diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25081 @@
+{
+  "best_metric": 1.599829912185669,
+  "best_model_checkpoint": "runs/deepseek_lora_/home/audrey/air2/runs/deepseek_lora_/home/audrey/air2/runs/deepseek_lora_/home/audrey/air2/runs/deepseek_lora_/home/audrey/air2/runs/deepseek_lora_20240420-031057/checkpoint-10000_20240420-061900/checkpoint-10000_20240420-141714/checkpoint-30000_20240421-001954/checkpoint-20000_20240421-063809/checkpoint-27000",
+  "epoch": 2.9166666666666665,
+  "eval_steps": 500,
+  "global_step": 35000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2827765941619873,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.2577,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5021369457244873,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.392,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0452425479888916,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.3256,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0641889572143555,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.3476,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.357707977294922,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.2706,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.017151355743408,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 2.125,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5771751403808594,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 2.2445,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8890554904937744,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 2.2334,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2498245239257812,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 2.2226,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.084829807281494,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.0497,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.884516716003418,
+      "learning_rate": 4.4e-06,
+      "loss": 2.0363,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7535431385040283,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 2.1011,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.637084722518921,
+      "learning_rate": 5.2e-06,
+      "loss": 2.0225,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6097259521484375,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.9933,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4520087242126465,
+      "learning_rate": 6e-06,
+      "loss": 2.0717,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.878129005432129,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 2.078,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6473491191864014,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.8528,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8662400245666504,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.9476,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6106956005096436,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.9378,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1151645183563232,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.8459,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.933845281600952,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.9992,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.989990472793579,
+      "learning_rate": 8.8e-06,
+      "loss": 1.9832,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.849925994873047,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.9316,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.09212875366211,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.9109,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.965531587600708,
+      "learning_rate": 1e-05,
+      "loss": 1.9034,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.259747266769409,
+      "learning_rate": 1.04e-05,
+      "loss": 1.9923,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.087090015411377,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 1.9494,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.418077230453491,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.8784,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9205658435821533,
+      "learning_rate": 1.16e-05,
+      "loss": 1.7895,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.845510959625244,
+      "learning_rate": 1.2e-05,
+      "loss": 1.8195,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.274329662322998,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 1.9255,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.376899242401123,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.9822,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.367288589477539,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 1.8333,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.786440849304199,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.8435,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.883002758026123,
+      "learning_rate": 1.4e-05,
+      "loss": 1.8751,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.210261583328247,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.8073,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7850048542022705,
+      "learning_rate": 1.48e-05,
+      "loss": 1.8612,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.726701259613037,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.9211,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.326230764389038,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 1.9087,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.577868938446045,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.0082,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0567996501922607,
+      "learning_rate": 1.64e-05,
+      "loss": 1.9232,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.262115478515625,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.9156,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7286055088043213,
+      "learning_rate": 1.72e-05,
+      "loss": 1.7268,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0500264167785645,
+      "learning_rate": 1.76e-05,
+      "loss": 1.9313,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8140387535095215,
+      "learning_rate": 1.8e-05,
+      "loss": 1.8847,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.911093235015869,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.6759,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.010791778564453,
+      "learning_rate": 1.88e-05,
+      "loss": 1.9112,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5784366130828857,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.9196,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.904146671295166,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.9137,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3381693363189697,
+      "learning_rate": 2e-05,
+      "loss": 1.8686,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 1.8578028678894043,
+      "eval_runtime": 107.4833,
+      "eval_samples_per_second": 9.304,
+      "eval_steps_per_second": 2.326,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6459054946899414,
+      "learning_rate": 1.9994202898550726e-05,
+      "loss": 1.8223,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.068203926086426,
+      "learning_rate": 1.998840579710145e-05,
+      "loss": 1.7466,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4843411445617676,
+      "learning_rate": 1.9982608695652174e-05,
+      "loss": 1.9179,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6055413484573364,
+      "learning_rate": 1.99768115942029e-05,
+      "loss": 1.9317,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9912831783294678,
+      "learning_rate": 1.9971014492753625e-05,
+      "loss": 1.9632,
+      "step": 550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7726389169692993,
+      "learning_rate": 1.996521739130435e-05,
+      "loss": 1.8542,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2357091903686523,
+      "learning_rate": 1.9959420289855073e-05,
+      "loss": 1.8574,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8849786520004272,
+      "learning_rate": 1.99536231884058e-05,
+      "loss": 1.8169,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.756426215171814,
+      "learning_rate": 1.9947826086956524e-05,
+      "loss": 1.8116,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3438060283660889,
+      "learning_rate": 1.994202898550725e-05,
+      "loss": 1.9238,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7150946855545044,
+      "learning_rate": 1.9936231884057972e-05,
+      "loss": 1.9171,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1640572547912598,
+      "learning_rate": 1.9930434782608696e-05,
+      "loss": 1.8279,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2766189575195312,
+      "learning_rate": 1.9924637681159424e-05,
+      "loss": 1.8898,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.365070819854736,
+      "learning_rate": 1.9918840579710144e-05,
+      "loss": 1.862,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1916489601135254,
+      "learning_rate": 1.9913043478260872e-05,
+      "loss": 1.8582,
+      "step": 650
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0066256523132324,
+      "learning_rate": 1.9907246376811596e-05,
+      "loss": 1.7583,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.382798671722412,
+      "learning_rate": 1.990144927536232e-05,
+      "loss": 1.8526,
+      "step": 670
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.918565273284912,
+      "learning_rate": 1.9895652173913044e-05,
+      "loss": 1.8112,
+      "step": 680
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8301721811294556,
+      "learning_rate": 1.988985507246377e-05,
+      "loss": 1.8805,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.482556104660034,
+      "learning_rate": 1.9884057971014495e-05,
+      "loss": 1.778,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.739922046661377,
+      "learning_rate": 1.987826086956522e-05,
+      "loss": 1.8817,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8400835990905762,
+      "learning_rate": 1.9872463768115943e-05,
+      "loss": 1.9506,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7623933553695679,
+      "learning_rate": 1.9866666666666667e-05,
+      "loss": 1.8342,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.191768169403076,
+      "learning_rate": 1.9860869565217395e-05,
+      "loss": 1.8549,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0716779232025146,
+      "learning_rate": 1.9855072463768115e-05,
+      "loss": 1.7888,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.673297643661499,
+      "learning_rate": 1.9849275362318843e-05,
+      "loss": 1.8287,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.45849609375,
+      "learning_rate": 1.9843478260869567e-05,
+      "loss": 1.8583,
+      "step": 770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.941355586051941,
+      "learning_rate": 1.983768115942029e-05,
+      "loss": 1.7381,
+      "step": 780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.047844886779785,
+      "learning_rate": 1.9831884057971015e-05,
+      "loss": 1.8842,
+      "step": 790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3743892908096313,
+      "learning_rate": 1.9826086956521742e-05,
+      "loss": 1.9104,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7510887384414673,
+      "learning_rate": 1.9820289855072466e-05,
+      "loss": 1.8223,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.888203740119934,
+      "learning_rate": 1.981449275362319e-05,
+      "loss": 1.7711,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0050196647644043,
+      "learning_rate": 1.9808695652173914e-05,
+      "loss": 1.756,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3720544576644897,
+      "learning_rate": 1.9802898550724638e-05,
+      "loss": 1.8173,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.654989242553711,
+      "learning_rate": 1.9797101449275366e-05,
+      "loss": 1.7568,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6716177463531494,
+      "learning_rate": 1.979130434782609e-05,
+      "loss": 1.9341,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.005023241043091,
+      "learning_rate": 1.9785507246376814e-05,
+      "loss": 1.9117,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9368737936019897,
+      "learning_rate": 1.9779710144927538e-05,
+      "loss": 1.6635,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.550541400909424,
+      "learning_rate": 1.9773913043478265e-05,
+      "loss": 1.8452,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.289937138557434,
+      "learning_rate": 1.9768115942028986e-05,
+      "loss": 1.8579,
+      "step": 900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2402184009552,
+      "learning_rate": 1.9762318840579713e-05,
+      "loss": 1.8988,
+      "step": 910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5080454349517822,
+      "learning_rate": 1.9756521739130437e-05,
+      "loss": 1.9026,
+      "step": 920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0446062088012695,
+      "learning_rate": 1.975072463768116e-05,
+      "loss": 1.8181,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.787261724472046,
+      "learning_rate": 1.9744927536231885e-05,
+      "loss": 1.8197,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6586248874664307,
+      "learning_rate": 1.973913043478261e-05,
+      "loss": 1.8774,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1193392276763916,
+      "learning_rate": 1.9733333333333336e-05,
+      "loss": 1.9113,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.299938917160034,
+      "learning_rate": 1.972753623188406e-05,
+      "loss": 1.8613,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.443349838256836,
+      "learning_rate": 1.9721739130434784e-05,
+      "loss": 1.8193,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3777451515197754,
+      "learning_rate": 1.971594202898551e-05,
+      "loss": 1.7872,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1168971061706543,
+      "learning_rate": 1.9710144927536236e-05,
+      "loss": 1.7879,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 1.8089497089385986,
+      "eval_runtime": 107.4884,
+      "eval_samples_per_second": 9.303,
+      "eval_steps_per_second": 2.326,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.817137360572815,
+      "learning_rate": 1.9704347826086956e-05,
+      "loss": 1.9,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.249089479446411,
+      "learning_rate": 1.969855072463768e-05,
+      "loss": 1.9053,
+      "step": 1020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.998011589050293,
+      "learning_rate": 1.9692753623188408e-05,
+      "loss": 1.8222,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.248562335968018,
+      "learning_rate": 1.9686956521739132e-05,
+      "loss": 1.7655,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.401443004608154,
+      "learning_rate": 1.9681159420289856e-05,
+      "loss": 1.9682,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3086190223693848,
+      "learning_rate": 1.967536231884058e-05,
+      "loss": 1.7689,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4858791828155518,
+      "learning_rate": 1.9669565217391307e-05,
+      "loss": 1.6351,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9979552030563354,
+      "learning_rate": 1.966376811594203e-05,
+      "loss": 1.8494,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.989320158958435,
+      "learning_rate": 1.9657971014492755e-05,
+      "loss": 1.7798,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.888331890106201,
+      "learning_rate": 1.965217391304348e-05,
+      "loss": 1.8133,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.666988730430603,
+      "learning_rate": 1.9646376811594207e-05,
+      "loss": 1.8949,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6761118173599243,
+      "learning_rate": 1.964057971014493e-05,
+      "loss": 1.9065,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6665714979171753,
+      "learning_rate": 1.963478260869565e-05,
+      "loss": 1.8475,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.513988971710205,
+      "learning_rate": 1.962898550724638e-05,
+      "loss": 1.6495,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7759839296340942,
+      "learning_rate": 1.9623188405797103e-05,
+      "loss": 1.7775,
+      "step": 1150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4246402978897095,
+      "learning_rate": 1.9617391304347827e-05,
+      "loss": 1.765,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7224540710449219,
+      "learning_rate": 1.961159420289855e-05,
+      "loss": 1.7599,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1602578163146973,
+      "learning_rate": 1.9605797101449278e-05,
+      "loss": 1.63,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.389148473739624,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.771,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9033640623092651,
+      "learning_rate": 1.9594202898550726e-05,
+      "loss": 1.8217,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8465853929519653,
+      "learning_rate": 1.958840579710145e-05,
+      "loss": 1.8974,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0622667074203491,
+      "learning_rate": 1.9582608695652177e-05,
+      "loss": 1.7574,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.802722692489624,
+      "learning_rate": 1.95768115942029e-05,
+      "loss": 1.7193,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6129810810089111,
+      "learning_rate": 1.9571014492753625e-05,
+      "loss": 1.8253,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9575114250183105,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 1.8882,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4238998889923096,
+      "learning_rate": 1.9559420289855074e-05,
+      "loss": 1.7379,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9338934421539307,
+      "learning_rate": 1.9553623188405798e-05,
+      "loss": 1.7347,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6793437004089355,
+      "learning_rate": 1.954782608695652e-05,
+      "loss": 1.9141,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.266021728515625,
+      "learning_rate": 1.954202898550725e-05,
+      "loss": 1.8423,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8466399908065796,
+      "learning_rate": 1.9536231884057973e-05,
+      "loss": 1.8266,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8755276203155518,
+      "learning_rate": 1.9530434782608697e-05,
+      "loss": 1.9117,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3602133989334106,
+      "learning_rate": 1.952463768115942e-05,
+      "loss": 1.7265,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.353096842765808,
+      "learning_rate": 1.9518840579710145e-05,
+      "loss": 1.8355,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.312878966331482,
+      "learning_rate": 1.9513043478260872e-05,
+      "loss": 1.7331,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.431149959564209,
+      "learning_rate": 1.9507246376811596e-05,
+      "loss": 1.7316,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3108808994293213,
+      "learning_rate": 1.950144927536232e-05,
+      "loss": 1.8539,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.86793851852417,
+      "learning_rate": 1.9495652173913044e-05,
+      "loss": 1.8277,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5965479612350464,
+      "learning_rate": 1.9489855072463772e-05,
+      "loss": 1.7548,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.187406063079834,
+      "learning_rate": 1.9484057971014492e-05,
+      "loss": 1.8396,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5422654151916504,
+      "learning_rate": 1.947826086956522e-05,
+      "loss": 1.7876,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.088440418243408,
+      "learning_rate": 1.9472463768115944e-05,
+      "loss": 1.8558,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8927093744277954,
+      "learning_rate": 1.9466666666666668e-05,
+      "loss": 1.8849,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.945380687713623,
+      "learning_rate": 1.9460869565217392e-05,
+      "loss": 1.7351,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.608578681945801,
+      "learning_rate": 1.9455072463768116e-05,
+      "loss": 1.7438,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4089462757110596,
+      "learning_rate": 1.9449275362318843e-05,
+      "loss": 1.8227,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.93408203125,
+      "learning_rate": 1.9443478260869567e-05,
+      "loss": 1.8358,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.577270269393921,
+      "learning_rate": 1.943768115942029e-05,
+      "loss": 1.8707,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3741328716278076,
+      "learning_rate": 1.9431884057971015e-05,
+      "loss": 1.764,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4061601161956787,
+      "learning_rate": 1.9426086956521743e-05,
+      "loss": 1.8593,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7175543308258057,
+      "learning_rate": 1.9420289855072467e-05,
+      "loss": 1.9067,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 1.7743674516677856,
+      "eval_runtime": 107.4624,
+      "eval_samples_per_second": 9.306,
+      "eval_steps_per_second": 2.326,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.384735107421875,
+      "learning_rate": 1.941449275362319e-05,
+      "loss": 1.6986,
+      "step": 1510
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8683741092681885,
+      "learning_rate": 1.9408695652173915e-05,
+      "loss": 1.8999,
+      "step": 1520
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.08429217338562,
+      "learning_rate": 1.9402898550724642e-05,
+      "loss": 1.8664,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.030617117881775,
+      "learning_rate": 1.9397101449275363e-05,
+      "loss": 1.5804,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.734713315963745,
+      "learning_rate": 1.9391304347826087e-05,
+      "loss": 1.8053,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2831270694732666,
+      "learning_rate": 1.9385507246376814e-05,
+      "loss": 1.7894,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.505608558654785,
+      "learning_rate": 1.9379710144927538e-05,
+      "loss": 1.6539,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0595582723617554,
+      "learning_rate": 1.9373913043478262e-05,
+      "loss": 1.8655,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5555408000946045,
+      "learning_rate": 1.9368115942028986e-05,
+      "loss": 1.763,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7931909561157227,
+      "learning_rate": 1.9362318840579713e-05,
+      "loss": 1.8121,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.358292579650879,
+      "learning_rate": 1.9356521739130437e-05,
+      "loss": 1.8876,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.060112953186035,
+      "learning_rate": 1.935072463768116e-05,
+      "loss": 1.7779,
+      "step": 1620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.94615638256073,
+      "learning_rate": 1.9344927536231885e-05,
+      "loss": 1.7276,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.230520009994507,
+      "learning_rate": 1.933913043478261e-05,
+      "loss": 1.7617,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.058115243911743,
+      "learning_rate": 1.9333333333333333e-05,
+      "loss": 1.8117,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2918510437011719,
+      "learning_rate": 1.9327536231884057e-05,
+      "loss": 1.6933,
+      "step": 1660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8761500120162964,
+      "learning_rate": 1.9321739130434785e-05,
+      "loss": 1.8075,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5313031673431396,
+      "learning_rate": 1.931594202898551e-05,
+      "loss": 1.6154,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8059382438659668,
+      "learning_rate": 1.9310144927536233e-05,
+      "loss": 1.7508,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9243950843811035,
+      "learning_rate": 1.9304347826086957e-05,
+      "loss": 1.832,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.17168664932251,
+      "learning_rate": 1.9298550724637684e-05,
+      "loss": 1.8438,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7666871547698975,
+      "learning_rate": 1.9292753623188408e-05,
+      "loss": 1.8314,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.743995666503906,
+      "learning_rate": 1.9286956521739132e-05,
+      "loss": 1.7588,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5909578800201416,
+      "learning_rate": 1.9281159420289856e-05,
+      "loss": 1.839,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.395579218864441,
+      "learning_rate": 1.927536231884058e-05,
+      "loss": 1.6756,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7446272373199463,
+      "learning_rate": 1.9269565217391308e-05,
+      "loss": 1.8533,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.249009132385254,
+      "learning_rate": 1.9263768115942028e-05,
+      "loss": 1.8051,
+      "step": 1770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8617609143257141,
+      "learning_rate": 1.9257971014492756e-05,
+      "loss": 1.7813,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3332220315933228,
+      "learning_rate": 1.925217391304348e-05,
+      "loss": 1.7937,
+      "step": 1790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5308120250701904,
+      "learning_rate": 1.9246376811594204e-05,
+      "loss": 1.7395,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.629775047302246,
+      "learning_rate": 1.9240579710144928e-05,
+      "loss": 1.7794,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5039477348327637,
+      "learning_rate": 1.9234782608695655e-05,
+      "loss": 1.7868,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4312188625335693,
+      "learning_rate": 1.922898550724638e-05,
+      "loss": 1.7787,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.299351930618286,
+      "learning_rate": 1.9223188405797103e-05,
+      "loss": 1.5264,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.888294219970703,
+      "learning_rate": 1.9217391304347827e-05,
+      "loss": 1.7851,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4942209720611572,
+      "learning_rate": 1.921159420289855e-05,
+      "loss": 1.8019,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9073644876480103,
+      "learning_rate": 1.920579710144928e-05,
+      "loss": 1.8111,
+      "step": 1870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.672957420349121,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.8331,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.514133095741272,
+      "learning_rate": 1.9194202898550727e-05,
+      "loss": 1.7383,
+      "step": 1890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5273349285125732,
+      "learning_rate": 1.918840579710145e-05,
+      "loss": 1.7974,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.798083782196045,
+      "learning_rate": 1.9182608695652175e-05,
+      "loss": 1.7457,
+      "step": 1910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.434926509857178,
+      "learning_rate": 1.91768115942029e-05,
+      "loss": 1.7104,
+      "step": 1920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5063635110855103,
+      "learning_rate": 1.9171014492753626e-05,
+      "loss": 1.7753,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0606720447540283,
+      "learning_rate": 1.916521739130435e-05,
+      "loss": 1.7012,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9884568452835083,
+      "learning_rate": 1.9159420289855074e-05,
+      "loss": 1.7646,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3289304971694946,
+      "learning_rate": 1.9153623188405798e-05,
+      "loss": 1.8317,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8612945079803467,
+      "learning_rate": 1.9147826086956522e-05,
+      "loss": 1.8826,
+      "step": 1970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.970140218734741,
+      "learning_rate": 1.914202898550725e-05,
+      "loss": 1.9036,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.007293939590454,
+      "learning_rate": 1.9136231884057973e-05,
+      "loss": 1.7557,
+      "step": 1990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1491663455963135,
+      "learning_rate": 1.9130434782608697e-05,
+      "loss": 1.7295,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 1.7665390968322754,
+      "eval_runtime": 107.4574,
+      "eval_samples_per_second": 9.306,
+      "eval_steps_per_second": 2.327,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.889491319656372,
+      "learning_rate": 1.912463768115942e-05,
+      "loss": 1.7628,
+      "step": 2010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.507429838180542,
+      "learning_rate": 1.911884057971015e-05,
+      "loss": 1.7585,
+      "step": 2020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2127914428710938,
+      "learning_rate": 1.911304347826087e-05,
+      "loss": 1.8121,
+      "step": 2030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9668962955474854,
+      "learning_rate": 1.9107246376811597e-05,
+      "loss": 1.8563,
+      "step": 2040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9941543936729431,
+      "learning_rate": 1.910144927536232e-05,
+      "loss": 1.8258,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4788460731506348,
+      "learning_rate": 1.9095652173913045e-05,
+      "loss": 1.7617,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.523179769515991,
+      "learning_rate": 1.908985507246377e-05,
+      "loss": 1.6908,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2159104347229004,
+      "learning_rate": 1.9084057971014493e-05,
+      "loss": 1.6652,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2359981536865234,
+      "learning_rate": 1.907826086956522e-05,
+      "loss": 1.8826,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3493943214416504,
+      "learning_rate": 1.9072463768115944e-05,
+      "loss": 1.8261,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5862370729446411,
+      "learning_rate": 1.9066666666666668e-05,
+      "loss": 1.6941,
+      "step": 2110
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1425403356552124,
+      "learning_rate": 1.9060869565217392e-05,
+      "loss": 1.8313,
+      "step": 2120
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.16150426864624,
+      "learning_rate": 1.905507246376812e-05,
+      "loss": 1.7649,
+      "step": 2130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.25124192237854,
+      "learning_rate": 1.9049275362318844e-05,
+      "loss": 1.8382,
+      "step": 2140
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4652185440063477,
+      "learning_rate": 1.9043478260869568e-05,
+      "loss": 1.8486,
+      "step": 2150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7186965942382812,
+      "learning_rate": 1.903768115942029e-05,
+      "loss": 1.8783,
+      "step": 2160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2005832195281982,
+      "learning_rate": 1.9031884057971016e-05,
+      "loss": 1.8612,
+      "step": 2170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.344748020172119,
+      "learning_rate": 1.902608695652174e-05,
+      "loss": 1.6858,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9262315034866333,
+      "learning_rate": 1.9020289855072464e-05,
+      "loss": 1.7719,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.292480945587158,
+      "learning_rate": 1.901449275362319e-05,
+      "loss": 1.883,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0080437660217285,
+      "learning_rate": 1.9008695652173915e-05,
+      "loss": 1.7567,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.655610203742981,
+      "learning_rate": 1.900289855072464e-05,
+      "loss": 1.7931,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8952635526657104,
+      "learning_rate": 1.8997101449275363e-05,
+      "loss": 1.7877,
+      "step": 2230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0049967765808105,
+      "learning_rate": 1.899130434782609e-05,
+      "loss": 1.8179,
+      "step": 2240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.809584379196167,
+      "learning_rate": 1.8985507246376814e-05,
+      "loss": 1.7893,
+      "step": 2250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3408210277557373,
+      "learning_rate": 1.8979710144927535e-05,
+      "loss": 1.76,
+      "step": 2260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.576713800430298,
+      "learning_rate": 1.8973913043478262e-05,
+      "loss": 1.6123,
+      "step": 2270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.999994993209839,
+      "learning_rate": 1.8968115942028986e-05,
+      "loss": 1.748,
+      "step": 2280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8222105503082275,
+      "learning_rate": 1.896231884057971e-05,
+      "loss": 1.6274,
+      "step": 2290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.565905213356018,
+      "learning_rate": 1.8956521739130434e-05,
+      "loss": 1.7498,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2533594369888306,
+      "learning_rate": 1.8950724637681162e-05,
+      "loss": 1.8822,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8874733448028564,
+      "learning_rate": 1.8944927536231886e-05,
+      "loss": 1.7679,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1580071449279785,
+      "learning_rate": 1.893913043478261e-05,
+      "loss": 1.829,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6889914274215698,
+      "learning_rate": 1.8933333333333334e-05,
+      "loss": 1.7765,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.708866000175476,
+      "learning_rate": 1.892753623188406e-05,
+      "loss": 1.7612,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0311877727508545,
+      "learning_rate": 1.8921739130434785e-05,
+      "loss": 1.8271,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4897469282150269,
+      "learning_rate": 1.891594202898551e-05,
+      "loss": 1.67,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0771639347076416,
+      "learning_rate": 1.8910144927536233e-05,
+      "loss": 1.8175,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.115084171295166,
+      "learning_rate": 1.8904347826086957e-05,
+      "loss": 1.8247,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.106081008911133,
+      "learning_rate": 1.8898550724637685e-05,
+      "loss": 1.833,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4720683097839355,
+      "learning_rate": 1.8892753623188405e-05,
+      "loss": 1.8494,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4541406631469727,
+      "learning_rate": 1.8886956521739133e-05,
+      "loss": 1.8464,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.226954698562622,
+      "learning_rate": 1.8881159420289857e-05,
+      "loss": 1.8729,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2664247751235962,
+      "learning_rate": 1.887536231884058e-05,
+      "loss": 1.8722,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0264010429382324,
+      "learning_rate": 1.8869565217391305e-05,
+      "loss": 1.7936,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.336003303527832,
+      "learning_rate": 1.8863768115942032e-05,
+      "loss": 1.9073,
+      "step": 2460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.730409622192383,
+      "learning_rate": 1.8857971014492756e-05,
+      "loss": 1.9041,
+      "step": 2470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9845330715179443,
+      "learning_rate": 1.885217391304348e-05,
+      "loss": 1.7753,
+      "step": 2480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5443974733352661,
+      "learning_rate": 1.8846376811594204e-05,
+      "loss": 1.7274,
+      "step": 2490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1884684562683105,
+      "learning_rate": 1.8840579710144928e-05,
+      "loss": 1.8284,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 1.7748754024505615,
+      "eval_runtime": 107.4612,
+      "eval_samples_per_second": 9.306,
+      "eval_steps_per_second": 2.326,
+      "step": 2500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1604175567626953,
+      "learning_rate": 1.8834782608695656e-05,
+      "loss": 1.7344,
+      "step": 2510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7605400085449219,
+      "learning_rate": 1.882898550724638e-05,
+      "loss": 1.7241,
+      "step": 2520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8347537517547607,
+      "learning_rate": 1.8823188405797104e-05,
+      "loss": 1.7196,
+      "step": 2530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.828438639640808,
+      "learning_rate": 1.8817391304347828e-05,
+      "loss": 1.8529,
+      "step": 2540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.69232976436615,
+      "learning_rate": 1.881159420289855e-05,
+      "loss": 1.7768,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.592120885848999,
+      "learning_rate": 1.8805797101449276e-05,
+      "loss": 1.7074,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.56288743019104,
+      "learning_rate": 1.88e-05,
+      "loss": 1.8631,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6752883791923523,
+      "learning_rate": 1.8794202898550727e-05,
+      "loss": 1.8049,
+      "step": 2580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.010446071624756,
+      "learning_rate": 1.878840579710145e-05,
+      "loss": 1.7486,
+      "step": 2590
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9133752584457397,
+      "learning_rate": 1.8782608695652175e-05,
+      "loss": 1.7831,
+      "step": 2600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1954097747802734,
+      "learning_rate": 1.87768115942029e-05,
+      "loss": 1.7329,
+      "step": 2610
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0870425701141357,
+      "learning_rate": 1.8771014492753626e-05,
+      "loss": 1.7143,
+      "step": 2620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.162560224533081,
+      "learning_rate": 1.876521739130435e-05,
+      "loss": 1.8418,
+      "step": 2630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2718247175216675,
+      "learning_rate": 1.8759420289855074e-05,
+      "loss": 1.7022,
+      "step": 2640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4909746646881104,
+      "learning_rate": 1.87536231884058e-05,
+      "loss": 1.8087,
+      "step": 2650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.101371765136719,
+      "learning_rate": 1.8747826086956526e-05,
+      "loss": 1.8178,
+      "step": 2660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9989445209503174,
+      "learning_rate": 1.8742028985507246e-05,
+      "loss": 1.7954,
+      "step": 2670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0736329555511475,
+      "learning_rate": 1.873623188405797e-05,
+      "loss": 1.8488,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.652923822402954,
+      "learning_rate": 1.8730434782608698e-05,
+      "loss": 1.7438,
+      "step": 2690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.146462917327881,
+      "learning_rate": 1.8724637681159422e-05,
+      "loss": 1.8169,
+      "step": 2700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5568904876708984,
+      "learning_rate": 1.8718840579710146e-05,
+      "loss": 1.7904,
+      "step": 2710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2244021892547607,
+      "learning_rate": 1.871304347826087e-05,
+      "loss": 1.8384,
+      "step": 2720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3431702852249146,
+      "learning_rate": 1.8707246376811597e-05,
+      "loss": 1.8376,
+      "step": 2730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.871310234069824,
+      "learning_rate": 1.870144927536232e-05,
+      "loss": 1.6912,
+      "step": 2740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3337674140930176,
+      "learning_rate": 1.8695652173913045e-05,
+      "loss": 1.7164,
+      "step": 2750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.844233751296997,
+      "learning_rate": 1.868985507246377e-05,
+      "loss": 1.7873,
+      "step": 2760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0465248823165894,
+      "learning_rate": 1.8684057971014497e-05,
+      "loss": 1.5893,
+      "step": 2770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0744643211364746,
+      "learning_rate": 1.867826086956522e-05,
+      "loss": 1.8043,
+      "step": 2780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2488594055175781,
+      "learning_rate": 1.867246376811594e-05,
+      "loss": 1.7215,
+      "step": 2790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.04681658744812,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 1.901,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6009609699249268,
+      "learning_rate": 1.8660869565217393e-05,
+      "loss": 1.7446,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.019435167312622,
+      "learning_rate": 1.8655072463768117e-05,
+      "loss": 1.8343,
+      "step": 2820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.385256290435791,
+      "learning_rate": 1.864927536231884e-05,
+      "loss": 1.6675,
+      "step": 2830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.222172737121582,
+      "learning_rate": 1.8643478260869568e-05,
+      "loss": 1.6239,
+      "step": 2840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.963768005371094,
+      "learning_rate": 1.8637681159420292e-05,
+      "loss": 1.711,
+      "step": 2850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1189117431640625,
+      "learning_rate": 1.8631884057971016e-05,
+      "loss": 1.886,
+      "step": 2860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5243103504180908,
+      "learning_rate": 1.862608695652174e-05,
+      "loss": 1.8398,
+      "step": 2870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8455665111541748,
+      "learning_rate": 1.8620289855072464e-05,
+      "loss": 1.8868,
+      "step": 2880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1385709047317505,
+      "learning_rate": 1.861449275362319e-05,
+      "loss": 1.8302,
+      "step": 2890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1462037563323975,
+      "learning_rate": 1.8608695652173912e-05,
+      "loss": 1.8159,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0207372903823853,
+      "learning_rate": 1.860289855072464e-05,
+      "loss": 1.8887,
+      "step": 2910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9228794574737549,
+      "learning_rate": 1.8597101449275363e-05,
+      "loss": 1.801,
+      "step": 2920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8184621334075928,
+      "learning_rate": 1.8591304347826087e-05,
+      "loss": 1.7667,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6740689277648926,
+      "learning_rate": 1.858550724637681e-05,
+      "loss": 1.8734,
+      "step": 2940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6178441047668457,
+      "learning_rate": 1.857971014492754e-05,
+      "loss": 1.7861,
+      "step": 2950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4509289264678955,
+      "learning_rate": 1.8573913043478263e-05,
+      "loss": 1.7024,
+      "step": 2960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.4058709144592285,
+      "learning_rate": 1.8568115942028987e-05,
+      "loss": 1.7561,
+      "step": 2970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0747387409210205,
+      "learning_rate": 1.856231884057971e-05,
+      "loss": 1.8715,
+      "step": 2980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.675367832183838,
+      "learning_rate": 1.8556521739130435e-05,
+      "loss": 1.6404,
+      "step": 2990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9367339611053467,
+      "learning_rate": 1.8550724637681162e-05,
+      "loss": 1.7286,
+      "step": 3000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.7500901222229004,
+      "eval_runtime": 107.4483,
+      "eval_samples_per_second": 9.307,
+      "eval_steps_per_second": 2.327,
+      "step": 3000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8943095207214355,
+      "learning_rate": 1.8544927536231886e-05,
+      "loss": 1.6788,
+      "step": 3010
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3797285556793213,
+      "learning_rate": 1.853913043478261e-05,
+      "loss": 1.8931,
+      "step": 3020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7746102809906006,
+      "learning_rate": 1.8533333333333334e-05,
+      "loss": 1.7836,
+      "step": 3030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9604554176330566,
+      "learning_rate": 1.8527536231884062e-05,
+      "loss": 1.704,
+      "step": 3040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5160584449768066,
+      "learning_rate": 1.8521739130434782e-05,
+      "loss": 1.7964,
+      "step": 3050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3620455265045166,
+      "learning_rate": 1.851594202898551e-05,
+      "loss": 1.7291,
+      "step": 3060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6815402507781982,
+      "learning_rate": 1.8510144927536234e-05,
+      "loss": 1.744,
+      "step": 3070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7348963022232056,
+      "learning_rate": 1.8504347826086958e-05,
+      "loss": 1.6285,
+      "step": 3080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5644665956497192,
+      "learning_rate": 1.8498550724637682e-05,
+      "loss": 1.6034,
+      "step": 3090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5588579177856445,
+      "learning_rate": 1.8492753623188406e-05,
+      "loss": 1.6455,
+      "step": 3100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.486201763153076,
+      "learning_rate": 1.8486956521739133e-05,
+      "loss": 1.6324,
+      "step": 3110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.734580039978027,
+      "learning_rate": 1.8481159420289857e-05,
+      "loss": 1.7564,
+      "step": 3120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.060638427734375,
+      "learning_rate": 1.847536231884058e-05,
+      "loss": 1.8093,
+      "step": 3130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.396253228187561,
+      "learning_rate": 1.8469565217391305e-05,
+      "loss": 1.8422,
+      "step": 3140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.647871494293213,
+      "learning_rate": 1.8463768115942033e-05,
+      "loss": 1.692,
+      "step": 3150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.792811870574951,
+      "learning_rate": 1.8457971014492753e-05,
+      "loss": 1.8721,
+      "step": 3160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6648412942886353,
+      "learning_rate": 1.845217391304348e-05,
+      "loss": 1.8176,
+      "step": 3170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2956972122192383,
+      "learning_rate": 1.8446376811594205e-05,
+      "loss": 1.7783,
+      "step": 3180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.968624472618103,
+      "learning_rate": 1.844057971014493e-05,
+      "loss": 1.75,
+      "step": 3190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.384540557861328,
+      "learning_rate": 1.8434782608695653e-05,
+      "loss": 1.8703,
+      "step": 3200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.963712751865387,
+      "learning_rate": 1.8428985507246377e-05,
+      "loss": 1.8838,
+      "step": 3210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5194199085235596,
+      "learning_rate": 1.8423188405797104e-05,
+      "loss": 1.7351,
+      "step": 3220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.223130226135254,
+      "learning_rate": 1.8417971014492754e-05,
+      "loss": 1.7526,
+      "step": 3230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9289439916610718,
+      "learning_rate": 1.841217391304348e-05,
+      "loss": 1.8648,
+      "step": 3240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.942191481590271,
+      "learning_rate": 1.8406376811594205e-05,
+      "loss": 1.7607,
+      "step": 3250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.61818265914917,
+      "learning_rate": 1.840057971014493e-05,
+      "loss": 1.6653,
+      "step": 3260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.776350736618042,
+      "learning_rate": 1.8394782608695653e-05,
+      "loss": 1.7829,
+      "step": 3270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.71359384059906,
+      "learning_rate": 1.8388985507246377e-05,
+      "loss": 1.8434,
+      "step": 3280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9720557928085327,
+      "learning_rate": 1.8383188405797104e-05,
+      "loss": 1.749,
+      "step": 3290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1935908794403076,
+      "learning_rate": 1.837739130434783e-05,
+      "loss": 1.8992,
+      "step": 3300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.199913024902344,
+      "learning_rate": 1.8371594202898552e-05,
+      "loss": 1.8549,
+      "step": 3310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7523651123046875,
+      "learning_rate": 1.8365797101449276e-05,
+      "loss": 1.8814,
+      "step": 3320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2951200008392334,
+      "learning_rate": 1.8360000000000004e-05,
+      "loss": 1.8184,
+      "step": 3330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8267765045166016,
+      "learning_rate": 1.8354202898550724e-05,
+      "loss": 1.8589,
+      "step": 3340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0535434484481812,
+      "learning_rate": 1.8348405797101452e-05,
+      "loss": 1.609,
+      "step": 3350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.798702716827393,
+      "learning_rate": 1.8342608695652176e-05,
+      "loss": 1.7119,
+      "step": 3360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.116357326507568,
+      "learning_rate": 1.83368115942029e-05,
+      "loss": 1.7238,
+      "step": 3370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4448415040969849,
+      "learning_rate": 1.8331014492753624e-05,
+      "loss": 1.6308,
+      "step": 3380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.775078296661377,
+      "learning_rate": 1.8325217391304348e-05,
+      "loss": 1.8911,
+      "step": 3390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1946187019348145,
+      "learning_rate": 1.8319420289855075e-05,
+      "loss": 1.7165,
+      "step": 3400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4208180904388428,
+      "learning_rate": 1.83136231884058e-05,
+      "loss": 1.851,
+      "step": 3410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1975769996643066,
+      "learning_rate": 1.8307826086956523e-05,
+      "loss": 1.7533,
+      "step": 3420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.09836483001709,
+      "learning_rate": 1.8302028985507247e-05,
+      "loss": 1.7399,
+      "step": 3430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.047780990600586,
+      "learning_rate": 1.8296231884057975e-05,
+      "loss": 1.6752,
+      "step": 3440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.692073106765747,
+      "learning_rate": 1.82904347826087e-05,
+      "loss": 1.649,
+      "step": 3450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.907693862915039,
+      "learning_rate": 1.8284637681159423e-05,
+      "loss": 1.7774,
+      "step": 3460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.013425350189209,
+      "learning_rate": 1.8278840579710147e-05,
+      "loss": 1.8186,
+      "step": 3470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.129022479057312,
+      "learning_rate": 1.827304347826087e-05,
+      "loss": 1.827,
+      "step": 3480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.979015350341797,
+      "learning_rate": 1.8267246376811595e-05,
+      "loss": 1.6826,
+      "step": 3490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.317847967147827,
+      "learning_rate": 1.826144927536232e-05,
+      "loss": 1.8459,
+      "step": 3500
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 1.7489032745361328,
+      "eval_runtime": 107.4616,
+      "eval_samples_per_second": 9.306,
+      "eval_steps_per_second": 2.326,
+      "step": 3500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7150217294692993,
+      "learning_rate": 1.8255652173913046e-05,
+      "loss": 1.734,
+      "step": 3510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6586185693740845,
+      "learning_rate": 1.824985507246377e-05,
+      "loss": 1.7426,
+      "step": 3520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2894079685211182,
+      "learning_rate": 1.8244057971014494e-05,
+      "loss": 1.7626,
+      "step": 3530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0136561393737793,
+      "learning_rate": 1.8238260869565218e-05,
+      "loss": 1.7033,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7215077877044678,
+      "learning_rate": 1.8232463768115945e-05,
+      "loss": 1.8255,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4265096187591553,
+      "learning_rate": 1.822666666666667e-05,
+      "loss": 1.7646,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2528915405273438,
+      "learning_rate": 1.822086956521739e-05,
+      "loss": 1.809,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1465145349502563,
+      "learning_rate": 1.8215072463768117e-05,
+      "loss": 1.7062,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.697144865989685,
+      "learning_rate": 1.820927536231884e-05,
+      "loss": 1.7734,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9666852951049805,
+      "learning_rate": 1.8203478260869565e-05,
+      "loss": 1.751,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8137166500091553,
+      "learning_rate": 1.819768115942029e-05,
+      "loss": 1.7801,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7446413040161133,
+      "learning_rate": 1.8191884057971017e-05,
+      "loss": 1.7313,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2775075435638428,
+      "learning_rate": 1.818608695652174e-05,
+      "loss": 1.7378,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.90742027759552,
+      "learning_rate": 1.8180289855072465e-05,
+      "loss": 1.8367,
+      "step": 3640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5931909084320068,
+      "learning_rate": 1.817449275362319e-05,
+      "loss": 1.7352,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.411388397216797,
+      "learning_rate": 1.8168695652173916e-05,
+      "loss": 1.8184,
+      "step": 3660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2610573768615723,
+      "learning_rate": 1.816289855072464e-05,
+      "loss": 1.8392,
+      "step": 3670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2999207973480225,
+      "learning_rate": 1.8157101449275364e-05,
+      "loss": 1.7333,
+      "step": 3680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7404083013534546,
+      "learning_rate": 1.815130434782609e-05,
+      "loss": 1.734,
+      "step": 3690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5067249536514282,
+      "learning_rate": 1.8145507246376812e-05,
+      "loss": 1.7091,
+      "step": 3700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4666568040847778,
+      "learning_rate": 1.813971014492754e-05,
+      "loss": 1.6971,
+      "step": 3710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8228498697280884,
+      "learning_rate": 1.813391304347826e-05,
+      "loss": 1.6056,
+      "step": 3720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.056992530822754,
+      "learning_rate": 1.8128115942028988e-05,
+      "loss": 1.7473,
+      "step": 3730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1190054416656494,
+      "learning_rate": 1.8122318840579712e-05,
+      "loss": 1.7543,
+      "step": 3740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7520599365234375,
+      "learning_rate": 1.8116521739130436e-05,
+      "loss": 1.8603,
+      "step": 3750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.523573637008667,
+      "learning_rate": 1.811072463768116e-05,
+      "loss": 1.8091,
+      "step": 3760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5726646184921265,
+      "learning_rate": 1.8104927536231887e-05,
+      "loss": 1.728,
+      "step": 3770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.465932846069336,
+      "learning_rate": 1.809913043478261e-05,
+      "loss": 1.5628,
+      "step": 3780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4633328914642334,
+      "learning_rate": 1.8093333333333335e-05,
+      "loss": 1.6955,
+      "step": 3790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3645209074020386,
+      "learning_rate": 1.808753623188406e-05,
+      "loss": 1.8479,
+      "step": 3800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8331495523452759,
+      "learning_rate": 1.8081739130434783e-05,
+      "loss": 1.7356,
+      "step": 3810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2319082021713257,
+      "learning_rate": 1.807594202898551e-05,
+      "loss": 1.6966,
+      "step": 3820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.204197645187378,
+      "learning_rate": 1.807014492753623e-05,
+      "loss": 1.8275,
+      "step": 3830
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6139984130859375,
+      "learning_rate": 1.806434782608696e-05,
+      "loss": 1.73,
+      "step": 3840
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1955862045288086,
+      "learning_rate": 1.8058550724637683e-05,
+      "loss": 1.7662,
+      "step": 3850
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9615164399147034,
+      "learning_rate": 1.8052753623188407e-05,
+      "loss": 1.8208,
+      "step": 3860
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.718482494354248,
+      "learning_rate": 1.804695652173913e-05,
+      "loss": 1.7241,
+      "step": 3870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.189242124557495,
+      "learning_rate": 1.8041159420289855e-05,
+      "loss": 1.8602,
+      "step": 3880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.950349807739258,
+      "learning_rate": 1.8035362318840582e-05,
+      "loss": 1.866,
+      "step": 3890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.566131591796875,
+      "learning_rate": 1.8029565217391306e-05,
+      "loss": 1.528,
+      "step": 3900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.409193515777588,
+      "learning_rate": 1.802376811594203e-05,
+      "loss": 1.7897,
+      "step": 3910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5493619441986084,
+      "learning_rate": 1.8017971014492754e-05,
+      "loss": 1.8192,
+      "step": 3920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.087979793548584,
+      "learning_rate": 1.801217391304348e-05,
+      "loss": 1.7661,
+      "step": 3930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.98085355758667,
+      "learning_rate": 1.8006376811594205e-05,
+      "loss": 1.7836,
+      "step": 3940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6179476976394653,
+      "learning_rate": 1.800057971014493e-05,
+      "loss": 1.8128,
+      "step": 3950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0330349206924438,
+      "learning_rate": 1.7994782608695653e-05,
+      "loss": 1.8616,
+      "step": 3960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4651768207550049,
+      "learning_rate": 1.798898550724638e-05,
+      "loss": 1.8311,
+      "step": 3970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7031122446060181,
+      "learning_rate": 1.79831884057971e-05,
+      "loss": 1.6809,
+      "step": 3980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.734285831451416,
+      "learning_rate": 1.7977391304347825e-05,
+      "loss": 1.7419,
+      "step": 3990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.079700469970703,
+      "learning_rate": 1.7971594202898553e-05,
+      "loss": 1.7631,
+      "step": 4000
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 1.7669097185134888,
+      "eval_runtime": 107.4754,
+      "eval_samples_per_second": 9.304,
+      "eval_steps_per_second": 2.326,
+      "step": 4000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9647717475891113,
+      "learning_rate": 1.7965797101449277e-05,
+      "loss": 1.723,
+      "step": 4010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0198094844818115,
+      "learning_rate": 1.796e-05,
+      "loss": 1.5237,
+      "step": 4020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.165492296218872,
+      "learning_rate": 1.7954202898550725e-05,
+      "loss": 1.7683,
+      "step": 4030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6819276809692383,
+      "learning_rate": 1.7948405797101452e-05,
+      "loss": 1.7101,
+      "step": 4040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1148452758789062,
+      "learning_rate": 1.7942608695652176e-05,
+      "loss": 1.6593,
+      "step": 4050
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5147534608840942,
+      "learning_rate": 1.79368115942029e-05,
+      "loss": 1.599,
+      "step": 4060
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.868163824081421,
+      "learning_rate": 1.7931014492753624e-05,
+      "loss": 1.7865,
+      "step": 4070
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7418978214263916,
+      "learning_rate": 1.792521739130435e-05,
+      "loss": 1.7432,
+      "step": 4080
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6911894083023071,
+      "learning_rate": 1.7919420289855076e-05,
+      "loss": 1.8255,
+      "step": 4090
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.916471004486084,
+      "learning_rate": 1.7913623188405796e-05,
+      "loss": 1.8335,
+      "step": 4100
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2466706037521362,
+      "learning_rate": 1.7907826086956524e-05,
+      "loss": 1.8451,
+      "step": 4110
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5247670412063599,
+      "learning_rate": 1.7902028985507248e-05,
+      "loss": 1.7443,
+      "step": 4120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.580941081047058,
+      "learning_rate": 1.789623188405797e-05,
+      "loss": 1.8384,
+      "step": 4130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.8174824714660645,
+      "learning_rate": 1.7890434782608696e-05,
+      "loss": 1.6805,
+      "step": 4140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.682310938835144,
+      "learning_rate": 1.7884637681159423e-05,
+      "loss": 1.7765,
+      "step": 4150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.392796039581299,
+      "learning_rate": 1.7878840579710147e-05,
+      "loss": 1.6769,
+      "step": 4160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4888694286346436,
+      "learning_rate": 1.787304347826087e-05,
+      "loss": 1.7414,
+      "step": 4170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9246952533721924,
+      "learning_rate": 1.7867246376811595e-05,
+      "loss": 1.7297,
+      "step": 4180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4315427541732788,
+      "learning_rate": 1.786144927536232e-05,
+      "loss": 1.7288,
+      "step": 4190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2288168668746948,
+      "learning_rate": 1.7855652173913046e-05,
+      "loss": 1.9408,
+      "step": 4200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8101353645324707,
+      "learning_rate": 1.7849855072463767e-05,
+      "loss": 1.7743,
+      "step": 4210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.800456166267395,
+      "learning_rate": 1.7844057971014495e-05,
+      "loss": 1.6975,
+      "step": 4220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9426621198654175,
+      "learning_rate": 1.783826086956522e-05,
+      "loss": 1.9247,
+      "step": 4230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1264923810958862,
+      "learning_rate": 1.7832463768115943e-05,
+      "loss": 1.8713,
+      "step": 4240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0976786613464355,
+      "learning_rate": 1.7826666666666667e-05,
+      "loss": 1.8226,
+      "step": 4250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.625213146209717,
+      "learning_rate": 1.7820869565217394e-05,
+      "loss": 1.7695,
+      "step": 4260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8756457567214966,
+      "learning_rate": 1.7815072463768118e-05,
+      "loss": 1.6924,
+      "step": 4270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0314109325408936,
+      "learning_rate": 1.7809275362318842e-05,
+      "loss": 1.8747,
+      "step": 4280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.296934127807617,
+      "learning_rate": 1.7803478260869566e-05,
+      "loss": 1.7441,
+      "step": 4290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1548571586608887,
+      "learning_rate": 1.779768115942029e-05,
+      "loss": 1.7059,
+      "step": 4300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.312896728515625,
+      "learning_rate": 1.7791884057971017e-05,
+      "loss": 1.8683,
+      "step": 4310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2402548789978027,
+      "learning_rate": 1.778608695652174e-05,
+      "loss": 1.8141,
+      "step": 4320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2827465534210205,
+      "learning_rate": 1.7780289855072465e-05,
+      "loss": 1.8275,
+      "step": 4330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1937191486358643,
+      "learning_rate": 1.777449275362319e-05,
+      "loss": 1.8033,
+      "step": 4340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4238903522491455,
+      "learning_rate": 1.7768695652173917e-05,
+      "loss": 1.8152,
+      "step": 4350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.212376117706299,
+      "learning_rate": 1.7762898550724637e-05,
+      "loss": 1.699,
+      "step": 4360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7652318477630615,
+      "learning_rate": 1.7757101449275365e-05,
+      "loss": 1.6854,
+      "step": 4370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9386508464813232,
+      "learning_rate": 1.775130434782609e-05,
+      "loss": 1.7143,
+      "step": 4380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.230910062789917,
+      "learning_rate": 1.7745507246376813e-05,
+      "loss": 1.7512,
+      "step": 4390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.213496685028076,
+      "learning_rate": 1.7739710144927537e-05,
+      "loss": 1.7486,
+      "step": 4400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4668760299682617,
+      "learning_rate": 1.773391304347826e-05,
+      "loss": 1.687,
+      "step": 4410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9463025331497192,
+      "learning_rate": 1.7728115942028988e-05,
+      "loss": 1.7829,
+      "step": 4420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4058449268341064,
+      "learning_rate": 1.7722318840579712e-05,
+      "loss": 1.6866,
+      "step": 4430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.426607370376587,
+      "learning_rate": 1.7716521739130436e-05,
+      "loss": 1.7626,
+      "step": 4440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6386098861694336,
+      "learning_rate": 1.771072463768116e-05,
+      "loss": 1.7471,
+      "step": 4450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.66159987449646,
+      "learning_rate": 1.7704927536231888e-05,
+      "loss": 1.6963,
+      "step": 4460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.790877103805542,
+      "learning_rate": 1.7699130434782608e-05,
+      "loss": 1.8013,
+      "step": 4470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4159287214279175,
+      "learning_rate": 1.7693333333333336e-05,
+      "loss": 1.6159,
+      "step": 4480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.394288539886475,
+      "learning_rate": 1.768753623188406e-05,
+      "loss": 1.7185,
+      "step": 4490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0887296199798584,
+      "learning_rate": 1.7681739130434784e-05,
+      "loss": 1.833,
+      "step": 4500
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.7682534456253052,
+      "eval_runtime": 107.4724,
+      "eval_samples_per_second": 9.305,
+      "eval_steps_per_second": 2.326,
+      "step": 4500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.085282325744629,
+      "learning_rate": 1.7675942028985508e-05,
+      "loss": 1.7992,
+      "step": 4510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.182236671447754,
+      "learning_rate": 1.767014492753623e-05,
+      "loss": 1.8982,
+      "step": 4520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9894583225250244,
+      "learning_rate": 1.766434782608696e-05,
+      "loss": 1.6623,
+      "step": 4530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1609718799591064,
+      "learning_rate": 1.7658550724637683e-05,
+      "loss": 1.7441,
+      "step": 4540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7761262655258179,
+      "learning_rate": 1.7652753623188407e-05,
+      "loss": 1.7824,
+      "step": 4550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.564272880554199,
+      "learning_rate": 1.764695652173913e-05,
+      "loss": 1.7627,
+      "step": 4560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6390734910964966,
+      "learning_rate": 1.764115942028986e-05,
+      "loss": 1.7783,
+      "step": 4570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8177413940429688,
+      "learning_rate": 1.7635362318840582e-05,
+      "loss": 1.6956,
+      "step": 4580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2485848665237427,
+      "learning_rate": 1.7629565217391306e-05,
+      "loss": 1.7598,
+      "step": 4590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.094052791595459,
+      "learning_rate": 1.762376811594203e-05,
+      "loss": 1.8477,
+      "step": 4600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.854745864868164,
+      "learning_rate": 1.7617971014492754e-05,
+      "loss": 1.6767,
+      "step": 4610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.275739908218384,
+      "learning_rate": 1.761217391304348e-05,
+      "loss": 1.7992,
+      "step": 4620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.860130786895752,
+      "learning_rate": 1.7606376811594202e-05,
+      "loss": 1.6274,
+      "step": 4630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5182762145996094,
+      "learning_rate": 1.760057971014493e-05,
+      "loss": 1.8214,
+      "step": 4640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3440704345703125,
+      "learning_rate": 1.7594782608695654e-05,
+      "loss": 1.778,
+      "step": 4650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2678601741790771,
+      "learning_rate": 1.7588985507246378e-05,
+      "loss": 1.6729,
+      "step": 4660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.45300030708313,
+      "learning_rate": 1.7583188405797102e-05,
+      "loss": 1.7167,
+      "step": 4670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4436949491500854,
+      "learning_rate": 1.757739130434783e-05,
+      "loss": 1.7222,
+      "step": 4680
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.70391047000885,
+      "learning_rate": 1.7571594202898553e-05,
+      "loss": 1.802,
+      "step": 4690
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.7926859855651855,
+      "learning_rate": 1.7565797101449277e-05,
+      "loss": 1.7951,
+      "step": 4700
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.777199387550354,
+      "learning_rate": 1.756e-05,
+      "loss": 1.7408,
+      "step": 4710
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9820401072502136,
+      "learning_rate": 1.7554202898550725e-05,
+      "loss": 1.8236,
+      "step": 4720
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.505861759185791,
+      "learning_rate": 1.754840579710145e-05,
+      "loss": 1.6859,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0576391220092773,
+      "learning_rate": 1.7542608695652173e-05,
+      "loss": 1.7222,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.96050763130188,
+      "learning_rate": 1.75368115942029e-05,
+      "loss": 1.8761,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.387930870056152,
+      "learning_rate": 1.7531014492753625e-05,
+      "loss": 1.6781,
+      "step": 4760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.898324012756348,
+      "learning_rate": 1.752521739130435e-05,
+      "loss": 1.783,
+      "step": 4770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.5490803718566895,
+      "learning_rate": 1.7519420289855073e-05,
+      "loss": 1.7627,
+      "step": 4780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.1455864906311035,
+      "learning_rate": 1.75136231884058e-05,
+      "loss": 1.7222,
+      "step": 4790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.632255792617798,
+      "learning_rate": 1.7507826086956524e-05,
+      "loss": 1.8319,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.351136565208435,
+      "learning_rate": 1.7502028985507248e-05,
+      "loss": 1.7507,
+      "step": 4810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6380447149276733,
+      "learning_rate": 1.7496231884057972e-05,
+      "loss": 1.6645,
+      "step": 4820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.601467490196228,
+      "learning_rate": 1.7490434782608696e-05,
+      "loss": 1.8176,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0333625078201294,
+      "learning_rate": 1.7484637681159424e-05,
+      "loss": 1.6752,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7758312225341797,
+      "learning_rate": 1.7478840579710144e-05,
+      "loss": 1.7522,
+      "step": 4850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.394230842590332,
+      "learning_rate": 1.747304347826087e-05,
+      "loss": 1.6703,
+      "step": 4860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6446850299835205,
+      "learning_rate": 1.7467246376811596e-05,
+      "loss": 1.6016,
+      "step": 4870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.261119842529297,
+      "learning_rate": 1.746144927536232e-05,
+      "loss": 1.598,
+      "step": 4880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5779495239257812,
+      "learning_rate": 1.7455652173913044e-05,
+      "loss": 1.7129,
+      "step": 4890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2901227474212646,
+      "learning_rate": 1.744985507246377e-05,
+      "loss": 1.6935,
+      "step": 4900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.433979034423828,
+      "learning_rate": 1.7444057971014495e-05,
+      "loss": 1.6954,
+      "step": 4910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8954328894615173,
+      "learning_rate": 1.743826086956522e-05,
+      "loss": 1.7721,
+      "step": 4920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.513370990753174,
+      "learning_rate": 1.7432463768115943e-05,
+      "loss": 1.6607,
+      "step": 4930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7021384239196777,
+      "learning_rate": 1.7426666666666667e-05,
+      "loss": 1.7342,
+      "step": 4940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.947293519973755,
+      "learning_rate": 1.7420869565217394e-05,
+      "loss": 1.6413,
+      "step": 4950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9602371454238892,
+      "learning_rate": 1.741507246376812e-05,
+      "loss": 1.6927,
+      "step": 4960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6332249641418457,
+      "learning_rate": 1.7409275362318842e-05,
+      "loss": 1.5976,
+      "step": 4970
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.293816328048706,
+      "learning_rate": 1.7403478260869566e-05,
+      "loss": 1.7207,
+      "step": 4980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.466865539550781,
+      "learning_rate": 1.7397681159420294e-05,
+      "loss": 1.7308,
+      "step": 4990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.4341559410095215,
+      "learning_rate": 1.7391884057971014e-05,
+      "loss": 1.7434,
+      "step": 5000
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.7550283670425415,
+      "eval_runtime": 107.4768,
+      "eval_samples_per_second": 9.304,
+      "eval_steps_per_second": 2.326,
+      "step": 5000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.294936180114746,
+      "learning_rate": 1.7386086956521742e-05,
+      "loss": 1.746,
+      "step": 5010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.660215139389038,
+      "learning_rate": 1.7380289855072466e-05,
+      "loss": 1.7183,
+      "step": 5020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8590058088302612,
+      "learning_rate": 1.737449275362319e-05,
+      "loss": 1.8525,
+      "step": 5030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5797920227050781,
+      "learning_rate": 1.7368695652173914e-05,
+      "loss": 1.7257,
+      "step": 5040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3411496877670288,
+      "learning_rate": 1.7362898550724638e-05,
+      "loss": 1.8922,
+      "step": 5050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5491766929626465,
+      "learning_rate": 1.7357101449275365e-05,
+      "loss": 1.7402,
+      "step": 5060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.72056245803833,
+      "learning_rate": 1.735130434782609e-05,
+      "loss": 1.7225,
+      "step": 5070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4798800945281982,
+      "learning_rate": 1.7345507246376813e-05,
+      "loss": 1.7624,
+      "step": 5080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2017393112182617,
+      "learning_rate": 1.7339710144927537e-05,
+      "loss": 1.7921,
+      "step": 5090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.185555934906006,
+      "learning_rate": 1.7333913043478265e-05,
+      "loss": 1.615,
+      "step": 5100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8740787506103516,
+      "learning_rate": 1.7328115942028985e-05,
+      "loss": 1.778,
+      "step": 5110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6810436248779297,
+      "learning_rate": 1.732231884057971e-05,
+      "loss": 1.7957,
+      "step": 5120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.130495309829712,
+      "learning_rate": 1.7316521739130437e-05,
+      "loss": 1.8177,
+      "step": 5130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.660665988922119,
+      "learning_rate": 1.731072463768116e-05,
+      "loss": 1.8078,
+      "step": 5140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.669716238975525,
+      "learning_rate": 1.7304927536231885e-05,
+      "loss": 1.7621,
+      "step": 5150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6844958066940308,
+      "learning_rate": 1.729913043478261e-05,
+      "loss": 1.6396,
+      "step": 5160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.097205877304077,
+      "learning_rate": 1.7293333333333336e-05,
+      "loss": 1.7409,
+      "step": 5170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8433555960655212,
+      "learning_rate": 1.728753623188406e-05,
+      "loss": 1.8663,
+      "step": 5180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1566624641418457,
+      "learning_rate": 1.7281739130434784e-05,
+      "loss": 1.5808,
+      "step": 5190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.130633592605591,
+      "learning_rate": 1.7275942028985508e-05,
+      "loss": 1.6708,
+      "step": 5200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3637614250183105,
+      "learning_rate": 1.7270144927536235e-05,
+      "loss": 1.7204,
+      "step": 5210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.3107686042785645,
+      "learning_rate": 1.726434782608696e-05,
+      "loss": 1.7913,
+      "step": 5220
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3103054761886597,
+      "learning_rate": 1.725855072463768e-05,
+      "loss": 1.7843,
+      "step": 5230
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.409878730773926,
+      "learning_rate": 1.7252753623188407e-05,
+      "loss": 1.5763,
+      "step": 5240
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2016843557357788,
+      "learning_rate": 1.724695652173913e-05,
+      "loss": 1.7474,
+      "step": 5250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6358120441436768,
+      "learning_rate": 1.7241159420289855e-05,
+      "loss": 1.6569,
+      "step": 5260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.816072702407837,
+      "learning_rate": 1.723536231884058e-05,
+      "loss": 1.7288,
+      "step": 5270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.151131629943848,
+      "learning_rate": 1.7229565217391307e-05,
+      "loss": 1.7776,
+      "step": 5280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.895945429801941,
+      "learning_rate": 1.722376811594203e-05,
+      "loss": 1.7824,
+      "step": 5290
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5758895874023438,
+      "learning_rate": 1.7217971014492755e-05,
+      "loss": 1.6245,
+      "step": 5300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6517353057861328,
+      "learning_rate": 1.721217391304348e-05,
+      "loss": 1.5666,
+      "step": 5310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3738932609558105,
+      "learning_rate": 1.7206376811594206e-05,
+      "loss": 1.6949,
+      "step": 5320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.818183898925781,
+      "learning_rate": 1.720057971014493e-05,
+      "loss": 1.7589,
+      "step": 5330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0807089805603027,
+      "learning_rate": 1.7194782608695654e-05,
+      "loss": 1.8133,
+      "step": 5340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9534027576446533,
+      "learning_rate": 1.7188985507246378e-05,
+      "loss": 1.7675,
+      "step": 5350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3471546173095703,
+      "learning_rate": 1.7183188405797102e-05,
+      "loss": 1.4739,
+      "step": 5360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1540451049804688,
+      "learning_rate": 1.7177391304347826e-05,
+      "loss": 1.7602,
+      "step": 5370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6220521926879883,
+      "learning_rate": 1.717159420289855e-05,
+      "loss": 1.8812,
+      "step": 5380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6644699573516846,
+      "learning_rate": 1.7165797101449278e-05,
+      "loss": 1.8191,
+      "step": 5390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.378482818603516,
+      "learning_rate": 1.7160000000000002e-05,
+      "loss": 1.7765,
+      "step": 5400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0857391357421875,
+      "learning_rate": 1.7154202898550726e-05,
+      "loss": 1.7762,
+      "step": 5410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7650538086891174,
+      "learning_rate": 1.714840579710145e-05,
+      "loss": 1.7468,
+      "step": 5420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3827682733535767,
+      "learning_rate": 1.7142608695652174e-05,
+      "loss": 1.8392,
+      "step": 5430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.6188459396362305,
+      "learning_rate": 1.71368115942029e-05,
+      "loss": 1.765,
+      "step": 5440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.591964602470398,
+      "learning_rate": 1.7131014492753625e-05,
+      "loss": 1.6735,
+      "step": 5450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4797956943511963,
+      "learning_rate": 1.712521739130435e-05,
+      "loss": 1.8693,
+      "step": 5460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.759324312210083,
+      "learning_rate": 1.7119420289855073e-05,
+      "loss": 1.691,
+      "step": 5470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8067173957824707,
+      "learning_rate": 1.71136231884058e-05,
+      "loss": 1.8131,
+      "step": 5480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9415967464447021,
+      "learning_rate": 1.710782608695652e-05,
+      "loss": 1.8269,
+      "step": 5490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.027404308319092,
+      "learning_rate": 1.710202898550725e-05,
+      "loss": 1.8457,
+      "step": 5500
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.7384296655654907,
+      "eval_runtime": 107.4854,
+      "eval_samples_per_second": 9.304,
+      "eval_steps_per_second": 2.326,
+      "step": 5500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0170934200286865,
+      "learning_rate": 1.7096231884057973e-05,
+      "loss": 1.7555,
+      "step": 5510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8536728620529175,
+      "learning_rate": 1.7090434782608697e-05,
+      "loss": 1.6274,
+      "step": 5520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8969680070877075,
+      "learning_rate": 1.708463768115942e-05,
+      "loss": 1.7977,
+      "step": 5530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3134697675704956,
+      "learning_rate": 1.7078840579710145e-05,
+      "loss": 1.7819,
+      "step": 5540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6828033924102783,
+      "learning_rate": 1.7073043478260872e-05,
+      "loss": 1.924,
+      "step": 5550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.980905294418335,
+      "learning_rate": 1.7067246376811596e-05,
+      "loss": 1.7677,
+      "step": 5560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.521270751953125,
+      "learning_rate": 1.706144927536232e-05,
+      "loss": 1.7319,
+      "step": 5570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6345350742340088,
+      "learning_rate": 1.7055652173913044e-05,
+      "loss": 1.8736,
+      "step": 5580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6569340229034424,
+      "learning_rate": 1.704985507246377e-05,
+      "loss": 1.7549,
+      "step": 5590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3906219005584717,
+      "learning_rate": 1.7044057971014495e-05,
+      "loss": 1.8223,
+      "step": 5600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6184358596801758,
+      "learning_rate": 1.703826086956522e-05,
+      "loss": 1.7187,
+      "step": 5610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9662846326828003,
+      "learning_rate": 1.7032463768115943e-05,
+      "loss": 1.6826,
+      "step": 5620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2513630390167236,
+      "learning_rate": 1.702666666666667e-05,
+      "loss": 1.8264,
+      "step": 5630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.757554769515991,
+      "learning_rate": 1.702086956521739e-05,
+      "loss": 1.7352,
+      "step": 5640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9750083088874817,
+      "learning_rate": 1.7015072463768115e-05,
+      "loss": 1.68,
+      "step": 5650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.763793706893921,
+      "learning_rate": 1.7009275362318843e-05,
+      "loss": 1.7913,
+      "step": 5660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.929349899291992,
+      "learning_rate": 1.7003478260869567e-05,
+      "loss": 1.7713,
+      "step": 5670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.068824768066406,
+      "learning_rate": 1.699768115942029e-05,
+      "loss": 1.6546,
+      "step": 5680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.953929901123047,
+      "learning_rate": 1.6991884057971015e-05,
+      "loss": 1.8524,
+      "step": 5690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.38455069065094,
+      "learning_rate": 1.6986086956521742e-05,
+      "loss": 1.8452,
+      "step": 5700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4616096019744873,
+      "learning_rate": 1.6980289855072466e-05,
+      "loss": 1.7802,
+      "step": 5710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.079864263534546,
+      "learning_rate": 1.697449275362319e-05,
+      "loss": 1.6293,
+      "step": 5720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.769149303436279,
+      "learning_rate": 1.6968695652173914e-05,
+      "loss": 1.7868,
+      "step": 5730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6357669830322266,
+      "learning_rate": 1.6962898550724638e-05,
+      "loss": 1.8268,
+      "step": 5740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.242451190948486,
+      "learning_rate": 1.6957101449275362e-05,
+      "loss": 1.8289,
+      "step": 5750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4025174379348755,
+      "learning_rate": 1.6951304347826086e-05,
+      "loss": 1.7262,
+      "step": 5760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.201425790786743,
+      "learning_rate": 1.6945507246376814e-05,
+      "loss": 1.7995,
+      "step": 5770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1672089099884033,
+      "learning_rate": 1.6939710144927538e-05,
+      "loss": 1.7804,
+      "step": 5780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.394217848777771,
+      "learning_rate": 1.693391304347826e-05,
+      "loss": 1.6865,
+      "step": 5790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.455097198486328,
+      "learning_rate": 1.6928115942028986e-05,
+      "loss": 1.6132,
+      "step": 5800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6794978380203247,
+      "learning_rate": 1.6922318840579713e-05,
+      "loss": 1.7146,
+      "step": 5810
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.268734455108643,
+      "learning_rate": 1.6916521739130437e-05,
+      "loss": 1.8458,
+      "step": 5820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9144909381866455,
+      "learning_rate": 1.691072463768116e-05,
+      "loss": 1.7298,
+      "step": 5830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2349727153778076,
+      "learning_rate": 1.6904927536231885e-05,
+      "loss": 1.6982,
+      "step": 5840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7180557250976562,
+      "learning_rate": 1.689913043478261e-05,
+      "loss": 1.8027,
+      "step": 5850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6468281745910645,
+      "learning_rate": 1.6893913043478262e-05,
+      "loss": 1.8093,
+      "step": 5860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.979691982269287,
+      "learning_rate": 1.6888115942028986e-05,
+      "loss": 1.5826,
+      "step": 5870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.700094699859619,
+      "learning_rate": 1.6882318840579713e-05,
+      "loss": 1.6295,
+      "step": 5880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.222003936767578,
+      "learning_rate": 1.6876521739130437e-05,
+      "loss": 1.655,
+      "step": 5890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3589376211166382,
+      "learning_rate": 1.687072463768116e-05,
+      "loss": 1.7627,
+      "step": 5900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.710470676422119,
+      "learning_rate": 1.6864927536231885e-05,
+      "loss": 1.7624,
+      "step": 5910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3515254259109497,
+      "learning_rate": 1.685913043478261e-05,
+      "loss": 1.754,
+      "step": 5920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.581727385520935,
+      "learning_rate": 1.6853333333333333e-05,
+      "loss": 1.7327,
+      "step": 5930
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4707586765289307,
+      "learning_rate": 1.6847536231884057e-05,
+      "loss": 1.796,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6579630374908447,
+      "learning_rate": 1.6841739130434785e-05,
+      "loss": 1.664,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.639225959777832,
+      "learning_rate": 1.683594202898551e-05,
+      "loss": 1.7041,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8888185024261475,
+      "learning_rate": 1.6830144927536233e-05,
+      "loss": 1.7493,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3176138401031494,
+      "learning_rate": 1.6824347826086957e-05,
+      "loss": 1.5773,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5189319849014282,
+      "learning_rate": 1.6818550724637684e-05,
+      "loss": 1.6852,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.535059928894043,
+      "learning_rate": 1.6812753623188408e-05,
+      "loss": 1.7166,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.727913498878479,
+      "eval_runtime": 107.4778,
+      "eval_samples_per_second": 9.304,
+      "eval_steps_per_second": 2.326,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6467275619506836,
+      "learning_rate": 1.6806956521739132e-05,
+      "loss": 1.8415,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4421344995498657,
+      "learning_rate": 1.6801159420289856e-05,
+      "loss": 1.6914,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6661970615386963,
+      "learning_rate": 1.679536231884058e-05,
+      "loss": 1.8147,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5970772504806519,
+      "learning_rate": 1.6789565217391304e-05,
+      "loss": 1.6534,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1191587448120117,
+      "learning_rate": 1.678376811594203e-05,
+      "loss": 1.7521,
+      "step": 6050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.529994249343872,
+      "learning_rate": 1.6777971014492756e-05,
+      "loss": 1.6383,
+      "step": 6060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4743109941482544,
+      "learning_rate": 1.677217391304348e-05,
+      "loss": 1.6432,
+      "step": 6070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2322564125061035,
+      "learning_rate": 1.6766376811594204e-05,
+      "loss": 1.6845,
+      "step": 6080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.006726264953613,
+      "learning_rate": 1.6760579710144928e-05,
+      "loss": 1.8527,
+      "step": 6090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6824820041656494,
+      "learning_rate": 1.6754782608695655e-05,
+      "loss": 1.7306,
+      "step": 6100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2515430450439453,
+      "learning_rate": 1.674898550724638e-05,
+      "loss": 1.6419,
+      "step": 6110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.548213005065918,
+      "learning_rate": 1.6743188405797103e-05,
+      "loss": 1.6744,
+      "step": 6120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4482362270355225,
+      "learning_rate": 1.6737391304347827e-05,
+      "loss": 1.8767,
+      "step": 6130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4814817905426025,
+      "learning_rate": 1.673159420289855e-05,
+      "loss": 1.71,
+      "step": 6140
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6053385734558105,
+      "learning_rate": 1.672579710144928e-05,
+      "loss": 1.557,
+      "step": 6150
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.154989004135132,
+      "learning_rate": 1.672e-05,
+      "loss": 1.7413,
+      "step": 6160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2812576293945312,
+      "learning_rate": 1.6714202898550727e-05,
+      "loss": 1.7674,
+      "step": 6170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1822229623794556,
+      "learning_rate": 1.670840579710145e-05,
+      "loss": 1.7635,
+      "step": 6180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.022081136703491,
+      "learning_rate": 1.6702608695652175e-05,
+      "loss": 1.9249,
+      "step": 6190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4836503267288208,
+      "learning_rate": 1.66968115942029e-05,
+      "loss": 1.7538,
+      "step": 6200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.186819314956665,
+      "learning_rate": 1.6691014492753626e-05,
+      "loss": 1.7048,
+      "step": 6210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.98888897895813,
+      "learning_rate": 1.668521739130435e-05,
+      "loss": 1.7195,
+      "step": 6220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2272439002990723,
+      "learning_rate": 1.6679420289855074e-05,
+      "loss": 1.7088,
+      "step": 6230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.195132255554199,
+      "learning_rate": 1.6673623188405798e-05,
+      "loss": 1.9109,
+      "step": 6240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4767738580703735,
+      "learning_rate": 1.6667826086956522e-05,
+      "loss": 1.7119,
+      "step": 6250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.870778560638428,
+      "learning_rate": 1.666202898550725e-05,
+      "loss": 1.6744,
+      "step": 6260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.495901584625244,
+      "learning_rate": 1.6656231884057973e-05,
+      "loss": 1.7129,
+      "step": 6270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2989490032196045,
+      "learning_rate": 1.6650434782608697e-05,
+      "loss": 1.7543,
+      "step": 6280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2560532093048096,
+      "learning_rate": 1.664463768115942e-05,
+      "loss": 1.6504,
+      "step": 6290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1828808784484863,
+      "learning_rate": 1.663884057971015e-05,
+      "loss": 1.8215,
+      "step": 6300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3935143947601318,
+      "learning_rate": 1.663304347826087e-05,
+      "loss": 1.7343,
+      "step": 6310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2186152935028076,
+      "learning_rate": 1.6627246376811597e-05,
+      "loss": 1.8291,
+      "step": 6320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0307185649871826,
+      "learning_rate": 1.662144927536232e-05,
+      "loss": 1.7496,
+      "step": 6330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4176080226898193,
+      "learning_rate": 1.6615652173913045e-05,
+      "loss": 1.7847,
+      "step": 6340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.357802391052246,
+      "learning_rate": 1.660985507246377e-05,
+      "loss": 1.7378,
+      "step": 6350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5888750553131104,
+      "learning_rate": 1.6604057971014493e-05,
+      "loss": 1.7028,
+      "step": 6360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2639836072921753,
+      "learning_rate": 1.659826086956522e-05,
+      "loss": 1.7333,
+      "step": 6370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8823935985565186,
+      "learning_rate": 1.6592463768115944e-05,
+      "loss": 1.7696,
+      "step": 6380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7665212154388428,
+      "learning_rate": 1.6586666666666668e-05,
+      "loss": 1.8186,
+      "step": 6390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.463416814804077,
+      "learning_rate": 1.6580869565217392e-05,
+      "loss": 1.8244,
+      "step": 6400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7725675106048584,
+      "learning_rate": 1.657507246376812e-05,
+      "loss": 1.7399,
+      "step": 6410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1055688858032227,
+      "learning_rate": 1.656927536231884e-05,
+      "loss": 1.6752,
+      "step": 6420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.611109972000122,
+      "learning_rate": 1.6563478260869568e-05,
+      "loss": 1.5842,
+      "step": 6430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2504234313964844,
+      "learning_rate": 1.655768115942029e-05,
+      "loss": 1.6229,
+      "step": 6440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.377387046813965,
+      "learning_rate": 1.6551884057971016e-05,
+      "loss": 1.7606,
+      "step": 6450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7958356142044067,
+      "learning_rate": 1.654608695652174e-05,
+      "loss": 1.7271,
+      "step": 6460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7075881958007812,
+      "learning_rate": 1.6540289855072464e-05,
+      "loss": 1.7918,
+      "step": 6470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.526883602142334,
+      "learning_rate": 1.653449275362319e-05,
+      "loss": 1.721,
+      "step": 6480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.219193458557129,
+      "learning_rate": 1.6528695652173915e-05,
+      "loss": 1.8068,
+      "step": 6490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7453542947769165,
+      "learning_rate": 1.652289855072464e-05,
+      "loss": 1.7639,
+      "step": 6500
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.727725625038147,
+      "eval_runtime": 107.4666,
+      "eval_samples_per_second": 9.305,
+      "eval_steps_per_second": 2.326,
+      "step": 6500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9024394750595093,
+      "learning_rate": 1.6517101449275363e-05,
+      "loss": 1.7855,
+      "step": 6510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.008007287979126,
+      "learning_rate": 1.651130434782609e-05,
+      "loss": 1.7057,
+      "step": 6520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9877010583877563,
+      "learning_rate": 1.6505507246376814e-05,
+      "loss": 1.7263,
+      "step": 6530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.027751445770264,
+      "learning_rate": 1.6499710144927535e-05,
+      "loss": 1.6714,
+      "step": 6540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.0006256103515625,
+      "learning_rate": 1.6493913043478262e-05,
+      "loss": 1.6872,
+      "step": 6550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7696921825408936,
+      "learning_rate": 1.6488115942028986e-05,
+      "loss": 1.7761,
+      "step": 6560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.299485683441162,
+      "learning_rate": 1.648231884057971e-05,
+      "loss": 1.6696,
+      "step": 6570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5291144847869873,
+      "learning_rate": 1.6476521739130435e-05,
+      "loss": 1.6788,
+      "step": 6580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.1017231941223145,
+      "learning_rate": 1.6470724637681162e-05,
+      "loss": 1.6339,
+      "step": 6590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.975684881210327,
+      "learning_rate": 1.6464927536231886e-05,
+      "loss": 1.63,
+      "step": 6600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5306545495986938,
+      "learning_rate": 1.645913043478261e-05,
+      "loss": 1.6622,
+      "step": 6610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.816344976425171,
+      "learning_rate": 1.6453333333333334e-05,
+      "loss": 1.8507,
+      "step": 6620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.180396556854248,
+      "learning_rate": 1.644753623188406e-05,
+      "loss": 1.6871,
+      "step": 6630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9468295574188232,
+      "learning_rate": 1.6441739130434785e-05,
+      "loss": 1.7409,
+      "step": 6640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6706266403198242,
+      "learning_rate": 1.6435942028985506e-05,
+      "loss": 1.6801,
+      "step": 6650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4290724992752075,
+      "learning_rate": 1.6430144927536233e-05,
+      "loss": 1.6932,
+      "step": 6660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8305716514587402,
+      "learning_rate": 1.6424347826086957e-05,
+      "loss": 1.8525,
+      "step": 6670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.744149088859558,
+      "learning_rate": 1.641855072463768e-05,
+      "loss": 1.6997,
+      "step": 6680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0578601360321045,
+      "learning_rate": 1.6412753623188405e-05,
+      "loss": 1.706,
+      "step": 6690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.352912425994873,
+      "learning_rate": 1.6406956521739133e-05,
+      "loss": 1.7743,
+      "step": 6700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.978705883026123,
+      "learning_rate": 1.6401159420289857e-05,
+      "loss": 1.7195,
+      "step": 6710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2666630744934082,
+      "learning_rate": 1.639536231884058e-05,
+      "loss": 1.6889,
+      "step": 6720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.681015133857727,
+      "learning_rate": 1.6389565217391305e-05,
+      "loss": 1.6464,
+      "step": 6730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7917625904083252,
+      "learning_rate": 1.6383768115942032e-05,
+      "loss": 1.7322,
+      "step": 6740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.438145875930786,
+      "learning_rate": 1.6377971014492756e-05,
+      "loss": 1.6975,
+      "step": 6750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.554532766342163,
+      "learning_rate": 1.637217391304348e-05,
+      "loss": 1.7592,
+      "step": 6760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3768386840820312,
+      "learning_rate": 1.6366376811594204e-05,
+      "loss": 1.6267,
+      "step": 6770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.807661533355713,
+      "learning_rate": 1.6360579710144928e-05,
+      "loss": 1.6814,
+      "step": 6780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.202757835388184,
+      "learning_rate": 1.6354782608695656e-05,
+      "loss": 1.7636,
+      "step": 6790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.915995121002197,
+      "learning_rate": 1.6348985507246376e-05,
+      "loss": 1.6552,
+      "step": 6800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9319915771484375,
+      "learning_rate": 1.6343188405797104e-05,
+      "loss": 1.6542,
+      "step": 6810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.227419137954712,
+      "learning_rate": 1.6337391304347828e-05,
+      "loss": 1.771,
+      "step": 6820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.618062973022461,
+      "learning_rate": 1.633159420289855e-05,
+      "loss": 1.7122,
+      "step": 6830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2178608179092407,
+      "learning_rate": 1.6325797101449276e-05,
+      "loss": 1.6398,
+      "step": 6840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5877292156219482,
+      "learning_rate": 1.632e-05,
+      "loss": 1.8347,
+      "step": 6850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.831631898880005,
+      "learning_rate": 1.6314202898550727e-05,
+      "loss": 1.6153,
+      "step": 6860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.031569242477417,
+      "learning_rate": 1.630840579710145e-05,
+      "loss": 1.5382,
+      "step": 6870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.036290645599365,
+      "learning_rate": 1.6302608695652175e-05,
+      "loss": 1.7124,
+      "step": 6880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.04699444770813,
+      "learning_rate": 1.62968115942029e-05,
+      "loss": 1.8437,
+      "step": 6890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.757121205329895,
+      "learning_rate": 1.6291014492753626e-05,
+      "loss": 1.7331,
+      "step": 6900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.107645273208618,
+      "learning_rate": 1.628521739130435e-05,
+      "loss": 1.7234,
+      "step": 6910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.109172821044922,
+      "learning_rate": 1.6279420289855074e-05,
+      "loss": 1.8055,
+      "step": 6920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.344881772994995,
+      "learning_rate": 1.62736231884058e-05,
+      "loss": 1.7585,
+      "step": 6930
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4652295112609863,
+      "learning_rate": 1.6267826086956522e-05,
+      "loss": 1.7634,
+      "step": 6940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3896074295043945,
+      "learning_rate": 1.6262028985507246e-05,
+      "loss": 1.8021,
+      "step": 6950
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5588115453720093,
+      "learning_rate": 1.625623188405797e-05,
+      "loss": 1.7534,
+      "step": 6960
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.033330202102661,
+      "learning_rate": 1.6250434782608698e-05,
+      "loss": 1.6904,
+      "step": 6970
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9910866022109985,
+      "learning_rate": 1.6244637681159422e-05,
+      "loss": 1.708,
+      "step": 6980
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4834866523742676,
+      "learning_rate": 1.6238840579710146e-05,
+      "loss": 1.717,
+      "step": 6990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1782163381576538,
+      "learning_rate": 1.623304347826087e-05,
+      "loss": 1.5833,
+      "step": 7000
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 1.7107688188552856,
+      "eval_runtime": 107.4755,
+      "eval_samples_per_second": 9.304,
+      "eval_steps_per_second": 2.326,
+      "step": 7000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3755738735198975,
+      "learning_rate": 1.6227246376811597e-05,
+      "loss": 1.845,
+      "step": 7010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7433241605758667,
+      "learning_rate": 1.622144927536232e-05,
+      "loss": 1.8534,
+      "step": 7020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.202053070068359,
+      "learning_rate": 1.6215652173913045e-05,
+      "loss": 1.6611,
+      "step": 7030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.540313720703125,
+      "learning_rate": 1.620985507246377e-05,
+      "loss": 1.7816,
+      "step": 7040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.980997085571289,
+      "learning_rate": 1.6204057971014497e-05,
+      "loss": 1.8204,
+      "step": 7050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4950547218322754,
+      "learning_rate": 1.6198260869565217e-05,
+      "loss": 1.779,
+      "step": 7060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1530729532241821,
+      "learning_rate": 1.619246376811594e-05,
+      "loss": 1.7435,
+      "step": 7070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9771350622177124,
+      "learning_rate": 1.618666666666667e-05,
+      "loss": 1.65,
+      "step": 7080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.06072735786438,
+      "learning_rate": 1.6180869565217393e-05,
+      "loss": 1.741,
+      "step": 7090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4323557615280151,
+      "learning_rate": 1.6175072463768117e-05,
+      "loss": 1.807,
+      "step": 7100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6237205266952515,
+      "learning_rate": 1.616927536231884e-05,
+      "loss": 1.742,
+      "step": 7110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.375201463699341,
+      "learning_rate": 1.6163478260869568e-05,
+      "loss": 1.6201,
+      "step": 7120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.091128349304199,
+      "learning_rate": 1.6157681159420292e-05,
+      "loss": 1.7904,
+      "step": 7130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4652012586593628,
+      "learning_rate": 1.6151884057971016e-05,
+      "loss": 1.6906,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8465373516082764,
+      "learning_rate": 1.614608695652174e-05,
+      "loss": 1.8106,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7708303928375244,
+      "learning_rate": 1.6140289855072464e-05,
+      "loss": 1.7154,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.694922924041748,
+      "learning_rate": 1.613449275362319e-05,
+      "loss": 1.738,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2050352096557617,
+      "learning_rate": 1.6128695652173912e-05,
+      "loss": 1.7701,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.498955249786377,
+      "learning_rate": 1.612289855072464e-05,
+      "loss": 1.7794,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.483665108680725,
+      "learning_rate": 1.6117101449275364e-05,
+      "loss": 1.7243,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6146090030670166,
+      "learning_rate": 1.6111304347826088e-05,
+      "loss": 1.7547,
+      "step": 7210
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9853881597518921,
+      "learning_rate": 1.610550724637681e-05,
+      "loss": 1.6421,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.687666654586792,
+      "learning_rate": 1.609971014492754e-05,
+      "loss": 1.6646,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8231513500213623,
+      "learning_rate": 1.6093913043478263e-05,
+      "loss": 1.8263,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8006614446640015,
+      "learning_rate": 1.6088115942028987e-05,
+      "loss": 1.6252,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6685547828674316,
+      "learning_rate": 1.608231884057971e-05,
+      "loss": 1.6308,
+      "step": 7260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.637762069702148,
+      "learning_rate": 1.6076521739130435e-05,
+      "loss": 1.6501,
+      "step": 7270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1397578716278076,
+      "learning_rate": 1.6070724637681162e-05,
+      "loss": 1.7779,
+      "step": 7280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.391406536102295,
+      "learning_rate": 1.6064927536231883e-05,
+      "loss": 1.7377,
+      "step": 7290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.646698474884033,
+      "learning_rate": 1.605913043478261e-05,
+      "loss": 1.7138,
+      "step": 7300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9890764951705933,
+      "learning_rate": 1.6053333333333334e-05,
+      "loss": 1.636,
+      "step": 7310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.770165205001831,
+      "learning_rate": 1.604753623188406e-05,
+      "loss": 1.8327,
+      "step": 7320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.544312834739685,
+      "learning_rate": 1.6041739130434782e-05,
+      "loss": 1.7628,
+      "step": 7330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.551856517791748,
+      "learning_rate": 1.603594202898551e-05,
+      "loss": 1.7112,
+      "step": 7340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.357561111450195,
+      "learning_rate": 1.6030144927536234e-05,
+      "loss": 1.8364,
+      "step": 7350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.165691375732422,
+      "learning_rate": 1.6024347826086958e-05,
+      "loss": 1.6699,
+      "step": 7360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3608006238937378,
+      "learning_rate": 1.6018550724637682e-05,
+      "loss": 1.8786,
+      "step": 7370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6346604824066162,
+      "learning_rate": 1.6012753623188406e-05,
+      "loss": 1.6912,
+      "step": 7380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.323948383331299,
+      "learning_rate": 1.6006956521739133e-05,
+      "loss": 1.7305,
+      "step": 7390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.645804762840271,
+      "learning_rate": 1.6001159420289857e-05,
+      "loss": 1.6702,
+      "step": 7400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5827347040176392,
+      "learning_rate": 1.599536231884058e-05,
+      "loss": 1.6253,
+      "step": 7410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5143661499023438,
+      "learning_rate": 1.5989565217391305e-05,
+      "loss": 1.6295,
+      "step": 7420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.025846242904663,
+      "learning_rate": 1.5983768115942033e-05,
+      "loss": 1.6297,
+      "step": 7430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.31910240650177,
+      "learning_rate": 1.5977971014492753e-05,
+      "loss": 1.66,
+      "step": 7440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5227108001708984,
+      "learning_rate": 1.597217391304348e-05,
+      "loss": 1.5631,
+      "step": 7450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1496291160583496,
+      "learning_rate": 1.5966376811594205e-05,
+      "loss": 1.7575,
+      "step": 7460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5027258396148682,
+      "learning_rate": 1.596057971014493e-05,
+      "loss": 1.7693,
+      "step": 7470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8915343284606934,
+      "learning_rate": 1.5954782608695653e-05,
+      "loss": 1.5916,
+      "step": 7480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5063656568527222,
+      "learning_rate": 1.5948985507246377e-05,
+      "loss": 1.8137,
+      "step": 7490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6433496475219727,
+      "learning_rate": 1.5943188405797104e-05,
+      "loss": 1.7322,
+      "step": 7500
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 1.705659031867981,
+      "eval_runtime": 107.4816,
+      "eval_samples_per_second": 9.304,
+      "eval_steps_per_second": 2.326,
+      "step": 7500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.358209609985352,
+      "learning_rate": 1.5937391304347828e-05,
+      "loss": 1.7619,
+      "step": 7510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.150690078735352,
+      "learning_rate": 1.5931594202898552e-05,
+      "loss": 1.6083,
+      "step": 7520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.220200061798096,
+      "learning_rate": 1.5925797101449276e-05,
+      "loss": 1.7934,
+      "step": 7530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5552024841308594,
+      "learning_rate": 1.5920000000000003e-05,
+      "loss": 1.7875,
+      "step": 7540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2470431327819824,
+      "learning_rate": 1.5914202898550727e-05,
+      "loss": 1.8069,
+      "step": 7550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.775547981262207,
+      "learning_rate": 1.590840579710145e-05,
+      "loss": 1.7393,
+      "step": 7560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7677903175354004,
+      "learning_rate": 1.5902608695652175e-05,
+      "loss": 1.7661,
+      "step": 7570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0370768308639526,
+      "learning_rate": 1.58968115942029e-05,
+      "loss": 1.7814,
+      "step": 7580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4217703342437744,
+      "learning_rate": 1.5891014492753623e-05,
+      "loss": 1.7731,
+      "step": 7590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.532466173171997,
+      "learning_rate": 1.5885217391304347e-05,
+      "loss": 1.68,
+      "step": 7600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8123104572296143,
+      "learning_rate": 1.5879420289855075e-05,
+      "loss": 1.6958,
+      "step": 7610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.266079425811768,
+      "learning_rate": 1.58736231884058e-05,
+      "loss": 1.8838,
+      "step": 7620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.8042216300964355,
+      "learning_rate": 1.5867826086956523e-05,
+      "loss": 1.7128,
+      "step": 7630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8047385215759277,
+      "learning_rate": 1.5862028985507247e-05,
+      "loss": 1.7068,
+      "step": 7640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.157292604446411,
+      "learning_rate": 1.5856231884057974e-05,
+      "loss": 1.8066,
+      "step": 7650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3345887660980225,
+      "learning_rate": 1.5850434782608698e-05,
+      "loss": 1.6732,
+      "step": 7660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.439162254333496,
+      "learning_rate": 1.5844637681159422e-05,
+      "loss": 1.698,
+      "step": 7670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.751528263092041,
+      "learning_rate": 1.5838840579710146e-05,
+      "loss": 1.7296,
+      "step": 7680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4270384311676025,
+      "learning_rate": 1.583304347826087e-05,
+      "loss": 1.84,
+      "step": 7690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0358874797821045,
+      "learning_rate": 1.5827246376811594e-05,
+      "loss": 1.7292,
+      "step": 7700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3515872955322266,
+      "learning_rate": 1.5821449275362318e-05,
+      "loss": 1.6988,
+      "step": 7710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5109755992889404,
+      "learning_rate": 1.5815652173913046e-05,
+      "loss": 1.6356,
+      "step": 7720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2800748348236084,
+      "learning_rate": 1.580985507246377e-05,
+      "loss": 1.6617,
+      "step": 7730
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.169689178466797,
+      "learning_rate": 1.5804057971014494e-05,
+      "loss": 1.8323,
+      "step": 7740
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6964682340621948,
+      "learning_rate": 1.5798260869565218e-05,
+      "loss": 1.7671,
+      "step": 7750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.151012420654297,
+      "learning_rate": 1.5792463768115945e-05,
+      "loss": 1.7451,
+      "step": 7760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.739316463470459,
+      "learning_rate": 1.578666666666667e-05,
+      "loss": 1.7101,
+      "step": 7770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8832248449325562,
+      "learning_rate": 1.5780869565217393e-05,
+      "loss": 1.7384,
+      "step": 7780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.235473155975342,
+      "learning_rate": 1.5775072463768117e-05,
+      "loss": 1.6944,
+      "step": 7790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9296913146972656,
+      "learning_rate": 1.576927536231884e-05,
+      "loss": 1.7979,
+      "step": 7800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7439193725585938,
+      "learning_rate": 1.576347826086957e-05,
+      "loss": 1.6823,
+      "step": 7810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.104919195175171,
+      "learning_rate": 1.575768115942029e-05,
+      "loss": 1.9278,
+      "step": 7820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0970284938812256,
+      "learning_rate": 1.5751884057971017e-05,
+      "loss": 1.6475,
+      "step": 7830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.109387159347534,
+      "learning_rate": 1.574608695652174e-05,
+      "loss": 1.7482,
+      "step": 7840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.093923807144165,
+      "learning_rate": 1.5740289855072465e-05,
+      "loss": 1.708,
+      "step": 7850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.431930422782898,
+      "learning_rate": 1.573449275362319e-05,
+      "loss": 1.7585,
+      "step": 7860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5843312740325928,
+      "learning_rate": 1.5728695652173916e-05,
+      "loss": 1.7017,
+      "step": 7870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.592013120651245,
+      "learning_rate": 1.572289855072464e-05,
+      "loss": 1.6862,
+      "step": 7880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1581757068634033,
+      "learning_rate": 1.5717101449275364e-05,
+      "loss": 1.8041,
+      "step": 7890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.719002604484558,
+      "learning_rate": 1.5711304347826088e-05,
+      "loss": 1.7885,
+      "step": 7900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8887274265289307,
+      "learning_rate": 1.5705507246376812e-05,
+      "loss": 1.741,
+      "step": 7910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5597342252731323,
+      "learning_rate": 1.569971014492754e-05,
+      "loss": 1.7687,
+      "step": 7920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3231523036956787,
+      "learning_rate": 1.569391304347826e-05,
+      "loss": 1.7038,
+      "step": 7930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.540226936340332,
+      "learning_rate": 1.5688115942028987e-05,
+      "loss": 1.8191,
+      "step": 7940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.614171266555786,
+      "learning_rate": 1.568231884057971e-05,
+      "loss": 1.5657,
+      "step": 7950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2686009407043457,
+      "learning_rate": 1.5676521739130435e-05,
+      "loss": 1.6592,
+      "step": 7960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4677610397338867,
+      "learning_rate": 1.567072463768116e-05,
+      "loss": 1.702,
+      "step": 7970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.712357521057129,
+      "learning_rate": 1.5664927536231887e-05,
+      "loss": 1.7096,
+      "step": 7980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5932676792144775,
+      "learning_rate": 1.565913043478261e-05,
+      "loss": 1.6477,
+      "step": 7990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4786651134490967,
+      "learning_rate": 1.5653333333333335e-05,
+      "loss": 1.7013,
+      "step": 8000
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 1.751250982284546,
+      "eval_runtime": 107.5079,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 8000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.696633815765381,
+      "learning_rate": 1.564753623188406e-05,
+      "loss": 1.7176,
+      "step": 8010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.963792324066162,
+      "learning_rate": 1.5641739130434783e-05,
+      "loss": 1.7605,
+      "step": 8020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.107274055480957,
+      "learning_rate": 1.563594202898551e-05,
+      "loss": 1.5939,
+      "step": 8030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7392399311065674,
+      "learning_rate": 1.5630144927536234e-05,
+      "loss": 1.7734,
+      "step": 8040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.943000078201294,
+      "learning_rate": 1.5624347826086958e-05,
+      "loss": 1.6703,
+      "step": 8050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6716084480285645,
+      "learning_rate": 1.5618550724637682e-05,
+      "loss": 1.7177,
+      "step": 8060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.429185390472412,
+      "learning_rate": 1.561275362318841e-05,
+      "loss": 1.7043,
+      "step": 8070
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.558910608291626,
+      "learning_rate": 1.560695652173913e-05,
+      "loss": 1.7066,
+      "step": 8080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.654998302459717,
+      "learning_rate": 1.5601159420289854e-05,
+      "loss": 1.7063,
+      "step": 8090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6076571941375732,
+      "learning_rate": 1.559536231884058e-05,
+      "loss": 1.6818,
+      "step": 8100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5298787355422974,
+      "learning_rate": 1.5589565217391306e-05,
+      "loss": 1.6751,
+      "step": 8110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7017557621002197,
+      "learning_rate": 1.558376811594203e-05,
+      "loss": 1.7395,
+      "step": 8120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.02172589302063,
+      "learning_rate": 1.5577971014492754e-05,
+      "loss": 1.777,
+      "step": 8130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.633596897125244,
+      "learning_rate": 1.557217391304348e-05,
+      "loss": 1.7152,
+      "step": 8140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3255687952041626,
+      "learning_rate": 1.5566376811594205e-05,
+      "loss": 1.6739,
+      "step": 8150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6454014778137207,
+      "learning_rate": 1.556057971014493e-05,
+      "loss": 1.7551,
+      "step": 8160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2562685012817383,
+      "learning_rate": 1.5554782608695653e-05,
+      "loss": 1.7421,
+      "step": 8170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.259690761566162,
+      "learning_rate": 1.554898550724638e-05,
+      "loss": 1.6761,
+      "step": 8180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7266426086425781,
+      "learning_rate": 1.55431884057971e-05,
+      "loss": 1.8321,
+      "step": 8190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9167896509170532,
+      "learning_rate": 1.5537391304347825e-05,
+      "loss": 1.7285,
+      "step": 8200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.338323593139648,
+      "learning_rate": 1.5531594202898552e-05,
+      "loss": 1.7145,
+      "step": 8210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5180749893188477,
+      "learning_rate": 1.5525797101449276e-05,
+      "loss": 1.7482,
+      "step": 8220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.429786682128906,
+      "learning_rate": 1.552e-05,
+      "loss": 1.7808,
+      "step": 8230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.005053997039795,
+      "learning_rate": 1.5514202898550724e-05,
+      "loss": 1.7514,
+      "step": 8240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.007596969604492,
+      "learning_rate": 1.5508405797101452e-05,
+      "loss": 1.739,
+      "step": 8250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8556976318359375,
+      "learning_rate": 1.5502608695652176e-05,
+      "loss": 1.6269,
+      "step": 8260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3058815002441406,
+      "learning_rate": 1.54968115942029e-05,
+      "loss": 1.8055,
+      "step": 8270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.482654094696045,
+      "learning_rate": 1.5491014492753624e-05,
+      "loss": 1.7335,
+      "step": 8280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.104722499847412,
+      "learning_rate": 1.548521739130435e-05,
+      "loss": 1.6697,
+      "step": 8290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4763132333755493,
+      "learning_rate": 1.5479420289855075e-05,
+      "loss": 1.7573,
+      "step": 8300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5669054985046387,
+      "learning_rate": 1.5473623188405796e-05,
+      "loss": 1.737,
+      "step": 8310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9973231554031372,
+      "learning_rate": 1.5467826086956523e-05,
+      "loss": 1.828,
+      "step": 8320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6836585998535156,
+      "learning_rate": 1.5462028985507247e-05,
+      "loss": 1.6509,
+      "step": 8330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6782662868499756,
+      "learning_rate": 1.545623188405797e-05,
+      "loss": 1.7843,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.171507835388184,
+      "learning_rate": 1.5450434782608695e-05,
+      "loss": 1.697,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.005525588989258,
+      "learning_rate": 1.5444637681159423e-05,
+      "loss": 1.6807,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1129069328308105,
+      "learning_rate": 1.5438840579710147e-05,
+      "loss": 1.6926,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.524050235748291,
+      "learning_rate": 1.543304347826087e-05,
+      "loss": 1.7259,
+      "step": 8380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.451939105987549,
+      "learning_rate": 1.5427246376811595e-05,
+      "loss": 1.8682,
+      "step": 8390
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.816455841064453,
+      "learning_rate": 1.542144927536232e-05,
+      "loss": 1.6865,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2041363716125488,
+      "learning_rate": 1.5415652173913046e-05,
+      "loss": 1.7065,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.183321475982666,
+      "learning_rate": 1.540985507246377e-05,
+      "loss": 1.7404,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7429885864257812,
+      "learning_rate": 1.5404057971014494e-05,
+      "loss": 1.7694,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.275554656982422,
+      "learning_rate": 1.5398260869565218e-05,
+      "loss": 1.6809,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.737149953842163,
+      "learning_rate": 1.5392463768115946e-05,
+      "loss": 1.8422,
+      "step": 8450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7504504919052124,
+      "learning_rate": 1.5386666666666666e-05,
+      "loss": 1.7135,
+      "step": 8460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.513226270675659,
+      "learning_rate": 1.5380869565217394e-05,
+      "loss": 1.7704,
+      "step": 8470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.151536226272583,
+      "learning_rate": 1.5375072463768118e-05,
+      "loss": 1.6754,
+      "step": 8480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.003200054168701,
+      "learning_rate": 1.536927536231884e-05,
+      "loss": 1.7605,
+      "step": 8490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3642451763153076,
+      "learning_rate": 1.5363478260869566e-05,
+      "loss": 1.7791,
+      "step": 8500
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 1.728246808052063,
+      "eval_runtime": 107.4982,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 8500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0081084966659546,
+      "learning_rate": 1.535768115942029e-05,
+      "loss": 1.5997,
+      "step": 8510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.5596394538879395,
+      "learning_rate": 1.5351884057971017e-05,
+      "loss": 1.5862,
+      "step": 8520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.325129508972168,
+      "learning_rate": 1.534608695652174e-05,
+      "loss": 1.6014,
+      "step": 8530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.541696071624756,
+      "learning_rate": 1.5340289855072465e-05,
+      "loss": 1.7766,
+      "step": 8540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9516916275024414,
+      "learning_rate": 1.533449275362319e-05,
+      "loss": 1.7535,
+      "step": 8550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0460009574890137,
+      "learning_rate": 1.5328695652173916e-05,
+      "loss": 1.6929,
+      "step": 8560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.826079845428467,
+      "learning_rate": 1.5322898550724637e-05,
+      "loss": 1.7532,
+      "step": 8570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.370847702026367,
+      "learning_rate": 1.5317101449275364e-05,
+      "loss": 1.8819,
+      "step": 8580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.229450702667236,
+      "learning_rate": 1.531130434782609e-05,
+      "loss": 1.7454,
+      "step": 8590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1140810251235962,
+      "learning_rate": 1.5305507246376812e-05,
+      "loss": 1.6904,
+      "step": 8600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1529414653778076,
+      "learning_rate": 1.5299710144927536e-05,
+      "loss": 1.8351,
+      "step": 8610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1377217769622803,
+      "learning_rate": 1.529391304347826e-05,
+      "loss": 1.7735,
+      "step": 8620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.921389102935791,
+      "learning_rate": 1.5288115942028988e-05,
+      "loss": 1.7565,
+      "step": 8630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.921605110168457,
+      "learning_rate": 1.5282318840579712e-05,
+      "loss": 1.8189,
+      "step": 8640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4307273626327515,
+      "learning_rate": 1.5276521739130436e-05,
+      "loss": 1.8747,
+      "step": 8650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.513711452484131,
+      "learning_rate": 1.527072463768116e-05,
+      "loss": 1.7317,
+      "step": 8660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3792582750320435,
+      "learning_rate": 1.5264927536231887e-05,
+      "loss": 1.8341,
+      "step": 8670
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2375993728637695,
+      "learning_rate": 1.525913043478261e-05,
+      "loss": 1.665,
+      "step": 8680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7629612684249878,
+      "learning_rate": 1.5253333333333335e-05,
+      "loss": 1.7347,
+      "step": 8690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4616599082946777,
+      "learning_rate": 1.524753623188406e-05,
+      "loss": 1.6149,
+      "step": 8700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.200507164001465,
+      "learning_rate": 1.5241739130434783e-05,
+      "loss": 1.6675,
+      "step": 8710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.254258632659912,
+      "learning_rate": 1.5235942028985509e-05,
+      "loss": 1.7182,
+      "step": 8720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.435335636138916,
+      "learning_rate": 1.5230144927536233e-05,
+      "loss": 1.738,
+      "step": 8730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.216738224029541,
+      "learning_rate": 1.5224347826086959e-05,
+      "loss": 1.6982,
+      "step": 8740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1532115936279297,
+      "learning_rate": 1.5218550724637681e-05,
+      "loss": 1.6887,
+      "step": 8750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.216334581375122,
+      "learning_rate": 1.5212753623188408e-05,
+      "loss": 1.7531,
+      "step": 8760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3834766149520874,
+      "learning_rate": 1.520695652173913e-05,
+      "loss": 1.7209,
+      "step": 8770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.171159267425537,
+      "learning_rate": 1.5201159420289856e-05,
+      "loss": 1.7489,
+      "step": 8780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.972515344619751,
+      "learning_rate": 1.519536231884058e-05,
+      "loss": 1.641,
+      "step": 8790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.618605136871338,
+      "learning_rate": 1.5189565217391306e-05,
+      "loss": 1.7289,
+      "step": 8800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3899489641189575,
+      "learning_rate": 1.518376811594203e-05,
+      "loss": 1.7706,
+      "step": 8810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.194946527481079,
+      "learning_rate": 1.5177971014492754e-05,
+      "loss": 1.8283,
+      "step": 8820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.15311598777771,
+      "learning_rate": 1.517217391304348e-05,
+      "loss": 1.704,
+      "step": 8830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4149380922317505,
+      "learning_rate": 1.5166376811594204e-05,
+      "loss": 1.6738,
+      "step": 8840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4329111576080322,
+      "learning_rate": 1.516057971014493e-05,
+      "loss": 1.7411,
+      "step": 8850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.406054973602295,
+      "learning_rate": 1.5154782608695654e-05,
+      "loss": 1.8027,
+      "step": 8860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.0179643630981445,
+      "learning_rate": 1.514898550724638e-05,
+      "loss": 1.7083,
+      "step": 8870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3201992511749268,
+      "learning_rate": 1.5143188405797103e-05,
+      "loss": 1.6307,
+      "step": 8880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.243584632873535,
+      "learning_rate": 1.5137391304347829e-05,
+      "loss": 1.7504,
+      "step": 8890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8807439804077148,
+      "learning_rate": 1.5131594202898551e-05,
+      "loss": 1.7607,
+      "step": 8900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.392615556716919,
+      "learning_rate": 1.5125797101449277e-05,
+      "loss": 1.6733,
+      "step": 8910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.388437032699585,
+      "learning_rate": 1.5120000000000001e-05,
+      "loss": 1.6187,
+      "step": 8920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.407155990600586,
+      "learning_rate": 1.5114202898550725e-05,
+      "loss": 1.6473,
+      "step": 8930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9519150257110596,
+      "learning_rate": 1.510840579710145e-05,
+      "loss": 1.7244,
+      "step": 8940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5977227687835693,
+      "learning_rate": 1.5102608695652175e-05,
+      "loss": 1.7157,
+      "step": 8950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.951735019683838,
+      "learning_rate": 1.50968115942029e-05,
+      "loss": 1.7765,
+      "step": 8960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0863535404205322,
+      "learning_rate": 1.5091014492753624e-05,
+      "loss": 1.5254,
+      "step": 8970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2637622356414795,
+      "learning_rate": 1.508521739130435e-05,
+      "loss": 1.808,
+      "step": 8980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3032875061035156,
+      "learning_rate": 1.5079420289855074e-05,
+      "loss": 1.7458,
+      "step": 8990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.13789701461792,
+      "learning_rate": 1.50736231884058e-05,
+      "loss": 1.6349,
+      "step": 9000
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.7237039804458618,
+      "eval_runtime": 107.5052,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 9000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9929821491241455,
+      "learning_rate": 1.5067826086956524e-05,
+      "loss": 1.7925,
+      "step": 9010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.004290580749512,
+      "learning_rate": 1.5062028985507246e-05,
+      "loss": 1.6467,
+      "step": 9020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.387837529182434,
+      "learning_rate": 1.5056231884057972e-05,
+      "loss": 1.6188,
+      "step": 9030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9127049446105957,
+      "learning_rate": 1.5050434782608696e-05,
+      "loss": 1.6898,
+      "step": 9040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.460002899169922,
+      "learning_rate": 1.5044637681159421e-05,
+      "loss": 1.7306,
+      "step": 9050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.018058776855469,
+      "learning_rate": 1.5038840579710145e-05,
+      "loss": 1.6259,
+      "step": 9060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8971534967422485,
+      "learning_rate": 1.5033043478260871e-05,
+      "loss": 1.6011,
+      "step": 9070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.7335357666015625,
+      "learning_rate": 1.5027246376811595e-05,
+      "loss": 1.7402,
+      "step": 9080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.477769613265991,
+      "learning_rate": 1.5021449275362321e-05,
+      "loss": 1.7765,
+      "step": 9090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.670304775238037,
+      "learning_rate": 1.5015652173913045e-05,
+      "loss": 1.8922,
+      "step": 9100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.747793197631836,
+      "learning_rate": 1.500985507246377e-05,
+      "loss": 1.6625,
+      "step": 9110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.719919204711914,
+      "learning_rate": 1.5004057971014495e-05,
+      "loss": 1.7383,
+      "step": 9120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1044466495513916,
+      "learning_rate": 1.4998260869565217e-05,
+      "loss": 1.8098,
+      "step": 9130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7055985927581787,
+      "learning_rate": 1.4992463768115944e-05,
+      "loss": 1.4545,
+      "step": 9140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7693852186203003,
+      "learning_rate": 1.4986666666666667e-05,
+      "loss": 1.6745,
+      "step": 9150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0475828647613525,
+      "learning_rate": 1.4980869565217392e-05,
+      "loss": 1.5843,
+      "step": 9160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5320639610290527,
+      "learning_rate": 1.4975072463768116e-05,
+      "loss": 1.6735,
+      "step": 9170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.41857647895813,
+      "learning_rate": 1.4969275362318842e-05,
+      "loss": 1.6797,
+      "step": 9180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8437910079956055,
+      "learning_rate": 1.4963478260869566e-05,
+      "loss": 1.7409,
+      "step": 9190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.104484796524048,
+      "learning_rate": 1.4957681159420292e-05,
+      "loss": 1.5646,
+      "step": 9200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2742903232574463,
+      "learning_rate": 1.4951884057971016e-05,
+      "loss": 1.8273,
+      "step": 9210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.759800434112549,
+      "learning_rate": 1.494608695652174e-05,
+      "loss": 1.5638,
+      "step": 9220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4003862142562866,
+      "learning_rate": 1.4940289855072465e-05,
+      "loss": 1.8205,
+      "step": 9230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.187140703201294,
+      "learning_rate": 1.493449275362319e-05,
+      "loss": 1.6775,
+      "step": 9240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.953629493713379,
+      "learning_rate": 1.4928695652173915e-05,
+      "loss": 1.7948,
+      "step": 9250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.153158664703369,
+      "learning_rate": 1.4922898550724637e-05,
+      "loss": 1.7641,
+      "step": 9260
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8816325664520264,
+      "learning_rate": 1.4917101449275365e-05,
+      "loss": 1.7967,
+      "step": 9270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1909523010253906,
+      "learning_rate": 1.4911304347826087e-05,
+      "loss": 1.7295,
+      "step": 9280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3025991916656494,
+      "learning_rate": 1.4905507246376813e-05,
+      "loss": 1.7594,
+      "step": 9290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.461934566497803,
+      "learning_rate": 1.4899710144927537e-05,
+      "loss": 1.8171,
+      "step": 9300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.337406158447266,
+      "learning_rate": 1.4893913043478263e-05,
+      "loss": 1.6085,
+      "step": 9310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.752158284187317,
+      "learning_rate": 1.4888115942028987e-05,
+      "loss": 1.6972,
+      "step": 9320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.966576099395752,
+      "learning_rate": 1.488231884057971e-05,
+      "loss": 1.6919,
+      "step": 9330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.894684076309204,
+      "learning_rate": 1.4876521739130436e-05,
+      "loss": 1.4976,
+      "step": 9340
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2293604612350464,
+      "learning_rate": 1.487072463768116e-05,
+      "loss": 1.6831,
+      "step": 9350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8790899515151978,
+      "learning_rate": 1.4864927536231886e-05,
+      "loss": 1.7635,
+      "step": 9360
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.168088436126709,
+      "learning_rate": 1.485913043478261e-05,
+      "loss": 1.6389,
+      "step": 9370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0390825271606445,
+      "learning_rate": 1.4853333333333336e-05,
+      "loss": 1.6824,
+      "step": 9380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.15370512008667,
+      "learning_rate": 1.4847536231884058e-05,
+      "loss": 1.6074,
+      "step": 9390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2792294025421143,
+      "learning_rate": 1.4841739130434785e-05,
+      "loss": 1.6977,
+      "step": 9400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7052159309387207,
+      "learning_rate": 1.4835942028985508e-05,
+      "loss": 1.6734,
+      "step": 9410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8690690994262695,
+      "learning_rate": 1.4830144927536233e-05,
+      "loss": 1.7114,
+      "step": 9420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2574138641357422,
+      "learning_rate": 1.4824347826086957e-05,
+      "loss": 1.4424,
+      "step": 9430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.904484748840332,
+      "learning_rate": 1.4818550724637681e-05,
+      "loss": 1.7067,
+      "step": 9440
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.294973611831665,
+      "learning_rate": 1.4812753623188407e-05,
+      "loss": 1.8665,
+      "step": 9450
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.394903182983398,
+      "learning_rate": 1.4806956521739131e-05,
+      "loss": 1.5927,
+      "step": 9460
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6690807342529297,
+      "learning_rate": 1.4801159420289857e-05,
+      "loss": 1.6617,
+      "step": 9470
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6983529329299927,
+      "learning_rate": 1.479536231884058e-05,
+      "loss": 1.7189,
+      "step": 9480
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5802037715911865,
+      "learning_rate": 1.4789565217391307e-05,
+      "loss": 1.4436,
+      "step": 9490
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2202868461608887,
+      "learning_rate": 1.478376811594203e-05,
+      "loss": 1.8035,
+      "step": 9500
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 1.7105869054794312,
+      "eval_runtime": 107.5846,
+      "eval_samples_per_second": 9.295,
+      "eval_steps_per_second": 2.324,
+      "step": 9500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4393227100372314,
+      "learning_rate": 1.4777971014492756e-05,
+      "loss": 1.5999,
+      "step": 9510
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.443208694458008,
+      "learning_rate": 1.4772173913043479e-05,
+      "loss": 1.6505,
+      "step": 9520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.468209743499756,
+      "learning_rate": 1.4766376811594203e-05,
+      "loss": 1.6897,
+      "step": 9530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3215961456298828,
+      "learning_rate": 1.4760579710144928e-05,
+      "loss": 1.8336,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.739179849624634,
+      "learning_rate": 1.4754782608695652e-05,
+      "loss": 1.8085,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8474160432815552,
+      "learning_rate": 1.4748985507246378e-05,
+      "loss": 1.8288,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9103896617889404,
+      "learning_rate": 1.4743188405797102e-05,
+      "loss": 1.7058,
+      "step": 9570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.432541847229004,
+      "learning_rate": 1.4737391304347828e-05,
+      "loss": 1.7497,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.380267381668091,
+      "learning_rate": 1.4731594202898552e-05,
+      "loss": 1.6908,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8983180522918701,
+      "learning_rate": 1.4725797101449277e-05,
+      "loss": 1.7913,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5178720951080322,
+      "learning_rate": 1.4720000000000001e-05,
+      "loss": 1.6481,
+      "step": 9610
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.430023670196533,
+      "learning_rate": 1.4714202898550727e-05,
+      "loss": 1.7796,
+      "step": 9620
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9804303646087646,
+      "learning_rate": 1.4708405797101451e-05,
+      "loss": 1.7676,
+      "step": 9630
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.044782638549805,
+      "learning_rate": 1.4702608695652173e-05,
+      "loss": 1.6331,
+      "step": 9640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2377848625183105,
+      "learning_rate": 1.46968115942029e-05,
+      "loss": 1.6992,
+      "step": 9650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.13010835647583,
+      "learning_rate": 1.4691014492753623e-05,
+      "loss": 1.7363,
+      "step": 9660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.638706684112549,
+      "learning_rate": 1.4685217391304349e-05,
+      "loss": 1.6856,
+      "step": 9670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5270848274230957,
+      "learning_rate": 1.4679420289855073e-05,
+      "loss": 1.694,
+      "step": 9680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1668355464935303,
+      "learning_rate": 1.4673623188405798e-05,
+      "loss": 1.5912,
+      "step": 9690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8836257457733154,
+      "learning_rate": 1.466840579710145e-05,
+      "loss": 1.7058,
+      "step": 9700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6809861660003662,
+      "learning_rate": 1.4662608695652174e-05,
+      "loss": 1.6679,
+      "step": 9710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5727224349975586,
+      "learning_rate": 1.46568115942029e-05,
+      "loss": 1.7444,
+      "step": 9720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.230935573577881,
+      "learning_rate": 1.4651014492753623e-05,
+      "loss": 1.649,
+      "step": 9730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9964046478271484,
+      "learning_rate": 1.464521739130435e-05,
+      "loss": 1.6816,
+      "step": 9740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.299757480621338,
+      "learning_rate": 1.4639420289855073e-05,
+      "loss": 1.7576,
+      "step": 9750
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.057912588119507,
+      "learning_rate": 1.4633623188405799e-05,
+      "loss": 1.7915,
+      "step": 9760
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.693455934524536,
+      "learning_rate": 1.4627826086956523e-05,
+      "loss": 1.7852,
+      "step": 9770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2792394161224365,
+      "learning_rate": 1.4622028985507249e-05,
+      "loss": 1.7702,
+      "step": 9780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9166826009750366,
+      "learning_rate": 1.4616231884057973e-05,
+      "loss": 1.5197,
+      "step": 9790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.76140832901001,
+      "learning_rate": 1.4610434782608698e-05,
+      "loss": 1.7217,
+      "step": 9800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.99151611328125,
+      "learning_rate": 1.4604637681159422e-05,
+      "loss": 1.8153,
+      "step": 9810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7407947778701782,
+      "learning_rate": 1.4598840579710145e-05,
+      "loss": 1.7537,
+      "step": 9820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.095398426055908,
+      "learning_rate": 1.459304347826087e-05,
+      "loss": 1.7267,
+      "step": 9830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.100191593170166,
+      "learning_rate": 1.4587246376811594e-05,
+      "loss": 1.6822,
+      "step": 9840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.774409294128418,
+      "learning_rate": 1.458144927536232e-05,
+      "loss": 1.7082,
+      "step": 9850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7792176008224487,
+      "learning_rate": 1.4575652173913044e-05,
+      "loss": 1.6193,
+      "step": 9860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.539458751678467,
+      "learning_rate": 1.456985507246377e-05,
+      "loss": 1.6587,
+      "step": 9870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.321563243865967,
+      "learning_rate": 1.4564057971014494e-05,
+      "loss": 1.7235,
+      "step": 9880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7294243574142456,
+      "learning_rate": 1.455826086956522e-05,
+      "loss": 1.7623,
+      "step": 9890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.641298532485962,
+      "learning_rate": 1.4552463768115943e-05,
+      "loss": 1.7657,
+      "step": 9900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3333618640899658,
+      "learning_rate": 1.4546666666666669e-05,
+      "loss": 1.6669,
+      "step": 9910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8383522033691406,
+      "learning_rate": 1.4540869565217393e-05,
+      "loss": 1.7376,
+      "step": 9920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.071063995361328,
+      "learning_rate": 1.4535072463768115e-05,
+      "loss": 1.6751,
+      "step": 9930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.008157730102539,
+      "learning_rate": 1.4529275362318843e-05,
+      "loss": 1.6904,
+      "step": 9940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.280172824859619,
+      "learning_rate": 1.4523478260869565e-05,
+      "loss": 1.5522,
+      "step": 9950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.36051607131958,
+      "learning_rate": 1.4517681159420291e-05,
+      "loss": 1.6922,
+      "step": 9960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.0582804679870605,
+      "learning_rate": 1.4511884057971015e-05,
+      "loss": 1.7,
+      "step": 9970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5071215629577637,
+      "learning_rate": 1.450608695652174e-05,
+      "loss": 1.6065,
+      "step": 9980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.5151567459106445,
+      "learning_rate": 1.4500289855072465e-05,
+      "loss": 1.646,
+      "step": 9990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.011928081512451,
+      "learning_rate": 1.449449275362319e-05,
+      "loss": 1.8767,
+      "step": 10000
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 1.7204252481460571,
+      "eval_runtime": 107.4981,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 10000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8452903032302856,
+      "learning_rate": 1.4488695652173914e-05,
+      "loss": 1.6305,
+      "step": 10010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8112026453018188,
+      "learning_rate": 1.4482898550724638e-05,
+      "loss": 1.6268,
+      "step": 10020
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9594013094902039,
+      "learning_rate": 1.4477101449275364e-05,
+      "loss": 1.4332,
+      "step": 10030
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1587038040161133,
+      "learning_rate": 1.4471304347826088e-05,
+      "loss": 1.8313,
+      "step": 10040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3806848526000977,
+      "learning_rate": 1.4465507246376814e-05,
+      "loss": 1.6507,
+      "step": 10050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.119511127471924,
+      "learning_rate": 1.4459710144927536e-05,
+      "loss": 1.8907,
+      "step": 10060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.085169553756714,
+      "learning_rate": 1.4453913043478263e-05,
+      "loss": 1.7827,
+      "step": 10070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.273642063140869,
+      "learning_rate": 1.4448115942028986e-05,
+      "loss": 1.8521,
+      "step": 10080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0806262493133545,
+      "learning_rate": 1.4442318840579711e-05,
+      "loss": 1.7231,
+      "step": 10090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6042985916137695,
+      "learning_rate": 1.4436521739130435e-05,
+      "loss": 1.6173,
+      "step": 10100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.670179843902588,
+      "learning_rate": 1.4430724637681161e-05,
+      "loss": 1.8075,
+      "step": 10110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.045022010803223,
+      "learning_rate": 1.4424927536231885e-05,
+      "loss": 1.7247,
+      "step": 10120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5309295654296875,
+      "learning_rate": 1.4419130434782609e-05,
+      "loss": 1.7216,
+      "step": 10130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8289453983306885,
+      "learning_rate": 1.4413333333333335e-05,
+      "loss": 1.7299,
+      "step": 10140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.870276927947998,
+      "learning_rate": 1.4407536231884059e-05,
+      "loss": 1.6632,
+      "step": 10150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4293287992477417,
+      "learning_rate": 1.4401739130434785e-05,
+      "loss": 1.7234,
+      "step": 10160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1435534954071045,
+      "learning_rate": 1.4395942028985509e-05,
+      "loss": 1.7715,
+      "step": 10170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4607239961624146,
+      "learning_rate": 1.4390144927536234e-05,
+      "loss": 1.728,
+      "step": 10180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.555734872817993,
+      "learning_rate": 1.4384347826086957e-05,
+      "loss": 1.8154,
+      "step": 10190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3336338996887207,
+      "learning_rate": 1.4378550724637684e-05,
+      "loss": 1.6378,
+      "step": 10200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9876394271850586,
+      "learning_rate": 1.4372753623188406e-05,
+      "loss": 1.6636,
+      "step": 10210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9591132402420044,
+      "learning_rate": 1.4366956521739132e-05,
+      "loss": 1.7456,
+      "step": 10220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.272648572921753,
+      "learning_rate": 1.4361159420289856e-05,
+      "loss": 1.8266,
+      "step": 10230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.858809947967529,
+      "learning_rate": 1.435536231884058e-05,
+      "loss": 1.6315,
+      "step": 10240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7406672835350037,
+      "learning_rate": 1.4349565217391306e-05,
+      "loss": 1.4722,
+      "step": 10250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0749115943908691,
+      "learning_rate": 1.434376811594203e-05,
+      "loss": 1.8539,
+      "step": 10260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3175392150878906,
+      "learning_rate": 1.4337971014492755e-05,
+      "loss": 1.7049,
+      "step": 10270
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.374159812927246,
+      "learning_rate": 1.433217391304348e-05,
+      "loss": 1.7605,
+      "step": 10280
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9364595413208008,
+      "learning_rate": 1.4326376811594205e-05,
+      "loss": 1.6288,
+      "step": 10290
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.389453887939453,
+      "learning_rate": 1.4320579710144929e-05,
+      "loss": 1.6014,
+      "step": 10300
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7157788276672363,
+      "learning_rate": 1.4314782608695655e-05,
+      "loss": 1.738,
+      "step": 10310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8316123485565186,
+      "learning_rate": 1.4308985507246377e-05,
+      "loss": 1.727,
+      "step": 10320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3950917720794678,
+      "learning_rate": 1.4303188405797101e-05,
+      "loss": 1.6789,
+      "step": 10330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.218912601470947,
+      "learning_rate": 1.4297391304347827e-05,
+      "loss": 1.8176,
+      "step": 10340
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7928733825683594,
+      "learning_rate": 1.429159420289855e-05,
+      "loss": 1.6744,
+      "step": 10350
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1876684427261353,
+      "learning_rate": 1.4285797101449276e-05,
+      "loss": 1.6813,
+      "step": 10360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4012842178344727,
+      "learning_rate": 1.428e-05,
+      "loss": 1.7983,
+      "step": 10370
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.704860210418701,
+      "learning_rate": 1.4274202898550726e-05,
+      "loss": 1.6289,
+      "step": 10380
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7611334323883057,
+      "learning_rate": 1.426840579710145e-05,
+      "loss": 1.6068,
+      "step": 10390
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.28679084777832,
+      "learning_rate": 1.4262608695652176e-05,
+      "loss": 1.644,
+      "step": 10400
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1040380001068115,
+      "learning_rate": 1.42568115942029e-05,
+      "loss": 1.8251,
+      "step": 10410
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5401835441589355,
+      "learning_rate": 1.4251014492753626e-05,
+      "loss": 1.6375,
+      "step": 10420
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4024248123168945,
+      "learning_rate": 1.424521739130435e-05,
+      "loss": 1.6935,
+      "step": 10430
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.5380167961120605,
+      "learning_rate": 1.4239420289855072e-05,
+      "loss": 1.8278,
+      "step": 10440
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.322249412536621,
+      "learning_rate": 1.42336231884058e-05,
+      "loss": 1.6748,
+      "step": 10450
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3199710845947266,
+      "learning_rate": 1.4227826086956522e-05,
+      "loss": 1.7673,
+      "step": 10460
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2281827926635742,
+      "learning_rate": 1.4222028985507247e-05,
+      "loss": 1.7763,
+      "step": 10470
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.571286678314209,
+      "learning_rate": 1.4216231884057971e-05,
+      "loss": 1.5969,
+      "step": 10480
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.81419038772583,
+      "learning_rate": 1.4210434782608697e-05,
+      "loss": 1.7263,
+      "step": 10490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.55985951423645,
+      "learning_rate": 1.4204637681159421e-05,
+      "loss": 1.7246,
+      "step": 10500
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 1.7039064168930054,
+      "eval_runtime": 107.5109,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 10500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4727416038513184,
+      "learning_rate": 1.4198840579710147e-05,
+      "loss": 1.7325,
+      "step": 10510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.254564046859741,
+      "learning_rate": 1.419304347826087e-05,
+      "loss": 1.6793,
+      "step": 10520
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.339053153991699,
+      "learning_rate": 1.4187246376811596e-05,
+      "loss": 1.8367,
+      "step": 10530
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8006272315979004,
+      "learning_rate": 1.418144927536232e-05,
+      "loss": 1.5286,
+      "step": 10540
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7751678228378296,
+      "learning_rate": 1.4175652173913044e-05,
+      "loss": 1.7536,
+      "step": 10550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.587690830230713,
+      "learning_rate": 1.416985507246377e-05,
+      "loss": 1.8107,
+      "step": 10560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8632152080535889,
+      "learning_rate": 1.4164057971014492e-05,
+      "loss": 1.6903,
+      "step": 10570
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.762665033340454,
+      "learning_rate": 1.415826086956522e-05,
+      "loss": 1.7828,
+      "step": 10580
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1041312217712402,
+      "learning_rate": 1.4152463768115942e-05,
+      "loss": 1.8221,
+      "step": 10590
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7822948694229126,
+      "learning_rate": 1.4146666666666668e-05,
+      "loss": 1.7623,
+      "step": 10600
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2196130752563477,
+      "learning_rate": 1.4140869565217392e-05,
+      "loss": 1.8126,
+      "step": 10610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2900569438934326,
+      "learning_rate": 1.4135072463768118e-05,
+      "loss": 1.7031,
+      "step": 10620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.033328056335449,
+      "learning_rate": 1.4129275362318842e-05,
+      "loss": 1.7968,
+      "step": 10630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2577106952667236,
+      "learning_rate": 1.4123478260869566e-05,
+      "loss": 1.69,
+      "step": 10640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.332272052764893,
+      "learning_rate": 1.4117681159420291e-05,
+      "loss": 1.6143,
+      "step": 10650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.976055383682251,
+      "learning_rate": 1.4111884057971015e-05,
+      "loss": 1.7698,
+      "step": 10660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.821053385734558,
+      "learning_rate": 1.4106086956521741e-05,
+      "loss": 1.8864,
+      "step": 10670
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.310410261154175,
+      "learning_rate": 1.4100289855072465e-05,
+      "loss": 1.583,
+      "step": 10680
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4083127975463867,
+      "learning_rate": 1.409449275362319e-05,
+      "loss": 1.7432,
+      "step": 10690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7858573198318481,
+      "learning_rate": 1.4088695652173913e-05,
+      "loss": 1.566,
+      "step": 10700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7986173629760742,
+      "learning_rate": 1.408289855072464e-05,
+      "loss": 1.7651,
+      "step": 10710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9657052755355835,
+      "learning_rate": 1.4077101449275363e-05,
+      "loss": 1.7637,
+      "step": 10720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.002399206161499,
+      "learning_rate": 1.4071304347826088e-05,
+      "loss": 1.8534,
+      "step": 10730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.677050828933716,
+      "learning_rate": 1.4065507246376812e-05,
+      "loss": 1.8067,
+      "step": 10740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9738922119140625,
+      "learning_rate": 1.4059710144927536e-05,
+      "loss": 1.7928,
+      "step": 10750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2202324867248535,
+      "learning_rate": 1.4053913043478262e-05,
+      "loss": 1.7436,
+      "step": 10760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8412764072418213,
+      "learning_rate": 1.4048115942028986e-05,
+      "loss": 1.7523,
+      "step": 10770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8097102046012878,
+      "learning_rate": 1.4042318840579712e-05,
+      "loss": 1.7362,
+      "step": 10780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.471013069152832,
+      "learning_rate": 1.4036521739130436e-05,
+      "loss": 1.7287,
+      "step": 10790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.12705659866333,
+      "learning_rate": 1.4030724637681162e-05,
+      "loss": 1.7023,
+      "step": 10800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3543355464935303,
+      "learning_rate": 1.4024927536231886e-05,
+      "loss": 1.7299,
+      "step": 10810
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.779575824737549,
+      "learning_rate": 1.4019130434782611e-05,
+      "loss": 1.668,
+      "step": 10820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1106367111206055,
+      "learning_rate": 1.4013333333333334e-05,
+      "loss": 1.713,
+      "step": 10830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9769501686096191,
+      "learning_rate": 1.4007536231884061e-05,
+      "loss": 1.6006,
+      "step": 10840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.193175792694092,
+      "learning_rate": 1.4001739130434783e-05,
+      "loss": 1.734,
+      "step": 10850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.645627737045288,
+      "learning_rate": 1.3995942028985507e-05,
+      "loss": 1.5706,
+      "step": 10860
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.674108982086182,
+      "learning_rate": 1.3990144927536233e-05,
+      "loss": 1.702,
+      "step": 10870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0062819719314575,
+      "learning_rate": 1.3984347826086957e-05,
+      "loss": 1.7267,
+      "step": 10880
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.037877559661865,
+      "learning_rate": 1.3978550724637683e-05,
+      "loss": 1.7847,
+      "step": 10890
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.889549493789673,
+      "learning_rate": 1.3972753623188407e-05,
+      "loss": 1.6711,
+      "step": 10900
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1683433055877686,
+      "learning_rate": 1.3966956521739132e-05,
+      "loss": 1.7289,
+      "step": 10910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.776911973953247,
+      "learning_rate": 1.3961159420289856e-05,
+      "loss": 1.6323,
+      "step": 10920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.037374973297119,
+      "learning_rate": 1.3955942028985508e-05,
+      "loss": 1.7624,
+      "step": 10930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.559633255004883,
+      "learning_rate": 1.3950144927536233e-05,
+      "loss": 1.6858,
+      "step": 10940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2170414924621582,
+      "learning_rate": 1.3944347826086957e-05,
+      "loss": 1.7849,
+      "step": 10950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.384916305541992,
+      "learning_rate": 1.3938550724637683e-05,
+      "loss": 1.6399,
+      "step": 10960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4139840602874756,
+      "learning_rate": 1.3932753623188407e-05,
+      "loss": 1.7363,
+      "step": 10970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0569369792938232,
+      "learning_rate": 1.3926956521739133e-05,
+      "loss": 1.8243,
+      "step": 10980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9499967098236084,
+      "learning_rate": 1.3921159420289855e-05,
+      "loss": 1.8001,
+      "step": 10990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.538534164428711,
+      "learning_rate": 1.3915362318840582e-05,
+      "loss": 1.5845,
+      "step": 11000
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 1.7137649059295654,
+      "eval_runtime": 107.4981,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 11000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2500182390213013,
+      "learning_rate": 1.3909565217391305e-05,
+      "loss": 1.7198,
+      "step": 11010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7554583549499512,
+      "learning_rate": 1.3903768115942029e-05,
+      "loss": 1.678,
+      "step": 11020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5758776664733887,
+      "learning_rate": 1.3897971014492754e-05,
+      "loss": 1.7378,
+      "step": 11030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.143527984619141,
+      "learning_rate": 1.3892173913043478e-05,
+      "loss": 1.7865,
+      "step": 11040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.270503282546997,
+      "learning_rate": 1.3886376811594204e-05,
+      "loss": 1.7676,
+      "step": 11050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5987863540649414,
+      "learning_rate": 1.3880579710144928e-05,
+      "loss": 1.7073,
+      "step": 11060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.498621702194214,
+      "learning_rate": 1.3874782608695654e-05,
+      "loss": 1.7672,
+      "step": 11070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1236436367034912,
+      "learning_rate": 1.3868985507246378e-05,
+      "loss": 1.8432,
+      "step": 11080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.279869079589844,
+      "learning_rate": 1.3863188405797104e-05,
+      "loss": 1.6825,
+      "step": 11090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.814366817474365,
+      "learning_rate": 1.3857391304347828e-05,
+      "loss": 1.6321,
+      "step": 11100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.510978102684021,
+      "learning_rate": 1.3851594202898553e-05,
+      "loss": 1.6527,
+      "step": 11110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.76588773727417,
+      "learning_rate": 1.3845797101449276e-05,
+      "loss": 1.7147,
+      "step": 11120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.148089408874512,
+      "learning_rate": 1.384e-05,
+      "loss": 1.7523,
+      "step": 11130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.291975736618042,
+      "learning_rate": 1.3834202898550725e-05,
+      "loss": 1.6122,
+      "step": 11140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4009017944335938,
+      "learning_rate": 1.382840579710145e-05,
+      "loss": 1.7563,
+      "step": 11150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2660974264144897,
+      "learning_rate": 1.3822608695652175e-05,
+      "loss": 1.8469,
+      "step": 11160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1221381425857544,
+      "learning_rate": 1.3816811594202899e-05,
+      "loss": 1.8096,
+      "step": 11170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3825074434280396,
+      "learning_rate": 1.3811014492753625e-05,
+      "loss": 1.7913,
+      "step": 11180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.367438316345215,
+      "learning_rate": 1.3805217391304349e-05,
+      "loss": 1.7117,
+      "step": 11190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.15596866607666,
+      "learning_rate": 1.3799420289855074e-05,
+      "loss": 1.7031,
+      "step": 11200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.4687042236328125,
+      "learning_rate": 1.3793623188405798e-05,
+      "loss": 1.6699,
+      "step": 11210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.699357509613037,
+      "learning_rate": 1.3787826086956524e-05,
+      "loss": 1.829,
+      "step": 11220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6832666397094727,
+      "learning_rate": 1.3782028985507248e-05,
+      "loss": 1.5892,
+      "step": 11230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3249785900115967,
+      "learning_rate": 1.377623188405797e-05,
+      "loss": 1.8073,
+      "step": 11240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6117970943450928,
+      "learning_rate": 1.3770434782608698e-05,
+      "loss": 1.684,
+      "step": 11250
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9994568824768066,
+      "learning_rate": 1.376463768115942e-05,
+      "loss": 1.6926,
+      "step": 11260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.465999126434326,
+      "learning_rate": 1.3758840579710146e-05,
+      "loss": 1.6104,
+      "step": 11270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.657724380493164,
+      "learning_rate": 1.375304347826087e-05,
+      "loss": 1.8288,
+      "step": 11280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.94265615940094,
+      "learning_rate": 1.3747246376811596e-05,
+      "loss": 1.7035,
+      "step": 11290
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6312084197998047,
+      "learning_rate": 1.374144927536232e-05,
+      "loss": 1.7352,
+      "step": 11300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.722726583480835,
+      "learning_rate": 1.3735652173913045e-05,
+      "loss": 1.7667,
+      "step": 11310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.727392196655273,
+      "learning_rate": 1.372985507246377e-05,
+      "loss": 1.5222,
+      "step": 11320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.453555583953857,
+      "learning_rate": 1.3724057971014493e-05,
+      "loss": 1.8861,
+      "step": 11330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.511663913726807,
+      "learning_rate": 1.3718260869565219e-05,
+      "loss": 1.836,
+      "step": 11340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.072493553161621,
+      "learning_rate": 1.3712463768115943e-05,
+      "loss": 1.5072,
+      "step": 11350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9541881084442139,
+      "learning_rate": 1.3706666666666669e-05,
+      "loss": 1.6182,
+      "step": 11360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.463501453399658,
+      "learning_rate": 1.3700869565217391e-05,
+      "loss": 1.8108,
+      "step": 11370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4841718673706055,
+      "learning_rate": 1.3695072463768118e-05,
+      "loss": 1.8205,
+      "step": 11380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.357886791229248,
+      "learning_rate": 1.368927536231884e-05,
+      "loss": 1.6904,
+      "step": 11390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.65535306930542,
+      "learning_rate": 1.3683478260869566e-05,
+      "loss": 1.7598,
+      "step": 11400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.2453694343566895,
+      "learning_rate": 1.367768115942029e-05,
+      "loss": 1.6009,
+      "step": 11410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.134276866912842,
+      "learning_rate": 1.3671884057971016e-05,
+      "loss": 1.7633,
+      "step": 11420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.813711643218994,
+      "learning_rate": 1.366608695652174e-05,
+      "loss": 1.5795,
+      "step": 11430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.364342451095581,
+      "learning_rate": 1.3660289855072464e-05,
+      "loss": 1.6675,
+      "step": 11440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.627918004989624,
+      "learning_rate": 1.365449275362319e-05,
+      "loss": 1.7405,
+      "step": 11450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7319889068603516,
+      "learning_rate": 1.3648695652173914e-05,
+      "loss": 1.6259,
+      "step": 11460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.867110013961792,
+      "learning_rate": 1.364289855072464e-05,
+      "loss": 1.5038,
+      "step": 11470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0498149394989014,
+      "learning_rate": 1.3637101449275364e-05,
+      "loss": 1.557,
+      "step": 11480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.804266929626465,
+      "learning_rate": 1.363130434782609e-05,
+      "loss": 1.4778,
+      "step": 11490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.833094120025635,
+      "learning_rate": 1.3625507246376812e-05,
+      "loss": 1.7432,
+      "step": 11500
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.6946756839752197,
+      "eval_runtime": 107.5258,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 11500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.91051185131073,
+      "learning_rate": 1.3619710144927539e-05,
+      "loss": 1.6593,
+      "step": 11510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2992429733276367,
+      "learning_rate": 1.3613913043478261e-05,
+      "loss": 1.6403,
+      "step": 11520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9277349710464478,
+      "learning_rate": 1.3608115942028987e-05,
+      "loss": 1.7522,
+      "step": 11530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0008699893951416,
+      "learning_rate": 1.3602318840579711e-05,
+      "loss": 1.6957,
+      "step": 11540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4071532487869263,
+      "learning_rate": 1.3596521739130435e-05,
+      "loss": 1.6937,
+      "step": 11550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.890350103378296,
+      "learning_rate": 1.359072463768116e-05,
+      "loss": 1.6852,
+      "step": 11560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.77126932144165,
+      "learning_rate": 1.3584927536231885e-05,
+      "loss": 1.6281,
+      "step": 11570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.499045729637146,
+      "learning_rate": 1.357913043478261e-05,
+      "loss": 1.6476,
+      "step": 11580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.360612154006958,
+      "learning_rate": 1.3573333333333334e-05,
+      "loss": 1.7523,
+      "step": 11590
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.634352922439575,
+      "learning_rate": 1.356753623188406e-05,
+      "loss": 1.7197,
+      "step": 11600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7271361351013184,
+      "learning_rate": 1.3561739130434784e-05,
+      "loss": 1.7322,
+      "step": 11610
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.323874473571777,
+      "learning_rate": 1.355594202898551e-05,
+      "loss": 1.6905,
+      "step": 11620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.2105393409729,
+      "learning_rate": 1.3550144927536232e-05,
+      "loss": 1.7726,
+      "step": 11630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.96962833404541,
+      "learning_rate": 1.3544347826086956e-05,
+      "loss": 1.6487,
+      "step": 11640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.641913652420044,
+      "learning_rate": 1.3538550724637682e-05,
+      "loss": 1.6343,
+      "step": 11650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.166054725646973,
+      "learning_rate": 1.3532753623188406e-05,
+      "loss": 1.6923,
+      "step": 11660
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.171774387359619,
+      "learning_rate": 1.3526956521739132e-05,
+      "loss": 1.8464,
+      "step": 11670
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8756415843963623,
+      "learning_rate": 1.3521159420289856e-05,
+      "loss": 1.6128,
+      "step": 11680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.395898818969727,
+      "learning_rate": 1.3515362318840581e-05,
+      "loss": 1.7561,
+      "step": 11690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.581686019897461,
+      "learning_rate": 1.3509565217391305e-05,
+      "loss": 1.6437,
+      "step": 11700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5616443157196045,
+      "learning_rate": 1.3503768115942031e-05,
+      "loss": 1.7065,
+      "step": 11710
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.503169536590576,
+      "learning_rate": 1.3497971014492755e-05,
+      "loss": 1.7289,
+      "step": 11720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8909196853637695,
+      "learning_rate": 1.349217391304348e-05,
+      "loss": 1.708,
+      "step": 11730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.47155499458313,
+      "learning_rate": 1.3486376811594205e-05,
+      "loss": 1.6653,
+      "step": 11740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7819576263427734,
+      "learning_rate": 1.3480579710144927e-05,
+      "loss": 1.765,
+      "step": 11750
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.265803337097168,
+      "learning_rate": 1.3474782608695653e-05,
+      "loss": 1.8526,
+      "step": 11760
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.643122673034668,
+      "learning_rate": 1.3468985507246377e-05,
+      "loss": 1.7633,
+      "step": 11770
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.063491821289062,
+      "learning_rate": 1.3463188405797102e-05,
+      "loss": 1.7809,
+      "step": 11780
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.299911618232727,
+      "learning_rate": 1.3457391304347826e-05,
+      "loss": 1.7441,
+      "step": 11790
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.312831401824951,
+      "learning_rate": 1.3451594202898552e-05,
+      "loss": 1.7987,
+      "step": 11800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.44124698638916,
+      "learning_rate": 1.3445797101449276e-05,
+      "loss": 1.6447,
+      "step": 11810
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1065027713775635,
+      "learning_rate": 1.3440000000000002e-05,
+      "loss": 1.8101,
+      "step": 11820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.687283039093018,
+      "learning_rate": 1.3434202898550726e-05,
+      "loss": 1.7435,
+      "step": 11830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.883983135223389,
+      "learning_rate": 1.3428405797101451e-05,
+      "loss": 1.6305,
+      "step": 11840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.454877257347107,
+      "learning_rate": 1.3422608695652175e-05,
+      "loss": 1.7328,
+      "step": 11850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.217712879180908,
+      "learning_rate": 1.34168115942029e-05,
+      "loss": 1.8219,
+      "step": 11860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0228896141052246,
+      "learning_rate": 1.3411014492753625e-05,
+      "loss": 1.6406,
+      "step": 11870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.98235285282135,
+      "learning_rate": 1.3405217391304347e-05,
+      "loss": 1.814,
+      "step": 11880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.972719192504883,
+      "learning_rate": 1.3399420289855073e-05,
+      "loss": 1.7487,
+      "step": 11890
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.730912446975708,
+      "learning_rate": 1.3393623188405797e-05,
+      "loss": 1.6201,
+      "step": 11900
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.768197774887085,
+      "learning_rate": 1.3387826086956523e-05,
+      "loss": 1.5983,
+      "step": 11910
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7705687284469604,
+      "learning_rate": 1.3382028985507247e-05,
+      "loss": 1.8149,
+      "step": 11920
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.303154230117798,
+      "learning_rate": 1.3376231884057973e-05,
+      "loss": 1.5987,
+      "step": 11930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.428690195083618,
+      "learning_rate": 1.3370434782608697e-05,
+      "loss": 1.7096,
+      "step": 11940
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.814437985420227,
+      "learning_rate": 1.336463768115942e-05,
+      "loss": 1.5782,
+      "step": 11950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.096653938293457,
+      "learning_rate": 1.3358840579710146e-05,
+      "loss": 1.7079,
+      "step": 11960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4044532775878906,
+      "learning_rate": 1.335304347826087e-05,
+      "loss": 1.5858,
+      "step": 11970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.206218719482422,
+      "learning_rate": 1.3347246376811596e-05,
+      "loss": 1.7468,
+      "step": 11980
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.609010219573975,
+      "learning_rate": 1.334144927536232e-05,
+      "loss": 1.7859,
+      "step": 11990
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.646367311477661,
+      "learning_rate": 1.3335652173913046e-05,
+      "loss": 1.4624,
+      "step": 12000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.7011674642562866,
+      "eval_runtime": 107.5473,
+      "eval_samples_per_second": 9.298,
+      "eval_steps_per_second": 2.325,
+      "step": 12000
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.173443794250488,
+      "learning_rate": 1.3329855072463768e-05,
+      "loss": 1.6229,
+      "step": 12010
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7814215421676636,
+      "learning_rate": 1.3324057971014495e-05,
+      "loss": 1.4929,
+      "step": 12020
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4045521020889282,
+      "learning_rate": 1.3318260869565218e-05,
+      "loss": 1.6919,
+      "step": 12030
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0634101629257202,
+      "learning_rate": 1.3312463768115943e-05,
+      "loss": 1.6938,
+      "step": 12040
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.312534213066101,
+      "learning_rate": 1.3306666666666667e-05,
+      "loss": 1.718,
+      "step": 12050
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.818319320678711,
+      "learning_rate": 1.3300869565217391e-05,
+      "loss": 1.7151,
+      "step": 12060
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.955862045288086,
+      "learning_rate": 1.3295072463768117e-05,
+      "loss": 1.7341,
+      "step": 12070
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.5620914697647095,
+      "learning_rate": 1.3289275362318841e-05,
+      "loss": 1.8083,
+      "step": 12080
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.745596170425415,
+      "learning_rate": 1.3283478260869567e-05,
+      "loss": 1.6492,
+      "step": 12090
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.607650279998779,
+      "learning_rate": 1.3277681159420291e-05,
+      "loss": 1.6268,
+      "step": 12100
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8814033269882202,
+      "learning_rate": 1.3271884057971017e-05,
+      "loss": 1.6127,
+      "step": 12110
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.7189909219741821,
+      "learning_rate": 1.326608695652174e-05,
+      "loss": 1.7186,
+      "step": 12120
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.2475483417510986,
+      "learning_rate": 1.3260289855072466e-05,
+      "loss": 1.7642,
+      "step": 12130
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.229576587677002,
+      "learning_rate": 1.3254492753623189e-05,
+      "loss": 1.7376,
+      "step": 12140
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.324844837188721,
+      "learning_rate": 1.3248695652173916e-05,
+      "loss": 1.753,
+      "step": 12150
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.727643370628357,
+      "learning_rate": 1.3242898550724638e-05,
+      "loss": 1.7612,
+      "step": 12160
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.785902976989746,
+      "learning_rate": 1.3237101449275362e-05,
+      "loss": 1.8341,
+      "step": 12170
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.65829598903656,
+      "learning_rate": 1.3231304347826088e-05,
+      "loss": 1.6544,
+      "step": 12180
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.451395511627197,
+      "learning_rate": 1.3225507246376812e-05,
+      "loss": 1.7604,
+      "step": 12190
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 11.722532272338867,
+      "learning_rate": 1.3219710144927538e-05,
+      "loss": 1.6377,
+      "step": 12200
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.6431243419647217,
+      "learning_rate": 1.3213913043478262e-05,
+      "loss": 1.6285,
+      "step": 12210
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.2399184703826904,
+      "learning_rate": 1.3208115942028987e-05,
+      "loss": 1.6122,
+      "step": 12220
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8696668148040771,
+      "learning_rate": 1.3202318840579711e-05,
+      "loss": 1.7179,
+      "step": 12230
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.199878692626953,
+      "learning_rate": 1.3196521739130437e-05,
+      "loss": 1.6551,
+      "step": 12240
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.326812744140625,
+      "learning_rate": 1.3190724637681161e-05,
+      "loss": 1.6373,
+      "step": 12250
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3237980604171753,
+      "learning_rate": 1.3184927536231883e-05,
+      "loss": 1.6698,
+      "step": 12260
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.768131971359253,
+      "learning_rate": 1.3179130434782609e-05,
+      "loss": 1.7116,
+      "step": 12270
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.411503791809082,
+      "learning_rate": 1.3173333333333333e-05,
+      "loss": 1.6156,
+      "step": 12280
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.180882930755615,
+      "learning_rate": 1.3167536231884059e-05,
+      "loss": 1.6427,
+      "step": 12290
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.48996639251709,
+      "learning_rate": 1.3161739130434783e-05,
+      "loss": 1.8268,
+      "step": 12300
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.8137004375457764,
+      "learning_rate": 1.3155942028985509e-05,
+      "loss": 1.6113,
+      "step": 12310
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.2274035215377808,
+      "learning_rate": 1.3150144927536233e-05,
+      "loss": 1.5674,
+      "step": 12320
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9615777730941772,
+      "learning_rate": 1.3144347826086958e-05,
+      "loss": 1.7014,
+      "step": 12330
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.100159168243408,
+      "learning_rate": 1.3138550724637682e-05,
+      "loss": 1.6939,
+      "step": 12340
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.2069636583328247,
+      "learning_rate": 1.3132753623188408e-05,
+      "loss": 1.8285,
+      "step": 12350
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.2240705490112305,
+      "learning_rate": 1.3126956521739132e-05,
+      "loss": 1.7967,
+      "step": 12360
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.67422342300415,
+      "learning_rate": 1.3121159420289856e-05,
+      "loss": 1.6853,
+      "step": 12370
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0639190673828125,
+      "learning_rate": 1.3115362318840582e-05,
+      "loss": 1.8051,
+      "step": 12380
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.48922598361969,
+      "learning_rate": 1.3109565217391304e-05,
+      "loss": 1.744,
+      "step": 12390
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7941759824752808,
+      "learning_rate": 1.310376811594203e-05,
+      "loss": 1.6758,
+      "step": 12400
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.099454402923584,
+      "learning_rate": 1.3097971014492754e-05,
+      "loss": 1.6345,
+      "step": 12410
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9024593830108643,
+      "learning_rate": 1.309217391304348e-05,
+      "loss": 1.5965,
+      "step": 12420
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.2126779556274414,
+      "learning_rate": 1.3086376811594203e-05,
+      "loss": 1.6138,
+      "step": 12430
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.7725675106048584,
+      "learning_rate": 1.3080579710144929e-05,
+      "loss": 1.7117,
+      "step": 12440
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.924130439758301,
+      "learning_rate": 1.3074782608695653e-05,
+      "loss": 1.7234,
+      "step": 12450
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2998852729797363,
+      "learning_rate": 1.3068985507246379e-05,
+      "loss": 1.6755,
+      "step": 12460
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.43101167678833,
+      "learning_rate": 1.3063188405797103e-05,
+      "loss": 1.6946,
+      "step": 12470
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.988603115081787,
+      "learning_rate": 1.3057391304347827e-05,
+      "loss": 1.6585,
+      "step": 12480
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.184678316116333,
+      "learning_rate": 1.3051594202898552e-05,
+      "loss": 1.6695,
+      "step": 12490
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.5645880699157715,
+      "learning_rate": 1.3045797101449277e-05,
+      "loss": 1.7914,
+      "step": 12500
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 1.7145652770996094,
+      "eval_runtime": 107.5024,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 12500
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6192374229431152,
+      "learning_rate": 1.3040000000000002e-05,
+      "loss": 1.7134,
+      "step": 12510
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7455005645751953,
+      "learning_rate": 1.3034202898550725e-05,
+      "loss": 1.7623,
+      "step": 12520
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.3086254596710205,
+      "learning_rate": 1.302840579710145e-05,
+      "loss": 1.6877,
+      "step": 12530
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5355310440063477,
+      "learning_rate": 1.3022608695652174e-05,
+      "loss": 1.7792,
+      "step": 12540
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7353932857513428,
+      "learning_rate": 1.30168115942029e-05,
+      "loss": 1.7148,
+      "step": 12550
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.468092918395996,
+      "learning_rate": 1.3011014492753624e-05,
+      "loss": 1.7166,
+      "step": 12560
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.603166580200195,
+      "learning_rate": 1.3005217391304348e-05,
+      "loss": 1.7689,
+      "step": 12570
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.521178960800171,
+      "learning_rate": 1.2999420289855074e-05,
+      "loss": 1.6698,
+      "step": 12580
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.06740665435791,
+      "learning_rate": 1.2993623188405798e-05,
+      "loss": 1.6722,
+      "step": 12590
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.7508833408355713,
+      "learning_rate": 1.2987826086956523e-05,
+      "loss": 1.6326,
+      "step": 12600
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.8367888927459717,
+      "learning_rate": 1.2982028985507247e-05,
+      "loss": 1.7824,
+      "step": 12610
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0329577922821045,
+      "learning_rate": 1.2976231884057973e-05,
+      "loss": 1.6271,
+      "step": 12620
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7697292566299438,
+      "learning_rate": 1.2970434782608697e-05,
+      "loss": 1.5485,
+      "step": 12630
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.019354820251465,
+      "learning_rate": 1.2964637681159423e-05,
+      "loss": 1.764,
+      "step": 12640
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.5531795024871826,
+      "learning_rate": 1.2958840579710145e-05,
+      "loss": 1.5918,
+      "step": 12650
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.986388206481934,
+      "learning_rate": 1.295304347826087e-05,
+      "loss": 1.7121,
+      "step": 12660
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.05776047706604,
+      "learning_rate": 1.2947246376811595e-05,
+      "loss": 1.8649,
+      "step": 12670
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 8.3735990524292,
+      "learning_rate": 1.2941449275362319e-05,
+      "loss": 1.7337,
+      "step": 12680
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.643716812133789,
+      "learning_rate": 1.2935652173913044e-05,
+      "loss": 1.8637,
+      "step": 12690
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.21591854095459,
+      "learning_rate": 1.2929855072463768e-05,
+      "loss": 1.6824,
+      "step": 12700
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.3029552698135376,
+      "learning_rate": 1.2924057971014494e-05,
+      "loss": 1.7475,
+      "step": 12710
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8332914113998413,
+      "learning_rate": 1.2918260869565218e-05,
+      "loss": 1.6892,
+      "step": 12720
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.6949455738067627,
+      "learning_rate": 1.2912463768115944e-05,
+      "loss": 1.6439,
+      "step": 12730
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.436410427093506,
+      "learning_rate": 1.2906666666666668e-05,
+      "loss": 1.7154,
+      "step": 12740
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0431625843048096,
+      "learning_rate": 1.2900869565217394e-05,
+      "loss": 1.6907,
+      "step": 12750
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.340989112854004,
+      "learning_rate": 1.2895072463768118e-05,
+      "loss": 1.6155,
+      "step": 12760
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4236230850219727,
+      "learning_rate": 1.2889275362318843e-05,
+      "loss": 1.7229,
+      "step": 12770
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.5978729724884033,
+      "learning_rate": 1.2883478260869566e-05,
+      "loss": 1.7474,
+      "step": 12780
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.729132056236267,
+      "learning_rate": 1.287768115942029e-05,
+      "loss": 1.6459,
+      "step": 12790
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.0741076469421387,
+      "learning_rate": 1.2871884057971015e-05,
+      "loss": 1.7362,
+      "step": 12800
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.056138038635254,
+      "learning_rate": 1.286608695652174e-05,
+      "loss": 1.6319,
+      "step": 12810
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.7434098720550537,
+      "learning_rate": 1.2860289855072465e-05,
+      "loss": 1.6161,
+      "step": 12820
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8511974811553955,
+      "learning_rate": 1.2854492753623189e-05,
+      "loss": 1.5639,
+      "step": 12830
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.42405891418457,
+      "learning_rate": 1.2848695652173915e-05,
+      "loss": 1.7232,
+      "step": 12840
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.329336166381836,
+      "learning_rate": 1.2842898550724639e-05,
+      "loss": 1.4743,
+      "step": 12850
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.943577527999878,
+      "learning_rate": 1.2837101449275364e-05,
+      "loss": 1.724,
+      "step": 12860
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.932284116744995,
+      "learning_rate": 1.2831304347826088e-05,
+      "loss": 1.7264,
+      "step": 12870
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.513105392456055,
+      "learning_rate": 1.282550724637681e-05,
+      "loss": 1.7304,
+      "step": 12880
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8422374129295349,
+      "learning_rate": 1.2819710144927538e-05,
+      "loss": 1.7715,
+      "step": 12890
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.7153878211975098,
+      "learning_rate": 1.281391304347826e-05,
+      "loss": 1.6772,
+      "step": 12900
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.201610088348389,
+      "learning_rate": 1.2808115942028986e-05,
+      "loss": 1.6741,
+      "step": 12910
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2562031745910645,
+      "learning_rate": 1.280231884057971e-05,
+      "loss": 1.7259,
+      "step": 12920
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.5883278846740723,
+      "learning_rate": 1.2797101449275365e-05,
+      "loss": 1.6095,
+      "step": 12930
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.405609130859375,
+      "learning_rate": 1.2791304347826087e-05,
+      "loss": 1.6715,
+      "step": 12940
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.966475486755371,
+      "learning_rate": 1.2785507246376815e-05,
+      "loss": 1.7278,
+      "step": 12950
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.7825913429260254,
+      "learning_rate": 1.2779710144927537e-05,
+      "loss": 1.7728,
+      "step": 12960
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6676304340362549,
+      "learning_rate": 1.277391304347826e-05,
+      "loss": 1.604,
+      "step": 12970
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 8.661015510559082,
+      "learning_rate": 1.2768115942028987e-05,
+      "loss": 1.6988,
+      "step": 12980
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4278024435043335,
+      "learning_rate": 1.276231884057971e-05,
+      "loss": 1.6335,
+      "step": 12990
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.0306954383850098,
+      "learning_rate": 1.2756521739130436e-05,
+      "loss": 1.7252,
+      "step": 13000
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 1.6973094940185547,
+      "eval_runtime": 107.5084,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 13000
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.311023712158203,
+      "learning_rate": 1.275072463768116e-05,
+      "loss": 1.6827,
+      "step": 13010
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.270779132843018,
+      "learning_rate": 1.2744927536231886e-05,
+      "loss": 1.6404,
+      "step": 13020
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.767482876777649,
+      "learning_rate": 1.273913043478261e-05,
+      "loss": 1.6437,
+      "step": 13030
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.334309101104736,
+      "learning_rate": 1.2733333333333336e-05,
+      "loss": 1.8447,
+      "step": 13040
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.177881240844727,
+      "learning_rate": 1.272753623188406e-05,
+      "loss": 1.7184,
+      "step": 13050
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.994053304195404,
+      "learning_rate": 1.2721739130434782e-05,
+      "loss": 1.5713,
+      "step": 13060
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.124035120010376,
+      "learning_rate": 1.2715942028985508e-05,
+      "loss": 1.7455,
+      "step": 13070
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.642811298370361,
+      "learning_rate": 1.2710144927536232e-05,
+      "loss": 1.6266,
+      "step": 13080
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.6701507568359375,
+      "learning_rate": 1.2704347826086957e-05,
+      "loss": 1.5264,
+      "step": 13090
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.8725638389587402,
+      "learning_rate": 1.2698550724637681e-05,
+      "loss": 1.6485,
+      "step": 13100
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.132758140563965,
+      "learning_rate": 1.2692753623188407e-05,
+      "loss": 1.7384,
+      "step": 13110
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.472156047821045,
+      "learning_rate": 1.2686956521739131e-05,
+      "loss": 1.7509,
+      "step": 13120
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.3745663166046143,
+      "learning_rate": 1.2681159420289857e-05,
+      "loss": 1.8063,
+      "step": 13130
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.400909423828125,
+      "learning_rate": 1.267536231884058e-05,
+      "loss": 1.502,
+      "step": 13140
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.341827154159546,
+      "learning_rate": 1.2669565217391306e-05,
+      "loss": 1.8114,
+      "step": 13150
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.864140033721924,
+      "learning_rate": 1.266376811594203e-05,
+      "loss": 1.6826,
+      "step": 13160
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.464040756225586,
+      "learning_rate": 1.2657971014492755e-05,
+      "loss": 1.8461,
+      "step": 13170
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.3931642770767212,
+      "learning_rate": 1.265217391304348e-05,
+      "loss": 1.5932,
+      "step": 13180
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9136666059494019,
+      "learning_rate": 1.2646376811594203e-05,
+      "loss": 1.655,
+      "step": 13190
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6585626006126404,
+      "learning_rate": 1.2640579710144928e-05,
+      "loss": 1.6953,
+      "step": 13200
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.4719610214233398,
+      "learning_rate": 1.2634782608695652e-05,
+      "loss": 1.7122,
+      "step": 13210
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.93149995803833,
+      "learning_rate": 1.2628985507246378e-05,
+      "loss": 1.7336,
+      "step": 13220
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.2544167041778564,
+      "learning_rate": 1.2623188405797102e-05,
+      "loss": 1.5404,
+      "step": 13230
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.97430682182312,
+      "learning_rate": 1.2617391304347828e-05,
+      "loss": 1.7408,
+      "step": 13240
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8780460357666016,
+      "learning_rate": 1.2611594202898552e-05,
+      "loss": 1.5946,
+      "step": 13250
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.108811140060425,
+      "learning_rate": 1.2605797101449277e-05,
+      "loss": 1.7989,
+      "step": 13260
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.7901716232299805,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 1.715,
+      "step": 13270
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.103832244873047,
+      "learning_rate": 1.2594202898550725e-05,
+      "loss": 1.5802,
+      "step": 13280
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.0903170108795166,
+      "learning_rate": 1.2588405797101451e-05,
+      "loss": 1.7531,
+      "step": 13290
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.20255184173584,
+      "learning_rate": 1.2582608695652175e-05,
+      "loss": 1.7651,
+      "step": 13300
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.1130000352859497,
+      "learning_rate": 1.25768115942029e-05,
+      "loss": 1.6944,
+      "step": 13310
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.6338019371032715,
+      "learning_rate": 1.2571014492753623e-05,
+      "loss": 1.6242,
+      "step": 13320
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.7077834606170654,
+      "learning_rate": 1.2565217391304349e-05,
+      "loss": 1.5732,
+      "step": 13330
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3346569538116455,
+      "learning_rate": 1.2559420289855073e-05,
+      "loss": 1.8445,
+      "step": 13340
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 9.315794944763184,
+      "learning_rate": 1.2553623188405798e-05,
+      "loss": 1.6678,
+      "step": 13350
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.115358591079712,
+      "learning_rate": 1.2547826086956522e-05,
+      "loss": 1.719,
+      "step": 13360
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.239322662353516,
+      "learning_rate": 1.2542028985507246e-05,
+      "loss": 1.6076,
+      "step": 13370
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.326850891113281,
+      "learning_rate": 1.2536231884057972e-05,
+      "loss": 1.722,
+      "step": 13380
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.592775344848633,
+      "learning_rate": 1.2530434782608696e-05,
+      "loss": 1.7467,
+      "step": 13390
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 7.9691667556762695,
+      "learning_rate": 1.2524637681159422e-05,
+      "loss": 1.5619,
+      "step": 13400
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5476248264312744,
+      "learning_rate": 1.2518840579710146e-05,
+      "loss": 1.7403,
+      "step": 13410
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.466348886489868,
+      "learning_rate": 1.2513043478260872e-05,
+      "loss": 1.7568,
+      "step": 13420
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7985725402832031,
+      "learning_rate": 1.2507246376811596e-05,
+      "loss": 1.6505,
+      "step": 13430
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.469550848007202,
+      "learning_rate": 1.2501449275362321e-05,
+      "loss": 1.686,
+      "step": 13440
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.672849178314209,
+      "learning_rate": 1.2495652173913044e-05,
+      "loss": 1.6282,
+      "step": 13450
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2976021766662598,
+      "learning_rate": 1.2489855072463771e-05,
+      "loss": 1.7232,
+      "step": 13460
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.909353256225586,
+      "learning_rate": 1.2484057971014493e-05,
+      "loss": 1.5368,
+      "step": 13470
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.343336582183838,
+      "learning_rate": 1.2478260869565217e-05,
+      "loss": 1.8103,
+      "step": 13480
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.372796058654785,
+      "learning_rate": 1.2472463768115943e-05,
+      "loss": 1.6098,
+      "step": 13490
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7971751689910889,
+      "learning_rate": 1.2466666666666667e-05,
+      "loss": 1.6064,
+      "step": 13500
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 1.6684225797653198,
+      "eval_runtime": 107.5063,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 13500
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.6352314949035645,
+      "learning_rate": 1.2460869565217393e-05,
+      "loss": 1.7123,
+      "step": 13510
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.39823317527771,
+      "learning_rate": 1.2455072463768117e-05,
+      "loss": 1.7258,
+      "step": 13520
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8198342323303223,
+      "learning_rate": 1.2449275362318842e-05,
+      "loss": 1.6146,
+      "step": 13530
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.5022692680358887,
+      "learning_rate": 1.2443478260869566e-05,
+      "loss": 1.6835,
+      "step": 13540
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.4644007682800293,
+      "learning_rate": 1.2437681159420292e-05,
+      "loss": 1.7186,
+      "step": 13550
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.09822416305542,
+      "learning_rate": 1.2431884057971016e-05,
+      "loss": 1.7713,
+      "step": 13560
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.5740549564361572,
+      "learning_rate": 1.2426086956521742e-05,
+      "loss": 1.7463,
+      "step": 13570
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3453164100646973,
+      "learning_rate": 1.2420289855072464e-05,
+      "loss": 1.6706,
+      "step": 13580
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.465335369110107,
+      "learning_rate": 1.2414492753623188e-05,
+      "loss": 1.5353,
+      "step": 13590
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.2128164768218994,
+      "learning_rate": 1.2408695652173914e-05,
+      "loss": 1.7146,
+      "step": 13600
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.6610260009765625,
+      "learning_rate": 1.2402898550724638e-05,
+      "loss": 1.6992,
+      "step": 13610
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.8867626190185547,
+      "learning_rate": 1.2397101449275364e-05,
+      "loss": 1.7159,
+      "step": 13620
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6658381223678589,
+      "learning_rate": 1.2391304347826088e-05,
+      "loss": 1.6678,
+      "step": 13630
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.239511728286743,
+      "learning_rate": 1.2385507246376813e-05,
+      "loss": 1.6895,
+      "step": 13640
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.459529399871826,
+      "learning_rate": 1.2379710144927537e-05,
+      "loss": 1.7424,
+      "step": 13650
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.84796142578125,
+      "learning_rate": 1.2373913043478263e-05,
+      "loss": 1.6314,
+      "step": 13660
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.1975581645965576,
+      "learning_rate": 1.2368115942028987e-05,
+      "loss": 1.7106,
+      "step": 13670
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.918569564819336,
+      "learning_rate": 1.236231884057971e-05,
+      "loss": 1.6561,
+      "step": 13680
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.9456100463867188,
+      "learning_rate": 1.2356521739130437e-05,
+      "loss": 1.5517,
+      "step": 13690
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2089602947235107,
+      "learning_rate": 1.2350724637681159e-05,
+      "loss": 1.6579,
+      "step": 13700
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.878429889678955,
+      "learning_rate": 1.2344927536231885e-05,
+      "loss": 1.6001,
+      "step": 13710
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9600847959518433,
+      "learning_rate": 1.2339130434782609e-05,
+      "loss": 1.6324,
+      "step": 13720
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.215534687042236,
+      "learning_rate": 1.2333333333333334e-05,
+      "loss": 1.76,
+      "step": 13730
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.4697761535644531,
+      "learning_rate": 1.2327536231884058e-05,
+      "loss": 1.6336,
+      "step": 13740
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4715230464935303,
+      "learning_rate": 1.2321739130434784e-05,
+      "loss": 1.7028,
+      "step": 13750
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2741434574127197,
+      "learning_rate": 1.2315942028985508e-05,
+      "loss": 1.713,
+      "step": 13760
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9270362854003906,
+      "learning_rate": 1.2310144927536234e-05,
+      "loss": 1.6415,
+      "step": 13770
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.036669731140137,
+      "learning_rate": 1.2304347826086958e-05,
+      "loss": 1.7856,
+      "step": 13780
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.498189926147461,
+      "learning_rate": 1.2298550724637682e-05,
+      "loss": 1.575,
+      "step": 13790
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.0799221992492676,
+      "learning_rate": 1.2292753623188408e-05,
+      "loss": 1.4947,
+      "step": 13800
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5803940296173096,
+      "learning_rate": 1.228695652173913e-05,
+      "loss": 1.7288,
+      "step": 13810
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.215254306793213,
+      "learning_rate": 1.2281159420289857e-05,
+      "loss": 1.76,
+      "step": 13820
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.722815990447998,
+      "learning_rate": 1.227536231884058e-05,
+      "loss": 1.6812,
+      "step": 13830
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.2459473609924316,
+      "learning_rate": 1.2269565217391305e-05,
+      "loss": 1.6047,
+      "step": 13840
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8425486087799072,
+      "learning_rate": 1.226376811594203e-05,
+      "loss": 1.6592,
+      "step": 13850
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.949090003967285,
+      "learning_rate": 1.2257971014492755e-05,
+      "loss": 1.644,
+      "step": 13860
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5830769538879395,
+      "learning_rate": 1.2252173913043479e-05,
+      "loss": 1.7824,
+      "step": 13870
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8728455305099487,
+      "learning_rate": 1.2246376811594205e-05,
+      "loss": 1.7766,
+      "step": 13880
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3987927436828613,
+      "learning_rate": 1.2240579710144929e-05,
+      "loss": 1.6303,
+      "step": 13890
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.897881031036377,
+      "learning_rate": 1.2234782608695653e-05,
+      "loss": 1.6986,
+      "step": 13900
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8493584394454956,
+      "learning_rate": 1.2228985507246378e-05,
+      "loss": 1.6711,
+      "step": 13910
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.1658079624176025,
+      "learning_rate": 1.2223188405797102e-05,
+      "loss": 1.5714,
+      "step": 13920
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 7.026312828063965,
+      "learning_rate": 1.2217391304347828e-05,
+      "loss": 1.6648,
+      "step": 13930
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.64369010925293,
+      "learning_rate": 1.2211594202898552e-05,
+      "loss": 1.622,
+      "step": 13940
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.4812980890274048,
+      "learning_rate": 1.2205797101449278e-05,
+      "loss": 1.6563,
+      "step": 13950
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.3375606536865234,
+      "learning_rate": 1.22e-05,
+      "loss": 1.7177,
+      "step": 13960
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.0398337841033936,
+      "learning_rate": 1.2194202898550726e-05,
+      "loss": 1.8483,
+      "step": 13970
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.477541446685791,
+      "learning_rate": 1.218840579710145e-05,
+      "loss": 1.7636,
+      "step": 13980
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.2809600830078125,
+      "learning_rate": 1.2182608695652174e-05,
+      "loss": 1.6272,
+      "step": 13990
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.5135672092437744,
+      "learning_rate": 1.21768115942029e-05,
+      "loss": 1.6053,
+      "step": 14000
+    },
+    {
+      "epoch": 1.17,
+      "eval_loss": 1.655966877937317,
+      "eval_runtime": 107.5028,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 14000
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.65315580368042,
+      "learning_rate": 1.2171014492753624e-05,
+      "loss": 1.6857,
+      "step": 14010
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.5551207065582275,
+      "learning_rate": 1.216521739130435e-05,
+      "loss": 1.6957,
+      "step": 14020
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.662346839904785,
+      "learning_rate": 1.2159420289855073e-05,
+      "loss": 1.709,
+      "step": 14030
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.7407190799713135,
+      "learning_rate": 1.2153623188405799e-05,
+      "loss": 1.7064,
+      "step": 14040
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.3801932334899902,
+      "learning_rate": 1.2147826086956523e-05,
+      "loss": 1.5048,
+      "step": 14050
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.1840766668319702,
+      "learning_rate": 1.2142028985507249e-05,
+      "loss": 1.6463,
+      "step": 14060
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.610103130340576,
+      "learning_rate": 1.2136231884057973e-05,
+      "loss": 1.6947,
+      "step": 14070
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.217325210571289,
+      "learning_rate": 1.2130434782608698e-05,
+      "loss": 1.6183,
+      "step": 14080
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.316359281539917,
+      "learning_rate": 1.212463768115942e-05,
+      "loss": 1.6692,
+      "step": 14090
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.7778801918029785,
+      "learning_rate": 1.2118840579710145e-05,
+      "loss": 1.7233,
+      "step": 14100
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.980316162109375,
+      "learning_rate": 1.211304347826087e-05,
+      "loss": 1.5753,
+      "step": 14110
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.6318607330322266,
+      "learning_rate": 1.2107246376811594e-05,
+      "loss": 1.5216,
+      "step": 14120
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.447054386138916,
+      "learning_rate": 1.210144927536232e-05,
+      "loss": 1.82,
+      "step": 14130
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.368971586227417,
+      "learning_rate": 1.2095652173913044e-05,
+      "loss": 1.7372,
+      "step": 14140
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.2000670433044434,
+      "learning_rate": 1.208985507246377e-05,
+      "loss": 1.7553,
+      "step": 14150
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.1410953998565674,
+      "learning_rate": 1.2084057971014494e-05,
+      "loss": 1.778,
+      "step": 14160
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6340982913970947,
+      "learning_rate": 1.207826086956522e-05,
+      "loss": 1.5523,
+      "step": 14170
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.945784568786621,
+      "learning_rate": 1.2072463768115943e-05,
+      "loss": 1.6813,
+      "step": 14180
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7603634595870972,
+      "learning_rate": 1.206666666666667e-05,
+      "loss": 1.7787,
+      "step": 14190
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8819363117218018,
+      "learning_rate": 1.2060869565217393e-05,
+      "loss": 1.7003,
+      "step": 14200
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.882131576538086,
+      "learning_rate": 1.2055072463768115e-05,
+      "loss": 1.7552,
+      "step": 14210
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.339003324508667,
+      "learning_rate": 1.2049275362318841e-05,
+      "loss": 1.6867,
+      "step": 14220
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.375877857208252,
+      "learning_rate": 1.2043478260869565e-05,
+      "loss": 1.7049,
+      "step": 14230
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.6439826488494873,
+      "learning_rate": 1.2037681159420291e-05,
+      "loss": 1.7307,
+      "step": 14240
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.504507541656494,
+      "learning_rate": 1.2031884057971015e-05,
+      "loss": 1.7403,
+      "step": 14250
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.381837844848633,
+      "learning_rate": 1.202608695652174e-05,
+      "loss": 1.6733,
+      "step": 14260
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.466141700744629,
+      "learning_rate": 1.2020289855072465e-05,
+      "loss": 1.7054,
+      "step": 14270
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.095202922821045,
+      "learning_rate": 1.201449275362319e-05,
+      "loss": 1.6852,
+      "step": 14280
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.33658766746521,
+      "learning_rate": 1.2008695652173914e-05,
+      "loss": 1.7658,
+      "step": 14290
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.439746379852295,
+      "learning_rate": 1.2002898550724638e-05,
+      "loss": 1.5291,
+      "step": 14300
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.367286682128906,
+      "learning_rate": 1.1997101449275364e-05,
+      "loss": 1.6469,
+      "step": 14310
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.05848503112793,
+      "learning_rate": 1.1991304347826086e-05,
+      "loss": 1.8017,
+      "step": 14320
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.569128036499023,
+      "learning_rate": 1.1985507246376814e-05,
+      "loss": 1.5023,
+      "step": 14330
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.5229172706604004,
+      "learning_rate": 1.1979710144927536e-05,
+      "loss": 1.7475,
+      "step": 14340
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.319218635559082,
+      "learning_rate": 1.1973913043478262e-05,
+      "loss": 1.6949,
+      "step": 14350
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.897936820983887,
+      "learning_rate": 1.1968115942028986e-05,
+      "loss": 1.6718,
+      "step": 14360
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.435079574584961,
+      "learning_rate": 1.1962318840579711e-05,
+      "loss": 1.6979,
+      "step": 14370
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.011115312576294,
+      "learning_rate": 1.1956521739130435e-05,
+      "loss": 1.4714,
+      "step": 14380
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.5290638208389282,
+      "learning_rate": 1.1950724637681161e-05,
+      "loss": 1.5394,
+      "step": 14390
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.277037620544434,
+      "learning_rate": 1.1944927536231885e-05,
+      "loss": 1.6627,
+      "step": 14400
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.786652088165283,
+      "learning_rate": 1.1939130434782609e-05,
+      "loss": 1.7029,
+      "step": 14410
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.4963030815124512,
+      "learning_rate": 1.1933333333333335e-05,
+      "loss": 1.6999,
+      "step": 14420
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0223753452301025,
+      "learning_rate": 1.1927536231884059e-05,
+      "loss": 1.6998,
+      "step": 14430
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.4138219356536865,
+      "learning_rate": 1.1921739130434785e-05,
+      "loss": 1.5385,
+      "step": 14440
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.297712802886963,
+      "learning_rate": 1.1915942028985507e-05,
+      "loss": 1.6142,
+      "step": 14450
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.5567455291748047,
+      "learning_rate": 1.1910144927536234e-05,
+      "loss": 1.7254,
+      "step": 14460
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.329671382904053,
+      "learning_rate": 1.1904347826086957e-05,
+      "loss": 1.7098,
+      "step": 14470
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1015174388885498,
+      "learning_rate": 1.1898550724637682e-05,
+      "loss": 1.7312,
+      "step": 14480
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.632756233215332,
+      "learning_rate": 1.1892753623188406e-05,
+      "loss": 1.5845,
+      "step": 14490
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.536245822906494,
+      "learning_rate": 1.1886956521739132e-05,
+      "loss": 1.7143,
+      "step": 14500
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 1.686023235321045,
+      "eval_runtime": 107.504,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 14500
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.551739454269409,
+      "learning_rate": 1.1881159420289856e-05,
+      "loss": 1.7145,
+      "step": 14510
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.037818431854248,
+      "learning_rate": 1.187536231884058e-05,
+      "loss": 1.4981,
+      "step": 14520
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1799204349517822,
+      "learning_rate": 1.1869565217391306e-05,
+      "loss": 1.634,
+      "step": 14530
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.238110065460205,
+      "learning_rate": 1.186376811594203e-05,
+      "loss": 1.5684,
+      "step": 14540
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.419713020324707,
+      "learning_rate": 1.1857971014492755e-05,
+      "loss": 1.6688,
+      "step": 14550
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3431668281555176,
+      "learning_rate": 1.185217391304348e-05,
+      "loss": 1.7666,
+      "step": 14560
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.332759857177734,
+      "learning_rate": 1.1846376811594205e-05,
+      "loss": 1.7597,
+      "step": 14570
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.4301503896713257,
+      "learning_rate": 1.1840579710144927e-05,
+      "loss": 1.742,
+      "step": 14580
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.501166343688965,
+      "learning_rate": 1.1834782608695655e-05,
+      "loss": 1.6943,
+      "step": 14590
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.4921094179153442,
+      "learning_rate": 1.1828985507246377e-05,
+      "loss": 1.7038,
+      "step": 14600
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.763652801513672,
+      "learning_rate": 1.1823188405797101e-05,
+      "loss": 1.683,
+      "step": 14610
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.096171855926514,
+      "learning_rate": 1.1817391304347827e-05,
+      "loss": 1.5952,
+      "step": 14620
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.0550990104675293,
+      "learning_rate": 1.181159420289855e-05,
+      "loss": 1.6841,
+      "step": 14630
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.7404937744140625,
+      "learning_rate": 1.1805797101449277e-05,
+      "loss": 1.3851,
+      "step": 14640
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.006895065307617,
+      "learning_rate": 1.18e-05,
+      "loss": 1.6884,
+      "step": 14650
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.734574317932129,
+      "learning_rate": 1.1794202898550726e-05,
+      "loss": 1.6566,
+      "step": 14660
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.7309789657592773,
+      "learning_rate": 1.178840579710145e-05,
+      "loss": 1.7182,
+      "step": 14670
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.692793846130371,
+      "learning_rate": 1.1782608695652176e-05,
+      "loss": 1.6907,
+      "step": 14680
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2033607959747314,
+      "learning_rate": 1.17768115942029e-05,
+      "loss": 1.6187,
+      "step": 14690
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.37400221824646,
+      "learning_rate": 1.1771014492753626e-05,
+      "loss": 1.7607,
+      "step": 14700
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.4525043964385986,
+      "learning_rate": 1.176521739130435e-05,
+      "loss": 1.5481,
+      "step": 14710
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.075309753417969,
+      "learning_rate": 1.1759420289855072e-05,
+      "loss": 1.6437,
+      "step": 14720
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6363898515701294,
+      "learning_rate": 1.1753623188405798e-05,
+      "loss": 1.6288,
+      "step": 14730
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.210326671600342,
+      "learning_rate": 1.1747826086956522e-05,
+      "loss": 1.6164,
+      "step": 14740
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.808457851409912,
+      "learning_rate": 1.1742028985507247e-05,
+      "loss": 1.7345,
+      "step": 14750
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.7805867195129395,
+      "learning_rate": 1.1736231884057971e-05,
+      "loss": 1.666,
+      "step": 14760
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.179888129234314,
+      "learning_rate": 1.1730434782608697e-05,
+      "loss": 1.66,
+      "step": 14770
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.300703287124634,
+      "learning_rate": 1.1724637681159421e-05,
+      "loss": 1.5873,
+      "step": 14780
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.175435543060303,
+      "learning_rate": 1.1718840579710147e-05,
+      "loss": 1.7415,
+      "step": 14790
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.4557175636291504,
+      "learning_rate": 1.171304347826087e-05,
+      "loss": 1.7369,
+      "step": 14800
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.879883289337158,
+      "learning_rate": 1.1707246376811596e-05,
+      "loss": 1.6883,
+      "step": 14810
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.861069679260254,
+      "learning_rate": 1.170144927536232e-05,
+      "loss": 1.6678,
+      "step": 14820
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.97892427444458,
+      "learning_rate": 1.1695652173913043e-05,
+      "loss": 1.5186,
+      "step": 14830
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9701839685440063,
+      "learning_rate": 1.168985507246377e-05,
+      "loss": 1.6847,
+      "step": 14840
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1920979022979736,
+      "learning_rate": 1.1684057971014492e-05,
+      "loss": 1.6583,
+      "step": 14850
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.745516777038574,
+      "learning_rate": 1.1678260869565218e-05,
+      "loss": 1.6884,
+      "step": 14860
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.854526042938232,
+      "learning_rate": 1.1672463768115942e-05,
+      "loss": 1.7469,
+      "step": 14870
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.85132098197937,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 1.5597,
+      "step": 14880
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9505503177642822,
+      "learning_rate": 1.1660869565217392e-05,
+      "loss": 1.701,
+      "step": 14890
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.181471824645996,
+      "learning_rate": 1.1655072463768118e-05,
+      "loss": 1.7843,
+      "step": 14900
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.77048659324646,
+      "learning_rate": 1.1649275362318842e-05,
+      "loss": 1.7381,
+      "step": 14910
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.85846471786499,
+      "learning_rate": 1.1643478260869566e-05,
+      "loss": 1.7673,
+      "step": 14920
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 7.585514545440674,
+      "learning_rate": 1.1637681159420291e-05,
+      "loss": 1.6633,
+      "step": 14930
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.48277759552002,
+      "learning_rate": 1.1631884057971015e-05,
+      "loss": 1.6939,
+      "step": 14940
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.1543049812316895,
+      "learning_rate": 1.1626086956521741e-05,
+      "loss": 1.6358,
+      "step": 14950
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.4650087356567383,
+      "learning_rate": 1.1620289855072463e-05,
+      "loss": 1.7219,
+      "step": 14960
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.0995783805847168,
+      "learning_rate": 1.161449275362319e-05,
+      "loss": 1.68,
+      "step": 14970
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.351716995239258,
+      "learning_rate": 1.1608695652173913e-05,
+      "loss": 1.7076,
+      "step": 14980
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.2592127323150635,
+      "learning_rate": 1.1602898550724639e-05,
+      "loss": 1.6001,
+      "step": 14990
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4201385974884033,
+      "learning_rate": 1.1597101449275363e-05,
+      "loss": 1.753,
+      "step": 15000
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 1.7132278680801392,
+      "eval_runtime": 107.4961,
+      "eval_samples_per_second": 9.303,
+      "eval_steps_per_second": 2.326,
+      "step": 15000
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.306769609451294,
+      "learning_rate": 1.1591304347826088e-05,
+      "loss": 1.6939,
+      "step": 15010
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9689505100250244,
+      "learning_rate": 1.1585507246376812e-05,
+      "loss": 1.6635,
+      "step": 15020
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.207001209259033,
+      "learning_rate": 1.1579710144927536e-05,
+      "loss": 1.7515,
+      "step": 15030
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.0428285598754883,
+      "learning_rate": 1.1573913043478262e-05,
+      "loss": 1.5935,
+      "step": 15040
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.358133792877197,
+      "learning_rate": 1.1568115942028986e-05,
+      "loss": 1.6732,
+      "step": 15050
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.284139633178711,
+      "learning_rate": 1.1562318840579712e-05,
+      "loss": 1.7963,
+      "step": 15060
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.9040706157684326,
+      "learning_rate": 1.1556521739130436e-05,
+      "loss": 1.7937,
+      "step": 15070
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.678374290466309,
+      "learning_rate": 1.1550724637681162e-05,
+      "loss": 1.5987,
+      "step": 15080
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.1551108360290527,
+      "learning_rate": 1.1544927536231884e-05,
+      "loss": 1.7292,
+      "step": 15090
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.153188943862915,
+      "learning_rate": 1.1539130434782611e-05,
+      "loss": 1.5968,
+      "step": 15100
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.321458339691162,
+      "learning_rate": 1.1533333333333334e-05,
+      "loss": 1.7164,
+      "step": 15110
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9099916219711304,
+      "learning_rate": 1.152753623188406e-05,
+      "loss": 1.7368,
+      "step": 15120
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.101009368896484,
+      "learning_rate": 1.1521739130434783e-05,
+      "loss": 1.7376,
+      "step": 15130
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 10.590901374816895,
+      "learning_rate": 1.1515942028985507e-05,
+      "loss": 1.7985,
+      "step": 15140
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.799911022186279,
+      "learning_rate": 1.1510144927536233e-05,
+      "loss": 1.7259,
+      "step": 15150
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.911226511001587,
+      "learning_rate": 1.1504347826086957e-05,
+      "loss": 1.7439,
+      "step": 15160
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5930674076080322,
+      "learning_rate": 1.1498550724637683e-05,
+      "loss": 1.7223,
+      "step": 15170
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.125276565551758,
+      "learning_rate": 1.1492753623188407e-05,
+      "loss": 1.645,
+      "step": 15180
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.448198914527893,
+      "learning_rate": 1.1486956521739132e-05,
+      "loss": 1.6675,
+      "step": 15190
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.842217445373535,
+      "learning_rate": 1.1481159420289856e-05,
+      "loss": 1.761,
+      "step": 15200
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.6608208417892456,
+      "learning_rate": 1.1475362318840582e-05,
+      "loss": 1.6394,
+      "step": 15210
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.386859178543091,
+      "learning_rate": 1.1469565217391304e-05,
+      "loss": 1.6165,
+      "step": 15220
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5258991718292236,
+      "learning_rate": 1.1463768115942028e-05,
+      "loss": 1.7252,
+      "step": 15230
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.106661081314087,
+      "learning_rate": 1.1457971014492754e-05,
+      "loss": 1.6343,
+      "step": 15240
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.260819435119629,
+      "learning_rate": 1.1452173913043478e-05,
+      "loss": 1.7971,
+      "step": 15250
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.5960564613342285,
+      "learning_rate": 1.1446376811594204e-05,
+      "loss": 1.6154,
+      "step": 15260
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.4843623638153076,
+      "learning_rate": 1.1440579710144928e-05,
+      "loss": 1.8285,
+      "step": 15270
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.6920154094696045,
+      "learning_rate": 1.1434782608695654e-05,
+      "loss": 1.7324,
+      "step": 15280
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.802539110183716,
+      "learning_rate": 1.1428985507246378e-05,
+      "loss": 1.6399,
+      "step": 15290
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.4101133346557617,
+      "learning_rate": 1.1423188405797103e-05,
+      "loss": 1.7186,
+      "step": 15300
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7906736135482788,
+      "learning_rate": 1.1417391304347827e-05,
+      "loss": 1.7397,
+      "step": 15310
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0107662677764893,
+      "learning_rate": 1.1411594202898553e-05,
+      "loss": 1.6297,
+      "step": 15320
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.830557107925415,
+      "learning_rate": 1.1405797101449277e-05,
+      "loss": 1.6362,
+      "step": 15330
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7907980680465698,
+      "learning_rate": 1.14e-05,
+      "loss": 1.6913,
+      "step": 15340
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5224626064300537,
+      "learning_rate": 1.1394202898550725e-05,
+      "loss": 1.7274,
+      "step": 15350
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 7.044021129608154,
+      "learning_rate": 1.1388405797101449e-05,
+      "loss": 1.6653,
+      "step": 15360
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0197882652282715,
+      "learning_rate": 1.1382608695652175e-05,
+      "loss": 1.7462,
+      "step": 15370
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.1112966537475586,
+      "learning_rate": 1.1376811594202899e-05,
+      "loss": 1.7024,
+      "step": 15380
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.9264869689941406,
+      "learning_rate": 1.1371014492753624e-05,
+      "loss": 1.6285,
+      "step": 15390
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.849034309387207,
+      "learning_rate": 1.1365217391304348e-05,
+      "loss": 1.7542,
+      "step": 15400
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.498842239379883,
+      "learning_rate": 1.1359420289855074e-05,
+      "loss": 1.6733,
+      "step": 15410
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.047708034515381,
+      "learning_rate": 1.1353623188405798e-05,
+      "loss": 1.6027,
+      "step": 15420
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.122775077819824,
+      "learning_rate": 1.1347826086956524e-05,
+      "loss": 1.7346,
+      "step": 15430
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.988992929458618,
+      "learning_rate": 1.1342028985507248e-05,
+      "loss": 1.5555,
+      "step": 15440
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.4278367757797241,
+      "learning_rate": 1.1336231884057972e-05,
+      "loss": 1.6886,
+      "step": 15450
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.543633222579956,
+      "learning_rate": 1.1330434782608698e-05,
+      "loss": 1.5712,
+      "step": 15460
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.563520908355713,
+      "learning_rate": 1.132463768115942e-05,
+      "loss": 1.6082,
+      "step": 15470
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.311037540435791,
+      "learning_rate": 1.1318840579710147e-05,
+      "loss": 1.56,
+      "step": 15480
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.979817271232605,
+      "learning_rate": 1.131304347826087e-05,
+      "loss": 1.7833,
+      "step": 15490
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.7615084648132324,
+      "learning_rate": 1.1307246376811595e-05,
+      "loss": 1.6641,
+      "step": 15500
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 1.6843485832214355,
+      "eval_runtime": 107.5057,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 15500
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.724392890930176,
+      "learning_rate": 1.130144927536232e-05,
+      "loss": 1.5888,
+      "step": 15510
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.149465560913086,
+      "learning_rate": 1.1295652173913045e-05,
+      "loss": 1.7467,
+      "step": 15520
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9117258787155151,
+      "learning_rate": 1.1289855072463769e-05,
+      "loss": 1.6763,
+      "step": 15530
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0557730197906494,
+      "learning_rate": 1.1284057971014493e-05,
+      "loss": 1.7202,
+      "step": 15540
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.089738130569458,
+      "learning_rate": 1.1278260869565219e-05,
+      "loss": 1.7192,
+      "step": 15550
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.925050735473633,
+      "learning_rate": 1.1272463768115943e-05,
+      "loss": 1.7204,
+      "step": 15560
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 7.140872001647949,
+      "learning_rate": 1.1266666666666668e-05,
+      "loss": 1.5706,
+      "step": 15570
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.26693058013916,
+      "learning_rate": 1.1260869565217392e-05,
+      "loss": 1.796,
+      "step": 15580
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.269399881362915,
+      "learning_rate": 1.1255072463768118e-05,
+      "loss": 1.5746,
+      "step": 15590
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.2779074907302856,
+      "learning_rate": 1.124927536231884e-05,
+      "loss": 1.5743,
+      "step": 15600
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.808041572570801,
+      "learning_rate": 1.1243478260869568e-05,
+      "loss": 1.7358,
+      "step": 15610
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.192412853240967,
+      "learning_rate": 1.123768115942029e-05,
+      "loss": 1.7665,
+      "step": 15620
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8821000456809998,
+      "learning_rate": 1.1231884057971016e-05,
+      "loss": 1.6442,
+      "step": 15630
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.8440635204315186,
+      "learning_rate": 1.122608695652174e-05,
+      "loss": 1.6318,
+      "step": 15640
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.6958911418914795,
+      "learning_rate": 1.1220289855072464e-05,
+      "loss": 1.7744,
+      "step": 15650
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 12.584587097167969,
+      "learning_rate": 1.121449275362319e-05,
+      "loss": 1.7242,
+      "step": 15660
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.3364410400390625,
+      "learning_rate": 1.1208695652173913e-05,
+      "loss": 1.7048,
+      "step": 15670
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.9475982189178467,
+      "learning_rate": 1.120289855072464e-05,
+      "loss": 1.6612,
+      "step": 15680
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.1554036140441895,
+      "learning_rate": 1.1197101449275363e-05,
+      "loss": 1.6869,
+      "step": 15690
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2033803462982178,
+      "learning_rate": 1.1191304347826089e-05,
+      "loss": 1.6598,
+      "step": 15700
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.7358901500701904,
+      "learning_rate": 1.1185507246376813e-05,
+      "loss": 1.7445,
+      "step": 15710
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.531160593032837,
+      "learning_rate": 1.1179710144927539e-05,
+      "loss": 1.6378,
+      "step": 15720
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.818709373474121,
+      "learning_rate": 1.1173913043478261e-05,
+      "loss": 1.6497,
+      "step": 15730
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.811004161834717,
+      "learning_rate": 1.1168115942028988e-05,
+      "loss": 1.7243,
+      "step": 15740
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.756585597991943,
+      "learning_rate": 1.116231884057971e-05,
+      "loss": 1.7284,
+      "step": 15750
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.713197946548462,
+      "learning_rate": 1.1156521739130435e-05,
+      "loss": 1.7706,
+      "step": 15760
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.140644073486328,
+      "learning_rate": 1.115072463768116e-05,
+      "loss": 1.6415,
+      "step": 15770
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6640381813049316,
+      "learning_rate": 1.1144927536231884e-05,
+      "loss": 1.7129,
+      "step": 15780
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.814182996749878,
+      "learning_rate": 1.113913043478261e-05,
+      "loss": 1.5169,
+      "step": 15790
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.703028678894043,
+      "learning_rate": 1.1133333333333334e-05,
+      "loss": 1.6786,
+      "step": 15800
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.120509147644043,
+      "learning_rate": 1.112753623188406e-05,
+      "loss": 1.7163,
+      "step": 15810
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0847809314727783,
+      "learning_rate": 1.1121739130434784e-05,
+      "loss": 1.6943,
+      "step": 15820
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 10.130949974060059,
+      "learning_rate": 1.111594202898551e-05,
+      "loss": 1.729,
+      "step": 15830
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.5162465572357178,
+      "learning_rate": 1.1110144927536233e-05,
+      "loss": 1.7251,
+      "step": 15840
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.79127311706543,
+      "learning_rate": 1.1104347826086956e-05,
+      "loss": 1.7665,
+      "step": 15850
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3719213008880615,
+      "learning_rate": 1.1098550724637681e-05,
+      "loss": 1.6865,
+      "step": 15860
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.254711151123047,
+      "learning_rate": 1.1092753623188405e-05,
+      "loss": 1.7149,
+      "step": 15870
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 11.944002151489258,
+      "learning_rate": 1.1086956521739131e-05,
+      "loss": 1.6525,
+      "step": 15880
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.941390037536621,
+      "learning_rate": 1.1081159420289855e-05,
+      "loss": 1.5516,
+      "step": 15890
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.373305797576904,
+      "learning_rate": 1.1075362318840581e-05,
+      "loss": 1.473,
+      "step": 15900
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.447077751159668,
+      "learning_rate": 1.1069565217391305e-05,
+      "loss": 1.6378,
+      "step": 15910
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.908672332763672,
+      "learning_rate": 1.106376811594203e-05,
+      "loss": 1.7512,
+      "step": 15920
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0174648761749268,
+      "learning_rate": 1.1057971014492755e-05,
+      "loss": 1.6252,
+      "step": 15930
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 7.651984214782715,
+      "learning_rate": 1.105217391304348e-05,
+      "loss": 1.5333,
+      "step": 15940
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.4102983474731445,
+      "learning_rate": 1.1046376811594204e-05,
+      "loss": 1.5242,
+      "step": 15950
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.826643943786621,
+      "learning_rate": 1.1040579710144928e-05,
+      "loss": 1.6655,
+      "step": 15960
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4821903705596924,
+      "learning_rate": 1.1034782608695654e-05,
+      "loss": 1.6926,
+      "step": 15970
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.6242338418960571,
+      "learning_rate": 1.1028985507246376e-05,
+      "loss": 1.7808,
+      "step": 15980
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 8.218697547912598,
+      "learning_rate": 1.1023188405797102e-05,
+      "loss": 1.648,
+      "step": 15990
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.398993730545044,
+      "learning_rate": 1.1017391304347826e-05,
+      "loss": 1.624,
+      "step": 16000
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 1.6754653453826904,
+      "eval_runtime": 107.521,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 16000
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.014492511749268,
+      "learning_rate": 1.1011594202898552e-05,
+      "loss": 1.5021,
+      "step": 16010
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.1234304904937744,
+      "learning_rate": 1.1006376811594203e-05,
+      "loss": 1.6802,
+      "step": 16020
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.326152801513672,
+      "learning_rate": 1.1000579710144927e-05,
+      "loss": 1.7546,
+      "step": 16030
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9750605821609497,
+      "learning_rate": 1.0994782608695653e-05,
+      "loss": 1.6459,
+      "step": 16040
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2335305213928223,
+      "learning_rate": 1.0988985507246377e-05,
+      "loss": 1.6248,
+      "step": 16050
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1429505348205566,
+      "learning_rate": 1.0983188405797102e-05,
+      "loss": 1.7079,
+      "step": 16060
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.822464942932129,
+      "learning_rate": 1.0977391304347826e-05,
+      "loss": 1.7218,
+      "step": 16070
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.145028114318848,
+      "learning_rate": 1.0971594202898552e-05,
+      "loss": 1.7098,
+      "step": 16080
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.912876129150391,
+      "learning_rate": 1.0965797101449276e-05,
+      "loss": 1.7075,
+      "step": 16090
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.154576778411865,
+      "learning_rate": 1.0960000000000002e-05,
+      "loss": 1.6158,
+      "step": 16100
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7767752408981323,
+      "learning_rate": 1.0954202898550726e-05,
+      "loss": 1.5413,
+      "step": 16110
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.51847243309021,
+      "learning_rate": 1.0948405797101451e-05,
+      "loss": 1.6848,
+      "step": 16120
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7433416843414307,
+      "learning_rate": 1.0942608695652176e-05,
+      "loss": 1.6288,
+      "step": 16130
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.0296177864074707,
+      "learning_rate": 1.0936811594202898e-05,
+      "loss": 1.5401,
+      "step": 16140
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.524282932281494,
+      "learning_rate": 1.0931014492753625e-05,
+      "loss": 1.5648,
+      "step": 16150
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 7.122012138366699,
+      "learning_rate": 1.0925217391304348e-05,
+      "loss": 1.7171,
+      "step": 16160
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0002593994140625,
+      "learning_rate": 1.0919420289855073e-05,
+      "loss": 1.7173,
+      "step": 16170
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.547085762023926,
+      "learning_rate": 1.0913623188405797e-05,
+      "loss": 1.7737,
+      "step": 16180
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.6324098110198975,
+      "learning_rate": 1.0907826086956523e-05,
+      "loss": 1.7222,
+      "step": 16190
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.268777847290039,
+      "learning_rate": 1.0902028985507247e-05,
+      "loss": 1.7525,
+      "step": 16200
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.096090793609619,
+      "learning_rate": 1.0896231884057973e-05,
+      "loss": 1.5166,
+      "step": 16210
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.225011110305786,
+      "learning_rate": 1.0890434782608697e-05,
+      "loss": 1.7065,
+      "step": 16220
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.468010425567627,
+      "learning_rate": 1.088463768115942e-05,
+      "loss": 1.7441,
+      "step": 16230
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.6063239574432373,
+      "learning_rate": 1.0878840579710146e-05,
+      "loss": 1.7374,
+      "step": 16240
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.4506964683532715,
+      "learning_rate": 1.087304347826087e-05,
+      "loss": 1.8498,
+      "step": 16250
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.358245611190796,
+      "learning_rate": 1.0867246376811596e-05,
+      "loss": 1.7641,
+      "step": 16260
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0673184394836426,
+      "learning_rate": 1.0861449275362318e-05,
+      "loss": 1.7407,
+      "step": 16270
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.860028624534607,
+      "learning_rate": 1.0855652173913046e-05,
+      "loss": 1.7653,
+      "step": 16280
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.493389129638672,
+      "learning_rate": 1.0849855072463768e-05,
+      "loss": 1.5432,
+      "step": 16290
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.1333208084106445,
+      "learning_rate": 1.0844057971014494e-05,
+      "loss": 1.5558,
+      "step": 16300
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.722856283187866,
+      "learning_rate": 1.0838260869565218e-05,
+      "loss": 1.7136,
+      "step": 16310
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.029232501983643,
+      "learning_rate": 1.0832463768115943e-05,
+      "loss": 1.7611,
+      "step": 16320
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.4627902507781982,
+      "learning_rate": 1.0826666666666667e-05,
+      "loss": 1.6839,
+      "step": 16330
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.860472559928894,
+      "learning_rate": 1.0820869565217391e-05,
+      "loss": 1.6267,
+      "step": 16340
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.3834178447723389,
+      "learning_rate": 1.0815072463768117e-05,
+      "loss": 1.6913,
+      "step": 16350
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.249950885772705,
+      "learning_rate": 1.0809275362318841e-05,
+      "loss": 1.6362,
+      "step": 16360
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.343896865844727,
+      "learning_rate": 1.0803478260869567e-05,
+      "loss": 1.7563,
+      "step": 16370
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.7473175525665283,
+      "learning_rate": 1.0797681159420291e-05,
+      "loss": 1.6708,
+      "step": 16380
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.320664882659912,
+      "learning_rate": 1.0791884057971017e-05,
+      "loss": 1.8427,
+      "step": 16390
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.2452778816223145,
+      "learning_rate": 1.0786086956521739e-05,
+      "loss": 1.6387,
+      "step": 16400
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.248746395111084,
+      "learning_rate": 1.0780289855072466e-05,
+      "loss": 1.5971,
+      "step": 16410
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.274169921875,
+      "learning_rate": 1.0774492753623189e-05,
+      "loss": 1.6122,
+      "step": 16420
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0745168924331665,
+      "learning_rate": 1.0768695652173914e-05,
+      "loss": 1.6966,
+      "step": 16430
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.370829463005066,
+      "learning_rate": 1.0762898550724638e-05,
+      "loss": 1.5828,
+      "step": 16440
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.591113567352295,
+      "learning_rate": 1.0757101449275362e-05,
+      "loss": 1.6895,
+      "step": 16450
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.2173213958740234,
+      "learning_rate": 1.0751304347826088e-05,
+      "loss": 1.734,
+      "step": 16460
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.926522970199585,
+      "learning_rate": 1.0745507246376812e-05,
+      "loss": 1.6714,
+      "step": 16470
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.1437100172042847,
+      "learning_rate": 1.0739710144927538e-05,
+      "loss": 1.7062,
+      "step": 16480
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.2627272605895996,
+      "learning_rate": 1.0733913043478262e-05,
+      "loss": 1.5612,
+      "step": 16490
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.8069777488708496,
+      "learning_rate": 1.0728115942028987e-05,
+      "loss": 1.7116,
+      "step": 16500
+    },
+    {
+      "epoch": 1.38,
+      "eval_loss": 1.7130539417266846,
+      "eval_runtime": 107.5132,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 16500
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 5.705499172210693,
+      "learning_rate": 1.0722318840579711e-05,
+      "loss": 1.6609,
+      "step": 16510
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.6209776401519775,
+      "learning_rate": 1.0716521739130437e-05,
+      "loss": 1.683,
+      "step": 16520
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.3433544635772705,
+      "learning_rate": 1.071072463768116e-05,
+      "loss": 1.7747,
+      "step": 16530
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.4690423011779785,
+      "learning_rate": 1.0704927536231883e-05,
+      "loss": 1.6671,
+      "step": 16540
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.0458269119262695,
+      "learning_rate": 1.069913043478261e-05,
+      "loss": 1.7513,
+      "step": 16550
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.192028522491455,
+      "learning_rate": 1.0693333333333333e-05,
+      "loss": 1.7646,
+      "step": 16560
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.764258861541748,
+      "learning_rate": 1.0687536231884059e-05,
+      "loss": 1.6782,
+      "step": 16570
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.255431890487671,
+      "learning_rate": 1.0681739130434783e-05,
+      "loss": 1.6417,
+      "step": 16580
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.955453634262085,
+      "learning_rate": 1.0675942028985509e-05,
+      "loss": 1.7354,
+      "step": 16590
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.478759765625,
+      "learning_rate": 1.0670144927536233e-05,
+      "loss": 1.7076,
+      "step": 16600
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 10.808612823486328,
+      "learning_rate": 1.0664347826086958e-05,
+      "loss": 1.7509,
+      "step": 16610
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.3443713188171387,
+      "learning_rate": 1.0658550724637682e-05,
+      "loss": 1.6233,
+      "step": 16620
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.435446262359619,
+      "learning_rate": 1.0652753623188408e-05,
+      "loss": 1.7493,
+      "step": 16630
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0918169021606445,
+      "learning_rate": 1.0646956521739132e-05,
+      "loss": 1.7184,
+      "step": 16640
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8072077631950378,
+      "learning_rate": 1.0641159420289854e-05,
+      "loss": 1.6724,
+      "step": 16650
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0545530319213867,
+      "learning_rate": 1.063536231884058e-05,
+      "loss": 1.7191,
+      "step": 16660
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.794948101043701,
+      "learning_rate": 1.0629565217391304e-05,
+      "loss": 1.6553,
+      "step": 16670
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 6.329624176025391,
+      "learning_rate": 1.062376811594203e-05,
+      "loss": 1.7232,
+      "step": 16680
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5596625804901123,
+      "learning_rate": 1.0617971014492754e-05,
+      "loss": 1.77,
+      "step": 16690
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.4740288257598877,
+      "learning_rate": 1.061217391304348e-05,
+      "loss": 1.6394,
+      "step": 16700
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8927739858627319,
+      "learning_rate": 1.0606376811594203e-05,
+      "loss": 1.7045,
+      "step": 16710
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.0956900119781494,
+      "learning_rate": 1.0600579710144929e-05,
+      "loss": 1.7058,
+      "step": 16720
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.9234437942504883,
+      "learning_rate": 1.0594782608695653e-05,
+      "loss": 1.6324,
+      "step": 16730
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.277024745941162,
+      "learning_rate": 1.0588985507246379e-05,
+      "loss": 1.7152,
+      "step": 16740
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.355858564376831,
+      "learning_rate": 1.0583188405797103e-05,
+      "loss": 1.7431,
+      "step": 16750
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.8377106189727783,
+      "learning_rate": 1.0577391304347827e-05,
+      "loss": 1.6996,
+      "step": 16760
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.7551558017730713,
+      "learning_rate": 1.0571594202898553e-05,
+      "loss": 1.597,
+      "step": 16770
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.006781578063965,
+      "learning_rate": 1.0565797101449275e-05,
+      "loss": 1.7018,
+      "step": 16780
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.4323341846466064,
+      "learning_rate": 1.056e-05,
+      "loss": 1.5812,
+      "step": 16790
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.4357855319976807,
+      "learning_rate": 1.0554202898550725e-05,
+      "loss": 1.7026,
+      "step": 16800
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.7937941551208496,
+      "learning_rate": 1.054840579710145e-05,
+      "loss": 1.7299,
+      "step": 16810
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.6717631816864014,
+      "learning_rate": 1.0542608695652174e-05,
+      "loss": 1.759,
+      "step": 16820
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.9999706745147705,
+      "learning_rate": 1.05368115942029e-05,
+      "loss": 1.736,
+      "step": 16830
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.3674368858337402,
+      "learning_rate": 1.0531014492753624e-05,
+      "loss": 1.7221,
+      "step": 16840
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.2170531749725342,
+      "learning_rate": 1.0525217391304348e-05,
+      "loss": 1.7482,
+      "step": 16850
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1290643215179443,
+      "learning_rate": 1.0519420289855074e-05,
+      "loss": 1.5598,
+      "step": 16860
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8863322734832764,
+      "learning_rate": 1.0513623188405798e-05,
+      "loss": 1.5603,
+      "step": 16870
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.208227634429932,
+      "learning_rate": 1.0507826086956523e-05,
+      "loss": 1.7284,
+      "step": 16880
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.908130168914795,
+      "learning_rate": 1.0502028985507247e-05,
+      "loss": 1.6787,
+      "step": 16890
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 7.316353797912598,
+      "learning_rate": 1.0496231884057973e-05,
+      "loss": 1.8262,
+      "step": 16900
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.9257173538208,
+      "learning_rate": 1.0490434782608695e-05,
+      "loss": 1.7177,
+      "step": 16910
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.56701946258545,
+      "learning_rate": 1.0484637681159423e-05,
+      "loss": 1.6473,
+      "step": 16920
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.9926609992980957,
+      "learning_rate": 1.0478840579710145e-05,
+      "loss": 1.7083,
+      "step": 16930
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.859083414077759,
+      "learning_rate": 1.047304347826087e-05,
+      "loss": 1.7799,
+      "step": 16940
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.5566657781600952,
+      "learning_rate": 1.0467246376811595e-05,
+      "loss": 1.7282,
+      "step": 16950
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 8.074483871459961,
+      "learning_rate": 1.0461449275362319e-05,
+      "loss": 1.6706,
+      "step": 16960
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.9916458129882812,
+      "learning_rate": 1.0455652173913045e-05,
+      "loss": 1.6624,
+      "step": 16970
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.076849937438965,
+      "learning_rate": 1.0449855072463769e-05,
+      "loss": 1.7047,
+      "step": 16980
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.859311819076538,
+      "learning_rate": 1.0444057971014494e-05,
+      "loss": 1.6748,
+      "step": 16990
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.4924099445343018,
+      "learning_rate": 1.0438260869565218e-05,
+      "loss": 1.7947,
+      "step": 17000
+    },
+    {
+      "epoch": 1.42,
+      "eval_loss": 1.6634085178375244,
+      "eval_runtime": 107.5078,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 17000
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.4697929620742798,
+      "learning_rate": 1.0432463768115944e-05,
+      "loss": 1.652,
+      "step": 17010
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3207297325134277,
+      "learning_rate": 1.0426666666666668e-05,
+      "loss": 1.6632,
+      "step": 17020
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.104315757751465,
+      "learning_rate": 1.0420869565217394e-05,
+      "loss": 1.6028,
+      "step": 17030
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.517398715019226,
+      "learning_rate": 1.0415072463768116e-05,
+      "loss": 1.7276,
+      "step": 17040
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.3302900791168213,
+      "learning_rate": 1.0409275362318843e-05,
+      "loss": 1.5883,
+      "step": 17050
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.5856385231018066,
+      "learning_rate": 1.0403478260869566e-05,
+      "loss": 1.6082,
+      "step": 17060
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.210336685180664,
+      "learning_rate": 1.039768115942029e-05,
+      "loss": 1.7695,
+      "step": 17070
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3868658542633057,
+      "learning_rate": 1.0391884057971015e-05,
+      "loss": 1.6391,
+      "step": 17080
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.809682369232178,
+      "learning_rate": 1.038608695652174e-05,
+      "loss": 1.7408,
+      "step": 17090
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.6190929412841797,
+      "learning_rate": 1.0380289855072465e-05,
+      "loss": 1.5544,
+      "step": 17100
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.5290706157684326,
+      "learning_rate": 1.0374492753623189e-05,
+      "loss": 1.6121,
+      "step": 17110
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6158972382545471,
+      "learning_rate": 1.0368695652173915e-05,
+      "loss": 1.6292,
+      "step": 17120
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.539661407470703,
+      "learning_rate": 1.0362898550724639e-05,
+      "loss": 1.6785,
+      "step": 17130
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.0767712593078613,
+      "learning_rate": 1.0357101449275364e-05,
+      "loss": 1.7784,
+      "step": 17140
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.4671566486358643,
+      "learning_rate": 1.0351304347826088e-05,
+      "loss": 1.6613,
+      "step": 17150
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.9954442977905273,
+      "learning_rate": 1.034550724637681e-05,
+      "loss": 1.6233,
+      "step": 17160
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9763154983520508,
+      "learning_rate": 1.0339710144927536e-05,
+      "loss": 1.6937,
+      "step": 17170
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.507520079612732,
+      "learning_rate": 1.033391304347826e-05,
+      "loss": 1.765,
+      "step": 17180
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.03114652633667,
+      "learning_rate": 1.0328115942028986e-05,
+      "loss": 1.6805,
+      "step": 17190
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.9964370727539062,
+      "learning_rate": 1.032231884057971e-05,
+      "loss": 1.7913,
+      "step": 17200
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.9785213470458984,
+      "learning_rate": 1.0316521739130436e-05,
+      "loss": 1.6369,
+      "step": 17210
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.543067455291748,
+      "learning_rate": 1.031072463768116e-05,
+      "loss": 1.6102,
+      "step": 17220
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.464430332183838,
+      "learning_rate": 1.0304927536231886e-05,
+      "loss": 1.6151,
+      "step": 17230
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.594529390335083,
+      "learning_rate": 1.029913043478261e-05,
+      "loss": 1.68,
+      "step": 17240
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.4555535316467285,
+      "learning_rate": 1.0293333333333335e-05,
+      "loss": 1.6307,
+      "step": 17250
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 8.041542053222656,
+      "learning_rate": 1.028753623188406e-05,
+      "loss": 1.7092,
+      "step": 17260
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2063286304473877,
+      "learning_rate": 1.0281739130434782e-05,
+      "loss": 1.7111,
+      "step": 17270
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.018704891204834,
+      "learning_rate": 1.0275942028985509e-05,
+      "loss": 1.6344,
+      "step": 17280
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.327304840087891,
+      "learning_rate": 1.0270144927536231e-05,
+      "loss": 1.6644,
+      "step": 17290
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.130178213119507,
+      "learning_rate": 1.0264347826086957e-05,
+      "loss": 1.7562,
+      "step": 17300
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.026088237762451,
+      "learning_rate": 1.0258550724637681e-05,
+      "loss": 1.6024,
+      "step": 17310
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.9865357875823975,
+      "learning_rate": 1.0252753623188407e-05,
+      "loss": 1.6829,
+      "step": 17320
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.843345046043396,
+      "learning_rate": 1.024695652173913e-05,
+      "loss": 1.7297,
+      "step": 17330
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.475865125656128,
+      "learning_rate": 1.0241159420289856e-05,
+      "loss": 1.6453,
+      "step": 17340
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.109663963317871,
+      "learning_rate": 1.023536231884058e-05,
+      "loss": 1.6039,
+      "step": 17350
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.058063507080078,
+      "learning_rate": 1.0229565217391306e-05,
+      "loss": 1.7368,
+      "step": 17360
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.7936859130859375,
+      "learning_rate": 1.022376811594203e-05,
+      "loss": 1.6527,
+      "step": 17370
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9645744562149048,
+      "learning_rate": 1.0217971014492754e-05,
+      "loss": 1.6359,
+      "step": 17380
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9280155897140503,
+      "learning_rate": 1.021217391304348e-05,
+      "loss": 1.6755,
+      "step": 17390
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 11.76801586151123,
+      "learning_rate": 1.0206376811594204e-05,
+      "loss": 1.5459,
+      "step": 17400
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.5407767295837402,
+      "learning_rate": 1.020057971014493e-05,
+      "loss": 1.6835,
+      "step": 17410
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.4772510528564453,
+      "learning_rate": 1.0194782608695652e-05,
+      "loss": 1.7516,
+      "step": 17420
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1580195426940918,
+      "learning_rate": 1.0188985507246378e-05,
+      "loss": 1.7234,
+      "step": 17430
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6305845975875854,
+      "learning_rate": 1.0183188405797102e-05,
+      "loss": 1.5738,
+      "step": 17440
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.306994676589966,
+      "learning_rate": 1.0177391304347827e-05,
+      "loss": 1.6702,
+      "step": 17450
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.651767253875732,
+      "learning_rate": 1.0171594202898551e-05,
+      "loss": 1.6585,
+      "step": 17460
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 8.080167770385742,
+      "learning_rate": 1.0165797101449275e-05,
+      "loss": 1.8123,
+      "step": 17470
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.453287363052368,
+      "learning_rate": 1.0160000000000001e-05,
+      "loss": 1.601,
+      "step": 17480
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.5909740924835205,
+      "learning_rate": 1.0154202898550725e-05,
+      "loss": 1.7291,
+      "step": 17490
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.380848407745361,
+      "learning_rate": 1.014840579710145e-05,
+      "loss": 1.6845,
+      "step": 17500
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 1.6659685373306274,
+      "eval_runtime": 107.5088,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 17500
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.1240538358688354,
+      "learning_rate": 1.0142608695652175e-05,
+      "loss": 1.6778,
+      "step": 17510
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.835226535797119,
+      "learning_rate": 1.01368115942029e-05,
+      "loss": 1.7074,
+      "step": 17520
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.039511680603027,
+      "learning_rate": 1.0131014492753624e-05,
+      "loss": 1.6447,
+      "step": 17530
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 11.439852714538574,
+      "learning_rate": 1.012521739130435e-05,
+      "loss": 1.5716,
+      "step": 17540
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 7.208695411682129,
+      "learning_rate": 1.0119420289855072e-05,
+      "loss": 1.5687,
+      "step": 17550
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 12.316939353942871,
+      "learning_rate": 1.0113623188405798e-05,
+      "loss": 1.6564,
+      "step": 17560
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.0833916664123535,
+      "learning_rate": 1.0107826086956522e-05,
+      "loss": 1.5983,
+      "step": 17570
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.681445121765137,
+      "learning_rate": 1.0102028985507246e-05,
+      "loss": 1.6755,
+      "step": 17580
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.685708045959473,
+      "learning_rate": 1.0096231884057972e-05,
+      "loss": 1.7461,
+      "step": 17590
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8818423748016357,
+      "learning_rate": 1.0090434782608696e-05,
+      "loss": 1.6818,
+      "step": 17600
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.1815342903137207,
+      "learning_rate": 1.0084637681159422e-05,
+      "loss": 1.7476,
+      "step": 17610
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.193720817565918,
+      "learning_rate": 1.0078840579710146e-05,
+      "loss": 1.8282,
+      "step": 17620
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.380415439605713,
+      "learning_rate": 1.0073043478260871e-05,
+      "loss": 1.5551,
+      "step": 17630
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.54988956451416,
+      "learning_rate": 1.0067246376811595e-05,
+      "loss": 1.6923,
+      "step": 17640
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.875814914703369,
+      "learning_rate": 1.0061449275362321e-05,
+      "loss": 1.6452,
+      "step": 17650
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.4553890228271484,
+      "learning_rate": 1.0055652173913045e-05,
+      "loss": 1.734,
+      "step": 17660
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.171555280685425,
+      "learning_rate": 1.004985507246377e-05,
+      "loss": 1.4397,
+      "step": 17670
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.2089016437530518,
+      "learning_rate": 1.0044057971014493e-05,
+      "loss": 1.5919,
+      "step": 17680
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.439237117767334,
+      "learning_rate": 1.0038260869565217e-05,
+      "loss": 1.7071,
+      "step": 17690
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.091012954711914,
+      "learning_rate": 1.0032463768115943e-05,
+      "loss": 1.6278,
+      "step": 17700
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.4860005378723145,
+      "learning_rate": 1.0026666666666667e-05,
+      "loss": 1.6406,
+      "step": 17710
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.64132308959961,
+      "learning_rate": 1.0020869565217392e-05,
+      "loss": 1.6317,
+      "step": 17720
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.587737798690796,
+      "learning_rate": 1.0015072463768116e-05,
+      "loss": 1.5913,
+      "step": 17730
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.59721040725708,
+      "learning_rate": 1.0009275362318842e-05,
+      "loss": 1.4954,
+      "step": 17740
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 7.057238578796387,
+      "learning_rate": 1.0003478260869566e-05,
+      "loss": 1.7462,
+      "step": 17750
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.500436782836914,
+      "learning_rate": 9.99768115942029e-06,
+      "loss": 1.6778,
+      "step": 17760
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.1724953651428223,
+      "learning_rate": 9.991884057971016e-06,
+      "loss": 1.6173,
+      "step": 17770
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.133427619934082,
+      "learning_rate": 9.98608695652174e-06,
+      "loss": 1.6226,
+      "step": 17780
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.147580623626709,
+      "learning_rate": 9.980289855072465e-06,
+      "loss": 1.6144,
+      "step": 17790
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.114604949951172,
+      "learning_rate": 9.97449275362319e-06,
+      "loss": 1.7495,
+      "step": 17800
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.3972349166870117,
+      "learning_rate": 9.968695652173913e-06,
+      "loss": 1.686,
+      "step": 17810
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.335039138793945,
+      "learning_rate": 9.96289855072464e-06,
+      "loss": 1.6874,
+      "step": 17820
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.4044220447540283,
+      "learning_rate": 9.957101449275363e-06,
+      "loss": 1.6847,
+      "step": 17830
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 7.1201982498168945,
+      "learning_rate": 9.951304347826087e-06,
+      "loss": 1.6875,
+      "step": 17840
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6048938035964966,
+      "learning_rate": 9.945507246376813e-06,
+      "loss": 1.6163,
+      "step": 17850
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.1629419326782227,
+      "learning_rate": 9.939710144927537e-06,
+      "loss": 1.7531,
+      "step": 17860
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 9.67171573638916,
+      "learning_rate": 9.933913043478261e-06,
+      "loss": 1.804,
+      "step": 17870
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4759604930877686,
+      "learning_rate": 9.928115942028987e-06,
+      "loss": 1.762,
+      "step": 17880
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.9004738330841064,
+      "learning_rate": 9.92231884057971e-06,
+      "loss": 1.6533,
+      "step": 17890
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.209875106811523,
+      "learning_rate": 9.916521739130436e-06,
+      "loss": 1.6523,
+      "step": 17900
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.8896372318267822,
+      "learning_rate": 9.91072463768116e-06,
+      "loss": 1.6475,
+      "step": 17910
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.572081208229065,
+      "learning_rate": 9.904927536231886e-06,
+      "loss": 1.5881,
+      "step": 17920
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.5894155502319336,
+      "learning_rate": 9.89913043478261e-06,
+      "loss": 1.6233,
+      "step": 17930
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.091580867767334,
+      "learning_rate": 9.893333333333334e-06,
+      "loss": 1.7261,
+      "step": 17940
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.4239258766174316,
+      "learning_rate": 9.887536231884058e-06,
+      "loss": 1.6872,
+      "step": 17950
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.6650853157043457,
+      "learning_rate": 9.881739130434784e-06,
+      "loss": 1.5972,
+      "step": 17960
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.779630661010742,
+      "learning_rate": 9.875942028985508e-06,
+      "loss": 1.5442,
+      "step": 17970
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8651297092437744,
+      "learning_rate": 9.870144927536233e-06,
+      "loss": 1.6394,
+      "step": 17980
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.072394371032715,
+      "learning_rate": 9.864347826086957e-06,
+      "loss": 1.6355,
+      "step": 17990
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.223601341247559,
+      "learning_rate": 9.858550724637681e-06,
+      "loss": 1.7034,
+      "step": 18000
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 1.6817396879196167,
+      "eval_runtime": 107.5133,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 18000
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.345555782318115,
+      "learning_rate": 9.852753623188407e-06,
+      "loss": 1.6598,
+      "step": 18010
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.217700958251953,
+      "learning_rate": 9.846956521739131e-06,
+      "loss": 1.7581,
+      "step": 18020
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.60097074508667,
+      "learning_rate": 9.841159420289857e-06,
+      "loss": 1.4824,
+      "step": 18030
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.772341012954712,
+      "learning_rate": 9.83536231884058e-06,
+      "loss": 1.6643,
+      "step": 18040
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.193320631980896,
+      "learning_rate": 9.829565217391305e-06,
+      "loss": 1.8004,
+      "step": 18050
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.566644191741943,
+      "learning_rate": 9.823768115942029e-06,
+      "loss": 1.7329,
+      "step": 18060
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.771097183227539,
+      "learning_rate": 9.817971014492755e-06,
+      "loss": 1.5573,
+      "step": 18070
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.004515647888184,
+      "learning_rate": 9.812173913043479e-06,
+      "loss": 1.5583,
+      "step": 18080
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.670802116394043,
+      "learning_rate": 9.806376811594204e-06,
+      "loss": 1.6425,
+      "step": 18090
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.591397762298584,
+      "learning_rate": 9.800579710144928e-06,
+      "loss": 1.7255,
+      "step": 18100
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.184049606323242,
+      "learning_rate": 9.794782608695654e-06,
+      "loss": 1.6388,
+      "step": 18110
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.081575870513916,
+      "learning_rate": 9.788985507246378e-06,
+      "loss": 1.741,
+      "step": 18120
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.1452643871307373,
+      "learning_rate": 9.783188405797102e-06,
+      "loss": 1.5816,
+      "step": 18130
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.788238048553467,
+      "learning_rate": 9.777391304347826e-06,
+      "loss": 1.6184,
+      "step": 18140
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.260552883148193,
+      "learning_rate": 9.771594202898552e-06,
+      "loss": 1.6836,
+      "step": 18150
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.803924083709717,
+      "learning_rate": 9.765797101449276e-06,
+      "loss": 1.5638,
+      "step": 18160
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.803575873374939,
+      "learning_rate": 9.760000000000001e-06,
+      "loss": 1.7536,
+      "step": 18170
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.703744888305664,
+      "learning_rate": 9.754202898550725e-06,
+      "loss": 1.6187,
+      "step": 18180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.61587381362915,
+      "learning_rate": 9.74840579710145e-06,
+      "loss": 1.5927,
+      "step": 18190
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.6593663692474365,
+      "learning_rate": 9.742608695652175e-06,
+      "loss": 1.8459,
+      "step": 18200
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.2275209426879883,
+      "learning_rate": 9.736811594202899e-06,
+      "loss": 1.5672,
+      "step": 18210
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.009672164916992,
+      "learning_rate": 9.731014492753625e-06,
+      "loss": 1.613,
+      "step": 18220
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9421452879905701,
+      "learning_rate": 9.725217391304349e-06,
+      "loss": 1.8051,
+      "step": 18230
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.675242900848389,
+      "learning_rate": 9.719420289855075e-06,
+      "loss": 1.5781,
+      "step": 18240
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.626055717468262,
+      "learning_rate": 9.713623188405797e-06,
+      "loss": 1.6158,
+      "step": 18250
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.181307554244995,
+      "learning_rate": 9.707826086956523e-06,
+      "loss": 1.6871,
+      "step": 18260
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.5483827590942383,
+      "learning_rate": 9.702028985507247e-06,
+      "loss": 1.6884,
+      "step": 18270
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.3285396099090576,
+      "learning_rate": 9.696231884057972e-06,
+      "loss": 1.7371,
+      "step": 18280
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.175654888153076,
+      "learning_rate": 9.690434782608696e-06,
+      "loss": 1.7425,
+      "step": 18290
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.972175359725952,
+      "learning_rate": 9.684637681159422e-06,
+      "loss": 1.838,
+      "step": 18300
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.2489099502563477,
+      "learning_rate": 9.678840579710146e-06,
+      "loss": 1.6443,
+      "step": 18310
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.820380210876465,
+      "learning_rate": 9.67304347826087e-06,
+      "loss": 1.6839,
+      "step": 18320
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.2533926963806152,
+      "learning_rate": 9.667246376811596e-06,
+      "loss": 1.6476,
+      "step": 18330
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.8515777587890625,
+      "learning_rate": 9.66144927536232e-06,
+      "loss": 1.6342,
+      "step": 18340
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8389453887939453,
+      "learning_rate": 9.655652173913044e-06,
+      "loss": 1.5585,
+      "step": 18350
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.700596570968628,
+      "learning_rate": 9.649855072463768e-06,
+      "loss": 1.5371,
+      "step": 18360
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.20552396774292,
+      "learning_rate": 9.644057971014493e-06,
+      "loss": 1.7413,
+      "step": 18370
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 9.809810638427734,
+      "learning_rate": 9.638260869565217e-06,
+      "loss": 1.6194,
+      "step": 18380
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.2184398174285889,
+      "learning_rate": 9.632463768115943e-06,
+      "loss": 1.7167,
+      "step": 18390
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.6306843757629395,
+      "learning_rate": 9.626666666666667e-06,
+      "loss": 1.7024,
+      "step": 18400
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.053318500518799,
+      "learning_rate": 9.620869565217393e-06,
+      "loss": 1.6515,
+      "step": 18410
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.242835998535156,
+      "learning_rate": 9.615072463768117e-06,
+      "loss": 1.542,
+      "step": 18420
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.169401168823242,
+      "learning_rate": 9.609275362318843e-06,
+      "loss": 1.6536,
+      "step": 18430
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.6384594440460205,
+      "learning_rate": 9.603478260869567e-06,
+      "loss": 1.6468,
+      "step": 18440
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.198095321655273,
+      "learning_rate": 9.59768115942029e-06,
+      "loss": 1.6955,
+      "step": 18450
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.319828987121582,
+      "learning_rate": 9.591884057971015e-06,
+      "loss": 1.7273,
+      "step": 18460
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.001225471496582,
+      "learning_rate": 9.58608695652174e-06,
+      "loss": 1.8093,
+      "step": 18470
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.059565544128418,
+      "learning_rate": 9.580289855072464e-06,
+      "loss": 1.6228,
+      "step": 18480
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9582866430282593,
+      "learning_rate": 9.574492753623188e-06,
+      "loss": 1.6258,
+      "step": 18490
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.2751723527908325,
+      "learning_rate": 9.568695652173914e-06,
+      "loss": 1.7295,
+      "step": 18500
+    },
+    {
+      "epoch": 1.54,
+      "eval_loss": 1.6792817115783691,
+      "eval_runtime": 107.5135,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 18500
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.7914340496063232,
+      "learning_rate": 9.562898550724638e-06,
+      "loss": 1.6173,
+      "step": 18510
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 7.133903980255127,
+      "learning_rate": 9.557101449275364e-06,
+      "loss": 1.6678,
+      "step": 18520
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.9746956825256348,
+      "learning_rate": 9.551304347826088e-06,
+      "loss": 1.6362,
+      "step": 18530
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.67676305770874,
+      "learning_rate": 9.545507246376813e-06,
+      "loss": 1.6312,
+      "step": 18540
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.4198997020721436,
+      "learning_rate": 9.539710144927537e-06,
+      "loss": 1.6763,
+      "step": 18550
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.5471702814102173,
+      "learning_rate": 9.533913043478261e-06,
+      "loss": 1.6033,
+      "step": 18560
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.1242103576660156,
+      "learning_rate": 9.528115942028985e-06,
+      "loss": 1.5981,
+      "step": 18570
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.2516043186187744,
+      "learning_rate": 9.522318840579711e-06,
+      "loss": 1.7269,
+      "step": 18580
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.795421600341797,
+      "learning_rate": 9.516521739130435e-06,
+      "loss": 1.6065,
+      "step": 18590
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.5633459091186523,
+      "learning_rate": 9.51072463768116e-06,
+      "loss": 1.621,
+      "step": 18600
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.669309616088867,
+      "learning_rate": 9.504927536231885e-06,
+      "loss": 1.6455,
+      "step": 18610
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.369694232940674,
+      "learning_rate": 9.49913043478261e-06,
+      "loss": 1.55,
+      "step": 18620
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.2723934650421143,
+      "learning_rate": 9.493333333333334e-06,
+      "loss": 1.7658,
+      "step": 18630
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.9826152324676514,
+      "learning_rate": 9.487536231884058e-06,
+      "loss": 1.6784,
+      "step": 18640
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.489523410797119,
+      "learning_rate": 9.481739130434784e-06,
+      "loss": 1.7932,
+      "step": 18650
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.9629557132720947,
+      "learning_rate": 9.475942028985508e-06,
+      "loss": 1.7041,
+      "step": 18660
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.74808406829834,
+      "learning_rate": 9.470144927536232e-06,
+      "loss": 1.6448,
+      "step": 18670
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.5019195079803467,
+      "learning_rate": 9.464347826086956e-06,
+      "loss": 1.7318,
+      "step": 18680
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.352245330810547,
+      "learning_rate": 9.458550724637682e-06,
+      "loss": 1.7178,
+      "step": 18690
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.7317593097686768,
+      "learning_rate": 9.452753623188406e-06,
+      "loss": 1.6883,
+      "step": 18700
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.4841550588607788,
+      "learning_rate": 9.446956521739132e-06,
+      "loss": 1.654,
+      "step": 18710
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.4384384155273438,
+      "learning_rate": 9.441159420289856e-06,
+      "loss": 1.6427,
+      "step": 18720
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.350482225418091,
+      "learning_rate": 9.435362318840581e-06,
+      "loss": 1.5817,
+      "step": 18730
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.0648393630981445,
+      "learning_rate": 9.429565217391305e-06,
+      "loss": 1.4933,
+      "step": 18740
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.213344573974609,
+      "learning_rate": 9.423768115942031e-06,
+      "loss": 1.575,
+      "step": 18750
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 7.273472309112549,
+      "learning_rate": 9.417971014492753e-06,
+      "loss": 1.726,
+      "step": 18760
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.119070053100586,
+      "learning_rate": 9.412173913043479e-06,
+      "loss": 1.7348,
+      "step": 18770
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.862486362457275,
+      "learning_rate": 9.406376811594203e-06,
+      "loss": 1.6862,
+      "step": 18780
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6140770316123962,
+      "learning_rate": 9.400579710144929e-06,
+      "loss": 1.5872,
+      "step": 18790
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.3810670375823975,
+      "learning_rate": 9.394782608695653e-06,
+      "loss": 1.6539,
+      "step": 18800
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.982848882675171,
+      "learning_rate": 9.388985507246377e-06,
+      "loss": 1.8119,
+      "step": 18810
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6384955644607544,
+      "learning_rate": 9.383188405797102e-06,
+      "loss": 1.5366,
+      "step": 18820
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.467543125152588,
+      "learning_rate": 9.377391304347826e-06,
+      "loss": 1.641,
+      "step": 18830
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.90090799331665,
+      "learning_rate": 9.371594202898552e-06,
+      "loss": 1.6931,
+      "step": 18840
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.378636837005615,
+      "learning_rate": 9.365797101449276e-06,
+      "loss": 1.7595,
+      "step": 18850
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.310837268829346,
+      "learning_rate": 9.360000000000002e-06,
+      "loss": 1.6787,
+      "step": 18860
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.714045524597168,
+      "learning_rate": 9.354202898550724e-06,
+      "loss": 1.7402,
+      "step": 18870
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.784644842147827,
+      "learning_rate": 9.34840579710145e-06,
+      "loss": 1.565,
+      "step": 18880
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 8.056758880615234,
+      "learning_rate": 9.342608695652174e-06,
+      "loss": 1.6375,
+      "step": 18890
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.494594097137451,
+      "learning_rate": 9.3368115942029e-06,
+      "loss": 1.6977,
+      "step": 18900
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.0827395915985107,
+      "learning_rate": 9.331014492753624e-06,
+      "loss": 1.6315,
+      "step": 18910
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.295031547546387,
+      "learning_rate": 9.32521739130435e-06,
+      "loss": 1.7304,
+      "step": 18920
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8913993835449219,
+      "learning_rate": 9.319420289855073e-06,
+      "loss": 1.7426,
+      "step": 18930
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0160393714904785,
+      "learning_rate": 9.313623188405799e-06,
+      "loss": 1.7997,
+      "step": 18940
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.568789482116699,
+      "learning_rate": 9.307826086956523e-06,
+      "loss": 1.699,
+      "step": 18950
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.200634479522705,
+      "learning_rate": 9.302028985507247e-06,
+      "loss": 1.5647,
+      "step": 18960
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.3525495529174805,
+      "learning_rate": 9.296231884057971e-06,
+      "loss": 1.6899,
+      "step": 18970
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.9189863204956055,
+      "learning_rate": 9.290434782608697e-06,
+      "loss": 1.7537,
+      "step": 18980
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.9687070846557617,
+      "learning_rate": 9.28463768115942e-06,
+      "loss": 1.6875,
+      "step": 18990
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.2662925720214844,
+      "learning_rate": 9.278840579710145e-06,
+      "loss": 1.6292,
+      "step": 19000
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 1.665460228919983,
+      "eval_runtime": 107.4967,
+      "eval_samples_per_second": 9.303,
+      "eval_steps_per_second": 2.326,
+      "step": 19000
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.05936336517334,
+      "learning_rate": 9.27304347826087e-06,
+      "loss": 1.5103,
+      "step": 19010
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.9469785690307617,
+      "learning_rate": 9.267246376811594e-06,
+      "loss": 1.6837,
+      "step": 19020
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.0714728832244873,
+      "learning_rate": 9.26144927536232e-06,
+      "loss": 1.725,
+      "step": 19030
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.3942904472351074,
+      "learning_rate": 9.255652173913044e-06,
+      "loss": 1.6261,
+      "step": 19040
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.1022446155548096,
+      "learning_rate": 9.24985507246377e-06,
+      "loss": 1.7056,
+      "step": 19050
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8929376602172852,
+      "learning_rate": 9.244057971014494e-06,
+      "loss": 1.666,
+      "step": 19060
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.3702325820922852,
+      "learning_rate": 9.238260869565218e-06,
+      "loss": 1.6023,
+      "step": 19070
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.481436014175415,
+      "learning_rate": 9.232463768115942e-06,
+      "loss": 1.6922,
+      "step": 19080
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.172421455383301,
+      "learning_rate": 9.226666666666668e-06,
+      "loss": 1.7799,
+      "step": 19090
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.3141672611236572,
+      "learning_rate": 9.220869565217392e-06,
+      "loss": 1.326,
+      "step": 19100
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.576709270477295,
+      "learning_rate": 9.215072463768117e-06,
+      "loss": 1.6309,
+      "step": 19110
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9501328468322754,
+      "learning_rate": 9.209275362318841e-06,
+      "loss": 1.6507,
+      "step": 19120
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.715846300125122,
+      "learning_rate": 9.203478260869565e-06,
+      "loss": 1.6828,
+      "step": 19130
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.534573554992676,
+      "learning_rate": 9.197681159420291e-06,
+      "loss": 1.619,
+      "step": 19140
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.9621031284332275,
+      "learning_rate": 9.191884057971015e-06,
+      "loss": 1.6573,
+      "step": 19150
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.051302194595337,
+      "learning_rate": 9.18608695652174e-06,
+      "loss": 1.71,
+      "step": 19160
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.671502113342285,
+      "learning_rate": 9.180289855072465e-06,
+      "loss": 1.6008,
+      "step": 19170
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.0547165870666504,
+      "learning_rate": 9.174492753623189e-06,
+      "loss": 1.7074,
+      "step": 19180
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.7989306449890137,
+      "learning_rate": 9.168695652173913e-06,
+      "loss": 1.7383,
+      "step": 19190
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.843214750289917,
+      "learning_rate": 9.162898550724638e-06,
+      "loss": 1.6794,
+      "step": 19200
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.1126368045806885,
+      "learning_rate": 9.157101449275362e-06,
+      "loss": 1.6367,
+      "step": 19210
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7680147886276245,
+      "learning_rate": 9.151304347826088e-06,
+      "loss": 1.6034,
+      "step": 19220
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.450879096984863,
+      "learning_rate": 9.145507246376812e-06,
+      "loss": 1.632,
+      "step": 19230
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.6349544525146484,
+      "learning_rate": 9.139710144927538e-06,
+      "loss": 1.6385,
+      "step": 19240
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9699334502220154,
+      "learning_rate": 9.133913043478262e-06,
+      "loss": 1.5493,
+      "step": 19250
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.04218053817749,
+      "learning_rate": 9.128115942028986e-06,
+      "loss": 1.6542,
+      "step": 19260
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.5029866695404053,
+      "learning_rate": 9.122318840579712e-06,
+      "loss": 1.8018,
+      "step": 19270
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.081104278564453,
+      "learning_rate": 9.116521739130436e-06,
+      "loss": 1.6778,
+      "step": 19280
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 6.336639881134033,
+      "learning_rate": 9.11072463768116e-06,
+      "loss": 1.5925,
+      "step": 19290
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.3441286087036133,
+      "learning_rate": 9.104927536231885e-06,
+      "loss": 1.6447,
+      "step": 19300
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.648870468139648,
+      "learning_rate": 9.09913043478261e-06,
+      "loss": 1.8095,
+      "step": 19310
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.4305033683776855,
+      "learning_rate": 9.093333333333333e-06,
+      "loss": 1.6808,
+      "step": 19320
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9530631303787231,
+      "learning_rate": 9.087536231884059e-06,
+      "loss": 1.6648,
+      "step": 19330
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.785933494567871,
+      "learning_rate": 9.081739130434783e-06,
+      "loss": 1.5965,
+      "step": 19340
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.8842251300811768,
+      "learning_rate": 9.075942028985509e-06,
+      "loss": 1.6147,
+      "step": 19350
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.127828121185303,
+      "learning_rate": 9.070144927536233e-06,
+      "loss": 1.7384,
+      "step": 19360
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9837595224380493,
+      "learning_rate": 9.064347826086958e-06,
+      "loss": 1.8099,
+      "step": 19370
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.733158588409424,
+      "learning_rate": 9.05855072463768e-06,
+      "loss": 1.5608,
+      "step": 19380
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.9853909015655518,
+      "learning_rate": 9.052753623188406e-06,
+      "loss": 1.6227,
+      "step": 19390
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.49697208404541,
+      "learning_rate": 9.04695652173913e-06,
+      "loss": 1.5671,
+      "step": 19400
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.7057849168777466,
+      "learning_rate": 9.041159420289856e-06,
+      "loss": 1.6561,
+      "step": 19410
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.0063562393188477,
+      "learning_rate": 9.03536231884058e-06,
+      "loss": 1.7223,
+      "step": 19420
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.740435600280762,
+      "learning_rate": 9.029565217391306e-06,
+      "loss": 1.6543,
+      "step": 19430
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.221389293670654,
+      "learning_rate": 9.02376811594203e-06,
+      "loss": 1.673,
+      "step": 19440
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.550647735595703,
+      "learning_rate": 9.017971014492754e-06,
+      "loss": 1.5946,
+      "step": 19450
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.3443706035614014,
+      "learning_rate": 9.01217391304348e-06,
+      "loss": 1.6832,
+      "step": 19460
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3949636220932007,
+      "learning_rate": 9.006376811594203e-06,
+      "loss": 1.5993,
+      "step": 19470
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.4200923442840576,
+      "learning_rate": 9.00057971014493e-06,
+      "loss": 1.6081,
+      "step": 19480
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.661345958709717,
+      "learning_rate": 8.994782608695653e-06,
+      "loss": 1.7607,
+      "step": 19490
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.108694314956665,
+      "learning_rate": 8.988985507246377e-06,
+      "loss": 1.6209,
+      "step": 19500
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 1.6830703020095825,
+      "eval_runtime": 107.5287,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 19500
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.532451629638672,
+      "learning_rate": 8.983188405797101e-06,
+      "loss": 1.6172,
+      "step": 19510
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.787123441696167,
+      "learning_rate": 8.977391304347827e-06,
+      "loss": 1.6074,
+      "step": 19520
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.405480146408081,
+      "learning_rate": 8.971594202898551e-06,
+      "loss": 1.6032,
+      "step": 19530
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.917876243591309,
+      "learning_rate": 8.965797101449277e-06,
+      "loss": 1.6732,
+      "step": 19540
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.5339460372924805,
+      "learning_rate": 8.96e-06,
+      "loss": 1.7914,
+      "step": 19550
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.5086684226989746,
+      "learning_rate": 8.954202898550726e-06,
+      "loss": 1.5816,
+      "step": 19560
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 9.773869514465332,
+      "learning_rate": 8.94840579710145e-06,
+      "loss": 1.67,
+      "step": 19570
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.713991165161133,
+      "learning_rate": 8.942608695652174e-06,
+      "loss": 1.7089,
+      "step": 19580
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.009612560272217,
+      "learning_rate": 8.936811594202898e-06,
+      "loss": 1.7485,
+      "step": 19590
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.547402858734131,
+      "learning_rate": 8.931014492753624e-06,
+      "loss": 1.5412,
+      "step": 19600
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.649285078048706,
+      "learning_rate": 8.925217391304348e-06,
+      "loss": 1.7208,
+      "step": 19610
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.029441833496094,
+      "learning_rate": 8.919420289855074e-06,
+      "loss": 1.6938,
+      "step": 19620
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.3328795433044434,
+      "learning_rate": 8.913623188405798e-06,
+      "loss": 1.5802,
+      "step": 19630
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.51261043548584,
+      "learning_rate": 8.907826086956522e-06,
+      "loss": 1.6576,
+      "step": 19640
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.171065330505371,
+      "learning_rate": 8.902028985507247e-06,
+      "loss": 1.5348,
+      "step": 19650
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 5.7769856452941895,
+      "learning_rate": 8.896231884057971e-06,
+      "loss": 1.6869,
+      "step": 19660
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.950839042663574,
+      "learning_rate": 8.890434782608697e-06,
+      "loss": 1.6537,
+      "step": 19670
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.615936279296875,
+      "learning_rate": 8.884637681159421e-06,
+      "loss": 1.7185,
+      "step": 19680
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.862908959388733,
+      "learning_rate": 8.878840579710145e-06,
+      "loss": 1.6145,
+      "step": 19690
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.8000476360321045,
+      "learning_rate": 8.87304347826087e-06,
+      "loss": 1.736,
+      "step": 19700
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.0812816619873047,
+      "learning_rate": 8.867246376811595e-06,
+      "loss": 1.4757,
+      "step": 19710
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.9957151412963867,
+      "learning_rate": 8.861449275362319e-06,
+      "loss": 1.6096,
+      "step": 19720
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.853338241577148,
+      "learning_rate": 8.855652173913045e-06,
+      "loss": 1.6642,
+      "step": 19730
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.507507085800171,
+      "learning_rate": 8.849855072463769e-06,
+      "loss": 1.6589,
+      "step": 19740
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.7916064262390137,
+      "learning_rate": 8.844057971014494e-06,
+      "loss": 1.6656,
+      "step": 19750
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 9.049138069152832,
+      "learning_rate": 8.838260869565218e-06,
+      "loss": 1.6074,
+      "step": 19760
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.988990306854248,
+      "learning_rate": 8.832463768115942e-06,
+      "loss": 1.697,
+      "step": 19770
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.35215163230896,
+      "learning_rate": 8.826666666666668e-06,
+      "loss": 1.6842,
+      "step": 19780
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.3260128498077393,
+      "learning_rate": 8.820869565217392e-06,
+      "loss": 1.6981,
+      "step": 19790
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1436783075332642,
+      "learning_rate": 8.815072463768116e-06,
+      "loss": 1.7511,
+      "step": 19800
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.495922565460205,
+      "learning_rate": 8.809275362318842e-06,
+      "loss": 1.5865,
+      "step": 19810
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.033712148666382,
+      "learning_rate": 8.803478260869566e-06,
+      "loss": 1.6375,
+      "step": 19820
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9689141511917114,
+      "learning_rate": 8.79768115942029e-06,
+      "loss": 1.5311,
+      "step": 19830
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.1666340827941895,
+      "learning_rate": 8.791884057971015e-06,
+      "loss": 1.7137,
+      "step": 19840
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.735472202301025,
+      "learning_rate": 8.78608695652174e-06,
+      "loss": 1.7429,
+      "step": 19850
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 7.482568264007568,
+      "learning_rate": 8.780289855072465e-06,
+      "loss": 1.5834,
+      "step": 19860
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.3337314128875732,
+      "learning_rate": 8.774492753623189e-06,
+      "loss": 1.7067,
+      "step": 19870
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.9348156452178955,
+      "learning_rate": 8.768695652173915e-06,
+      "loss": 1.7376,
+      "step": 19880
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.5238430500030518,
+      "learning_rate": 8.762898550724639e-06,
+      "loss": 1.5539,
+      "step": 19890
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.653454303741455,
+      "learning_rate": 8.757101449275363e-06,
+      "loss": 1.7208,
+      "step": 19900
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.208953857421875,
+      "learning_rate": 8.751304347826087e-06,
+      "loss": 1.5903,
+      "step": 19910
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.089755535125732,
+      "learning_rate": 8.745507246376813e-06,
+      "loss": 1.4553,
+      "step": 19920
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.7506258487701416,
+      "learning_rate": 8.739710144927537e-06,
+      "loss": 1.6508,
+      "step": 19930
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.436467170715332,
+      "learning_rate": 8.733913043478262e-06,
+      "loss": 1.8371,
+      "step": 19940
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.3517704010009766,
+      "learning_rate": 8.728115942028986e-06,
+      "loss": 1.7193,
+      "step": 19950
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.238973379135132,
+      "learning_rate": 8.72231884057971e-06,
+      "loss": 1.7621,
+      "step": 19960
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.6009724140167236,
+      "learning_rate": 8.716521739130436e-06,
+      "loss": 1.8128,
+      "step": 19970
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.952020645141602,
+      "learning_rate": 8.71072463768116e-06,
+      "loss": 1.7867,
+      "step": 19980
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.133454322814941,
+      "learning_rate": 8.704927536231886e-06,
+      "loss": 1.7099,
+      "step": 19990
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.35225510597229,
+      "learning_rate": 8.69913043478261e-06,
+      "loss": 1.7125,
+      "step": 20000
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 1.6788996458053589,
+      "eval_runtime": 107.4956,
+      "eval_samples_per_second": 9.303,
+      "eval_steps_per_second": 2.326,
+      "step": 20000
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.7249436378479004,
+      "learning_rate": 8.693333333333334e-06,
+      "loss": 1.6412,
+      "step": 20010
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.052044868469238,
+      "learning_rate": 8.687536231884058e-06,
+      "loss": 1.4474,
+      "step": 20020
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.286026954650879,
+      "learning_rate": 8.681739130434783e-06,
+      "loss": 1.7352,
+      "step": 20030
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 9.765252113342285,
+      "learning_rate": 8.675942028985507e-06,
+      "loss": 1.6815,
+      "step": 20040
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.1208224296569824,
+      "learning_rate": 8.670144927536233e-06,
+      "loss": 1.7426,
+      "step": 20050
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.312246799468994,
+      "learning_rate": 8.664347826086957e-06,
+      "loss": 1.7116,
+      "step": 20060
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.1685314178466797,
+      "learning_rate": 8.658550724637683e-06,
+      "loss": 1.7183,
+      "step": 20070
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.873128890991211,
+      "learning_rate": 8.652753623188407e-06,
+      "loss": 1.811,
+      "step": 20080
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.21943473815918,
+      "learning_rate": 8.64695652173913e-06,
+      "loss": 1.6304,
+      "step": 20090
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.778602123260498,
+      "learning_rate": 8.641159420289857e-06,
+      "loss": 1.6324,
+      "step": 20100
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.6876754760742188,
+      "learning_rate": 8.63536231884058e-06,
+      "loss": 1.6039,
+      "step": 20110
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.973613739013672,
+      "learning_rate": 8.629565217391305e-06,
+      "loss": 1.7671,
+      "step": 20120
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4355108737945557,
+      "learning_rate": 8.62376811594203e-06,
+      "loss": 1.6335,
+      "step": 20130
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.8951032161712646,
+      "learning_rate": 8.617971014492754e-06,
+      "loss": 1.7804,
+      "step": 20140
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.68106746673584,
+      "learning_rate": 8.612173913043478e-06,
+      "loss": 1.5194,
+      "step": 20150
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.569166660308838,
+      "learning_rate": 8.606376811594204e-06,
+      "loss": 1.6177,
+      "step": 20160
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.555330276489258,
+      "learning_rate": 8.600579710144928e-06,
+      "loss": 1.6004,
+      "step": 20170
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8976370096206665,
+      "learning_rate": 8.59536231884058e-06,
+      "loss": 1.5999,
+      "step": 20180
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.5247642993927002,
+      "learning_rate": 8.589565217391305e-06,
+      "loss": 1.6919,
+      "step": 20190
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.274435997009277,
+      "learning_rate": 8.583768115942029e-06,
+      "loss": 1.5242,
+      "step": 20200
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.763355255126953,
+      "learning_rate": 8.577971014492755e-06,
+      "loss": 1.4803,
+      "step": 20210
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.4387519359588623,
+      "learning_rate": 8.572173913043479e-06,
+      "loss": 1.5794,
+      "step": 20220
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.496255874633789,
+      "learning_rate": 8.566376811594204e-06,
+      "loss": 1.6001,
+      "step": 20230
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4461989402770996,
+      "learning_rate": 8.560579710144928e-06,
+      "loss": 1.7451,
+      "step": 20240
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.702919006347656,
+      "learning_rate": 8.554782608695652e-06,
+      "loss": 1.6566,
+      "step": 20250
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9157620668411255,
+      "learning_rate": 8.548985507246378e-06,
+      "loss": 1.5958,
+      "step": 20260
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.364008903503418,
+      "learning_rate": 8.543188405797102e-06,
+      "loss": 1.6164,
+      "step": 20270
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.2819979190826416,
+      "learning_rate": 8.537391304347826e-06,
+      "loss": 1.7358,
+      "step": 20280
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.845746994018555,
+      "learning_rate": 8.531594202898552e-06,
+      "loss": 1.6241,
+      "step": 20290
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.371389627456665,
+      "learning_rate": 8.525797101449276e-06,
+      "loss": 1.6148,
+      "step": 20300
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.634834289550781,
+      "learning_rate": 8.52e-06,
+      "loss": 1.7313,
+      "step": 20310
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 11.510677337646484,
+      "learning_rate": 8.514202898550725e-06,
+      "loss": 1.6605,
+      "step": 20320
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.1439342498779297,
+      "learning_rate": 8.50840579710145e-06,
+      "loss": 1.5686,
+      "step": 20330
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.642844796180725,
+      "learning_rate": 8.502608695652175e-06,
+      "loss": 1.6276,
+      "step": 20340
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.143886089324951,
+      "learning_rate": 8.496811594202899e-06,
+      "loss": 1.5472,
+      "step": 20350
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.869246006011963,
+      "learning_rate": 8.491014492753625e-06,
+      "loss": 1.6785,
+      "step": 20360
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.9599814414978027,
+      "learning_rate": 8.485217391304349e-06,
+      "loss": 1.6191,
+      "step": 20370
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.2835745811462402,
+      "learning_rate": 8.479420289855073e-06,
+      "loss": 1.7113,
+      "step": 20380
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.400454998016357,
+      "learning_rate": 8.473623188405797e-06,
+      "loss": 1.7069,
+      "step": 20390
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.021839141845703,
+      "learning_rate": 8.467826086956523e-06,
+      "loss": 1.7014,
+      "step": 20400
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.3753602504730225,
+      "learning_rate": 8.462028985507247e-06,
+      "loss": 1.6844,
+      "step": 20410
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.501394271850586,
+      "learning_rate": 8.456231884057972e-06,
+      "loss": 1.7469,
+      "step": 20420
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 16.001710891723633,
+      "learning_rate": 8.450434782608696e-06,
+      "loss": 1.7221,
+      "step": 20430
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.816986083984375,
+      "learning_rate": 8.44463768115942e-06,
+      "loss": 1.7513,
+      "step": 20440
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.2568550109863281,
+      "learning_rate": 8.438840579710146e-06,
+      "loss": 1.5909,
+      "step": 20450
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2014676332473755,
+      "learning_rate": 8.43304347826087e-06,
+      "loss": 1.5539,
+      "step": 20460
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.841648817062378,
+      "learning_rate": 8.427246376811596e-06,
+      "loss": 1.6945,
+      "step": 20470
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.8036043643951416,
+      "learning_rate": 8.42144927536232e-06,
+      "loss": 1.6231,
+      "step": 20480
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.289719343185425,
+      "learning_rate": 8.415652173913044e-06,
+      "loss": 1.5679,
+      "step": 20490
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.5245754718780518,
+      "learning_rate": 8.409855072463768e-06,
+      "loss": 1.5689,
+      "step": 20500
+    },
+    {
+      "epoch": 1.71,
+      "eval_loss": 1.6623921394348145,
+      "eval_runtime": 107.5043,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 20500
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 9.230620384216309,
+      "learning_rate": 8.404057971014493e-06,
+      "loss": 1.4972,
+      "step": 20510
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.701317310333252,
+      "learning_rate": 8.398260869565217e-06,
+      "loss": 1.7549,
+      "step": 20520
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.25839900970459,
+      "learning_rate": 8.392463768115943e-06,
+      "loss": 1.66,
+      "step": 20530
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.2549502849578857,
+      "learning_rate": 8.386666666666667e-06,
+      "loss": 1.8211,
+      "step": 20540
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.639694690704346,
+      "learning_rate": 8.380869565217393e-06,
+      "loss": 1.6038,
+      "step": 20550
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.658710956573486,
+      "learning_rate": 8.375072463768117e-06,
+      "loss": 1.7501,
+      "step": 20560
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.85872745513916,
+      "learning_rate": 8.36927536231884e-06,
+      "loss": 1.5652,
+      "step": 20570
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.761110544204712,
+      "learning_rate": 8.363478260869567e-06,
+      "loss": 1.6325,
+      "step": 20580
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.664401054382324,
+      "learning_rate": 8.35768115942029e-06,
+      "loss": 1.6404,
+      "step": 20590
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.662811279296875,
+      "learning_rate": 8.351884057971015e-06,
+      "loss": 1.5619,
+      "step": 20600
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.5120227336883545,
+      "learning_rate": 8.34608695652174e-06,
+      "loss": 1.7325,
+      "step": 20610
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.742555618286133,
+      "learning_rate": 8.340289855072464e-06,
+      "loss": 1.7704,
+      "step": 20620
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6222686767578125,
+      "learning_rate": 8.334492753623188e-06,
+      "loss": 1.6901,
+      "step": 20630
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.6312642097473145,
+      "learning_rate": 8.328695652173914e-06,
+      "loss": 1.728,
+      "step": 20640
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7402911186218262,
+      "learning_rate": 8.322898550724638e-06,
+      "loss": 1.6404,
+      "step": 20650
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.214266777038574,
+      "learning_rate": 8.317101449275364e-06,
+      "loss": 1.7745,
+      "step": 20660
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.402463674545288,
+      "learning_rate": 8.311304347826088e-06,
+      "loss": 1.6937,
+      "step": 20670
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.309727668762207,
+      "learning_rate": 8.305507246376813e-06,
+      "loss": 1.633,
+      "step": 20680
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6282193660736084,
+      "learning_rate": 8.299710144927537e-06,
+      "loss": 1.6727,
+      "step": 20690
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.9179277420043945,
+      "learning_rate": 8.293913043478261e-06,
+      "loss": 1.6977,
+      "step": 20700
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.5713701248168945,
+      "learning_rate": 8.288115942028985e-06,
+      "loss": 1.7506,
+      "step": 20710
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.8336901664733887,
+      "learning_rate": 8.282318840579711e-06,
+      "loss": 1.6477,
+      "step": 20720
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.1530611515045166,
+      "learning_rate": 8.276521739130435e-06,
+      "loss": 1.6213,
+      "step": 20730
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.452496290206909,
+      "learning_rate": 8.27072463768116e-06,
+      "loss": 1.5587,
+      "step": 20740
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 10.321528434753418,
+      "learning_rate": 8.264927536231885e-06,
+      "loss": 1.6628,
+      "step": 20750
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.9270588159561157,
+      "learning_rate": 8.259130434782609e-06,
+      "loss": 1.6388,
+      "step": 20760
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.518328905105591,
+      "learning_rate": 8.253333333333334e-06,
+      "loss": 1.6791,
+      "step": 20770
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.031456470489502,
+      "learning_rate": 8.247536231884059e-06,
+      "loss": 1.5683,
+      "step": 20780
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.8340535163879395,
+      "learning_rate": 8.241739130434784e-06,
+      "loss": 1.5371,
+      "step": 20790
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.309947967529297,
+      "learning_rate": 8.235942028985508e-06,
+      "loss": 1.5707,
+      "step": 20800
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.8457183837890625,
+      "learning_rate": 8.230144927536232e-06,
+      "loss": 1.6439,
+      "step": 20810
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.240163803100586,
+      "learning_rate": 8.224347826086956e-06,
+      "loss": 1.7126,
+      "step": 20820
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.0228397846221924,
+      "learning_rate": 8.218550724637682e-06,
+      "loss": 1.5457,
+      "step": 20830
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.194091796875,
+      "learning_rate": 8.212753623188406e-06,
+      "loss": 1.6615,
+      "step": 20840
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.5026867389678955,
+      "learning_rate": 8.206956521739132e-06,
+      "loss": 1.5305,
+      "step": 20850
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.275241374969482,
+      "learning_rate": 8.201159420289856e-06,
+      "loss": 1.5742,
+      "step": 20860
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.040112257003784,
+      "learning_rate": 8.195362318840581e-06,
+      "loss": 1.6762,
+      "step": 20870
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.858863592147827,
+      "learning_rate": 8.189565217391305e-06,
+      "loss": 1.7056,
+      "step": 20880
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.349377155303955,
+      "learning_rate": 8.18376811594203e-06,
+      "loss": 1.5865,
+      "step": 20890
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.167051792144775,
+      "learning_rate": 8.177971014492753e-06,
+      "loss": 1.712,
+      "step": 20900
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.692874431610107,
+      "learning_rate": 8.172173913043479e-06,
+      "loss": 1.6484,
+      "step": 20910
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 13.255517959594727,
+      "learning_rate": 8.166376811594203e-06,
+      "loss": 1.7715,
+      "step": 20920
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.60584831237793,
+      "learning_rate": 8.160579710144929e-06,
+      "loss": 1.5116,
+      "step": 20930
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.476940631866455,
+      "learning_rate": 8.154782608695653e-06,
+      "loss": 1.6632,
+      "step": 20940
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.8701999187469482,
+      "learning_rate": 8.148985507246377e-06,
+      "loss": 1.6628,
+      "step": 20950
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.236228346824646,
+      "learning_rate": 8.143188405797102e-06,
+      "loss": 1.4092,
+      "step": 20960
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.624843120574951,
+      "learning_rate": 8.137391304347826e-06,
+      "loss": 1.742,
+      "step": 20970
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.9257400035858154,
+      "learning_rate": 8.131594202898552e-06,
+      "loss": 1.6487,
+      "step": 20980
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.001676082611084,
+      "learning_rate": 8.125797101449276e-06,
+      "loss": 1.6715,
+      "step": 20990
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8285595178604126,
+      "learning_rate": 8.120000000000002e-06,
+      "loss": 1.5451,
+      "step": 21000
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 1.6759874820709229,
+      "eval_runtime": 107.5007,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 21000
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 6.6983747482299805,
+      "learning_rate": 8.114202898550724e-06,
+      "loss": 1.6217,
+      "step": 21010
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.789233446121216,
+      "learning_rate": 8.10840579710145e-06,
+      "loss": 1.5388,
+      "step": 21020
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 9.76033878326416,
+      "learning_rate": 8.102608695652174e-06,
+      "loss": 1.7604,
+      "step": 21030
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.5290467739105225,
+      "learning_rate": 8.0968115942029e-06,
+      "loss": 1.6859,
+      "step": 21040
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.0937581062316895,
+      "learning_rate": 8.091014492753624e-06,
+      "loss": 1.481,
+      "step": 21050
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.91987943649292,
+      "learning_rate": 8.08521739130435e-06,
+      "loss": 1.5735,
+      "step": 21060
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.2947590351104736,
+      "learning_rate": 8.079420289855073e-06,
+      "loss": 1.6143,
+      "step": 21070
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.225331783294678,
+      "learning_rate": 8.073623188405797e-06,
+      "loss": 1.6024,
+      "step": 21080
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.730823993682861,
+      "learning_rate": 8.067826086956523e-06,
+      "loss": 1.667,
+      "step": 21090
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8926206827163696,
+      "learning_rate": 8.062028985507247e-06,
+      "loss": 1.6495,
+      "step": 21100
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.650490760803223,
+      "learning_rate": 8.056231884057971e-06,
+      "loss": 1.7194,
+      "step": 21110
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.5820090770721436,
+      "learning_rate": 8.050434782608697e-06,
+      "loss": 1.591,
+      "step": 21120
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5368926525115967,
+      "learning_rate": 8.04463768115942e-06,
+      "loss": 1.6234,
+      "step": 21130
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6347370147705078,
+      "learning_rate": 8.038840579710145e-06,
+      "loss": 1.7416,
+      "step": 21140
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 10.997393608093262,
+      "learning_rate": 8.03304347826087e-06,
+      "loss": 1.7531,
+      "step": 21150
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.327144145965576,
+      "learning_rate": 8.027246376811594e-06,
+      "loss": 1.4644,
+      "step": 21160
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6970939636230469,
+      "learning_rate": 8.02144927536232e-06,
+      "loss": 1.761,
+      "step": 21170
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.163623332977295,
+      "learning_rate": 8.015652173913044e-06,
+      "loss": 1.6956,
+      "step": 21180
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5294394493103027,
+      "learning_rate": 8.00985507246377e-06,
+      "loss": 1.7231,
+      "step": 21190
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.195105075836182,
+      "learning_rate": 8.004057971014494e-06,
+      "loss": 1.7115,
+      "step": 21200
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.7161829471588135,
+      "learning_rate": 7.998260869565218e-06,
+      "loss": 1.7947,
+      "step": 21210
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7141447067260742,
+      "learning_rate": 7.992463768115942e-06,
+      "loss": 1.6888,
+      "step": 21220
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.2814693450927734,
+      "learning_rate": 7.986666666666668e-06,
+      "loss": 1.5235,
+      "step": 21230
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.9310901165008545,
+      "learning_rate": 7.980869565217392e-06,
+      "loss": 1.6429,
+      "step": 21240
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.654700994491577,
+      "learning_rate": 7.975072463768117e-06,
+      "loss": 1.6168,
+      "step": 21250
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.43727445602417,
+      "learning_rate": 7.969275362318841e-06,
+      "loss": 1.6492,
+      "step": 21260
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 9.46625804901123,
+      "learning_rate": 7.963478260869565e-06,
+      "loss": 1.735,
+      "step": 21270
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.2169768810272217,
+      "learning_rate": 7.957681159420291e-06,
+      "loss": 1.5743,
+      "step": 21280
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.560244560241699,
+      "learning_rate": 7.951884057971015e-06,
+      "loss": 1.5637,
+      "step": 21290
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.38496208190918,
+      "learning_rate": 7.94608695652174e-06,
+      "loss": 1.6104,
+      "step": 21300
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.6433284282684326,
+      "learning_rate": 7.940289855072465e-06,
+      "loss": 1.644,
+      "step": 21310
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.7574374675750732,
+      "learning_rate": 7.934492753623189e-06,
+      "loss": 1.6456,
+      "step": 21320
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.709298610687256,
+      "learning_rate": 7.928695652173913e-06,
+      "loss": 1.6377,
+      "step": 21330
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.04289174079895,
+      "learning_rate": 7.922898550724638e-06,
+      "loss": 1.6613,
+      "step": 21340
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.023538112640381,
+      "learning_rate": 7.917101449275362e-06,
+      "loss": 1.7339,
+      "step": 21350
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.419352054595947,
+      "learning_rate": 7.911304347826088e-06,
+      "loss": 1.7423,
+      "step": 21360
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 6.251063823699951,
+      "learning_rate": 7.905507246376812e-06,
+      "loss": 1.5134,
+      "step": 21370
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.289116859436035,
+      "learning_rate": 7.899710144927538e-06,
+      "loss": 1.6577,
+      "step": 21380
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.699251651763916,
+      "learning_rate": 7.893913043478262e-06,
+      "loss": 1.6832,
+      "step": 21390
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.5587689876556396,
+      "learning_rate": 7.888115942028986e-06,
+      "loss": 1.6295,
+      "step": 21400
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.560650110244751,
+      "learning_rate": 7.882318840579712e-06,
+      "loss": 1.6037,
+      "step": 21410
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.8277647495269775,
+      "learning_rate": 7.876521739130436e-06,
+      "loss": 1.6725,
+      "step": 21420
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.283670425415039,
+      "learning_rate": 7.87072463768116e-06,
+      "loss": 1.6771,
+      "step": 21430
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.659524917602539,
+      "learning_rate": 7.864927536231885e-06,
+      "loss": 1.7715,
+      "step": 21440
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.301861763000488,
+      "learning_rate": 7.85913043478261e-06,
+      "loss": 1.5021,
+      "step": 21450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.5721640586853027,
+      "learning_rate": 7.853333333333333e-06,
+      "loss": 1.5547,
+      "step": 21460
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.1962008476257324,
+      "learning_rate": 7.847536231884059e-06,
+      "loss": 1.6434,
+      "step": 21470
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.1904305219650269,
+      "learning_rate": 7.841739130434783e-06,
+      "loss": 1.7324,
+      "step": 21480
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.97566556930542,
+      "learning_rate": 7.835942028985509e-06,
+      "loss": 1.5438,
+      "step": 21490
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.0194718837738037,
+      "learning_rate": 7.830144927536233e-06,
+      "loss": 1.8216,
+      "step": 21500
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 1.6438640356063843,
+      "eval_runtime": 107.5066,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 21500
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.1328859329223633,
+      "learning_rate": 7.824347826086958e-06,
+      "loss": 1.6511,
+      "step": 21510
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 16.312145233154297,
+      "learning_rate": 7.81855072463768e-06,
+      "loss": 1.6316,
+      "step": 21520
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.9844260215759277,
+      "learning_rate": 7.812753623188406e-06,
+      "loss": 1.6512,
+      "step": 21530
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.2778491973876953,
+      "learning_rate": 7.80695652173913e-06,
+      "loss": 1.5582,
+      "step": 21540
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.904211044311523,
+      "learning_rate": 7.801159420289856e-06,
+      "loss": 1.55,
+      "step": 21550
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.345255374908447,
+      "learning_rate": 7.79536231884058e-06,
+      "loss": 1.6164,
+      "step": 21560
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.088725566864014,
+      "learning_rate": 7.789565217391306e-06,
+      "loss": 1.5831,
+      "step": 21570
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.778326988220215,
+      "learning_rate": 7.78376811594203e-06,
+      "loss": 1.6626,
+      "step": 21580
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.572787284851074,
+      "learning_rate": 7.777971014492754e-06,
+      "loss": 1.7164,
+      "step": 21590
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.807311534881592,
+      "learning_rate": 7.77217391304348e-06,
+      "loss": 1.5888,
+      "step": 21600
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.578456401824951,
+      "learning_rate": 7.766376811594203e-06,
+      "loss": 1.7378,
+      "step": 21610
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.514986276626587,
+      "learning_rate": 7.76057971014493e-06,
+      "loss": 1.72,
+      "step": 21620
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.335219383239746,
+      "learning_rate": 7.754782608695653e-06,
+      "loss": 1.6237,
+      "step": 21630
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.087151527404785,
+      "learning_rate": 7.748985507246377e-06,
+      "loss": 1.6961,
+      "step": 21640
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.020003318786621,
+      "learning_rate": 7.743188405797101e-06,
+      "loss": 1.5885,
+      "step": 21650
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.0945472717285156,
+      "learning_rate": 7.737391304347827e-06,
+      "loss": 1.656,
+      "step": 21660
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 10.43069076538086,
+      "learning_rate": 7.731594202898551e-06,
+      "loss": 1.7332,
+      "step": 21670
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.4677326679229736,
+      "learning_rate": 7.725797101449277e-06,
+      "loss": 1.5584,
+      "step": 21680
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4591469764709473,
+      "learning_rate": 7.72e-06,
+      "loss": 1.777,
+      "step": 21690
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.6109650135040283,
+      "learning_rate": 7.714202898550726e-06,
+      "loss": 1.7946,
+      "step": 21700
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.215765953063965,
+      "learning_rate": 7.70840579710145e-06,
+      "loss": 1.5901,
+      "step": 21710
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.505645751953125,
+      "learning_rate": 7.702608695652174e-06,
+      "loss": 1.6892,
+      "step": 21720
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5885299444198608,
+      "learning_rate": 7.696811594202898e-06,
+      "loss": 1.7118,
+      "step": 21730
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.334027051925659,
+      "learning_rate": 7.691014492753624e-06,
+      "loss": 1.647,
+      "step": 21740
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 8.775703430175781,
+      "learning_rate": 7.685217391304348e-06,
+      "loss": 1.6092,
+      "step": 21750
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.955008506774902,
+      "learning_rate": 7.679420289855074e-06,
+      "loss": 1.5739,
+      "step": 21760
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.728649139404297,
+      "learning_rate": 7.673623188405798e-06,
+      "loss": 1.532,
+      "step": 21770
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.9815008640289307,
+      "learning_rate": 7.667826086956522e-06,
+      "loss": 1.6954,
+      "step": 21780
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.196691513061523,
+      "learning_rate": 7.662028985507247e-06,
+      "loss": 1.6589,
+      "step": 21790
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.2602643966674805,
+      "learning_rate": 7.656231884057971e-06,
+      "loss": 1.6467,
+      "step": 21800
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.570413112640381,
+      "learning_rate": 7.650434782608697e-06,
+      "loss": 1.6406,
+      "step": 21810
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 11.799811363220215,
+      "learning_rate": 7.644637681159421e-06,
+      "loss": 1.7035,
+      "step": 21820
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.307993412017822,
+      "learning_rate": 7.638840579710145e-06,
+      "loss": 1.6075,
+      "step": 21830
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.131196975708008,
+      "learning_rate": 7.63304347826087e-06,
+      "loss": 1.5969,
+      "step": 21840
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.6842682361602783,
+      "learning_rate": 7.627246376811595e-06,
+      "loss": 1.6489,
+      "step": 21850
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.3321924209594727,
+      "learning_rate": 7.621449275362319e-06,
+      "loss": 1.5213,
+      "step": 21860
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 8.997515678405762,
+      "learning_rate": 7.615652173913044e-06,
+      "loss": 1.7068,
+      "step": 21870
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.286634922027588,
+      "learning_rate": 7.609855072463769e-06,
+      "loss": 1.6143,
+      "step": 21880
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.148158073425293,
+      "learning_rate": 7.6040579710144934e-06,
+      "loss": 1.7376,
+      "step": 21890
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.665143013000488,
+      "learning_rate": 7.598260869565218e-06,
+      "loss": 1.6027,
+      "step": 21900
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.7268582582473755,
+      "learning_rate": 7.592463768115943e-06,
+      "loss": 1.5526,
+      "step": 21910
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.673620581626892,
+      "learning_rate": 7.586666666666668e-06,
+      "loss": 1.7982,
+      "step": 21920
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.819336175918579,
+      "learning_rate": 7.580869565217393e-06,
+      "loss": 1.7283,
+      "step": 21930
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.207809925079346,
+      "learning_rate": 7.575072463768116e-06,
+      "loss": 1.5976,
+      "step": 21940
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.5421364307403564,
+      "learning_rate": 7.569275362318841e-06,
+      "loss": 1.7216,
+      "step": 21950
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.2121992111206055,
+      "learning_rate": 7.563478260869566e-06,
+      "loss": 1.5512,
+      "step": 21960
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.52865743637085,
+      "learning_rate": 7.5576811594202906e-06,
+      "loss": 1.6024,
+      "step": 21970
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 6.420586109161377,
+      "learning_rate": 7.5518840579710154e-06,
+      "loss": 1.6298,
+      "step": 21980
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 9.54375171661377,
+      "learning_rate": 7.5460869565217394e-06,
+      "loss": 1.7235,
+      "step": 21990
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.1675570011138916,
+      "learning_rate": 7.540289855072464e-06,
+      "loss": 1.8252,
+      "step": 22000
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 1.6268221139907837,
+      "eval_runtime": 107.5262,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 22000
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.670650005340576,
+      "learning_rate": 7.534492753623189e-06,
+      "loss": 1.7059,
+      "step": 22010
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6870843172073364,
+      "learning_rate": 7.528695652173914e-06,
+      "loss": 1.8291,
+      "step": 22020
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.0037596225738525,
+      "learning_rate": 7.522898550724639e-06,
+      "loss": 1.7288,
+      "step": 22030
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.6715378761291504,
+      "learning_rate": 7.517101449275363e-06,
+      "loss": 1.5929,
+      "step": 22040
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.457519769668579,
+      "learning_rate": 7.511304347826087e-06,
+      "loss": 1.4734,
+      "step": 22050
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 7.1646342277526855,
+      "learning_rate": 7.505507246376812e-06,
+      "loss": 1.7495,
+      "step": 22060
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.050849437713623,
+      "learning_rate": 7.4997101449275366e-06,
+      "loss": 1.667,
+      "step": 22070
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.6044504642486572,
+      "learning_rate": 7.493913043478261e-06,
+      "loss": 1.4592,
+      "step": 22080
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.669327974319458,
+      "learning_rate": 7.488115942028986e-06,
+      "loss": 1.563,
+      "step": 22090
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.431074142456055,
+      "learning_rate": 7.482318840579711e-06,
+      "loss": 1.5672,
+      "step": 22100
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.121443748474121,
+      "learning_rate": 7.476521739130436e-06,
+      "loss": 1.6698,
+      "step": 22110
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.799599289894104,
+      "learning_rate": 7.470724637681161e-06,
+      "loss": 1.6929,
+      "step": 22120
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.3442468643188477,
+      "learning_rate": 7.464927536231885e-06,
+      "loss": 1.7485,
+      "step": 22130
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.691938638687134,
+      "learning_rate": 7.459130434782609e-06,
+      "loss": 1.727,
+      "step": 22140
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.82914400100708,
+      "learning_rate": 7.453333333333334e-06,
+      "loss": 1.5514,
+      "step": 22150
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.622535467147827,
+      "learning_rate": 7.4475362318840585e-06,
+      "loss": 1.7807,
+      "step": 22160
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.5094232559204102,
+      "learning_rate": 7.441739130434783e-06,
+      "loss": 1.7863,
+      "step": 22170
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.016897201538086,
+      "learning_rate": 7.435942028985507e-06,
+      "loss": 1.7699,
+      "step": 22180
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.2854208946228027,
+      "learning_rate": 7.430144927536232e-06,
+      "loss": 1.7514,
+      "step": 22190
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.095964431762695,
+      "learning_rate": 7.424347826086957e-06,
+      "loss": 1.6181,
+      "step": 22200
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.0803048610687256,
+      "learning_rate": 7.418550724637682e-06,
+      "loss": 1.6808,
+      "step": 22210
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.521609306335449,
+      "learning_rate": 7.413333333333333e-06,
+      "loss": 1.641,
+      "step": 22220
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.716726303100586,
+      "learning_rate": 7.407536231884058e-06,
+      "loss": 1.7521,
+      "step": 22230
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.207428455352783,
+      "learning_rate": 7.401739130434783e-06,
+      "loss": 1.6185,
+      "step": 22240
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.840510368347168,
+      "learning_rate": 7.395942028985508e-06,
+      "loss": 1.4995,
+      "step": 22250
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.4733941555023193,
+      "learning_rate": 7.390144927536233e-06,
+      "loss": 1.5198,
+      "step": 22260
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.7702128887176514,
+      "learning_rate": 7.3843478260869575e-06,
+      "loss": 1.543,
+      "step": 22270
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.0251762866973877,
+      "learning_rate": 7.378550724637682e-06,
+      "loss": 1.5978,
+      "step": 22280
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.0895233154296875,
+      "learning_rate": 7.372753623188406e-06,
+      "loss": 1.6949,
+      "step": 22290
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.8260159492492676,
+      "learning_rate": 7.366956521739131e-06,
+      "loss": 1.7792,
+      "step": 22300
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.7007031440734863,
+      "learning_rate": 7.361159420289856e-06,
+      "loss": 1.6979,
+      "step": 22310
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.845219612121582,
+      "learning_rate": 7.35536231884058e-06,
+      "loss": 1.3902,
+      "step": 22320
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.411277770996094,
+      "learning_rate": 7.349565217391305e-06,
+      "loss": 1.6546,
+      "step": 22330
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.6635563373565674,
+      "learning_rate": 7.343768115942029e-06,
+      "loss": 1.7551,
+      "step": 22340
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.20759391784668,
+      "learning_rate": 7.337971014492754e-06,
+      "loss": 1.574,
+      "step": 22350
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.9890128374099731,
+      "learning_rate": 7.332173913043479e-06,
+      "loss": 1.6738,
+      "step": 22360
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 8.436697959899902,
+      "learning_rate": 7.3263768115942035e-06,
+      "loss": 1.6415,
+      "step": 22370
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.9131646156311035,
+      "learning_rate": 7.320579710144928e-06,
+      "loss": 1.5153,
+      "step": 22380
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.4065988063812256,
+      "learning_rate": 7.314782608695653e-06,
+      "loss": 1.6153,
+      "step": 22390
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.279060125350952,
+      "learning_rate": 7.308985507246378e-06,
+      "loss": 1.6871,
+      "step": 22400
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.864788770675659,
+      "learning_rate": 7.303188405797103e-06,
+      "loss": 1.6629,
+      "step": 22410
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 6.575710296630859,
+      "learning_rate": 7.297391304347826e-06,
+      "loss": 1.5627,
+      "step": 22420
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.330643653869629,
+      "learning_rate": 7.291594202898551e-06,
+      "loss": 1.7939,
+      "step": 22430
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.9671368598937988,
+      "learning_rate": 7.285797101449276e-06,
+      "loss": 1.6338,
+      "step": 22440
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.052875518798828,
+      "learning_rate": 7.280000000000001e-06,
+      "loss": 1.7247,
+      "step": 22450
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 8.003440856933594,
+      "learning_rate": 7.2742028985507255e-06,
+      "loss": 1.6002,
+      "step": 22460
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.555729866027832,
+      "learning_rate": 7.26840579710145e-06,
+      "loss": 1.7289,
+      "step": 22470
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.3797450065612793,
+      "learning_rate": 7.262608695652174e-06,
+      "loss": 1.5819,
+      "step": 22480
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.605469226837158,
+      "learning_rate": 7.256811594202899e-06,
+      "loss": 1.6578,
+      "step": 22490
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.348262786865234,
+      "learning_rate": 7.251014492753624e-06,
+      "loss": 1.6862,
+      "step": 22500
+    },
+    {
+      "epoch": 1.88,
+      "eval_loss": 1.624435544013977,
+      "eval_runtime": 107.5057,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 22500
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7244126796722412,
+      "learning_rate": 7.245217391304349e-06,
+      "loss": 1.6801,
+      "step": 22510
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.5117290019989014,
+      "learning_rate": 7.239420289855073e-06,
+      "loss": 1.6471,
+      "step": 22520
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.5990796089172363,
+      "learning_rate": 7.233623188405797e-06,
+      "loss": 1.625,
+      "step": 22530
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.40592622756958,
+      "learning_rate": 7.227826086956522e-06,
+      "loss": 1.6896,
+      "step": 22540
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.524301528930664,
+      "learning_rate": 7.222028985507247e-06,
+      "loss": 1.6493,
+      "step": 22550
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.01413631439209,
+      "learning_rate": 7.2162318840579714e-06,
+      "loss": 1.6456,
+      "step": 22560
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.1266565322875977,
+      "learning_rate": 7.210434782608696e-06,
+      "loss": 1.6527,
+      "step": 22570
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.7390304803848267,
+      "learning_rate": 7.204637681159421e-06,
+      "loss": 1.6855,
+      "step": 22580
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.171137571334839,
+      "learning_rate": 7.198840579710146e-06,
+      "loss": 1.6591,
+      "step": 22590
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.5304112434387207,
+      "learning_rate": 7.193043478260871e-06,
+      "loss": 1.62,
+      "step": 22600
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.766373872756958,
+      "learning_rate": 7.187246376811595e-06,
+      "loss": 1.7098,
+      "step": 22610
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.5186986923217773,
+      "learning_rate": 7.18144927536232e-06,
+      "loss": 1.5517,
+      "step": 22620
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9023741483688354,
+      "learning_rate": 7.175652173913044e-06,
+      "loss": 1.7088,
+      "step": 22630
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.54264497756958,
+      "learning_rate": 7.1698550724637686e-06,
+      "loss": 1.6399,
+      "step": 22640
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 7.001004695892334,
+      "learning_rate": 7.164057971014493e-06,
+      "loss": 1.5971,
+      "step": 22650
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.7402238845825195,
+      "learning_rate": 7.1582608695652174e-06,
+      "loss": 1.6201,
+      "step": 22660
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.761517524719238,
+      "learning_rate": 7.152463768115942e-06,
+      "loss": 1.6743,
+      "step": 22670
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6771907806396484,
+      "learning_rate": 7.146666666666667e-06,
+      "loss": 1.5944,
+      "step": 22680
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 9.115729331970215,
+      "learning_rate": 7.140869565217392e-06,
+      "loss": 1.6292,
+      "step": 22690
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.5204763412475586,
+      "learning_rate": 7.135072463768117e-06,
+      "loss": 1.6079,
+      "step": 22700
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9801039695739746,
+      "learning_rate": 7.129275362318842e-06,
+      "loss": 1.6728,
+      "step": 22710
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.535811424255371,
+      "learning_rate": 7.1234782608695665e-06,
+      "loss": 1.5367,
+      "step": 22720
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.705811023712158,
+      "learning_rate": 7.11768115942029e-06,
+      "loss": 1.6359,
+      "step": 22730
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.373073577880859,
+      "learning_rate": 7.1118840579710146e-06,
+      "loss": 1.5461,
+      "step": 22740
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.482778549194336,
+      "learning_rate": 7.106086956521739e-06,
+      "loss": 1.6968,
+      "step": 22750
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 7.682448387145996,
+      "learning_rate": 7.100289855072464e-06,
+      "loss": 1.6213,
+      "step": 22760
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.669218063354492,
+      "learning_rate": 7.094492753623189e-06,
+      "loss": 1.6408,
+      "step": 22770
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.680028200149536,
+      "learning_rate": 7.088695652173914e-06,
+      "loss": 1.6225,
+      "step": 22780
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 9.829727172851562,
+      "learning_rate": 7.082898550724638e-06,
+      "loss": 1.5706,
+      "step": 22790
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.793923377990723,
+      "learning_rate": 7.077101449275363e-06,
+      "loss": 1.6649,
+      "step": 22800
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.1820645332336426,
+      "learning_rate": 7.071304347826088e-06,
+      "loss": 1.518,
+      "step": 22810
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.055821418762207,
+      "learning_rate": 7.0655072463768125e-06,
+      "loss": 1.6835,
+      "step": 22820
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.942118167877197,
+      "learning_rate": 7.0597101449275365e-06,
+      "loss": 1.5547,
+      "step": 22830
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.1956920623779297,
+      "learning_rate": 7.053913043478261e-06,
+      "loss": 1.7392,
+      "step": 22840
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6930139064788818,
+      "learning_rate": 7.048115942028985e-06,
+      "loss": 1.7377,
+      "step": 22850
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.8593344688415527,
+      "learning_rate": 7.04231884057971e-06,
+      "loss": 1.6003,
+      "step": 22860
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.152258396148682,
+      "learning_rate": 7.036521739130435e-06,
+      "loss": 1.671,
+      "step": 22870
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.3631834983825684,
+      "learning_rate": 7.03072463768116e-06,
+      "loss": 1.5857,
+      "step": 22880
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.4059417247772217,
+      "learning_rate": 7.024927536231885e-06,
+      "loss": 1.7629,
+      "step": 22890
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.9621164798736572,
+      "learning_rate": 7.01913043478261e-06,
+      "loss": 1.7076,
+      "step": 22900
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.742729127407074,
+      "learning_rate": 7.0133333333333345e-06,
+      "loss": 1.6253,
+      "step": 22910
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.2152140140533447,
+      "learning_rate": 7.007536231884059e-06,
+      "loss": 1.6246,
+      "step": 22920
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.5290796756744385,
+      "learning_rate": 7.001739130434783e-06,
+      "loss": 1.4322,
+      "step": 22930
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.470486164093018,
+      "learning_rate": 6.995942028985507e-06,
+      "loss": 1.6678,
+      "step": 22940
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.923020124435425,
+      "learning_rate": 6.990144927536232e-06,
+      "loss": 1.5589,
+      "step": 22950
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.6417629718780518,
+      "learning_rate": 6.984347826086957e-06,
+      "loss": 1.5959,
+      "step": 22960
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.574911117553711,
+      "learning_rate": 6.978550724637682e-06,
+      "loss": 1.7439,
+      "step": 22970
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8391579389572144,
+      "learning_rate": 6.972753623188406e-06,
+      "loss": 1.6204,
+      "step": 22980
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.0558955669403076,
+      "learning_rate": 6.966956521739131e-06,
+      "loss": 1.6833,
+      "step": 22990
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.424144744873047,
+      "learning_rate": 6.961159420289856e-06,
+      "loss": 1.7008,
+      "step": 23000
+    },
+    {
+      "epoch": 1.92,
+      "eval_loss": 1.6694403886795044,
+      "eval_runtime": 107.5169,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 23000
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.5954649448394775,
+      "learning_rate": 6.9553623188405805e-06,
+      "loss": 1.6293,
+      "step": 23010
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.2413156032562256,
+      "learning_rate": 6.949565217391305e-06,
+      "loss": 1.6196,
+      "step": 23020
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.4454963207244873,
+      "learning_rate": 6.94376811594203e-06,
+      "loss": 1.689,
+      "step": 23030
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.627885580062866,
+      "learning_rate": 6.937971014492753e-06,
+      "loss": 1.623,
+      "step": 23040
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9731435775756836,
+      "learning_rate": 6.932173913043478e-06,
+      "loss": 1.5804,
+      "step": 23050
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.2060045003890991,
+      "learning_rate": 6.926376811594203e-06,
+      "loss": 1.5979,
+      "step": 23060
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.452988147735596,
+      "learning_rate": 6.920579710144928e-06,
+      "loss": 1.693,
+      "step": 23070
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.145307779312134,
+      "learning_rate": 6.914782608695653e-06,
+      "loss": 1.7325,
+      "step": 23080
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.082767009735107,
+      "learning_rate": 6.908985507246378e-06,
+      "loss": 1.6337,
+      "step": 23090
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.040972948074341,
+      "learning_rate": 6.9031884057971025e-06,
+      "loss": 1.6223,
+      "step": 23100
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.870182752609253,
+      "learning_rate": 6.8973913043478265e-06,
+      "loss": 1.7026,
+      "step": 23110
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9211017489433289,
+      "learning_rate": 6.891594202898551e-06,
+      "loss": 1.642,
+      "step": 23120
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.97402286529541,
+      "learning_rate": 6.885797101449276e-06,
+      "loss": 1.6483,
+      "step": 23130
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.9657325744628906,
+      "learning_rate": 6.88e-06,
+      "loss": 1.6167,
+      "step": 23140
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.580372333526611,
+      "learning_rate": 6.874202898550725e-06,
+      "loss": 1.8135,
+      "step": 23150
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.190333843231201,
+      "learning_rate": 6.86840579710145e-06,
+      "loss": 1.669,
+      "step": 23160
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 9.4367094039917,
+      "learning_rate": 6.862608695652174e-06,
+      "loss": 1.5822,
+      "step": 23170
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.261363506317139,
+      "learning_rate": 6.856811594202899e-06,
+      "loss": 1.5777,
+      "step": 23180
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.0194172859191895,
+      "learning_rate": 6.851014492753624e-06,
+      "loss": 1.7024,
+      "step": 23190
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.917288780212402,
+      "learning_rate": 6.8452173913043485e-06,
+      "loss": 1.565,
+      "step": 23200
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.0884828567504883,
+      "learning_rate": 6.839420289855073e-06,
+      "loss": 1.7616,
+      "step": 23210
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 8.573901176452637,
+      "learning_rate": 6.833623188405798e-06,
+      "loss": 1.7546,
+      "step": 23220
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.3873114585876465,
+      "learning_rate": 6.827826086956523e-06,
+      "loss": 1.69,
+      "step": 23230
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.1748428344726562,
+      "learning_rate": 6.822028985507248e-06,
+      "loss": 1.5809,
+      "step": 23240
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.9165292978286743,
+      "learning_rate": 6.816231884057971e-06,
+      "loss": 1.6822,
+      "step": 23250
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.971045970916748,
+      "learning_rate": 6.810434782608696e-06,
+      "loss": 1.7906,
+      "step": 23260
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 7.549450397491455,
+      "learning_rate": 6.804637681159421e-06,
+      "loss": 1.6903,
+      "step": 23270
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.463688611984253,
+      "learning_rate": 6.798840579710146e-06,
+      "loss": 1.601,
+      "step": 23280
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.8097610473632812,
+      "learning_rate": 6.7930434782608704e-06,
+      "loss": 1.6345,
+      "step": 23290
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.737390518188477,
+      "learning_rate": 6.7872463768115945e-06,
+      "loss": 1.5124,
+      "step": 23300
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.170960426330566,
+      "learning_rate": 6.781449275362319e-06,
+      "loss": 1.6134,
+      "step": 23310
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.446510076522827,
+      "learning_rate": 6.775652173913044e-06,
+      "loss": 1.6565,
+      "step": 23320
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.7264490127563477,
+      "learning_rate": 6.769855072463769e-06,
+      "loss": 1.7299,
+      "step": 23330
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.8352794647216797,
+      "learning_rate": 6.764057971014494e-06,
+      "loss": 1.6823,
+      "step": 23340
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.742753982543945,
+      "learning_rate": 6.758260869565217e-06,
+      "loss": 1.6586,
+      "step": 23350
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.161562204360962,
+      "learning_rate": 6.752463768115942e-06,
+      "loss": 1.5763,
+      "step": 23360
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.118688583374023,
+      "learning_rate": 6.746666666666667e-06,
+      "loss": 1.6764,
+      "step": 23370
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.455047369003296,
+      "learning_rate": 6.740869565217392e-06,
+      "loss": 1.7486,
+      "step": 23380
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.586467742919922,
+      "learning_rate": 6.7350724637681164e-06,
+      "loss": 1.6002,
+      "step": 23390
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.8399118185043335,
+      "learning_rate": 6.729275362318841e-06,
+      "loss": 1.617,
+      "step": 23400
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.191316604614258,
+      "learning_rate": 6.723478260869566e-06,
+      "loss": 1.6777,
+      "step": 23410
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.110720634460449,
+      "learning_rate": 6.717681159420291e-06,
+      "loss": 1.627,
+      "step": 23420
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.99454927444458,
+      "learning_rate": 6.711884057971015e-06,
+      "loss": 1.6708,
+      "step": 23430
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.553925514221191,
+      "learning_rate": 6.70608695652174e-06,
+      "loss": 1.5284,
+      "step": 23440
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.1461501121521,
+      "learning_rate": 6.700289855072464e-06,
+      "loss": 1.6239,
+      "step": 23450
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.479620933532715,
+      "learning_rate": 6.694492753623189e-06,
+      "loss": 1.7803,
+      "step": 23460
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.193335056304932,
+      "learning_rate": 6.6886956521739136e-06,
+      "loss": 1.726,
+      "step": 23470
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.909477710723877,
+      "learning_rate": 6.682898550724638e-06,
+      "loss": 1.7277,
+      "step": 23480
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2803231477737427,
+      "learning_rate": 6.677101449275362e-06,
+      "loss": 1.6677,
+      "step": 23490
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.0764338970184326,
+      "learning_rate": 6.671304347826087e-06,
+      "loss": 1.6192,
+      "step": 23500
+    },
+    {
+      "epoch": 1.96,
+      "eval_loss": 1.663988471031189,
+      "eval_runtime": 107.5268,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 23500
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.041726112365723,
+      "learning_rate": 6.665507246376812e-06,
+      "loss": 1.6856,
+      "step": 23510
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 7.944361209869385,
+      "learning_rate": 6.659710144927537e-06,
+      "loss": 1.7495,
+      "step": 23520
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.918194532394409,
+      "learning_rate": 6.653913043478262e-06,
+      "loss": 1.6635,
+      "step": 23530
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 8.257862091064453,
+      "learning_rate": 6.648115942028987e-06,
+      "loss": 1.717,
+      "step": 23540
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 6.422058582305908,
+      "learning_rate": 6.6423188405797115e-06,
+      "loss": 1.6294,
+      "step": 23550
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.261116027832031,
+      "learning_rate": 6.636521739130435e-06,
+      "loss": 1.712,
+      "step": 23560
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.29148006439209,
+      "learning_rate": 6.6307246376811595e-06,
+      "loss": 1.7626,
+      "step": 23570
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.524306774139404,
+      "learning_rate": 6.624927536231884e-06,
+      "loss": 1.4419,
+      "step": 23580
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.170078277587891,
+      "learning_rate": 6.619130434782609e-06,
+      "loss": 1.7131,
+      "step": 23590
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 7.53287410736084,
+      "learning_rate": 6.613333333333334e-06,
+      "loss": 1.6708,
+      "step": 23600
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.6637678146362305,
+      "learning_rate": 6.607536231884059e-06,
+      "loss": 1.8157,
+      "step": 23610
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.637990117073059,
+      "learning_rate": 6.601739130434783e-06,
+      "loss": 1.5326,
+      "step": 23620
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.433206081390381,
+      "learning_rate": 6.595942028985508e-06,
+      "loss": 1.8064,
+      "step": 23630
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.04767370223999,
+      "learning_rate": 6.590144927536233e-06,
+      "loss": 1.5902,
+      "step": 23640
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.667262554168701,
+      "learning_rate": 6.5843478260869575e-06,
+      "loss": 1.5682,
+      "step": 23650
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.9290482997894287,
+      "learning_rate": 6.5785507246376815e-06,
+      "loss": 1.6237,
+      "step": 23660
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.297651529312134,
+      "learning_rate": 6.5727536231884055e-06,
+      "loss": 1.537,
+      "step": 23670
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.4150543212890625,
+      "learning_rate": 6.56695652173913e-06,
+      "loss": 1.6064,
+      "step": 23680
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4117975234985352,
+      "learning_rate": 6.561159420289855e-06,
+      "loss": 1.6012,
+      "step": 23690
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.487409591674805,
+      "learning_rate": 6.55536231884058e-06,
+      "loss": 1.6483,
+      "step": 23700
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.014012575149536,
+      "learning_rate": 6.549565217391305e-06,
+      "loss": 1.592,
+      "step": 23710
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.575296401977539,
+      "learning_rate": 6.54376811594203e-06,
+      "loss": 1.574,
+      "step": 23720
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.4668456315994263,
+      "learning_rate": 6.537971014492755e-06,
+      "loss": 1.6368,
+      "step": 23730
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.705700874328613,
+      "learning_rate": 6.5321739130434795e-06,
+      "loss": 1.6754,
+      "step": 23740
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.414182424545288,
+      "learning_rate": 6.5263768115942035e-06,
+      "loss": 1.5873,
+      "step": 23750
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.606837749481201,
+      "learning_rate": 6.5205797101449275e-06,
+      "loss": 1.7331,
+      "step": 23760
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 6.832721710205078,
+      "learning_rate": 6.514782608695652e-06,
+      "loss": 1.668,
+      "step": 23770
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.788400173187256,
+      "learning_rate": 6.508985507246377e-06,
+      "loss": 1.5374,
+      "step": 23780
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.2363767623901367,
+      "learning_rate": 6.503188405797102e-06,
+      "loss": 1.6609,
+      "step": 23790
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.4389679431915283,
+      "learning_rate": 6.497391304347826e-06,
+      "loss": 1.6405,
+      "step": 23800
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.902224540710449,
+      "learning_rate": 6.491594202898551e-06,
+      "loss": 1.6073,
+      "step": 23810
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.991961717605591,
+      "learning_rate": 6.485797101449276e-06,
+      "loss": 1.6646,
+      "step": 23820
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.761397838592529,
+      "learning_rate": 6.480000000000001e-06,
+      "loss": 1.6553,
+      "step": 23830
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.5798428058624268,
+      "learning_rate": 6.4742028985507255e-06,
+      "loss": 1.6409,
+      "step": 23840
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.4175989627838135,
+      "learning_rate": 6.46840579710145e-06,
+      "loss": 1.7199,
+      "step": 23850
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.976370334625244,
+      "learning_rate": 6.462608695652175e-06,
+      "loss": 1.6925,
+      "step": 23860
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.479163646697998,
+      "learning_rate": 6.456811594202898e-06,
+      "loss": 1.6045,
+      "step": 23870
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.5710270404815674,
+      "learning_rate": 6.451014492753623e-06,
+      "loss": 1.6237,
+      "step": 23880
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.615454196929932,
+      "learning_rate": 6.445217391304348e-06,
+      "loss": 1.7599,
+      "step": 23890
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 7.336389064788818,
+      "learning_rate": 6.439420289855073e-06,
+      "loss": 1.4353,
+      "step": 23900
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 8.621050834655762,
+      "learning_rate": 6.433623188405798e-06,
+      "loss": 1.6964,
+      "step": 23910
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.770620346069336,
+      "learning_rate": 6.427826086956523e-06,
+      "loss": 1.467,
+      "step": 23920
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.004281520843506,
+      "learning_rate": 6.4220289855072475e-06,
+      "loss": 1.6323,
+      "step": 23930
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.508981704711914,
+      "learning_rate": 6.4162318840579715e-06,
+      "loss": 1.648,
+      "step": 23940
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.081500053405762,
+      "learning_rate": 6.410434782608696e-06,
+      "loss": 1.5768,
+      "step": 23950
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.746595859527588,
+      "learning_rate": 6.404637681159421e-06,
+      "loss": 1.6595,
+      "step": 23960
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.1742167472839355,
+      "learning_rate": 6.398840579710145e-06,
+      "loss": 1.6841,
+      "step": 23970
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.1276462078094482,
+      "learning_rate": 6.39304347826087e-06,
+      "loss": 1.5213,
+      "step": 23980
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.458132028579712,
+      "learning_rate": 6.387246376811594e-06,
+      "loss": 1.5896,
+      "step": 23990
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.596000671386719,
+      "learning_rate": 6.381449275362319e-06,
+      "loss": 1.6449,
+      "step": 24000
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.6730765104293823,
+      "eval_runtime": 107.5278,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 24000
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.354638576507568,
+      "learning_rate": 6.375652173913044e-06,
+      "loss": 1.5624,
+      "step": 24010
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6478346586227417,
+      "learning_rate": 6.369855072463769e-06,
+      "loss": 1.7343,
+      "step": 24020
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.3177080154418945,
+      "learning_rate": 6.3640579710144935e-06,
+      "loss": 1.4786,
+      "step": 24030
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.199551582336426,
+      "learning_rate": 6.358260869565218e-06,
+      "loss": 1.5448,
+      "step": 24040
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.634474754333496,
+      "learning_rate": 6.352463768115943e-06,
+      "loss": 1.4954,
+      "step": 24050
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.2182674407958984,
+      "learning_rate": 6.346666666666668e-06,
+      "loss": 1.6519,
+      "step": 24060
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 12.482495307922363,
+      "learning_rate": 6.340869565217391e-06,
+      "loss": 1.5864,
+      "step": 24070
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.257279872894287,
+      "learning_rate": 6.335072463768116e-06,
+      "loss": 1.4682,
+      "step": 24080
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.081076622009277,
+      "learning_rate": 6.329275362318841e-06,
+      "loss": 1.5972,
+      "step": 24090
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.3276896476745605,
+      "learning_rate": 6.323478260869566e-06,
+      "loss": 1.658,
+      "step": 24100
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.144803524017334,
+      "learning_rate": 6.317681159420291e-06,
+      "loss": 1.4938,
+      "step": 24110
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 11.347249031066895,
+      "learning_rate": 6.311884057971015e-06,
+      "loss": 1.6704,
+      "step": 24120
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.875922679901123,
+      "learning_rate": 6.3060869565217394e-06,
+      "loss": 1.5384,
+      "step": 24130
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.232929229736328,
+      "learning_rate": 6.300289855072464e-06,
+      "loss": 1.6183,
+      "step": 24140
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.2293903827667236,
+      "learning_rate": 6.294492753623189e-06,
+      "loss": 1.4965,
+      "step": 24150
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 7.873164653778076,
+      "learning_rate": 6.288695652173914e-06,
+      "loss": 1.6817,
+      "step": 24160
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.767528057098389,
+      "learning_rate": 6.282898550724639e-06,
+      "loss": 1.7035,
+      "step": 24170
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.819685220718384,
+      "learning_rate": 6.277101449275362e-06,
+      "loss": 1.6476,
+      "step": 24180
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 8.485639572143555,
+      "learning_rate": 6.271304347826087e-06,
+      "loss": 1.6737,
+      "step": 24190
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.753501892089844,
+      "learning_rate": 6.265507246376812e-06,
+      "loss": 1.5851,
+      "step": 24200
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.462150812149048,
+      "learning_rate": 6.2597101449275366e-06,
+      "loss": 1.5416,
+      "step": 24210
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.648767948150635,
+      "learning_rate": 6.2539130434782614e-06,
+      "loss": 1.6021,
+      "step": 24220
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.275688648223877,
+      "learning_rate": 6.248115942028986e-06,
+      "loss": 1.6221,
+      "step": 24230
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.949949264526367,
+      "learning_rate": 6.242318840579711e-06,
+      "loss": 1.5643,
+      "step": 24240
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.8903957605361938,
+      "learning_rate": 6.236521739130436e-06,
+      "loss": 1.8374,
+      "step": 24250
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 12.835528373718262,
+      "learning_rate": 6.23072463768116e-06,
+      "loss": 1.7176,
+      "step": 24260
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.02567195892334,
+      "learning_rate": 6.224927536231885e-06,
+      "loss": 1.7012,
+      "step": 24270
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.2769625186920166,
+      "learning_rate": 6.219130434782609e-06,
+      "loss": 1.6408,
+      "step": 24280
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.010924339294434,
+      "learning_rate": 6.213333333333334e-06,
+      "loss": 1.7554,
+      "step": 24290
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.4019651412963867,
+      "learning_rate": 6.2075362318840586e-06,
+      "loss": 1.6405,
+      "step": 24300
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 12.412097930908203,
+      "learning_rate": 6.2017391304347826e-06,
+      "loss": 1.5947,
+      "step": 24310
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.544475555419922,
+      "learning_rate": 6.195942028985507e-06,
+      "loss": 1.6161,
+      "step": 24320
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.096646308898926,
+      "learning_rate": 6.190144927536232e-06,
+      "loss": 1.5265,
+      "step": 24330
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.934004783630371,
+      "learning_rate": 6.184347826086957e-06,
+      "loss": 1.6078,
+      "step": 24340
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.017855405807495,
+      "learning_rate": 6.178550724637682e-06,
+      "loss": 1.7315,
+      "step": 24350
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.79693603515625,
+      "learning_rate": 6.172753623188407e-06,
+      "loss": 1.6719,
+      "step": 24360
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.320878267288208,
+      "learning_rate": 6.166956521739132e-06,
+      "loss": 1.7665,
+      "step": 24370
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.4704517126083374,
+      "learning_rate": 6.161159420289855e-06,
+      "loss": 1.6487,
+      "step": 24380
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.9392638206481934,
+      "learning_rate": 6.15536231884058e-06,
+      "loss": 1.6703,
+      "step": 24390
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.3862284421920776,
+      "learning_rate": 6.1495652173913045e-06,
+      "loss": 1.5763,
+      "step": 24400
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0127170085906982,
+      "learning_rate": 6.143768115942029e-06,
+      "loss": 1.6736,
+      "step": 24410
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.7247025966644287,
+      "learning_rate": 6.137971014492754e-06,
+      "loss": 1.5715,
+      "step": 24420
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.8624944686889648,
+      "learning_rate": 6.132173913043479e-06,
+      "loss": 1.7204,
+      "step": 24430
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 9.080060005187988,
+      "learning_rate": 6.126376811594203e-06,
+      "loss": 1.6873,
+      "step": 24440
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.190478801727295,
+      "learning_rate": 6.120579710144928e-06,
+      "loss": 1.6602,
+      "step": 24450
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.000903606414795,
+      "learning_rate": 6.114782608695653e-06,
+      "loss": 1.6189,
+      "step": 24460
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.4447052478790283,
+      "learning_rate": 6.108985507246378e-06,
+      "loss": 1.6426,
+      "step": 24470
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.6354033946990967,
+      "learning_rate": 6.1031884057971025e-06,
+      "loss": 1.5703,
+      "step": 24480
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.4452903270721436,
+      "learning_rate": 6.0973913043478265e-06,
+      "loss": 1.6759,
+      "step": 24490
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.570955991744995,
+      "learning_rate": 6.0915942028985505e-06,
+      "loss": 1.7965,
+      "step": 24500
+    },
+    {
+      "epoch": 2.04,
+      "eval_loss": 1.6831989288330078,
+      "eval_runtime": 107.5466,
+      "eval_samples_per_second": 9.298,
+      "eval_steps_per_second": 2.325,
+      "step": 24500
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.475395917892456,
+      "learning_rate": 6.085797101449275e-06,
+      "loss": 1.5667,
+      "step": 24510
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.197578430175781,
+      "learning_rate": 6.08e-06,
+      "loss": 1.443,
+      "step": 24520
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.9446558952331543,
+      "learning_rate": 6.074202898550725e-06,
+      "loss": 1.6287,
+      "step": 24530
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.6334280967712402,
+      "learning_rate": 6.06840579710145e-06,
+      "loss": 1.6424,
+      "step": 24540
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.000389575958252,
+      "learning_rate": 6.062608695652175e-06,
+      "loss": 1.5955,
+      "step": 24550
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.637186050415039,
+      "learning_rate": 6.0568115942029e-06,
+      "loss": 1.5686,
+      "step": 24560
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.8098504543304443,
+      "learning_rate": 6.051014492753624e-06,
+      "loss": 1.6036,
+      "step": 24570
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.713962078094482,
+      "learning_rate": 6.0452173913043485e-06,
+      "loss": 1.7327,
+      "step": 24580
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.344076156616211,
+      "learning_rate": 6.0394202898550725e-06,
+      "loss": 1.7235,
+      "step": 24590
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.279098033905029,
+      "learning_rate": 6.0342028985507255e-06,
+      "loss": 1.7064,
+      "step": 24600
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.025627851486206,
+      "learning_rate": 6.0284057971014495e-06,
+      "loss": 1.6894,
+      "step": 24610
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 22.400253295898438,
+      "learning_rate": 6.022608695652174e-06,
+      "loss": 1.6231,
+      "step": 24620
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.277951955795288,
+      "learning_rate": 6.016811594202899e-06,
+      "loss": 1.6483,
+      "step": 24630
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.1129698753356934,
+      "learning_rate": 6.011014492753624e-06,
+      "loss": 1.6069,
+      "step": 24640
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.6595231294631958,
+      "learning_rate": 6.005217391304349e-06,
+      "loss": 1.5796,
+      "step": 24650
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.5862345695495605,
+      "learning_rate": 5.999420289855072e-06,
+      "loss": 1.6568,
+      "step": 24660
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.170677661895752,
+      "learning_rate": 5.993623188405797e-06,
+      "loss": 1.608,
+      "step": 24670
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.7852559089660645,
+      "learning_rate": 5.987826086956522e-06,
+      "loss": 1.6639,
+      "step": 24680
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.2292609214782715,
+      "learning_rate": 5.982028985507247e-06,
+      "loss": 1.7244,
+      "step": 24690
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.7636544704437256,
+      "learning_rate": 5.9762318840579715e-06,
+      "loss": 1.6089,
+      "step": 24700
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.0442698001861572,
+      "learning_rate": 5.970434782608696e-06,
+      "loss": 1.6019,
+      "step": 24710
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.134725093841553,
+      "learning_rate": 5.964637681159421e-06,
+      "loss": 1.5547,
+      "step": 24720
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.921815872192383,
+      "learning_rate": 5.958840579710146e-06,
+      "loss": 1.5884,
+      "step": 24730
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.0440871715545654,
+      "learning_rate": 5.95304347826087e-06,
+      "loss": 1.7173,
+      "step": 24740
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 9.600371360778809,
+      "learning_rate": 5.947246376811595e-06,
+      "loss": 1.6929,
+      "step": 24750
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.437139987945557,
+      "learning_rate": 5.94144927536232e-06,
+      "loss": 1.5867,
+      "step": 24760
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.7451493740081787,
+      "learning_rate": 5.935652173913044e-06,
+      "loss": 1.5753,
+      "step": 24770
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.03071928024292,
+      "learning_rate": 5.929855072463769e-06,
+      "loss": 1.7336,
+      "step": 24780
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.349041938781738,
+      "learning_rate": 5.924057971014493e-06,
+      "loss": 1.5073,
+      "step": 24790
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.7859678268432617,
+      "learning_rate": 5.9182608695652174e-06,
+      "loss": 1.5372,
+      "step": 24800
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.020723819732666,
+      "learning_rate": 5.912463768115942e-06,
+      "loss": 1.4775,
+      "step": 24810
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.465567111968994,
+      "learning_rate": 5.906666666666667e-06,
+      "loss": 1.6493,
+      "step": 24820
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.622159004211426,
+      "learning_rate": 5.900869565217392e-06,
+      "loss": 1.5543,
+      "step": 24830
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.285824775695801,
+      "learning_rate": 5.895072463768117e-06,
+      "loss": 1.5926,
+      "step": 24840
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 18.595064163208008,
+      "learning_rate": 5.889275362318842e-06,
+      "loss": 1.5055,
+      "step": 24850
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.7584556341171265,
+      "learning_rate": 5.8834782608695666e-06,
+      "loss": 1.6876,
+      "step": 24860
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 9.670427322387695,
+      "learning_rate": 5.87768115942029e-06,
+      "loss": 1.5141,
+      "step": 24870
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.9229793548583984,
+      "learning_rate": 5.8718840579710146e-06,
+      "loss": 1.5989,
+      "step": 24880
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.915713310241699,
+      "learning_rate": 5.866086956521739e-06,
+      "loss": 1.6928,
+      "step": 24890
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.840363025665283,
+      "learning_rate": 5.860289855072464e-06,
+      "loss": 1.6718,
+      "step": 24900
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 9.295182228088379,
+      "learning_rate": 5.854492753623189e-06,
+      "loss": 1.6016,
+      "step": 24910
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.412475109100342,
+      "learning_rate": 5.848695652173913e-06,
+      "loss": 1.6354,
+      "step": 24920
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 8.369576454162598,
+      "learning_rate": 5.842898550724638e-06,
+      "loss": 1.6169,
+      "step": 24930
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.795441746711731,
+      "learning_rate": 5.837101449275363e-06,
+      "loss": 1.6285,
+      "step": 24940
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.794939994812012,
+      "learning_rate": 5.831304347826088e-06,
+      "loss": 1.5579,
+      "step": 24950
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.6949840784072876,
+      "learning_rate": 5.8255072463768125e-06,
+      "loss": 1.6978,
+      "step": 24960
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.829167366027832,
+      "learning_rate": 5.8197101449275366e-06,
+      "loss": 1.6054,
+      "step": 24970
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.521337509155273,
+      "learning_rate": 5.8139130434782606e-06,
+      "loss": 1.6029,
+      "step": 24980
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 7.663822174072266,
+      "learning_rate": 5.808115942028985e-06,
+      "loss": 1.5795,
+      "step": 24990
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.8638572692871094,
+      "learning_rate": 5.80231884057971e-06,
+      "loss": 1.715,
+      "step": 25000
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 1.62440824508667,
+      "eval_runtime": 107.4912,
+      "eval_samples_per_second": 9.303,
+      "eval_steps_per_second": 2.326,
+      "step": 25000
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.584785223007202,
+      "learning_rate": 5.796521739130435e-06,
+      "loss": 1.6872,
+      "step": 25010
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.3847389221191406,
+      "learning_rate": 5.79072463768116e-06,
+      "loss": 1.6624,
+      "step": 25020
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.764808654785156,
+      "learning_rate": 5.784927536231885e-06,
+      "loss": 1.682,
+      "step": 25030
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.870251178741455,
+      "learning_rate": 5.77913043478261e-06,
+      "loss": 1.4977,
+      "step": 25040
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.905124187469482,
+      "learning_rate": 5.7733333333333345e-06,
+      "loss": 1.573,
+      "step": 25050
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 11.395575523376465,
+      "learning_rate": 5.7675362318840585e-06,
+      "loss": 1.6053,
+      "step": 25060
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.989179849624634,
+      "learning_rate": 5.761739130434783e-06,
+      "loss": 1.6168,
+      "step": 25070
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.6165571212768555,
+      "learning_rate": 5.755942028985507e-06,
+      "loss": 1.6171,
+      "step": 25080
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.178078651428223,
+      "learning_rate": 5.750144927536232e-06,
+      "loss": 1.6461,
+      "step": 25090
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.245847702026367,
+      "learning_rate": 5.744347826086957e-06,
+      "loss": 1.581,
+      "step": 25100
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 8.20195484161377,
+      "learning_rate": 5.738550724637681e-06,
+      "loss": 1.6275,
+      "step": 25110
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.7283331155776978,
+      "learning_rate": 5.732753623188406e-06,
+      "loss": 1.6411,
+      "step": 25120
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.998704195022583,
+      "learning_rate": 5.726956521739131e-06,
+      "loss": 1.6231,
+      "step": 25130
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.490156412124634,
+      "learning_rate": 5.721159420289856e-06,
+      "loss": 1.6043,
+      "step": 25140
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.0524210929870605,
+      "learning_rate": 5.7153623188405805e-06,
+      "loss": 1.7545,
+      "step": 25150
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 9.426039695739746,
+      "learning_rate": 5.709565217391305e-06,
+      "loss": 1.5325,
+      "step": 25160
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.069500207901001,
+      "learning_rate": 5.70376811594203e-06,
+      "loss": 1.6505,
+      "step": 25170
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.134079933166504,
+      "learning_rate": 5.697971014492753e-06,
+      "loss": 1.5572,
+      "step": 25180
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.0880463123321533,
+      "learning_rate": 5.692173913043478e-06,
+      "loss": 1.6467,
+      "step": 25190
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.196481227874756,
+      "learning_rate": 5.686376811594203e-06,
+      "loss": 1.5883,
+      "step": 25200
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.574937343597412,
+      "learning_rate": 5.680579710144928e-06,
+      "loss": 1.7104,
+      "step": 25210
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.061581611633301,
+      "learning_rate": 5.674782608695653e-06,
+      "loss": 1.6283,
+      "step": 25220
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.538878917694092,
+      "learning_rate": 5.668985507246378e-06,
+      "loss": 1.5797,
+      "step": 25230
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.407559394836426,
+      "learning_rate": 5.663188405797102e-06,
+      "loss": 1.6253,
+      "step": 25240
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.123929500579834,
+      "learning_rate": 5.6573913043478265e-06,
+      "loss": 1.7386,
+      "step": 25250
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 10.668411254882812,
+      "learning_rate": 5.651594202898551e-06,
+      "loss": 1.6129,
+      "step": 25260
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.001271724700928,
+      "learning_rate": 5.645797101449276e-06,
+      "loss": 1.6302,
+      "step": 25270
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 9.053343772888184,
+      "learning_rate": 5.64e-06,
+      "loss": 1.7988,
+      "step": 25280
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.4710533618927,
+      "learning_rate": 5.634202898550725e-06,
+      "loss": 1.6616,
+      "step": 25290
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.005622148513794,
+      "learning_rate": 5.628405797101449e-06,
+      "loss": 1.5469,
+      "step": 25300
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.569432258605957,
+      "learning_rate": 5.623188405797102e-06,
+      "loss": 1.5047,
+      "step": 25310
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.52546763420105,
+      "learning_rate": 5.617391304347827e-06,
+      "loss": 1.6252,
+      "step": 25320
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.454789400100708,
+      "learning_rate": 5.611594202898552e-06,
+      "loss": 1.6849,
+      "step": 25330
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.595395565032959,
+      "learning_rate": 5.605797101449277e-06,
+      "loss": 1.5866,
+      "step": 25340
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.6814996004104614,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.6539,
+      "step": 25350
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.3363173007965088,
+      "learning_rate": 5.594202898550725e-06,
+      "loss": 1.7953,
+      "step": 25360
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.2266364097595215,
+      "learning_rate": 5.5884057971014495e-06,
+      "loss": 1.459,
+      "step": 25370
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.90019702911377,
+      "learning_rate": 5.582608695652174e-06,
+      "loss": 1.5548,
+      "step": 25380
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.731317520141602,
+      "learning_rate": 5.576811594202899e-06,
+      "loss": 1.6977,
+      "step": 25390
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.7743234634399414,
+      "learning_rate": 5.571014492753624e-06,
+      "loss": 1.632,
+      "step": 25400
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.33520245552063,
+      "learning_rate": 5.565217391304348e-06,
+      "loss": 1.634,
+      "step": 25410
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 8.20616340637207,
+      "learning_rate": 5.559420289855073e-06,
+      "loss": 1.6429,
+      "step": 25420
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.349891662597656,
+      "learning_rate": 5.553623188405798e-06,
+      "loss": 1.5035,
+      "step": 25430
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.2954518795013428,
+      "learning_rate": 5.5478260869565226e-06,
+      "loss": 1.6812,
+      "step": 25440
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 7.884774684906006,
+      "learning_rate": 5.5420289855072474e-06,
+      "loss": 1.6622,
+      "step": 25450
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.3407180309295654,
+      "learning_rate": 5.536231884057971e-06,
+      "loss": 1.7026,
+      "step": 25460
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.300926685333252,
+      "learning_rate": 5.5304347826086954e-06,
+      "loss": 1.4354,
+      "step": 25470
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.3809895515441895,
+      "learning_rate": 5.52463768115942e-06,
+      "loss": 1.5977,
+      "step": 25480
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.672784447669983,
+      "learning_rate": 5.518840579710145e-06,
+      "loss": 1.69,
+      "step": 25490
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.11064875125885,
+      "learning_rate": 5.51304347826087e-06,
+      "loss": 1.4795,
+      "step": 25500
+    },
+    {
+      "epoch": 2.12,
+      "eval_loss": 1.6643449068069458,
+      "eval_runtime": 107.487,
+      "eval_samples_per_second": 9.303,
+      "eval_steps_per_second": 2.326,
+      "step": 25500
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4832613468170166,
+      "learning_rate": 5.507246376811595e-06,
+      "loss": 1.631,
+      "step": 25510
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.866186141967773,
+      "learning_rate": 5.50144927536232e-06,
+      "loss": 1.6299,
+      "step": 25520
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.978271961212158,
+      "learning_rate": 5.4956521739130446e-06,
+      "loss": 1.69,
+      "step": 25530
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.5575902462005615,
+      "learning_rate": 5.4898550724637686e-06,
+      "loss": 1.7588,
+      "step": 25540
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.088501214981079,
+      "learning_rate": 5.484057971014493e-06,
+      "loss": 1.5889,
+      "step": 25550
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.234321594238281,
+      "learning_rate": 5.478260869565217e-06,
+      "loss": 1.615,
+      "step": 25560
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.050870895385742,
+      "learning_rate": 5.472463768115942e-06,
+      "loss": 1.8112,
+      "step": 25570
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.9914329051971436,
+      "learning_rate": 5.466666666666667e-06,
+      "loss": 1.4678,
+      "step": 25580
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.666996479034424,
+      "learning_rate": 5.460869565217391e-06,
+      "loss": 1.6458,
+      "step": 25590
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.144287109375,
+      "learning_rate": 5.455072463768116e-06,
+      "loss": 1.6335,
+      "step": 25600
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.561944007873535,
+      "learning_rate": 5.449275362318841e-06,
+      "loss": 1.7007,
+      "step": 25610
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 8.11923599243164,
+      "learning_rate": 5.443478260869566e-06,
+      "loss": 1.5806,
+      "step": 25620
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.6465091705322266,
+      "learning_rate": 5.4376811594202905e-06,
+      "loss": 1.5382,
+      "step": 25630
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.644728183746338,
+      "learning_rate": 5.431884057971015e-06,
+      "loss": 1.6434,
+      "step": 25640
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.5903024673461914,
+      "learning_rate": 5.42608695652174e-06,
+      "loss": 1.7184,
+      "step": 25650
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.686862945556641,
+      "learning_rate": 5.420289855072465e-06,
+      "loss": 1.6501,
+      "step": 25660
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.454892158508301,
+      "learning_rate": 5.414492753623188e-06,
+      "loss": 1.4907,
+      "step": 25670
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.291064262390137,
+      "learning_rate": 5.408695652173913e-06,
+      "loss": 1.6432,
+      "step": 25680
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.9967716336250305,
+      "learning_rate": 5.402898550724638e-06,
+      "loss": 1.6613,
+      "step": 25690
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 10.572330474853516,
+      "learning_rate": 5.397101449275363e-06,
+      "loss": 1.6701,
+      "step": 25700
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.3549177646636963,
+      "learning_rate": 5.391304347826088e-06,
+      "loss": 1.5943,
+      "step": 25710
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.504870891571045,
+      "learning_rate": 5.3855072463768125e-06,
+      "loss": 1.5729,
+      "step": 25720
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.8512842655181885,
+      "learning_rate": 5.3797101449275365e-06,
+      "loss": 1.653,
+      "step": 25730
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.5188934803009033,
+      "learning_rate": 5.373913043478261e-06,
+      "loss": 1.6456,
+      "step": 25740
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.3979270458221436,
+      "learning_rate": 5.368115942028986e-06,
+      "loss": 1.7375,
+      "step": 25750
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.565988540649414,
+      "learning_rate": 5.362318840579711e-06,
+      "loss": 1.6154,
+      "step": 25760
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.995846271514893,
+      "learning_rate": 5.356521739130435e-06,
+      "loss": 1.6018,
+      "step": 25770
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.048307418823242,
+      "learning_rate": 5.350724637681159e-06,
+      "loss": 1.5733,
+      "step": 25780
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.6430954933166504,
+      "learning_rate": 5.344927536231884e-06,
+      "loss": 1.6623,
+      "step": 25790
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.271862983703613,
+      "learning_rate": 5.339130434782609e-06,
+      "loss": 1.7376,
+      "step": 25800
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.284571647644043,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 1.6058,
+      "step": 25810
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.8240787982940674,
+      "learning_rate": 5.3275362318840585e-06,
+      "loss": 1.6588,
+      "step": 25820
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.922074556350708,
+      "learning_rate": 5.321739130434783e-06,
+      "loss": 1.504,
+      "step": 25830
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.5163557529449463,
+      "learning_rate": 5.315942028985508e-06,
+      "loss": 1.7208,
+      "step": 25840
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.59926176071167,
+      "learning_rate": 5.310144927536233e-06,
+      "loss": 1.5188,
+      "step": 25850
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.922865867614746,
+      "learning_rate": 5.304347826086957e-06,
+      "loss": 1.7082,
+      "step": 25860
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.3443105220794678,
+      "learning_rate": 5.298550724637681e-06,
+      "loss": 1.5892,
+      "step": 25870
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.4057115316390991,
+      "learning_rate": 5.292753623188406e-06,
+      "loss": 1.5478,
+      "step": 25880
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.410768985748291,
+      "learning_rate": 5.286956521739131e-06,
+      "loss": 1.6926,
+      "step": 25890
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.0512008666992188,
+      "learning_rate": 5.281159420289856e-06,
+      "loss": 1.6573,
+      "step": 25900
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.070111274719238,
+      "learning_rate": 5.27536231884058e-06,
+      "loss": 1.6467,
+      "step": 25910
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 13.7697114944458,
+      "learning_rate": 5.2695652173913045e-06,
+      "loss": 1.4524,
+      "step": 25920
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 6.01429557800293,
+      "learning_rate": 5.263768115942029e-06,
+      "loss": 1.6422,
+      "step": 25930
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.9737797975540161,
+      "learning_rate": 5.257971014492754e-06,
+      "loss": 1.8127,
+      "step": 25940
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.113332271575928,
+      "learning_rate": 5.252173913043479e-06,
+      "loss": 1.6502,
+      "step": 25950
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.877610683441162,
+      "learning_rate": 5.246376811594204e-06,
+      "loss": 1.7218,
+      "step": 25960
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.620281219482422,
+      "learning_rate": 5.240579710144929e-06,
+      "loss": 1.7076,
+      "step": 25970
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.5956878662109375,
+      "learning_rate": 5.234782608695652e-06,
+      "loss": 1.5189,
+      "step": 25980
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.458510398864746,
+      "learning_rate": 5.228985507246377e-06,
+      "loss": 1.5011,
+      "step": 25990
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.703575372695923,
+      "learning_rate": 5.223188405797102e-06,
+      "loss": 1.6217,
+      "step": 26000
+    },
+    {
+      "epoch": 2.17,
+      "eval_loss": 1.641511082649231,
+      "eval_runtime": 107.5147,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 26000
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.613325595855713,
+      "learning_rate": 5.2173913043478265e-06,
+      "loss": 1.5841,
+      "step": 26010
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.493524551391602,
+      "learning_rate": 5.211594202898551e-06,
+      "loss": 1.6712,
+      "step": 26020
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.726990699768066,
+      "learning_rate": 5.205797101449276e-06,
+      "loss": 1.562,
+      "step": 26030
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.116297721862793,
+      "learning_rate": 5.2e-06,
+      "loss": 1.6151,
+      "step": 26040
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.906617641448975,
+      "learning_rate": 5.194202898550725e-06,
+      "loss": 1.8011,
+      "step": 26050
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.110503196716309,
+      "learning_rate": 5.18840579710145e-06,
+      "loss": 1.5389,
+      "step": 26060
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.075992584228516,
+      "learning_rate": 5.182608695652175e-06,
+      "loss": 1.6278,
+      "step": 26070
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.633796215057373,
+      "learning_rate": 5.176811594202899e-06,
+      "loss": 1.6312,
+      "step": 26080
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 8.259605407714844,
+      "learning_rate": 5.171014492753624e-06,
+      "loss": 1.6943,
+      "step": 26090
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.173551082611084,
+      "learning_rate": 5.165217391304348e-06,
+      "loss": 1.6731,
+      "step": 26100
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.1462539434432983,
+      "learning_rate": 5.1594202898550725e-06,
+      "loss": 1.6857,
+      "step": 26110
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.7535749673843384,
+      "learning_rate": 5.153623188405797e-06,
+      "loss": 1.6063,
+      "step": 26120
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7708317041397095,
+      "learning_rate": 5.147826086956522e-06,
+      "loss": 1.5231,
+      "step": 26130
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.939055919647217,
+      "learning_rate": 5.142028985507247e-06,
+      "loss": 1.5663,
+      "step": 26140
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.7642502784729004,
+      "learning_rate": 5.136231884057972e-06,
+      "loss": 1.4762,
+      "step": 26150
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.4374587535858154,
+      "learning_rate": 5.130434782608697e-06,
+      "loss": 1.5594,
+      "step": 26160
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.011373996734619,
+      "learning_rate": 5.124637681159422e-06,
+      "loss": 1.6385,
+      "step": 26170
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.366337299346924,
+      "learning_rate": 5.118840579710145e-06,
+      "loss": 1.6437,
+      "step": 26180
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.652083873748779,
+      "learning_rate": 5.11304347826087e-06,
+      "loss": 1.6124,
+      "step": 26190
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 7.6200690269470215,
+      "learning_rate": 5.1072463768115944e-06,
+      "loss": 1.6403,
+      "step": 26200
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.9343698620796204,
+      "learning_rate": 5.101449275362319e-06,
+      "loss": 1.5959,
+      "step": 26210
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.836380481719971,
+      "learning_rate": 5.095652173913044e-06,
+      "loss": 1.6459,
+      "step": 26220
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.8471882343292236,
+      "learning_rate": 5.089855072463768e-06,
+      "loss": 1.7146,
+      "step": 26230
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.5582016706466675,
+      "learning_rate": 5.084057971014493e-06,
+      "loss": 1.752,
+      "step": 26240
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.58816385269165,
+      "learning_rate": 5.078260869565218e-06,
+      "loss": 1.7379,
+      "step": 26250
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 12.132075309753418,
+      "learning_rate": 5.072463768115943e-06,
+      "loss": 1.7138,
+      "step": 26260
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 7.812301158905029,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 1.522,
+      "step": 26270
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 8.638223648071289,
+      "learning_rate": 5.060869565217392e-06,
+      "loss": 1.4687,
+      "step": 26280
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.7802847623825073,
+      "learning_rate": 5.055072463768116e-06,
+      "loss": 1.6814,
+      "step": 26290
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.118692398071289,
+      "learning_rate": 5.0492753623188404e-06,
+      "loss": 1.5805,
+      "step": 26300
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.812072515487671,
+      "learning_rate": 5.043478260869565e-06,
+      "loss": 1.5966,
+      "step": 26310
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.6763594150543213,
+      "learning_rate": 5.03768115942029e-06,
+      "loss": 1.5277,
+      "step": 26320
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.058526992797852,
+      "learning_rate": 5.031884057971015e-06,
+      "loss": 1.5608,
+      "step": 26330
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 6.4017653465271,
+      "learning_rate": 5.02608695652174e-06,
+      "loss": 1.702,
+      "step": 26340
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.419919490814209,
+      "learning_rate": 5.020289855072465e-06,
+      "loss": 1.7232,
+      "step": 26350
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.014277458190918,
+      "learning_rate": 5.014492753623189e-06,
+      "loss": 1.7648,
+      "step": 26360
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.1701581478118896,
+      "learning_rate": 5.0086956521739136e-06,
+      "loss": 1.3927,
+      "step": 26370
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.6772942543029785,
+      "learning_rate": 5.002898550724638e-06,
+      "loss": 1.5294,
+      "step": 26380
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.949331283569336,
+      "learning_rate": 4.997101449275362e-06,
+      "loss": 1.7033,
+      "step": 26390
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 7.254538536071777,
+      "learning_rate": 4.991304347826087e-06,
+      "loss": 1.6756,
+      "step": 26400
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.6317763328552246,
+      "learning_rate": 4.985507246376812e-06,
+      "loss": 1.5334,
+      "step": 26410
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.098857402801514,
+      "learning_rate": 4.979710144927536e-06,
+      "loss": 1.5211,
+      "step": 26420
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 10.028970718383789,
+      "learning_rate": 4.973913043478261e-06,
+      "loss": 1.6545,
+      "step": 26430
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.9619600772857666,
+      "learning_rate": 4.968115942028986e-06,
+      "loss": 1.5913,
+      "step": 26440
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.192411422729492,
+      "learning_rate": 4.962318840579711e-06,
+      "loss": 1.5367,
+      "step": 26450
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.78745698928833,
+      "learning_rate": 4.9565217391304355e-06,
+      "loss": 1.6372,
+      "step": 26460
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.531749963760376,
+      "learning_rate": 4.9507246376811595e-06,
+      "loss": 1.6329,
+      "step": 26470
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.566132068634033,
+      "learning_rate": 4.944927536231884e-06,
+      "loss": 1.6664,
+      "step": 26480
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 11.072998046875,
+      "learning_rate": 4.939130434782609e-06,
+      "loss": 1.6001,
+      "step": 26490
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 14.129816055297852,
+      "learning_rate": 4.933333333333334e-06,
+      "loss": 1.6051,
+      "step": 26500
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 1.6546690464019775,
+      "eval_runtime": 107.5057,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 26500
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.938653945922852,
+      "learning_rate": 4.927536231884059e-06,
+      "loss": 1.549,
+      "step": 26510
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.820674419403076,
+      "learning_rate": 4.921739130434783e-06,
+      "loss": 1.5996,
+      "step": 26520
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.9461727142333984,
+      "learning_rate": 4.915942028985508e-06,
+      "loss": 1.5242,
+      "step": 26530
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.117321729660034,
+      "learning_rate": 4.910144927536233e-06,
+      "loss": 1.6593,
+      "step": 26540
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.951183795928955,
+      "learning_rate": 4.904347826086957e-06,
+      "loss": 1.6066,
+      "step": 26550
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.954775333404541,
+      "learning_rate": 4.8985507246376815e-06,
+      "loss": 1.6093,
+      "step": 26560
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.857492208480835,
+      "learning_rate": 4.892753623188406e-06,
+      "loss": 1.6337,
+      "step": 26570
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 6.1911091804504395,
+      "learning_rate": 4.88695652173913e-06,
+      "loss": 1.6241,
+      "step": 26580
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.39035701751709,
+      "learning_rate": 4.881159420289855e-06,
+      "loss": 1.4798,
+      "step": 26590
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.794005870819092,
+      "learning_rate": 4.87536231884058e-06,
+      "loss": 1.6761,
+      "step": 26600
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.688593626022339,
+      "learning_rate": 4.869565217391305e-06,
+      "loss": 1.7455,
+      "step": 26610
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.9753758907318115,
+      "learning_rate": 4.863768115942029e-06,
+      "loss": 1.6043,
+      "step": 26620
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.76633882522583,
+      "learning_rate": 4.857971014492754e-06,
+      "loss": 1.5899,
+      "step": 26630
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 9.436808586120605,
+      "learning_rate": 4.852173913043479e-06,
+      "loss": 1.5456,
+      "step": 26640
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.1801443099975586,
+      "learning_rate": 4.8463768115942035e-06,
+      "loss": 1.4973,
+      "step": 26650
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.6295244693756104,
+      "learning_rate": 4.840579710144928e-06,
+      "loss": 1.6056,
+      "step": 26660
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.5441739559173584,
+      "learning_rate": 4.834782608695652e-06,
+      "loss": 1.5993,
+      "step": 26670
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.536238193511963,
+      "learning_rate": 4.828985507246377e-06,
+      "loss": 1.7003,
+      "step": 26680
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 8.867695808410645,
+      "learning_rate": 4.823188405797102e-06,
+      "loss": 1.5816,
+      "step": 26690
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.4776854515075684,
+      "learning_rate": 4.817391304347827e-06,
+      "loss": 1.6089,
+      "step": 26700
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.4541239738464355,
+      "learning_rate": 4.811594202898551e-06,
+      "loss": 1.5486,
+      "step": 26710
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.7615251541137695,
+      "learning_rate": 4.805797101449276e-06,
+      "loss": 1.4767,
+      "step": 26720
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.553129196166992,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.6036,
+      "step": 26730
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.3022050857543945,
+      "learning_rate": 4.794202898550725e-06,
+      "loss": 1.5677,
+      "step": 26740
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.267906427383423,
+      "learning_rate": 4.7884057971014495e-06,
+      "loss": 1.5816,
+      "step": 26750
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.8131909370422363,
+      "learning_rate": 4.782608695652174e-06,
+      "loss": 1.7683,
+      "step": 26760
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.468419551849365,
+      "learning_rate": 4.776811594202899e-06,
+      "loss": 1.4521,
+      "step": 26770
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.726469039916992,
+      "learning_rate": 4.771014492753623e-06,
+      "loss": 1.6083,
+      "step": 26780
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.375027894973755,
+      "learning_rate": 4.765217391304348e-06,
+      "loss": 1.6481,
+      "step": 26790
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 8.728473663330078,
+      "learning_rate": 4.759420289855073e-06,
+      "loss": 1.6519,
+      "step": 26800
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 13.2547607421875,
+      "learning_rate": 4.753623188405798e-06,
+      "loss": 1.646,
+      "step": 26810
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.739991664886475,
+      "learning_rate": 4.747826086956523e-06,
+      "loss": 1.6067,
+      "step": 26820
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.9073076248168945,
+      "learning_rate": 4.742028985507247e-06,
+      "loss": 1.4971,
+      "step": 26830
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 8.832313537597656,
+      "learning_rate": 4.7362318840579715e-06,
+      "loss": 1.6724,
+      "step": 26840
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.5870254039764404,
+      "learning_rate": 4.730434782608696e-06,
+      "loss": 1.7331,
+      "step": 26850
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.8640055656433105,
+      "learning_rate": 4.724637681159421e-06,
+      "loss": 1.5264,
+      "step": 26860
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.005427360534668,
+      "learning_rate": 4.718840579710145e-06,
+      "loss": 1.5717,
+      "step": 26870
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.7698235511779785,
+      "learning_rate": 4.71304347826087e-06,
+      "loss": 1.7427,
+      "step": 26880
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.162291049957275,
+      "learning_rate": 4.707246376811595e-06,
+      "loss": 1.7923,
+      "step": 26890
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.3494528532028198,
+      "learning_rate": 4.701449275362319e-06,
+      "loss": 1.7313,
+      "step": 26900
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.9067726135253906,
+      "learning_rate": 4.695652173913044e-06,
+      "loss": 1.7465,
+      "step": 26910
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.578123092651367,
+      "learning_rate": 4.689855072463769e-06,
+      "loss": 1.5809,
+      "step": 26920
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.2738120555877686,
+      "learning_rate": 4.684057971014493e-06,
+      "loss": 1.662,
+      "step": 26930
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.7239420413970947,
+      "learning_rate": 4.6782608695652175e-06,
+      "loss": 1.5232,
+      "step": 26940
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.876070022583008,
+      "learning_rate": 4.672463768115942e-06,
+      "loss": 1.7607,
+      "step": 26950
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.976984977722168,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.6693,
+      "step": 26960
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.821426868438721,
+      "learning_rate": 4.660869565217392e-06,
+      "loss": 1.6281,
+      "step": 26970
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.057045936584473,
+      "learning_rate": 4.655072463768116e-06,
+      "loss": 1.6141,
+      "step": 26980
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.272705554962158,
+      "learning_rate": 4.649275362318841e-06,
+      "loss": 1.6509,
+      "step": 26990
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.8042986392974854,
+      "learning_rate": 4.643478260869566e-06,
+      "loss": 1.6283,
+      "step": 27000
+    },
+    {
+      "epoch": 2.25,
+      "eval_loss": 1.599829912185669,
+      "eval_runtime": 107.518,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 27000
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.672848701477051,
+      "learning_rate": 4.637681159420291e-06,
+      "loss": 1.587,
+      "step": 27010
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.5648651123046875,
+      "learning_rate": 4.6318840579710154e-06,
+      "loss": 1.6535,
+      "step": 27020
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.619153022766113,
+      "learning_rate": 4.6260869565217394e-06,
+      "loss": 1.7095,
+      "step": 27030
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.661830425262451,
+      "learning_rate": 4.620289855072464e-06,
+      "loss": 1.5421,
+      "step": 27040
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.601088285446167,
+      "learning_rate": 4.614492753623188e-06,
+      "loss": 1.6169,
+      "step": 27050
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.6670879125595093,
+      "learning_rate": 4.608695652173913e-06,
+      "loss": 1.7808,
+      "step": 27060
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.285208702087402,
+      "learning_rate": 4.602898550724638e-06,
+      "loss": 1.6018,
+      "step": 27070
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 13.955137252807617,
+      "learning_rate": 4.597101449275363e-06,
+      "loss": 1.6071,
+      "step": 27080
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.246000289916992,
+      "learning_rate": 4.591304347826087e-06,
+      "loss": 1.74,
+      "step": 27090
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.297258377075195,
+      "learning_rate": 4.585507246376812e-06,
+      "loss": 1.6233,
+      "step": 27100
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 6.08992338180542,
+      "learning_rate": 4.5797101449275366e-06,
+      "loss": 1.6511,
+      "step": 27110
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.511499404907227,
+      "learning_rate": 4.573913043478261e-06,
+      "loss": 1.5344,
+      "step": 27120
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.989239454269409,
+      "learning_rate": 4.568115942028986e-06,
+      "loss": 1.6172,
+      "step": 27130
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.358445882797241,
+      "learning_rate": 4.56231884057971e-06,
+      "loss": 1.6004,
+      "step": 27140
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.533210039138794,
+      "learning_rate": 4.556521739130435e-06,
+      "loss": 1.6651,
+      "step": 27150
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.6375526189804077,
+      "learning_rate": 4.55072463768116e-06,
+      "loss": 1.603,
+      "step": 27160
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 16.163049697875977,
+      "learning_rate": 4.544927536231885e-06,
+      "loss": 1.7045,
+      "step": 27170
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.7427773475646973,
+      "learning_rate": 4.53913043478261e-06,
+      "loss": 1.7164,
+      "step": 27180
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.3884432315826416,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 1.5997,
+      "step": 27190
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 11.96393871307373,
+      "learning_rate": 4.5275362318840585e-06,
+      "loss": 1.6458,
+      "step": 27200
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.4181969165802,
+      "learning_rate": 4.5217391304347826e-06,
+      "loss": 1.5166,
+      "step": 27210
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.070342540740967,
+      "learning_rate": 4.515942028985507e-06,
+      "loss": 1.5037,
+      "step": 27220
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.330051898956299,
+      "learning_rate": 4.510144927536232e-06,
+      "loss": 1.6322,
+      "step": 27230
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.178266525268555,
+      "learning_rate": 4.504347826086956e-06,
+      "loss": 1.5712,
+      "step": 27240
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 10.042227745056152,
+      "learning_rate": 4.498550724637681e-06,
+      "loss": 1.5195,
+      "step": 27250
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.68449592590332,
+      "learning_rate": 4.492753623188406e-06,
+      "loss": 1.5767,
+      "step": 27260
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.098611354827881,
+      "learning_rate": 4.486956521739131e-06,
+      "loss": 1.5826,
+      "step": 27270
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.7514631748199463,
+      "learning_rate": 4.481159420289856e-06,
+      "loss": 1.6519,
+      "step": 27280
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 6.086874961853027,
+      "learning_rate": 4.47536231884058e-06,
+      "loss": 1.6356,
+      "step": 27290
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.969055414199829,
+      "learning_rate": 4.4695652173913045e-06,
+      "loss": 1.5372,
+      "step": 27300
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.0818963050842285,
+      "learning_rate": 4.463768115942029e-06,
+      "loss": 1.6195,
+      "step": 27310
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.355165481567383,
+      "learning_rate": 4.457971014492754e-06,
+      "loss": 1.5498,
+      "step": 27320
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.69412899017334,
+      "learning_rate": 4.452173913043479e-06,
+      "loss": 1.6927,
+      "step": 27330
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.266238212585449,
+      "learning_rate": 4.446376811594204e-06,
+      "loss": 1.5444,
+      "step": 27340
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 10.938898086547852,
+      "learning_rate": 4.440579710144928e-06,
+      "loss": 1.5767,
+      "step": 27350
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.8983266353607178,
+      "learning_rate": 4.434782608695653e-06,
+      "loss": 1.6385,
+      "step": 27360
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.039658069610596,
+      "learning_rate": 4.428985507246377e-06,
+      "loss": 1.6902,
+      "step": 27370
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.6886088848114014,
+      "learning_rate": 4.423188405797102e-06,
+      "loss": 1.5464,
+      "step": 27380
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.805041551589966,
+      "learning_rate": 4.4173913043478265e-06,
+      "loss": 1.5516,
+      "step": 27390
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.767727851867676,
+      "learning_rate": 4.4115942028985505e-06,
+      "loss": 1.6857,
+      "step": 27400
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.386350631713867,
+      "learning_rate": 4.405797101449275e-06,
+      "loss": 1.7113,
+      "step": 27410
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.8526190519332886,
+      "learning_rate": 4.4e-06,
+      "loss": 1.6934,
+      "step": 27420
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.962111949920654,
+      "learning_rate": 4.394202898550725e-06,
+      "loss": 1.5782,
+      "step": 27430
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.9486684799194336,
+      "learning_rate": 4.38840579710145e-06,
+      "loss": 1.5233,
+      "step": 27440
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.742575645446777,
+      "learning_rate": 4.382608695652174e-06,
+      "loss": 1.5672,
+      "step": 27450
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 7.8938422203063965,
+      "learning_rate": 4.376811594202899e-06,
+      "loss": 1.5662,
+      "step": 27460
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.794543743133545,
+      "learning_rate": 4.371014492753624e-06,
+      "loss": 1.6878,
+      "step": 27470
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.3197507858276367,
+      "learning_rate": 4.3652173913043485e-06,
+      "loss": 1.6168,
+      "step": 27480
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.5373730659484863,
+      "learning_rate": 4.359420289855073e-06,
+      "loss": 1.6601,
+      "step": 27490
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5637263059616089,
+      "learning_rate": 4.353623188405797e-06,
+      "loss": 1.6436,
+      "step": 27500
+    },
+    {
+      "epoch": 2.29,
+      "eval_loss": 1.637457251548767,
+      "eval_runtime": 107.504,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 27500
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.249775409698486,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 1.5759,
+      "step": 27510
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 9.582161903381348,
+      "learning_rate": 4.342028985507247e-06,
+      "loss": 1.5566,
+      "step": 27520
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.517526626586914,
+      "learning_rate": 4.336231884057971e-06,
+      "loss": 1.5944,
+      "step": 27530
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.537724494934082,
+      "learning_rate": 4.330434782608696e-06,
+      "loss": 1.5976,
+      "step": 27540
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.978813886642456,
+      "learning_rate": 4.324637681159421e-06,
+      "loss": 1.6186,
+      "step": 27550
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.7505924701690674,
+      "learning_rate": 4.318840579710145e-06,
+      "loss": 1.5513,
+      "step": 27560
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.5586419105529785,
+      "learning_rate": 4.31304347826087e-06,
+      "loss": 1.7299,
+      "step": 27570
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.848520278930664,
+      "learning_rate": 4.3072463768115945e-06,
+      "loss": 1.6996,
+      "step": 27580
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 10.454072952270508,
+      "learning_rate": 4.301449275362319e-06,
+      "loss": 1.748,
+      "step": 27590
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.605503559112549,
+      "learning_rate": 4.295652173913043e-06,
+      "loss": 1.6378,
+      "step": 27600
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.994022369384766,
+      "learning_rate": 4.289855072463768e-06,
+      "loss": 1.5866,
+      "step": 27610
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.65685772895813,
+      "learning_rate": 4.284057971014493e-06,
+      "loss": 1.5671,
+      "step": 27620
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 8.096505165100098,
+      "learning_rate": 4.278260869565218e-06,
+      "loss": 1.5329,
+      "step": 27630
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.837905406951904,
+      "learning_rate": 4.272463768115943e-06,
+      "loss": 1.7046,
+      "step": 27640
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.230941295623779,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 1.6219,
+      "step": 27650
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.498970031738281,
+      "learning_rate": 4.260869565217392e-06,
+      "loss": 1.5791,
+      "step": 27660
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.494067430496216,
+      "learning_rate": 4.2550724637681165e-06,
+      "loss": 1.5099,
+      "step": 27670
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.1975343227386475,
+      "learning_rate": 4.249275362318841e-06,
+      "loss": 1.5382,
+      "step": 27680
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.870430946350098,
+      "learning_rate": 4.243478260869565e-06,
+      "loss": 1.4734,
+      "step": 27690
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.4920201301574707,
+      "learning_rate": 4.23768115942029e-06,
+      "loss": 1.5764,
+      "step": 27700
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.826338052749634,
+      "learning_rate": 4.231884057971015e-06,
+      "loss": 1.766,
+      "step": 27710
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 9.284537315368652,
+      "learning_rate": 4.226086956521739e-06,
+      "loss": 1.5152,
+      "step": 27720
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.703794479370117,
+      "learning_rate": 4.220289855072464e-06,
+      "loss": 1.5506,
+      "step": 27730
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.4055094718933105,
+      "learning_rate": 4.214492753623189e-06,
+      "loss": 1.6593,
+      "step": 27740
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.2604644298553467,
+      "learning_rate": 4.208695652173914e-06,
+      "loss": 1.5203,
+      "step": 27750
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.278348445892334,
+      "learning_rate": 4.202898550724638e-06,
+      "loss": 1.6086,
+      "step": 27760
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 8.626958847045898,
+      "learning_rate": 4.1971014492753624e-06,
+      "loss": 1.569,
+      "step": 27770
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.3467583656311035,
+      "learning_rate": 4.191304347826087e-06,
+      "loss": 1.5253,
+      "step": 27780
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.055905818939209,
+      "learning_rate": 4.185507246376812e-06,
+      "loss": 1.6537,
+      "step": 27790
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.519817352294922,
+      "learning_rate": 4.179710144927537e-06,
+      "loss": 1.5329,
+      "step": 27800
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.1581406593322754,
+      "learning_rate": 4.173913043478261e-06,
+      "loss": 1.616,
+      "step": 27810
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.114212155342102,
+      "learning_rate": 4.168115942028986e-06,
+      "loss": 1.4248,
+      "step": 27820
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.742372989654541,
+      "learning_rate": 4.162318840579711e-06,
+      "loss": 1.7195,
+      "step": 27830
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.707331895828247,
+      "learning_rate": 4.1565217391304356e-06,
+      "loss": 1.5344,
+      "step": 27840
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.827779293060303,
+      "learning_rate": 4.1507246376811596e-06,
+      "loss": 1.6523,
+      "step": 27850
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.977830410003662,
+      "learning_rate": 4.1449275362318844e-06,
+      "loss": 1.6839,
+      "step": 27860
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 8.55379867553711,
+      "learning_rate": 4.139130434782609e-06,
+      "loss": 1.6007,
+      "step": 27870
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.7034173011779785,
+      "learning_rate": 4.133333333333333e-06,
+      "loss": 1.5704,
+      "step": 27880
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.772855520248413,
+      "learning_rate": 4.127536231884058e-06,
+      "loss": 1.7209,
+      "step": 27890
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.952080011367798,
+      "learning_rate": 4.121739130434783e-06,
+      "loss": 1.5282,
+      "step": 27900
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.7901113033294678,
+      "learning_rate": 4.115942028985507e-06,
+      "loss": 1.6545,
+      "step": 27910
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.063865661621094,
+      "learning_rate": 4.110144927536232e-06,
+      "loss": 1.4981,
+      "step": 27920
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.868477821350098,
+      "learning_rate": 4.104347826086957e-06,
+      "loss": 1.6642,
+      "step": 27930
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.399648666381836,
+      "learning_rate": 4.0985507246376816e-06,
+      "loss": 1.6422,
+      "step": 27940
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.0602734088897705,
+      "learning_rate": 4.092753623188406e-06,
+      "loss": 1.7643,
+      "step": 27950
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.221512794494629,
+      "learning_rate": 4.086956521739131e-06,
+      "loss": 1.5439,
+      "step": 27960
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.30639386177063,
+      "learning_rate": 4.081159420289855e-06,
+      "loss": 1.5125,
+      "step": 27970
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.630527973175049,
+      "learning_rate": 4.07536231884058e-06,
+      "loss": 1.5665,
+      "step": 27980
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 9.667155265808105,
+      "learning_rate": 4.069565217391305e-06,
+      "loss": 1.6917,
+      "step": 27990
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.4528236389160156,
+      "learning_rate": 4.06376811594203e-06,
+      "loss": 1.6995,
+      "step": 28000
+    },
+    {
+      "epoch": 2.33,
+      "eval_loss": 1.644856333732605,
+      "eval_runtime": 107.5008,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 28000
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.805037260055542,
+      "learning_rate": 4.057971014492754e-06,
+      "loss": 1.5207,
+      "step": 28010
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.819047451019287,
+      "learning_rate": 4.052173913043479e-06,
+      "loss": 1.6518,
+      "step": 28020
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.75721263885498,
+      "learning_rate": 4.0463768115942035e-06,
+      "loss": 1.6758,
+      "step": 28030
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.432452201843262,
+      "learning_rate": 4.0405797101449275e-06,
+      "loss": 1.7066,
+      "step": 28040
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.6831490993499756,
+      "learning_rate": 4.034782608695652e-06,
+      "loss": 1.6379,
+      "step": 28050
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.150071620941162,
+      "learning_rate": 4.028985507246377e-06,
+      "loss": 1.5458,
+      "step": 28060
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.020603179931641,
+      "learning_rate": 4.023188405797101e-06,
+      "loss": 1.6909,
+      "step": 28070
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.10090446472168,
+      "learning_rate": 4.017391304347826e-06,
+      "loss": 1.7057,
+      "step": 28080
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.3139498233795166,
+      "learning_rate": 4.011594202898551e-06,
+      "loss": 1.6334,
+      "step": 28090
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 11.909808158874512,
+      "learning_rate": 4.005797101449276e-06,
+      "loss": 1.6384,
+      "step": 28100
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.950678586959839,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.6356,
+      "step": 28110
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.596022844314575,
+      "learning_rate": 3.994202898550725e-06,
+      "loss": 1.4085,
+      "step": 28120
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 9.143681526184082,
+      "learning_rate": 3.9884057971014495e-06,
+      "loss": 1.5866,
+      "step": 28130
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.4474549293518066,
+      "learning_rate": 3.982608695652174e-06,
+      "loss": 1.5971,
+      "step": 28140
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.5778965950012207,
+      "learning_rate": 3.976811594202899e-06,
+      "loss": 1.5654,
+      "step": 28150
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.3869068622589111,
+      "learning_rate": 3.971014492753624e-06,
+      "loss": 1.6479,
+      "step": 28160
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 6.334060192108154,
+      "learning_rate": 3.965217391304348e-06,
+      "loss": 1.4869,
+      "step": 28170
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.896120071411133,
+      "learning_rate": 3.959420289855073e-06,
+      "loss": 1.5473,
+      "step": 28180
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.137752532958984,
+      "learning_rate": 3.953623188405798e-06,
+      "loss": 1.5977,
+      "step": 28190
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.7057723999023438,
+      "learning_rate": 3.947826086956522e-06,
+      "loss": 1.6333,
+      "step": 28200
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 7.279668807983398,
+      "learning_rate": 3.942028985507247e-06,
+      "loss": 1.6963,
+      "step": 28210
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.471526145935059,
+      "learning_rate": 3.936231884057971e-06,
+      "loss": 1.6287,
+      "step": 28220
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.1095166206359863,
+      "learning_rate": 3.9304347826086955e-06,
+      "loss": 1.5774,
+      "step": 28230
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.295337677001953,
+      "learning_rate": 3.92463768115942e-06,
+      "loss": 1.6255,
+      "step": 28240
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.934601306915283,
+      "learning_rate": 3.918840579710145e-06,
+      "loss": 1.6194,
+      "step": 28250
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.069929361343384,
+      "learning_rate": 3.91304347826087e-06,
+      "loss": 1.7143,
+      "step": 28260
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.211582660675049,
+      "learning_rate": 3.907246376811595e-06,
+      "loss": 1.5541,
+      "step": 28270
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.216347694396973,
+      "learning_rate": 3.901449275362319e-06,
+      "loss": 1.5638,
+      "step": 28280
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.743631362915039,
+      "learning_rate": 3.895652173913044e-06,
+      "loss": 1.5687,
+      "step": 28290
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.041175842285156,
+      "learning_rate": 3.889855072463769e-06,
+      "loss": 1.5939,
+      "step": 28300
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.4274263381958,
+      "learning_rate": 3.8840579710144935e-06,
+      "loss": 1.7864,
+      "step": 28310
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.043384552001953,
+      "learning_rate": 3.878260869565218e-06,
+      "loss": 1.5686,
+      "step": 28320
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.899559736251831,
+      "learning_rate": 3.872463768115942e-06,
+      "loss": 1.7651,
+      "step": 28330
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 6.566895008087158,
+      "learning_rate": 3.866666666666667e-06,
+      "loss": 1.6548,
+      "step": 28340
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.242386341094971,
+      "learning_rate": 3.860869565217392e-06,
+      "loss": 1.5816,
+      "step": 28350
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.0503904819488525,
+      "learning_rate": 3.855072463768116e-06,
+      "loss": 1.7181,
+      "step": 28360
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 8.445423126220703,
+      "learning_rate": 3.849275362318841e-06,
+      "loss": 1.6133,
+      "step": 28370
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.893218040466309,
+      "learning_rate": 3.843478260869565e-06,
+      "loss": 1.6204,
+      "step": 28380
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.2958295345306396,
+      "learning_rate": 3.83768115942029e-06,
+      "loss": 1.524,
+      "step": 28390
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.6230056285858154,
+      "learning_rate": 3.831884057971015e-06,
+      "loss": 1.7141,
+      "step": 28400
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.484030246734619,
+      "learning_rate": 3.8260869565217395e-06,
+      "loss": 1.7142,
+      "step": 28410
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.462039470672607,
+      "learning_rate": 3.820289855072464e-06,
+      "loss": 1.6588,
+      "step": 28420
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.6934198141098022,
+      "learning_rate": 3.8144927536231883e-06,
+      "loss": 1.6298,
+      "step": 28430
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.632734298706055,
+      "learning_rate": 3.808695652173913e-06,
+      "loss": 1.523,
+      "step": 28440
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.3441051244735718,
+      "learning_rate": 3.802898550724638e-06,
+      "loss": 1.6716,
+      "step": 28450
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 14.253397941589355,
+      "learning_rate": 3.797101449275363e-06,
+      "loss": 1.6655,
+      "step": 28460
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.6277294158935547,
+      "learning_rate": 3.7913043478260873e-06,
+      "loss": 1.6434,
+      "step": 28470
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.960604429244995,
+      "learning_rate": 3.7855072463768117e-06,
+      "loss": 1.6302,
+      "step": 28480
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 8.853973388671875,
+      "learning_rate": 3.7797101449275366e-06,
+      "loss": 1.5669,
+      "step": 28490
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.044206619262695,
+      "learning_rate": 3.773913043478261e-06,
+      "loss": 1.6932,
+      "step": 28500
+    },
+    {
+      "epoch": 2.38,
+      "eval_loss": 1.6286858320236206,
+      "eval_runtime": 107.5089,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.325,
+      "step": 28500
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.0954174995422363,
+      "learning_rate": 3.768115942028986e-06,
+      "loss": 1.7568,
+      "step": 28510
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 7.709628105163574,
+      "learning_rate": 3.7623188405797107e-06,
+      "loss": 1.4997,
+      "step": 28520
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.7862284183502197,
+      "learning_rate": 3.7565217391304347e-06,
+      "loss": 1.6189,
+      "step": 28530
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.443877696990967,
+      "learning_rate": 3.7507246376811596e-06,
+      "loss": 1.6114,
+      "step": 28540
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.974430084228516,
+      "learning_rate": 3.7449275362318844e-06,
+      "loss": 1.7863,
+      "step": 28550
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.032052516937256,
+      "learning_rate": 3.739130434782609e-06,
+      "loss": 1.5454,
+      "step": 28560
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 15.427571296691895,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 1.583,
+      "step": 28570
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.56199836730957,
+      "learning_rate": 3.727536231884058e-06,
+      "loss": 1.5224,
+      "step": 28580
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.051842212677002,
+      "learning_rate": 3.7217391304347826e-06,
+      "loss": 1.4863,
+      "step": 28590
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.057842493057251,
+      "learning_rate": 3.7159420289855074e-06,
+      "loss": 1.6266,
+      "step": 28600
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.723554849624634,
+      "learning_rate": 3.7101449275362323e-06,
+      "loss": 1.4752,
+      "step": 28610
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.5356404781341553,
+      "learning_rate": 3.704347826086957e-06,
+      "loss": 1.6376,
+      "step": 28620
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.806687593460083,
+      "learning_rate": 3.6985507246376816e-06,
+      "loss": 1.5874,
+      "step": 28630
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.9221794605255127,
+      "learning_rate": 3.692753623188406e-06,
+      "loss": 1.5031,
+      "step": 28640
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.143265724182129,
+      "learning_rate": 3.686956521739131e-06,
+      "loss": 1.5727,
+      "step": 28650
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.832207679748535,
+      "learning_rate": 3.6811594202898553e-06,
+      "loss": 1.6113,
+      "step": 28660
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.17288875579834,
+      "learning_rate": 3.67536231884058e-06,
+      "loss": 1.593,
+      "step": 28670
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.8944125175476074,
+      "learning_rate": 3.669565217391305e-06,
+      "loss": 1.6084,
+      "step": 28680
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 9.046699523925781,
+      "learning_rate": 3.663768115942029e-06,
+      "loss": 1.5391,
+      "step": 28690
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.482266902923584,
+      "learning_rate": 3.657971014492754e-06,
+      "loss": 1.6621,
+      "step": 28700
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.573151588439941,
+      "learning_rate": 3.6521739130434787e-06,
+      "loss": 1.6467,
+      "step": 28710
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 13.082232475280762,
+      "learning_rate": 3.646376811594203e-06,
+      "loss": 1.6243,
+      "step": 28720
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.799623012542725,
+      "learning_rate": 3.640579710144928e-06,
+      "loss": 1.5777,
+      "step": 28730
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.449925422668457,
+      "learning_rate": 3.6347826086956524e-06,
+      "loss": 1.7436,
+      "step": 28740
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.415322780609131,
+      "learning_rate": 3.628985507246377e-06,
+      "loss": 1.5015,
+      "step": 28750
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.3753433227539062,
+      "learning_rate": 3.6231884057971017e-06,
+      "loss": 1.696,
+      "step": 28760
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.1241776943206787,
+      "learning_rate": 3.6173913043478265e-06,
+      "loss": 1.5431,
+      "step": 28770
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.052039623260498,
+      "learning_rate": 3.6115942028985514e-06,
+      "loss": 1.5878,
+      "step": 28780
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.090257167816162,
+      "learning_rate": 3.6057971014492754e-06,
+      "loss": 1.5928,
+      "step": 28790
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.91934871673584,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.6831,
+      "step": 28800
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 8.889466285705566,
+      "learning_rate": 3.594202898550725e-06,
+      "loss": 1.599,
+      "step": 28810
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.158334255218506,
+      "learning_rate": 3.5884057971014495e-06,
+      "loss": 1.6809,
+      "step": 28820
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.268106460571289,
+      "learning_rate": 3.5826086956521744e-06,
+      "loss": 1.6967,
+      "step": 28830
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.080434322357178,
+      "learning_rate": 3.5768115942028984e-06,
+      "loss": 1.7103,
+      "step": 28840
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.549793243408203,
+      "learning_rate": 3.5710144927536233e-06,
+      "loss": 1.5933,
+      "step": 28850
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.401309013366699,
+      "learning_rate": 3.565217391304348e-06,
+      "loss": 1.6199,
+      "step": 28860
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.3040978908538818,
+      "learning_rate": 3.559420289855073e-06,
+      "loss": 1.7081,
+      "step": 28870
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.473288059234619,
+      "learning_rate": 3.5536231884057974e-06,
+      "loss": 1.626,
+      "step": 28880
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.599831461906433,
+      "learning_rate": 3.547826086956522e-06,
+      "loss": 1.7197,
+      "step": 28890
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.284329414367676,
+      "learning_rate": 3.5420289855072467e-06,
+      "loss": 1.6332,
+      "step": 28900
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 8.579666137695312,
+      "learning_rate": 3.536231884057971e-06,
+      "loss": 1.7461,
+      "step": 28910
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.893390953540802,
+      "learning_rate": 3.530434782608696e-06,
+      "loss": 1.588,
+      "step": 28920
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 6.217680931091309,
+      "learning_rate": 3.524637681159421e-06,
+      "loss": 1.6019,
+      "step": 28930
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.5240437984466553,
+      "learning_rate": 3.5188405797101457e-06,
+      "loss": 1.5526,
+      "step": 28940
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.2045577764511108,
+      "learning_rate": 3.5130434782608697e-06,
+      "loss": 1.5679,
+      "step": 28950
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.309528350830078,
+      "learning_rate": 3.5072463768115945e-06,
+      "loss": 1.6655,
+      "step": 28960
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 7.580850601196289,
+      "learning_rate": 3.501449275362319e-06,
+      "loss": 1.6961,
+      "step": 28970
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.6850571632385254,
+      "learning_rate": 3.495652173913044e-06,
+      "loss": 1.6288,
+      "step": 28980
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.09687614440918,
+      "learning_rate": 3.4898550724637687e-06,
+      "loss": 1.7291,
+      "step": 28990
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.748039960861206,
+      "learning_rate": 3.4840579710144927e-06,
+      "loss": 1.5591,
+      "step": 29000
+    },
+    {
+      "epoch": 2.42,
+      "eval_loss": 1.657044529914856,
+      "eval_runtime": 107.5181,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 29000
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.007520914077759,
+      "learning_rate": 3.4782608695652175e-06,
+      "loss": 1.6379,
+      "step": 29010
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.748697280883789,
+      "learning_rate": 3.4724637681159424e-06,
+      "loss": 1.7399,
+      "step": 29020
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 15.37971305847168,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 1.5297,
+      "step": 29030
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.962806463241577,
+      "learning_rate": 3.4608695652173916e-06,
+      "loss": 1.6266,
+      "step": 29040
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 13.86978530883789,
+      "learning_rate": 3.455072463768116e-06,
+      "loss": 1.6452,
+      "step": 29050
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.6884708404541016,
+      "learning_rate": 3.449275362318841e-06,
+      "loss": 1.63,
+      "step": 29060
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.2268710136413574,
+      "learning_rate": 3.4434782608695654e-06,
+      "loss": 1.7602,
+      "step": 29070
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 8.4721040725708,
+      "learning_rate": 3.43768115942029e-06,
+      "loss": 1.7621,
+      "step": 29080
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 9.810726165771484,
+      "learning_rate": 3.431884057971015e-06,
+      "loss": 1.5289,
+      "step": 29090
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.837118625640869,
+      "learning_rate": 3.426086956521739e-06,
+      "loss": 1.6023,
+      "step": 29100
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.719472885131836,
+      "learning_rate": 3.420289855072464e-06,
+      "loss": 1.6571,
+      "step": 29110
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.3916150331497192,
+      "learning_rate": 3.4144927536231888e-06,
+      "loss": 1.6231,
+      "step": 29120
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.7587890625,
+      "learning_rate": 3.408695652173913e-06,
+      "loss": 1.6458,
+      "step": 29130
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.0899200439453125,
+      "learning_rate": 3.402898550724638e-06,
+      "loss": 1.681,
+      "step": 29140
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.8998475074768066,
+      "learning_rate": 3.3971014492753625e-06,
+      "loss": 1.5843,
+      "step": 29150
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 13.847112655639648,
+      "learning_rate": 3.391304347826087e-06,
+      "loss": 1.5932,
+      "step": 29160
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 16.733882904052734,
+      "learning_rate": 3.3855072463768118e-06,
+      "loss": 1.6131,
+      "step": 29170
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.589210033416748,
+      "learning_rate": 3.3797101449275366e-06,
+      "loss": 1.5347,
+      "step": 29180
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.7056937217712402,
+      "learning_rate": 3.3739130434782615e-06,
+      "loss": 1.6423,
+      "step": 29190
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.761937618255615,
+      "learning_rate": 3.3681159420289855e-06,
+      "loss": 1.5252,
+      "step": 29200
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.621154308319092,
+      "learning_rate": 3.3623188405797103e-06,
+      "loss": 1.4666,
+      "step": 29210
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 9.775867462158203,
+      "learning_rate": 3.356521739130435e-06,
+      "loss": 1.6301,
+      "step": 29220
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.965156078338623,
+      "learning_rate": 3.3507246376811596e-06,
+      "loss": 1.7043,
+      "step": 29230
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 11.6044282913208,
+      "learning_rate": 3.3449275362318845e-06,
+      "loss": 1.6566,
+      "step": 29240
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.9858810901641846,
+      "learning_rate": 3.3391304347826093e-06,
+      "loss": 1.5687,
+      "step": 29250
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.0615386962890625,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.4778,
+      "step": 29260
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 8.376899719238281,
+      "learning_rate": 3.327536231884058e-06,
+      "loss": 1.6104,
+      "step": 29270
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.015613555908203,
+      "learning_rate": 3.321739130434783e-06,
+      "loss": 1.5788,
+      "step": 29280
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.4846994876861572,
+      "learning_rate": 3.3159420289855075e-06,
+      "loss": 1.5098,
+      "step": 29290
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.8240742683410645,
+      "learning_rate": 3.3101449275362323e-06,
+      "loss": 1.7382,
+      "step": 29300
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.34401798248291,
+      "learning_rate": 3.3043478260869567e-06,
+      "loss": 1.6224,
+      "step": 29310
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.242111682891846,
+      "learning_rate": 3.298550724637681e-06,
+      "loss": 1.7168,
+      "step": 29320
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.714269638061523,
+      "learning_rate": 3.292753623188406e-06,
+      "loss": 1.6101,
+      "step": 29330
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.2656619548797607,
+      "learning_rate": 3.286956521739131e-06,
+      "loss": 1.6284,
+      "step": 29340
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.77830696105957,
+      "learning_rate": 3.2811594202898557e-06,
+      "loss": 1.6186,
+      "step": 29350
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.6399919986724854,
+      "learning_rate": 3.2753623188405797e-06,
+      "loss": 1.5704,
+      "step": 29360
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.7771320343017578,
+      "learning_rate": 3.2695652173913046e-06,
+      "loss": 1.6302,
+      "step": 29370
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.038367748260498,
+      "learning_rate": 3.2637681159420294e-06,
+      "loss": 1.4699,
+      "step": 29380
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.5462366342544556,
+      "learning_rate": 3.257971014492754e-06,
+      "loss": 1.5656,
+      "step": 29390
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 8.673677444458008,
+      "learning_rate": 3.252753623188406e-06,
+      "loss": 1.5968,
+      "step": 29400
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.5238194465637207,
+      "learning_rate": 3.246956521739131e-06,
+      "loss": 1.7341,
+      "step": 29410
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.633085012435913,
+      "learning_rate": 3.2411594202898557e-06,
+      "loss": 1.6154,
+      "step": 29420
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.964749336242676,
+      "learning_rate": 3.2353623188405797e-06,
+      "loss": 1.4897,
+      "step": 29430
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.036418437957764,
+      "learning_rate": 3.2295652173913045e-06,
+      "loss": 1.74,
+      "step": 29440
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.448636054992676,
+      "learning_rate": 3.2237681159420294e-06,
+      "loss": 1.7668,
+      "step": 29450
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.758767604827881,
+      "learning_rate": 3.217971014492754e-06,
+      "loss": 1.5285,
+      "step": 29460
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 13.795671463012695,
+      "learning_rate": 3.2121739130434787e-06,
+      "loss": 1.5928,
+      "step": 29470
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.950065553188324,
+      "learning_rate": 3.2063768115942027e-06,
+      "loss": 1.6188,
+      "step": 29480
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 9.130352020263672,
+      "learning_rate": 3.2005797101449275e-06,
+      "loss": 1.5059,
+      "step": 29490
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.847170114517212,
+      "learning_rate": 3.1947826086956524e-06,
+      "loss": 1.6167,
+      "step": 29500
+    },
+    {
+      "epoch": 2.46,
+      "eval_loss": 1.6451491117477417,
+      "eval_runtime": 107.4982,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 29500
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.0313570499420166,
+      "learning_rate": 3.1889855072463772e-06,
+      "loss": 1.6151,
+      "step": 29510
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.9217429161071777,
+      "learning_rate": 3.1831884057971017e-06,
+      "loss": 1.643,
+      "step": 29520
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.5794498920440674,
+      "learning_rate": 3.1773913043478265e-06,
+      "loss": 1.6273,
+      "step": 29530
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.837756395339966,
+      "learning_rate": 3.171594202898551e-06,
+      "loss": 1.5733,
+      "step": 29540
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.105989933013916,
+      "learning_rate": 3.1657971014492754e-06,
+      "loss": 1.5915,
+      "step": 29550
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 9.32128620147705,
+      "learning_rate": 3.1600000000000002e-06,
+      "loss": 1.5463,
+      "step": 29560
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.723803520202637,
+      "learning_rate": 3.154202898550725e-06,
+      "loss": 1.7413,
+      "step": 29570
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 8.316760063171387,
+      "learning_rate": 3.14840579710145e-06,
+      "loss": 1.6757,
+      "step": 29580
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.694721698760986,
+      "learning_rate": 3.142608695652174e-06,
+      "loss": 1.5203,
+      "step": 29590
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.9127633571624756,
+      "learning_rate": 3.136811594202899e-06,
+      "loss": 1.5992,
+      "step": 29600
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.171841144561768,
+      "learning_rate": 3.1310144927536237e-06,
+      "loss": 1.6754,
+      "step": 29610
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.929067134857178,
+      "learning_rate": 3.125217391304348e-06,
+      "loss": 1.6738,
+      "step": 29620
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.294098854064941,
+      "learning_rate": 3.119420289855073e-06,
+      "loss": 1.6663,
+      "step": 29630
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.76629376411438,
+      "learning_rate": 3.113623188405797e-06,
+      "loss": 1.6625,
+      "step": 29640
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.9984612464904785,
+      "learning_rate": 3.107826086956522e-06,
+      "loss": 1.5415,
+      "step": 29650
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.3784775733947754,
+      "learning_rate": 3.1020289855072466e-06,
+      "loss": 1.7637,
+      "step": 29660
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.289485454559326,
+      "learning_rate": 3.0962318840579715e-06,
+      "loss": 1.7363,
+      "step": 29670
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.363936424255371,
+      "learning_rate": 3.090434782608696e-06,
+      "loss": 1.7829,
+      "step": 29680
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.904037356376648,
+      "learning_rate": 3.0846376811594204e-06,
+      "loss": 1.6108,
+      "step": 29690
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.6849846839904785,
+      "learning_rate": 3.0788405797101452e-06,
+      "loss": 1.53,
+      "step": 29700
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 13.439961433410645,
+      "learning_rate": 3.0730434782608696e-06,
+      "loss": 1.6294,
+      "step": 29710
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 14.236470222473145,
+      "learning_rate": 3.0672463768115945e-06,
+      "loss": 1.5566,
+      "step": 29720
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.827075481414795,
+      "learning_rate": 3.0614492753623193e-06,
+      "loss": 1.6481,
+      "step": 29730
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.0929653644561768,
+      "learning_rate": 3.0556521739130434e-06,
+      "loss": 1.6433,
+      "step": 29740
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.4234700202941895,
+      "learning_rate": 3.049855072463768e-06,
+      "loss": 1.5093,
+      "step": 29750
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.105968952178955,
+      "learning_rate": 3.044057971014493e-06,
+      "loss": 1.5319,
+      "step": 29760
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.3987698554992676,
+      "learning_rate": 3.038260869565218e-06,
+      "loss": 1.5574,
+      "step": 29770
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.3514299392700195,
+      "learning_rate": 3.0324637681159423e-06,
+      "loss": 1.5673,
+      "step": 29780
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.782695770263672,
+      "learning_rate": 3.0266666666666668e-06,
+      "loss": 1.5443,
+      "step": 29790
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.090881824493408,
+      "learning_rate": 3.020869565217391e-06,
+      "loss": 1.5812,
+      "step": 29800
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 11.7606782913208,
+      "learning_rate": 3.015072463768116e-06,
+      "loss": 1.6019,
+      "step": 29810
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 10.391364097595215,
+      "learning_rate": 3.009275362318841e-06,
+      "loss": 1.5642,
+      "step": 29820
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.301925182342529,
+      "learning_rate": 3.0034782608695658e-06,
+      "loss": 1.575,
+      "step": 29830
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.747443675994873,
+      "learning_rate": 2.99768115942029e-06,
+      "loss": 1.7891,
+      "step": 29840
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.40224027633667,
+      "learning_rate": 2.9918840579710146e-06,
+      "loss": 1.7931,
+      "step": 29850
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.9095501899719238,
+      "learning_rate": 2.9860869565217395e-06,
+      "loss": 1.727,
+      "step": 29860
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.217514514923096,
+      "learning_rate": 2.980289855072464e-06,
+      "loss": 1.5056,
+      "step": 29870
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.045649766921997,
+      "learning_rate": 2.9744927536231888e-06,
+      "loss": 1.5775,
+      "step": 29880
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.6917686462402344,
+      "learning_rate": 2.9686956521739136e-06,
+      "loss": 1.6197,
+      "step": 29890
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.3253226280212402,
+      "learning_rate": 2.9628985507246376e-06,
+      "loss": 1.7079,
+      "step": 29900
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.5302798748016357,
+      "learning_rate": 2.9571014492753625e-06,
+      "loss": 1.5915,
+      "step": 29910
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.9401729106903076,
+      "learning_rate": 2.9513043478260873e-06,
+      "loss": 1.5172,
+      "step": 29920
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.63379430770874,
+      "learning_rate": 2.945507246376812e-06,
+      "loss": 1.6475,
+      "step": 29930
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.7005679607391357,
+      "learning_rate": 2.9397101449275366e-06,
+      "loss": 1.383,
+      "step": 29940
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.5383400917053223,
+      "learning_rate": 2.933913043478261e-06,
+      "loss": 1.6116,
+      "step": 29950
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.307799816131592,
+      "learning_rate": 2.9281159420289855e-06,
+      "loss": 1.7254,
+      "step": 29960
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.233972549438477,
+      "learning_rate": 2.9223188405797103e-06,
+      "loss": 1.5307,
+      "step": 29970
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.475644588470459,
+      "learning_rate": 2.916521739130435e-06,
+      "loss": 1.6835,
+      "step": 29980
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.9054574966430664,
+      "learning_rate": 2.91072463768116e-06,
+      "loss": 1.5383,
+      "step": 29990
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.8980517387390137,
+      "learning_rate": 2.904927536231884e-06,
+      "loss": 1.6759,
+      "step": 30000
+    },
+    {
+      "epoch": 2.5,
+      "eval_loss": 1.6585332155227661,
+      "eval_runtime": 107.501,
+      "eval_samples_per_second": 9.302,
+      "eval_steps_per_second": 2.326,
+      "step": 30000
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.9723654985427856,
+      "learning_rate": 2.899130434782609e-06,
+      "loss": 1.6399,
+      "step": 30010
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 7.082981586456299,
+      "learning_rate": 2.8933333333333337e-06,
+      "loss": 1.6246,
+      "step": 30020
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.984813690185547,
+      "learning_rate": 2.887536231884058e-06,
+      "loss": 1.6357,
+      "step": 30030
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 9.565591812133789,
+      "learning_rate": 2.881739130434783e-06,
+      "loss": 1.6206,
+      "step": 30040
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 13.118734359741211,
+      "learning_rate": 2.8759420289855074e-06,
+      "loss": 1.6352,
+      "step": 30050
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.744248867034912,
+      "learning_rate": 2.870144927536232e-06,
+      "loss": 1.4582,
+      "step": 30060
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.0195579528808594,
+      "learning_rate": 2.8643478260869567e-06,
+      "loss": 1.6441,
+      "step": 30070
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.9167885780334473,
+      "learning_rate": 2.8585507246376816e-06,
+      "loss": 1.7603,
+      "step": 30080
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.419768810272217,
+      "learning_rate": 2.852753623188406e-06,
+      "loss": 1.598,
+      "step": 30090
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.9617881774902344,
+      "learning_rate": 2.8469565217391304e-06,
+      "loss": 1.639,
+      "step": 30100
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.755096673965454,
+      "learning_rate": 2.8411594202898553e-06,
+      "loss": 1.6888,
+      "step": 30110
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 9.04216194152832,
+      "learning_rate": 2.8353623188405797e-06,
+      "loss": 1.6008,
+      "step": 30120
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.4999070167541504,
+      "learning_rate": 2.8295652173913046e-06,
+      "loss": 1.6094,
+      "step": 30130
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.484206199645996,
+      "learning_rate": 2.8237681159420294e-06,
+      "loss": 1.6802,
+      "step": 30140
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 7.458893299102783,
+      "learning_rate": 2.8179710144927543e-06,
+      "loss": 1.559,
+      "step": 30150
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 8.366874694824219,
+      "learning_rate": 2.8121739130434783e-06,
+      "loss": 1.5715,
+      "step": 30160
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 10.303510665893555,
+      "learning_rate": 2.806376811594203e-06,
+      "loss": 1.5516,
+      "step": 30170
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.16677713394165,
+      "learning_rate": 2.800579710144928e-06,
+      "loss": 1.5497,
+      "step": 30180
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.9423246383667,
+      "learning_rate": 2.7947826086956524e-06,
+      "loss": 1.7121,
+      "step": 30190
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.328890562057495,
+      "learning_rate": 2.7889855072463773e-06,
+      "loss": 1.5159,
+      "step": 30200
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.184723854064941,
+      "learning_rate": 2.7831884057971013e-06,
+      "loss": 1.6757,
+      "step": 30210
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.9649858474731445,
+      "learning_rate": 2.777391304347826e-06,
+      "loss": 1.4885,
+      "step": 30220
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.443271160125732,
+      "learning_rate": 2.771594202898551e-06,
+      "loss": 1.5998,
+      "step": 30230
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.292788028717041,
+      "learning_rate": 2.765797101449276e-06,
+      "loss": 1.5722,
+      "step": 30240
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 9.991342544555664,
+      "learning_rate": 2.7600000000000003e-06,
+      "loss": 1.6356,
+      "step": 30250
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.1068317890167236,
+      "learning_rate": 2.7542028985507247e-06,
+      "loss": 1.651,
+      "step": 30260
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 7.125411033630371,
+      "learning_rate": 2.7484057971014495e-06,
+      "loss": 1.5846,
+      "step": 30270
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.9197959899902344,
+      "learning_rate": 2.742608695652174e-06,
+      "loss": 1.6979,
+      "step": 30280
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.4163668155670166,
+      "learning_rate": 2.736811594202899e-06,
+      "loss": 1.6169,
+      "step": 30290
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 8.496438026428223,
+      "learning_rate": 2.7310144927536237e-06,
+      "loss": 1.6185,
+      "step": 30300
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.996656894683838,
+      "learning_rate": 2.7252173913043477e-06,
+      "loss": 1.747,
+      "step": 30310
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.3623204231262207,
+      "learning_rate": 2.7194202898550725e-06,
+      "loss": 1.6652,
+      "step": 30320
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.751153945922852,
+      "learning_rate": 2.7136231884057974e-06,
+      "loss": 1.5636,
+      "step": 30330
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.7132070064544678,
+      "learning_rate": 2.7078260869565222e-06,
+      "loss": 1.5571,
+      "step": 30340
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 12.558267593383789,
+      "learning_rate": 2.7020289855072467e-06,
+      "loss": 1.6319,
+      "step": 30350
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.379528522491455,
+      "learning_rate": 2.696231884057971e-06,
+      "loss": 1.6902,
+      "step": 30360
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.271668910980225,
+      "learning_rate": 2.6904347826086955e-06,
+      "loss": 1.7105,
+      "step": 30370
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.071247577667236,
+      "learning_rate": 2.6846376811594204e-06,
+      "loss": 1.5997,
+      "step": 30380
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.980905532836914,
+      "learning_rate": 2.6788405797101452e-06,
+      "loss": 1.6008,
+      "step": 30390
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.2732436656951904,
+      "learning_rate": 2.67304347826087e-06,
+      "loss": 1.6783,
+      "step": 30400
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.015439748764038,
+      "learning_rate": 2.667246376811594e-06,
+      "loss": 1.6118,
+      "step": 30410
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.7114615440368652,
+      "learning_rate": 2.661449275362319e-06,
+      "loss": 1.6513,
+      "step": 30420
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.484541893005371,
+      "learning_rate": 2.655652173913044e-06,
+      "loss": 1.6789,
+      "step": 30430
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.272510528564453,
+      "learning_rate": 2.6498550724637682e-06,
+      "loss": 1.5341,
+      "step": 30440
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.3156094551086426,
+      "learning_rate": 2.644057971014493e-06,
+      "loss": 1.7266,
+      "step": 30450
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 15.670822143554688,
+      "learning_rate": 2.638260869565218e-06,
+      "loss": 1.6492,
+      "step": 30460
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 10.792664527893066,
+      "learning_rate": 2.632463768115942e-06,
+      "loss": 1.7282,
+      "step": 30470
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.0802708864212036,
+      "learning_rate": 2.6266666666666668e-06,
+      "loss": 1.6272,
+      "step": 30480
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.8151583671569824,
+      "learning_rate": 2.6208695652173916e-06,
+      "loss": 1.6041,
+      "step": 30490
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.1548194885253906,
+      "learning_rate": 2.6150724637681165e-06,
+      "loss": 1.6457,
+      "step": 30500
+    },
+    {
+      "epoch": 2.54,
+      "eval_loss": 1.642120599746704,
+      "eval_runtime": 107.4944,
+      "eval_samples_per_second": 9.303,
+      "eval_steps_per_second": 2.326,
+      "step": 30500
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.8505381345748901,
+      "learning_rate": 2.609275362318841e-06,
+      "loss": 1.5916,
+      "step": 30510
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.077845096588135,
+      "learning_rate": 2.6034782608695654e-06,
+      "loss": 1.5464,
+      "step": 30520
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.143275022506714,
+      "learning_rate": 2.5976811594202898e-06,
+      "loss": 1.5341,
+      "step": 30530
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.190448760986328,
+      "learning_rate": 2.5918840579710146e-06,
+      "loss": 1.5817,
+      "step": 30540
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.009246587753296,
+      "learning_rate": 2.5860869565217395e-06,
+      "loss": 1.5787,
+      "step": 30550
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.875110149383545,
+      "learning_rate": 2.5802898550724643e-06,
+      "loss": 1.5827,
+      "step": 30560
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.125142097473145,
+      "learning_rate": 2.5744927536231883e-06,
+      "loss": 1.4877,
+      "step": 30570
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 6.384122848510742,
+      "learning_rate": 2.568695652173913e-06,
+      "loss": 1.6345,
+      "step": 30580
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 10.49345874786377,
+      "learning_rate": 2.562898550724638e-06,
+      "loss": 1.6619,
+      "step": 30590
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.870846152305603,
+      "learning_rate": 2.5571014492753625e-06,
+      "loss": 1.5736,
+      "step": 30600
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.0048367977142334,
+      "learning_rate": 2.5513043478260873e-06,
+      "loss": 1.6575,
+      "step": 30610
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.404007434844971,
+      "learning_rate": 2.5455072463768118e-06,
+      "loss": 1.6789,
+      "step": 30620
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.966566562652588,
+      "learning_rate": 2.539710144927536e-06,
+      "loss": 1.5581,
+      "step": 30630
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.988835096359253,
+      "learning_rate": 2.533913043478261e-06,
+      "loss": 1.4879,
+      "step": 30640
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 7.500990390777588,
+      "learning_rate": 2.528115942028986e-06,
+      "loss": 1.6214,
+      "step": 30650
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 6.991439342498779,
+      "learning_rate": 2.5223188405797107e-06,
+      "loss": 1.5002,
+      "step": 30660
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.588336706161499,
+      "learning_rate": 2.5165217391304348e-06,
+      "loss": 1.6576,
+      "step": 30670
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.5451226234436035,
+      "learning_rate": 2.5107246376811596e-06,
+      "loss": 1.6218,
+      "step": 30680
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 10.67144775390625,
+      "learning_rate": 2.504927536231884e-06,
+      "loss": 1.6439,
+      "step": 30690
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 14.395882606506348,
+      "learning_rate": 2.499130434782609e-06,
+      "loss": 1.5812,
+      "step": 30700
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.2323532104492188,
+      "learning_rate": 2.4933333333333333e-06,
+      "loss": 1.6072,
+      "step": 30710
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.6720938682556152,
+      "learning_rate": 2.487536231884058e-06,
+      "loss": 1.5967,
+      "step": 30720
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.9183871746063232,
+      "learning_rate": 2.481739130434783e-06,
+      "loss": 1.659,
+      "step": 30730
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 8.462002754211426,
+      "learning_rate": 2.4759420289855075e-06,
+      "loss": 1.6629,
+      "step": 30740
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 9.175357818603516,
+      "learning_rate": 2.4701449275362323e-06,
+      "loss": 1.6591,
+      "step": 30750
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 11.072102546691895,
+      "learning_rate": 2.4643478260869567e-06,
+      "loss": 1.5515,
+      "step": 30760
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.013851642608643,
+      "learning_rate": 2.458550724637681e-06,
+      "loss": 1.5813,
+      "step": 30770
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.7383862733840942,
+      "learning_rate": 2.452753623188406e-06,
+      "loss": 1.7057,
+      "step": 30780
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.308183670043945,
+      "learning_rate": 2.4469565217391304e-06,
+      "loss": 1.5989,
+      "step": 30790
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.486476898193359,
+      "learning_rate": 2.4411594202898553e-06,
+      "loss": 1.6127,
+      "step": 30800
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.1956095695495605,
+      "learning_rate": 2.4353623188405797e-06,
+      "loss": 1.515,
+      "step": 30810
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.074836730957031,
+      "learning_rate": 2.4295652173913046e-06,
+      "loss": 1.5636,
+      "step": 30820
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.5480659008026123,
+      "learning_rate": 2.4237681159420294e-06,
+      "loss": 1.6296,
+      "step": 30830
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.6157848834991455,
+      "learning_rate": 2.417971014492754e-06,
+      "loss": 1.7141,
+      "step": 30840
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.7036499977111816,
+      "learning_rate": 2.4121739130434783e-06,
+      "loss": 1.6024,
+      "step": 30850
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 9.184311866760254,
+      "learning_rate": 2.406376811594203e-06,
+      "loss": 1.6096,
+      "step": 30860
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 10.717764854431152,
+      "learning_rate": 2.4005797101449276e-06,
+      "loss": 1.6807,
+      "step": 30870
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 6.877434253692627,
+      "learning_rate": 2.3947826086956524e-06,
+      "loss": 1.6401,
+      "step": 30880
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.13861083984375,
+      "learning_rate": 2.388985507246377e-06,
+      "loss": 1.6329,
+      "step": 30890
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.996332883834839,
+      "learning_rate": 2.3831884057971017e-06,
+      "loss": 1.5384,
+      "step": 30900
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.9450478553771973,
+      "learning_rate": 2.3773913043478266e-06,
+      "loss": 1.6154,
+      "step": 30910
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.448505401611328,
+      "learning_rate": 2.371594202898551e-06,
+      "loss": 1.7094,
+      "step": 30920
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 8.060430526733398,
+      "learning_rate": 2.3657971014492754e-06,
+      "loss": 1.5228,
+      "step": 30930
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.698649406433105,
+      "learning_rate": 2.3600000000000003e-06,
+      "loss": 1.6043,
+      "step": 30940
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.175755262374878,
+      "learning_rate": 2.3542028985507247e-06,
+      "loss": 1.5744,
+      "step": 30950
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.9357259273529053,
+      "learning_rate": 2.3484057971014496e-06,
+      "loss": 1.5971,
+      "step": 30960
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.1497623920440674,
+      "learning_rate": 2.342608695652174e-06,
+      "loss": 1.5522,
+      "step": 30970
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.6525659561157227,
+      "learning_rate": 2.336811594202899e-06,
+      "loss": 1.6543,
+      "step": 30980
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 11.570157051086426,
+      "learning_rate": 2.3310144927536237e-06,
+      "loss": 1.5808,
+      "step": 30990
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 7.399476528167725,
+      "learning_rate": 2.325217391304348e-06,
+      "loss": 1.701,
+      "step": 31000
+    },
+    {
+      "epoch": 2.58,
+      "eval_loss": 1.6337002515792847,
+      "eval_runtime": 107.5309,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 31000
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.558446407318115,
+      "learning_rate": 2.3194202898550725e-06,
+      "loss": 1.6179,
+      "step": 31010
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.1192896366119385,
+      "learning_rate": 2.3142028985507247e-06,
+      "loss": 1.4463,
+      "step": 31020
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.965945839881897,
+      "learning_rate": 2.3084057971014495e-06,
+      "loss": 1.6299,
+      "step": 31030
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.286752223968506,
+      "learning_rate": 2.302608695652174e-06,
+      "loss": 1.5018,
+      "step": 31040
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.3863645792007446,
+      "learning_rate": 2.296811594202899e-06,
+      "loss": 1.6844,
+      "step": 31050
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 13.46044921875,
+      "learning_rate": 2.2910144927536237e-06,
+      "loss": 1.6116,
+      "step": 31060
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 7.819109916687012,
+      "learning_rate": 2.285217391304348e-06,
+      "loss": 1.5284,
+      "step": 31070
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.452699899673462,
+      "learning_rate": 2.2794202898550725e-06,
+      "loss": 1.6716,
+      "step": 31080
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.367595195770264,
+      "learning_rate": 2.2736231884057974e-06,
+      "loss": 1.4822,
+      "step": 31090
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 9.822129249572754,
+      "learning_rate": 2.267826086956522e-06,
+      "loss": 1.6761,
+      "step": 31100
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.4133310317993164,
+      "learning_rate": 2.2620289855072466e-06,
+      "loss": 1.655,
+      "step": 31110
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.704021453857422,
+      "learning_rate": 2.256231884057971e-06,
+      "loss": 1.531,
+      "step": 31120
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 6.551307201385498,
+      "learning_rate": 2.250434782608696e-06,
+      "loss": 1.6892,
+      "step": 31130
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.172097682952881,
+      "learning_rate": 2.2446376811594208e-06,
+      "loss": 1.6129,
+      "step": 31140
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.622472286224365,
+      "learning_rate": 2.238840579710145e-06,
+      "loss": 1.5993,
+      "step": 31150
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.656231641769409,
+      "learning_rate": 2.2330434782608696e-06,
+      "loss": 1.6874,
+      "step": 31160
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.335791826248169,
+      "learning_rate": 2.2272463768115945e-06,
+      "loss": 1.5399,
+      "step": 31170
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.9562406539916992,
+      "learning_rate": 2.221449275362319e-06,
+      "loss": 1.7015,
+      "step": 31180
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.210826873779297,
+      "learning_rate": 2.2156521739130438e-06,
+      "loss": 1.5114,
+      "step": 31190
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 9.8193998336792,
+      "learning_rate": 2.209855072463768e-06,
+      "loss": 1.509,
+      "step": 31200
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.946864128112793,
+      "learning_rate": 2.204057971014493e-06,
+      "loss": 1.6648,
+      "step": 31210
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.112027168273926,
+      "learning_rate": 2.1982608695652175e-06,
+      "loss": 1.6629,
+      "step": 31220
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.7467082738876343,
+      "learning_rate": 2.1924637681159423e-06,
+      "loss": 1.637,
+      "step": 31230
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.126824855804443,
+      "learning_rate": 2.1866666666666668e-06,
+      "loss": 1.6365,
+      "step": 31240
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 7.7713398933410645,
+      "learning_rate": 2.180869565217391e-06,
+      "loss": 1.4703,
+      "step": 31250
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.900889873504639,
+      "learning_rate": 2.175072463768116e-06,
+      "loss": 1.5412,
+      "step": 31260
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.582507848739624,
+      "learning_rate": 2.169275362318841e-06,
+      "loss": 1.51,
+      "step": 31270
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.354831218719482,
+      "learning_rate": 2.1634782608695653e-06,
+      "loss": 1.7147,
+      "step": 31280
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.068592071533203,
+      "learning_rate": 2.15768115942029e-06,
+      "loss": 1.5849,
+      "step": 31290
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.196958065032959,
+      "learning_rate": 2.1518840579710146e-06,
+      "loss": 1.6493,
+      "step": 31300
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.429135322570801,
+      "learning_rate": 2.1460869565217395e-06,
+      "loss": 1.5347,
+      "step": 31310
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.728844165802002,
+      "learning_rate": 2.140289855072464e-06,
+      "loss": 1.5515,
+      "step": 31320
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 13.872916221618652,
+      "learning_rate": 2.1344927536231883e-06,
+      "loss": 1.602,
+      "step": 31330
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.392234802246094,
+      "learning_rate": 2.128695652173913e-06,
+      "loss": 1.7038,
+      "step": 31340
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.907491683959961,
+      "learning_rate": 2.1228985507246376e-06,
+      "loss": 1.5619,
+      "step": 31350
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.335078239440918,
+      "learning_rate": 2.1171014492753625e-06,
+      "loss": 1.5478,
+      "step": 31360
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.8956576585769653,
+      "learning_rate": 2.1113043478260873e-06,
+      "loss": 1.6199,
+      "step": 31370
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 12.366223335266113,
+      "learning_rate": 2.1055072463768117e-06,
+      "loss": 1.4663,
+      "step": 31380
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.7240118980407715,
+      "learning_rate": 2.0997101449275366e-06,
+      "loss": 1.7283,
+      "step": 31390
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.118774652481079,
+      "learning_rate": 2.093913043478261e-06,
+      "loss": 1.6955,
+      "step": 31400
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.2114875316619873,
+      "learning_rate": 2.0881159420289855e-06,
+      "loss": 1.5495,
+      "step": 31410
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.9266295433044434,
+      "learning_rate": 2.0823188405797103e-06,
+      "loss": 1.5095,
+      "step": 31420
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.1518819332122803,
+      "learning_rate": 2.0765217391304347e-06,
+      "loss": 1.6809,
+      "step": 31430
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.437044620513916,
+      "learning_rate": 2.0707246376811596e-06,
+      "loss": 1.6229,
+      "step": 31440
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.866549491882324,
+      "learning_rate": 2.0649275362318844e-06,
+      "loss": 1.6808,
+      "step": 31450
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.811148166656494,
+      "learning_rate": 2.059130434782609e-06,
+      "loss": 1.4217,
+      "step": 31460
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.14778470993042,
+      "learning_rate": 2.0533333333333337e-06,
+      "loss": 1.6761,
+      "step": 31470
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.273214340209961,
+      "learning_rate": 2.047536231884058e-06,
+      "loss": 1.5311,
+      "step": 31480
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.9031133055686951,
+      "learning_rate": 2.0417391304347826e-06,
+      "loss": 1.7013,
+      "step": 31490
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 10.49181079864502,
+      "learning_rate": 2.0359420289855074e-06,
+      "loss": 1.6723,
+      "step": 31500
+    },
+    {
+      "epoch": 2.62,
+      "eval_loss": 1.6217410564422607,
+      "eval_runtime": 107.5528,
+      "eval_samples_per_second": 9.298,
+      "eval_steps_per_second": 2.324,
+      "step": 31500
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.9454545974731445,
+      "learning_rate": 2.030144927536232e-06,
+      "loss": 1.6591,
+      "step": 31510
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 9.571122169494629,
+      "learning_rate": 2.0243478260869567e-06,
+      "loss": 1.6785,
+      "step": 31520
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 7.118087291717529,
+      "learning_rate": 2.018550724637681e-06,
+      "loss": 1.5377,
+      "step": 31530
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 16.46897315979004,
+      "learning_rate": 2.012753623188406e-06,
+      "loss": 1.5683,
+      "step": 31540
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.300513505935669,
+      "learning_rate": 2.006956521739131e-06,
+      "loss": 1.6441,
+      "step": 31550
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.414223551750183,
+      "learning_rate": 2.0011594202898553e-06,
+      "loss": 1.4831,
+      "step": 31560
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.0008349418640137,
+      "learning_rate": 1.9953623188405797e-06,
+      "loss": 1.6988,
+      "step": 31570
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.3853042125701904,
+      "learning_rate": 1.9895652173913046e-06,
+      "loss": 1.6968,
+      "step": 31580
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 5.788661956787109,
+      "learning_rate": 1.983768115942029e-06,
+      "loss": 1.7416,
+      "step": 31590
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.9878652095794678,
+      "learning_rate": 1.977971014492754e-06,
+      "loss": 1.733,
+      "step": 31600
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.0681121349334717,
+      "learning_rate": 1.9721739130434783e-06,
+      "loss": 1.5268,
+      "step": 31610
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 6.360716342926025,
+      "learning_rate": 1.966376811594203e-06,
+      "loss": 1.6611,
+      "step": 31620
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.8834378719329834,
+      "learning_rate": 1.960579710144928e-06,
+      "loss": 1.59,
+      "step": 31630
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.510819911956787,
+      "learning_rate": 1.9547826086956524e-06,
+      "loss": 1.6145,
+      "step": 31640
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.6274240016937256,
+      "learning_rate": 1.948985507246377e-06,
+      "loss": 1.5678,
+      "step": 31650
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.6434683799743652,
+      "learning_rate": 1.9431884057971017e-06,
+      "loss": 1.6534,
+      "step": 31660
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.2217583656311035,
+      "learning_rate": 1.937391304347826e-06,
+      "loss": 1.5893,
+      "step": 31670
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.410333156585693,
+      "learning_rate": 1.931594202898551e-06,
+      "loss": 1.7107,
+      "step": 31680
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.742589473724365,
+      "learning_rate": 1.9257971014492754e-06,
+      "loss": 1.7014,
+      "step": 31690
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.723726272583008,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 1.4988,
+      "step": 31700
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.500274181365967,
+      "learning_rate": 1.9142028985507247e-06,
+      "loss": 1.5023,
+      "step": 31710
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.310852527618408,
+      "learning_rate": 1.9084057971014495e-06,
+      "loss": 1.65,
+      "step": 31720
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 8.708599090576172,
+      "learning_rate": 1.9026086956521742e-06,
+      "loss": 1.6035,
+      "step": 31730
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.076088905334473,
+      "learning_rate": 1.8968115942028986e-06,
+      "loss": 1.5787,
+      "step": 31740
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.023421287536621,
+      "learning_rate": 1.8910144927536235e-06,
+      "loss": 1.6103,
+      "step": 31750
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.380520820617676,
+      "learning_rate": 1.885217391304348e-06,
+      "loss": 1.5645,
+      "step": 31760
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 12.353821754455566,
+      "learning_rate": 1.8794202898550725e-06,
+      "loss": 1.487,
+      "step": 31770
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 10.361377716064453,
+      "learning_rate": 1.8736231884057974e-06,
+      "loss": 1.7336,
+      "step": 31780
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.4413888454437256,
+      "learning_rate": 1.8678260869565218e-06,
+      "loss": 1.6445,
+      "step": 31790
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.693152904510498,
+      "learning_rate": 1.8620289855072465e-06,
+      "loss": 1.5021,
+      "step": 31800
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 27.886554718017578,
+      "learning_rate": 1.8562318840579713e-06,
+      "loss": 1.5851,
+      "step": 31810
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.8801918029785156,
+      "learning_rate": 1.8504347826086957e-06,
+      "loss": 1.6223,
+      "step": 31820
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.8942666053771973,
+      "learning_rate": 1.8446376811594206e-06,
+      "loss": 1.7608,
+      "step": 31830
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.6183860301971436,
+      "learning_rate": 1.838840579710145e-06,
+      "loss": 1.644,
+      "step": 31840
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.7478418350219727,
+      "learning_rate": 1.8330434782608697e-06,
+      "loss": 1.4282,
+      "step": 31850
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.082428216934204,
+      "learning_rate": 1.8272463768115945e-06,
+      "loss": 1.6814,
+      "step": 31860
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.891972064971924,
+      "learning_rate": 1.821449275362319e-06,
+      "loss": 1.5985,
+      "step": 31870
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 10.067206382751465,
+      "learning_rate": 1.8156521739130436e-06,
+      "loss": 1.411,
+      "step": 31880
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.1874613761901855,
+      "learning_rate": 1.8098550724637682e-06,
+      "loss": 1.5593,
+      "step": 31890
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.310758590698242,
+      "learning_rate": 1.8040579710144929e-06,
+      "loss": 1.5513,
+      "step": 31900
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.2715983390808105,
+      "learning_rate": 1.7982608695652177e-06,
+      "loss": 1.7019,
+      "step": 31910
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.090458869934082,
+      "learning_rate": 1.7924637681159421e-06,
+      "loss": 1.6117,
+      "step": 31920
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 12.502359390258789,
+      "learning_rate": 1.7866666666666668e-06,
+      "loss": 1.4899,
+      "step": 31930
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 13.158916473388672,
+      "learning_rate": 1.7808695652173916e-06,
+      "loss": 1.7233,
+      "step": 31940
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 15.319182395935059,
+      "learning_rate": 1.775072463768116e-06,
+      "loss": 1.5413,
+      "step": 31950
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.916179180145264,
+      "learning_rate": 1.7692753623188407e-06,
+      "loss": 1.6006,
+      "step": 31960
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 8.421045303344727,
+      "learning_rate": 1.7634782608695653e-06,
+      "loss": 1.5972,
+      "step": 31970
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.742832660675049,
+      "learning_rate": 1.75768115942029e-06,
+      "loss": 1.6625,
+      "step": 31980
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5600188970565796,
+      "learning_rate": 1.7518840579710148e-06,
+      "loss": 1.647,
+      "step": 31990
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.127601385116577,
+      "learning_rate": 1.7460869565217393e-06,
+      "loss": 1.5884,
+      "step": 32000
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 1.6017138957977295,
+      "eval_runtime": 107.5185,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 32000
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.60046672821045,
+      "learning_rate": 1.740289855072464e-06,
+      "loss": 1.5752,
+      "step": 32010
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 14.514521598815918,
+      "learning_rate": 1.7344927536231883e-06,
+      "loss": 1.6261,
+      "step": 32020
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 8.017278671264648,
+      "learning_rate": 1.7286956521739132e-06,
+      "loss": 1.6809,
+      "step": 32030
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.6506872177124023,
+      "learning_rate": 1.7228985507246378e-06,
+      "loss": 1.6235,
+      "step": 32040
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.44380521774292,
+      "learning_rate": 1.7171014492753625e-06,
+      "loss": 1.6831,
+      "step": 32050
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.318562984466553,
+      "learning_rate": 1.7113043478260871e-06,
+      "loss": 1.726,
+      "step": 32060
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.484785556793213,
+      "learning_rate": 1.705507246376812e-06,
+      "loss": 1.63,
+      "step": 32070
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.5618247985839844,
+      "learning_rate": 1.6997101449275364e-06,
+      "loss": 1.6426,
+      "step": 32080
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5634952783584595,
+      "learning_rate": 1.693913043478261e-06,
+      "loss": 1.506,
+      "step": 32090
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.788817882537842,
+      "learning_rate": 1.6881159420289855e-06,
+      "loss": 1.6246,
+      "step": 32100
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 7.975259780883789,
+      "learning_rate": 1.6823188405797103e-06,
+      "loss": 1.5248,
+      "step": 32110
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 8.401017189025879,
+      "learning_rate": 1.676521739130435e-06,
+      "loss": 1.5491,
+      "step": 32120
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3177763223648071,
+      "learning_rate": 1.6707246376811596e-06,
+      "loss": 1.6666,
+      "step": 32130
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.550060749053955,
+      "learning_rate": 1.6649275362318842e-06,
+      "loss": 1.4496,
+      "step": 32140
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.780557155609131,
+      "learning_rate": 1.6591304347826087e-06,
+      "loss": 1.5778,
+      "step": 32150
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.2793478965759277,
+      "learning_rate": 1.6533333333333335e-06,
+      "loss": 1.7047,
+      "step": 32160
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.5238006114959717,
+      "learning_rate": 1.6475362318840582e-06,
+      "loss": 1.6762,
+      "step": 32170
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.78254508972168,
+      "learning_rate": 1.6417391304347826e-06,
+      "loss": 1.473,
+      "step": 32180
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.1378421783447266,
+      "learning_rate": 1.6359420289855074e-06,
+      "loss": 1.5478,
+      "step": 32190
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.800199508666992,
+      "learning_rate": 1.6301449275362319e-06,
+      "loss": 1.7282,
+      "step": 32200
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 11.791644096374512,
+      "learning_rate": 1.6243478260869565e-06,
+      "loss": 1.6786,
+      "step": 32210
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.185345649719238,
+      "learning_rate": 1.6185507246376814e-06,
+      "loss": 1.5419,
+      "step": 32220
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.825332164764404,
+      "learning_rate": 1.6127536231884058e-06,
+      "loss": 1.7859,
+      "step": 32230
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.289442300796509,
+      "learning_rate": 1.6069565217391307e-06,
+      "loss": 1.7831,
+      "step": 32240
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 11.469537734985352,
+      "learning_rate": 1.6011594202898553e-06,
+      "loss": 1.591,
+      "step": 32250
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.8224618434906006,
+      "learning_rate": 1.5953623188405797e-06,
+      "loss": 1.7497,
+      "step": 32260
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.616290807723999,
+      "learning_rate": 1.5895652173913046e-06,
+      "loss": 1.6872,
+      "step": 32270
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.3341703414916992,
+      "learning_rate": 1.583768115942029e-06,
+      "loss": 1.593,
+      "step": 32280
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 8.595649719238281,
+      "learning_rate": 1.5779710144927536e-06,
+      "loss": 1.5657,
+      "step": 32290
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.221670389175415,
+      "learning_rate": 1.5721739130434785e-06,
+      "loss": 1.6259,
+      "step": 32300
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.5493180751800537,
+      "learning_rate": 1.566376811594203e-06,
+      "loss": 1.6628,
+      "step": 32310
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.4357565641403198,
+      "learning_rate": 1.5605797101449278e-06,
+      "loss": 1.5747,
+      "step": 32320
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 16.11993408203125,
+      "learning_rate": 1.5547826086956522e-06,
+      "loss": 1.4607,
+      "step": 32330
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.597362995147705,
+      "learning_rate": 1.5489855072463769e-06,
+      "loss": 1.5912,
+      "step": 32340
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.872979164123535,
+      "learning_rate": 1.5431884057971017e-06,
+      "loss": 1.4533,
+      "step": 32350
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.2023985385894775,
+      "learning_rate": 1.5373913043478261e-06,
+      "loss": 1.5058,
+      "step": 32360
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.435876369476318,
+      "learning_rate": 1.5315942028985508e-06,
+      "loss": 1.5454,
+      "step": 32370
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.7985036373138428,
+      "learning_rate": 1.5257971014492756e-06,
+      "loss": 1.6741,
+      "step": 32380
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.3253607749938965,
+      "learning_rate": 1.52e-06,
+      "loss": 1.7431,
+      "step": 32390
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4724886417388916,
+      "learning_rate": 1.514202898550725e-06,
+      "loss": 1.6594,
+      "step": 32400
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.5811171531677246,
+      "learning_rate": 1.5084057971014493e-06,
+      "loss": 1.5847,
+      "step": 32410
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.18812084197998,
+      "learning_rate": 1.502608695652174e-06,
+      "loss": 1.6295,
+      "step": 32420
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 9.978080749511719,
+      "learning_rate": 1.4968115942028988e-06,
+      "loss": 1.5189,
+      "step": 32430
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.765017509460449,
+      "learning_rate": 1.4910144927536233e-06,
+      "loss": 1.6167,
+      "step": 32440
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.3332769870758057,
+      "learning_rate": 1.485217391304348e-06,
+      "loss": 1.6536,
+      "step": 32450
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.815036773681641,
+      "learning_rate": 1.4794202898550725e-06,
+      "loss": 1.6844,
+      "step": 32460
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 8.239996910095215,
+      "learning_rate": 1.4736231884057972e-06,
+      "loss": 1.6551,
+      "step": 32470
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.380647659301758,
+      "learning_rate": 1.467826086956522e-06,
+      "loss": 1.6455,
+      "step": 32480
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 14.243300437927246,
+      "learning_rate": 1.4620289855072465e-06,
+      "loss": 1.6603,
+      "step": 32490
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.9138994216918945,
+      "learning_rate": 1.4562318840579711e-06,
+      "loss": 1.56,
+      "step": 32500
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 1.6212100982666016,
+      "eval_runtime": 107.5288,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 32500
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 10.24506664276123,
+      "learning_rate": 1.4504347826086955e-06,
+      "loss": 1.6814,
+      "step": 32510
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 12.093935012817383,
+      "learning_rate": 1.4446376811594204e-06,
+      "loss": 1.611,
+      "step": 32520
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.0004658699035645,
+      "learning_rate": 1.438840579710145e-06,
+      "loss": 1.5318,
+      "step": 32530
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.960986614227295,
+      "learning_rate": 1.4330434782608697e-06,
+      "loss": 1.6179,
+      "step": 32540
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.640979290008545,
+      "learning_rate": 1.4272463768115943e-06,
+      "loss": 1.6522,
+      "step": 32550
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.157996892929077,
+      "learning_rate": 1.4214492753623192e-06,
+      "loss": 1.6118,
+      "step": 32560
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.887393474578857,
+      "learning_rate": 1.4156521739130436e-06,
+      "loss": 1.6207,
+      "step": 32570
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 11.028023719787598,
+      "learning_rate": 1.4098550724637682e-06,
+      "loss": 1.5071,
+      "step": 32580
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.2174723148345947,
+      "learning_rate": 1.4040579710144927e-06,
+      "loss": 1.6467,
+      "step": 32590
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.945136785507202,
+      "learning_rate": 1.3982608695652175e-06,
+      "loss": 1.6025,
+      "step": 32600
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.605086326599121,
+      "learning_rate": 1.3924637681159422e-06,
+      "loss": 1.5723,
+      "step": 32610
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.6031261682510376,
+      "learning_rate": 1.3866666666666668e-06,
+      "loss": 1.6667,
+      "step": 32620
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 10.576859474182129,
+      "learning_rate": 1.3808695652173914e-06,
+      "loss": 1.5214,
+      "step": 32630
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.951624870300293,
+      "learning_rate": 1.3750724637681159e-06,
+      "loss": 1.4939,
+      "step": 32640
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.097296714782715,
+      "learning_rate": 1.3692753623188407e-06,
+      "loss": 1.4366,
+      "step": 32650
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.388517379760742,
+      "learning_rate": 1.3634782608695654e-06,
+      "loss": 1.5237,
+      "step": 32660
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.370208263397217,
+      "learning_rate": 1.3576811594202898e-06,
+      "loss": 1.7699,
+      "step": 32670
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.332535982131958,
+      "learning_rate": 1.3518840579710146e-06,
+      "loss": 1.6375,
+      "step": 32680
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.344327449798584,
+      "learning_rate": 1.3460869565217393e-06,
+      "loss": 1.4128,
+      "step": 32690
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 9.12246036529541,
+      "learning_rate": 1.340289855072464e-06,
+      "loss": 1.4211,
+      "step": 32700
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.78950572013855,
+      "learning_rate": 1.3344927536231886e-06,
+      "loss": 1.6012,
+      "step": 32710
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.5925843715667725,
+      "learning_rate": 1.328695652173913e-06,
+      "loss": 1.5014,
+      "step": 32720
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.6471080780029297,
+      "learning_rate": 1.3228985507246379e-06,
+      "loss": 1.6523,
+      "step": 32730
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 12.4561128616333,
+      "learning_rate": 1.3171014492753625e-06,
+      "loss": 1.6558,
+      "step": 32740
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.051928997039795,
+      "learning_rate": 1.311304347826087e-06,
+      "loss": 1.5797,
+      "step": 32750
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.791071891784668,
+      "learning_rate": 1.3055072463768118e-06,
+      "loss": 1.5058,
+      "step": 32760
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.921985149383545,
+      "learning_rate": 1.2997101449275362e-06,
+      "loss": 1.727,
+      "step": 32770
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.559007406234741,
+      "learning_rate": 1.293913043478261e-06,
+      "loss": 1.5479,
+      "step": 32780
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 7.684724807739258,
+      "learning_rate": 1.2881159420289857e-06,
+      "loss": 1.5711,
+      "step": 32790
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.558244228363037,
+      "learning_rate": 1.2823188405797101e-06,
+      "loss": 1.5848,
+      "step": 32800
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.056134223937988,
+      "learning_rate": 1.276521739130435e-06,
+      "loss": 1.6693,
+      "step": 32810
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 16.4197998046875,
+      "learning_rate": 1.2707246376811594e-06,
+      "loss": 1.5769,
+      "step": 32820
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.05332612991333,
+      "learning_rate": 1.264927536231884e-06,
+      "loss": 1.6521,
+      "step": 32830
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.956727027893066,
+      "learning_rate": 1.259130434782609e-06,
+      "loss": 1.596,
+      "step": 32840
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.413317680358887,
+      "learning_rate": 1.2533333333333333e-06,
+      "loss": 1.496,
+      "step": 32850
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.767711639404297,
+      "learning_rate": 1.2475362318840582e-06,
+      "loss": 1.5371,
+      "step": 32860
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.680372714996338,
+      "learning_rate": 1.2417391304347826e-06,
+      "loss": 1.531,
+      "step": 32870
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.4916908740997314,
+      "learning_rate": 1.2359420289855073e-06,
+      "loss": 1.6898,
+      "step": 32880
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.154963493347168,
+      "learning_rate": 1.2301449275362321e-06,
+      "loss": 1.4688,
+      "step": 32890
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.380986213684082,
+      "learning_rate": 1.2243478260869567e-06,
+      "loss": 1.7119,
+      "step": 32900
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 10.677800178527832,
+      "learning_rate": 1.2185507246376812e-06,
+      "loss": 1.616,
+      "step": 32910
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 7.30240535736084,
+      "learning_rate": 1.2127536231884058e-06,
+      "loss": 1.5754,
+      "step": 32920
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.45409870147705,
+      "learning_rate": 1.2069565217391305e-06,
+      "loss": 1.6461,
+      "step": 32930
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.102117538452148,
+      "learning_rate": 1.2011594202898553e-06,
+      "loss": 1.4394,
+      "step": 32940
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.18507194519043,
+      "learning_rate": 1.1953623188405797e-06,
+      "loss": 1.6139,
+      "step": 32950
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 7.034895896911621,
+      "learning_rate": 1.1895652173913044e-06,
+      "loss": 1.571,
+      "step": 32960
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.3450069427490234,
+      "learning_rate": 1.183768115942029e-06,
+      "loss": 1.4893,
+      "step": 32970
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.201149940490723,
+      "learning_rate": 1.1779710144927539e-06,
+      "loss": 1.4029,
+      "step": 32980
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.9617148637771606,
+      "learning_rate": 1.1721739130434783e-06,
+      "loss": 1.6637,
+      "step": 32990
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.6870609521865845,
+      "learning_rate": 1.166376811594203e-06,
+      "loss": 1.5653,
+      "step": 33000
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 1.617643117904663,
+      "eval_runtime": 107.5322,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 33000
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.035808563232422,
+      "learning_rate": 1.1605797101449276e-06,
+      "loss": 1.5895,
+      "step": 33010
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.8778083324432373,
+      "learning_rate": 1.1547826086956522e-06,
+      "loss": 1.5143,
+      "step": 33020
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.143531322479248,
+      "learning_rate": 1.1489855072463769e-06,
+      "loss": 1.5878,
+      "step": 33030
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.5766236782073975,
+      "learning_rate": 1.1431884057971015e-06,
+      "loss": 1.5166,
+      "step": 33040
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.0880119800567627,
+      "learning_rate": 1.1373913043478262e-06,
+      "loss": 1.5383,
+      "step": 33050
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.4174723625183105,
+      "learning_rate": 1.1315942028985508e-06,
+      "loss": 1.6729,
+      "step": 33060
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.1581320762634277,
+      "learning_rate": 1.1257971014492754e-06,
+      "loss": 1.7279,
+      "step": 33070
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 4.984561443328857,
+      "learning_rate": 1.12e-06,
+      "loss": 1.651,
+      "step": 33080
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.201277732849121,
+      "learning_rate": 1.1142028985507247e-06,
+      "loss": 1.6384,
+      "step": 33090
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.131202459335327,
+      "learning_rate": 1.1084057971014494e-06,
+      "loss": 1.5721,
+      "step": 33100
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 10.104864120483398,
+      "learning_rate": 1.102608695652174e-06,
+      "loss": 1.7424,
+      "step": 33110
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 6.0055012702941895,
+      "learning_rate": 1.0968115942028986e-06,
+      "loss": 1.5323,
+      "step": 33120
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 5.302622318267822,
+      "learning_rate": 1.0910144927536233e-06,
+      "loss": 1.4631,
+      "step": 33130
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 7.973222732543945,
+      "learning_rate": 1.085217391304348e-06,
+      "loss": 1.7163,
+      "step": 33140
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.29218578338623,
+      "learning_rate": 1.0794202898550726e-06,
+      "loss": 1.5899,
+      "step": 33150
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 9.235881805419922,
+      "learning_rate": 1.0736231884057972e-06,
+      "loss": 1.612,
+      "step": 33160
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 11.965916633605957,
+      "learning_rate": 1.0678260869565218e-06,
+      "loss": 1.602,
+      "step": 33170
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.200899124145508,
+      "learning_rate": 1.0620289855072465e-06,
+      "loss": 1.5964,
+      "step": 33180
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 10.006152153015137,
+      "learning_rate": 1.0562318840579711e-06,
+      "loss": 1.541,
+      "step": 33190
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.9720447063446045,
+      "learning_rate": 1.0504347826086958e-06,
+      "loss": 1.5875,
+      "step": 33200
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.2912840843200684,
+      "learning_rate": 1.0446376811594204e-06,
+      "loss": 1.5263,
+      "step": 33210
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.361422538757324,
+      "learning_rate": 1.038840579710145e-06,
+      "loss": 1.6695,
+      "step": 33220
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.8953531980514526,
+      "learning_rate": 1.0330434782608697e-06,
+      "loss": 1.6833,
+      "step": 33230
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.783647537231445,
+      "learning_rate": 1.0272463768115941e-06,
+      "loss": 1.483,
+      "step": 33240
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 4.266748905181885,
+      "learning_rate": 1.021449275362319e-06,
+      "loss": 1.5473,
+      "step": 33250
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.104844093322754,
+      "learning_rate": 1.0156521739130436e-06,
+      "loss": 1.5915,
+      "step": 33260
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.9186527729034424,
+      "learning_rate": 1.0098550724637683e-06,
+      "loss": 1.6114,
+      "step": 33270
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.935939311981201,
+      "learning_rate": 1.0040579710144927e-06,
+      "loss": 1.4601,
+      "step": 33280
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 5.332785129547119,
+      "learning_rate": 9.982608695652175e-07,
+      "loss": 1.7318,
+      "step": 33290
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 7.0045485496521,
+      "learning_rate": 9.924637681159422e-07,
+      "loss": 1.5863,
+      "step": 33300
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 10.329780578613281,
+      "learning_rate": 9.866666666666668e-07,
+      "loss": 1.5792,
+      "step": 33310
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.1837329864501953,
+      "learning_rate": 9.808695652173912e-07,
+      "loss": 1.4957,
+      "step": 33320
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 14.454916954040527,
+      "learning_rate": 9.750724637681159e-07,
+      "loss": 1.6237,
+      "step": 33330
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.9528446197509766,
+      "learning_rate": 9.692753623188407e-07,
+      "loss": 1.6035,
+      "step": 33340
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.9662082195281982,
+      "learning_rate": 9.634782608695654e-07,
+      "loss": 1.5402,
+      "step": 33350
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.538402557373047,
+      "learning_rate": 9.576811594202898e-07,
+      "loss": 1.6336,
+      "step": 33360
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.312148571014404,
+      "learning_rate": 9.518840579710146e-07,
+      "loss": 1.5634,
+      "step": 33370
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.3434953689575195,
+      "learning_rate": 9.460869565217393e-07,
+      "loss": 1.587,
+      "step": 33380
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 4.069890975952148,
+      "learning_rate": 9.402898550724638e-07,
+      "loss": 1.5212,
+      "step": 33390
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 19.578651428222656,
+      "learning_rate": 9.344927536231885e-07,
+      "loss": 1.6608,
+      "step": 33400
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 5.462759494781494,
+      "learning_rate": 9.286956521739131e-07,
+      "loss": 1.6739,
+      "step": 33410
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.227196216583252,
+      "learning_rate": 9.228985507246377e-07,
+      "loss": 1.6326,
+      "step": 33420
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.722791910171509,
+      "learning_rate": 9.171014492753624e-07,
+      "loss": 1.6126,
+      "step": 33430
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 15.881592750549316,
+      "learning_rate": 9.11304347826087e-07,
+      "loss": 1.7239,
+      "step": 33440
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 6.390254020690918,
+      "learning_rate": 9.055072463768117e-07,
+      "loss": 1.6419,
+      "step": 33450
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.134054660797119,
+      "learning_rate": 8.997101449275362e-07,
+      "loss": 1.6066,
+      "step": 33460
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 5.273614406585693,
+      "learning_rate": 8.93913043478261e-07,
+      "loss": 1.6421,
+      "step": 33470
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 4.4810967445373535,
+      "learning_rate": 8.881159420289856e-07,
+      "loss": 1.705,
+      "step": 33480
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.226357460021973,
+      "learning_rate": 8.823188405797103e-07,
+      "loss": 1.5615,
+      "step": 33490
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.2726452350616455,
+      "learning_rate": 8.765217391304348e-07,
+      "loss": 1.713,
+      "step": 33500
+    },
+    {
+      "epoch": 2.79,
+      "eval_loss": 1.6301745176315308,
+      "eval_runtime": 107.5184,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 33500
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 9.75317096710205,
+      "learning_rate": 8.707246376811595e-07,
+      "loss": 1.6305,
+      "step": 33510
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 8.24269962310791,
+      "learning_rate": 8.649275362318842e-07,
+      "loss": 1.6059,
+      "step": 33520
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 7.880953788757324,
+      "learning_rate": 8.591304347826088e-07,
+      "loss": 1.6878,
+      "step": 33530
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.4583754539489746,
+      "learning_rate": 8.533333333333334e-07,
+      "loss": 1.7179,
+      "step": 33540
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.8554558753967285,
+      "learning_rate": 8.47536231884058e-07,
+      "loss": 1.535,
+      "step": 33550
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 5.77476167678833,
+      "learning_rate": 8.417391304347827e-07,
+      "loss": 1.6241,
+      "step": 33560
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.786972999572754,
+      "learning_rate": 8.359420289855074e-07,
+      "loss": 1.6576,
+      "step": 33570
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.2922704219818115,
+      "learning_rate": 8.301449275362319e-07,
+      "loss": 1.7354,
+      "step": 33580
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.593795776367188,
+      "learning_rate": 8.243478260869566e-07,
+      "loss": 1.5815,
+      "step": 33590
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.0302324295043945,
+      "learning_rate": 8.185507246376813e-07,
+      "loss": 1.6433,
+      "step": 33600
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 10.300827026367188,
+      "learning_rate": 8.127536231884059e-07,
+      "loss": 1.613,
+      "step": 33610
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.340428352355957,
+      "learning_rate": 8.069565217391305e-07,
+      "loss": 1.5346,
+      "step": 33620
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 6.048191547393799,
+      "learning_rate": 8.011594202898551e-07,
+      "loss": 1.5197,
+      "step": 33630
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.3901071548461914,
+      "learning_rate": 7.953623188405798e-07,
+      "loss": 1.4772,
+      "step": 33640
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.204463481903076,
+      "learning_rate": 7.895652173913045e-07,
+      "loss": 1.6636,
+      "step": 33650
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.664523124694824,
+      "learning_rate": 7.83768115942029e-07,
+      "loss": 1.5472,
+      "step": 33660
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.5395827293396,
+      "learning_rate": 7.779710144927537e-07,
+      "loss": 1.6717,
+      "step": 33670
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.011569976806641,
+      "learning_rate": 7.721739130434783e-07,
+      "loss": 1.7751,
+      "step": 33680
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 6.7899556159973145,
+      "learning_rate": 7.66376811594203e-07,
+      "loss": 1.7148,
+      "step": 33690
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 5.4738054275512695,
+      "learning_rate": 7.605797101449276e-07,
+      "loss": 1.4787,
+      "step": 33700
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 8.042850494384766,
+      "learning_rate": 7.547826086956522e-07,
+      "loss": 1.675,
+      "step": 33710
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 13.439699172973633,
+      "learning_rate": 7.489855072463768e-07,
+      "loss": 1.6719,
+      "step": 33720
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.596892833709717,
+      "learning_rate": 7.431884057971014e-07,
+      "loss": 1.6386,
+      "step": 33730
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 4.695061683654785,
+      "learning_rate": 7.373913043478262e-07,
+      "loss": 1.6444,
+      "step": 33740
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.027362108230591,
+      "learning_rate": 7.315942028985508e-07,
+      "loss": 1.6984,
+      "step": 33750
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.8642385005950928,
+      "learning_rate": 7.257971014492753e-07,
+      "loss": 1.5937,
+      "step": 33760
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 2.800645112991333,
+      "learning_rate": 7.2e-07,
+      "loss": 1.5741,
+      "step": 33770
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 10.626157760620117,
+      "learning_rate": 7.142028985507247e-07,
+      "loss": 1.5336,
+      "step": 33780
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 9.157116889953613,
+      "learning_rate": 7.084057971014494e-07,
+      "loss": 1.5535,
+      "step": 33790
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 7.302629470825195,
+      "learning_rate": 7.026086956521739e-07,
+      "loss": 1.5688,
+      "step": 33800
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.5727436542510986,
+      "learning_rate": 6.968115942028986e-07,
+      "loss": 1.6339,
+      "step": 33810
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 16.979459762573242,
+      "learning_rate": 6.910144927536233e-07,
+      "loss": 1.5552,
+      "step": 33820
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.843989133834839,
+      "learning_rate": 6.852173913043479e-07,
+      "loss": 1.5359,
+      "step": 33830
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.522932767868042,
+      "learning_rate": 6.794202898550725e-07,
+      "loss": 1.7011,
+      "step": 33840
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 5.601334095001221,
+      "learning_rate": 6.736231884057971e-07,
+      "loss": 1.6746,
+      "step": 33850
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.354949712753296,
+      "learning_rate": 6.678260869565218e-07,
+      "loss": 1.589,
+      "step": 33860
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.3252222537994385,
+      "learning_rate": 6.620289855072465e-07,
+      "loss": 1.5887,
+      "step": 33870
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.7831883430480957,
+      "learning_rate": 6.56231884057971e-07,
+      "loss": 1.5805,
+      "step": 33880
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 8.868619918823242,
+      "learning_rate": 6.504347826086957e-07,
+      "loss": 1.6854,
+      "step": 33890
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 7.663196563720703,
+      "learning_rate": 6.446376811594203e-07,
+      "loss": 1.7296,
+      "step": 33900
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.567215919494629,
+      "learning_rate": 6.388405797101451e-07,
+      "loss": 1.5271,
+      "step": 33910
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.213868141174316,
+      "learning_rate": 6.330434782608696e-07,
+      "loss": 1.4958,
+      "step": 33920
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.9527573585510254,
+      "learning_rate": 6.272463768115942e-07,
+      "loss": 1.5046,
+      "step": 33930
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.8709187507629395,
+      "learning_rate": 6.214492753623189e-07,
+      "loss": 1.5993,
+      "step": 33940
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 17.194610595703125,
+      "learning_rate": 6.156521739130435e-07,
+      "loss": 1.6132,
+      "step": 33950
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 6.302470684051514,
+      "learning_rate": 6.098550724637682e-07,
+      "loss": 1.6321,
+      "step": 33960
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.782871127128601,
+      "learning_rate": 6.040579710144928e-07,
+      "loss": 1.4957,
+      "step": 33970
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.105221748352051,
+      "learning_rate": 5.982608695652174e-07,
+      "loss": 1.6038,
+      "step": 33980
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.2038047313690186,
+      "learning_rate": 5.924637681159421e-07,
+      "loss": 1.454,
+      "step": 33990
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 11.589030265808105,
+      "learning_rate": 5.866666666666667e-07,
+      "loss": 1.583,
+      "step": 34000
+    },
+    {
+      "epoch": 2.83,
+      "eval_loss": 1.6165950298309326,
+      "eval_runtime": 107.5234,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 34000
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 5.653379917144775,
+      "learning_rate": 5.808695652173914e-07,
+      "loss": 1.5756,
+      "step": 34010
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.994700908660889,
+      "learning_rate": 5.75072463768116e-07,
+      "loss": 1.6044,
+      "step": 34020
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.27285099029541,
+      "learning_rate": 5.692753623188407e-07,
+      "loss": 1.5561,
+      "step": 34030
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.3184444904327393,
+      "learning_rate": 5.634782608695653e-07,
+      "loss": 1.5874,
+      "step": 34040
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.3386433124542236,
+      "learning_rate": 5.576811594202898e-07,
+      "loss": 1.6027,
+      "step": 34050
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 9.001840591430664,
+      "learning_rate": 5.518840579710146e-07,
+      "loss": 1.5911,
+      "step": 34060
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 11.12852954864502,
+      "learning_rate": 5.460869565217391e-07,
+      "loss": 1.5201,
+      "step": 34070
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 5.407654285430908,
+      "learning_rate": 5.402898550724639e-07,
+      "loss": 1.5665,
+      "step": 34080
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 4.4106831550598145,
+      "learning_rate": 5.344927536231884e-07,
+      "loss": 1.7836,
+      "step": 34090
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 8.831493377685547,
+      "learning_rate": 5.286956521739131e-07,
+      "loss": 1.5497,
+      "step": 34100
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.7488794326782227,
+      "learning_rate": 5.228985507246377e-07,
+      "loss": 1.6616,
+      "step": 34110
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.5988351702690125,
+      "learning_rate": 5.171014492753624e-07,
+      "loss": 1.4637,
+      "step": 34120
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 7.661751747131348,
+      "learning_rate": 5.11304347826087e-07,
+      "loss": 1.6349,
+      "step": 34130
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.7503852844238281,
+      "learning_rate": 5.055072463768116e-07,
+      "loss": 1.7236,
+      "step": 34140
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.78981876373291,
+      "learning_rate": 4.997101449275362e-07,
+      "loss": 1.6712,
+      "step": 34150
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.7767815589904785,
+      "learning_rate": 4.939130434782609e-07,
+      "loss": 1.5735,
+      "step": 34160
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 8.948434829711914,
+      "learning_rate": 4.881159420289855e-07,
+      "loss": 1.4895,
+      "step": 34170
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.406666278839111,
+      "learning_rate": 4.823188405797102e-07,
+      "loss": 1.5253,
+      "step": 34180
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.939326524734497,
+      "learning_rate": 4.7652173913043486e-07,
+      "loss": 1.6299,
+      "step": 34190
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 12.841042518615723,
+      "learning_rate": 4.7072463768115945e-07,
+      "loss": 1.6818,
+      "step": 34200
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 2.7751150131225586,
+      "learning_rate": 4.6492753623188414e-07,
+      "loss": 1.7415,
+      "step": 34210
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 5.986341953277588,
+      "learning_rate": 4.5913043478260873e-07,
+      "loss": 1.5771,
+      "step": 34220
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 7.021378517150879,
+      "learning_rate": 4.533333333333334e-07,
+      "loss": 1.4319,
+      "step": 34230
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 6.669992446899414,
+      "learning_rate": 4.47536231884058e-07,
+      "loss": 1.6399,
+      "step": 34240
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.224604845046997,
+      "learning_rate": 4.417391304347826e-07,
+      "loss": 1.7313,
+      "step": 34250
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 4.330286026000977,
+      "learning_rate": 4.359420289855073e-07,
+      "loss": 1.658,
+      "step": 34260
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.8469319343566895,
+      "learning_rate": 4.301449275362319e-07,
+      "loss": 1.5495,
+      "step": 34270
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5525469779968262,
+      "learning_rate": 4.243478260869566e-07,
+      "loss": 1.5885,
+      "step": 34280
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.59576416015625,
+      "learning_rate": 4.1855072463768116e-07,
+      "loss": 1.6725,
+      "step": 34290
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 14.655567169189453,
+      "learning_rate": 4.1275362318840586e-07,
+      "loss": 1.4917,
+      "step": 34300
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.645742416381836,
+      "learning_rate": 4.0695652173913044e-07,
+      "loss": 1.5779,
+      "step": 34310
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.458324670791626,
+      "learning_rate": 4.0115942028985514e-07,
+      "loss": 1.6004,
+      "step": 34320
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.773531913757324,
+      "learning_rate": 3.9536231884057973e-07,
+      "loss": 1.504,
+      "step": 34330
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 4.06483268737793,
+      "learning_rate": 3.8956521739130437e-07,
+      "loss": 1.6038,
+      "step": 34340
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.4933018684387207,
+      "learning_rate": 3.83768115942029e-07,
+      "loss": 1.6455,
+      "step": 34350
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.6547389030456543,
+      "learning_rate": 3.7797101449275365e-07,
+      "loss": 1.5391,
+      "step": 34360
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.740145683288574,
+      "learning_rate": 3.721739130434783e-07,
+      "loss": 1.6146,
+      "step": 34370
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.028878688812256,
+      "learning_rate": 3.6637681159420293e-07,
+      "loss": 1.5921,
+      "step": 34380
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 6.614234447479248,
+      "learning_rate": 3.6057971014492757e-07,
+      "loss": 1.5552,
+      "step": 34390
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.7156872749328613,
+      "learning_rate": 3.547826086956522e-07,
+      "loss": 1.6545,
+      "step": 34400
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.235097646713257,
+      "learning_rate": 3.4898550724637685e-07,
+      "loss": 1.5333,
+      "step": 34410
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.5841474533081055,
+      "learning_rate": 3.431884057971015e-07,
+      "loss": 1.6009,
+      "step": 34420
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.2173547744750977,
+      "learning_rate": 3.3739130434782614e-07,
+      "loss": 1.5737,
+      "step": 34430
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.977945327758789,
+      "learning_rate": 3.315942028985508e-07,
+      "loss": 1.6099,
+      "step": 34440
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.717473983764648,
+      "learning_rate": 3.2579710144927537e-07,
+      "loss": 1.6938,
+      "step": 34450
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.8826589584350586,
+      "learning_rate": 3.2e-07,
+      "loss": 1.5333,
+      "step": 34460
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.7628260850906372,
+      "learning_rate": 3.1420289855072465e-07,
+      "loss": 1.4657,
+      "step": 34470
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.397793769836426,
+      "learning_rate": 3.084057971014493e-07,
+      "loss": 1.7186,
+      "step": 34480
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 13.669360160827637,
+      "learning_rate": 3.0260869565217393e-07,
+      "loss": 1.6144,
+      "step": 34490
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.7966532707214355,
+      "learning_rate": 2.9681159420289857e-07,
+      "loss": 1.7335,
+      "step": 34500
+    },
+    {
+      "epoch": 2.88,
+      "eval_loss": 1.6453466415405273,
+      "eval_runtime": 107.5244,
+      "eval_samples_per_second": 9.3,
+      "eval_steps_per_second": 2.325,
+      "step": 34500
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.828636646270752,
+      "learning_rate": 2.910144927536232e-07,
+      "loss": 1.4402,
+      "step": 34510
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.665623664855957,
+      "learning_rate": 2.8521739130434785e-07,
+      "loss": 1.5562,
+      "step": 34520
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 7.800908088684082,
+      "learning_rate": 2.794202898550725e-07,
+      "loss": 1.5813,
+      "step": 34530
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.734397888183594,
+      "learning_rate": 2.7362318840579713e-07,
+      "loss": 1.7721,
+      "step": 34540
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.1839168071746826,
+      "learning_rate": 2.678260869565218e-07,
+      "loss": 1.5236,
+      "step": 34550
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.9042229652404785,
+      "learning_rate": 2.620289855072464e-07,
+      "loss": 1.6205,
+      "step": 34560
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.442493438720703,
+      "learning_rate": 2.5623188405797106e-07,
+      "loss": 1.6334,
+      "step": 34570
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 11.60261058807373,
+      "learning_rate": 2.504347826086957e-07,
+      "loss": 1.6631,
+      "step": 34580
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.5220028162002563,
+      "learning_rate": 2.4463768115942034e-07,
+      "loss": 1.5903,
+      "step": 34590
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 6.463686466217041,
+      "learning_rate": 2.3884057971014493e-07,
+      "loss": 1.505,
+      "step": 34600
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.3802415132522583,
+      "learning_rate": 2.3304347826086957e-07,
+      "loss": 1.5574,
+      "step": 34610
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 5.931606292724609,
+      "learning_rate": 2.272463768115942e-07,
+      "loss": 1.6049,
+      "step": 34620
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.055948734283447,
+      "learning_rate": 2.2144927536231885e-07,
+      "loss": 1.7223,
+      "step": 34630
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.3893649578094482,
+      "learning_rate": 2.156521739130435e-07,
+      "loss": 1.6336,
+      "step": 34640
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.007665157318115,
+      "learning_rate": 2.0985507246376813e-07,
+      "loss": 1.5563,
+      "step": 34650
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.499795913696289,
+      "learning_rate": 2.0405797101449277e-07,
+      "loss": 1.6813,
+      "step": 34660
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.31160569190979,
+      "learning_rate": 1.9826086956521742e-07,
+      "loss": 1.5491,
+      "step": 34670
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 2.493201732635498,
+      "learning_rate": 1.9246376811594206e-07,
+      "loss": 1.7265,
+      "step": 34680
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 6.865158557891846,
+      "learning_rate": 1.866666666666667e-07,
+      "loss": 1.6564,
+      "step": 34690
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 2.9603042602539062,
+      "learning_rate": 1.808695652173913e-07,
+      "loss": 1.6632,
+      "step": 34700
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 10.82306957244873,
+      "learning_rate": 1.7507246376811595e-07,
+      "loss": 1.555,
+      "step": 34710
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 4.0846428871154785,
+      "learning_rate": 1.692753623188406e-07,
+      "loss": 1.6608,
+      "step": 34720
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 5.046102523803711,
+      "learning_rate": 1.6347826086956523e-07,
+      "loss": 1.6355,
+      "step": 34730
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 5.843890190124512,
+      "learning_rate": 1.5768115942028988e-07,
+      "loss": 1.7457,
+      "step": 34740
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.0817549228668213,
+      "learning_rate": 1.5188405797101452e-07,
+      "loss": 1.5093,
+      "step": 34750
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.4835991859436035,
+      "learning_rate": 1.4608695652173916e-07,
+      "loss": 1.5516,
+      "step": 34760
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.763504505157471,
+      "learning_rate": 1.402898550724638e-07,
+      "loss": 1.5014,
+      "step": 34770
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 7.537784576416016,
+      "learning_rate": 1.344927536231884e-07,
+      "loss": 1.5698,
+      "step": 34780
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.3694698810577393,
+      "learning_rate": 1.2869565217391305e-07,
+      "loss": 1.7388,
+      "step": 34790
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.1657655239105225,
+      "learning_rate": 1.228985507246377e-07,
+      "loss": 1.6657,
+      "step": 34800
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.10614013671875,
+      "learning_rate": 1.1710144927536234e-07,
+      "loss": 1.6406,
+      "step": 34810
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 8.151602745056152,
+      "learning_rate": 1.1130434782608698e-07,
+      "loss": 1.7237,
+      "step": 34820
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.7785089015960693,
+      "learning_rate": 1.0550724637681159e-07,
+      "loss": 1.6641,
+      "step": 34830
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.2298667430877686,
+      "learning_rate": 9.971014492753623e-08,
+      "loss": 1.659,
+      "step": 34840
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 4.855597972869873,
+      "learning_rate": 9.391304347826087e-08,
+      "loss": 1.5468,
+      "step": 34850
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 7.356712818145752,
+      "learning_rate": 8.811594202898551e-08,
+      "loss": 1.5559,
+      "step": 34860
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.330800533294678,
+      "learning_rate": 8.231884057971016e-08,
+      "loss": 1.5451,
+      "step": 34870
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 4.4921183586120605,
+      "learning_rate": 7.65217391304348e-08,
+      "loss": 1.6324,
+      "step": 34880
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.110346794128418,
+      "learning_rate": 7.072463768115942e-08,
+      "loss": 1.5751,
+      "step": 34890
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 5.403293609619141,
+      "learning_rate": 6.492753623188407e-08,
+      "loss": 1.6089,
+      "step": 34900
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 11.329370498657227,
+      "learning_rate": 5.9130434782608707e-08,
+      "loss": 1.7421,
+      "step": 34910
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.819499969482422,
+      "learning_rate": 5.3333333333333334e-08,
+      "loss": 1.6757,
+      "step": 34920
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.4054958820343018,
+      "learning_rate": 4.7536231884057975e-08,
+      "loss": 1.6065,
+      "step": 34930
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.741161823272705,
+      "learning_rate": 4.173913043478261e-08,
+      "loss": 1.705,
+      "step": 34940
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 12.431832313537598,
+      "learning_rate": 3.594202898550725e-08,
+      "loss": 1.637,
+      "step": 34950
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.2292697429656982,
+      "learning_rate": 3.0144927536231885e-08,
+      "loss": 1.6626,
+      "step": 34960
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 6.705583095550537,
+      "learning_rate": 2.4347826086956523e-08,
+      "loss": 1.7878,
+      "step": 34970
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 4.077023506164551,
+      "learning_rate": 1.855072463768116e-08,
+      "loss": 1.5233,
+      "step": 34980
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.2491979598999023,
+      "learning_rate": 1.2753623188405798e-08,
+      "loss": 1.6806,
+      "step": 34990
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 5.141303062438965,
+      "learning_rate": 6.956521739130436e-09,
+      "loss": 1.6406,
+      "step": 35000
+    },
+    {
+      "epoch": 2.92,
+      "eval_loss": 1.6195024251937866,
+      "eval_runtime": 107.5117,
+      "eval_samples_per_second": 9.301,
+      "eval_steps_per_second": 2.325,
+      "step": 35000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 35000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 5.6357440978944e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}