e-05,
+      "loss": 0.3126,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8355967574483257,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.3385665529010242e-05,
+      "loss": 0.3046,
+      "step": 9960
+    },
+    {
+      "epoch": 0.836435710016045,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.331740614334471e-05,
+      "loss": 0.3024,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8372746625837642,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3249146757679182e-05,
+      "loss": 0.3161,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8381136151514834,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.3180887372013654e-05,
+      "loss": 0.3139,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8389525677192026,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.3112627986348123e-05,
+      "loss": 0.3115,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8397915202869218,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3044368600682595e-05,
+      "loss": 0.3043,
+      "step": 10010
+    },
+    {
+      "epoch": 0.840630472854641,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.2976109215017065e-05,
+      "loss": 0.3056,
+      "step": 10020
+    },
+    {
+      "epoch": 0.8414694254223601,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.2907849829351537e-05,
+      "loss": 0.2873,
+      "step": 10030
+    },
+    {
+      "epoch": 0.8423083779900794,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2839590443686008e-05,
+      "loss": 0.2969,
+      "step": 10040
+    },
+    {
+      "epoch": 0.8431473305577986,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.277133105802048e-05,
+      "loss": 0.2774,
+      "step": 10050
+    },
+    {
+      "epoch": 0.8439862831255178,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.270307167235495e-05,
+      "loss": 0.276,
+      "step": 10060
+    },
+    {
+      "epoch": 0.844825235693237,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.2634812286689422e-05,
+      "loss": 0.3276,
+      "step": 10070
+    },
+    {
+      "epoch": 0.8456641882609562,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2566552901023892e-05,
+      "loss": 0.3473,
+      "step": 10080
+    },
+    {
+      "epoch": 0.8465031408286754,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.2498293515358362e-05,
+      "loss": 0.297,
+      "step": 10090
+    },
+    {
+      "epoch": 0.8473420933963945,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.2430034129692834e-05,
+      "loss": 0.2876,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8481810459641138,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.2361774744027305e-05,
+      "loss": 0.277,
+      "step": 10110
+    },
+    {
+      "epoch": 0.849019998531833,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.2293515358361776e-05,
+      "loss": 0.3034,
+      "step": 10120
+    },
+    {
+      "epoch": 0.8498589510995522,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2225255972696246e-05,
+      "loss": 0.2847,
+      "step": 10130
+    },
+    {
+      "epoch": 0.8506979036672714,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.2156996587030719e-05,
+      "loss": 0.3247,
+      "step": 10140
+    },
+    {
+      "epoch": 0.8515368562349906,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2088737201365188e-05,
+      "loss": 0.3008,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8523758088027098,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.202047781569966e-05,
+      "loss": 0.3411,
+      "step": 10160
+    },
+    {
+      "epoch": 0.8532147613704291,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.1952218430034131e-05,
+      "loss": 0.3085,
+      "step": 10170
+    },
+    {
+      "epoch": 0.8540537139381482,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.1883959044368602e-05,
+      "loss": 0.3263,
+      "step": 10180
+    },
+    {
+      "epoch": 0.8548926665058674,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.1815699658703073e-05,
+      "loss": 0.3052,
+      "step": 10190
+    },
+    {
+      "epoch": 0.8557316190735866,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.1747440273037543e-05,
+      "loss": 0.266,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8565705716413058,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.1679180887372014e-05,
+      "loss": 0.3017,
+      "step": 10210
+    },
+    {
+      "epoch": 0.857409524209025,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1610921501706485e-05,
+      "loss": 0.3197,
+      "step": 10220
+    },
+    {
+      "epoch": 0.8582484767767442,
+      "grad_norm": 0.375,
+      "learning_rate": 1.1542662116040957e-05,
+      "loss": 0.3272,
+      "step": 10230
+    },
+    {
+      "epoch": 0.8590874293444635,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.1474402730375426e-05,
+      "loss": 0.2973,
+      "step": 10240
+    },
+    {
+      "epoch": 0.8599263819121826,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.1406143344709899e-05,
+      "loss": 0.3108,
+      "step": 10250
+    },
+    {
+      "epoch": 0.8607653344799019,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.133788395904437e-05,
+      "loss": 0.2892,
+      "step": 10260
+    },
+    {
+      "epoch": 0.861604287047621,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.1269624573378842e-05,
+      "loss": 0.31,
+      "step": 10270
+    },
+    {
+      "epoch": 0.8624432396153402,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.1201365187713311e-05,
+      "loss": 0.3138,
+      "step": 10280
+    },
+    {
+      "epoch": 0.8632821921830595,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.1133105802047784e-05,
+      "loss": 0.302,
+      "step": 10290
+    },
+    {
+      "epoch": 0.8641211447507786,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.1064846416382253e-05,
+      "loss": 0.303,
+      "step": 10300
+    },
+    {
+      "epoch": 0.8649600973184979,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.0996587030716725e-05,
+      "loss": 0.3183,
+      "step": 10310
+    },
+    {
+      "epoch": 0.865799049886217,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.0928327645051196e-05,
+      "loss": 0.3013,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8666380024539363,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.0860068259385665e-05,
+      "loss": 0.2986,
+      "step": 10330
+    },
+    {
+      "epoch": 0.8674769550216554,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.0791808873720137e-05,
+      "loss": 0.3251,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8683159075893747,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.0723549488054608e-05,
+      "loss": 0.3291,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8691548601570939,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.065529010238908e-05,
+      "loss": 0.2905,
+      "step": 10360
+    },
+    {
+      "epoch": 0.869993812724813,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.058703071672355e-05,
+      "loss": 0.3071,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8708327652925323,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.0518771331058022e-05,
+      "loss": 0.3146,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8716717178602514,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.0450511945392491e-05,
+      "loss": 0.2967,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8725106704279707,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.0382252559726964e-05,
+      "loss": 0.2679,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8733496229956899,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.0313993174061434e-05,
+      "loss": 0.3069,
+      "step": 10410
+    },
+    {
+      "epoch": 0.8741885755634091,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.0245733788395907e-05,
+      "loss": 0.2903,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8750275281311283,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.0177474402730376e-05,
+      "loss": 0.2657,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8758664806988475,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.0109215017064847e-05,
+      "loss": 0.3163,
+      "step": 10440
+    },
+    {
+      "epoch": 0.8767054332665667,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.0040955631399319e-05,
+      "loss": 0.3044,
+      "step": 10450
+    },
+    {
+      "epoch": 0.8775443858342858,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.97269624573379e-06,
+      "loss": 0.3208,
+      "step": 10460
+    },
+    {
+      "epoch": 0.8783833384020051,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.90443686006826e-06,
+      "loss": 0.2919,
+      "step": 10470
+    },
+    {
+      "epoch": 0.8792222909697243,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.836177474402731e-06,
+      "loss": 0.309,
+      "step": 10480
+    },
+    {
+      "epoch": 0.8800612435374435,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.767918088737202e-06,
+      "loss": 0.3021,
+      "step": 10490
+    },
+    {
+      "epoch": 0.8809001961051627,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.699658703071673e-06,
+      "loss": 0.3156,
+      "step": 10500
+    },
+    {
+      "epoch": 0.881739148672882,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.631399317406144e-06,
+      "loss": 0.2878,
+      "step": 10510
+    },
+    {
+      "epoch": 0.8825781012406011,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.563139931740614e-06,
+      "loss": 0.3264,
+      "step": 10520
+    },
+    {
+      "epoch": 0.8834170538083204,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.494880546075085e-06,
+      "loss": 0.3046,
+      "step": 10530
+    },
+    {
+      "epoch": 0.8842560063760395,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.426621160409558e-06,
+      "loss": 0.3141,
+      "step": 10540
+    },
+    {
+      "epoch": 0.8850949589437587,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.358361774744028e-06,
+      "loss": 0.3095,
+      "step": 10550
+    },
+    {
+      "epoch": 0.8859339115114779,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.290102389078499e-06,
+      "loss": 0.3005,
+      "step": 10560
+    },
+    {
+      "epoch": 0.8867728640791971,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.22184300341297e-06,
+      "loss": 0.2936,
+      "step": 10570
+    },
+    {
+      "epoch": 0.8876118166469164,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.15358361774744e-06,
+      "loss": 0.3146,
+      "step": 10580
+    },
+    {
+      "epoch": 0.8884507692146355,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.085324232081913e-06,
+      "loss": 0.2872,
+      "step": 10590
+    },
+    {
+      "epoch": 0.8892897217823548,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.017064846416384e-06,
+      "loss": 0.297,
+      "step": 10600
+    },
+    {
+      "epoch": 0.8901286743500739,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.948805460750853e-06,
+      "loss": 0.2811,
+      "step": 10610
+    },
+    {
+      "epoch": 0.8909676269177932,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.880546075085325e-06,
+      "loss": 0.3066,
+      "step": 10620
+    },
+    {
+      "epoch": 0.8918065794855123,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.812286689419796e-06,
+      "loss": 0.2821,
+      "step": 10630
+    },
+    {
+      "epoch": 0.8926455320532315,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.744027303754267e-06,
+      "loss": 0.311,
+      "step": 10640
+    },
+    {
+      "epoch": 0.8934844846209508,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.675767918088738e-06,
+      "loss": 0.2785,
+      "step": 10650
+    },
+    {
+      "epoch": 0.8943234371886699,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.607508532423208e-06,
+      "loss": 0.3074,
+      "step": 10660
+    },
+    {
+      "epoch": 0.8951623897563892,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.539249146757679e-06,
+      "loss": 0.2939,
+      "step": 10670
+    },
+    {
+      "epoch": 0.8960013423241083,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.470989761092152e-06,
+      "loss": 0.3236,
+      "step": 10680
+    },
+    {
+      "epoch": 0.8968402948918276,
+      "grad_norm": 0.3125,
+      "learning_rate": 8.402730375426622e-06,
+      "loss": 0.2945,
+      "step": 10690
+    },
+    {
+      "epoch": 0.8976792474595467,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.334470989761093e-06,
+      "loss": 0.3205,
+      "step": 10700
+    },
+    {
+      "epoch": 0.898518200027266,
+      "grad_norm": 0.380859375,
+      "learning_rate": 8.266211604095564e-06,
+      "loss": 0.3004,
+      "step": 10710
+    },
+    {
+      "epoch": 0.8993571525949852,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.197952218430035e-06,
+      "loss": 0.2815,
+      "step": 10720
+    },
+    {
+      "epoch": 0.9001961051627043,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.129692832764505e-06,
+      "loss": 0.3134,
+      "step": 10730
+    },
+    {
+      "epoch": 0.9010350577304236,
+      "grad_norm": 0.37890625,
+      "learning_rate": 8.061433447098976e-06,
+      "loss": 0.299,
+      "step": 10740
+    },
+    {
+      "epoch": 0.9018740102981427,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.993174061433447e-06,
+      "loss": 0.3134,
+      "step": 10750
+    },
+    {
+      "epoch": 0.902712962865862,
+      "grad_norm": 0.30078125,
+      "learning_rate": 7.92491467576792e-06,
+      "loss": 0.3037,
+      "step": 10760
+    },
+    {
+      "epoch": 0.9035519154335812,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.85665529010239e-06,
+      "loss": 0.3105,
+      "step": 10770
+    },
+    {
+      "epoch": 0.9043908680013004,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.788395904436861e-06,
+      "loss": 0.2843,
+      "step": 10780
+    },
+    {
+      "epoch": 0.9052298205690196,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.720136518771332e-06,
+      "loss": 0.309,
+      "step": 10790
+    },
+    {
+      "epoch": 0.9060687731367388,
+      "grad_norm": 0.376953125,
+      "learning_rate": 7.651877133105802e-06,
+      "loss": 0.3122,
+      "step": 10800
+    },
+    {
+      "epoch": 0.906907725704458,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.583617747440274e-06,
+      "loss": 0.2958,
+      "step": 10810
+    },
+    {
+      "epoch": 0.9077466782721771,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.515358361774745e-06,
+      "loss": 0.2904,
+      "step": 10820
+    },
+    {
+      "epoch": 0.9085856308398964,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.447098976109216e-06,
+      "loss": 0.314,
+      "step": 10830
+    },
+    {
+      "epoch": 0.9094245834076156,
+      "grad_norm": 0.380859375,
+      "learning_rate": 7.378839590443687e-06,
+      "loss": 0.2947,
+      "step": 10840
+    },
+    {
+      "epoch": 0.9102635359753348,
+      "grad_norm": 0.380859375,
+      "learning_rate": 7.310580204778157e-06,
+      "loss": 0.2921,
+      "step": 10850
+    },
+    {
+      "epoch": 0.911102488543054,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.242320819112628e-06,
+      "loss": 0.3032,
+      "step": 10860
+    },
+    {
+      "epoch": 0.9119414411107732,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.174061433447099e-06,
+      "loss": 0.2936,
+      "step": 10870
+    },
+    {
+      "epoch": 0.9127803936784924,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.10580204778157e-06,
+      "loss": 0.3046,
+      "step": 10880
+    },
+    {
+      "epoch": 0.9136193462462115,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.037542662116042e-06,
+      "loss": 0.3087,
+      "step": 10890
+    },
+    {
+      "epoch": 0.9144582988139308,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.9692832764505125e-06,
+      "loss": 0.2993,
+      "step": 10900
+    },
+    {
+      "epoch": 0.91529725138165,
+      "grad_norm": 0.41015625,
+      "learning_rate": 6.901023890784983e-06,
+      "loss": 0.3018,
+      "step": 10910
+    },
+    {
+      "epoch": 0.9161362039493692,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.832764505119455e-06,
+      "loss": 0.2988,
+      "step": 10920
+    },
+    {
+      "epoch": 0.9169751565170884,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.764505119453926e-06,
+      "loss": 0.2871,
+      "step": 10930
+    },
+    {
+      "epoch": 0.9178141090848076,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.696245733788397e-06,
+      "loss": 0.3117,
+      "step": 10940
+    },
+    {
+      "epoch": 0.9186530616525268,
+      "grad_norm": 0.376953125,
+      "learning_rate": 6.627986348122868e-06,
+      "loss": 0.2891,
+      "step": 10950
+    },
+    {
+      "epoch": 0.9194920142202461,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.559726962457338e-06,
+      "loss": 0.305,
+      "step": 10960
+    },
+    {
+      "epoch": 0.9203309667879652,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.491467576791809e-06,
+      "loss": 0.323,
+      "step": 10970
+    },
+    {
+      "epoch": 0.9211699193556844,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.42320819112628e-06,
+      "loss": 0.2954,
+      "step": 10980
+    },
+    {
+      "epoch": 0.9220088719234036,
+      "grad_norm": 0.3203125,
+      "learning_rate": 6.354948805460751e-06,
+      "loss": 0.2797,
+      "step": 10990
+    },
+    {
+      "epoch": 0.9228478244911228,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.286689419795222e-06,
+      "loss": 0.2958,
+      "step": 11000
+    },
+    {
+      "epoch": 0.9236867770588421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.218430034129693e-06,
+      "loss": 0.3124,
+      "step": 11010
+    },
+    {
+      "epoch": 0.9245257296265612,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.150170648464164e-06,
+      "loss": 0.2913,
+      "step": 11020
+    },
+    {
+      "epoch": 0.9253646821942805,
+      "grad_norm": 0.314453125,
+      "learning_rate": 6.081911262798636e-06,
+      "loss": 0.2895,
+      "step": 11030
+    },
+    {
+      "epoch": 0.9262036347619996,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.0136518771331065e-06,
+      "loss": 0.3067,
+      "step": 11040
+    },
+    {
+      "epoch": 0.9270425873297189,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.945392491467577e-06,
+      "loss": 0.3171,
+      "step": 11050
+    },
+    {
+      "epoch": 0.927881539897438,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.877133105802049e-06,
+      "loss": 0.3025,
+      "step": 11060
+    },
+    {
+      "epoch": 0.9287204924651572,
+      "grad_norm": 0.375,
+      "learning_rate": 5.80887372013652e-06,
+      "loss": 0.2783,
+      "step": 11070
+    },
+    {
+      "epoch": 0.9295594450328765,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.7406143344709895e-06,
+      "loss": 0.3012,
+      "step": 11080
+    },
+    {
+      "epoch": 0.9303983976005956,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.672354948805461e-06,
+      "loss": 0.3142,
+      "step": 11090
+    },
+    {
+      "epoch": 0.9312373501683149,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.604095563139932e-06,
+      "loss": 0.3134,
+      "step": 11100
+    },
+    {
+      "epoch": 0.932076302736034,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.535836177474403e-06,
+      "loss": 0.2859,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9329152553037533,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.467576791808874e-06,
+      "loss": 0.3346,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9337542078714725,
+      "grad_norm": 0.375,
+      "learning_rate": 5.399317406143345e-06,
+      "loss": 0.3141,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9345931604391917,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.331058020477816e-06,
+      "loss": 0.3127,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9354321130069109,
+      "grad_norm": 0.369140625,
+      "learning_rate": 5.262798634812287e-06,
+      "loss": 0.2821,
+      "step": 11150
+    },
+    {
+      "epoch": 0.93627106557463,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.194539249146758e-06,
+      "loss": 0.2895,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9371100181423493,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.12627986348123e-06,
+      "loss": 0.3041,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9379489707100684,
+      "grad_norm": 0.37109375,
+      "learning_rate": 5.0580204778157005e-06,
+      "loss": 0.3035,
+      "step": 11180
+    },
+    {
+      "epoch": 0.9387879232777877,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.989761092150171e-06,
+      "loss": 0.2749,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9396268758455069,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.921501706484642e-06,
+      "loss": 0.3236,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9404658284132261,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.853242320819113e-06,
+      "loss": 0.3138,
+      "step": 11210
+    },
+    {
+      "epoch": 0.9413047809809453,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.784982935153584e-06,
+      "loss": 0.2989,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9421437335486645,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.716723549488055e-06,
+      "loss": 0.3034,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9429826861163837,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.648464163822526e-06,
+      "loss": 0.3193,
+      "step": 11240
+    },
+    {
+      "epoch": 0.9438216386841028,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.580204778156997e-06,
+      "loss": 0.2978,
+      "step": 11250
+    },
+    {
+      "epoch": 0.9446605912518221,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.511945392491468e-06,
+      "loss": 0.3112,
+      "step": 11260
+    },
+    {
+      "epoch": 0.9454995438195413,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.443686006825939e-06,
+      "loss": 0.2959,
+      "step": 11270
+    },
+    {
+      "epoch": 0.9463384963872605,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.37542662116041e-06,
+      "loss": 0.2823,
+      "step": 11280
+    },
+    {
+      "epoch": 0.9471774489549797,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.3071672354948806e-06,
+      "loss": 0.326,
+      "step": 11290
+    },
+    {
+      "epoch": 0.948016401522699,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.238907849829352e-06,
+      "loss": 0.3031,
+      "step": 11300
+    },
+    {
+      "epoch": 0.9488553540904181,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.170648464163823e-06,
+      "loss": 0.3171,
+      "step": 11310
+    },
+    {
+      "epoch": 0.9496943066581374,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.102389078498294e-06,
+      "loss": 0.3,
+      "step": 11320
+    },
+    {
+      "epoch": 0.9505332592258565,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.034129692832765e-06,
+      "loss": 0.3284,
+      "step": 11330
+    },
+    {
+      "epoch": 0.9513722117935757,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.965870307167236e-06,
+      "loss": 0.291,
+      "step": 11340
+    },
+    {
+      "epoch": 0.9522111643612949,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.897610921501707e-06,
+      "loss": 0.3076,
+      "step": 11350
+    },
+    {
+      "epoch": 0.9530501169290141,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.8293515358361776e-06,
+      "loss": 0.3199,
+      "step": 11360
+    },
+    {
+      "epoch": 0.9538890694967334,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.7610921501706487e-06,
+      "loss": 0.2884,
+      "step": 11370
+    },
+    {
+      "epoch": 0.9547280220644525,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.69283276450512e-06,
+      "loss": 0.2989,
+      "step": 11380
+    },
+    {
+      "epoch": 0.9555669746321718,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.624573378839591e-06,
+      "loss": 0.3027,
+      "step": 11390
+    },
+    {
+      "epoch": 0.9564059271998909,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.5563139931740614e-06,
+      "loss": 0.291,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9572448797676102,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.4880546075085326e-06,
+      "loss": 0.3165,
+      "step": 11410
+    },
+    {
+      "epoch": 0.9580838323353293,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.419795221843004e-06,
+      "loss": 0.3191,
+      "step": 11420
+    },
+    {
+      "epoch": 0.9589227849030485,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.3515358361774746e-06,
+      "loss": 0.3036,
+      "step": 11430
+    },
+    {
+      "epoch": 0.9597617374707678,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.2832764505119457e-06,
+      "loss": 0.3068,
+      "step": 11440
+    },
+    {
+      "epoch": 0.9606006900384869,
+      "grad_norm": 0.375,
+      "learning_rate": 3.215017064846417e-06,
+      "loss": 0.294,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9614396426062062,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.1467576791808873e-06,
+      "loss": 0.2933,
+      "step": 11460
+    },
+    {
+      "epoch": 0.9622785951739253,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.0784982935153584e-06,
+      "loss": 0.3109,
+      "step": 11470
+    },
+    {
+      "epoch": 0.9631175477416446,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.0102389078498296e-06,
+      "loss": 0.3117,
+      "step": 11480
+    },
+    {
+      "epoch": 0.9639565003093638,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.941979522184301e-06,
+      "loss": 0.2911,
+      "step": 11490
+    },
+    {
+      "epoch": 0.964795452877083,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.8737201365187716e-06,
+      "loss": 0.283,
+      "step": 11500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 11920,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5775843924708995e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}