diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5705 @@
+{
+  "best_metric": 2.4689557552337646,
+  "best_model_checkpoint": "./output/training_results/C017_random_sample_llama3-8b-base_pretrain_20240504_182259/checkpoint-800",
+  "epoch": 4.0,
+  "eval_steps": 200,
+  "global_step": 3944,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0010141987829614604,
+      "grad_norm": 4.267137538119642,
+      "learning_rate": 7.5e-07,
+      "loss": 2.7134,
+      "step": 1
+    },
+    {
+      "epoch": 0.005070993914807302,
+      "grad_norm": 4.879489677016923,
+      "learning_rate": 2.25e-06,
+      "loss": 2.7254,
+      "step": 5
+    },
+    {
+      "epoch": 0.010141987829614604,
+      "grad_norm": 2.7621009561709564,
+      "learning_rate": 6e-06,
+      "loss": 2.707,
+      "step": 10
+    },
+    {
+      "epoch": 0.015212981744421906,
+      "grad_norm": 2.404100845677231,
+      "learning_rate": 9e-06,
+      "loss": 2.6421,
+      "step": 15
+    },
+    {
+      "epoch": 0.02028397565922921,
+      "grad_norm": 2.4429846538599254,
+      "learning_rate": 1.275e-05,
+      "loss": 2.6682,
+      "step": 20
+    },
+    {
+      "epoch": 0.02535496957403651,
+      "grad_norm": 2.8575493026010625,
+      "learning_rate": 1.4916395742870319e-05,
+      "loss": 2.6639,
+      "step": 25
+    },
+    {
+      "epoch": 0.030425963488843813,
+      "grad_norm": 2.4347171369214538,
+      "learning_rate": 1.4709241308404976e-05,
+      "loss": 2.6624,
+      "step": 30
+    },
+    {
+      "epoch": 0.035496957403651115,
+      "grad_norm": 2.5792627004512942,
+      "learning_rate": 1.4504714365262738e-05,
+      "loss": 2.6351,
+      "step": 35
+    },
+    {
+      "epoch": 0.04056795131845842,
+      "grad_norm": 2.1789139866654366,
+      "learning_rate": 1.4302784881547452e-05,
+      "loss": 2.6055,
+      "step": 40
+    },
+    {
+      "epoch": 0.04563894523326572,
+      "grad_norm": 2.232485210798856,
+      "learning_rate": 1.4103423130872168e-05,
+      "loss": 2.5938,
+      "step": 45
+    },
+    {
+      "epoch": 0.05070993914807302,
+      "grad_norm": 2.2896589926745814,
+      "learning_rate": 1.390659968963626e-05,
+      "loss": 2.6334,
+      "step": 50
+    },
+    {
+      "epoch": 0.055780933062880324,
+      "grad_norm": 2.7780457428021985,
+      "learning_rate": 1.3712285434323396e-05,
+      "loss": 2.646,
+      "step": 55
+    },
+    {
+      "epoch": 0.060851926977687626,
+      "grad_norm": 1.9399001575023072,
+      "learning_rate": 1.352045153882017e-05,
+      "loss": 2.6182,
+      "step": 60
+    },
+    {
+      "epoch": 0.06592292089249494,
+      "grad_norm": 1.9083156579424998,
+      "learning_rate": 1.3331069471755332e-05,
+      "loss": 2.6056,
+      "step": 65
+    },
+    {
+      "epoch": 0.07099391480730223,
+      "grad_norm": 2.2298396560554683,
+      "learning_rate": 1.314411099385942e-05,
+      "loss": 2.6043,
+      "step": 70
+    },
+    {
+      "epoch": 0.07606490872210954,
+      "grad_norm": 1.9661711744318215,
+      "learning_rate": 1.2959548155344706e-05,
+      "loss": 2.6321,
+      "step": 75
+    },
+    {
+      "epoch": 0.08113590263691683,
+      "grad_norm": 2.1260634398939438,
+      "learning_rate": 1.2777353293305311e-05,
+      "loss": 2.5744,
+      "step": 80
+    },
+    {
+      "epoch": 0.08620689655172414,
+      "grad_norm": 2.171189842092272,
+      "learning_rate": 1.2597499029137354e-05,
+      "loss": 2.6102,
+      "step": 85
+    },
+    {
+      "epoch": 0.09127789046653144,
+      "grad_norm": 2.118995328928547,
+      "learning_rate": 1.2419958265979023e-05,
+      "loss": 2.6056,
+      "step": 90
+    },
+    {
+      "epoch": 0.09634888438133875,
+      "grad_norm": 2.1743656445294466,
+      "learning_rate": 1.2244704186170414e-05,
+      "loss": 2.591,
+      "step": 95
+    },
+    {
+      "epoch": 0.10141987829614604,
+      "grad_norm": 2.100620832387391,
+      "learning_rate": 1.2106129489565247e-05,
+      "loss": 2.6461,
+      "step": 100
+    },
+    {
+      "epoch": 0.10649087221095335,
+      "grad_norm": 2.02911049207023,
+      "learning_rate": 1.1934924740853141e-05,
+      "loss": 2.5878,
+      "step": 105
+    },
+    {
+      "epoch": 0.11156186612576065,
+      "grad_norm": 2.12870974325018,
+      "learning_rate": 1.1765933050017452e-05,
+      "loss": 2.5793,
+      "step": 110
+    },
+    {
+      "epoch": 0.11663286004056796,
+      "grad_norm": 1.9038783159180614,
+      "learning_rate": 1.1599128637544344e-05,
+      "loss": 2.5612,
+      "step": 115
+    },
+    {
+      "epoch": 0.12170385395537525,
+      "grad_norm": 1.9647399779959451,
+      "learning_rate": 1.1434485991200533e-05,
+      "loss": 2.6083,
+      "step": 120
+    },
+    {
+      "epoch": 0.12677484787018256,
+      "grad_norm": 1.88937427094592,
+      "learning_rate": 1.1271979863605386e-05,
+      "loss": 2.5561,
+      "step": 125
+    },
+    {
+      "epoch": 0.13184584178498987,
+      "grad_norm": 1.8208051471693176,
+      "learning_rate": 1.111158526982193e-05,
+      "loss": 2.5884,
+      "step": 130
+    },
+    {
+      "epoch": 0.13691683569979715,
+      "grad_norm": 1.771422341312915,
+      "learning_rate": 1.0953277484966689e-05,
+      "loss": 2.5509,
+      "step": 135
+    },
+    {
+      "epoch": 0.14198782961460446,
+      "grad_norm": 1.8296701053391813,
+      "learning_rate": 1.0797032041838185e-05,
+      "loss": 2.5784,
+      "step": 140
+    },
+    {
+      "epoch": 0.14705882352941177,
+      "grad_norm": 1.8139046565289612,
+      "learning_rate": 1.0642824728564022e-05,
+      "loss": 2.5624,
+      "step": 145
+    },
+    {
+      "epoch": 0.15212981744421908,
+      "grad_norm": 1.9862915107502803,
+      "learning_rate": 1.0490631586266381e-05,
+      "loss": 2.6007,
+      "step": 150
+    },
+    {
+      "epoch": 0.15720081135902636,
+      "grad_norm": 1.8392246134083736,
+      "learning_rate": 1.0340428906745863e-05,
+      "loss": 2.5775,
+      "step": 155
+    },
+    {
+      "epoch": 0.16227180527383367,
+      "grad_norm": 1.9250085841598776,
+      "learning_rate": 1.0192193230183505e-05,
+      "loss": 2.6045,
+      "step": 160
+    },
+    {
+      "epoch": 0.16734279918864098,
+      "grad_norm": 2.1119936162911825,
+      "learning_rate": 1.0045901342860905e-05,
+      "loss": 2.5838,
+      "step": 165
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 1.9416866546338962,
+      "learning_rate": 9.901530274898272e-06,
+      "loss": 2.5643,
+      "step": 170
+    },
+    {
+      "epoch": 0.17748478701825557,
+      "grad_norm": 1.871570899679003,
+      "learning_rate": 9.75905729801036e-06,
+      "loss": 2.5549,
+      "step": 175
+    },
+    {
+      "epoch": 0.18255578093306288,
+      "grad_norm": 2.0672616615182897,
+      "learning_rate": 9.61845992328009e-06,
+      "loss": 2.561,
+      "step": 180
+    },
+    {
+      "epoch": 0.1876267748478702,
+      "grad_norm": 1.8373271363293353,
+      "learning_rate": 9.479715898949807e-06,
+      "loss": 2.5728,
+      "step": 185
+    },
+    {
+      "epoch": 0.1926977687626775,
+      "grad_norm": 1.9497106449021773,
+      "learning_rate": 9.342803208230014e-06,
+      "loss": 2.5535,
+      "step": 190
+    },
+    {
+      "epoch": 0.19776876267748478,
+      "grad_norm": 1.913646357656738,
+      "learning_rate": 9.207700067125492e-06,
+      "loss": 2.5411,
+      "step": 195
+    },
+    {
+      "epoch": 0.2028397565922921,
+      "grad_norm": 1.7027113982701332,
+      "learning_rate": 9.074384922278684e-06,
+      "loss": 2.5442,
+      "step": 200
+    },
+    {
+      "epoch": 0.2028397565922921,
+      "eval_loss": 2.55521821975708,
+      "eval_runtime": 81.0607,
+      "eval_samples_per_second": 86.429,
+      "eval_steps_per_second": 0.679,
+      "step": 200
+    },
+    {
+      "epoch": 0.2079107505070994,
+      "grad_norm": 1.753576639879344,
+      "learning_rate": 8.942836448830213e-06,
+      "loss": 2.5264,
+      "step": 205
+    },
+    {
+      "epoch": 0.2129817444219067,
+      "grad_norm": 1.7785092188900598,
+      "learning_rate": 8.813033548296443e-06,
+      "loss": 2.5645,
+      "step": 210
+    },
+    {
+      "epoch": 0.21805273833671399,
+      "grad_norm": 1.7915296631060966,
+      "learning_rate": 8.684955346463971e-06,
+      "loss": 2.555,
+      "step": 215
+    },
+    {
+      "epoch": 0.2231237322515213,
+      "grad_norm": 1.7452346531223148,
+      "learning_rate": 8.558581191300906e-06,
+      "loss": 2.6118,
+      "step": 220
+    },
+    {
+      "epoch": 0.2281947261663286,
+      "grad_norm": 2.339774136223256,
+      "learning_rate": 8.433890650884857e-06,
+      "loss": 2.5284,
+      "step": 225
+    },
+    {
+      "epoch": 0.2332657200811359,
+      "grad_norm": 1.7961229516339332,
+      "learning_rate": 8.310863511347508e-06,
+      "loss": 2.558,
+      "step": 230
+    },
+    {
+      "epoch": 0.2383367139959432,
+      "grad_norm": 2.000305491613022,
+      "learning_rate": 8.189479774835651e-06,
+      "loss": 2.5312,
+      "step": 235
+    },
+    {
+      "epoch": 0.2434077079107505,
+      "grad_norm": 1.9162907270104979,
+      "learning_rate": 8.069719657488614e-06,
+      "loss": 2.4983,
+      "step": 240
+    },
+    {
+      "epoch": 0.2484787018255578,
+      "grad_norm": 1.9447544732938296,
+      "learning_rate": 7.951563587431902e-06,
+      "loss": 2.5462,
+      "step": 245
+    },
+    {
+      "epoch": 0.2535496957403651,
+      "grad_norm": 1.8244106804572084,
+      "learning_rate": 7.834992202787018e-06,
+      "loss": 2.5354,
+      "step": 250
+    },
+    {
+      "epoch": 0.25862068965517243,
+      "grad_norm": 1.714609238517639,
+      "learning_rate": 7.719986349697309e-06,
+      "loss": 2.5386,
+      "step": 255
+    },
+    {
+      "epoch": 0.26369168356997974,
+      "grad_norm": 1.795436681758725,
+      "learning_rate": 7.606527080369728e-06,
+      "loss": 2.5388,
+      "step": 260
+    },
+    {
+      "epoch": 0.268762677484787,
+      "grad_norm": 1.7081706265027667,
+      "learning_rate": 7.494595651132443e-06,
+      "loss": 2.568,
+      "step": 265
+    },
+    {
+      "epoch": 0.2738336713995943,
+      "grad_norm": 1.6958291617768828,
+      "learning_rate": 7.384173520508138e-06,
+      "loss": 2.5489,
+      "step": 270
+    },
+    {
+      "epoch": 0.2789046653144016,
+      "grad_norm": 1.6677502189962874,
+      "learning_rate": 7.275242347302937e-06,
+      "loss": 2.5666,
+      "step": 275
+    },
+    {
+      "epoch": 0.2839756592292089,
+      "grad_norm": 1.6916519769077745,
+      "learning_rate": 7.167783988710829e-06,
+      "loss": 2.5161,
+      "step": 280
+    },
+    {
+      "epoch": 0.28904665314401623,
+      "grad_norm": 1.9276199368209956,
+      "learning_rate": 7.061780498433485e-06,
+      "loss": 2.5461,
+      "step": 285
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 1.721858200785338,
+      "learning_rate": 6.957214124815376e-06,
+      "loss": 2.56,
+      "step": 290
+    },
+    {
+      "epoch": 0.29918864097363085,
+      "grad_norm": 1.7023218265873687,
+      "learning_rate": 6.854067308994081e-06,
+      "loss": 2.5252,
+      "step": 295
+    },
+    {
+      "epoch": 0.30425963488843816,
+      "grad_norm": 1.7702063060142263,
+      "learning_rate": 6.752322683065677e-06,
+      "loss": 2.5365,
+      "step": 300
+    },
+    {
+      "epoch": 0.3093306288032454,
+      "grad_norm": 1.807175965887596,
+      "learning_rate": 6.651963068265119e-06,
+      "loss": 2.5351,
+      "step": 305
+    },
+    {
+      "epoch": 0.3144016227180527,
+      "grad_norm": 1.7687398862728192,
+      "learning_rate": 6.5529714731614995e-06,
+      "loss": 2.5184,
+      "step": 310
+    },
+    {
+      "epoch": 0.31947261663286003,
+      "grad_norm": 1.808664958617461,
+      "learning_rate": 6.455331091868087e-06,
+      "loss": 2.5062,
+      "step": 315
+    },
+    {
+      "epoch": 0.32454361054766734,
+      "grad_norm": 1.9021979000655393,
+      "learning_rate": 6.359025302267049e-06,
+      "loss": 2.5225,
+      "step": 320
+    },
+    {
+      "epoch": 0.32961460446247465,
+      "grad_norm": 1.704473391712384,
+      "learning_rate": 6.264037664248752e-06,
+      "loss": 2.5233,
+      "step": 325
+    },
+    {
+      "epoch": 0.33468559837728196,
+      "grad_norm": 1.751379362669565,
+      "learning_rate": 6.17035191796554e-06,
+      "loss": 2.4854,
+      "step": 330
+    },
+    {
+      "epoch": 0.33975659229208927,
+      "grad_norm": 1.6980009285341724,
+      "learning_rate": 6.077951982099886e-06,
+      "loss": 2.5008,
+      "step": 335
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 1.6987141788770321,
+      "learning_rate": 5.986821952146847e-06,
+      "loss": 2.5438,
+      "step": 340
+    },
+    {
+      "epoch": 0.34989858012170383,
+      "grad_norm": 1.6781775461943316,
+      "learning_rate": 5.89694609871067e-06,
+      "loss": 2.5417,
+      "step": 345
+    },
+    {
+      "epoch": 0.35496957403651114,
+      "grad_norm": 1.7326892052245193,
+      "learning_rate": 5.808308865815513e-06,
+      "loss": 2.5185,
+      "step": 350
+    },
+    {
+      "epoch": 0.36004056795131845,
+      "grad_norm": 1.743645811121294,
+      "learning_rate": 5.720894869230136e-06,
+      "loss": 2.5094,
+      "step": 355
+    },
+    {
+      "epoch": 0.36511156186612576,
+      "grad_norm": 1.7256678519147217,
+      "learning_rate": 5.634688894806482e-06,
+      "loss": 2.5316,
+      "step": 360
+    },
+    {
+      "epoch": 0.37018255578093306,
+      "grad_norm": 1.6209115792712339,
+      "learning_rate": 5.549675896832072e-06,
+      "loss": 2.5164,
+      "step": 365
+    },
+    {
+      "epoch": 0.3752535496957404,
+      "grad_norm": 1.6497735310259896,
+      "learning_rate": 5.465840996396076e-06,
+      "loss": 2.5363,
+      "step": 370
+    },
+    {
+      "epoch": 0.3803245436105477,
+      "grad_norm": 1.665747208014539,
+      "learning_rate": 5.383169479769005e-06,
+      "loss": 2.5015,
+      "step": 375
+    },
+    {
+      "epoch": 0.385395537525355,
+      "grad_norm": 1.8360023746562857,
+      "learning_rate": 5.301646796795905e-06,
+      "loss": 2.4465,
+      "step": 380
+    },
+    {
+      "epoch": 0.39046653144016225,
+      "grad_norm": 1.721788501212322,
+      "learning_rate": 5.221258559302969e-06,
+      "loss": 2.5104,
+      "step": 385
+    },
+    {
+      "epoch": 0.39553752535496955,
+      "grad_norm": 1.7896539066797603,
+      "learning_rate": 5.141990539517474e-06,
+      "loss": 2.5406,
+      "step": 390
+    },
+    {
+      "epoch": 0.40060851926977686,
+      "grad_norm": 1.7026594592165973,
+      "learning_rate": 5.0638286685009445e-06,
+      "loss": 2.5403,
+      "step": 395
+    },
+    {
+      "epoch": 0.4056795131845842,
+      "grad_norm": 1.7666645373608338,
+      "learning_rate": 4.986759034595453e-06,
+      "loss": 2.5376,
+      "step": 400
+    },
+    {
+      "epoch": 0.4056795131845842,
+      "eval_loss": 2.509550094604492,
+      "eval_runtime": 81.0126,
+      "eval_samples_per_second": 86.48,
+      "eval_steps_per_second": 0.679,
+      "step": 400
+    },
+    {
+      "epoch": 0.4107505070993915,
+      "grad_norm": 1.702454460655481,
+      "learning_rate": 4.910767881882966e-06,
+      "loss": 2.5017,
+      "step": 405
+    },
+    {
+      "epoch": 0.4158215010141988,
+      "grad_norm": 1.6625424708509573,
+      "learning_rate": 4.83584160865765e-06,
+      "loss": 2.5271,
+      "step": 410
+    },
+    {
+      "epoch": 0.4208924949290061,
+      "grad_norm": 1.6622717975288752,
+      "learning_rate": 4.761966765911026e-06,
+      "loss": 2.5238,
+      "step": 415
+    },
+    {
+      "epoch": 0.4259634888438134,
+      "grad_norm": 1.6256800857720881,
+      "learning_rate": 4.689130055829907e-06,
+      "loss": 2.5191,
+      "step": 420
+    },
+    {
+      "epoch": 0.43103448275862066,
+      "grad_norm": 1.7950911413498376,
+      "learning_rate": 4.617318330307044e-06,
+      "loss": 2.4909,
+      "step": 425
+    },
+    {
+      "epoch": 0.43610547667342797,
+      "grad_norm": 1.5866160053351177,
+      "learning_rate": 4.5465185894642715e-06,
+      "loss": 2.5128,
+      "step": 430
+    },
+    {
+      "epoch": 0.4411764705882353,
+      "grad_norm": 1.6754882575554404,
+      "learning_rate": 4.476717980188313e-06,
+      "loss": 2.5028,
+      "step": 435
+    },
+    {
+      "epoch": 0.4462474645030426,
+      "grad_norm": 1.6606915353792953,
+      "learning_rate": 4.407903794678819e-06,
+      "loss": 2.5207,
+      "step": 440
+    },
+    {
+      "epoch": 0.4513184584178499,
+      "grad_norm": 1.8160247477825882,
+      "learning_rate": 4.340063469008923e-06,
+      "loss": 2.5017,
+      "step": 445
+    },
+    {
+      "epoch": 0.4563894523326572,
+      "grad_norm": 1.7663094048322825,
+      "learning_rate": 4.2731845816978475e-06,
+      "loss": 2.5021,
+      "step": 450
+    },
+    {
+      "epoch": 0.4614604462474645,
+      "grad_norm": 1.7799998175038592,
+      "learning_rate": 4.207254852295854e-06,
+      "loss": 2.4953,
+      "step": 455
+    },
+    {
+      "epoch": 0.4665314401622718,
+      "grad_norm": 1.6715645487953392,
+      "learning_rate": 4.142262139981073e-06,
+      "loss": 2.4435,
+      "step": 460
+    },
+    {
+      "epoch": 0.4716024340770791,
+      "grad_norm": 1.7256265015398793,
+      "learning_rate": 4.078194442168494e-06,
+      "loss": 2.5146,
+      "step": 465
+    },
+    {
+      "epoch": 0.4766734279918864,
+      "grad_norm": 1.6662015811964308,
+      "learning_rate": 4.015039893130705e-06,
+      "loss": 2.5187,
+      "step": 470
+    },
+    {
+      "epoch": 0.4817444219066937,
+      "grad_norm": 1.7649431318197315,
+      "learning_rate": 3.952786762630535e-06,
+      "loss": 2.5223,
+      "step": 475
+    },
+    {
+      "epoch": 0.486815415821501,
+      "grad_norm": 1.679617464261057,
+      "learning_rate": 3.891423454565385e-06,
+      "loss": 2.4394,
+      "step": 480
+    },
+    {
+      "epoch": 0.4918864097363083,
+      "grad_norm": 1.6233085596184735,
+      "learning_rate": 3.830938505623211e-06,
+      "loss": 2.512,
+      "step": 485
+    },
+    {
+      "epoch": 0.4969574036511156,
+      "grad_norm": 1.7195900327055993,
+      "learning_rate": 3.7713205839500707e-06,
+      "loss": 2.4649,
+      "step": 490
+    },
+    {
+      "epoch": 0.5020283975659229,
+      "grad_norm": 1.7034828407083669,
+      "learning_rate": 3.7125584878291374e-06,
+      "loss": 2.497,
+      "step": 495
+    },
+    {
+      "epoch": 0.5070993914807302,
+      "grad_norm": 1.7618287486879018,
+      "learning_rate": 3.6546411443711164e-06,
+      "loss": 2.5353,
+      "step": 500
+    },
+    {
+      "epoch": 0.5121703853955375,
+      "grad_norm": 1.6191614066287776,
+      "learning_rate": 3.597557608215969e-06,
+      "loss": 2.5052,
+      "step": 505
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 1.6450813134062763,
+      "learning_rate": 3.54129706024587e-06,
+      "loss": 2.5106,
+      "step": 510
+    },
+    {
+      "epoch": 0.5223123732251521,
+      "grad_norm": 1.7767916102532666,
+      "learning_rate": 3.4858488063093135e-06,
+      "loss": 2.4651,
+      "step": 515
+    },
+    {
+      "epoch": 0.5273833671399595,
+      "grad_norm": 1.6720237829560067,
+      "learning_rate": 3.431202275956285e-06,
+      "loss": 2.4908,
+      "step": 520
+    },
+    {
+      "epoch": 0.5324543610547667,
+      "grad_norm": 1.6484154917054958,
+      "learning_rate": 3.3773470211844283e-06,
+      "loss": 2.4856,
+      "step": 525
+    },
+    {
+      "epoch": 0.537525354969574,
+      "grad_norm": 1.651838194240797,
+      "learning_rate": 3.324272715196116e-06,
+      "loss": 2.4675,
+      "step": 530
+    },
+    {
+      "epoch": 0.5425963488843814,
+      "grad_norm": 1.6241151521510617,
+      "learning_rate": 3.2719691511663524e-06,
+      "loss": 2.4896,
+      "step": 535
+    },
+    {
+      "epoch": 0.5476673427991886,
+      "grad_norm": 1.6894175077795812,
+      "learning_rate": 3.2204262410214273e-06,
+      "loss": 2.4556,
+      "step": 540
+    },
+    {
+      "epoch": 0.552738336713996,
+      "grad_norm": 1.6686417855987385,
+      "learning_rate": 3.1696340142282437e-06,
+      "loss": 2.5062,
+      "step": 545
+    },
+    {
+      "epoch": 0.5578093306288032,
+      "grad_norm": 1.7200856267540612,
+      "learning_rate": 3.119582616594238e-06,
+      "loss": 2.4878,
+      "step": 550
+    },
+    {
+      "epoch": 0.5628803245436106,
+      "grad_norm": 1.672252633477676,
+      "learning_rate": 3.0702623090778174e-06,
+      "loss": 2.5077,
+      "step": 555
+    },
+    {
+      "epoch": 0.5679513184584178,
+      "grad_norm": 1.7008466667698958,
+      "learning_rate": 3.021663466609246e-06,
+      "loss": 2.4837,
+      "step": 560
+    },
+    {
+      "epoch": 0.5730223123732252,
+      "grad_norm": 1.6805676799462346,
+      "learning_rate": 2.973776576921883e-06,
+      "loss": 2.5062,
+      "step": 565
+    },
+    {
+      "epoch": 0.5780933062880325,
+      "grad_norm": 1.6136103005628197,
+      "learning_rate": 2.9265922393937183e-06,
+      "loss": 2.5035,
+      "step": 570
+    },
+    {
+      "epoch": 0.5831643002028397,
+      "grad_norm": 1.6014078073339035,
+      "learning_rate": 2.880101163899116e-06,
+      "loss": 2.5101,
+      "step": 575
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 1.7220406203120746,
+      "learning_rate": 2.8342941696706994e-06,
+      "loss": 2.5217,
+      "step": 580
+    },
+    {
+      "epoch": 0.5933062880324543,
+      "grad_norm": 1.6605964063316545,
+      "learning_rate": 2.789162184171294e-06,
+      "loss": 2.4756,
+      "step": 585
+    },
+    {
+      "epoch": 0.5983772819472617,
+      "grad_norm": 1.6566249973518374,
+      "learning_rate": 2.7446962419758632e-06,
+      "loss": 2.4739,
+      "step": 590
+    },
+    {
+      "epoch": 0.603448275862069,
+      "grad_norm": 1.6340883136536262,
+      "learning_rate": 2.700887483663357e-06,
+      "loss": 2.4869,
+      "step": 595
+    },
+    {
+      "epoch": 0.6085192697768763,
+      "grad_norm": 1.6233109361058542,
+      "learning_rate": 2.657727154718401e-06,
+      "loss": 2.4487,
+      "step": 600
+    },
+    {
+      "epoch": 0.6085192697768763,
+      "eval_loss": 2.4831416606903076,
+      "eval_runtime": 80.984,
+      "eval_samples_per_second": 86.511,
+      "eval_steps_per_second": 0.679,
+      "step": 600
+    },
+    {
+      "epoch": 0.6135902636916836,
+      "grad_norm": 1.616769928055098,
+      "learning_rate": 2.615206604442756e-06,
+      "loss": 2.4638,
+      "step": 605
+    },
+    {
+      "epoch": 0.6186612576064908,
+      "grad_norm": 1.6396235170920117,
+      "learning_rate": 2.5733172848764733e-06,
+      "loss": 2.4891,
+      "step": 610
+    },
+    {
+      "epoch": 0.6237322515212982,
+      "grad_norm": 1.5936144163067276,
+      "learning_rate": 2.5320507497286705e-06,
+      "loss": 2.4902,
+      "step": 615
+    },
+    {
+      "epoch": 0.6288032454361054,
+      "grad_norm": 1.6679977682798468,
+      "learning_rate": 2.491398653317866e-06,
+      "loss": 2.4695,
+      "step": 620
+    },
+    {
+      "epoch": 0.6338742393509128,
+      "grad_norm": 1.7008178983911084,
+      "learning_rate": 2.4513527495217875e-06,
+      "loss": 2.4626,
+      "step": 625
+    },
+    {
+      "epoch": 0.6389452332657201,
+      "grad_norm": 1.610985443276998,
+      "learning_rate": 2.4119048907365937e-06,
+      "loss": 2.4934,
+      "step": 630
+    },
+    {
+      "epoch": 0.6440162271805274,
+      "grad_norm": 1.6323121910464156,
+      "learning_rate": 2.3730470268454385e-06,
+      "loss": 2.4819,
+      "step": 635
+    },
+    {
+      "epoch": 0.6490872210953347,
+      "grad_norm": 1.6525382291119861,
+      "learning_rate": 2.3347712041962997e-06,
+      "loss": 2.5046,
+      "step": 640
+    },
+    {
+      "epoch": 0.654158215010142,
+      "grad_norm": 1.6380351817927594,
+      "learning_rate": 2.297069564589013e-06,
+      "loss": 2.4864,
+      "step": 645
+    },
+    {
+      "epoch": 0.6592292089249493,
+      "grad_norm": 1.6579813340009797,
+      "learning_rate": 2.259934344271433e-06,
+      "loss": 2.4715,
+      "step": 650
+    },
+    {
+      "epoch": 0.6643002028397565,
+      "grad_norm": 1.7919239246160015,
+      "learning_rate": 2.22335787294466e-06,
+      "loss": 2.4972,
+      "step": 655
+    },
+    {
+      "epoch": 0.6693711967545639,
+      "grad_norm": 1.586961409961355,
+      "learning_rate": 2.18733257277726e-06,
+      "loss": 2.4787,
+      "step": 660
+    },
+    {
+      "epoch": 0.6744421906693712,
+      "grad_norm": 1.684301176230389,
+      "learning_rate": 2.1518509574284106e-06,
+      "loss": 2.4158,
+      "step": 665
+    },
+    {
+      "epoch": 0.6795131845841785,
+      "grad_norm": 1.6178388175493554,
+      "learning_rate": 2.123852145211829e-06,
+      "loss": 2.5152,
+      "step": 670
+    },
+    {
+      "epoch": 0.6845841784989858,
+      "grad_norm": 1.704137336957441,
+      "learning_rate": 2.089330585293108e-06,
+      "loss": 2.4807,
+      "step": 675
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 1.653288753563856,
+      "learning_rate": 2.055332226962747e-06,
+      "loss": 2.4781,
+      "step": 680
+    },
+    {
+      "epoch": 0.6947261663286004,
+      "grad_norm": 1.6910190620923418,
+      "learning_rate": 2.0218499227907136e-06,
+      "loss": 2.5114,
+      "step": 685
+    },
+    {
+      "epoch": 0.6997971602434077,
+      "grad_norm": 1.6297896630103186,
+      "learning_rate": 1.988876612270826e-06,
+      "loss": 2.4963,
+      "step": 690
+    },
+    {
+      "epoch": 0.704868154158215,
+      "grad_norm": 1.6254042637268307,
+      "learning_rate": 1.9564053208943578e-06,
+      "loss": 2.4651,
+      "step": 695
+    },
+    {
+      "epoch": 0.7099391480730223,
+      "grad_norm": 1.849820644961665,
+      "learning_rate": 1.924429159232111e-06,
+      "loss": 2.4625,
+      "step": 700
+    },
+    {
+      "epoch": 0.7150101419878296,
+      "grad_norm": 1.6947938784926828,
+      "learning_rate": 1.892941322024907e-06,
+      "loss": 2.4683,
+      "step": 705
+    },
+    {
+      "epoch": 0.7200811359026369,
+      "grad_norm": 1.6500218076608433,
+      "learning_rate": 1.861935087282421e-06,
+      "loss": 2.474,
+      "step": 710
+    },
+    {
+      "epoch": 0.7251521298174443,
+      "grad_norm": 1.5695461599237197,
+      "learning_rate": 1.8314038153902991e-06,
+      "loss": 2.4626,
+      "step": 715
+    },
+    {
+      "epoch": 0.7302231237322515,
+      "grad_norm": 1.661274439764298,
+      "learning_rate": 1.8013409482254947e-06,
+      "loss": 2.4901,
+      "step": 720
+    },
+    {
+      "epoch": 0.7352941176470589,
+      "grad_norm": 1.5971717624468098,
+      "learning_rate": 1.7717400082797614e-06,
+      "loss": 2.498,
+      "step": 725
+    },
+    {
+      "epoch": 0.7403651115618661,
+      "grad_norm": 1.6006841184664817,
+      "learning_rate": 1.7425945977912387e-06,
+      "loss": 2.5096,
+      "step": 730
+    },
+    {
+      "epoch": 0.7454361054766734,
+      "grad_norm": 1.8078007149616142,
+      "learning_rate": 1.7138983978840686e-06,
+      "loss": 2.4733,
+      "step": 735
+    },
+    {
+      "epoch": 0.7505070993914807,
+      "grad_norm": 1.6080637102108633,
+      "learning_rate": 1.685645167715982e-06,
+      "loss": 2.4645,
+      "step": 740
+    },
+    {
+      "epoch": 0.755578093306288,
+      "grad_norm": 1.6034092883417612,
+      "learning_rate": 1.6578287436337897e-06,
+      "loss": 2.4874,
+      "step": 745
+    },
+    {
+      "epoch": 0.7606490872210954,
+      "grad_norm": 1.6562691168973722,
+      "learning_rate": 1.6304430383367233e-06,
+      "loss": 2.5147,
+      "step": 750
+    },
+    {
+      "epoch": 0.7657200811359026,
+      "grad_norm": 1.631836734297837,
+      "learning_rate": 1.6034820400475576e-06,
+      "loss": 2.449,
+      "step": 755
+    },
+    {
+      "epoch": 0.77079107505071,
+      "grad_norm": 2.633902381426751,
+      "learning_rate": 1.5769398116914607e-06,
+      "loss": 2.4502,
+      "step": 760
+    },
+    {
+      "epoch": 0.7758620689655172,
+      "grad_norm": 1.6338196504524252,
+      "learning_rate": 1.550810490082507e-06,
+      "loss": 2.4375,
+      "step": 765
+    },
+    {
+      "epoch": 0.7809330628803245,
+      "grad_norm": 1.6881605246261733,
+      "learning_rate": 1.5250882851177956e-06,
+      "loss": 2.4623,
+      "step": 770
+    },
+    {
+      "epoch": 0.7860040567951319,
+      "grad_norm": 1.7430128340035491,
+      "learning_rate": 1.4997674789791142e-06,
+      "loss": 2.4592,
+      "step": 775
+    },
+    {
+      "epoch": 0.7910750507099391,
+      "grad_norm": 1.6974037503954427,
+      "learning_rate": 1.4748424253420905e-06,
+      "loss": 2.5001,
+      "step": 780
+    },
+    {
+      "epoch": 0.7961460446247465,
+      "grad_norm": 1.6057434981804433,
+      "learning_rate": 1.4503075485927704e-06,
+      "loss": 2.4603,
+      "step": 785
+    },
+    {
+      "epoch": 0.8012170385395537,
+      "grad_norm": 1.5564356238507298,
+      "learning_rate": 1.4261573430515669e-06,
+      "loss": 2.4357,
+      "step": 790
+    },
+    {
+      "epoch": 0.8062880324543611,
+      "grad_norm": 1.7042405076576008,
+      "learning_rate": 1.4023863722045201e-06,
+      "loss": 2.4747,
+      "step": 795
+    },
+    {
+      "epoch": 0.8113590263691683,
+      "grad_norm": 1.5640034942530554,
+      "learning_rate": 1.3789892679418134e-06,
+      "loss": 2.5324,
+      "step": 800
+    },
+    {
+      "epoch": 0.8113590263691683,
+      "eval_loss": 2.4689557552337646,
+      "eval_runtime": 81.0232,
+      "eval_samples_per_second": 86.469,
+      "eval_steps_per_second": 0.679,
+      "step": 800
+    },
+    {
+      "epoch": 0.8164300202839757,
+      "grad_norm": 1.7227060519078905,
+      "learning_rate": 1.3559607298034838e-06,
+      "loss": 2.4806,
+      "step": 805
+    },
+    {
+      "epoch": 0.821501014198783,
+      "grad_norm": 1.5855673393298833,
+      "learning_rate": 1.333295524232277e-06,
+      "loss": 2.4642,
+      "step": 810
+    },
+    {
+      "epoch": 0.8265720081135902,
+      "grad_norm": 1.8155636812941185,
+      "learning_rate": 1.310988483833583e-06,
+      "loss": 2.4746,
+      "step": 815
+    },
+    {
+      "epoch": 0.8316430020283976,
+      "grad_norm": 1.6824796691575312,
+      "learning_rate": 1.289034506642401e-06,
+      "loss": 2.5168,
+      "step": 820
+    },
+    {
+      "epoch": 0.8367139959432048,
+      "grad_norm": 1.6084122349859742,
+      "learning_rate": 1.2674285553972776e-06,
+      "loss": 2.4112,
+      "step": 825
+    },
+    {
+      "epoch": 0.8417849898580122,
+      "grad_norm": 1.6807591569306923,
+      "learning_rate": 1.2461656568211607e-06,
+      "loss": 2.4555,
+      "step": 830
+    },
+    {
+      "epoch": 0.8468559837728195,
+      "grad_norm": 1.64520194930749,
+      "learning_rate": 1.2252409009091154e-06,
+      "loss": 2.5222,
+      "step": 835
+    },
+    {
+      "epoch": 0.8519269776876268,
+      "grad_norm": 1.642941398726877,
+      "learning_rate": 1.2046494402228485e-06,
+      "loss": 2.4607,
+      "step": 840
+    },
+    {
+      "epoch": 0.8569979716024341,
+      "grad_norm": 1.6323907187692908,
+      "learning_rate": 1.1843864891919843e-06,
+      "loss": 2.4724,
+      "step": 845
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 1.6489728444762863,
+      "learning_rate": 1.1644473234220412e-06,
+      "loss": 2.483,
+      "step": 850
+    },
+    {
+      "epoch": 0.8671399594320487,
+      "grad_norm": 1.5735584816022383,
+      "learning_rate": 1.1448272790090529e-06,
+      "loss": 2.4423,
+      "step": 855
+    },
+    {
+      "epoch": 0.8722109533468559,
+      "grad_norm": 1.6290164674794758,
+      "learning_rate": 1.1255217518607806e-06,
+      "loss": 2.4745,
+      "step": 860
+    },
+    {
+      "epoch": 0.8772819472616633,
+      "grad_norm": 1.9631129344699565,
+      "learning_rate": 1.1065261970244678e-06,
+      "loss": 2.4595,
+      "step": 865
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 1.8876833985138877,
+      "learning_rate": 1.0878361280210782e-06,
+      "loss": 2.4761,
+      "step": 870
+    },
+    {
+      "epoch": 0.8874239350912779,
+      "grad_norm": 1.7449962901668679,
+      "learning_rate": 1.0694471161859696e-06,
+      "loss": 2.4726,
+      "step": 875
+    },
+    {
+      "epoch": 0.8924949290060852,
+      "grad_norm": 1.6608657901001447,
+      "learning_rate": 1.051354790015952e-06,
+      "loss": 2.4817,
+      "step": 880
+    },
+    {
+      "epoch": 0.8975659229208925,
+      "grad_norm": 1.6370419920913908,
+      "learning_rate": 1.0335548345226733e-06,
+      "loss": 2.4861,
+      "step": 885
+    },
+    {
+      "epoch": 0.9026369168356998,
+      "grad_norm": 1.6266725844295284,
+      "learning_rate": 1.016042990592287e-06,
+      "loss": 2.4437,
+      "step": 890
+    },
+    {
+      "epoch": 0.907707910750507,
+      "grad_norm": 1.5909779389607082,
+      "learning_rate": 9.988150543513476e-07,
+      "loss": 2.4605,
+      "step": 895
+    },
+    {
+      "epoch": 0.9127789046653144,
+      "grad_norm": 1.5796802186393568,
+      "learning_rate": 9.818668765388872e-07,
+      "loss": 2.4863,
+      "step": 900
+    },
+    {
+      "epoch": 0.9178498985801217,
+      "grad_norm": 1.5779871460684796,
+      "learning_rate": 9.651943618846152e-07,
+      "loss": 2.4514,
+      "step": 905
+    },
+    {
+      "epoch": 0.922920892494929,
+      "grad_norm": 1.605102383763968,
+      "learning_rate": 9.487934684931995e-07,
+      "loss": 2.474,
+      "step": 910
+    },
+    {
+      "epoch": 0.9279918864097363,
+      "grad_norm": 1.6069103870683263,
+      "learning_rate": 9.326602072345758e-07,
+      "loss": 2.4828,
+      "step": 915
+    },
+    {
+      "epoch": 0.9330628803245437,
+      "grad_norm": 1.6236038441464034,
+      "learning_rate": 9.167906411402357e-07,
+      "loss": 2.4501,
+      "step": 920
+    },
+    {
+      "epoch": 0.9381338742393509,
+      "grad_norm": 1.6140284100171378,
+      "learning_rate": 9.011808848054445e-07,
+      "loss": 2.4441,
+      "step": 925
+    },
+    {
+      "epoch": 0.9432048681541582,
+      "grad_norm": 1.9823289784825078,
+      "learning_rate": 8.858271037973411e-07,
+      "loss": 2.4834,
+      "step": 930
+    },
+    {
+      "epoch": 0.9482758620689655,
+      "grad_norm": 1.7094985628575186,
+      "learning_rate": 8.707255140688767e-07,
+      "loss": 2.4428,
+      "step": 935
+    },
+    {
+      "epoch": 0.9533468559837728,
+      "grad_norm": 1.5851821971427773,
+      "learning_rate": 8.558723813785198e-07,
+      "loss": 2.4459,
+      "step": 940
+    },
+    {
+      "epoch": 0.9584178498985801,
+      "grad_norm": 1.8489283203955083,
+      "learning_rate": 8.412640207157327e-07,
+      "loss": 2.4671,
+      "step": 945
+    },
+    {
+      "epoch": 0.9634888438133874,
+      "grad_norm": 1.565327828926634,
+      "learning_rate": 8.268967957320976e-07,
+      "loss": 2.4762,
+      "step": 950
+    },
+    {
+      "epoch": 0.9685598377281948,
+      "grad_norm": 1.5753092524917698,
+      "learning_rate": 8.127671181781262e-07,
+      "loss": 2.487,
+      "step": 955
+    },
+    {
+      "epoch": 0.973630831643002,
+      "grad_norm": 1.5627741498336793,
+      "learning_rate": 7.988714473456279e-07,
+      "loss": 2.4899,
+      "step": 960
+    },
+    {
+      "epoch": 0.9787018255578094,
+      "grad_norm": 1.7322054425536324,
+      "learning_rate": 7.852062895156654e-07,
+      "loss": 2.4328,
+      "step": 965
+    },
+    {
+      "epoch": 0.9837728194726166,
+      "grad_norm": 1.5912533141539165,
+      "learning_rate": 7.717681974119764e-07,
+      "loss": 2.4887,
+      "step": 970
+    },
+    {
+      "epoch": 0.9888438133874239,
+      "grad_norm": 1.7127177872013957,
+      "learning_rate": 7.585537696598922e-07,
+      "loss": 2.4414,
+      "step": 975
+    },
+    {
+      "epoch": 0.9939148073022313,
+      "grad_norm": 1.6239111267541033,
+      "learning_rate": 7.455596502506312e-07,
+      "loss": 2.4962,
+      "step": 980
+    },
+    {
+      "epoch": 0.9989858012170385,
+      "grad_norm": 1.6117561424503084,
+      "learning_rate": 7.327825280109957e-07,
+      "loss": 2.4738,
+      "step": 985
+    },
+    {
+      "epoch": 1.0040567951318458,
+      "grad_norm": 1.9019039739296713,
+      "learning_rate": 7.20219136078357e-07,
+      "loss": 2.27,
+      "step": 990
+    },
+    {
+      "epoch": 1.0091277890466532,
+      "grad_norm": 1.7075178009820928,
+      "learning_rate": 7.078662513809528e-07,
+      "loss": 2.3072,
+      "step": 995
+    },
+    {
+      "epoch": 1.0141987829614605,
+      "grad_norm": 1.7844249258995124,
+      "learning_rate": 6.957206941233838e-07,
+      "loss": 2.265,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0141987829614605,
+      "eval_loss": 2.473280668258667,
+      "eval_runtime": 81.0085,
+      "eval_samples_per_second": 86.485,
+      "eval_steps_per_second": 0.679,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0192697768762677,
+      "grad_norm": 1.833316481131949,
+      "learning_rate": 6.837793272773345e-07,
+      "loss": 2.3069,
+      "step": 1005
+    },
+    {
+      "epoch": 1.024340770791075,
+      "grad_norm": 1.7388775994426842,
+      "learning_rate": 6.720390560774066e-07,
+      "loss": 2.266,
+      "step": 1010
+    },
+    {
+      "epoch": 1.0294117647058822,
+      "grad_norm": 1.6270190329782648,
+      "learning_rate": 6.604968275220875e-07,
+      "loss": 2.2664,
+      "step": 1015
+    },
+    {
+      "epoch": 1.0344827586206897,
+      "grad_norm": 1.7956207149367391,
+      "learning_rate": 6.491496298797458e-07,
+      "loss": 2.2394,
+      "step": 1020
+    },
+    {
+      "epoch": 1.039553752535497,
+      "grad_norm": 1.6994135825189252,
+      "learning_rate": 6.379944921996764e-07,
+      "loss": 2.2727,
+      "step": 1025
+    },
+    {
+      "epoch": 1.0446247464503042,
+      "grad_norm": 1.677197538792328,
+      "learning_rate": 6.270284838280882e-07,
+      "loss": 2.2072,
+      "step": 1030
+    },
+    {
+      "epoch": 1.0496957403651115,
+      "grad_norm": 1.719327046611783,
+      "learning_rate": 6.162487139290532e-07,
+      "loss": 2.3021,
+      "step": 1035
+    },
+    {
+      "epoch": 1.054766734279919,
+      "grad_norm": 1.7292340128968464,
+      "learning_rate": 6.056523310103172e-07,
+      "loss": 2.2737,
+      "step": 1040
+    },
+    {
+      "epoch": 1.0598377281947262,
+      "grad_norm": 1.7428974260955565,
+      "learning_rate": 5.95236522453988e-07,
+      "loss": 2.2556,
+      "step": 1045
+    },
+    {
+      "epoch": 1.0649087221095335,
+      "grad_norm": 1.694959472171586,
+      "learning_rate": 5.849985140519998e-07,
+      "loss": 2.2992,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0699797160243407,
+      "grad_norm": 1.7439692178448947,
+      "learning_rate": 5.749355695463754e-07,
+      "loss": 2.2557,
+      "step": 1055
+    },
+    {
+      "epoch": 1.075050709939148,
+      "grad_norm": 1.7558636029085997,
+      "learning_rate": 5.650449901741813e-07,
+      "loss": 2.2474,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0801217038539555,
+      "grad_norm": 1.785367595963534,
+      "learning_rate": 5.553241142171985e-07,
+      "loss": 2.267,
+      "step": 1065
+    },
+    {
+      "epoch": 1.0851926977687627,
+      "grad_norm": 1.7537584511707027,
+      "learning_rate": 5.45770316556211e-07,
+      "loss": 2.2823,
+      "step": 1070
+    },
+    {
+      "epoch": 1.09026369168357,
+      "grad_norm": 1.6825060417395732,
+      "learning_rate": 5.363810082299148e-07,
+      "loss": 2.2525,
+      "step": 1075
+    },
+    {
+      "epoch": 1.0953346855983772,
+      "grad_norm": 1.7339475460772475,
+      "learning_rate": 5.27153635998387e-07,
+      "loss": 2.3006,
+      "step": 1080
+    },
+    {
+      "epoch": 1.1004056795131847,
+      "grad_norm": 1.6977028436147512,
+      "learning_rate": 5.180856819110773e-07,
+      "loss": 2.2862,
+      "step": 1085
+    },
+    {
+      "epoch": 1.105476673427992,
+      "grad_norm": 1.7119437312783958,
+      "learning_rate": 5.091746628792904e-07,
+      "loss": 2.243,
+      "step": 1090
+    },
+    {
+      "epoch": 1.1105476673427992,
+      "grad_norm": 1.7918277133466605,
+      "learning_rate": 5.004181302531108e-07,
+      "loss": 2.2653,
+      "step": 1095
+    },
+    {
+      "epoch": 1.1156186612576064,
+      "grad_norm": 1.7198038075584687,
+      "learning_rate": 4.918136694027396e-07,
+      "loss": 2.2741,
+      "step": 1100
+    },
+    {
+      "epoch": 1.1206896551724137,
+      "grad_norm": 1.7122122501534425,
+      "learning_rate": 4.833588993041994e-07,
+      "loss": 2.2757,
+      "step": 1105
+    },
+    {
+      "epoch": 1.1257606490872212,
+      "grad_norm": 1.6934117050919777,
+      "learning_rate": 4.750514721293719e-07,
+      "loss": 2.2484,
+      "step": 1110
+    },
+    {
+      "epoch": 1.1308316430020284,
+      "grad_norm": 1.8096755323665539,
+      "learning_rate": 4.6688907284032994e-07,
+      "loss": 2.2329,
+      "step": 1115
+    },
+    {
+      "epoch": 1.1359026369168357,
+      "grad_norm": 1.7732841203420067,
+      "learning_rate": 4.588694187879258e-07,
+      "loss": 2.2636,
+      "step": 1120
+    },
+    {
+      "epoch": 1.140973630831643,
+      "grad_norm": 1.70514589311023,
+      "learning_rate": 4.5099025931459913e-07,
+      "loss": 2.2778,
+      "step": 1125
+    },
+    {
+      "epoch": 1.1460446247464504,
+      "grad_norm": 1.7135354540773058,
+      "learning_rate": 4.4324937536136735e-07,
+      "loss": 2.2905,
+      "step": 1130
+    },
+    {
+      "epoch": 1.1511156186612577,
+      "grad_norm": 1.6901713268949445,
+      "learning_rate": 4.3564457907896125e-07,
+      "loss": 2.302,
+      "step": 1135
+    },
+    {
+      "epoch": 1.156186612576065,
+      "grad_norm": 1.7350424488382163,
+      "learning_rate": 4.281737134430704e-07,
+      "loss": 2.2441,
+      "step": 1140
+    },
+    {
+      "epoch": 1.1612576064908722,
+      "grad_norm": 1.7433418190612922,
+      "learning_rate": 4.208346518736604e-07,
+      "loss": 2.2639,
+      "step": 1145
+    },
+    {
+      "epoch": 1.1663286004056794,
+      "grad_norm": 1.7278183208713844,
+      "learning_rate": 4.136252978583281e-07,
+      "loss": 2.272,
+      "step": 1150
+    },
+    {
+      "epoch": 1.171399594320487,
+      "grad_norm": 1.7049575091462312,
+      "learning_rate": 4.0654358457965706e-07,
+      "loss": 2.2822,
+      "step": 1155
+    },
+    {
+      "epoch": 1.1764705882352942,
+      "grad_norm": 1.7614119208994081,
+      "learning_rate": 3.995874745465392e-07,
+      "loss": 2.2882,
+      "step": 1160
+    },
+    {
+      "epoch": 1.1815415821501014,
+      "grad_norm": 1.7783667378053016,
+      "learning_rate": 3.927549592294267e-07,
+      "loss": 2.2779,
+      "step": 1165
+    },
+    {
+      "epoch": 1.1866125760649087,
+      "grad_norm": 1.7857803604726208,
+      "learning_rate": 3.8604405869947905e-07,
+      "loss": 2.2504,
+      "step": 1170
+    },
+    {
+      "epoch": 1.1916835699797161,
+      "grad_norm": 1.7894737586957659,
+      "learning_rate": 3.794528212715714e-07,
+      "loss": 2.2896,
+      "step": 1175
+    },
+    {
+      "epoch": 1.1967545638945234,
+      "grad_norm": 1.7605294591830605,
+      "learning_rate": 3.7297932315112855e-07,
+      "loss": 2.2803,
+      "step": 1180
+    },
+    {
+      "epoch": 1.2018255578093306,
+      "grad_norm": 1.7037189312181982,
+      "learning_rate": 3.6662166808475126e-07,
+      "loss": 2.2595,
+      "step": 1185
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 1.802568691083643,
+      "learning_rate": 3.6037798701460037e-07,
+      "loss": 2.3097,
+      "step": 1190
+    },
+    {
+      "epoch": 1.2119675456389452,
+      "grad_norm": 1.7227242510965723,
+      "learning_rate": 3.5424643773650545e-07,
+      "loss": 2.2473,
+      "step": 1195
+    },
+    {
+      "epoch": 1.2170385395537526,
+      "grad_norm": 1.7126735182979083,
+      "learning_rate": 3.482252045617637e-07,
+      "loss": 2.3002,
+      "step": 1200
+    },
+    {
+      "epoch": 1.2170385395537526,
+      "eval_loss": 2.4735846519470215,
+      "eval_runtime": 81.0924,
+      "eval_samples_per_second": 86.395,
+      "eval_steps_per_second": 0.678,
+      "step": 1200
+    },
+    {
+      "epoch": 1.2221095334685599,
+      "grad_norm": 1.7418672417675343,
+      "learning_rate": 3.423124979825969e-07,
+      "loss": 2.2259,
+      "step": 1205
+    },
+    {
+      "epoch": 1.2271805273833671,
+      "grad_norm": 1.7536106052680211,
+      "learning_rate": 3.365065543412324e-07,
+      "loss": 2.2625,
+      "step": 1210
+    },
+    {
+      "epoch": 1.2322515212981744,
+      "grad_norm": 1.6738354256007202,
+      "learning_rate": 3.3080563550257607e-07,
+      "loss": 2.2762,
+      "step": 1215
+    },
+    {
+      "epoch": 1.2373225152129819,
+      "grad_norm": 1.7304199756653005,
+      "learning_rate": 3.2520802853044393e-07,
+      "loss": 2.2864,
+      "step": 1220
+    },
+    {
+      "epoch": 1.2423935091277891,
+      "grad_norm": 1.761088776037141,
+      "learning_rate": 3.197120453673215e-07,
+      "loss": 2.2665,
+      "step": 1225
+    },
+    {
+      "epoch": 1.2474645030425964,
+      "grad_norm": 1.7101358055188194,
+      "learning_rate": 3.143160225176168e-07,
+      "loss": 2.2775,
+      "step": 1230
+    },
+    {
+      "epoch": 1.2525354969574036,
+      "grad_norm": 1.7571854143932952,
+      "learning_rate": 3.0901832073437713e-07,
+      "loss": 2.2979,
+      "step": 1235
+    },
+    {
+      "epoch": 1.2576064908722109,
+      "grad_norm": 1.7216743809437804,
+      "learning_rate": 3.0381732470943653e-07,
+      "loss": 2.3094,
+      "step": 1240
+    },
+    {
+      "epoch": 1.2626774847870181,
+      "grad_norm": 1.6935950803242086,
+      "learning_rate": 2.9871144276696387e-07,
+      "loss": 2.2707,
+      "step": 1245
+    },
+    {
+      "epoch": 1.2677484787018256,
+      "grad_norm": 1.7158452472154153,
+      "learning_rate": 2.9369910656037903e-07,
+      "loss": 2.2532,
+      "step": 1250
+    },
+    {
+      "epoch": 1.2728194726166329,
+      "grad_norm": 1.7587458046328184,
+      "learning_rate": 2.8877877077260676e-07,
+      "loss": 2.2968,
+      "step": 1255
+    },
+    {
+      "epoch": 1.2778904665314401,
+      "grad_norm": 1.7348605445713965,
+      "learning_rate": 2.839489128196406e-07,
+      "loss": 2.2596,
+      "step": 1260
+    },
+    {
+      "epoch": 1.2829614604462476,
+      "grad_norm": 1.6962275978449755,
+      "learning_rate": 2.7920803255737635e-07,
+      "loss": 2.2579,
+      "step": 1265
+    },
+    {
+      "epoch": 1.2880324543610548,
+      "grad_norm": 1.7562952815143784,
+      "learning_rate": 2.7455465199170286e-07,
+      "loss": 2.2518,
+      "step": 1270
+    },
+    {
+      "epoch": 1.293103448275862,
+      "grad_norm": 1.6974150722131578,
+      "learning_rate": 2.699873149917968e-07,
+      "loss": 2.2504,
+      "step": 1275
+    },
+    {
+      "epoch": 1.2981744421906694,
+      "grad_norm": 1.7036916845012207,
+      "learning_rate": 2.655045870066172e-07,
+      "loss": 2.2861,
+      "step": 1280
+    },
+    {
+      "epoch": 1.3032454361054766,
+      "grad_norm": 1.7486208966066876,
+      "learning_rate": 2.6110505478454324e-07,
+      "loss": 2.2467,
+      "step": 1285
+    },
+    {
+      "epoch": 1.3083164300202839,
+      "grad_norm": 1.712258524308874,
+      "learning_rate": 2.5678732609615423e-07,
+      "loss": 2.2515,
+      "step": 1290
+    },
+    {
+      "epoch": 1.3133874239350913,
+      "grad_norm": 1.7341023622582277,
+      "learning_rate": 2.525500294600939e-07,
+      "loss": 2.2757,
+      "step": 1295
+    },
+    {
+      "epoch": 1.3184584178498986,
+      "grad_norm": 1.889990239211246,
+      "learning_rate": 2.4839181387201796e-07,
+      "loss": 2.2791,
+      "step": 1300
+    },
+    {
+      "epoch": 1.3235294117647058,
+      "grad_norm": 1.798861207791198,
+      "learning_rate": 2.4431134853656976e-07,
+      "loss": 2.2817,
+      "step": 1305
+    },
+    {
+      "epoch": 1.3286004056795133,
+      "grad_norm": 1.7472239831698717,
+      "learning_rate": 2.4030732260238086e-07,
+      "loss": 2.2521,
+      "step": 1310
+    },
+    {
+      "epoch": 1.3336713995943206,
+      "grad_norm": 1.782522588407923,
+      "learning_rate": 2.3637844490004408e-07,
+      "loss": 2.2316,
+      "step": 1315
+    },
+    {
+      "epoch": 1.3387423935091278,
+      "grad_norm": 1.6996053792107884,
+      "learning_rate": 2.325234436830538e-07,
+      "loss": 2.2734,
+      "step": 1320
+    },
+    {
+      "epoch": 1.343813387423935,
+      "grad_norm": 1.7994805518930097,
+      "learning_rate": 2.2874106637166403e-07,
+      "loss": 2.2484,
+      "step": 1325
+    },
+    {
+      "epoch": 1.3488843813387423,
+      "grad_norm": 1.7489331509437775,
+      "learning_rate": 2.2503007929965749e-07,
+      "loss": 2.28,
+      "step": 1330
+    },
+    {
+      "epoch": 1.3539553752535496,
+      "grad_norm": 1.7160678233869127,
+      "learning_rate": 2.2138926746397777e-07,
+      "loss": 2.2565,
+      "step": 1335
+    },
+    {
+      "epoch": 1.359026369168357,
+      "grad_norm": 1.814687918697313,
+      "learning_rate": 2.178174342772177e-07,
+      "loss": 2.2517,
+      "step": 1340
+    },
+    {
+      "epoch": 1.3640973630831643,
+      "grad_norm": 1.6987256946879317,
+      "learning_rate": 2.143134013229167e-07,
+      "loss": 2.2672,
+      "step": 1345
+    },
+    {
+      "epoch": 1.3691683569979716,
+      "grad_norm": 1.7371785897491874,
+      "learning_rate": 2.1087600811366032e-07,
+      "loss": 2.2628,
+      "step": 1350
+    },
+    {
+      "epoch": 1.3742393509127788,
+      "grad_norm": 1.745926263655127,
+      "learning_rate": 2.075041118519355e-07,
+      "loss": 2.2532,
+      "step": 1355
+    },
+    {
+      "epoch": 1.3793103448275863,
+      "grad_norm": 1.700613383279488,
+      "learning_rate": 2.0419658719373504e-07,
+      "loss": 2.2617,
+      "step": 1360
+    },
+    {
+      "epoch": 1.3843813387423936,
+      "grad_norm": 1.691103098158946,
+      "learning_rate": 2.009523260148652e-07,
+      "loss": 2.2391,
+      "step": 1365
+    },
+    {
+      "epoch": 1.3894523326572008,
+      "grad_norm": 1.6917956046319294,
+      "learning_rate": 1.977702371799498e-07,
+      "loss": 2.2973,
+      "step": 1370
+    },
+    {
+      "epoch": 1.394523326572008,
+      "grad_norm": 1.7504566996070137,
+      "learning_rate": 1.946492463140869e-07,
+      "loss": 2.3102,
+      "step": 1375
+    },
+    {
+      "epoch": 1.3995943204868153,
+      "grad_norm": 1.838843879022522,
+      "learning_rate": 1.9158829557714903e-07,
+      "loss": 2.2819,
+      "step": 1380
+    },
+    {
+      "epoch": 1.4046653144016228,
+      "grad_norm": 1.7034157869918263,
+      "learning_rate": 1.8858634344068625e-07,
+      "loss": 2.2463,
+      "step": 1385
+    },
+    {
+      "epoch": 1.40973630831643,
+      "grad_norm": 1.7726664220307162,
+      "learning_rate": 1.8564236446742146e-07,
+      "loss": 2.2458,
+      "step": 1390
+    },
+    {
+      "epoch": 1.4148073022312373,
+      "grad_norm": 1.7584441947795304,
+      "learning_rate": 1.8275534909329853e-07,
+      "loss": 2.2663,
+      "step": 1395
+    },
+    {
+      "epoch": 1.4198782961460445,
+      "grad_norm": 1.7548926938859895,
+      "learning_rate": 1.7992430341207304e-07,
+      "loss": 2.29,
+      "step": 1400
+    },
+    {
+      "epoch": 1.4198782961460445,
+      "eval_loss": 2.4734323024749756,
+      "eval_runtime": 81.002,
+      "eval_samples_per_second": 86.492,
+      "eval_steps_per_second": 0.679,
+      "step": 1400
+    },
+    {
+      "epoch": 1.424949290060852,
+      "grad_norm": 1.691411914276979,
+      "learning_rate": 1.7714824896240595e-07,
+      "loss": 2.2565,
+      "step": 1405
+    },
+    {
+      "epoch": 1.4300202839756593,
+      "grad_norm": 1.7523279327159709,
+      "learning_rate": 1.7442622251745125e-07,
+      "loss": 2.2582,
+      "step": 1410
+    },
+    {
+      "epoch": 1.4350912778904665,
+      "grad_norm": 1.6844227513504313,
+      "learning_rate": 1.717572758768978e-07,
+      "loss": 2.2416,
+      "step": 1415
+    },
+    {
+      "epoch": 1.4401622718052738,
+      "grad_norm": 2.2030630647830245,
+      "learning_rate": 1.6914047566145662e-07,
+      "loss": 2.2289,
+      "step": 1420
+    },
+    {
+      "epoch": 1.445233265720081,
+      "grad_norm": 1.7795541841017355,
+      "learning_rate": 1.6657490310975468e-07,
+      "loss": 2.2841,
+      "step": 1425
+    },
+    {
+      "epoch": 1.4503042596348885,
+      "grad_norm": 1.8134633165357201,
+      "learning_rate": 1.6405965387762636e-07,
+      "loss": 2.2542,
+      "step": 1430
+    },
+    {
+      "epoch": 1.4553752535496958,
+      "grad_norm": 1.7604092301048675,
+      "learning_rate": 1.615938378397648e-07,
+      "loss": 2.2493,
+      "step": 1435
+    },
+    {
+      "epoch": 1.460446247464503,
+      "grad_norm": 1.8595724042593027,
+      "learning_rate": 1.5917657889372315e-07,
+      "loss": 2.2484,
+      "step": 1440
+    },
+    {
+      "epoch": 1.4655172413793103,
+      "grad_norm": 1.7081713686615858,
+      "learning_rate": 1.568070147662311e-07,
+      "loss": 2.2744,
+      "step": 1445
+    },
+    {
+      "epoch": 1.4705882352941178,
+      "grad_norm": 14.41030902656843,
+      "learning_rate": 1.5448429682181186e-07,
+      "loss": 2.2609,
+      "step": 1450
+    },
+    {
+      "epoch": 1.475659229208925,
+      "grad_norm": 1.7702111899429174,
+      "learning_rate": 1.5220758987367309e-07,
+      "loss": 2.2955,
+      "step": 1455
+    },
+    {
+      "epoch": 1.4807302231237323,
+      "grad_norm": 1.7932941724173908,
+      "learning_rate": 1.4997607199684964e-07,
+      "loss": 2.2478,
+      "step": 1460
+    },
+    {
+      "epoch": 1.4858012170385395,
+      "grad_norm": 1.7327449633169845,
+      "learning_rate": 1.477889343435765e-07,
+      "loss": 2.2713,
+      "step": 1465
+    },
+    {
+      "epoch": 1.4908722109533468,
+      "grad_norm": 1.7047486187689578,
+      "learning_rate": 1.456453809608691e-07,
+      "loss": 2.2586,
+      "step": 1470
+    },
+    {
+      "epoch": 1.495943204868154,
+      "grad_norm": 1.7085975289965103,
+      "learning_rate": 1.4354462861028889e-07,
+      "loss": 2.2602,
+      "step": 1475
+    },
+    {
+      "epoch": 1.5010141987829615,
+      "grad_norm": 1.7708851051604204,
+      "learning_rate": 1.414859065898731e-07,
+      "loss": 2.2913,
+      "step": 1480
+    },
+    {
+      "epoch": 1.5060851926977687,
+      "grad_norm": 1.6849008491575197,
+      "learning_rate": 1.3946845655820588e-07,
+      "loss": 2.2129,
+      "step": 1485
+    },
+    {
+      "epoch": 1.5111561866125762,
+      "grad_norm": 1.6770410018579935,
+      "learning_rate": 1.374915323606102e-07,
+      "loss": 2.2641,
+      "step": 1490
+    },
+    {
+      "epoch": 1.5162271805273835,
+      "grad_norm": 1.7333889728562109,
+      "learning_rate": 1.3555439985743863e-07,
+      "loss": 2.3096,
+      "step": 1495
+    },
+    {
+      "epoch": 1.5212981744421907,
+      "grad_norm": 1.7381149429179856,
+      "learning_rate": 1.3365633675444236e-07,
+      "loss": 2.2449,
+      "step": 1500
+    },
+    {
+      "epoch": 1.526369168356998,
+      "grad_norm": 1.7508604376509869,
+      "learning_rate": 1.317966324351968e-07,
+      "loss": 2.3006,
+      "step": 1505
+    },
+    {
+      "epoch": 1.5314401622718052,
+      "grad_norm": 1.731173156378831,
+      "learning_rate": 1.2997458779556342e-07,
+      "loss": 2.2721,
+      "step": 1510
+    },
+    {
+      "epoch": 1.5365111561866125,
+      "grad_norm": 1.7880722742651989,
+      "learning_rate": 1.2818951508016706e-07,
+      "loss": 2.2839,
+      "step": 1515
+    },
+    {
+      "epoch": 1.5415821501014197,
+      "grad_norm": 1.766456825336907,
+      "learning_rate": 1.264407377208682e-07,
+      "loss": 2.2542,
+      "step": 1520
+    },
+    {
+      "epoch": 1.5466531440162272,
+      "grad_norm": 1.793293076179441,
+      "learning_rate": 1.2472759017720967e-07,
+      "loss": 2.2345,
+      "step": 1525
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 1.7255231286858488,
+      "learning_rate": 1.2304941777881816e-07,
+      "loss": 2.2587,
+      "step": 1530
+    },
+    {
+      "epoch": 1.556795131845842,
+      "grad_norm": 1.7107497208562314,
+      "learning_rate": 1.214055765697399e-07,
+      "loss": 2.2587,
+      "step": 1535
+    },
+    {
+      "epoch": 1.5618661257606492,
+      "grad_norm": 1.7448234273922532,
+      "learning_rate": 1.197954331546911e-07,
+      "loss": 2.2493,
+      "step": 1540
+    },
+    {
+      "epoch": 1.5669371196754565,
+      "grad_norm": 1.713933005233849,
+      "learning_rate": 1.1821836454720342e-07,
+      "loss": 2.3028,
+      "step": 1545
+    },
+    {
+      "epoch": 1.5720081135902637,
+      "grad_norm": 1.8430768650069782,
+      "learning_rate": 1.1667375801964492e-07,
+      "loss": 2.2595,
+      "step": 1550
+    },
+    {
+      "epoch": 1.577079107505071,
+      "grad_norm": 1.7903141506679578,
+      "learning_rate": 1.15161010955097e-07,
+      "loss": 2.2555,
+      "step": 1555
+    },
+    {
+      "epoch": 1.5821501014198782,
+      "grad_norm": 1.810165731715535,
+      "learning_rate": 1.136795307010685e-07,
+      "loss": 2.2728,
+      "step": 1560
+    },
+    {
+      "epoch": 1.5872210953346855,
+      "grad_norm": 1.7357274884238136,
+      "learning_rate": 1.1222873442502753e-07,
+      "loss": 2.2741,
+      "step": 1565
+    },
+    {
+      "epoch": 1.592292089249493,
+      "grad_norm": 1.7545984913046129,
+      "learning_rate": 1.108080489717326e-07,
+      "loss": 2.2609,
+      "step": 1570
+    },
+    {
+      "epoch": 1.5973630831643002,
+      "grad_norm": 1.8639925458297812,
+      "learning_rate": 1.0941691072234387e-07,
+      "loss": 2.2349,
+      "step": 1575
+    },
+    {
+      "epoch": 1.6024340770791075,
+      "grad_norm": 1.7125402909483072,
+      "learning_rate": 1.080547654552963e-07,
+      "loss": 2.2929,
+      "step": 1580
+    },
+    {
+      "epoch": 1.607505070993915,
+      "grad_norm": 1.7300627575439524,
+      "learning_rate": 1.0672106820891631e-07,
+      "loss": 2.2823,
+      "step": 1585
+    },
+    {
+      "epoch": 1.6125760649087222,
+      "grad_norm": 1.7190554348875562,
+      "learning_rate": 1.0541528314576339e-07,
+      "loss": 2.2708,
+      "step": 1590
+    },
+    {
+      "epoch": 1.6176470588235294,
+      "grad_norm": 1.724918915538896,
+      "learning_rate": 1.04136883418679e-07,
+      "loss": 2.2491,
+      "step": 1595
+    },
+    {
+      "epoch": 1.6227180527383367,
+      "grad_norm": 1.7342048226287368,
+      "learning_rate": 1.0288535103852444e-07,
+      "loss": 2.2566,
+      "step": 1600
+    },
+    {
+      "epoch": 1.6227180527383367,
+      "eval_loss": 2.472487688064575,
+      "eval_runtime": 81.0795,
+      "eval_samples_per_second": 86.409,
+      "eval_steps_per_second": 0.678,
+      "step": 1600
+    },
+    {
+      "epoch": 1.627789046653144,
+      "grad_norm": 1.752725508386252,
+      "learning_rate": 1.0166017674359012e-07,
+      "loss": 2.2115,
+      "step": 1605
+    },
+    {
+      "epoch": 1.6328600405679512,
+      "grad_norm": 1.7053034674622713,
+      "learning_rate": 1.0046085987065856e-07,
+      "loss": 2.2349,
+      "step": 1610
+    },
+    {
+      "epoch": 1.6379310344827587,
+      "grad_norm": 1.6910767224745546,
+      "learning_rate": 9.928690822770361e-08,
+      "loss": 2.2661,
+      "step": 1615
+    },
+    {
+      "epoch": 1.643002028397566,
+      "grad_norm": 1.9415101732879068,
+      "learning_rate": 9.81378379682085e-08,
+      "loss": 2.2355,
+      "step": 1620
+    },
+    {
+      "epoch": 1.6480730223123732,
+      "grad_norm": 1.7692640477521646,
+      "learning_rate": 9.70131734670856e-08,
+      "loss": 2.2605,
+      "step": 1625
+    },
+    {
+      "epoch": 1.6531440162271807,
+      "grad_norm": 1.7825871200246013,
+      "learning_rate": 9.59124471981808e-08,
+      "loss": 2.2842,
+      "step": 1630
+    },
+    {
+      "epoch": 1.658215010141988,
+      "grad_norm": 1.805395258521555,
+      "learning_rate": 9.483519961334607e-08,
+      "loss": 2.2543,
+      "step": 1635
+    },
+    {
+      "epoch": 1.6632860040567952,
+      "grad_norm": 1.7151309029731219,
+      "learning_rate": 9.378097902306157e-08,
+      "loss": 2.2507,
+      "step": 1640
+    },
+    {
+      "epoch": 1.6683569979716024,
+      "grad_norm": 1.7662462146082336,
+      "learning_rate": 9.274934147859458e-08,
+      "loss": 2.2822,
+      "step": 1645
+    },
+    {
+      "epoch": 1.6734279918864097,
+      "grad_norm": 1.7065430440445857,
+      "learning_rate": 9.173985065567343e-08,
+      "loss": 2.2727,
+      "step": 1650
+    },
+    {
+      "epoch": 1.678498985801217,
+      "grad_norm": 1.8167004072102202,
+      "learning_rate": 9.075207773966592e-08,
+      "loss": 2.2582,
+      "step": 1655
+    },
+    {
+      "epoch": 1.6835699797160242,
+      "grad_norm": 1.7276973068156511,
+      "learning_rate": 8.978560131224021e-08,
+      "loss": 2.2451,
+      "step": 1660
+    },
+    {
+      "epoch": 1.6886409736308317,
+      "grad_norm": 1.7787413203893692,
+      "learning_rate": 8.88400072394981e-08,
+      "loss": 2.2421,
+      "step": 1665
+    },
+    {
+      "epoch": 1.693711967545639,
+      "grad_norm": 0.8868153668800921,
+      "learning_rate": 8.791488856155857e-08,
+      "loss": 2.2354,
+      "step": 1670
+    },
+    {
+      "epoch": 1.6987829614604464,
+      "grad_norm": 1.6998265742091707,
+      "learning_rate": 8.700984538358205e-08,
+      "loss": 2.264,
+      "step": 1675
+    },
+    {
+      "epoch": 1.7038539553752536,
+      "grad_norm": 1.7045446815412617,
+      "learning_rate": 8.612448476821393e-08,
+      "loss": 2.2775,
+      "step": 1680
+    },
+    {
+      "epoch": 1.708924949290061,
+      "grad_norm": 1.7898247009022359,
+      "learning_rate": 8.525842062943714e-08,
+      "loss": 2.2733,
+      "step": 1685
+    },
+    {
+      "epoch": 1.7139959432048681,
+      "grad_norm": 1.7604334600933766,
+      "learning_rate": 8.441127362781345e-08,
+      "loss": 2.2704,
+      "step": 1690
+    },
+    {
+      "epoch": 1.7190669371196754,
+      "grad_norm": 1.8108867949678853,
+      "learning_rate": 8.358267106710315e-08,
+      "loss": 2.2626,
+      "step": 1695
+    },
+    {
+      "epoch": 1.7241379310344827,
+      "grad_norm": 1.6881452920332736,
+      "learning_rate": 8.277224679224312e-08,
+      "loss": 2.2694,
+      "step": 1700
+    },
+    {
+      "epoch": 1.72920892494929,
+      "grad_norm": 1.7530216839199022,
+      "learning_rate": 8.197964108867328e-08,
+      "loss": 2.2622,
+      "step": 1705
+    },
+    {
+      "epoch": 1.7342799188640974,
+      "grad_norm": 1.7278497657123897,
+      "learning_rate": 8.12045005829916e-08,
+      "loss": 2.2471,
+      "step": 1710
+    },
+    {
+      "epoch": 1.7393509127789046,
+      "grad_norm": 1.8213327178561642,
+      "learning_rate": 8.044647814492792e-08,
+      "loss": 2.2313,
+      "step": 1715
+    },
+    {
+      "epoch": 1.744421906693712,
+      "grad_norm": 1.8304362576609268,
+      "learning_rate": 7.970523279061717e-08,
+      "loss": 2.2738,
+      "step": 1720
+    },
+    {
+      "epoch": 1.7494929006085194,
+      "grad_norm": 1.7718300765439339,
+      "learning_rate": 7.898042958716228e-08,
+      "loss": 2.2308,
+      "step": 1725
+    },
+    {
+      "epoch": 1.7545638945233266,
+      "grad_norm": 1.7305535723288619,
+      "learning_rate": 7.827173955846786e-08,
+      "loss": 2.2513,
+      "step": 1730
+    },
+    {
+      "epoch": 1.7596348884381339,
+      "grad_norm": 1.7402125464421778,
+      "learning_rate": 7.757883959233495e-08,
+      "loss": 2.2429,
+      "step": 1735
+    },
+    {
+      "epoch": 1.7647058823529411,
+      "grad_norm": 1.8175975710441392,
+      "learning_rate": 7.690141234879847e-08,
+      "loss": 2.288,
+      "step": 1740
+    },
+    {
+      "epoch": 1.7697768762677484,
+      "grad_norm": 1.851991292226803,
+      "learning_rate": 7.623914616969753e-08,
+      "loss": 2.2644,
+      "step": 1745
+    },
+    {
+      "epoch": 1.7748478701825556,
+      "grad_norm": 1.6602366231900278,
+      "learning_rate": 7.559173498946088e-08,
+      "loss": 2.2733,
+      "step": 1750
+    },
+    {
+      "epoch": 1.779918864097363,
+      "grad_norm": 1.7034994512549433,
+      "learning_rate": 7.495887824709769e-08,
+      "loss": 2.2674,
+      "step": 1755
+    },
+    {
+      "epoch": 1.7849898580121704,
+      "grad_norm": 1.7102833212058115,
+      "learning_rate": 7.434028079937624e-08,
+      "loss": 2.2752,
+      "step": 1760
+    },
+    {
+      "epoch": 1.7900608519269778,
+      "grad_norm": 2.1016603731428067,
+      "learning_rate": 7.373565283518085e-08,
+      "loss": 2.2726,
+      "step": 1765
+    },
+    {
+      "epoch": 1.795131845841785,
+      "grad_norm": 1.7876491597075783,
+      "learning_rate": 7.314470979103019e-08,
+      "loss": 2.2188,
+      "step": 1770
+    },
+    {
+      "epoch": 1.8002028397565923,
+      "grad_norm": 1.7984832581935817,
+      "learning_rate": 7.256717226774701e-08,
+      "loss": 2.2772,
+      "step": 1775
+    },
+    {
+      "epoch": 1.8052738336713996,
+      "grad_norm": 1.7621637378160073,
+      "learning_rate": 7.200276594826329e-08,
+      "loss": 2.2466,
+      "step": 1780
+    },
+    {
+      "epoch": 1.8103448275862069,
+      "grad_norm": 1.7255493399444854,
+      "learning_rate": 7.145122151655066e-08,
+      "loss": 2.2633,
+      "step": 1785
+    },
+    {
+      "epoch": 1.815415821501014,
+      "grad_norm": 1.7774418294615342,
+      "learning_rate": 7.101906869364121e-08,
+      "loss": 2.2966,
+      "step": 1790
+    },
+    {
+      "epoch": 1.8204868154158214,
+      "grad_norm": 1.7397631305330485,
+      "learning_rate": 7.049001264123894e-08,
+      "loss": 2.2644,
+      "step": 1795
+    },
+    {
+      "epoch": 1.8255578093306288,
+      "grad_norm": 1.7641738767791946,
+      "learning_rate": 6.997309032084255e-08,
+      "loss": 2.3052,
+      "step": 1800
+    },
+    {
+      "epoch": 1.8255578093306288,
+      "eval_loss": 2.4720866680145264,
+      "eval_runtime": 81.0596,
+      "eval_samples_per_second": 86.43,
+      "eval_steps_per_second": 0.679,
+      "step": 1800
+    },
+    {
+      "epoch": 1.830628803245436,
+      "grad_norm": 1.730995593445214,
+      "learning_rate": 6.946805070044455e-08,
+      "loss": 2.2748,
+      "step": 1805
+    },
+    {
+      "epoch": 1.8356997971602436,
+      "grad_norm": 1.708076665562477,
+      "learning_rate": 6.897464737518235e-08,
+      "loss": 2.2709,
+      "step": 1810
+    },
+    {
+      "epoch": 1.8407707910750508,
+      "grad_norm": 1.7961247246527527,
+      "learning_rate": 6.849263849253629e-08,
+      "loss": 2.2756,
+      "step": 1815
+    },
+    {
+      "epoch": 1.845841784989858,
+      "grad_norm": 1.7873259024447121,
+      "learning_rate": 6.802178667856782e-08,
+      "loss": 2.2619,
+      "step": 1820
+    },
+    {
+      "epoch": 1.8509127789046653,
+      "grad_norm": 1.7208578483390204,
+      "learning_rate": 6.756185896518329e-08,
+      "loss": 2.2563,
+      "step": 1825
+    },
+    {
+      "epoch": 1.8559837728194726,
+      "grad_norm": 1.6824119656694438,
+      "learning_rate": 6.711262671841385e-08,
+      "loss": 2.2524,
+      "step": 1830
+    },
+    {
+      "epoch": 1.8610547667342798,
+      "grad_norm": 1.717042060961093,
+      "learning_rate": 6.667386556769717e-08,
+      "loss": 2.3135,
+      "step": 1835
+    },
+    {
+      "epoch": 1.866125760649087,
+      "grad_norm": 1.736419652896857,
+      "learning_rate": 6.624535533615173e-08,
+      "loss": 2.288,
+      "step": 1840
+    },
+    {
+      "epoch": 1.8711967545638946,
+      "grad_norm": 1.75637188785577,
+      "learning_rate": 6.582687997182971e-08,
+      "loss": 2.2392,
+      "step": 1845
+    },
+    {
+      "epoch": 1.8762677484787018,
+      "grad_norm": 1.7282509939601418,
+      "learning_rate": 6.54182274799391e-08,
+      "loss": 2.2662,
+      "step": 1850
+    },
+    {
+      "epoch": 1.8813387423935093,
+      "grad_norm": 1.7060962855685544,
+      "learning_rate": 6.501918985602177e-08,
+      "loss": 2.2935,
+      "step": 1855
+    },
+    {
+      "epoch": 1.8864097363083165,
+      "grad_norm": 1.7581616823404618,
+      "learning_rate": 6.462956302007797e-08,
+      "loss": 2.2478,
+      "step": 1860
+    },
+    {
+      "epoch": 1.8914807302231238,
+      "grad_norm": 1.7987997676993257,
+      "learning_rate": 6.424914675162432e-08,
+      "loss": 2.2853,
+      "step": 1865
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 1.7116689993633696,
+      "learning_rate": 6.387774462567602e-08,
+      "loss": 2.2503,
+      "step": 1870
+    },
+    {
+      "epoch": 1.9016227180527383,
+      "grad_norm": 1.7086258587789072,
+      "learning_rate": 6.351516394964051e-08,
+      "loss": 2.2822,
+      "step": 1875
+    },
+    {
+      "epoch": 1.9066937119675456,
+      "grad_norm": 1.8235148496074345,
+      "learning_rate": 6.31612157011135e-08,
+      "loss": 2.2879,
+      "step": 1880
+    },
+    {
+      "epoch": 1.9117647058823528,
+      "grad_norm": 1.7448709638927917,
+      "learning_rate": 6.281571446656485e-08,
+      "loss": 2.2586,
+      "step": 1885
+    },
+    {
+      "epoch": 1.9168356997971603,
+      "grad_norm": 1.7421662505581106,
+      "learning_rate": 6.247847838090545e-08,
+      "loss": 2.2791,
+      "step": 1890
+    },
+    {
+      "epoch": 1.9219066937119675,
+      "grad_norm": 1.825830026911039,
+      "learning_rate": 6.21493290679226e-08,
+      "loss": 2.2385,
+      "step": 1895
+    },
+    {
+      "epoch": 1.9269776876267748,
+      "grad_norm": 1.796187481606512,
+      "learning_rate": 6.182809158157558e-08,
+      "loss": 2.2756,
+      "step": 1900
+    },
+    {
+      "epoch": 1.9320486815415823,
+      "grad_norm": 1.7552941496595575,
+      "learning_rate": 6.151459434813879e-08,
+      "loss": 2.2587,
+      "step": 1905
+    },
+    {
+      "epoch": 1.9371196754563895,
+      "grad_norm": 1.7522494947057408,
+      "learning_rate": 6.120866910918446e-08,
+      "loss": 2.2585,
+      "step": 1910
+    },
+    {
+      "epoch": 1.9421906693711968,
+      "grad_norm": 1.7522459962159465,
+      "learning_rate": 6.091015086539273e-08,
+      "loss": 2.251,
+      "step": 1915
+    },
+    {
+      "epoch": 1.947261663286004,
+      "grad_norm": 1.702096284758162,
+      "learning_rate": 6.061887782118077e-08,
+      "loss": 2.285,
+      "step": 1920
+    },
+    {
+      "epoch": 1.9523326572008113,
+      "grad_norm": 1.7643281133012019,
+      "learning_rate": 6.033469133013957e-08,
+      "loss": 2.2846,
+      "step": 1925
+    },
+    {
+      "epoch": 1.9574036511156185,
+      "grad_norm": 1.6926355627529537,
+      "learning_rate": 6.005743584126981e-08,
+      "loss": 2.2124,
+      "step": 1930
+    },
+    {
+      "epoch": 1.962474645030426,
+      "grad_norm": 1.6991484085258466,
+      "learning_rate": 5.984051918509233e-08,
+      "loss": 2.2919,
+      "step": 1935
+    },
+    {
+      "epoch": 1.9675456389452333,
+      "grad_norm": 1.6959402402475394,
+      "learning_rate": 5.957535718971899e-08,
+      "loss": 2.2133,
+      "step": 1940
+    },
+    {
+      "epoch": 1.9726166328600405,
+      "grad_norm": 1.7435422008262311,
+      "learning_rate": 5.931670667334593e-08,
+      "loss": 2.2272,
+      "step": 1945
+    },
+    {
+      "epoch": 1.977687626774848,
+      "grad_norm": 1.7235339509485863,
+      "learning_rate": 5.906442337098544e-08,
+      "loss": 2.2566,
+      "step": 1950
+    },
+    {
+      "epoch": 1.9827586206896552,
+      "grad_norm": 1.8046591422600013,
+      "learning_rate": 5.881836586579961e-08,
+      "loss": 2.295,
+      "step": 1955
+    },
+    {
+      "epoch": 1.9878296146044625,
+      "grad_norm": 1.8447312096680564,
+      "learning_rate": 5.8578395539777033e-08,
+      "loss": 2.29,
+      "step": 1960
+    },
+    {
+      "epoch": 1.9929006085192698,
+      "grad_norm": 1.6943108398877464,
+      "learning_rate": 5.834437652514426e-08,
+      "loss": 2.2188,
+      "step": 1965
+    },
+    {
+      "epoch": 1.997971602434077,
+      "grad_norm": 1.7174652428188777,
+      "learning_rate": 5.811617565650129e-08,
+      "loss": 2.2692,
+      "step": 1970
+    },
+    {
+      "epoch": 2.0030425963488843,
+      "grad_norm": 1.6831299340128894,
+      "learning_rate": 5.7893662423673665e-08,
+      "loss": 2.2025,
+      "step": 1975
+    },
+    {
+      "epoch": 2.0081135902636915,
+      "grad_norm": 1.826795197065323,
+      "learning_rate": 5.767670892527061e-08,
+      "loss": 2.2579,
+      "step": 1980
+    },
+    {
+      "epoch": 2.0131845841784988,
+      "grad_norm": 1.7520235012361185,
+      "learning_rate": 5.746518982294192e-08,
+      "loss": 2.2388,
+      "step": 1985
+    },
+    {
+      "epoch": 2.0182555780933065,
+      "grad_norm": 1.8440219249964744,
+      "learning_rate": 5.72589822963234e-08,
+      "loss": 2.2582,
+      "step": 1990
+    },
+    {
+      "epoch": 2.0233265720081137,
+      "grad_norm": 1.7151060194819,
+      "learning_rate": 5.705796599866345e-08,
+      "loss": 2.2156,
+      "step": 1995
+    },
+    {
+      "epoch": 2.028397565922921,
+      "grad_norm": 1.7333738899068507,
+      "learning_rate": 5.686202301312118e-08,
+      "loss": 2.2702,
+      "step": 2000
+    },
+    {
+      "epoch": 2.028397565922921,
+      "eval_loss": 2.4733877182006836,
+      "eval_runtime": 81.1205,
+      "eval_samples_per_second": 86.365,
+      "eval_steps_per_second": 0.678,
+      "step": 2000
+    },
+    {
+      "epoch": 2.0334685598377282,
+      "grad_norm": 1.7637474983877708,
+      "learning_rate": 5.667103780972823e-08,
+      "loss": 2.2378,
+      "step": 2005
+    },
+    {
+      "epoch": 2.0385395537525355,
+      "grad_norm": 1.7730571315134518,
+      "learning_rate": 5.648489720300554e-08,
+      "loss": 2.2513,
+      "step": 2010
+    },
+    {
+      "epoch": 2.0436105476673427,
+      "grad_norm": 1.774271074894755,
+      "learning_rate": 5.630349031022691e-08,
+      "loss": 2.2518,
+      "step": 2015
+    },
+    {
+      "epoch": 2.04868154158215,
+      "grad_norm": 1.6997020509374097,
+      "learning_rate": 5.6126708510320976e-08,
+      "loss": 2.2464,
+      "step": 2020
+    },
+    {
+      "epoch": 2.0537525354969572,
+      "grad_norm": 1.7833382557650153,
+      "learning_rate": 5.595444540340353e-08,
+      "loss": 2.2317,
+      "step": 2025
+    },
+    {
+      "epoch": 2.0588235294117645,
+      "grad_norm": 1.7296871432561252,
+      "learning_rate": 5.578659677093205e-08,
+      "loss": 2.231,
+      "step": 2030
+    },
+    {
+      "epoch": 2.063894523326572,
+      "grad_norm": 1.7166463945290173,
+      "learning_rate": 5.562306053647459e-08,
+      "loss": 2.2347,
+      "step": 2035
+    },
+    {
+      "epoch": 2.0689655172413794,
+      "grad_norm": 1.7948324654757548,
+      "learning_rate": 5.546373672708482e-08,
+      "loss": 2.2458,
+      "step": 2040
+    },
+    {
+      "epoch": 2.0740365111561867,
+      "grad_norm": 1.745646645076283,
+      "learning_rate": 5.530852743527571e-08,
+      "loss": 2.2504,
+      "step": 2045
+    },
+    {
+      "epoch": 2.079107505070994,
+      "grad_norm": 1.7778201657756552,
+      "learning_rate": 5.515733678158393e-08,
+      "loss": 2.26,
+      "step": 2050
+    },
+    {
+      "epoch": 2.084178498985801,
+      "grad_norm": 1.7226724662159607,
+      "learning_rate": 5.5010070877717374e-08,
+      "loss": 2.24,
+      "step": 2055
+    },
+    {
+      "epoch": 2.0892494929006085,
+      "grad_norm": 1.737085412071484,
+      "learning_rate": 5.486663779027808e-08,
+      "loss": 2.2138,
+      "step": 2060
+    },
+    {
+      "epoch": 2.0943204868154157,
+      "grad_norm": 1.7680067007098665,
+      "learning_rate": 5.4726947505053265e-08,
+      "loss": 2.2688,
+      "step": 2065
+    },
+    {
+      "epoch": 2.099391480730223,
+      "grad_norm": 1.7414742255329991,
+      "learning_rate": 5.459091189186688e-08,
+      "loss": 2.2591,
+      "step": 2070
+    },
+    {
+      "epoch": 2.1044624746450302,
+      "grad_norm": 1.7804223600059563,
+      "learning_rate": 5.4458444669984314e-08,
+      "loss": 2.2337,
+      "step": 2075
+    },
+    {
+      "epoch": 2.109533468559838,
+      "grad_norm": 1.7481822321590552,
+      "learning_rate": 5.432946137406314e-08,
+      "loss": 2.2792,
+      "step": 2080
+    },
+    {
+      "epoch": 2.114604462474645,
+      "grad_norm": 1.7497391573214505,
+      "learning_rate": 5.420387932064249e-08,
+      "loss": 2.2927,
+      "step": 2085
+    },
+    {
+      "epoch": 2.1196754563894524,
+      "grad_norm": 1.7279168540890797,
+      "learning_rate": 5.408161757516413e-08,
+      "loss": 2.2451,
+      "step": 2090
+    },
+    {
+      "epoch": 2.1247464503042597,
+      "grad_norm": 1.7394662730899328,
+      "learning_rate": 5.396259691951805e-08,
+      "loss": 2.2424,
+      "step": 2095
+    },
+    {
+      "epoch": 2.129817444219067,
+      "grad_norm": 1.77875077601377,
+      "learning_rate": 5.384673982010568e-08,
+      "loss": 2.2402,
+      "step": 2100
+    },
+    {
+      "epoch": 2.134888438133874,
+      "grad_norm": 1.7319261658863345,
+      "learning_rate": 5.373397039641377e-08,
+      "loss": 2.2287,
+      "step": 2105
+    },
+    {
+      "epoch": 2.1399594320486814,
+      "grad_norm": 1.751571162082358,
+      "learning_rate": 5.362421439009217e-08,
+      "loss": 2.2334,
+      "step": 2110
+    },
+    {
+      "epoch": 2.1450304259634887,
+      "grad_norm": 1.8093044605440316,
+      "learning_rate": 5.351739913452874e-08,
+      "loss": 2.271,
+      "step": 2115
+    },
+    {
+      "epoch": 2.150101419878296,
+      "grad_norm": 1.8469881188013633,
+      "learning_rate": 5.341345352491468e-08,
+      "loss": 2.2284,
+      "step": 2120
+    },
+    {
+      "epoch": 2.1551724137931036,
+      "grad_norm": 1.7711139740473771,
+      "learning_rate": 5.331230798879373e-08,
+      "loss": 2.2644,
+      "step": 2125
+    },
+    {
+      "epoch": 2.160243407707911,
+      "grad_norm": 1.7271859975777568,
+      "learning_rate": 5.3213894457088646e-08,
+      "loss": 2.2378,
+      "step": 2130
+    },
+    {
+      "epoch": 2.165314401622718,
+      "grad_norm": 1.8925272013685321,
+      "learning_rate": 5.3118146335598536e-08,
+      "loss": 2.265,
+      "step": 2135
+    },
+    {
+      "epoch": 2.1703853955375254,
+      "grad_norm": 1.7527393142771752,
+      "learning_rate": 5.3024998476960626e-08,
+      "loss": 2.2183,
+      "step": 2140
+    },
+    {
+      "epoch": 2.1754563894523327,
+      "grad_norm": 1.7698628867396988,
+      "learning_rate": 5.293438715307019e-08,
+      "loss": 2.233,
+      "step": 2145
+    },
+    {
+      "epoch": 2.18052738336714,
+      "grad_norm": 1.724950058777004,
+      "learning_rate": 5.2846250027952295e-08,
+      "loss": 2.249,
+      "step": 2150
+    },
+    {
+      "epoch": 2.185598377281947,
+      "grad_norm": 1.9072718835854334,
+      "learning_rate": 5.276052613107927e-08,
+      "loss": 2.2342,
+      "step": 2155
+    },
+    {
+      "epoch": 2.1906693711967544,
+      "grad_norm": 1.7983471937343785,
+      "learning_rate": 5.2677155831127696e-08,
+      "loss": 2.2707,
+      "step": 2160
+    },
+    {
+      "epoch": 2.1957403651115617,
+      "grad_norm": 1.7092533410568467,
+      "learning_rate": 5.259608081016899e-08,
+      "loss": 2.2479,
+      "step": 2165
+    },
+    {
+      "epoch": 2.2008113590263694,
+      "grad_norm": 1.7921254707864127,
+      "learning_rate": 5.2517244038287416e-08,
+      "loss": 2.229,
+      "step": 2170
+    },
+    {
+      "epoch": 2.2058823529411766,
+      "grad_norm": 1.75489401951672,
+      "learning_rate": 5.244058974861976e-08,
+      "loss": 2.2772,
+      "step": 2175
+    },
+    {
+      "epoch": 2.210953346855984,
+      "grad_norm": 1.8175479517709452,
+      "learning_rate": 5.236606341281078e-08,
+      "loss": 2.2356,
+      "step": 2180
+    },
+    {
+      "epoch": 2.216024340770791,
+      "grad_norm": 1.808556074117745,
+      "learning_rate": 5.229361171687859e-08,
+      "loss": 2.2553,
+      "step": 2185
+    },
+    {
+      "epoch": 2.2210953346855984,
+      "grad_norm": 1.7664667006627157,
+      "learning_rate": 5.2223182537484316e-08,
+      "loss": 2.2719,
+      "step": 2190
+    },
+    {
+      "epoch": 2.2261663286004056,
+      "grad_norm": 1.7502392717778497,
+      "learning_rate": 5.2154724918600314e-08,
+      "loss": 2.2583,
+      "step": 2195
+    },
+    {
+      "epoch": 2.231237322515213,
+      "grad_norm": 1.7242967584463027,
+      "learning_rate": 5.208818904857144e-08,
+      "loss": 2.2411,
+      "step": 2200
+    },
+    {
+      "epoch": 2.231237322515213,
+      "eval_loss": 2.474597930908203,
+      "eval_runtime": 81.0438,
+      "eval_samples_per_second": 86.447,
+      "eval_steps_per_second": 0.679,
+      "step": 2200
+    },
+    {
+      "epoch": 2.23630831643002,
+      "grad_norm": 1.760326712726159,
+      "learning_rate": 5.202352623756371e-08,
+      "loss": 2.2356,
+      "step": 2205
+    },
+    {
+      "epoch": 2.2413793103448274,
+      "grad_norm": 1.7625638663030738,
+      "learning_rate": 5.1960688895395006e-08,
+      "loss": 2.2441,
+      "step": 2210
+    },
+    {
+      "epoch": 2.2464503042596347,
+      "grad_norm": 1.7518142596486186,
+      "learning_rate": 5.189963050974238e-08,
+      "loss": 2.2674,
+      "step": 2215
+    },
+    {
+      "epoch": 2.2515212981744424,
+      "grad_norm": 1.8040378121090448,
+      "learning_rate": 5.184030562472053e-08,
+      "loss": 2.2233,
+      "step": 2220
+    },
+    {
+      "epoch": 2.2565922920892496,
+      "grad_norm": 1.769147010660197,
+      "learning_rate": 5.1782669819826294e-08,
+      "loss": 2.2445,
+      "step": 2225
+    },
+    {
+      "epoch": 2.261663286004057,
+      "grad_norm": 1.802360281392845,
+      "learning_rate": 5.1726679689243875e-08,
+      "loss": 2.234,
+      "step": 2230
+    },
+    {
+      "epoch": 2.266734279918864,
+      "grad_norm": 1.763707867667644,
+      "learning_rate": 5.1672292821505586e-08,
+      "loss": 2.2132,
+      "step": 2235
+    },
+    {
+      "epoch": 2.2718052738336714,
+      "grad_norm": 1.75034581686763,
+      "learning_rate": 5.161946777950308e-08,
+      "loss": 2.2381,
+      "step": 2240
+    },
+    {
+      "epoch": 2.2768762677484786,
+      "grad_norm": 1.7401836199474783,
+      "learning_rate": 5.1568164080844036e-08,
+      "loss": 2.2416,
+      "step": 2245
+    },
+    {
+      "epoch": 2.281947261663286,
+      "grad_norm": 1.7713650977668527,
+      "learning_rate": 5.1518342178549174e-08,
+      "loss": 2.224,
+      "step": 2250
+    },
+    {
+      "epoch": 2.287018255578093,
+      "grad_norm": 1.7671231076913356,
+      "learning_rate": 5.146996344208486e-08,
+      "loss": 2.2183,
+      "step": 2255
+    },
+    {
+      "epoch": 2.292089249492901,
+      "grad_norm": 1.7464419032652747,
+      "learning_rate": 5.142299013872629e-08,
+      "loss": 2.2419,
+      "step": 2260
+    },
+    {
+      "epoch": 2.297160243407708,
+      "grad_norm": 1.7990294085116565,
+      "learning_rate": 5.1377385415246445e-08,
+      "loss": 2.2311,
+      "step": 2265
+    },
+    {
+      "epoch": 2.3022312373225153,
+      "grad_norm": 1.7543351264072877,
+      "learning_rate": 5.1333113279926185e-08,
+      "loss": 2.238,
+      "step": 2270
+    },
+    {
+      "epoch": 2.3073022312373226,
+      "grad_norm": 1.6898279670163325,
+      "learning_rate": 5.129013858488057e-08,
+      "loss": 2.2308,
+      "step": 2275
+    },
+    {
+      "epoch": 2.31237322515213,
+      "grad_norm": 1.7334567047607963,
+      "learning_rate": 5.124842700869695e-08,
+      "loss": 2.3031,
+      "step": 2280
+    },
+    {
+      "epoch": 2.317444219066937,
+      "grad_norm": 1.760983319309442,
+      "learning_rate": 5.120794503938012e-08,
+      "loss": 2.2455,
+      "step": 2285
+    },
+    {
+      "epoch": 2.3225152129817443,
+      "grad_norm": 1.7621675205518297,
+      "learning_rate": 5.116865995760006e-08,
+      "loss": 2.228,
+      "step": 2290
+    },
+    {
+      "epoch": 2.3275862068965516,
+      "grad_norm": 1.8080633887862172,
+      "learning_rate": 5.113053982023768e-08,
+      "loss": 2.284,
+      "step": 2295
+    },
+    {
+      "epoch": 2.332657200811359,
+      "grad_norm": 1.7592998081055247,
+      "learning_rate": 5.1093553444224286e-08,
+      "loss": 2.2196,
+      "step": 2300
+    },
+    {
+      "epoch": 2.337728194726166,
+      "grad_norm": 1.7831607571885368,
+      "learning_rate": 5.105767039067024e-08,
+      "loss": 2.269,
+      "step": 2305
+    },
+    {
+      "epoch": 2.342799188640974,
+      "grad_norm": 1.7176459519033709,
+      "learning_rate": 5.102286094927856e-08,
+      "loss": 2.2435,
+      "step": 2310
+    },
+    {
+      "epoch": 2.347870182555781,
+      "grad_norm": 1.7512756209003166,
+      "learning_rate": 5.098909612303925e-08,
+      "loss": 2.2579,
+      "step": 2315
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 1.7419259056225642,
+      "learning_rate": 5.095634761319991e-08,
+      "loss": 2.268,
+      "step": 2320
+    },
+    {
+      "epoch": 2.3580121703853956,
+      "grad_norm": 1.7461469979215953,
+      "learning_rate": 5.092458780450876e-08,
+      "loss": 2.2252,
+      "step": 2325
+    },
+    {
+      "epoch": 2.363083164300203,
+      "grad_norm": 1.745083473021831,
+      "learning_rate": 5.089378975072569e-08,
+      "loss": 2.2591,
+      "step": 2330
+    },
+    {
+      "epoch": 2.36815415821501,
+      "grad_norm": 1.8343705825023535,
+      "learning_rate": 5.086392716039744e-08,
+      "loss": 2.2626,
+      "step": 2335
+    },
+    {
+      "epoch": 2.3732251521298173,
+      "grad_norm": 1.7515682941502182,
+      "learning_rate": 5.0834974382892763e-08,
+      "loss": 2.2378,
+      "step": 2340
+    },
+    {
+      "epoch": 2.3782961460446246,
+      "grad_norm": 1.772483228062822,
+      "learning_rate": 5.080690639469371e-08,
+      "loss": 2.2906,
+      "step": 2345
+    },
+    {
+      "epoch": 2.3833671399594323,
+      "grad_norm": 1.8298309311035177,
+      "learning_rate": 5.077969878593903e-08,
+      "loss": 2.2782,
+      "step": 2350
+    },
+    {
+      "epoch": 2.3884381338742395,
+      "grad_norm": 1.778228901931638,
+      "learning_rate": 5.0753327747215805e-08,
+      "loss": 2.2687,
+      "step": 2355
+    },
+    {
+      "epoch": 2.393509127789047,
+      "grad_norm": 1.9355725485663295,
+      "learning_rate": 5.0727770056595594e-08,
+      "loss": 2.25,
+      "step": 2360
+    },
+    {
+      "epoch": 2.398580121703854,
+      "grad_norm": 1.7876677525732199,
+      "learning_rate": 5.070300306691114e-08,
+      "loss": 2.2811,
+      "step": 2365
+    },
+    {
+      "epoch": 2.4036511156186613,
+      "grad_norm": 1.766450812020173,
+      "learning_rate": 5.067900469327011e-08,
+      "loss": 2.265,
+      "step": 2370
+    },
+    {
+      "epoch": 2.4087221095334685,
+      "grad_norm": 1.6988211316677768,
+      "learning_rate": 5.065575340080193e-08,
+      "loss": 2.2458,
+      "step": 2375
+    },
+    {
+      "epoch": 2.413793103448276,
+      "grad_norm": 1.777565241311822,
+      "learning_rate": 5.063322819263436e-08,
+      "loss": 2.289,
+      "step": 2380
+    },
+    {
+      "epoch": 2.418864097363083,
+      "grad_norm": 1.766648317811343,
+      "learning_rate": 5.061140859809592e-08,
+      "loss": 2.2263,
+      "step": 2385
+    },
+    {
+      "epoch": 2.4239350912778903,
+      "grad_norm": 1.760808570512941,
+      "learning_rate": 5.059027466114087e-08,
+      "loss": 2.2371,
+      "step": 2390
+    },
+    {
+      "epoch": 2.4290060851926976,
+      "grad_norm": 1.7497881623660254,
+      "learning_rate": 5.056980692899308e-08,
+      "loss": 2.2186,
+      "step": 2395
+    },
+    {
+      "epoch": 2.4340770791075053,
+      "grad_norm": 1.904368651484495,
+      "learning_rate": 5.0549986441005356e-08,
+      "loss": 2.2413,
+      "step": 2400
+    },
+    {
+      "epoch": 2.4340770791075053,
+      "eval_loss": 2.4748759269714355,
+      "eval_runtime": 81.0832,
+      "eval_samples_per_second": 86.405,
+      "eval_steps_per_second": 0.678,
+      "step": 2400
+    },
+    {
+      "epoch": 2.4391480730223125,
+      "grad_norm": 1.7410363640013542,
+      "learning_rate": 5.053079471773089e-08,
+      "loss": 2.2531,
+      "step": 2405
+    },
+    {
+      "epoch": 2.4442190669371198,
+      "grad_norm": 1.7518018775000213,
+      "learning_rate": 5.0512213750203305e-08,
+      "loss": 2.2473,
+      "step": 2410
+    },
+    {
+      "epoch": 2.449290060851927,
+      "grad_norm": 1.7662222396602074,
+      "learning_rate": 5.049422598942212e-08,
+      "loss": 2.2389,
+      "step": 2415
+    },
+    {
+      "epoch": 2.4543610547667343,
+      "grad_norm": 1.780666367007688,
+      "learning_rate": 5.0476814336040274e-08,
+      "loss": 2.197,
+      "step": 2420
+    },
+    {
+      "epoch": 2.4594320486815415,
+      "grad_norm": 1.7499711395815145,
+      "learning_rate": 5.04599621302504e-08,
+      "loss": 2.2261,
+      "step": 2425
+    },
+    {
+      "epoch": 2.464503042596349,
+      "grad_norm": 1.7882713122146334,
+      "learning_rate": 5.04436531418668e-08,
+      "loss": 2.2393,
+      "step": 2430
+    },
+    {
+      "epoch": 2.469574036511156,
+      "grad_norm": 1.75643986036064,
+      "learning_rate": 5.042787156059982e-08,
+      "loss": 2.2439,
+      "step": 2435
+    },
+    {
+      "epoch": 2.4746450304259637,
+      "grad_norm": 1.7353199942499,
+      "learning_rate": 5.041260198651953e-08,
+      "loss": 2.2275,
+      "step": 2440
+    },
+    {
+      "epoch": 2.479716024340771,
+      "grad_norm": 1.7683236873580634,
+      "learning_rate": 5.039782942070575e-08,
+      "loss": 2.2378,
+      "step": 2445
+    },
+    {
+      "epoch": 2.4847870182555782,
+      "grad_norm": 1.7482878827223234,
+      "learning_rate": 5.038353925608112e-08,
+      "loss": 2.2655,
+      "step": 2450
+    },
+    {
+      "epoch": 2.4898580121703855,
+      "grad_norm": 1.7553465772492238,
+      "learning_rate": 5.036971726842454e-08,
+      "loss": 2.2509,
+      "step": 2455
+    },
+    {
+      "epoch": 2.4949290060851927,
+      "grad_norm": 1.7194051175937297,
+      "learning_rate": 5.035634960756173e-08,
+      "loss": 2.2246,
+      "step": 2460
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.780820717878673,
+      "learning_rate": 5.0345973520341744e-08,
+      "loss": 2.3116,
+      "step": 2465
+    },
+    {
+      "epoch": 2.5050709939148073,
+      "grad_norm": 1.7092302368812895,
+      "learning_rate": 5.0333389906255366e-08,
+      "loss": 2.2434,
+      "step": 2470
+    },
+    {
+      "epoch": 2.5101419878296145,
+      "grad_norm": 1.6995993050400164,
+      "learning_rate": 5.03212237555571e-08,
+      "loss": 2.234,
+      "step": 2475
+    },
+    {
+      "epoch": 2.5152129817444218,
+      "grad_norm": 1.7916125090755124,
+      "learning_rate": 5.030946256214713e-08,
+      "loss": 2.2365,
+      "step": 2480
+    },
+    {
+      "epoch": 2.520283975659229,
+      "grad_norm": 1.743409123646943,
+      "learning_rate": 5.0298094154063516e-08,
+      "loss": 2.2778,
+      "step": 2485
+    },
+    {
+      "epoch": 2.5253549695740363,
+      "grad_norm": 1.7989761193864806,
+      "learning_rate": 5.028710668564437e-08,
+      "loss": 2.2698,
+      "step": 2490
+    },
+    {
+      "epoch": 2.530425963488844,
+      "grad_norm": 1.768436463277154,
+      "learning_rate": 5.027648862984817e-08,
+      "loss": 2.2295,
+      "step": 2495
+    },
+    {
+      "epoch": 2.535496957403651,
+      "grad_norm": 1.7762161444449078,
+      "learning_rate": 5.026622877072948e-08,
+      "loss": 2.2772,
+      "step": 2500
+    },
+    {
+      "epoch": 2.5405679513184585,
+      "grad_norm": 1.7325943514517332,
+      "learning_rate": 5.0256316196067565e-08,
+      "loss": 2.2326,
+      "step": 2505
+    },
+    {
+      "epoch": 2.5456389452332657,
+      "grad_norm": 1.7568007182157335,
+      "learning_rate": 5.024674029014512e-08,
+      "loss": 2.2575,
+      "step": 2510
+    },
+    {
+      "epoch": 2.550709939148073,
+      "grad_norm": 1.7465474101311085,
+      "learning_rate": 5.023749072667476e-08,
+      "loss": 2.2398,
+      "step": 2515
+    },
+    {
+      "epoch": 2.5557809330628802,
+      "grad_norm": 1.7105972624166814,
+      "learning_rate": 5.022855746187064e-08,
+      "loss": 2.2348,
+      "step": 2520
+    },
+    {
+      "epoch": 2.5608519269776875,
+      "grad_norm": 1.759196327867933,
+      "learning_rate": 5.021993072766265e-08,
+      "loss": 2.2302,
+      "step": 2525
+    },
+    {
+      "epoch": 2.565922920892495,
+      "grad_norm": 1.7618696598564434,
+      "learning_rate": 5.0211601025050875e-08,
+      "loss": 2.2783,
+      "step": 2530
+    },
+    {
+      "epoch": 2.5709939148073024,
+      "grad_norm": 1.7357397604845723,
+      "learning_rate": 5.020355911759782e-08,
+      "loss": 2.2399,
+      "step": 2535
+    },
+    {
+      "epoch": 2.5760649087221097,
+      "grad_norm": 1.7797963559349856,
+      "learning_rate": 5.019579602505595e-08,
+      "loss": 2.3119,
+      "step": 2540
+    },
+    {
+      "epoch": 2.581135902636917,
+      "grad_norm": 1.7476476267237637,
+      "learning_rate": 5.0188303017128396e-08,
+      "loss": 2.2362,
+      "step": 2545
+    },
+    {
+      "epoch": 2.586206896551724,
+      "grad_norm": 1.7871655712678034,
+      "learning_rate": 5.018107160736018e-08,
+      "loss": 2.2684,
+      "step": 2550
+    },
+    {
+      "epoch": 2.5912778904665315,
+      "grad_norm": 1.8564365985849263,
+      "learning_rate": 5.0174093547158035e-08,
+      "loss": 2.2683,
+      "step": 2555
+    },
+    {
+      "epoch": 2.5963488843813387,
+      "grad_norm": 1.7498854511370805,
+      "learning_rate": 5.016736081993624e-08,
+      "loss": 2.2518,
+      "step": 2560
+    },
+    {
+      "epoch": 2.601419878296146,
+      "grad_norm": 1.7966977533010748,
+      "learning_rate": 5.016086563538651e-08,
+      "loss": 2.2218,
+      "step": 2565
+    },
+    {
+      "epoch": 2.606490872210953,
+      "grad_norm": 1.7558979371137615,
+      "learning_rate": 5.015460042386951e-08,
+      "loss": 2.2658,
+      "step": 2570
+    },
+    {
+      "epoch": 2.6115618661257605,
+      "grad_norm": 1.7805268954368878,
+      "learning_rate": 5.014855783092602e-08,
+      "loss": 2.2324,
+      "step": 2575
+    },
+    {
+      "epoch": 2.6166328600405677,
+      "grad_norm": 1.7547744035144406,
+      "learning_rate": 5.0142730711905564e-08,
+      "loss": 2.2635,
+      "step": 2580
+    },
+    {
+      "epoch": 2.6217038539553754,
+      "grad_norm": 1.7892043381738651,
+      "learning_rate": 5.013711212671024e-08,
+      "loss": 2.2174,
+      "step": 2585
+    },
+    {
+      "epoch": 2.6267748478701827,
+      "grad_norm": 1.7661048483256172,
+      "learning_rate": 5.013169533465201e-08,
+      "loss": 2.2411,
+      "step": 2590
+    },
+    {
+      "epoch": 2.63184584178499,
+      "grad_norm": 1.7714992602824393,
+      "learning_rate": 5.012647378942108e-08,
+      "loss": 2.2379,
+      "step": 2595
+    },
+    {
+      "epoch": 2.636916835699797,
+      "grad_norm": 1.757980523509378,
+      "learning_rate": 5.0121441134163554e-08,
+      "loss": 2.216,
+      "step": 2600
+    },
+    {
+      "epoch": 2.636916835699797,
+      "eval_loss": 2.4749209880828857,
+      "eval_runtime": 81.0391,
+      "eval_samples_per_second": 86.452,
+      "eval_steps_per_second": 0.679,
+      "step": 2600
+    },
+    {
+      "epoch": 2.6419878296146044,
+      "grad_norm": 1.8485215916583273,
+      "learning_rate": 5.011659119666631e-08,
+      "loss": 2.2233,
+      "step": 2605
+    },
+    {
+      "epoch": 2.6470588235294117,
+      "grad_norm": 1.7863067371305124,
+      "learning_rate": 5.0111917984647157e-08,
+      "loss": 2.244,
+      "step": 2610
+    },
+    {
+      "epoch": 2.652129817444219,
+      "grad_norm": 1.7146816358296353,
+      "learning_rate": 5.010741568114834e-08,
+      "loss": 2.2351,
+      "step": 2615
+    },
+    {
+      "epoch": 2.6572008113590266,
+      "grad_norm": 1.831188399230356,
+      "learning_rate": 5.0103078640031516e-08,
+      "loss": 2.2269,
+      "step": 2620
+    },
+    {
+      "epoch": 2.662271805273834,
+      "grad_norm": 1.7724728214531387,
+      "learning_rate": 5.009890138157231e-08,
+      "loss": 2.2075,
+      "step": 2625
+    },
+    {
+      "epoch": 2.667342799188641,
+      "grad_norm": 1.782021890238949,
+      "learning_rate": 5.009487858815262e-08,
+      "loss": 2.217,
+      "step": 2630
+    },
+    {
+      "epoch": 2.6724137931034484,
+      "grad_norm": 1.7481328251498853,
+      "learning_rate": 5.0091005100048845e-08,
+      "loss": 2.2719,
+      "step": 2635
+    },
+    {
+      "epoch": 2.6774847870182557,
+      "grad_norm": 1.7906104909059064,
+      "learning_rate": 5.0087275911314286e-08,
+      "loss": 2.236,
+      "step": 2640
+    },
+    {
+      "epoch": 2.682555780933063,
+      "grad_norm": 1.7602535674283515,
+      "learning_rate": 5.008368616575389e-08,
+      "loss": 2.2479,
+      "step": 2645
+    },
+    {
+      "epoch": 2.68762677484787,
+      "grad_norm": 1.775336072092801,
+      "learning_rate": 5.00802311529897e-08,
+      "loss": 2.2651,
+      "step": 2650
+    },
+    {
+      "epoch": 2.6926977687626774,
+      "grad_norm": 1.7553544981528668,
+      "learning_rate": 5.00769063046152e-08,
+      "loss": 2.2695,
+      "step": 2655
+    },
+    {
+      "epoch": 2.6977687626774847,
+      "grad_norm": 1.827043155040219,
+      "learning_rate": 5.0073707190436947e-08,
+      "loss": 2.2565,
+      "step": 2660
+    },
+    {
+      "epoch": 2.702839756592292,
+      "grad_norm": 1.7286161050152862,
+      "learning_rate": 5.00706295148018e-08,
+      "loss": 2.2447,
+      "step": 2665
+    },
+    {
+      "epoch": 2.707910750507099,
+      "grad_norm": 1.818175461042268,
+      "learning_rate": 5.0067669113008144e-08,
+      "loss": 2.2437,
+      "step": 2670
+    },
+    {
+      "epoch": 2.7129817444219064,
+      "grad_norm": 1.8017061603291116,
+      "learning_rate": 5.006482194779946e-08,
+      "loss": 2.2557,
+      "step": 2675
+    },
+    {
+      "epoch": 2.718052738336714,
+      "grad_norm": 1.7866064039916518,
+      "learning_rate": 5.006208410593867e-08,
+      "loss": 2.2752,
+      "step": 2680
+    },
+    {
+      "epoch": 2.7231237322515214,
+      "grad_norm": 1.7655940160674672,
+      "learning_rate": 5.0059451794861766e-08,
+      "loss": 2.2834,
+      "step": 2685
+    },
+    {
+      "epoch": 2.7281947261663286,
+      "grad_norm": 1.7936324116014108,
+      "learning_rate": 5.005692133940906e-08,
+      "loss": 2.2634,
+      "step": 2690
+    },
+    {
+      "epoch": 2.733265720081136,
+      "grad_norm": 1.7857563825463283,
+      "learning_rate": 5.00544891786327e-08,
+      "loss": 2.2741,
+      "step": 2695
+    },
+    {
+      "epoch": 2.738336713995943,
+      "grad_norm": 1.7472045814339527,
+      "learning_rate": 5.005215186267882e-08,
+      "loss": 2.2644,
+      "step": 2700
+    },
+    {
+      "epoch": 2.7434077079107504,
+      "grad_norm": 1.8795177703424921,
+      "learning_rate": 5.0049906049743e-08,
+      "loss": 2.3007,
+      "step": 2705
+    },
+    {
+      "epoch": 2.7484787018255576,
+      "grad_norm": 1.8521743861085576,
+      "learning_rate": 5.004774850309745e-08,
+      "loss": 2.2366,
+      "step": 2710
+    },
+    {
+      "epoch": 2.7535496957403653,
+      "grad_norm": 1.7735396381086006,
+      "learning_rate": 5.0045676088188616e-08,
+      "loss": 2.2481,
+      "step": 2715
+    },
+    {
+      "epoch": 2.7586206896551726,
+      "grad_norm": 1.750426755759642,
+      "learning_rate": 5.004368576980381e-08,
+      "loss": 2.2235,
+      "step": 2720
+    },
+    {
+      "epoch": 2.76369168356998,
+      "grad_norm": 1.7041388090684644,
+      "learning_rate": 5.004177460930539e-08,
+      "loss": 2.2231,
+      "step": 2725
+    },
+    {
+      "epoch": 2.768762677484787,
+      "grad_norm": 1.8140115420681437,
+      "learning_rate": 5.003993976193124e-08,
+      "loss": 2.2138,
+      "step": 2730
+    },
+    {
+      "epoch": 2.7738336713995944,
+      "grad_norm": 1.822513477317258,
+      "learning_rate": 5.0038178474160234e-08,
+      "loss": 2.2612,
+      "step": 2735
+    },
+    {
+      "epoch": 2.7789046653144016,
+      "grad_norm": 1.7108014551704207,
+      "learning_rate": 5.003648808114121e-08,
+      "loss": 2.2464,
+      "step": 2740
+    },
+    {
+      "epoch": 2.783975659229209,
+      "grad_norm": 1.7880353168056893,
+      "learning_rate": 5.0034866004184443e-08,
+      "loss": 2.2571,
+      "step": 2745
+    },
+    {
+      "epoch": 2.789046653144016,
+      "grad_norm": 1.738078469289302,
+      "learning_rate": 5.003330974831406e-08,
+      "loss": 2.2712,
+      "step": 2750
+    },
+    {
+      "epoch": 2.7941176470588234,
+      "grad_norm": 1.851997917147577,
+      "learning_rate": 5.0031816899880413e-08,
+      "loss": 2.266,
+      "step": 2755
+    },
+    {
+      "epoch": 2.7991886409736306,
+      "grad_norm": 1.7297614602052127,
+      "learning_rate": 5.0030385124230966e-08,
+      "loss": 2.2423,
+      "step": 2760
+    },
+    {
+      "epoch": 2.804259634888438,
+      "grad_norm": 1.8006816107770167,
+      "learning_rate": 5.002901216343864e-08,
+      "loss": 2.2506,
+      "step": 2765
+    },
+    {
+      "epoch": 2.8093306288032456,
+      "grad_norm": 1.8037373860257597,
+      "learning_rate": 5.002769583408638e-08,
+      "loss": 2.2504,
+      "step": 2770
+    },
+    {
+      "epoch": 2.814401622718053,
+      "grad_norm": 1.7406557827783702,
+      "learning_rate": 5.002643402510677e-08,
+      "loss": 2.2676,
+      "step": 2775
+    },
+    {
+      "epoch": 2.81947261663286,
+      "grad_norm": 1.7784795193672072,
+      "learning_rate": 5.0025224695675576e-08,
+      "loss": 2.2052,
+      "step": 2780
+    },
+    {
+      "epoch": 2.8245436105476673,
+      "grad_norm": 1.7627831810019972,
+      "learning_rate": 5.002406587315805e-08,
+      "loss": 2.2315,
+      "step": 2785
+    },
+    {
+      "epoch": 2.8296146044624746,
+      "grad_norm": 1.798869752268086,
+      "learning_rate": 5.0022955651106973e-08,
+      "loss": 2.2436,
+      "step": 2790
+    },
+    {
+      "epoch": 2.834685598377282,
+      "grad_norm": 1.712097491290732,
+      "learning_rate": 5.00218921873112e-08,
+      "loss": 2.274,
+      "step": 2795
+    },
+    {
+      "epoch": 2.839756592292089,
+      "grad_norm": 1.8197661388888422,
+      "learning_rate": 5.002087370189384e-08,
+      "loss": 2.2696,
+      "step": 2800
+    },
+    {
+      "epoch": 2.839756592292089,
+      "eval_loss": 2.4746689796447754,
+      "eval_runtime": 80.933,
+      "eval_samples_per_second": 86.565,
+      "eval_steps_per_second": 0.68,
+      "step": 2800
+    },
+    {
+      "epoch": 2.844827586206897,
+      "grad_norm": 1.7693694208988924,
+      "learning_rate": 5.001989847545882e-08,
+      "loss": 2.2054,
+      "step": 2805
+    },
+    {
+      "epoch": 2.849898580121704,
+      "grad_norm": 1.8223549799119019,
+      "learning_rate": 5.001896484728491e-08,
+      "loss": 2.2656,
+      "step": 2810
+    },
+    {
+      "epoch": 2.8549695740365113,
+      "grad_norm": 1.805868445642325,
+      "learning_rate": 5.00180712135662e-08,
+      "loss": 2.26,
+      "step": 2815
+    },
+    {
+      "epoch": 2.8600405679513186,
+      "grad_norm": 1.7505054153674502,
+      "learning_rate": 5.001721602569797e-08,
+      "loss": 2.2465,
+      "step": 2820
+    },
+    {
+      "epoch": 2.865111561866126,
+      "grad_norm": 1.8486977309170785,
+      "learning_rate": 5.0016397788606984e-08,
+      "loss": 2.2764,
+      "step": 2825
+    },
+    {
+      "epoch": 2.870182555780933,
+      "grad_norm": 1.7740829866432102,
+      "learning_rate": 5.0015615059125324e-08,
+      "loss": 2.2303,
+      "step": 2830
+    },
+    {
+      "epoch": 2.8752535496957403,
+      "grad_norm": 1.7656514305652502,
+      "learning_rate": 5.00148664444067e-08,
+      "loss": 2.238,
+      "step": 2835
+    },
+    {
+      "epoch": 2.8803245436105476,
+      "grad_norm": 1.7634420973902674,
+      "learning_rate": 5.001415060038435e-08,
+      "loss": 2.2489,
+      "step": 2840
+    },
+    {
+      "epoch": 2.885395537525355,
+      "grad_norm": 1.8143454888420456,
+      "learning_rate": 5.0013466230269694e-08,
+      "loss": 2.2607,
+      "step": 2845
+    },
+    {
+      "epoch": 2.890466531440162,
+      "grad_norm": 1.7405623983796592,
+      "learning_rate": 5.001281208309067e-08,
+      "loss": 2.2677,
+      "step": 2850
+    },
+    {
+      "epoch": 2.8955375253549693,
+      "grad_norm": 1.7692613071607504,
+      "learning_rate": 5.0012186952269086e-08,
+      "loss": 2.2499,
+      "step": 2855
+    },
+    {
+      "epoch": 2.900608519269777,
+      "grad_norm": 1.8007487263191868,
+      "learning_rate": 5.0011589674235926e-08,
+      "loss": 2.277,
+      "step": 2860
+    },
+    {
+      "epoch": 2.9056795131845843,
+      "grad_norm": 1.7487914626739638,
+      "learning_rate": 5.001101912708386e-08,
+      "loss": 2.2377,
+      "step": 2865
+    },
+    {
+      "epoch": 2.9107505070993915,
+      "grad_norm": 1.7555747509644022,
+      "learning_rate": 5.0010474229256126e-08,
+      "loss": 2.2532,
+      "step": 2870
+    },
+    {
+      "epoch": 2.915821501014199,
+      "grad_norm": 1.791874000591728,
+      "learning_rate": 5.0009953938270927e-08,
+      "loss": 2.234,
+      "step": 2875
+    },
+    {
+      "epoch": 2.920892494929006,
+      "grad_norm": 1.8071787232301668,
+      "learning_rate": 5.0009457249480536e-08,
+      "loss": 2.2316,
+      "step": 2880
+    },
+    {
+      "epoch": 2.9259634888438133,
+      "grad_norm": 1.7814343272445903,
+      "learning_rate": 5.000898319486436e-08,
+      "loss": 2.2427,
+      "step": 2885
+    },
+    {
+      "epoch": 2.9310344827586206,
+      "grad_norm": 1.8248593697919109,
+      "learning_rate": 5.000853084185513e-08,
+      "loss": 2.2027,
+      "step": 2890
+    },
+    {
+      "epoch": 2.9361054766734282,
+      "grad_norm": 1.7986268547334479,
+      "learning_rate": 5.00080992921975e-08,
+      "loss": 2.244,
+      "step": 2895
+    },
+    {
+      "epoch": 2.9411764705882355,
+      "grad_norm": 1.8701642658692874,
+      "learning_rate": 5.0007687680838296e-08,
+      "loss": 2.2341,
+      "step": 2900
+    },
+    {
+      "epoch": 2.9462474645030428,
+      "grad_norm": 1.7265239787323012,
+      "learning_rate": 5.000729517484766e-08,
+      "loss": 2.2781,
+      "step": 2905
+    },
+    {
+      "epoch": 2.95131845841785,
+      "grad_norm": 1.7596094154490194,
+      "learning_rate": 5.0006920972370384e-08,
+      "loss": 2.2184,
+      "step": 2910
+    },
+    {
+      "epoch": 2.9563894523326573,
+      "grad_norm": 1.775542548895703,
+      "learning_rate": 5.000656430160671e-08,
+      "loss": 2.2404,
+      "step": 2915
+    },
+    {
+      "epoch": 2.9614604462474645,
+      "grad_norm": 1.7859302210997496,
+      "learning_rate": 5.0006224419821984e-08,
+      "loss": 2.2567,
+      "step": 2920
+    },
+    {
+      "epoch": 2.9665314401622718,
+      "grad_norm": 1.8410867262560875,
+      "learning_rate": 5.000590061238431e-08,
+      "loss": 2.2288,
+      "step": 2925
+    },
+    {
+      "epoch": 2.971602434077079,
+      "grad_norm": 1.79261063919542,
+      "learning_rate": 5.0005592191829755e-08,
+      "loss": 2.2421,
+      "step": 2930
+    },
+    {
+      "epoch": 2.9766734279918863,
+      "grad_norm": 1.787266539908181,
+      "learning_rate": 5.0005298496954236e-08,
+      "loss": 2.2713,
+      "step": 2935
+    },
+    {
+      "epoch": 2.9817444219066935,
+      "grad_norm": 1.8046073077938924,
+      "learning_rate": 5.000501889193161e-08,
+      "loss": 2.2292,
+      "step": 2940
+    },
+    {
+      "epoch": 2.986815415821501,
+      "grad_norm": 1.785150585134779,
+      "learning_rate": 5.0004752765457286e-08,
+      "loss": 2.2557,
+      "step": 2945
+    },
+    {
+      "epoch": 2.991886409736308,
+      "grad_norm": 1.7007836630596234,
+      "learning_rate": 5.000449952991666e-08,
+      "loss": 2.2913,
+      "step": 2950
+    },
+    {
+      "epoch": 2.9969574036511157,
+      "grad_norm": 1.7834634363941848,
+      "learning_rate": 5.000425862057791e-08,
+      "loss": 2.2178,
+      "step": 2955
+    },
+    {
+      "epoch": 3.002028397565923,
+      "grad_norm": 1.7711499203458665,
+      "learning_rate": 5.000402949480845e-08,
+      "loss": 2.2302,
+      "step": 2960
+    },
+    {
+      "epoch": 3.0070993914807302,
+      "grad_norm": 1.757400702100505,
+      "learning_rate": 5.000381163131448e-08,
+      "loss": 2.228,
+      "step": 2965
+    },
+    {
+      "epoch": 3.0121703853955375,
+      "grad_norm": 1.7587243027978727,
+      "learning_rate": 5.0003604529403105e-08,
+      "loss": 2.2532,
+      "step": 2970
+    },
+    {
+      "epoch": 3.0172413793103448,
+      "grad_norm": 1.8076763012567914,
+      "learning_rate": 5.000340770826644e-08,
+      "loss": 2.2812,
+      "step": 2975
+    },
+    {
+      "epoch": 3.022312373225152,
+      "grad_norm": 1.7710168575859588,
+      "learning_rate": 5.000322070628711e-08,
+      "loss": 2.2227,
+      "step": 2980
+    },
+    {
+      "epoch": 3.0273833671399593,
+      "grad_norm": 1.7518567665908418,
+      "learning_rate": 5.0003043080364665e-08,
+      "loss": 2.267,
+      "step": 2985
+    },
+    {
+      "epoch": 3.032454361054767,
+      "grad_norm": 1.75371879782544,
+      "learning_rate": 5.0002874405262365e-08,
+      "loss": 2.2748,
+      "step": 2990
+    },
+    {
+      "epoch": 3.037525354969574,
+      "grad_norm": 1.7604102341237111,
+      "learning_rate": 5.000271427297382e-08,
+      "loss": 2.244,
+      "step": 2995
+    },
+    {
+      "epoch": 3.0425963488843815,
+      "grad_norm": 1.7473066315528492,
+      "learning_rate": 5.0002562292108974e-08,
+      "loss": 2.2455,
+      "step": 3000
+    },
+    {
+      "epoch": 3.0425963488843815,
+      "eval_loss": 2.475208282470703,
+      "eval_runtime": 81.0816,
+      "eval_samples_per_second": 86.407,
+      "eval_steps_per_second": 0.678,
+      "step": 3000
+    },
+    {
+      "epoch": 3.0476673427991887,
+      "grad_norm": 1.8183626105425974,
+      "learning_rate": 5.000241808729891e-08,
+      "loss": 2.2598,
+      "step": 3005
+    },
+    {
+      "epoch": 3.052738336713996,
+      "grad_norm": 1.776003383845723,
+      "learning_rate": 5.00022812986191e-08,
+      "loss": 2.2749,
+      "step": 3010
+    },
+    {
+      "epoch": 3.0578093306288032,
+      "grad_norm": 1.8405505191800016,
+      "learning_rate": 5.0002151581030434e-08,
+      "loss": 2.2201,
+      "step": 3015
+    },
+    {
+      "epoch": 3.0628803245436105,
+      "grad_norm": 1.7687042107524293,
+      "learning_rate": 5.00020286038378e-08,
+      "loss": 2.2398,
+      "step": 3020
+    },
+    {
+      "epoch": 3.0679513184584177,
+      "grad_norm": 1.7504153888466234,
+      "learning_rate": 5.000191205016553e-08,
+      "loss": 2.2221,
+      "step": 3025
+    },
+    {
+      "epoch": 3.073022312373225,
+      "grad_norm": 1.7642074409964643,
+      "learning_rate": 5.000180161644944e-08,
+      "loss": 2.2223,
+      "step": 3030
+    },
+    {
+      "epoch": 3.0780933062880322,
+      "grad_norm": 1.7392036544850287,
+      "learning_rate": 5.000169701194494e-08,
+      "loss": 2.2192,
+      "step": 3035
+    },
+    {
+      "epoch": 3.08316430020284,
+      "grad_norm": 1.720350344708903,
+      "learning_rate": 5.0001597958250776e-08,
+      "loss": 2.2315,
+      "step": 3040
+    },
+    {
+      "epoch": 3.088235294117647,
+      "grad_norm": 1.7724706443214726,
+      "learning_rate": 5.000150418884808e-08,
+      "loss": 2.2501,
+      "step": 3045
+    },
+    {
+      "epoch": 3.0933062880324544,
+      "grad_norm": 1.7924639073969963,
+      "learning_rate": 5.000141544865421e-08,
+      "loss": 2.2446,
+      "step": 3050
+    },
+    {
+      "epoch": 3.0983772819472617,
+      "grad_norm": 1.736852243176053,
+      "learning_rate": 5.000133149359102e-08,
+      "loss": 2.2457,
+      "step": 3055
+    },
+    {
+      "epoch": 3.103448275862069,
+      "grad_norm": 1.784090807966895,
+      "learning_rate": 5.000125209016723e-08,
+      "loss": 2.2521,
+      "step": 3060
+    },
+    {
+      "epoch": 3.108519269776876,
+      "grad_norm": 1.7552195819841987,
+      "learning_rate": 5.000117701507439e-08,
+      "loss": 2.2331,
+      "step": 3065
+    },
+    {
+      "epoch": 3.1135902636916835,
+      "grad_norm": 1.7588419707647238,
+      "learning_rate": 5.0001106054796176e-08,
+      "loss": 2.2465,
+      "step": 3070
+    },
+    {
+      "epoch": 3.1186612576064907,
+      "grad_norm": 1.731249391051153,
+      "learning_rate": 5.000103900523059e-08,
+      "loss": 2.2154,
+      "step": 3075
+    },
+    {
+      "epoch": 3.123732251521298,
+      "grad_norm": 1.86107961069035,
+      "learning_rate": 5.0000975671324725e-08,
+      "loss": 2.2498,
+      "step": 3080
+    },
+    {
+      "epoch": 3.1288032454361057,
+      "grad_norm": 1.7453958505335196,
+      "learning_rate": 5.000091586672176e-08,
+      "loss": 2.213,
+      "step": 3085
+    },
+    {
+      "epoch": 3.133874239350913,
+      "grad_norm": 1.739107722358469,
+      "learning_rate": 5.000085941341981e-08,
+      "loss": 2.2703,
+      "step": 3090
+    },
+    {
+      "epoch": 3.13894523326572,
+      "grad_norm": 1.723031377500322,
+      "learning_rate": 5.000080614144228e-08,
+      "loss": 2.256,
+      "step": 3095
+    },
+    {
+      "epoch": 3.1440162271805274,
+      "grad_norm": 1.7859618571141844,
+      "learning_rate": 5.0000755888519526e-08,
+      "loss": 2.2446,
+      "step": 3100
+    },
+    {
+      "epoch": 3.1490872210953347,
+      "grad_norm": 1.7642645902841112,
+      "learning_rate": 5.0000708499781274e-08,
+      "loss": 2.2365,
+      "step": 3105
+    },
+    {
+      "epoch": 3.154158215010142,
+      "grad_norm": 1.8188951223969028,
+      "learning_rate": 5.000066382745973e-08,
+      "loss": 2.2743,
+      "step": 3110
+    },
+    {
+      "epoch": 3.159229208924949,
+      "grad_norm": 1.8017937041457348,
+      "learning_rate": 5.000062173060291e-08,
+      "loss": 2.2501,
+      "step": 3115
+    },
+    {
+      "epoch": 3.1643002028397564,
+      "grad_norm": 1.7816544045204796,
+      "learning_rate": 5.0000582074797944e-08,
+      "loss": 2.2025,
+      "step": 3120
+    },
+    {
+      "epoch": 3.1693711967545637,
+      "grad_norm": 1.7911385432695703,
+      "learning_rate": 5.0000544731904076e-08,
+      "loss": 2.2284,
+      "step": 3125
+    },
+    {
+      "epoch": 3.1744421906693714,
+      "grad_norm": 1.9232399576032946,
+      "learning_rate": 5.000050957979507e-08,
+      "loss": 2.2407,
+      "step": 3130
+    },
+    {
+      "epoch": 3.1795131845841786,
+      "grad_norm": 1.7293397524348884,
+      "learning_rate": 5.000047650211071e-08,
+      "loss": 2.2468,
+      "step": 3135
+    },
+    {
+      "epoch": 3.184584178498986,
+      "grad_norm": 1.7870474846773756,
+      "learning_rate": 5.000044538801721e-08,
+      "loss": 2.2432,
+      "step": 3140
+    },
+    {
+      "epoch": 3.189655172413793,
+      "grad_norm": 1.7179456705770244,
+      "learning_rate": 5.000041613197611e-08,
+      "loss": 2.2478,
+      "step": 3145
+    },
+    {
+      "epoch": 3.1947261663286004,
+      "grad_norm": 1.782930312662543,
+      "learning_rate": 5.0000388633521626e-08,
+      "loss": 2.219,
+      "step": 3150
+    },
+    {
+      "epoch": 3.1997971602434077,
+      "grad_norm": 1.8396726211182168,
+      "learning_rate": 5.000036279704598e-08,
+      "loss": 2.2131,
+      "step": 3155
+    },
+    {
+      "epoch": 3.204868154158215,
+      "grad_norm": 1.7441223394696925,
+      "learning_rate": 5.000033853159261e-08,
+      "loss": 2.216,
+      "step": 3160
+    },
+    {
+      "epoch": 3.209939148073022,
+      "grad_norm": 1.79701015495686,
+      "learning_rate": 5.000031575065695e-08,
+      "loss": 2.2423,
+      "step": 3165
+    },
+    {
+      "epoch": 3.2150101419878294,
+      "grad_norm": 1.7824241551117812,
+      "learning_rate": 5.000029437199458e-08,
+      "loss": 2.245,
+      "step": 3170
+    },
+    {
+      "epoch": 3.220081135902637,
+      "grad_norm": 1.7859671284571614,
+      "learning_rate": 5.000027431743653e-08,
+      "loss": 2.2466,
+      "step": 3175
+    },
+    {
+      "epoch": 3.2251521298174444,
+      "grad_norm": 1.7508641392805016,
+      "learning_rate": 5.000025551271141e-08,
+      "loss": 2.2123,
+      "step": 3180
+    },
+    {
+      "epoch": 3.2302231237322516,
+      "grad_norm": 1.790375251718636,
+      "learning_rate": 5.000023788727435e-08,
+      "loss": 2.2387,
+      "step": 3185
+    },
+    {
+      "epoch": 3.235294117647059,
+      "grad_norm": 1.8347285573544698,
+      "learning_rate": 5.0000221374142326e-08,
+      "loss": 2.2024,
+      "step": 3190
+    },
+    {
+      "epoch": 3.240365111561866,
+      "grad_norm": 1.766020664546832,
+      "learning_rate": 5.0000205909735805e-08,
+      "loss": 2.25,
+      "step": 3195
+    },
+    {
+      "epoch": 3.2454361054766734,
+      "grad_norm": 1.7685652184853669,
+      "learning_rate": 5.000019143372644e-08,
+      "loss": 2.216,
+      "step": 3200
+    },
+    {
+      "epoch": 3.2454361054766734,
+      "eval_loss": 2.475315809249878,
+      "eval_runtime": 81.0728,
+      "eval_samples_per_second": 86.416,
+      "eval_steps_per_second": 0.678,
+      "step": 3200
+    },
+    {
+      "epoch": 3.2505070993914806,
+      "grad_norm": 1.8114020440458831,
+      "learning_rate": 5.000017788889067e-08,
+      "loss": 2.2909,
+      "step": 3205
+    },
+    {
+      "epoch": 3.255578093306288,
+      "grad_norm": 1.8044780174846506,
+      "learning_rate": 5.0000165220969006e-08,
+      "loss": 2.2682,
+      "step": 3210
+    },
+    {
+      "epoch": 3.260649087221095,
+      "grad_norm": 1.8227060747974817,
+      "learning_rate": 5.0000153378530776e-08,
+      "loss": 2.2551,
+      "step": 3215
+    },
+    {
+      "epoch": 3.2657200811359024,
+      "grad_norm": 1.712746112733307,
+      "learning_rate": 5.000014231284425e-08,
+      "loss": 2.2085,
+      "step": 3220
+    },
+    {
+      "epoch": 3.27079107505071,
+      "grad_norm": 1.7693643379563115,
+      "learning_rate": 5.000013197775189e-08,
+      "loss": 2.2089,
+      "step": 3225
+    },
+    {
+      "epoch": 3.2758620689655173,
+      "grad_norm": 1.742416891486272,
+      "learning_rate": 5.000012232955056e-08,
+      "loss": 2.2256,
+      "step": 3230
+    },
+    {
+      "epoch": 3.2809330628803246,
+      "grad_norm": 1.7588332712006007,
+      "learning_rate": 5.000011332687656e-08,
+      "loss": 2.2411,
+      "step": 3235
+    },
+    {
+      "epoch": 3.286004056795132,
+      "grad_norm": 1.748987632844159,
+      "learning_rate": 5.000010493059533e-08,
+      "loss": 2.2161,
+      "step": 3240
+    },
+    {
+      "epoch": 3.291075050709939,
+      "grad_norm": 1.7730209178260556,
+      "learning_rate": 5.000009710369558e-08,
+      "loss": 2.2454,
+      "step": 3245
+    },
+    {
+      "epoch": 3.2961460446247464,
+      "grad_norm": 1.7638994477476329,
+      "learning_rate": 5.000008981118782e-08,
+      "loss": 2.2762,
+      "step": 3250
+    },
+    {
+      "epoch": 3.3012170385395536,
+      "grad_norm": 1.8306906774843352,
+      "learning_rate": 5.000008302000705e-08,
+      "loss": 2.2484,
+      "step": 3255
+    },
+    {
+      "epoch": 3.306288032454361,
+      "grad_norm": 1.8155910247025784,
+      "learning_rate": 5.0000076698919504e-08,
+      "loss": 2.2172,
+      "step": 3260
+    },
+    {
+      "epoch": 3.3113590263691686,
+      "grad_norm": 1.9000838772092157,
+      "learning_rate": 5.0000070818433264e-08,
+      "loss": 2.2639,
+      "step": 3265
+    },
+    {
+      "epoch": 3.316430020283976,
+      "grad_norm": 1.8182257588876376,
+      "learning_rate": 5.000006535071267e-08,
+      "loss": 2.2302,
+      "step": 3270
+    },
+    {
+      "epoch": 3.321501014198783,
+      "grad_norm": 1.7421030430480422,
+      "learning_rate": 5.0000060269496374e-08,
+      "loss": 2.2618,
+      "step": 3275
+    },
+    {
+      "epoch": 3.3265720081135903,
+      "grad_norm": 1.7545361773998456,
+      "learning_rate": 5.0000055550018825e-08,
+      "loss": 2.2174,
+      "step": 3280
+    },
+    {
+      "epoch": 3.3316430020283976,
+      "grad_norm": 1.7382589137313635,
+      "learning_rate": 5.000005116893524e-08,
+      "loss": 2.2497,
+      "step": 3285
+    },
+    {
+      "epoch": 3.336713995943205,
+      "grad_norm": 1.7544110577796528,
+      "learning_rate": 5.000004710424972e-08,
+      "loss": 2.2386,
+      "step": 3290
+    },
+    {
+      "epoch": 3.341784989858012,
+      "grad_norm": 1.7756370830140873,
+      "learning_rate": 5.0000043335246576e-08,
+      "loss": 2.2124,
+      "step": 3295
+    },
+    {
+      "epoch": 3.3468559837728193,
+      "grad_norm": 1.7647740914276824,
+      "learning_rate": 5.0000039842424645e-08,
+      "loss": 2.2357,
+      "step": 3300
+    },
+    {
+      "epoch": 3.3519269776876266,
+      "grad_norm": 1.7614092517536837,
+      "learning_rate": 5.000003660743452e-08,
+      "loss": 2.2823,
+      "step": 3305
+    },
+    {
+      "epoch": 3.356997971602434,
+      "grad_norm": 1.7889494130903192,
+      "learning_rate": 5.000003361301858e-08,
+      "loss": 2.1835,
+      "step": 3310
+    },
+    {
+      "epoch": 3.3620689655172415,
+      "grad_norm": 1.7154434994558871,
+      "learning_rate": 5.000003084295374e-08,
+      "loss": 2.2724,
+      "step": 3315
+    },
+    {
+      "epoch": 3.367139959432049,
+      "grad_norm": 1.8155130093382392,
+      "learning_rate": 5.0000028281996743e-08,
+      "loss": 2.2823,
+      "step": 3320
+    },
+    {
+      "epoch": 3.372210953346856,
+      "grad_norm": 1.880078020122213,
+      "learning_rate": 5.0000025915832e-08,
+      "loss": 2.2421,
+      "step": 3325
+    },
+    {
+      "epoch": 3.3772819472616633,
+      "grad_norm": 1.7913171122885942,
+      "learning_rate": 5.000002373102181e-08,
+      "loss": 2.1806,
+      "step": 3330
+    },
+    {
+      "epoch": 3.3823529411764706,
+      "grad_norm": 1.8110141267464457,
+      "learning_rate": 5.000002171495887e-08,
+      "loss": 2.2315,
+      "step": 3335
+    },
+    {
+      "epoch": 3.387423935091278,
+      "grad_norm": 1.8187945379716748,
+      "learning_rate": 5.000001985582107e-08,
+      "loss": 2.2207,
+      "step": 3340
+    },
+    {
+      "epoch": 3.392494929006085,
+      "grad_norm": 1.7822827152937282,
+      "learning_rate": 5.000001814252828e-08,
+      "loss": 2.2411,
+      "step": 3345
+    },
+    {
+      "epoch": 3.3975659229208923,
+      "grad_norm": 1.7281310183638643,
+      "learning_rate": 5.0000016564701364e-08,
+      "loss": 2.2415,
+      "step": 3350
+    },
+    {
+      "epoch": 3.4026369168357,
+      "grad_norm": 1.7550793470914747,
+      "learning_rate": 5.000001511262302e-08,
+      "loss": 2.2464,
+      "step": 3355
+    },
+    {
+      "epoch": 3.4077079107505073,
+      "grad_norm": 1.7459578038518018,
+      "learning_rate": 5.0000013777200565e-08,
+      "loss": 2.2504,
+      "step": 3360
+    },
+    {
+      "epoch": 3.4127789046653145,
+      "grad_norm": 1.740338062654503,
+      "learning_rate": 5.000001254993049e-08,
+      "loss": 2.2292,
+      "step": 3365
+    },
+    {
+      "epoch": 3.417849898580122,
+      "grad_norm": 1.8005446847395141,
+      "learning_rate": 5.000001142286484e-08,
+      "loss": 2.2646,
+      "step": 3370
+    },
+    {
+      "epoch": 3.422920892494929,
+      "grad_norm": 1.8075984781615184,
+      "learning_rate": 5.000001038857911e-08,
+      "loss": 2.2549,
+      "step": 3375
+    },
+    {
+      "epoch": 3.4279918864097363,
+      "grad_norm": 1.7944612854944237,
+      "learning_rate": 5.000000944014192e-08,
+      "loss": 2.2607,
+      "step": 3380
+    },
+    {
+      "epoch": 3.4330628803245435,
+      "grad_norm": 1.8042996177778357,
+      "learning_rate": 5.000000857108604e-08,
+      "loss": 2.2129,
+      "step": 3385
+    },
+    {
+      "epoch": 3.438133874239351,
+      "grad_norm": 1.812331539187214,
+      "learning_rate": 5.0000007775380984e-08,
+      "loss": 2.247,
+      "step": 3390
+    },
+    {
+      "epoch": 3.443204868154158,
+      "grad_norm": 1.7634101221518121,
+      "learning_rate": 5.0000007047407e-08,
+      "loss": 2.2454,
+      "step": 3395
+    },
+    {
+      "epoch": 3.4482758620689653,
+      "grad_norm": 1.8137979752467785,
+      "learning_rate": 5.000000638193037e-08,
+      "loss": 2.2348,
+      "step": 3400
+    },
+    {
+      "epoch": 3.4482758620689653,
+      "eval_loss": 2.475677013397217,
+      "eval_runtime": 81.0429,
+      "eval_samples_per_second": 86.448,
+      "eval_steps_per_second": 0.679,
+      "step": 3400
+    },
+    {
+      "epoch": 3.453346855983773,
+      "grad_norm": 1.7639358988388496,
+      "learning_rate": 5.0000005774079994e-08,
+      "loss": 2.2434,
+      "step": 3405
+    },
+    {
+      "epoch": 3.4584178498985803,
+      "grad_norm": 1.8860372894717414,
+      "learning_rate": 5.0000005219325215e-08,
+      "loss": 2.2184,
+      "step": 3410
+    },
+    {
+      "epoch": 3.4634888438133875,
+      "grad_norm": 1.792302245525526,
+      "learning_rate": 5.000000471345483e-08,
+      "loss": 2.2405,
+      "step": 3415
+    },
+    {
+      "epoch": 3.4685598377281948,
+      "grad_norm": 1.7326646638681342,
+      "learning_rate": 5.000000425255718e-08,
+      "loss": 2.2582,
+      "step": 3420
+    },
+    {
+      "epoch": 3.473630831643002,
+      "grad_norm": 1.7944771245301363,
+      "learning_rate": 5.0000003833001365e-08,
+      "loss": 2.202,
+      "step": 3425
+    },
+    {
+      "epoch": 3.4787018255578093,
+      "grad_norm": 1.8158606522431084,
+      "learning_rate": 5.000000345141943e-08,
+      "loss": 2.2533,
+      "step": 3430
+    },
+    {
+      "epoch": 3.4837728194726165,
+      "grad_norm": 1.8541024781685664,
+      "learning_rate": 5.0000003104689555e-08,
+      "loss": 2.2387,
+      "step": 3435
+    },
+    {
+      "epoch": 3.4888438133874238,
+      "grad_norm": 1.7999921658917655,
+      "learning_rate": 5.0000002789920174e-08,
+      "loss": 2.2441,
+      "step": 3440
+    },
+    {
+      "epoch": 3.4939148073022315,
+      "grad_norm": 1.7685774604287066,
+      "learning_rate": 5.000000250443497e-08,
+      "loss": 2.3018,
+      "step": 3445
+    },
+    {
+      "epoch": 3.4989858012170387,
+      "grad_norm": 1.7777470112493552,
+      "learning_rate": 5.000000224575872e-08,
+      "loss": 2.2433,
+      "step": 3450
+    },
+    {
+      "epoch": 3.504056795131846,
+      "grad_norm": 1.7748253950125374,
+      "learning_rate": 5.000000201160396e-08,
+      "loss": 2.2782,
+      "step": 3455
+    },
+    {
+      "epoch": 3.5091277890466532,
+      "grad_norm": 1.7842700957790634,
+      "learning_rate": 5.000000179985839e-08,
+      "loss": 2.2659,
+      "step": 3460
+    },
+    {
+      "epoch": 3.5141987829614605,
+      "grad_norm": 1.798939281745875,
+      "learning_rate": 5.000000160857302e-08,
+      "loss": 2.2396,
+      "step": 3465
+    },
+    {
+      "epoch": 3.5192697768762677,
+      "grad_norm": 1.8045276757468276,
+      "learning_rate": 5.000000143595102e-08,
+      "loss": 2.2325,
+      "step": 3470
+    },
+    {
+      "epoch": 3.524340770791075,
+      "grad_norm": 1.7262031285233723,
+      "learning_rate": 5.0000001280337235e-08,
+      "loss": 2.243,
+      "step": 3475
+    },
+    {
+      "epoch": 3.5294117647058822,
+      "grad_norm": 1.8375261220518257,
+      "learning_rate": 5.000000114020828e-08,
+      "loss": 2.2075,
+      "step": 3480
+    },
+    {
+      "epoch": 3.5344827586206895,
+      "grad_norm": 1.8163152606406519,
+      "learning_rate": 5.0000001014163305e-08,
+      "loss": 2.2494,
+      "step": 3485
+    },
+    {
+      "epoch": 3.5395537525354968,
+      "grad_norm": 1.8525927462335219,
+      "learning_rate": 5.0000000900915245e-08,
+      "loss": 2.2163,
+      "step": 3490
+    },
+    {
+      "epoch": 3.544624746450304,
+      "grad_norm": 1.7805165281974848,
+      "learning_rate": 5.000000079928269e-08,
+      "loss": 2.2525,
+      "step": 3495
+    },
+    {
+      "epoch": 3.5496957403651117,
+      "grad_norm": 1.7990737454408499,
+      "learning_rate": 5.000000070818217e-08,
+      "loss": 2.2874,
+      "step": 3500
+    },
+    {
+      "epoch": 3.554766734279919,
+      "grad_norm": 1.8247781997920414,
+      "learning_rate": 5.000000062662102e-08,
+      "loss": 2.2215,
+      "step": 3505
+    },
+    {
+      "epoch": 3.559837728194726,
+      "grad_norm": 1.9826615858248522,
+      "learning_rate": 5.000000055369062e-08,
+      "loss": 2.2443,
+      "step": 3510
+    },
+    {
+      "epoch": 3.5649087221095335,
+      "grad_norm": 1.799487216606698,
+      "learning_rate": 5.000000048856012e-08,
+      "loss": 2.2266,
+      "step": 3515
+    },
+    {
+      "epoch": 3.5699797160243407,
+      "grad_norm": 1.8091696515518445,
+      "learning_rate": 5.0000000430470526e-08,
+      "loss": 2.2517,
+      "step": 3520
+    },
+    {
+      "epoch": 3.575050709939148,
+      "grad_norm": 1.7814535925288772,
+      "learning_rate": 5.0000000378729234e-08,
+      "loss": 2.2321,
+      "step": 3525
+    },
+    {
+      "epoch": 3.5801217038539552,
+      "grad_norm": 1.850742214981416,
+      "learning_rate": 5.000000033270488e-08,
+      "loss": 2.2597,
+      "step": 3530
+    },
+    {
+      "epoch": 3.585192697768763,
+      "grad_norm": 1.7822355084719033,
+      "learning_rate": 5.000000029182252e-08,
+      "loss": 2.2963,
+      "step": 3535
+    },
+    {
+      "epoch": 3.59026369168357,
+      "grad_norm": 1.7548584963433536,
+      "learning_rate": 5.0000000255559235e-08,
+      "loss": 2.2669,
+      "step": 3540
+    },
+    {
+      "epoch": 3.5953346855983774,
+      "grad_norm": 1.8526633444752874,
+      "learning_rate": 5.0000000223439884e-08,
+      "loss": 2.2367,
+      "step": 3545
+    },
+    {
+      "epoch": 3.6004056795131847,
+      "grad_norm": 1.8813498033155052,
+      "learning_rate": 5.0000000195033304e-08,
+      "loss": 2.2373,
+      "step": 3550
+    },
+    {
+      "epoch": 3.605476673427992,
+      "grad_norm": 1.7670822667081592,
+      "learning_rate": 5.0000000169948675e-08,
+      "loss": 2.2705,
+      "step": 3555
+    },
+    {
+      "epoch": 3.610547667342799,
+      "grad_norm": 1.7756286276528583,
+      "learning_rate": 5.000000014783217e-08,
+      "loss": 2.2979,
+      "step": 3560
+    },
+    {
+      "epoch": 3.6156186612576064,
+      "grad_norm": 1.7467172856710016,
+      "learning_rate": 5.000000012836387e-08,
+      "loss": 2.2538,
+      "step": 3565
+    },
+    {
+      "epoch": 3.6206896551724137,
+      "grad_norm": 1.7107623358426811,
+      "learning_rate": 5.000000011125491e-08,
+      "loss": 2.2807,
+      "step": 3570
+    },
+    {
+      "epoch": 3.625760649087221,
+      "grad_norm": 1.8431542462448438,
+      "learning_rate": 5.000000009624475e-08,
+      "loss": 2.252,
+      "step": 3575
+    },
+    {
+      "epoch": 3.630831643002028,
+      "grad_norm": 1.7683303237840782,
+      "learning_rate": 5.000000008309876e-08,
+      "loss": 2.2722,
+      "step": 3580
+    },
+    {
+      "epoch": 3.6359026369168355,
+      "grad_norm": 1.7463535795755278,
+      "learning_rate": 5.000000007160591e-08,
+      "loss": 2.2712,
+      "step": 3585
+    },
+    {
+      "epoch": 3.640973630831643,
+      "grad_norm": 1.8412435208194315,
+      "learning_rate": 5.0000000061576706e-08,
+      "loss": 2.2438,
+      "step": 3590
+    },
+    {
+      "epoch": 3.6460446247464504,
+      "grad_norm": 1.7731354966851007,
+      "learning_rate": 5.000000005284119e-08,
+      "loss": 2.2305,
+      "step": 3595
+    },
+    {
+      "epoch": 3.6511156186612577,
+      "grad_norm": 1.7263977118619886,
+      "learning_rate": 5.0000000045247174e-08,
+      "loss": 2.238,
+      "step": 3600
+    },
+    {
+      "epoch": 3.6511156186612577,
+      "eval_loss": 2.475299596786499,
+      "eval_runtime": 81.0503,
+      "eval_samples_per_second": 86.44,
+      "eval_steps_per_second": 0.679,
+      "step": 3600
+    },
+    {
+      "epoch": 3.656186612576065,
+      "grad_norm": 1.725184319305705,
+      "learning_rate": 5.000000003865863e-08,
+      "loss": 2.2283,
+      "step": 3605
+    },
+    {
+      "epoch": 3.661257606490872,
+      "grad_norm": 1.9023050674895976,
+      "learning_rate": 5.000000003295409e-08,
+      "loss": 2.21,
+      "step": 3610
+    },
+    {
+      "epoch": 3.6663286004056794,
+      "grad_norm": 1.8044353617499143,
+      "learning_rate": 5.0000000028025353e-08,
+      "loss": 2.2658,
+      "step": 3615
+    },
+    {
+      "epoch": 3.6713995943204867,
+      "grad_norm": 1.7560239895320502,
+      "learning_rate": 5.0000000023776127e-08,
+      "loss": 2.2558,
+      "step": 3620
+    },
+    {
+      "epoch": 3.6764705882352944,
+      "grad_norm": 1.9019670084185585,
+      "learning_rate": 5.00000000201209e-08,
+      "loss": 2.2154,
+      "step": 3625
+    },
+    {
+      "epoch": 3.6815415821501016,
+      "grad_norm": 1.835689830804529,
+      "learning_rate": 5.0000000016983875e-08,
+      "loss": 2.2586,
+      "step": 3630
+    },
+    {
+      "epoch": 3.686612576064909,
+      "grad_norm": 1.8589538257906977,
+      "learning_rate": 5.000000001429796e-08,
+      "loss": 2.2388,
+      "step": 3635
+    },
+    {
+      "epoch": 3.691683569979716,
+      "grad_norm": 1.8068715773945243,
+      "learning_rate": 5.000000001200391e-08,
+      "loss": 2.2571,
+      "step": 3640
+    },
+    {
+      "epoch": 3.6967545638945234,
+      "grad_norm": 1.7775448603509494,
+      "learning_rate": 5.0000000010049494e-08,
+      "loss": 2.2751,
+      "step": 3645
+    },
+    {
+      "epoch": 3.7018255578093306,
+      "grad_norm": 1.748064680879759,
+      "learning_rate": 5.0000000008388774e-08,
+      "loss": 2.2183,
+      "step": 3650
+    },
+    {
+      "epoch": 3.706896551724138,
+      "grad_norm": 1.752057568304335,
+      "learning_rate": 5.000000000698141e-08,
+      "loss": 2.2532,
+      "step": 3655
+    },
+    {
+      "epoch": 3.711967545638945,
+      "grad_norm": 1.7976874660325244,
+      "learning_rate": 5.000000000579206e-08,
+      "loss": 2.2447,
+      "step": 3660
+    },
+    {
+      "epoch": 3.7170385395537524,
+      "grad_norm": 1.8361658170177098,
+      "learning_rate": 5.000000000478986e-08,
+      "loss": 2.2274,
+      "step": 3665
+    },
+    {
+      "epoch": 3.7221095334685597,
+      "grad_norm": 1.7595086838837224,
+      "learning_rate": 5.0000000003947866e-08,
+      "loss": 2.2704,
+      "step": 3670
+    },
+    {
+      "epoch": 3.727180527383367,
+      "grad_norm": 1.7772692374868122,
+      "learning_rate": 5.0000000003242645e-08,
+      "loss": 2.2394,
+      "step": 3675
+    },
+    {
+      "epoch": 3.732251521298174,
+      "grad_norm": 1.7860835232171102,
+      "learning_rate": 5.000000000265387e-08,
+      "loss": 2.238,
+      "step": 3680
+    },
+    {
+      "epoch": 3.737322515212982,
+      "grad_norm": 1.7590689183822192,
+      "learning_rate": 5.000000000216394e-08,
+      "loss": 2.2764,
+      "step": 3685
+    },
+    {
+      "epoch": 3.742393509127789,
+      "grad_norm": 1.7659065260707336,
+      "learning_rate": 5.0000000001757664e-08,
+      "loss": 2.2459,
+      "step": 3690
+    },
+    {
+      "epoch": 3.7474645030425964,
+      "grad_norm": 1.8153822365083379,
+      "learning_rate": 5.0000000001421954e-08,
+      "loss": 2.2299,
+      "step": 3695
+    },
+    {
+      "epoch": 3.7525354969574036,
+      "grad_norm": 1.7704864144251407,
+      "learning_rate": 5.0000000001145583e-08,
+      "loss": 2.247,
+      "step": 3700
+    },
+    {
+      "epoch": 3.757606490872211,
+      "grad_norm": 1.7268180675977047,
+      "learning_rate": 5.000000000091894e-08,
+      "loss": 2.2483,
+      "step": 3705
+    },
+    {
+      "epoch": 3.762677484787018,
+      "grad_norm": 1.7808473093189052,
+      "learning_rate": 5.000000000073382e-08,
+      "loss": 2.2774,
+      "step": 3710
+    },
+    {
+      "epoch": 3.767748478701826,
+      "grad_norm": 1.7999930755140212,
+      "learning_rate": 5.0000000000583246e-08,
+      "loss": 2.2209,
+      "step": 3715
+    },
+    {
+      "epoch": 3.772819472616633,
+      "grad_norm": 1.7741565202241085,
+      "learning_rate": 5.0000000000461306e-08,
+      "loss": 2.2353,
+      "step": 3720
+    },
+    {
+      "epoch": 3.7778904665314403,
+      "grad_norm": 1.8046657930760326,
+      "learning_rate": 5.0000000000363e-08,
+      "loss": 2.2255,
+      "step": 3725
+    },
+    {
+      "epoch": 3.7829614604462476,
+      "grad_norm": 1.8038566902418574,
+      "learning_rate": 5.000000000028412e-08,
+      "loss": 2.2781,
+      "step": 3730
+    },
+    {
+      "epoch": 3.788032454361055,
+      "grad_norm": 1.7944584001789026,
+      "learning_rate": 5.0000000000221146e-08,
+      "loss": 2.272,
+      "step": 3735
+    },
+    {
+      "epoch": 3.793103448275862,
+      "grad_norm": 1.7491739462315397,
+      "learning_rate": 5.0000000000171125e-08,
+      "loss": 2.2293,
+      "step": 3740
+    },
+    {
+      "epoch": 3.7981744421906694,
+      "grad_norm": 1.7505007397811716,
+      "learning_rate": 5.000000000013161e-08,
+      "loss": 2.2373,
+      "step": 3745
+    },
+    {
+      "epoch": 3.8032454361054766,
+      "grad_norm": 1.8014769703402196,
+      "learning_rate": 5.000000000010057e-08,
+      "loss": 2.2552,
+      "step": 3750
+    },
+    {
+      "epoch": 3.808316430020284,
+      "grad_norm": 1.7608287864741985,
+      "learning_rate": 5.0000000000076337e-08,
+      "loss": 2.2277,
+      "step": 3755
+    },
+    {
+      "epoch": 3.813387423935091,
+      "grad_norm": 1.8323757058256038,
+      "learning_rate": 5.0000000000057536e-08,
+      "loss": 2.2341,
+      "step": 3760
+    },
+    {
+      "epoch": 3.8184584178498984,
+      "grad_norm": 1.7574657806555387,
+      "learning_rate": 5.000000000004304e-08,
+      "loss": 2.2223,
+      "step": 3765
+    },
+    {
+      "epoch": 3.8235294117647056,
+      "grad_norm": 1.7900689784727426,
+      "learning_rate": 5.000000000003194e-08,
+      "loss": 2.2445,
+      "step": 3770
+    },
+    {
+      "epoch": 3.8286004056795133,
+      "grad_norm": 1.7873969080046235,
+      "learning_rate": 5.000000000002351e-08,
+      "loss": 2.2692,
+      "step": 3775
+    },
+    {
+      "epoch": 3.8336713995943206,
+      "grad_norm": 1.7693343584107923,
+      "learning_rate": 5.000000000001716e-08,
+      "loss": 2.1982,
+      "step": 3780
+    },
+    {
+      "epoch": 3.838742393509128,
+      "grad_norm": 1.782049247288072,
+      "learning_rate": 5.00000000000124e-08,
+      "loss": 2.2417,
+      "step": 3785
+    },
+    {
+      "epoch": 3.843813387423935,
+      "grad_norm": 1.8357614780582354,
+      "learning_rate": 5.000000000000888e-08,
+      "loss": 2.2414,
+      "step": 3790
+    },
+    {
+      "epoch": 3.8488843813387423,
+      "grad_norm": 1.7593131764821546,
+      "learning_rate": 5.0000000000006284e-08,
+      "loss": 2.2721,
+      "step": 3795
+    },
+    {
+      "epoch": 3.8539553752535496,
+      "grad_norm": 1.8355045282767246,
+      "learning_rate": 5.0000000000004405e-08,
+      "loss": 2.2349,
+      "step": 3800
+    },
+    {
+      "epoch": 3.8539553752535496,
+      "eval_loss": 2.475205421447754,
+      "eval_runtime": 81.089,
+      "eval_samples_per_second": 86.399,
+      "eval_steps_per_second": 0.678,
+      "step": 3800
+    },
+    {
+      "epoch": 3.859026369168357,
+      "grad_norm": 1.7617334472370734,
+      "learning_rate": 5.000000000000305e-08,
+      "loss": 2.2796,
+      "step": 3805
+    },
+    {
+      "epoch": 3.8640973630831645,
+      "grad_norm": 1.7655616354078496,
+      "learning_rate": 5.000000000000208e-08,
+      "loss": 2.2904,
+      "step": 3810
+    },
+    {
+      "epoch": 3.869168356997972,
+      "grad_norm": 1.7499887502905194,
+      "learning_rate": 5.00000000000014e-08,
+      "loss": 2.2289,
+      "step": 3815
+    },
+    {
+      "epoch": 3.874239350912779,
+      "grad_norm": 1.7552158736441676,
+      "learning_rate": 5.000000000000093e-08,
+      "loss": 2.2524,
+      "step": 3820
+    },
+    {
+      "epoch": 3.8793103448275863,
+      "grad_norm": 1.779864718453615,
+      "learning_rate": 5.0000000000000607e-08,
+      "loss": 2.2557,
+      "step": 3825
+    },
+    {
+      "epoch": 3.8843813387423936,
+      "grad_norm": 1.8326086874257492,
+      "learning_rate": 5.000000000000039e-08,
+      "loss": 2.2642,
+      "step": 3830
+    },
+    {
+      "epoch": 3.889452332657201,
+      "grad_norm": 1.7709614684441606,
+      "learning_rate": 5.000000000000024e-08,
+      "loss": 2.2316,
+      "step": 3835
+    },
+    {
+      "epoch": 3.894523326572008,
+      "grad_norm": 1.8053802580849208,
+      "learning_rate": 5.000000000000015e-08,
+      "loss": 2.2568,
+      "step": 3840
+    },
+    {
+      "epoch": 3.8995943204868153,
+      "grad_norm": 1.7935470548184194,
+      "learning_rate": 5.0000000000000104e-08,
+      "loss": 2.2993,
+      "step": 3845
+    },
+    {
+      "epoch": 3.9046653144016226,
+      "grad_norm": 1.7497664491493299,
+      "learning_rate": 5.000000000000006e-08,
+      "loss": 2.1989,
+      "step": 3850
+    },
+    {
+      "epoch": 3.90973630831643,
+      "grad_norm": 1.754972418650299,
+      "learning_rate": 5.000000000000003e-08,
+      "loss": 2.2424,
+      "step": 3855
+    },
+    {
+      "epoch": 3.914807302231237,
+      "grad_norm": 1.7589479994346042,
+      "learning_rate": 5.000000000000002e-08,
+      "loss": 2.2632,
+      "step": 3860
+    },
+    {
+      "epoch": 3.9198782961460448,
+      "grad_norm": 1.7971848831669277,
+      "learning_rate": 5.000000000000001e-08,
+      "loss": 2.2336,
+      "step": 3865
+    },
+    {
+      "epoch": 3.924949290060852,
+      "grad_norm": 1.7639968737695348,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 2.2296,
+      "step": 3870
+    },
+    {
+      "epoch": 3.9300202839756593,
+      "grad_norm": 1.72827012743299,
+      "learning_rate": 5e-08,
+      "loss": 2.2451,
+      "step": 3875
+    },
+    {
+      "epoch": 3.9350912778904665,
+      "grad_norm": 1.749153588059136,
+      "learning_rate": 5e-08,
+      "loss": 2.258,
+      "step": 3880
+    },
+    {
+      "epoch": 3.940162271805274,
+      "grad_norm": 1.753206456867822,
+      "learning_rate": 5e-08,
+      "loss": 2.2587,
+      "step": 3885
+    },
+    {
+      "epoch": 3.945233265720081,
+      "grad_norm": 1.7816747777928572,
+      "learning_rate": 5e-08,
+      "loss": 2.2532,
+      "step": 3890
+    },
+    {
+      "epoch": 3.9503042596348883,
+      "grad_norm": 1.7762615930524053,
+      "learning_rate": 5e-08,
+      "loss": 2.2331,
+      "step": 3895
+    },
+    {
+      "epoch": 3.955375253549696,
+      "grad_norm": 1.8039115341801395,
+      "learning_rate": 5e-08,
+      "loss": 2.2271,
+      "step": 3900
+    },
+    {
+      "epoch": 3.9604462474645032,
+      "grad_norm": 1.7530354888252304,
+      "learning_rate": 5e-08,
+      "loss": 2.2191,
+      "step": 3905
+    },
+    {
+      "epoch": 3.9655172413793105,
+      "grad_norm": 1.883699780217342,
+      "learning_rate": 5e-08,
+      "loss": 2.2059,
+      "step": 3910
+    },
+    {
+      "epoch": 3.9705882352941178,
+      "grad_norm": 1.7246634345468168,
+      "learning_rate": 5e-08,
+      "loss": 2.2482,
+      "step": 3915
+    },
+    {
+      "epoch": 3.975659229208925,
+      "grad_norm": 1.762677648630269,
+      "learning_rate": 5e-08,
+      "loss": 2.2521,
+      "step": 3920
+    },
+    {
+      "epoch": 3.9807302231237323,
+      "grad_norm": 1.786354894638501,
+      "learning_rate": 5e-08,
+      "loss": 2.2763,
+      "step": 3925
+    },
+    {
+      "epoch": 3.9858012170385395,
+      "grad_norm": 1.81100838850099,
+      "learning_rate": 5e-08,
+      "loss": 2.2326,
+      "step": 3930
+    },
+    {
+      "epoch": 3.9908722109533468,
+      "grad_norm": 1.8115971845880692,
+      "learning_rate": 5e-08,
+      "loss": 2.2409,
+      "step": 3935
+    },
+    {
+      "epoch": 3.995943204868154,
+      "grad_norm": 1.901268059775357,
+      "learning_rate": 5e-08,
+      "loss": 2.2217,
+      "step": 3940
+    },
+    {
+      "epoch": 4.0,
+      "step": 3944,
+      "total_flos": 411954472550400.0,
+      "train_loss": 2.318100369605283,
+      "train_runtime": 14372.236,
+      "train_samples_per_second": 17.546,
+      "train_steps_per_second": 0.274
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 3944,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 200,
+  "total_flos": 411954472550400.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}