diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17882 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 19.0,
+  "eval_steps": 133,
+  "global_step": 2527,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.007518796992481203,
+      "grad_norm": 3.9574085902303486,
+      "learning_rate": 1.5151515151515152e-08,
+      "loss": 0.7852,
+      "step": 1
+    },
+    {
+      "epoch": 0.007518796992481203,
+      "eval_loss": 0.5364012122154236,
+      "eval_runtime": 1.7189,
+      "eval_samples_per_second": 190.242,
+      "eval_steps_per_second": 3.491,
+      "step": 1
+    },
+    {
+      "epoch": 0.015037593984962405,
+      "grad_norm": 4.152977194472038,
+      "learning_rate": 3.0303030303030305e-08,
+      "loss": 0.8071,
+      "step": 2
+    },
+    {
+      "epoch": 0.022556390977443608,
+      "grad_norm": 4.269466427516056,
+      "learning_rate": 4.545454545454545e-08,
+      "loss": 0.8131,
+      "step": 3
+    },
+    {
+      "epoch": 0.03007518796992481,
+      "grad_norm": 4.1964962162371044,
+      "learning_rate": 6.060606060606061e-08,
+      "loss": 0.8171,
+      "step": 4
+    },
+    {
+      "epoch": 0.03759398496240601,
+      "grad_norm": 4.30206270033078,
+      "learning_rate": 7.575757575757576e-08,
+      "loss": 0.823,
+      "step": 5
+    },
+    {
+      "epoch": 0.045112781954887216,
+      "grad_norm": 4.307545934750609,
+      "learning_rate": 9.09090909090909e-08,
+      "loss": 0.8088,
+      "step": 6
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.193976900412803,
+      "learning_rate": 1.0606060606060605e-07,
+      "loss": 0.8009,
+      "step": 7
+    },
+    {
+      "epoch": 0.06015037593984962,
+      "grad_norm": 4.009889950503695,
+      "learning_rate": 1.2121212121212122e-07,
+      "loss": 0.7987,
+      "step": 8
+    },
+    {
+      "epoch": 0.06766917293233082,
+      "grad_norm": 4.1316532774783,
+      "learning_rate": 1.3636363636363635e-07,
+      "loss": 0.8099,
+      "step": 9
+    },
+    {
+      "epoch": 0.07518796992481203,
+      "grad_norm": 4.165348106357401,
+      "learning_rate": 1.5151515151515152e-07,
+      "loss": 0.7966,
+      "step": 10
+    },
+    {
+      "epoch": 0.08270676691729323,
+      "grad_norm": 4.3132563073138686,
+      "learning_rate": 1.6666666666666665e-07,
+      "loss": 0.805,
+      "step": 11
+    },
+    {
+      "epoch": 0.09022556390977443,
+      "grad_norm": 4.136784903554356,
+      "learning_rate": 1.818181818181818e-07,
+      "loss": 0.8019,
+      "step": 12
+    },
+    {
+      "epoch": 0.09774436090225563,
+      "grad_norm": 4.264349126909646,
+      "learning_rate": 1.9696969696969696e-07,
+      "loss": 0.8109,
+      "step": 13
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 4.093651237952913,
+      "learning_rate": 2.121212121212121e-07,
+      "loss": 0.7985,
+      "step": 14
+    },
+    {
+      "epoch": 0.11278195488721804,
+      "grad_norm": 4.090865167402975,
+      "learning_rate": 2.2727272727272726e-07,
+      "loss": 0.7961,
+      "step": 15
+    },
+    {
+      "epoch": 0.12030075187969924,
+      "grad_norm": 4.0848586658316695,
+      "learning_rate": 2.4242424242424244e-07,
+      "loss": 0.7868,
+      "step": 16
+    },
+    {
+      "epoch": 0.12781954887218044,
+      "grad_norm": 4.106021663821251,
+      "learning_rate": 2.5757575757575754e-07,
+      "loss": 0.7961,
+      "step": 17
+    },
+    {
+      "epoch": 0.13533834586466165,
+      "grad_norm": 3.9797472957653444,
+      "learning_rate": 2.727272727272727e-07,
+      "loss": 0.7916,
+      "step": 18
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 4.234623519952662,
+      "learning_rate": 2.878787878787879e-07,
+      "loss": 0.8019,
+      "step": 19
+    },
+    {
+      "epoch": 0.15037593984962405,
+      "grad_norm": 4.183079244524941,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 0.8121,
+      "step": 20
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 4.058184851918076,
+      "learning_rate": 3.1818181818181815e-07,
+      "loss": 0.806,
+      "step": 21
+    },
+    {
+      "epoch": 0.16541353383458646,
+      "grad_norm": 4.050776781622013,
+      "learning_rate": 3.333333333333333e-07,
+      "loss": 0.7881,
+      "step": 22
+    },
+    {
+      "epoch": 0.17293233082706766,
+      "grad_norm": 4.071829473331961,
+      "learning_rate": 3.484848484848485e-07,
+      "loss": 0.7951,
+      "step": 23
+    },
+    {
+      "epoch": 0.18045112781954886,
+      "grad_norm": 3.887017041491382,
+      "learning_rate": 3.636363636363636e-07,
+      "loss": 0.7796,
+      "step": 24
+    },
+    {
+      "epoch": 0.18796992481203006,
+      "grad_norm": 3.7070163551391544,
+      "learning_rate": 3.7878787878787876e-07,
+      "loss": 0.7941,
+      "step": 25
+    },
+    {
+      "epoch": 0.19548872180451127,
+      "grad_norm": 3.738331706257787,
+      "learning_rate": 3.939393939393939e-07,
+      "loss": 0.7751,
+      "step": 26
+    },
+    {
+      "epoch": 0.20300751879699247,
+      "grad_norm": 3.781402761368936,
+      "learning_rate": 4.090909090909091e-07,
+      "loss": 0.788,
+      "step": 27
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 3.806056278029616,
+      "learning_rate": 4.242424242424242e-07,
+      "loss": 0.7698,
+      "step": 28
+    },
+    {
+      "epoch": 0.21804511278195488,
+      "grad_norm": 3.5728448565789828,
+      "learning_rate": 4.3939393939393937e-07,
+      "loss": 0.7622,
+      "step": 29
+    },
+    {
+      "epoch": 0.22556390977443608,
+      "grad_norm": 3.6232601345193864,
+      "learning_rate": 4.545454545454545e-07,
+      "loss": 0.7665,
+      "step": 30
+    },
+    {
+      "epoch": 0.23308270676691728,
+      "grad_norm": 3.594864819018081,
+      "learning_rate": 4.696969696969697e-07,
+      "loss": 0.7653,
+      "step": 31
+    },
+    {
+      "epoch": 0.24060150375939848,
+      "grad_norm": 3.5778962556257072,
+      "learning_rate": 4.848484848484849e-07,
+      "loss": 0.7634,
+      "step": 32
+    },
+    {
+      "epoch": 0.24812030075187969,
+      "grad_norm": 3.6429629635488663,
+      "learning_rate": 5e-07,
+      "loss": 0.7664,
+      "step": 33
+    },
+    {
+      "epoch": 0.2556390977443609,
+      "grad_norm": 3.0458880303947655,
+      "learning_rate": 5.151515151515151e-07,
+      "loss": 0.7126,
+      "step": 34
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 2.88712025714805,
+      "learning_rate": 5.303030303030303e-07,
+      "loss": 0.6935,
+      "step": 35
+    },
+    {
+      "epoch": 0.2706766917293233,
+      "grad_norm": 2.9700087736058487,
+      "learning_rate": 5.454545454545454e-07,
+      "loss": 0.7066,
+      "step": 36
+    },
+    {
+      "epoch": 0.2781954887218045,
+      "grad_norm": 2.9845380270132345,
+      "learning_rate": 5.606060606060605e-07,
+      "loss": 0.7085,
+      "step": 37
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 2.970880877241169,
+      "learning_rate": 5.757575757575758e-07,
+      "loss": 0.7102,
+      "step": 38
+    },
+    {
+      "epoch": 0.2932330827067669,
+      "grad_norm": 2.929604687958302,
+      "learning_rate": 5.909090909090909e-07,
+      "loss": 0.7018,
+      "step": 39
+    },
+    {
+      "epoch": 0.3007518796992481,
+      "grad_norm": 2.8694688045311456,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.6924,
+      "step": 40
+    },
+    {
+      "epoch": 0.3082706766917293,
+      "grad_norm": 2.9817739826423617,
+      "learning_rate": 6.212121212121212e-07,
+      "loss": 0.7108,
+      "step": 41
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 2.914047632531219,
+      "learning_rate": 6.363636363636363e-07,
+      "loss": 0.6924,
+      "step": 42
+    },
+    {
+      "epoch": 0.3233082706766917,
+      "grad_norm": 2.9274521249374845,
+      "learning_rate": 6.515151515151515e-07,
+      "loss": 0.6917,
+      "step": 43
+    },
+    {
+      "epoch": 0.3308270676691729,
+      "grad_norm": 2.9620633096203557,
+      "learning_rate": 6.666666666666666e-07,
+      "loss": 0.6706,
+      "step": 44
+    },
+    {
+      "epoch": 0.3383458646616541,
+      "grad_norm": 2.821343903541344,
+      "learning_rate": 6.818181818181817e-07,
+      "loss": 0.6594,
+      "step": 45
+    },
+    {
+      "epoch": 0.3458646616541353,
+      "grad_norm": 2.3858334153306133,
+      "learning_rate": 6.96969696969697e-07,
+      "loss": 0.6187,
+      "step": 46
+    },
+    {
+      "epoch": 0.3533834586466165,
+      "grad_norm": 2.1380956194098375,
+      "learning_rate": 7.121212121212121e-07,
+      "loss": 0.5962,
+      "step": 47
+    },
+    {
+      "epoch": 0.3609022556390977,
+      "grad_norm": 2.082345655853642,
+      "learning_rate": 7.272727272727272e-07,
+      "loss": 0.6053,
+      "step": 48
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 2.007894727709351,
+      "learning_rate": 7.424242424242424e-07,
+      "loss": 0.5812,
+      "step": 49
+    },
+    {
+      "epoch": 0.37593984962406013,
+      "grad_norm": 1.9999733254720435,
+      "learning_rate": 7.575757575757575e-07,
+      "loss": 0.5758,
+      "step": 50
+    },
+    {
+      "epoch": 0.38345864661654133,
+      "grad_norm": 1.9749427508560604,
+      "learning_rate": 7.727272727272727e-07,
+      "loss": 0.5651,
+      "step": 51
+    },
+    {
+      "epoch": 0.39097744360902253,
+      "grad_norm": 1.9952022079088583,
+      "learning_rate": 7.878787878787878e-07,
+      "loss": 0.5688,
+      "step": 52
+    },
+    {
+      "epoch": 0.39849624060150374,
+      "grad_norm": 1.97503990427937,
+      "learning_rate": 8.030303030303029e-07,
+      "loss": 0.5731,
+      "step": 53
+    },
+    {
+      "epoch": 0.40601503759398494,
+      "grad_norm": 1.8528704242680778,
+      "learning_rate": 8.181818181818182e-07,
+      "loss": 0.5582,
+      "step": 54
+    },
+    {
+      "epoch": 0.41353383458646614,
+      "grad_norm": 1.9175926204992202,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.5718,
+      "step": 55
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.8543235931575899,
+      "learning_rate": 8.484848484848484e-07,
+      "loss": 0.5592,
+      "step": 56
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.7508834138788243,
+      "learning_rate": 8.636363636363636e-07,
+      "loss": 0.5638,
+      "step": 57
+    },
+    {
+      "epoch": 0.43609022556390975,
+      "grad_norm": 1.7366194416919294,
+      "learning_rate": 8.787878787878787e-07,
+      "loss": 0.5482,
+      "step": 58
+    },
+    {
+      "epoch": 0.44360902255639095,
+      "grad_norm": 1.659920899105035,
+      "learning_rate": 8.939393939393938e-07,
+      "loss": 0.5442,
+      "step": 59
+    },
+    {
+      "epoch": 0.45112781954887216,
+      "grad_norm": 1.575527744126036,
+      "learning_rate": 9.09090909090909e-07,
+      "loss": 0.5274,
+      "step": 60
+    },
+    {
+      "epoch": 0.45864661654135336,
+      "grad_norm": 1.4235369002904827,
+      "learning_rate": 9.242424242424241e-07,
+      "loss": 0.5262,
+      "step": 61
+    },
+    {
+      "epoch": 0.46616541353383456,
+      "grad_norm": 1.2496356311046746,
+      "learning_rate": 9.393939393939395e-07,
+      "loss": 0.5206,
+      "step": 62
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.1000085568339923,
+      "learning_rate": 9.545454545454546e-07,
+      "loss": 0.5049,
+      "step": 63
+    },
+    {
+      "epoch": 0.48120300751879697,
+      "grad_norm": 1.0741625426091794,
+      "learning_rate": 9.696969696969698e-07,
+      "loss": 0.5086,
+      "step": 64
+    },
+    {
+      "epoch": 0.48872180451127817,
+      "grad_norm": 0.9903127899591331,
+      "learning_rate": 9.848484848484847e-07,
+      "loss": 0.4972,
+      "step": 65
+    },
+    {
+      "epoch": 0.49624060150375937,
+      "grad_norm": 0.9102001212678049,
+      "learning_rate": 1e-06,
+      "loss": 0.4831,
+      "step": 66
+    },
+    {
+      "epoch": 0.5037593984962406,
+      "grad_norm": 0.8579814331376515,
+      "learning_rate": 1.0151515151515152e-06,
+      "loss": 0.4719,
+      "step": 67
+    },
+    {
+      "epoch": 0.5112781954887218,
+      "grad_norm": 0.9025297083130085,
+      "learning_rate": 1.0303030303030302e-06,
+      "loss": 0.4899,
+      "step": 68
+    },
+    {
+      "epoch": 0.518796992481203,
+      "grad_norm": 0.766481103987309,
+      "learning_rate": 1.0454545454545454e-06,
+      "loss": 0.4637,
+      "step": 69
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.8024714712158572,
+      "learning_rate": 1.0606060606060606e-06,
+      "loss": 0.4697,
+      "step": 70
+    },
+    {
+      "epoch": 0.5338345864661654,
+      "grad_norm": 0.7354814210368072,
+      "learning_rate": 1.0757575757575756e-06,
+      "loss": 0.4753,
+      "step": 71
+    },
+    {
+      "epoch": 0.5413533834586466,
+      "grad_norm": 0.677941162133797,
+      "learning_rate": 1.0909090909090908e-06,
+      "loss": 0.4653,
+      "step": 72
+    },
+    {
+      "epoch": 0.5488721804511278,
+      "grad_norm": 0.7081746607392825,
+      "learning_rate": 1.106060606060606e-06,
+      "loss": 0.4616,
+      "step": 73
+    },
+    {
+      "epoch": 0.556390977443609,
+      "grad_norm": 0.6455062471842155,
+      "learning_rate": 1.121212121212121e-06,
+      "loss": 0.4613,
+      "step": 74
+    },
+    {
+      "epoch": 0.5639097744360902,
+      "grad_norm": 0.6028379671239867,
+      "learning_rate": 1.1363636363636364e-06,
+      "loss": 0.4455,
+      "step": 75
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.587266759347893,
+      "learning_rate": 1.1515151515151516e-06,
+      "loss": 0.456,
+      "step": 76
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.5706967003264505,
+      "learning_rate": 1.1666666666666668e-06,
+      "loss": 0.4554,
+      "step": 77
+    },
+    {
+      "epoch": 0.5864661654135338,
+      "grad_norm": 0.5844181517344095,
+      "learning_rate": 1.1818181818181818e-06,
+      "loss": 0.4563,
+      "step": 78
+    },
+    {
+      "epoch": 0.5939849624060151,
+      "grad_norm": 0.5090738582943606,
+      "learning_rate": 1.196969696969697e-06,
+      "loss": 0.4577,
+      "step": 79
+    },
+    {
+      "epoch": 0.6015037593984962,
+      "grad_norm": 0.5113675160685818,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.4575,
+      "step": 80
+    },
+    {
+      "epoch": 0.6090225563909775,
+      "grad_norm": 0.4785571923326216,
+      "learning_rate": 1.2272727272727272e-06,
+      "loss": 0.4426,
+      "step": 81
+    },
+    {
+      "epoch": 0.6165413533834586,
+      "grad_norm": 0.48139632215215283,
+      "learning_rate": 1.2424242424242424e-06,
+      "loss": 0.4632,
+      "step": 82
+    },
+    {
+      "epoch": 0.6240601503759399,
+      "grad_norm": 0.41816489414463653,
+      "learning_rate": 1.2575757575757576e-06,
+      "loss": 0.4232,
+      "step": 83
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.44124996636502223,
+      "learning_rate": 1.2727272727272726e-06,
+      "loss": 0.4325,
+      "step": 84
+    },
+    {
+      "epoch": 0.6390977443609023,
+      "grad_norm": 0.3998734716031043,
+      "learning_rate": 1.2878787878787878e-06,
+      "loss": 0.4178,
+      "step": 85
+    },
+    {
+      "epoch": 0.6466165413533834,
+      "grad_norm": 0.4011354713280704,
+      "learning_rate": 1.303030303030303e-06,
+      "loss": 0.4296,
+      "step": 86
+    },
+    {
+      "epoch": 0.6541353383458647,
+      "grad_norm": 0.36318360366948405,
+      "learning_rate": 1.318181818181818e-06,
+      "loss": 0.4398,
+      "step": 87
+    },
+    {
+      "epoch": 0.6616541353383458,
+      "grad_norm": 0.3768739367242804,
+      "learning_rate": 1.3333333333333332e-06,
+      "loss": 0.4354,
+      "step": 88
+    },
+    {
+      "epoch": 0.6691729323308271,
+      "grad_norm": 0.37215262280826256,
+      "learning_rate": 1.3484848484848484e-06,
+      "loss": 0.4329,
+      "step": 89
+    },
+    {
+      "epoch": 0.6766917293233082,
+      "grad_norm": 0.35820525434603706,
+      "learning_rate": 1.3636363636363634e-06,
+      "loss": 0.428,
+      "step": 90
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.3454536411045649,
+      "learning_rate": 1.3787878787878788e-06,
+      "loss": 0.4219,
+      "step": 91
+    },
+    {
+      "epoch": 0.6917293233082706,
+      "grad_norm": 0.34781590669974216,
+      "learning_rate": 1.393939393939394e-06,
+      "loss": 0.4135,
+      "step": 92
+    },
+    {
+      "epoch": 0.6992481203007519,
+      "grad_norm": 0.345098527101657,
+      "learning_rate": 1.409090909090909e-06,
+      "loss": 0.4254,
+      "step": 93
+    },
+    {
+      "epoch": 0.706766917293233,
+      "grad_norm": 0.3324646013142274,
+      "learning_rate": 1.4242424242424242e-06,
+      "loss": 0.4278,
+      "step": 94
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.34629001654483704,
+      "learning_rate": 1.4393939393939394e-06,
+      "loss": 0.4202,
+      "step": 95
+    },
+    {
+      "epoch": 0.7218045112781954,
+      "grad_norm": 0.3399738348931957,
+      "learning_rate": 1.4545454545454544e-06,
+      "loss": 0.4316,
+      "step": 96
+    },
+    {
+      "epoch": 0.7293233082706767,
+      "grad_norm": 0.317137489102149,
+      "learning_rate": 1.4696969696969696e-06,
+      "loss": 0.421,
+      "step": 97
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.30792167880140764,
+      "learning_rate": 1.4848484848484848e-06,
+      "loss": 0.4176,
+      "step": 98
+    },
+    {
+      "epoch": 0.7443609022556391,
+      "grad_norm": 0.30122408122192973,
+      "learning_rate": 1.5e-06,
+      "loss": 0.4115,
+      "step": 99
+    },
+    {
+      "epoch": 0.7518796992481203,
+      "grad_norm": 0.32578697399967477,
+      "learning_rate": 1.515151515151515e-06,
+      "loss": 0.4102,
+      "step": 100
+    },
+    {
+      "epoch": 0.7593984962406015,
+      "grad_norm": 0.3149775494672908,
+      "learning_rate": 1.5303030303030302e-06,
+      "loss": 0.4035,
+      "step": 101
+    },
+    {
+      "epoch": 0.7669172932330827,
+      "grad_norm": 0.3029210823657488,
+      "learning_rate": 1.5454545454545454e-06,
+      "loss": 0.4193,
+      "step": 102
+    },
+    {
+      "epoch": 0.7744360902255639,
+      "grad_norm": 0.3080280296204915,
+      "learning_rate": 1.5606060606060604e-06,
+      "loss": 0.416,
+      "step": 103
+    },
+    {
+      "epoch": 0.7819548872180451,
+      "grad_norm": 0.2871264310979226,
+      "learning_rate": 1.5757575757575756e-06,
+      "loss": 0.4191,
+      "step": 104
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.25676500736531105,
+      "learning_rate": 1.5909090909090908e-06,
+      "loss": 0.4025,
+      "step": 105
+    },
+    {
+      "epoch": 0.7969924812030075,
+      "grad_norm": 0.2553993306775827,
+      "learning_rate": 1.6060606060606058e-06,
+      "loss": 0.4014,
+      "step": 106
+    },
+    {
+      "epoch": 0.8045112781954887,
+      "grad_norm": 0.25199439392050066,
+      "learning_rate": 1.621212121212121e-06,
+      "loss": 0.3898,
+      "step": 107
+    },
+    {
+      "epoch": 0.8120300751879699,
+      "grad_norm": 0.25423232214885616,
+      "learning_rate": 1.6363636363636365e-06,
+      "loss": 0.4028,
+      "step": 108
+    },
+    {
+      "epoch": 0.8195488721804511,
+      "grad_norm": 0.2654888836125121,
+      "learning_rate": 1.6515151515151515e-06,
+      "loss": 0.4054,
+      "step": 109
+    },
+    {
+      "epoch": 0.8270676691729323,
+      "grad_norm": 0.2756893850917718,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.4018,
+      "step": 110
+    },
+    {
+      "epoch": 0.8345864661654135,
+      "grad_norm": 0.23960171360047544,
+      "learning_rate": 1.6818181818181819e-06,
+      "loss": 0.4081,
+      "step": 111
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.2413615125039257,
+      "learning_rate": 1.6969696969696969e-06,
+      "loss": 0.4057,
+      "step": 112
+    },
+    {
+      "epoch": 0.849624060150376,
+      "grad_norm": 0.24743801198836796,
+      "learning_rate": 1.712121212121212e-06,
+      "loss": 0.4024,
+      "step": 113
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.2415484883476495,
+      "learning_rate": 1.7272727272727273e-06,
+      "loss": 0.401,
+      "step": 114
+    },
+    {
+      "epoch": 0.8646616541353384,
+      "grad_norm": 0.24500246093342584,
+      "learning_rate": 1.7424242424242423e-06,
+      "loss": 0.4028,
+      "step": 115
+    },
+    {
+      "epoch": 0.8721804511278195,
+      "grad_norm": 0.2662830897185873,
+      "learning_rate": 1.7575757575757575e-06,
+      "loss": 0.4073,
+      "step": 116
+    },
+    {
+      "epoch": 0.8796992481203008,
+      "grad_norm": 0.23909822901815306,
+      "learning_rate": 1.7727272727272727e-06,
+      "loss": 0.4023,
+      "step": 117
+    },
+    {
+      "epoch": 0.8872180451127819,
+      "grad_norm": 0.2636288837692075,
+      "learning_rate": 1.7878787878787877e-06,
+      "loss": 0.42,
+      "step": 118
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.24645460321502913,
+      "learning_rate": 1.8030303030303029e-06,
+      "loss": 0.4086,
+      "step": 119
+    },
+    {
+      "epoch": 0.9022556390977443,
+      "grad_norm": 0.2372434612402873,
+      "learning_rate": 1.818181818181818e-06,
+      "loss": 0.3902,
+      "step": 120
+    },
+    {
+      "epoch": 0.9097744360902256,
+      "grad_norm": 0.22984882639021978,
+      "learning_rate": 1.833333333333333e-06,
+      "loss": 0.3817,
+      "step": 121
+    },
+    {
+      "epoch": 0.9172932330827067,
+      "grad_norm": 0.23412727392489194,
+      "learning_rate": 1.8484848484848483e-06,
+      "loss": 0.3882,
+      "step": 122
+    },
+    {
+      "epoch": 0.924812030075188,
+      "grad_norm": 0.24622959680743484,
+      "learning_rate": 1.8636363636363635e-06,
+      "loss": 0.4086,
+      "step": 123
+    },
+    {
+      "epoch": 0.9323308270676691,
+      "grad_norm": 0.22452028540322352,
+      "learning_rate": 1.878787878787879e-06,
+      "loss": 0.3915,
+      "step": 124
+    },
+    {
+      "epoch": 0.9398496240601504,
+      "grad_norm": 0.22047337702446065,
+      "learning_rate": 1.893939393939394e-06,
+      "loss": 0.3954,
+      "step": 125
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.2341253045607213,
+      "learning_rate": 1.909090909090909e-06,
+      "loss": 0.3991,
+      "step": 126
+    },
+    {
+      "epoch": 0.9548872180451128,
+      "grad_norm": 0.24600750032305052,
+      "learning_rate": 1.924242424242424e-06,
+      "loss": 0.3994,
+      "step": 127
+    },
+    {
+      "epoch": 0.9624060150375939,
+      "grad_norm": 0.26427181635552316,
+      "learning_rate": 1.9393939393939395e-06,
+      "loss": 0.4024,
+      "step": 128
+    },
+    {
+      "epoch": 0.9699248120300752,
+      "grad_norm": 0.24749604166329323,
+      "learning_rate": 1.9545454545454545e-06,
+      "loss": 0.3979,
+      "step": 129
+    },
+    {
+      "epoch": 0.9774436090225563,
+      "grad_norm": 0.2435430160630499,
+      "learning_rate": 1.9696969696969695e-06,
+      "loss": 0.391,
+      "step": 130
+    },
+    {
+      "epoch": 0.9849624060150376,
+      "grad_norm": 0.245143614150133,
+      "learning_rate": 1.984848484848485e-06,
+      "loss": 0.3874,
+      "step": 131
+    },
+    {
+      "epoch": 0.9924812030075187,
+      "grad_norm": 0.2178191761468034,
+      "learning_rate": 2e-06,
+      "loss": 0.3824,
+      "step": 132
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.22902584701598977,
+      "learning_rate": 1.9999992278253237e-06,
+      "loss": 0.3967,
+      "step": 133
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.38156241178512573,
+      "eval_runtime": 1.7509,
+      "eval_samples_per_second": 186.764,
+      "eval_steps_per_second": 3.427,
+      "step": 133
+    },
+    {
+      "epoch": 1.0075187969924813,
+      "grad_norm": 0.22251021566669277,
+      "learning_rate": 1.999996911302488e-06,
+      "loss": 0.4016,
+      "step": 134
+    },
+    {
+      "epoch": 1.0150375939849625,
+      "grad_norm": 0.2321273012804345,
+      "learning_rate": 1.99999305043507e-06,
+      "loss": 0.3824,
+      "step": 135
+    },
+    {
+      "epoch": 1.0225563909774436,
+      "grad_norm": 0.22893861405308363,
+      "learning_rate": 1.9999876452290317e-06,
+      "loss": 0.3902,
+      "step": 136
+    },
+    {
+      "epoch": 1.0300751879699248,
+      "grad_norm": 0.21429013823823248,
+      "learning_rate": 1.999980695692722e-06,
+      "loss": 0.3692,
+      "step": 137
+    },
+    {
+      "epoch": 1.037593984962406,
+      "grad_norm": 0.24165365825817373,
+      "learning_rate": 1.999972201836872e-06,
+      "loss": 0.3795,
+      "step": 138
+    },
+    {
+      "epoch": 1.045112781954887,
+      "grad_norm": 0.21472358382821435,
+      "learning_rate": 1.9999621636746e-06,
+      "loss": 0.381,
+      "step": 139
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.2313475017536494,
+      "learning_rate": 1.999950581221408e-06,
+      "loss": 0.383,
+      "step": 140
+    },
+    {
+      "epoch": 1.0601503759398496,
+      "grad_norm": 0.22056073484924088,
+      "learning_rate": 1.999937454495184e-06,
+      "loss": 0.3824,
+      "step": 141
+    },
+    {
+      "epoch": 1.0676691729323309,
+      "grad_norm": 0.23982841430407453,
+      "learning_rate": 1.9999227835162e-06,
+      "loss": 0.3744,
+      "step": 142
+    },
+    {
+      "epoch": 1.0751879699248121,
+      "grad_norm": 0.23515180314732004,
+      "learning_rate": 1.9999065683071128e-06,
+      "loss": 0.3921,
+      "step": 143
+    },
+    {
+      "epoch": 1.0827067669172932,
+      "grad_norm": 0.22551414249890428,
+      "learning_rate": 1.9998888088929643e-06,
+      "loss": 0.3856,
+      "step": 144
+    },
+    {
+      "epoch": 1.0902255639097744,
+      "grad_norm": 0.21558886490679555,
+      "learning_rate": 1.9998695053011815e-06,
+      "loss": 0.3751,
+      "step": 145
+    },
+    {
+      "epoch": 1.0977443609022557,
+      "grad_norm": 0.22228032831357214,
+      "learning_rate": 1.9998486575615758e-06,
+      "loss": 0.3825,
+      "step": 146
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.21905107141270422,
+      "learning_rate": 1.9998262657063435e-06,
+      "loss": 0.369,
+      "step": 147
+    },
+    {
+      "epoch": 1.112781954887218,
+      "grad_norm": 0.22656649015473823,
+      "learning_rate": 1.9998023297700654e-06,
+      "loss": 0.3938,
+      "step": 148
+    },
+    {
+      "epoch": 1.1203007518796992,
+      "grad_norm": 0.203639302416817,
+      "learning_rate": 1.999776849789707e-06,
+      "loss": 0.3734,
+      "step": 149
+    },
+    {
+      "epoch": 1.1278195488721805,
+      "grad_norm": 0.2187762802783149,
+      "learning_rate": 1.999749825804618e-06,
+      "loss": 0.3875,
+      "step": 150
+    },
+    {
+      "epoch": 1.1353383458646618,
+      "grad_norm": 0.22150708148698264,
+      "learning_rate": 1.9997212578565333e-06,
+      "loss": 0.386,
+      "step": 151
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.22960911718928317,
+      "learning_rate": 1.9996911459895713e-06,
+      "loss": 0.3849,
+      "step": 152
+    },
+    {
+      "epoch": 1.150375939849624,
+      "grad_norm": 0.23922917991060036,
+      "learning_rate": 1.999659490250236e-06,
+      "loss": 0.3963,
+      "step": 153
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.21667417626317434,
+      "learning_rate": 1.9996262906874136e-06,
+      "loss": 0.3686,
+      "step": 154
+    },
+    {
+      "epoch": 1.1654135338345863,
+      "grad_norm": 0.22151302791273322,
+      "learning_rate": 1.9995915473523774e-06,
+      "loss": 0.3834,
+      "step": 155
+    },
+    {
+      "epoch": 1.1729323308270676,
+      "grad_norm": 0.21935043583503394,
+      "learning_rate": 1.9995552602987826e-06,
+      "loss": 0.3827,
+      "step": 156
+    },
+    {
+      "epoch": 1.1804511278195489,
+      "grad_norm": 0.22798150425807281,
+      "learning_rate": 1.9995174295826686e-06,
+      "loss": 0.3711,
+      "step": 157
+    },
+    {
+      "epoch": 1.1879699248120301,
+      "grad_norm": 0.23112793116641064,
+      "learning_rate": 1.9994780552624593e-06,
+      "loss": 0.3834,
+      "step": 158
+    },
+    {
+      "epoch": 1.1954887218045114,
+      "grad_norm": 0.2333766174047214,
+      "learning_rate": 1.9994371373989633e-06,
+      "loss": 0.3957,
+      "step": 159
+    },
+    {
+      "epoch": 1.2030075187969924,
+      "grad_norm": 0.22385502007728467,
+      "learning_rate": 1.9993946760553714e-06,
+      "loss": 0.374,
+      "step": 160
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.20214308501873934,
+      "learning_rate": 1.9993506712972588e-06,
+      "loss": 0.367,
+      "step": 161
+    },
+    {
+      "epoch": 1.218045112781955,
+      "grad_norm": 0.20425303820861057,
+      "learning_rate": 1.9993051231925845e-06,
+      "loss": 0.3686,
+      "step": 162
+    },
+    {
+      "epoch": 1.225563909774436,
+      "grad_norm": 0.22480209047438862,
+      "learning_rate": 1.9992580318116905e-06,
+      "loss": 0.3663,
+      "step": 163
+    },
+    {
+      "epoch": 1.2330827067669172,
+      "grad_norm": 0.2218362857833691,
+      "learning_rate": 1.9992093972273017e-06,
+      "loss": 0.3799,
+      "step": 164
+    },
+    {
+      "epoch": 1.2406015037593985,
+      "grad_norm": 0.211631836385828,
+      "learning_rate": 1.999159219514528e-06,
+      "loss": 0.3727,
+      "step": 165
+    },
+    {
+      "epoch": 1.2481203007518797,
+      "grad_norm": 0.2271171988141363,
+      "learning_rate": 1.9991074987508608e-06,
+      "loss": 0.373,
+      "step": 166
+    },
+    {
+      "epoch": 1.255639097744361,
+      "grad_norm": 0.2283689245987733,
+      "learning_rate": 1.999054235016175e-06,
+      "loss": 0.3799,
+      "step": 167
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.21800599913929358,
+      "learning_rate": 1.9989994283927284e-06,
+      "loss": 0.3823,
+      "step": 168
+    },
+    {
+      "epoch": 1.2706766917293233,
+      "grad_norm": 0.22632962802803167,
+      "learning_rate": 1.9989430789651617e-06,
+      "loss": 0.3771,
+      "step": 169
+    },
+    {
+      "epoch": 1.2781954887218046,
+      "grad_norm": 0.2087977207875661,
+      "learning_rate": 1.9988851868204982e-06,
+      "loss": 0.3669,
+      "step": 170
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.2263412408714628,
+      "learning_rate": 1.9988257520481433e-06,
+      "loss": 0.3888,
+      "step": 171
+    },
+    {
+      "epoch": 1.2932330827067668,
+      "grad_norm": 0.22170980528971979,
+      "learning_rate": 1.998764774739885e-06,
+      "loss": 0.3634,
+      "step": 172
+    },
+    {
+      "epoch": 1.300751879699248,
+      "grad_norm": 0.2503501125736598,
+      "learning_rate": 1.9987022549898943e-06,
+      "loss": 0.3974,
+      "step": 173
+    },
+    {
+      "epoch": 1.3082706766917294,
+      "grad_norm": 0.22162488112605222,
+      "learning_rate": 1.9986381928947225e-06,
+      "loss": 0.3736,
+      "step": 174
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.21759144407285752,
+      "learning_rate": 1.9985725885533043e-06,
+      "loss": 0.397,
+      "step": 175
+    },
+    {
+      "epoch": 1.3233082706766917,
+      "grad_norm": 0.23957919551858087,
+      "learning_rate": 1.998505442066956e-06,
+      "loss": 0.3716,
+      "step": 176
+    },
+    {
+      "epoch": 1.330827067669173,
+      "grad_norm": 0.20922603377295165,
+      "learning_rate": 1.998436753539375e-06,
+      "loss": 0.3564,
+      "step": 177
+    },
+    {
+      "epoch": 1.3383458646616542,
+      "grad_norm": 0.223904168498734,
+      "learning_rate": 1.9983665230766404e-06,
+      "loss": 0.3721,
+      "step": 178
+    },
+    {
+      "epoch": 1.3458646616541352,
+      "grad_norm": 0.21913206351864345,
+      "learning_rate": 1.9982947507872127e-06,
+      "loss": 0.3652,
+      "step": 179
+    },
+    {
+      "epoch": 1.3533834586466165,
+      "grad_norm": 0.2343066814204347,
+      "learning_rate": 1.998221436781933e-06,
+      "loss": 0.3793,
+      "step": 180
+    },
+    {
+      "epoch": 1.3609022556390977,
+      "grad_norm": 0.22418373610201703,
+      "learning_rate": 1.998146581174024e-06,
+      "loss": 0.3829,
+      "step": 181
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.20600300816083308,
+      "learning_rate": 1.998070184079089e-06,
+      "loss": 0.3616,
+      "step": 182
+    },
+    {
+      "epoch": 1.3759398496240602,
+      "grad_norm": 0.2215211900786857,
+      "learning_rate": 1.9979922456151114e-06,
+      "loss": 0.3688,
+      "step": 183
+    },
+    {
+      "epoch": 1.3834586466165413,
+      "grad_norm": 0.21965009723257603,
+      "learning_rate": 1.997912765902456e-06,
+      "loss": 0.3655,
+      "step": 184
+    },
+    {
+      "epoch": 1.3909774436090225,
+      "grad_norm": 0.21824942571801875,
+      "learning_rate": 1.997831745063867e-06,
+      "loss": 0.3942,
+      "step": 185
+    },
+    {
+      "epoch": 1.3984962406015038,
+      "grad_norm": 0.21713411174478128,
+      "learning_rate": 1.9977491832244686e-06,
+      "loss": 0.3756,
+      "step": 186
+    },
+    {
+      "epoch": 1.4060150375939848,
+      "grad_norm": 0.20599775936174017,
+      "learning_rate": 1.9976650805117658e-06,
+      "loss": 0.3666,
+      "step": 187
+    },
+    {
+      "epoch": 1.413533834586466,
+      "grad_norm": 0.2139162411845165,
+      "learning_rate": 1.9975794370556416e-06,
+      "loss": 0.3584,
+      "step": 188
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.20738357079041192,
+      "learning_rate": 1.99749225298836e-06,
+      "loss": 0.3684,
+      "step": 189
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.20943506957677496,
+      "learning_rate": 1.9974035284445638e-06,
+      "loss": 0.3653,
+      "step": 190
+    },
+    {
+      "epoch": 1.4360902255639099,
+      "grad_norm": 0.22092760927819696,
+      "learning_rate": 1.997313263561275e-06,
+      "loss": 0.3764,
+      "step": 191
+    },
+    {
+      "epoch": 1.443609022556391,
+      "grad_norm": 0.22153559852279348,
+      "learning_rate": 1.9972214584778924e-06,
+      "loss": 0.3585,
+      "step": 192
+    },
+    {
+      "epoch": 1.4511278195488722,
+      "grad_norm": 0.21867396004887982,
+      "learning_rate": 1.9971281133361973e-06,
+      "loss": 0.3619,
+      "step": 193
+    },
+    {
+      "epoch": 1.4586466165413534,
+      "grad_norm": 0.22220305152370654,
+      "learning_rate": 1.997033228280346e-06,
+      "loss": 0.3737,
+      "step": 194
+    },
+    {
+      "epoch": 1.4661654135338344,
+      "grad_norm": 0.23859045282429994,
+      "learning_rate": 1.996936803456874e-06,
+      "loss": 0.3757,
+      "step": 195
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.22850464544447588,
+      "learning_rate": 1.9968388390146957e-06,
+      "loss": 0.3493,
+      "step": 196
+    },
+    {
+      "epoch": 1.481203007518797,
+      "grad_norm": 0.2148004066317054,
+      "learning_rate": 1.996739335105102e-06,
+      "loss": 0.3675,
+      "step": 197
+    },
+    {
+      "epoch": 1.4887218045112782,
+      "grad_norm": 0.22522318879070655,
+      "learning_rate": 1.996638291881762e-06,
+      "loss": 0.3674,
+      "step": 198
+    },
+    {
+      "epoch": 1.4962406015037595,
+      "grad_norm": 0.2028092311804893,
+      "learning_rate": 1.996535709500721e-06,
+      "loss": 0.3732,
+      "step": 199
+    },
+    {
+      "epoch": 1.5037593984962405,
+      "grad_norm": 0.20625254362232118,
+      "learning_rate": 1.9964315881204026e-06,
+      "loss": 0.3688,
+      "step": 200
+    },
+    {
+      "epoch": 1.5112781954887218,
+      "grad_norm": 0.21303077565999026,
+      "learning_rate": 1.996325927901607e-06,
+      "loss": 0.3594,
+      "step": 201
+    },
+    {
+      "epoch": 1.518796992481203,
+      "grad_norm": 0.20926361439535102,
+      "learning_rate": 1.9962187290075095e-06,
+      "loss": 0.3742,
+      "step": 202
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.213581506969076,
+      "learning_rate": 1.996109991603663e-06,
+      "loss": 0.3675,
+      "step": 203
+    },
+    {
+      "epoch": 1.5338345864661656,
+      "grad_norm": 0.20769966533064474,
+      "learning_rate": 1.9959997158579965e-06,
+      "loss": 0.3626,
+      "step": 204
+    },
+    {
+      "epoch": 1.5413533834586466,
+      "grad_norm": 0.1977662671264703,
+      "learning_rate": 1.995887901940814e-06,
+      "loss": 0.3641,
+      "step": 205
+    },
+    {
+      "epoch": 1.5488721804511278,
+      "grad_norm": 0.21189038064320026,
+      "learning_rate": 1.9957745500247954e-06,
+      "loss": 0.36,
+      "step": 206
+    },
+    {
+      "epoch": 1.556390977443609,
+      "grad_norm": 0.20381024646532855,
+      "learning_rate": 1.995659660284995e-06,
+      "loss": 0.3611,
+      "step": 207
+    },
+    {
+      "epoch": 1.5639097744360901,
+      "grad_norm": 0.21242140163413198,
+      "learning_rate": 1.9955432328988433e-06,
+      "loss": 0.3513,
+      "step": 208
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.21885998180744873,
+      "learning_rate": 1.995425268046145e-06,
+      "loss": 0.3649,
+      "step": 209
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.21134826515868899,
+      "learning_rate": 1.9953057659090784e-06,
+      "loss": 0.3623,
+      "step": 210
+    },
+    {
+      "epoch": 1.5864661654135337,
+      "grad_norm": 0.2104172225594599,
+      "learning_rate": 1.9951847266721967e-06,
+      "loss": 0.3605,
+      "step": 211
+    },
+    {
+      "epoch": 1.5939849624060152,
+      "grad_norm": 0.21397356027280193,
+      "learning_rate": 1.9950621505224274e-06,
+      "loss": 0.3781,
+      "step": 212
+    },
+    {
+      "epoch": 1.6015037593984962,
+      "grad_norm": 0.23291839504147976,
+      "learning_rate": 1.9949380376490703e-06,
+      "loss": 0.3611,
+      "step": 213
+    },
+    {
+      "epoch": 1.6090225563909775,
+      "grad_norm": 0.21866731264986877,
+      "learning_rate": 1.9948123882437994e-06,
+      "loss": 0.3479,
+      "step": 214
+    },
+    {
+      "epoch": 1.6165413533834587,
+      "grad_norm": 0.21601250442640677,
+      "learning_rate": 1.9946852025006605e-06,
+      "loss": 0.3704,
+      "step": 215
+    },
+    {
+      "epoch": 1.6240601503759398,
+      "grad_norm": 0.2246593414400215,
+      "learning_rate": 1.994556480616074e-06,
+      "loss": 0.367,
+      "step": 216
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.24174033142518786,
+      "learning_rate": 1.9944262227888307e-06,
+      "loss": 0.3706,
+      "step": 217
+    },
+    {
+      "epoch": 1.6390977443609023,
+      "grad_norm": 0.2069409923761465,
+      "learning_rate": 1.9942944292200944e-06,
+      "loss": 0.3644,
+      "step": 218
+    },
+    {
+      "epoch": 1.6466165413533833,
+      "grad_norm": 0.21383767422476158,
+      "learning_rate": 1.9941611001134e-06,
+      "loss": 0.3686,
+      "step": 219
+    },
+    {
+      "epoch": 1.6541353383458648,
+      "grad_norm": 0.21471119163618935,
+      "learning_rate": 1.9940262356746553e-06,
+      "loss": 0.362,
+      "step": 220
+    },
+    {
+      "epoch": 1.6616541353383458,
+      "grad_norm": 0.20317512271758342,
+      "learning_rate": 1.993889836112137e-06,
+      "loss": 0.3669,
+      "step": 221
+    },
+    {
+      "epoch": 1.669172932330827,
+      "grad_norm": 0.20357297977664585,
+      "learning_rate": 1.9937519016364938e-06,
+      "loss": 0.3553,
+      "step": 222
+    },
+    {
+      "epoch": 1.6766917293233083,
+      "grad_norm": 0.22178945647640003,
+      "learning_rate": 1.9936124324607453e-06,
+      "loss": 0.3619,
+      "step": 223
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.20408572993783505,
+      "learning_rate": 1.9934714288002807e-06,
+      "loss": 0.3577,
+      "step": 224
+    },
+    {
+      "epoch": 1.6917293233082706,
+      "grad_norm": 0.220360201725772,
+      "learning_rate": 1.9933288908728577e-06,
+      "loss": 0.3741,
+      "step": 225
+    },
+    {
+      "epoch": 1.699248120300752,
+      "grad_norm": 0.21418750613898777,
+      "learning_rate": 1.993184818898606e-06,
+      "loss": 0.3665,
+      "step": 226
+    },
+    {
+      "epoch": 1.706766917293233,
+      "grad_norm": 0.22719511963431602,
+      "learning_rate": 1.9930392131000224e-06,
+      "loss": 0.3601,
+      "step": 227
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.22947587753674983,
+      "learning_rate": 1.992892073701973e-06,
+      "loss": 0.3689,
+      "step": 228
+    },
+    {
+      "epoch": 1.7218045112781954,
+      "grad_norm": 0.21318596838189396,
+      "learning_rate": 1.9927434009316933e-06,
+      "loss": 0.359,
+      "step": 229
+    },
+    {
+      "epoch": 1.7293233082706767,
+      "grad_norm": 0.22169673305686793,
+      "learning_rate": 1.9925931950187853e-06,
+      "loss": 0.3672,
+      "step": 230
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.22758491440731635,
+      "learning_rate": 1.9924414561952193e-06,
+      "loss": 0.3521,
+      "step": 231
+    },
+    {
+      "epoch": 1.744360902255639,
+      "grad_norm": 0.22282118273950066,
+      "learning_rate": 1.992288184695333e-06,
+      "loss": 0.3567,
+      "step": 232
+    },
+    {
+      "epoch": 1.7518796992481203,
+      "grad_norm": 0.22975953062051802,
+      "learning_rate": 1.9921333807558316e-06,
+      "loss": 0.367,
+      "step": 233
+    },
+    {
+      "epoch": 1.7593984962406015,
+      "grad_norm": 0.20414752218120835,
+      "learning_rate": 1.9919770446157865e-06,
+      "loss": 0.3487,
+      "step": 234
+    },
+    {
+      "epoch": 1.7669172932330826,
+      "grad_norm": 0.21098895928566336,
+      "learning_rate": 1.991819176516635e-06,
+      "loss": 0.3618,
+      "step": 235
+    },
+    {
+      "epoch": 1.774436090225564,
+      "grad_norm": 0.21781619831505467,
+      "learning_rate": 1.9916597767021807e-06,
+      "loss": 0.3669,
+      "step": 236
+    },
+    {
+      "epoch": 1.781954887218045,
+      "grad_norm": 0.20529379519071542,
+      "learning_rate": 1.991498845418592e-06,
+      "loss": 0.3631,
+      "step": 237
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.19512298185899943,
+      "learning_rate": 1.991336382914404e-06,
+      "loss": 0.3574,
+      "step": 238
+    },
+    {
+      "epoch": 1.7969924812030076,
+      "grad_norm": 0.2049941121472269,
+      "learning_rate": 1.9911723894405154e-06,
+      "loss": 0.3713,
+      "step": 239
+    },
+    {
+      "epoch": 1.8045112781954886,
+      "grad_norm": 0.21071824402710027,
+      "learning_rate": 1.991006865250189e-06,
+      "loss": 0.3467,
+      "step": 240
+    },
+    {
+      "epoch": 1.8120300751879699,
+      "grad_norm": 0.21886227868774683,
+      "learning_rate": 1.990839810599052e-06,
+      "loss": 0.3607,
+      "step": 241
+    },
+    {
+      "epoch": 1.8195488721804511,
+      "grad_norm": 0.21233535817724378,
+      "learning_rate": 1.990671225745096e-06,
+      "loss": 0.3573,
+      "step": 242
+    },
+    {
+      "epoch": 1.8270676691729322,
+      "grad_norm": 0.21199403033367978,
+      "learning_rate": 1.9905011109486733e-06,
+      "loss": 0.3607,
+      "step": 243
+    },
+    {
+      "epoch": 1.8345864661654137,
+      "grad_norm": 0.21751216694093514,
+      "learning_rate": 1.990329466472502e-06,
+      "loss": 0.3631,
+      "step": 244
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.2101788322254525,
+      "learning_rate": 1.9901562925816604e-06,
+      "loss": 0.3539,
+      "step": 245
+    },
+    {
+      "epoch": 1.849624060150376,
+      "grad_norm": 0.19747285512224255,
+      "learning_rate": 1.9899815895435898e-06,
+      "loss": 0.3533,
+      "step": 246
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.22100881139080455,
+      "learning_rate": 1.9898053576280926e-06,
+      "loss": 0.3584,
+      "step": 247
+    },
+    {
+      "epoch": 1.8646616541353382,
+      "grad_norm": 0.22141915915140906,
+      "learning_rate": 1.9896275971073322e-06,
+      "loss": 0.3768,
+      "step": 248
+    },
+    {
+      "epoch": 1.8721804511278195,
+      "grad_norm": 0.21474474623004797,
+      "learning_rate": 1.9894483082558335e-06,
+      "loss": 0.3543,
+      "step": 249
+    },
+    {
+      "epoch": 1.8796992481203008,
+      "grad_norm": 0.2335244510926359,
+      "learning_rate": 1.9892674913504807e-06,
+      "loss": 0.3655,
+      "step": 250
+    },
+    {
+      "epoch": 1.8872180451127818,
+      "grad_norm": 0.20051360164492335,
+      "learning_rate": 1.9890851466705183e-06,
+      "loss": 0.3607,
+      "step": 251
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.21889554023687538,
+      "learning_rate": 1.9889012744975504e-06,
+      "loss": 0.3581,
+      "step": 252
+    },
+    {
+      "epoch": 1.9022556390977443,
+      "grad_norm": 0.21817125412279237,
+      "learning_rate": 1.98871587511554e-06,
+      "loss": 0.3728,
+      "step": 253
+    },
+    {
+      "epoch": 1.9097744360902256,
+      "grad_norm": 0.22616335236296678,
+      "learning_rate": 1.9885289488108084e-06,
+      "loss": 0.3567,
+      "step": 254
+    },
+    {
+      "epoch": 1.9172932330827068,
+      "grad_norm": 0.22483784008868193,
+      "learning_rate": 1.988340495872035e-06,
+      "loss": 0.3613,
+      "step": 255
+    },
+    {
+      "epoch": 1.9248120300751879,
+      "grad_norm": 0.22820161802257335,
+      "learning_rate": 1.9881505165902565e-06,
+      "loss": 0.3575,
+      "step": 256
+    },
+    {
+      "epoch": 1.9323308270676691,
+      "grad_norm": 0.21616987028351764,
+      "learning_rate": 1.987959011258868e-06,
+      "loss": 0.371,
+      "step": 257
+    },
+    {
+      "epoch": 1.9398496240601504,
+      "grad_norm": 0.2190480056578399,
+      "learning_rate": 1.9877659801736203e-06,
+      "loss": 0.3593,
+      "step": 258
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.2168657220192092,
+      "learning_rate": 1.987571423632621e-06,
+      "loss": 0.3695,
+      "step": 259
+    },
+    {
+      "epoch": 1.954887218045113,
+      "grad_norm": 0.2078445497847116,
+      "learning_rate": 1.987375341936333e-06,
+      "loss": 0.3534,
+      "step": 260
+    },
+    {
+      "epoch": 1.962406015037594,
+      "grad_norm": 0.21202119931028565,
+      "learning_rate": 1.9871777353875756e-06,
+      "loss": 0.3449,
+      "step": 261
+    },
+    {
+      "epoch": 1.9699248120300752,
+      "grad_norm": 0.20702065458323546,
+      "learning_rate": 1.986978604291522e-06,
+      "loss": 0.3619,
+      "step": 262
+    },
+    {
+      "epoch": 1.9774436090225564,
+      "grad_norm": 0.2114879467999042,
+      "learning_rate": 1.9867779489557003e-06,
+      "loss": 0.3505,
+      "step": 263
+    },
+    {
+      "epoch": 1.9849624060150375,
+      "grad_norm": 0.20550407926725847,
+      "learning_rate": 1.986575769689992e-06,
+      "loss": 0.3565,
+      "step": 264
+    },
+    {
+      "epoch": 1.9924812030075187,
+      "grad_norm": 0.21830003513580257,
+      "learning_rate": 1.9863720668066327e-06,
+      "loss": 0.3496,
+      "step": 265
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.21811976222225188,
+      "learning_rate": 1.986166840620211e-06,
+      "loss": 0.3574,
+      "step": 266
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.3915803134441376,
+      "eval_runtime": 1.7089,
+      "eval_samples_per_second": 191.346,
+      "eval_steps_per_second": 3.511,
+      "step": 266
+    },
+    {
+      "epoch": 2.007518796992481,
+      "grad_norm": 0.22095155610704018,
+      "learning_rate": 1.985960091447668e-06,
+      "loss": 0.3508,
+      "step": 267
+    },
+    {
+      "epoch": 2.0150375939849625,
+      "grad_norm": 0.23205298402439492,
+      "learning_rate": 1.9857518196082962e-06,
+      "loss": 0.3723,
+      "step": 268
+    },
+    {
+      "epoch": 2.0225563909774436,
+      "grad_norm": 0.2015969224867406,
+      "learning_rate": 1.9855420254237407e-06,
+      "loss": 0.3461,
+      "step": 269
+    },
+    {
+      "epoch": 2.030075187969925,
+      "grad_norm": 0.21418711150702877,
+      "learning_rate": 1.985330709217996e-06,
+      "loss": 0.3594,
+      "step": 270
+    },
+    {
+      "epoch": 2.037593984962406,
+      "grad_norm": 0.2188548107769725,
+      "learning_rate": 1.985117871317409e-06,
+      "loss": 0.3538,
+      "step": 271
+    },
+    {
+      "epoch": 2.045112781954887,
+      "grad_norm": 0.2195803363132177,
+      "learning_rate": 1.9849035120506753e-06,
+      "loss": 0.3587,
+      "step": 272
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.21374550172887663,
+      "learning_rate": 1.984687631748841e-06,
+      "loss": 0.3469,
+      "step": 273
+    },
+    {
+      "epoch": 2.0601503759398496,
+      "grad_norm": 0.22114188334663823,
+      "learning_rate": 1.9844702307453005e-06,
+      "loss": 0.3567,
+      "step": 274
+    },
+    {
+      "epoch": 2.0676691729323307,
+      "grad_norm": 0.23392079799356308,
+      "learning_rate": 1.9842513093757964e-06,
+      "loss": 0.3694,
+      "step": 275
+    },
+    {
+      "epoch": 2.075187969924812,
+      "grad_norm": 0.20941221061436996,
+      "learning_rate": 1.9840308679784207e-06,
+      "loss": 0.3417,
+      "step": 276
+    },
+    {
+      "epoch": 2.082706766917293,
+      "grad_norm": 0.2125075661776674,
+      "learning_rate": 1.983808906893611e-06,
+      "loss": 0.3576,
+      "step": 277
+    },
+    {
+      "epoch": 2.090225563909774,
+      "grad_norm": 0.22041918398930718,
+      "learning_rate": 1.9835854264641535e-06,
+      "loss": 0.3485,
+      "step": 278
+    },
+    {
+      "epoch": 2.0977443609022557,
+      "grad_norm": 0.22976488943921297,
+      "learning_rate": 1.9833604270351795e-06,
+      "loss": 0.3576,
+      "step": 279
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.21613002296099998,
+      "learning_rate": 1.983133908954167e-06,
+      "loss": 0.3452,
+      "step": 280
+    },
+    {
+      "epoch": 2.112781954887218,
+      "grad_norm": 0.22427308721545586,
+      "learning_rate": 1.982905872570939e-06,
+      "loss": 0.3589,
+      "step": 281
+    },
+    {
+      "epoch": 2.1203007518796992,
+      "grad_norm": 0.20505263992191758,
+      "learning_rate": 1.9826763182376634e-06,
+      "loss": 0.3414,
+      "step": 282
+    },
+    {
+      "epoch": 2.1278195488721803,
+      "grad_norm": 0.20241560175687237,
+      "learning_rate": 1.9824452463088522e-06,
+      "loss": 0.355,
+      "step": 283
+    },
+    {
+      "epoch": 2.1353383458646618,
+      "grad_norm": 0.21520490391852498,
+      "learning_rate": 1.9822126571413612e-06,
+      "loss": 0.356,
+      "step": 284
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.2050503303696051,
+      "learning_rate": 1.9819785510943896e-06,
+      "loss": 0.3588,
+      "step": 285
+    },
+    {
+      "epoch": 2.1503759398496243,
+      "grad_norm": 0.20029812354629564,
+      "learning_rate": 1.981742928529478e-06,
+      "loss": 0.3515,
+      "step": 286
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.22137452199988633,
+      "learning_rate": 1.9815057898105116e-06,
+      "loss": 0.3509,
+      "step": 287
+    },
+    {
+      "epoch": 2.1654135338345863,
+      "grad_norm": 0.22121725711665702,
+      "learning_rate": 1.9812671353037137e-06,
+      "loss": 0.3466,
+      "step": 288
+    },
+    {
+      "epoch": 2.172932330827068,
+      "grad_norm": 0.21342126868254194,
+      "learning_rate": 1.9810269653776514e-06,
+      "loss": 0.3634,
+      "step": 289
+    },
+    {
+      "epoch": 2.180451127819549,
+      "grad_norm": 0.22157667317570492,
+      "learning_rate": 1.98078528040323e-06,
+      "loss": 0.3326,
+      "step": 290
+    },
+    {
+      "epoch": 2.18796992481203,
+      "grad_norm": 0.20951150243889824,
+      "learning_rate": 1.980542080753697e-06,
+      "loss": 0.3457,
+      "step": 291
+    },
+    {
+      "epoch": 2.1954887218045114,
+      "grad_norm": 0.22624761008317051,
+      "learning_rate": 1.9802973668046363e-06,
+      "loss": 0.3581,
+      "step": 292
+    },
+    {
+      "epoch": 2.2030075187969924,
+      "grad_norm": 0.23216481321395555,
+      "learning_rate": 1.980051138933972e-06,
+      "loss": 0.3519,
+      "step": 293
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.2074818769870995,
+      "learning_rate": 1.979803397521966e-06,
+      "loss": 0.3602,
+      "step": 294
+    },
+    {
+      "epoch": 2.218045112781955,
+      "grad_norm": 0.24211448372787903,
+      "learning_rate": 1.9795541429512175e-06,
+      "loss": 0.3472,
+      "step": 295
+    },
+    {
+      "epoch": 2.225563909774436,
+      "grad_norm": 0.21867442543665017,
+      "learning_rate": 1.979303375606663e-06,
+      "loss": 0.3597,
+      "step": 296
+    },
+    {
+      "epoch": 2.2330827067669174,
+      "grad_norm": 0.21210581597198833,
+      "learning_rate": 1.9790510958755754e-06,
+      "loss": 0.3513,
+      "step": 297
+    },
+    {
+      "epoch": 2.2406015037593985,
+      "grad_norm": 0.2136409987977701,
+      "learning_rate": 1.9787973041475616e-06,
+      "loss": 0.3481,
+      "step": 298
+    },
+    {
+      "epoch": 2.2481203007518795,
+      "grad_norm": 0.19856987002481488,
+      "learning_rate": 1.978542000814565e-06,
+      "loss": 0.3433,
+      "step": 299
+    },
+    {
+      "epoch": 2.255639097744361,
+      "grad_norm": 0.20043253275988965,
+      "learning_rate": 1.9782851862708634e-06,
+      "loss": 0.3476,
+      "step": 300
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.20111671576193807,
+      "learning_rate": 1.9780268609130676e-06,
+      "loss": 0.3474,
+      "step": 301
+    },
+    {
+      "epoch": 2.2706766917293235,
+      "grad_norm": 0.19618404716572016,
+      "learning_rate": 1.977767025140123e-06,
+      "loss": 0.3493,
+      "step": 302
+    },
+    {
+      "epoch": 2.2781954887218046,
+      "grad_norm": 0.2301817485660158,
+      "learning_rate": 1.9775056793533064e-06,
+      "loss": 0.3369,
+      "step": 303
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.20163023809835484,
+      "learning_rate": 1.9772428239562273e-06,
+      "loss": 0.3477,
+      "step": 304
+    },
+    {
+      "epoch": 2.293233082706767,
+      "grad_norm": 0.22630637150003374,
+      "learning_rate": 1.9769784593548257e-06,
+      "loss": 0.3581,
+      "step": 305
+    },
+    {
+      "epoch": 2.300751879699248,
+      "grad_norm": 0.2252232253668197,
+      "learning_rate": 1.9767125859573733e-06,
+      "loss": 0.3579,
+      "step": 306
+    },
+    {
+      "epoch": 2.308270676691729,
+      "grad_norm": 0.2066982898753058,
+      "learning_rate": 1.9764452041744713e-06,
+      "loss": 0.3481,
+      "step": 307
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.2079146796480719,
+      "learning_rate": 1.976176314419051e-06,
+      "loss": 0.3447,
+      "step": 308
+    },
+    {
+      "epoch": 2.3233082706766917,
+      "grad_norm": 0.20758604477980225,
+      "learning_rate": 1.9759059171063714e-06,
+      "loss": 0.3517,
+      "step": 309
+    },
+    {
+      "epoch": 2.3308270676691727,
+      "grad_norm": 0.21735271529141797,
+      "learning_rate": 1.975634012654021e-06,
+      "loss": 0.3444,
+      "step": 310
+    },
+    {
+      "epoch": 2.338345864661654,
+      "grad_norm": 0.2240596866247531,
+      "learning_rate": 1.9753606014819155e-06,
+      "loss": 0.3435,
+      "step": 311
+    },
+    {
+      "epoch": 2.345864661654135,
+      "grad_norm": 0.21539433488662113,
+      "learning_rate": 1.9750856840122965e-06,
+      "loss": 0.3578,
+      "step": 312
+    },
+    {
+      "epoch": 2.3533834586466167,
+      "grad_norm": 0.22803083437572866,
+      "learning_rate": 1.9748092606697327e-06,
+      "loss": 0.3462,
+      "step": 313
+    },
+    {
+      "epoch": 2.3609022556390977,
+      "grad_norm": 0.2201334881858541,
+      "learning_rate": 1.9745313318811194e-06,
+      "loss": 0.3413,
+      "step": 314
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.20999853254239798,
+      "learning_rate": 1.974251898075674e-06,
+      "loss": 0.3526,
+      "step": 315
+    },
+    {
+      "epoch": 2.3759398496240602,
+      "grad_norm": 0.20568743493730382,
+      "learning_rate": 1.9739709596849416e-06,
+      "loss": 0.3518,
+      "step": 316
+    },
+    {
+      "epoch": 2.3834586466165413,
+      "grad_norm": 0.21394250830502207,
+      "learning_rate": 1.973688517142788e-06,
+      "loss": 0.3497,
+      "step": 317
+    },
+    {
+      "epoch": 2.3909774436090228,
+      "grad_norm": 0.2086014995806524,
+      "learning_rate": 1.9734045708854043e-06,
+      "loss": 0.3585,
+      "step": 318
+    },
+    {
+      "epoch": 2.398496240601504,
+      "grad_norm": 0.2230768060220669,
+      "learning_rate": 1.9731191213513014e-06,
+      "loss": 0.34,
+      "step": 319
+    },
+    {
+      "epoch": 2.406015037593985,
+      "grad_norm": 0.22446743442432016,
+      "learning_rate": 1.9728321689813137e-06,
+      "loss": 0.3468,
+      "step": 320
+    },
+    {
+      "epoch": 2.4135338345864663,
+      "grad_norm": 0.20699490455742026,
+      "learning_rate": 1.9725437142185965e-06,
+      "loss": 0.346,
+      "step": 321
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.20938383827053192,
+      "learning_rate": 1.972253757508624e-06,
+      "loss": 0.3472,
+      "step": 322
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.22152838435428374,
+      "learning_rate": 1.9719622992991907e-06,
+      "loss": 0.3531,
+      "step": 323
+    },
+    {
+      "epoch": 2.43609022556391,
+      "grad_norm": 0.20337914645534588,
+      "learning_rate": 1.9716693400404097e-06,
+      "loss": 0.342,
+      "step": 324
+    },
+    {
+      "epoch": 2.443609022556391,
+      "grad_norm": 0.20566607035376627,
+      "learning_rate": 1.9713748801847136e-06,
+      "loss": 0.3391,
+      "step": 325
+    },
+    {
+      "epoch": 2.451127819548872,
+      "grad_norm": 0.21551579618986477,
+      "learning_rate": 1.97107892018685e-06,
+      "loss": 0.3694,
+      "step": 326
+    },
+    {
+      "epoch": 2.4586466165413534,
+      "grad_norm": 0.2087427232510144,
+      "learning_rate": 1.970781460503885e-06,
+      "loss": 0.333,
+      "step": 327
+    },
+    {
+      "epoch": 2.4661654135338344,
+      "grad_norm": 0.2086944893621931,
+      "learning_rate": 1.9704825015952003e-06,
+      "loss": 0.3501,
+      "step": 328
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.20910301704743575,
+      "learning_rate": 1.970182043922493e-06,
+      "loss": 0.3438,
+      "step": 329
+    },
+    {
+      "epoch": 2.481203007518797,
+      "grad_norm": 0.23498397227301066,
+      "learning_rate": 1.9698800879497745e-06,
+      "loss": 0.349,
+      "step": 330
+    },
+    {
+      "epoch": 2.488721804511278,
+      "grad_norm": 0.2212834176092268,
+      "learning_rate": 1.96957663414337e-06,
+      "loss": 0.3362,
+      "step": 331
+    },
+    {
+      "epoch": 2.4962406015037595,
+      "grad_norm": 0.24171895494176854,
+      "learning_rate": 1.9692716829719194e-06,
+      "loss": 0.3578,
+      "step": 332
+    },
+    {
+      "epoch": 2.5037593984962405,
+      "grad_norm": 0.21808786186027995,
+      "learning_rate": 1.9689652349063723e-06,
+      "loss": 0.3459,
+      "step": 333
+    },
+    {
+      "epoch": 2.511278195488722,
+      "grad_norm": 0.1997348571042542,
+      "learning_rate": 1.9686572904199926e-06,
+      "loss": 0.3415,
+      "step": 334
+    },
+    {
+      "epoch": 2.518796992481203,
+      "grad_norm": 0.20378126040298905,
+      "learning_rate": 1.9683478499883537e-06,
+      "loss": 0.3529,
+      "step": 335
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.2650611360632721,
+      "learning_rate": 1.9680369140893403e-06,
+      "loss": 0.35,
+      "step": 336
+    },
+    {
+      "epoch": 2.5338345864661656,
+      "grad_norm": 0.20172820062562308,
+      "learning_rate": 1.9677244832031454e-06,
+      "loss": 0.3486,
+      "step": 337
+    },
+    {
+      "epoch": 2.5413533834586466,
+      "grad_norm": 0.22373045832272115,
+      "learning_rate": 1.9674105578122716e-06,
+      "loss": 0.3494,
+      "step": 338
+    },
+    {
+      "epoch": 2.548872180451128,
+      "grad_norm": 0.204439937940755,
+      "learning_rate": 1.9670951384015297e-06,
+      "loss": 0.3345,
+      "step": 339
+    },
+    {
+      "epoch": 2.556390977443609,
+      "grad_norm": 0.21935052012922732,
+      "learning_rate": 1.9667782254580374e-06,
+      "loss": 0.35,
+      "step": 340
+    },
+    {
+      "epoch": 2.56390977443609,
+      "grad_norm": 0.2108801795350368,
+      "learning_rate": 1.966459819471218e-06,
+      "loss": 0.3446,
+      "step": 341
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.20956494127839848,
+      "learning_rate": 1.9661399209328027e-06,
+      "loss": 0.3456,
+      "step": 342
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.21732264757412872,
+      "learning_rate": 1.965818530336827e-06,
+      "loss": 0.328,
+      "step": 343
+    },
+    {
+      "epoch": 2.5864661654135337,
+      "grad_norm": 0.21041637087794404,
+      "learning_rate": 1.965495648179629e-06,
+      "loss": 0.3488,
+      "step": 344
+    },
+    {
+      "epoch": 2.593984962406015,
+      "grad_norm": 0.21753110229520498,
+      "learning_rate": 1.9651712749598523e-06,
+      "loss": 0.3501,
+      "step": 345
+    },
+    {
+      "epoch": 2.601503759398496,
+      "grad_norm": 0.21715390612171026,
+      "learning_rate": 1.9648454111784418e-06,
+      "loss": 0.345,
+      "step": 346
+    },
+    {
+      "epoch": 2.6090225563909772,
+      "grad_norm": 0.22581672657766216,
+      "learning_rate": 1.964518057338646e-06,
+      "loss": 0.3484,
+      "step": 347
+    },
+    {
+      "epoch": 2.6165413533834587,
+      "grad_norm": 0.22359760894541875,
+      "learning_rate": 1.964189213946013e-06,
+      "loss": 0.3648,
+      "step": 348
+    },
+    {
+      "epoch": 2.6240601503759398,
+      "grad_norm": 0.2125240604251561,
+      "learning_rate": 1.963858881508392e-06,
+      "loss": 0.3306,
+      "step": 349
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.22227282248297905,
+      "learning_rate": 1.9635270605359315e-06,
+      "loss": 0.3526,
+      "step": 350
+    },
+    {
+      "epoch": 2.6390977443609023,
+      "grad_norm": 0.20467866571919213,
+      "learning_rate": 1.963193751541079e-06,
+      "loss": 0.3395,
+      "step": 351
+    },
+    {
+      "epoch": 2.6466165413533833,
+      "grad_norm": 0.20360730347782205,
+      "learning_rate": 1.962858955038581e-06,
+      "loss": 0.3437,
+      "step": 352
+    },
+    {
+      "epoch": 2.654135338345865,
+      "grad_norm": 0.21453628999930813,
+      "learning_rate": 1.9625226715454787e-06,
+      "loss": 0.3518,
+      "step": 353
+    },
+    {
+      "epoch": 2.661654135338346,
+      "grad_norm": 0.21508687621009656,
+      "learning_rate": 1.9621849015811122e-06,
+      "loss": 0.3297,
+      "step": 354
+    },
+    {
+      "epoch": 2.6691729323308273,
+      "grad_norm": 0.20289430726639016,
+      "learning_rate": 1.9618456456671163e-06,
+      "loss": 0.3495,
+      "step": 355
+    },
+    {
+      "epoch": 2.6766917293233083,
+      "grad_norm": 0.21816877514504923,
+      "learning_rate": 1.9615049043274204e-06,
+      "loss": 0.3402,
+      "step": 356
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.2047743180620807,
+      "learning_rate": 1.9611626780882484e-06,
+      "loss": 0.3428,
+      "step": 357
+    },
+    {
+      "epoch": 2.6917293233082704,
+      "grad_norm": 0.21139207118163825,
+      "learning_rate": 1.960818967478117e-06,
+      "loss": 0.3418,
+      "step": 358
+    },
+    {
+      "epoch": 2.699248120300752,
+      "grad_norm": 0.21261646713086174,
+      "learning_rate": 1.9604737730278354e-06,
+      "loss": 0.3388,
+      "step": 359
+    },
+    {
+      "epoch": 2.706766917293233,
+      "grad_norm": 0.21085143215695795,
+      "learning_rate": 1.960127095270505e-06,
+      "loss": 0.3459,
+      "step": 360
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.20692474979767714,
+      "learning_rate": 1.9597789347415167e-06,
+      "loss": 0.3409,
+      "step": 361
+    },
+    {
+      "epoch": 2.7218045112781954,
+      "grad_norm": 0.21255939808574717,
+      "learning_rate": 1.959429291978552e-06,
+      "loss": 0.3439,
+      "step": 362
+    },
+    {
+      "epoch": 2.7293233082706765,
+      "grad_norm": 0.20881501797785726,
+      "learning_rate": 1.959078167521582e-06,
+      "loss": 0.3411,
+      "step": 363
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.22203353446338803,
+      "learning_rate": 1.9587255619128646e-06,
+      "loss": 0.3219,
+      "step": 364
+    },
+    {
+      "epoch": 2.744360902255639,
+      "grad_norm": 0.20798902099716723,
+      "learning_rate": 1.9583714756969473e-06,
+      "loss": 0.35,
+      "step": 365
+    },
+    {
+      "epoch": 2.7518796992481205,
+      "grad_norm": 0.21962501725060773,
+      "learning_rate": 1.9580159094206617e-06,
+      "loss": 0.3415,
+      "step": 366
+    },
+    {
+      "epoch": 2.7593984962406015,
+      "grad_norm": 0.21158718507866742,
+      "learning_rate": 1.9576588636331273e-06,
+      "loss": 0.3495,
+      "step": 367
+    },
+    {
+      "epoch": 2.7669172932330826,
+      "grad_norm": 0.23764614893819733,
+      "learning_rate": 1.9573003388857475e-06,
+      "loss": 0.3535,
+      "step": 368
+    },
+    {
+      "epoch": 2.774436090225564,
+      "grad_norm": 0.21321226623484887,
+      "learning_rate": 1.956940335732209e-06,
+      "loss": 0.345,
+      "step": 369
+    },
+    {
+      "epoch": 2.781954887218045,
+      "grad_norm": 0.24523048888896556,
+      "learning_rate": 1.9565788547284824e-06,
+      "loss": 0.342,
+      "step": 370
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.22558348654313412,
+      "learning_rate": 1.956215896432822e-06,
+      "loss": 0.3409,
+      "step": 371
+    },
+    {
+      "epoch": 2.7969924812030076,
+      "grad_norm": 0.20373691548745282,
+      "learning_rate": 1.9558514614057607e-06,
+      "loss": 0.3477,
+      "step": 372
+    },
+    {
+      "epoch": 2.8045112781954886,
+      "grad_norm": 0.2213839935631959,
+      "learning_rate": 1.955485550210114e-06,
+      "loss": 0.3581,
+      "step": 373
+    },
+    {
+      "epoch": 2.8120300751879697,
+      "grad_norm": 0.2134493790831214,
+      "learning_rate": 1.955118163410977e-06,
+      "loss": 0.3402,
+      "step": 374
+    },
+    {
+      "epoch": 2.819548872180451,
+      "grad_norm": 0.2059425981395109,
+      "learning_rate": 1.9547493015757233e-06,
+      "loss": 0.3442,
+      "step": 375
+    },
+    {
+      "epoch": 2.827067669172932,
+      "grad_norm": 0.22475916032896573,
+      "learning_rate": 1.954378965274004e-06,
+      "loss": 0.3291,
+      "step": 376
+    },
+    {
+      "epoch": 2.8345864661654137,
+      "grad_norm": 0.2172714119703267,
+      "learning_rate": 1.9540071550777475e-06,
+      "loss": 0.3491,
+      "step": 377
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.22436181554202905,
+      "learning_rate": 1.9536338715611593e-06,
+      "loss": 0.3372,
+      "step": 378
+    },
+    {
+      "epoch": 2.8496240601503757,
+      "grad_norm": 0.21248780344771265,
+      "learning_rate": 1.953259115300719e-06,
+      "loss": 0.3351,
+      "step": 379
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.20959521555061097,
+      "learning_rate": 1.9528828868751815e-06,
+      "loss": 0.3506,
+      "step": 380
+    },
+    {
+      "epoch": 2.8646616541353382,
+      "grad_norm": 0.23064107219658858,
+      "learning_rate": 1.9525051868655753e-06,
+      "loss": 0.3448,
+      "step": 381
+    },
+    {
+      "epoch": 2.8721804511278197,
+      "grad_norm": 0.23350097222445898,
+      "learning_rate": 1.9521260158552004e-06,
+      "loss": 0.3483,
+      "step": 382
+    },
+    {
+      "epoch": 2.8796992481203008,
+      "grad_norm": 0.20886282626009017,
+      "learning_rate": 1.9517453744296294e-06,
+      "loss": 0.3565,
+      "step": 383
+    },
+    {
+      "epoch": 2.887218045112782,
+      "grad_norm": 0.20276559616256357,
+      "learning_rate": 1.9513632631767062e-06,
+      "loss": 0.3368,
+      "step": 384
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.2219462453889891,
+      "learning_rate": 1.9509796826865433e-06,
+      "loss": 0.3424,
+      "step": 385
+    },
+    {
+      "epoch": 2.9022556390977443,
+      "grad_norm": 0.20949378746508884,
+      "learning_rate": 1.950594633551524e-06,
+      "loss": 0.3415,
+      "step": 386
+    },
+    {
+      "epoch": 2.909774436090226,
+      "grad_norm": 0.2172978839539696,
+      "learning_rate": 1.950208116366298e-06,
+      "loss": 0.3438,
+      "step": 387
+    },
+    {
+      "epoch": 2.917293233082707,
+      "grad_norm": 0.22267053523538627,
+      "learning_rate": 1.949820131727783e-06,
+      "loss": 0.3388,
+      "step": 388
+    },
+    {
+      "epoch": 2.924812030075188,
+      "grad_norm": 0.21716281383566371,
+      "learning_rate": 1.949430680235162e-06,
+      "loss": 0.346,
+      "step": 389
+    },
+    {
+      "epoch": 2.932330827067669,
+      "grad_norm": 0.21121370711029158,
+      "learning_rate": 1.9490397624898857e-06,
+      "loss": 0.3457,
+      "step": 390
+    },
+    {
+      "epoch": 2.9398496240601504,
+      "grad_norm": 0.21395509670746496,
+      "learning_rate": 1.9486473790956668e-06,
+      "loss": 0.3466,
+      "step": 391
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.21942839384084808,
+      "learning_rate": 1.9482535306584824e-06,
+      "loss": 0.3361,
+      "step": 392
+    },
+    {
+      "epoch": 2.954887218045113,
+      "grad_norm": 0.21874557548231818,
+      "learning_rate": 1.947858217786572e-06,
+      "loss": 0.3479,
+      "step": 393
+    },
+    {
+      "epoch": 2.962406015037594,
+      "grad_norm": 0.2190618505898401,
+      "learning_rate": 1.947461441090437e-06,
+      "loss": 0.3497,
+      "step": 394
+    },
+    {
+      "epoch": 2.969924812030075,
+      "grad_norm": 0.20230877254650048,
+      "learning_rate": 1.9470632011828395e-06,
+      "loss": 0.34,
+      "step": 395
+    },
+    {
+      "epoch": 2.9774436090225564,
+      "grad_norm": 0.21413982282789149,
+      "learning_rate": 1.9466634986788002e-06,
+      "loss": 0.328,
+      "step": 396
+    },
+    {
+      "epoch": 2.9849624060150375,
+      "grad_norm": 0.22519605329087744,
+      "learning_rate": 1.9462623341956005e-06,
+      "loss": 0.3455,
+      "step": 397
+    },
+    {
+      "epoch": 2.992481203007519,
+      "grad_norm": 0.22102364613210643,
+      "learning_rate": 1.945859708352777e-06,
+      "loss": 0.3401,
+      "step": 398
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.20586509978549808,
+      "learning_rate": 1.945455621772126e-06,
+      "loss": 0.3464,
+      "step": 399
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.39734145998954773,
+      "eval_runtime": 1.7339,
+      "eval_samples_per_second": 188.597,
+      "eval_steps_per_second": 3.46,
+      "step": 399
+    },
+    {
+      "epoch": 3.007518796992481,
+      "grad_norm": 0.23053411043640512,
+      "learning_rate": 1.9450500750776984e-06,
+      "loss": 0.3349,
+      "step": 400
+    },
+    {
+      "epoch": 3.0150375939849625,
+      "grad_norm": 0.2135569178536534,
+      "learning_rate": 1.9446430688957987e-06,
+      "loss": 0.3365,
+      "step": 401
+    },
+    {
+      "epoch": 3.0225563909774436,
+      "grad_norm": 0.21219160279501142,
+      "learning_rate": 1.944234603854988e-06,
+      "loss": 0.3424,
+      "step": 402
+    },
+    {
+      "epoch": 3.030075187969925,
+      "grad_norm": 0.2044596344704826,
+      "learning_rate": 1.9438246805860783e-06,
+      "loss": 0.3208,
+      "step": 403
+    },
+    {
+      "epoch": 3.037593984962406,
+      "grad_norm": 0.2153336746545348,
+      "learning_rate": 1.9434132997221345e-06,
+      "loss": 0.3365,
+      "step": 404
+    },
+    {
+      "epoch": 3.045112781954887,
+      "grad_norm": 0.21142878658827913,
+      "learning_rate": 1.943000461898472e-06,
+      "loss": 0.3437,
+      "step": 405
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.20730997454728087,
+      "learning_rate": 1.9425861677526575e-06,
+      "loss": 0.3394,
+      "step": 406
+    },
+    {
+      "epoch": 3.0601503759398496,
+      "grad_norm": 0.20262588024279637,
+      "learning_rate": 1.942170417924505e-06,
+      "loss": 0.3328,
+      "step": 407
+    },
+    {
+      "epoch": 3.0676691729323307,
+      "grad_norm": 0.20966476083697377,
+      "learning_rate": 1.941753213056078e-06,
+      "loss": 0.3377,
+      "step": 408
+    },
+    {
+      "epoch": 3.075187969924812,
+      "grad_norm": 0.22171762802997932,
+      "learning_rate": 1.9413345537916864e-06,
+      "loss": 0.331,
+      "step": 409
+    },
+    {
+      "epoch": 3.082706766917293,
+      "grad_norm": 0.2515682304775118,
+      "learning_rate": 1.9409144407778865e-06,
+      "loss": 0.351,
+      "step": 410
+    },
+    {
+      "epoch": 3.090225563909774,
+      "grad_norm": 0.2262971412001475,
+      "learning_rate": 1.9404928746634793e-06,
+      "loss": 0.3397,
+      "step": 411
+    },
+    {
+      "epoch": 3.0977443609022557,
+      "grad_norm": 0.21541116097434343,
+      "learning_rate": 1.94006985609951e-06,
+      "loss": 0.3516,
+      "step": 412
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.24808505876572906,
+      "learning_rate": 1.9396453857392677e-06,
+      "loss": 0.331,
+      "step": 413
+    },
+    {
+      "epoch": 3.112781954887218,
+      "grad_norm": 0.21648025878383922,
+      "learning_rate": 1.9392194642382825e-06,
+      "loss": 0.3287,
+      "step": 414
+    },
+    {
+      "epoch": 3.1203007518796992,
+      "grad_norm": 0.21712583773244648,
+      "learning_rate": 1.938792092254326e-06,
+      "loss": 0.3519,
+      "step": 415
+    },
+    {
+      "epoch": 3.1278195488721803,
+      "grad_norm": 0.23282618571269065,
+      "learning_rate": 1.9383632704474103e-06,
+      "loss": 0.3386,
+      "step": 416
+    },
+    {
+      "epoch": 3.1353383458646618,
+      "grad_norm": 0.19994117695229985,
+      "learning_rate": 1.9379329994797854e-06,
+      "loss": 0.3317,
+      "step": 417
+    },
+    {
+      "epoch": 3.142857142857143,
+      "grad_norm": 0.21258979378039913,
+      "learning_rate": 1.9375012800159404e-06,
+      "loss": 0.3438,
+      "step": 418
+    },
+    {
+      "epoch": 3.1503759398496243,
+      "grad_norm": 0.22143692917602045,
+      "learning_rate": 1.9370681127226004e-06,
+      "loss": 0.3218,
+      "step": 419
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.21582228168215079,
+      "learning_rate": 1.936633498268728e-06,
+      "loss": 0.3431,
+      "step": 420
+    },
+    {
+      "epoch": 3.1654135338345863,
+      "grad_norm": 0.22497765476269319,
+      "learning_rate": 1.9361974373255187e-06,
+      "loss": 0.3236,
+      "step": 421
+    },
+    {
+      "epoch": 3.172932330827068,
+      "grad_norm": 0.24678375908677114,
+      "learning_rate": 1.935759930566404e-06,
+      "loss": 0.343,
+      "step": 422
+    },
+    {
+      "epoch": 3.180451127819549,
+      "grad_norm": 0.20650360396097048,
+      "learning_rate": 1.9353209786670465e-06,
+      "loss": 0.3307,
+      "step": 423
+    },
+    {
+      "epoch": 3.18796992481203,
+      "grad_norm": 0.20754792152721838,
+      "learning_rate": 1.934880582305341e-06,
+      "loss": 0.3355,
+      "step": 424
+    },
+    {
+      "epoch": 3.1954887218045114,
+      "grad_norm": 0.22756426501307186,
+      "learning_rate": 1.934438742161414e-06,
+      "loss": 0.3417,
+      "step": 425
+    },
+    {
+      "epoch": 3.2030075187969924,
+      "grad_norm": 0.22041567883012617,
+      "learning_rate": 1.933995458917621e-06,
+      "loss": 0.3442,
+      "step": 426
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.20662971111891107,
+      "learning_rate": 1.933550733258546e-06,
+      "loss": 0.329,
+      "step": 427
+    },
+    {
+      "epoch": 3.218045112781955,
+      "grad_norm": 0.2610866086417569,
+      "learning_rate": 1.9331045658710007e-06,
+      "loss": 0.3367,
+      "step": 428
+    },
+    {
+      "epoch": 3.225563909774436,
+      "grad_norm": 0.2365021784797832,
+      "learning_rate": 1.9326569574440237e-06,
+      "loss": 0.3422,
+      "step": 429
+    },
+    {
+      "epoch": 3.2330827067669174,
+      "grad_norm": 0.22233258001049008,
+      "learning_rate": 1.9322079086688784e-06,
+      "loss": 0.339,
+      "step": 430
+    },
+    {
+      "epoch": 3.2406015037593985,
+      "grad_norm": 0.2210806151938647,
+      "learning_rate": 1.931757420239053e-06,
+      "loss": 0.3358,
+      "step": 431
+    },
+    {
+      "epoch": 3.2481203007518795,
+      "grad_norm": 0.2280174556380608,
+      "learning_rate": 1.9313054928502594e-06,
+      "loss": 0.3303,
+      "step": 432
+    },
+    {
+      "epoch": 3.255639097744361,
+      "grad_norm": 0.20216371534007024,
+      "learning_rate": 1.930852127200431e-06,
+      "loss": 0.3353,
+      "step": 433
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.22870294868881944,
+      "learning_rate": 1.930397323989723e-06,
+      "loss": 0.3416,
+      "step": 434
+    },
+    {
+      "epoch": 3.2706766917293235,
+      "grad_norm": 0.1958936617501032,
+      "learning_rate": 1.9299410839205105e-06,
+      "loss": 0.3214,
+      "step": 435
+    },
+    {
+      "epoch": 3.2781954887218046,
+      "grad_norm": 0.2226100082588076,
+      "learning_rate": 1.9294834076973868e-06,
+      "loss": 0.3316,
+      "step": 436
+    },
+    {
+      "epoch": 3.2857142857142856,
+      "grad_norm": 0.21262603802637065,
+      "learning_rate": 1.929024296027165e-06,
+      "loss": 0.3374,
+      "step": 437
+    },
+    {
+      "epoch": 3.293233082706767,
+      "grad_norm": 0.21314911134708017,
+      "learning_rate": 1.9285637496188733e-06,
+      "loss": 0.323,
+      "step": 438
+    },
+    {
+      "epoch": 3.300751879699248,
+      "grad_norm": 0.2168636925050245,
+      "learning_rate": 1.9281017691837564e-06,
+      "loss": 0.3439,
+      "step": 439
+    },
+    {
+      "epoch": 3.308270676691729,
+      "grad_norm": 0.2228791385421588,
+      "learning_rate": 1.927638355435273e-06,
+      "loss": 0.3331,
+      "step": 440
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.22278249315332818,
+      "learning_rate": 1.9271735090890967e-06,
+      "loss": 0.338,
+      "step": 441
+    },
+    {
+      "epoch": 3.3233082706766917,
+      "grad_norm": 0.23520252671216407,
+      "learning_rate": 1.926707230863112e-06,
+      "loss": 0.3299,
+      "step": 442
+    },
+    {
+      "epoch": 3.3308270676691727,
+      "grad_norm": 0.20000960312539778,
+      "learning_rate": 1.9262395214774157e-06,
+      "loss": 0.3428,
+      "step": 443
+    },
+    {
+      "epoch": 3.338345864661654,
+      "grad_norm": 0.21169274516979636,
+      "learning_rate": 1.925770381654314e-06,
+      "loss": 0.3298,
+      "step": 444
+    },
+    {
+      "epoch": 3.345864661654135,
+      "grad_norm": 0.2011385640803676,
+      "learning_rate": 1.9252998121183235e-06,
+      "loss": 0.3378,
+      "step": 445
+    },
+    {
+      "epoch": 3.3533834586466167,
+      "grad_norm": 0.21312648590894295,
+      "learning_rate": 1.9248278135961674e-06,
+      "loss": 0.3388,
+      "step": 446
+    },
+    {
+      "epoch": 3.3609022556390977,
+      "grad_norm": 0.217231256956089,
+      "learning_rate": 1.9243543868167766e-06,
+      "loss": 0.3369,
+      "step": 447
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.20363278929539713,
+      "learning_rate": 1.9238795325112867e-06,
+      "loss": 0.343,
+      "step": 448
+    },
+    {
+      "epoch": 3.3759398496240602,
+      "grad_norm": 0.21992443499067824,
+      "learning_rate": 1.9234032514130392e-06,
+      "loss": 0.3489,
+      "step": 449
+    },
+    {
+      "epoch": 3.3834586466165413,
+      "grad_norm": 0.22293333056490677,
+      "learning_rate": 1.922925544257579e-06,
+      "loss": 0.3384,
+      "step": 450
+    },
+    {
+      "epoch": 3.3909774436090228,
+      "grad_norm": 0.2320232960961485,
+      "learning_rate": 1.922446411782652e-06,
+      "loss": 0.3224,
+      "step": 451
+    },
+    {
+      "epoch": 3.398496240601504,
+      "grad_norm": 0.20735409976850963,
+      "learning_rate": 1.9219658547282065e-06,
+      "loss": 0.3248,
+      "step": 452
+    },
+    {
+      "epoch": 3.406015037593985,
+      "grad_norm": 0.1898492922566569,
+      "learning_rate": 1.9214838738363904e-06,
+      "loss": 0.3224,
+      "step": 453
+    },
+    {
+      "epoch": 3.4135338345864663,
+      "grad_norm": 0.24327959918609202,
+      "learning_rate": 1.921000469851551e-06,
+      "loss": 0.3486,
+      "step": 454
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.19901485893006232,
+      "learning_rate": 1.920515643520232e-06,
+      "loss": 0.3242,
+      "step": 455
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.20778540847397078,
+      "learning_rate": 1.9200293955911755e-06,
+      "loss": 0.3155,
+      "step": 456
+    },
+    {
+      "epoch": 3.43609022556391,
+      "grad_norm": 0.23473135153423755,
+      "learning_rate": 1.919541726815318e-06,
+      "loss": 0.3429,
+      "step": 457
+    },
+    {
+      "epoch": 3.443609022556391,
+      "grad_norm": 0.20902404069407018,
+      "learning_rate": 1.91905263794579e-06,
+      "loss": 0.3351,
+      "step": 458
+    },
+    {
+      "epoch": 3.451127819548872,
+      "grad_norm": 0.23317522982179784,
+      "learning_rate": 1.9185621297379155e-06,
+      "loss": 0.3397,
+      "step": 459
+    },
+    {
+      "epoch": 3.4586466165413534,
+      "grad_norm": 0.21300234695531156,
+      "learning_rate": 1.9180702029492114e-06,
+      "loss": 0.3223,
+      "step": 460
+    },
+    {
+      "epoch": 3.4661654135338344,
+      "grad_norm": 0.2249693853422599,
+      "learning_rate": 1.9175768583393843e-06,
+      "loss": 0.3567,
+      "step": 461
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.22983527513324306,
+      "learning_rate": 1.9170820966703297e-06,
+      "loss": 0.3473,
+      "step": 462
+    },
+    {
+      "epoch": 3.481203007518797,
+      "grad_norm": 0.22114415392483444,
+      "learning_rate": 1.9165859187061336e-06,
+      "loss": 0.3336,
+      "step": 463
+    },
+    {
+      "epoch": 3.488721804511278,
+      "grad_norm": 0.2152687083439769,
+      "learning_rate": 1.9160883252130674e-06,
+      "loss": 0.3299,
+      "step": 464
+    },
+    {
+      "epoch": 3.4962406015037595,
+      "grad_norm": 0.20830543497209236,
+      "learning_rate": 1.9155893169595898e-06,
+      "loss": 0.3334,
+      "step": 465
+    },
+    {
+      "epoch": 3.5037593984962405,
+      "grad_norm": 0.2225749526606335,
+      "learning_rate": 1.9150888947163436e-06,
+      "loss": 0.3317,
+      "step": 466
+    },
+    {
+      "epoch": 3.511278195488722,
+      "grad_norm": 0.2180426536362119,
+      "learning_rate": 1.914587059256155e-06,
+      "loss": 0.3289,
+      "step": 467
+    },
+    {
+      "epoch": 3.518796992481203,
+      "grad_norm": 0.21369337975097136,
+      "learning_rate": 1.9140838113540346e-06,
+      "loss": 0.3275,
+      "step": 468
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.19194373594086006,
+      "learning_rate": 1.913579151787172e-06,
+      "loss": 0.318,
+      "step": 469
+    },
+    {
+      "epoch": 3.5338345864661656,
+      "grad_norm": 0.2163251555121206,
+      "learning_rate": 1.913073081334938e-06,
+      "loss": 0.3328,
+      "step": 470
+    },
+    {
+      "epoch": 3.5413533834586466,
+      "grad_norm": 0.21112026539941212,
+      "learning_rate": 1.912565600778882e-06,
+      "loss": 0.3258,
+      "step": 471
+    },
+    {
+      "epoch": 3.548872180451128,
+      "grad_norm": 0.2199466380762351,
+      "learning_rate": 1.912056710902732e-06,
+      "loss": 0.3414,
+      "step": 472
+    },
+    {
+      "epoch": 3.556390977443609,
+      "grad_norm": 0.22639118546753417,
+      "learning_rate": 1.911546412492391e-06,
+      "loss": 0.3391,
+      "step": 473
+    },
+    {
+      "epoch": 3.56390977443609,
+      "grad_norm": 0.2275351069768873,
+      "learning_rate": 1.9110347063359382e-06,
+      "loss": 0.3383,
+      "step": 474
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.20562381551625405,
+      "learning_rate": 1.910521593223627e-06,
+      "loss": 0.328,
+      "step": 475
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.20505532659496206,
+      "learning_rate": 1.910007073947883e-06,
+      "loss": 0.3272,
+      "step": 476
+    },
+    {
+      "epoch": 3.5864661654135337,
+      "grad_norm": 0.20711414205144435,
+      "learning_rate": 1.9094911493033035e-06,
+      "loss": 0.3291,
+      "step": 477
+    },
+    {
+      "epoch": 3.593984962406015,
+      "grad_norm": 0.21988340282251057,
+      "learning_rate": 1.908973820086657e-06,
+      "loss": 0.3396,
+      "step": 478
+    },
+    {
+      "epoch": 3.601503759398496,
+      "grad_norm": 0.2234402994624171,
+      "learning_rate": 1.9084550870968805e-06,
+      "loss": 0.3318,
+      "step": 479
+    },
+    {
+      "epoch": 3.6090225563909772,
+      "grad_norm": 0.22062845204918163,
+      "learning_rate": 1.9079349511350783e-06,
+      "loss": 0.3352,
+      "step": 480
+    },
+    {
+      "epoch": 3.6165413533834587,
+      "grad_norm": 0.21393552446547787,
+      "learning_rate": 1.9074134130045223e-06,
+      "loss": 0.3299,
+      "step": 481
+    },
+    {
+      "epoch": 3.6240601503759398,
+      "grad_norm": 0.21600780268405162,
+      "learning_rate": 1.9068904735106499e-06,
+      "loss": 0.3407,
+      "step": 482
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.1985301741538567,
+      "learning_rate": 1.9063661334610622e-06,
+      "loss": 0.3234,
+      "step": 483
+    },
+    {
+      "epoch": 3.6390977443609023,
+      "grad_norm": 0.2026474383963244,
+      "learning_rate": 1.9058403936655232e-06,
+      "loss": 0.3267,
+      "step": 484
+    },
+    {
+      "epoch": 3.6466165413533833,
+      "grad_norm": 0.2343507963816079,
+      "learning_rate": 1.905313254935959e-06,
+      "loss": 0.3362,
+      "step": 485
+    },
+    {
+      "epoch": 3.654135338345865,
+      "grad_norm": 0.21130572752024684,
+      "learning_rate": 1.9047847180864558e-06,
+      "loss": 0.334,
+      "step": 486
+    },
+    {
+      "epoch": 3.661654135338346,
+      "grad_norm": 0.21748731087007897,
+      "learning_rate": 1.9042547839332595e-06,
+      "loss": 0.3265,
+      "step": 487
+    },
+    {
+      "epoch": 3.6691729323308273,
+      "grad_norm": 0.2077749330904562,
+      "learning_rate": 1.9037234532947735e-06,
+      "loss": 0.3311,
+      "step": 488
+    },
+    {
+      "epoch": 3.6766917293233083,
+      "grad_norm": 0.19284597189773822,
+      "learning_rate": 1.9031907269915574e-06,
+      "loss": 0.3276,
+      "step": 489
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.2334839104691021,
+      "learning_rate": 1.9026566058463274e-06,
+      "loss": 0.3461,
+      "step": 490
+    },
+    {
+      "epoch": 3.6917293233082704,
+      "grad_norm": 0.23316731459324055,
+      "learning_rate": 1.9021210906839527e-06,
+      "loss": 0.3347,
+      "step": 491
+    },
+    {
+      "epoch": 3.699248120300752,
+      "grad_norm": 0.24050545802424847,
+      "learning_rate": 1.9015841823314558e-06,
+      "loss": 0.3384,
+      "step": 492
+    },
+    {
+      "epoch": 3.706766917293233,
+      "grad_norm": 0.2130701904002796,
+      "learning_rate": 1.901045881618011e-06,
+      "loss": 0.3214,
+      "step": 493
+    },
+    {
+      "epoch": 3.7142857142857144,
+      "grad_norm": 0.22128657971863466,
+      "learning_rate": 1.9005061893749427e-06,
+      "loss": 0.3514,
+      "step": 494
+    },
+    {
+      "epoch": 3.7218045112781954,
+      "grad_norm": 0.2169562309171576,
+      "learning_rate": 1.899965106435724e-06,
+      "loss": 0.3291,
+      "step": 495
+    },
+    {
+      "epoch": 3.7293233082706765,
+      "grad_norm": 0.2187018840340956,
+      "learning_rate": 1.899422633635976e-06,
+      "loss": 0.3341,
+      "step": 496
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.22087665503617396,
+      "learning_rate": 1.8988787718134664e-06,
+      "loss": 0.3445,
+      "step": 497
+    },
+    {
+      "epoch": 3.744360902255639,
+      "grad_norm": 0.23887388574425591,
+      "learning_rate": 1.8983335218081078e-06,
+      "loss": 0.3521,
+      "step": 498
+    },
+    {
+      "epoch": 3.7518796992481205,
+      "grad_norm": 0.21330538161187873,
+      "learning_rate": 1.8977868844619569e-06,
+      "loss": 0.344,
+      "step": 499
+    },
+    {
+      "epoch": 3.7593984962406015,
+      "grad_norm": 0.21793545755309388,
+      "learning_rate": 1.8972388606192122e-06,
+      "loss": 0.3288,
+      "step": 500
+    },
+    {
+      "epoch": 3.7669172932330826,
+      "grad_norm": 0.24892926600867735,
+      "learning_rate": 1.8966894511262144e-06,
+      "loss": 0.3304,
+      "step": 501
+    },
+    {
+      "epoch": 3.774436090225564,
+      "grad_norm": 0.21390310070654378,
+      "learning_rate": 1.8961386568314435e-06,
+      "loss": 0.3293,
+      "step": 502
+    },
+    {
+      "epoch": 3.781954887218045,
+      "grad_norm": 0.23359180422685968,
+      "learning_rate": 1.8955864785855185e-06,
+      "loss": 0.34,
+      "step": 503
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.20309318237905918,
+      "learning_rate": 1.8950329172411951e-06,
+      "loss": 0.3146,
+      "step": 504
+    },
+    {
+      "epoch": 3.7969924812030076,
+      "grad_norm": 0.2110201987511458,
+      "learning_rate": 1.8944779736533661e-06,
+      "loss": 0.3411,
+      "step": 505
+    },
+    {
+      "epoch": 3.8045112781954886,
+      "grad_norm": 0.2334955194899201,
+      "learning_rate": 1.8939216486790574e-06,
+      "loss": 0.3181,
+      "step": 506
+    },
+    {
+      "epoch": 3.8120300751879697,
+      "grad_norm": 0.22874042132852251,
+      "learning_rate": 1.8933639431774298e-06,
+      "loss": 0.3304,
+      "step": 507
+    },
+    {
+      "epoch": 3.819548872180451,
+      "grad_norm": 0.21090686472207962,
+      "learning_rate": 1.8928048580097756e-06,
+      "loss": 0.3198,
+      "step": 508
+    },
+    {
+      "epoch": 3.827067669172932,
+      "grad_norm": 0.22684233882121987,
+      "learning_rate": 1.8922443940395168e-06,
+      "loss": 0.3417,
+      "step": 509
+    },
+    {
+      "epoch": 3.8345864661654137,
+      "grad_norm": 0.21179465664262911,
+      "learning_rate": 1.891682552132206e-06,
+      "loss": 0.3329,
+      "step": 510
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.22598056964811206,
+      "learning_rate": 1.8911193331555232e-06,
+      "loss": 0.3326,
+      "step": 511
+    },
+    {
+      "epoch": 3.8496240601503757,
+      "grad_norm": 0.21610947265600972,
+      "learning_rate": 1.8905547379792757e-06,
+      "loss": 0.3319,
+      "step": 512
+    },
+    {
+      "epoch": 3.857142857142857,
+      "grad_norm": 0.21779291602588044,
+      "learning_rate": 1.8899887674753957e-06,
+      "loss": 0.3251,
+      "step": 513
+    },
+    {
+      "epoch": 3.8646616541353382,
+      "grad_norm": 0.2513852835058881,
+      "learning_rate": 1.8894214225179387e-06,
+      "loss": 0.3254,
+      "step": 514
+    },
+    {
+      "epoch": 3.8721804511278197,
+      "grad_norm": 0.22475609694155102,
+      "learning_rate": 1.8888527039830841e-06,
+      "loss": 0.317,
+      "step": 515
+    },
+    {
+      "epoch": 3.8796992481203008,
+      "grad_norm": 0.21266437501327956,
+      "learning_rate": 1.8882826127491318e-06,
+      "loss": 0.3403,
+      "step": 516
+    },
+    {
+      "epoch": 3.887218045112782,
+      "grad_norm": 0.23185631196652215,
+      "learning_rate": 1.887711149696502e-06,
+      "loss": 0.3368,
+      "step": 517
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.22528722260561196,
+      "learning_rate": 1.887138315707733e-06,
+      "loss": 0.3351,
+      "step": 518
+    },
+    {
+      "epoch": 3.9022556390977443,
+      "grad_norm": 0.23307245152914094,
+      "learning_rate": 1.8865641116674808e-06,
+      "loss": 0.3441,
+      "step": 519
+    },
+    {
+      "epoch": 3.909774436090226,
+      "grad_norm": 0.21897232804319286,
+      "learning_rate": 1.885988538462517e-06,
+      "loss": 0.3286,
+      "step": 520
+    },
+    {
+      "epoch": 3.917293233082707,
+      "grad_norm": 0.21331210508477158,
+      "learning_rate": 1.8854115969817276e-06,
+      "loss": 0.3406,
+      "step": 521
+    },
+    {
+      "epoch": 3.924812030075188,
+      "grad_norm": 0.21413336215847548,
+      "learning_rate": 1.8848332881161121e-06,
+      "loss": 0.3392,
+      "step": 522
+    },
+    {
+      "epoch": 3.932330827067669,
+      "grad_norm": 0.2196889146022895,
+      "learning_rate": 1.8842536127587812e-06,
+      "loss": 0.3422,
+      "step": 523
+    },
+    {
+      "epoch": 3.9398496240601504,
+      "grad_norm": 0.2099742309523454,
+      "learning_rate": 1.883672571804956e-06,
+      "loss": 0.3327,
+      "step": 524
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.21348444271818598,
+      "learning_rate": 1.8830901661519672e-06,
+      "loss": 0.3398,
+      "step": 525
+    },
+    {
+      "epoch": 3.954887218045113,
+      "grad_norm": 0.21850786409887793,
+      "learning_rate": 1.8825063966992523e-06,
+      "loss": 0.3107,
+      "step": 526
+    },
+    {
+      "epoch": 3.962406015037594,
+      "grad_norm": 0.1974575375798906,
+      "learning_rate": 1.8819212643483548e-06,
+      "loss": 0.3416,
+      "step": 527
+    },
+    {
+      "epoch": 3.969924812030075,
+      "grad_norm": 0.23094255823931867,
+      "learning_rate": 1.8813347700029242e-06,
+      "loss": 0.3315,
+      "step": 528
+    },
+    {
+      "epoch": 3.9774436090225564,
+      "grad_norm": 0.21295296318298487,
+      "learning_rate": 1.8807469145687127e-06,
+      "loss": 0.3195,
+      "step": 529
+    },
+    {
+      "epoch": 3.9849624060150375,
+      "grad_norm": 0.22557620628092856,
+      "learning_rate": 1.8801576989535741e-06,
+      "loss": 0.3247,
+      "step": 530
+    },
+    {
+      "epoch": 3.992481203007519,
+      "grad_norm": 0.21951807479235183,
+      "learning_rate": 1.8795671240674631e-06,
+      "loss": 0.3464,
+      "step": 531
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2122438357956682,
+      "learning_rate": 1.8789751908224336e-06,
+      "loss": 0.3362,
+      "step": 532
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.4019816219806671,
+      "eval_runtime": 1.6754,
+      "eval_samples_per_second": 195.173,
+      "eval_steps_per_second": 3.581,
+      "step": 532
+    },
+    {
+      "epoch": 4.007518796992481,
+      "grad_norm": 0.19683304314282413,
+      "learning_rate": 1.8783819001326378e-06,
+      "loss": 0.3088,
+      "step": 533
+    },
+    {
+      "epoch": 4.015037593984962,
+      "grad_norm": 0.2227809413446793,
+      "learning_rate": 1.8777872529143233e-06,
+      "loss": 0.3348,
+      "step": 534
+    },
+    {
+      "epoch": 4.022556390977444,
+      "grad_norm": 0.21921109198433691,
+      "learning_rate": 1.8771912500858333e-06,
+      "loss": 0.3287,
+      "step": 535
+    },
+    {
+      "epoch": 4.030075187969925,
+      "grad_norm": 0.3245274940627274,
+      "learning_rate": 1.8765938925676044e-06,
+      "loss": 0.328,
+      "step": 536
+    },
+    {
+      "epoch": 4.037593984962406,
+      "grad_norm": 0.19851784518823865,
+      "learning_rate": 1.8759951812821654e-06,
+      "loss": 0.309,
+      "step": 537
+    },
+    {
+      "epoch": 4.045112781954887,
+      "grad_norm": 0.21207942295902033,
+      "learning_rate": 1.8753951171541357e-06,
+      "loss": 0.3165,
+      "step": 538
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.21443563234599688,
+      "learning_rate": 1.8747937011102237e-06,
+      "loss": 0.3311,
+      "step": 539
+    },
+    {
+      "epoch": 4.06015037593985,
+      "grad_norm": 0.21713048310239663,
+      "learning_rate": 1.8741909340792259e-06,
+      "loss": 0.3336,
+      "step": 540
+    },
+    {
+      "epoch": 4.067669172932331,
+      "grad_norm": 0.21182033226493166,
+      "learning_rate": 1.8735868169920255e-06,
+      "loss": 0.3206,
+      "step": 541
+    },
+    {
+      "epoch": 4.075187969924812,
+      "grad_norm": 0.23136010576597213,
+      "learning_rate": 1.8729813507815901e-06,
+      "loss": 0.3272,
+      "step": 542
+    },
+    {
+      "epoch": 4.082706766917293,
+      "grad_norm": 0.2084029729611335,
+      "learning_rate": 1.8723745363829711e-06,
+      "loss": 0.3248,
+      "step": 543
+    },
+    {
+      "epoch": 4.090225563909774,
+      "grad_norm": 0.2145151109206258,
+      "learning_rate": 1.8717663747333016e-06,
+      "loss": 0.3298,
+      "step": 544
+    },
+    {
+      "epoch": 4.097744360902255,
+      "grad_norm": 0.21163175270183687,
+      "learning_rate": 1.871156866771796e-06,
+      "loss": 0.3199,
+      "step": 545
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.22443712132378746,
+      "learning_rate": 1.870546013439748e-06,
+      "loss": 0.3231,
+      "step": 546
+    },
+    {
+      "epoch": 4.112781954887218,
+      "grad_norm": 0.20430874581933128,
+      "learning_rate": 1.8699338156805275e-06,
+      "loss": 0.3223,
+      "step": 547
+    },
+    {
+      "epoch": 4.120300751879699,
+      "grad_norm": 0.20680275648912785,
+      "learning_rate": 1.8693202744395827e-06,
+      "loss": 0.3371,
+      "step": 548
+    },
+    {
+      "epoch": 4.12781954887218,
+      "grad_norm": 0.22463384914471105,
+      "learning_rate": 1.8687053906644347e-06,
+      "loss": 0.3341,
+      "step": 549
+    },
+    {
+      "epoch": 4.135338345864661,
+      "grad_norm": 0.2181399665790619,
+      "learning_rate": 1.8680891653046796e-06,
+      "loss": 0.3368,
+      "step": 550
+    },
+    {
+      "epoch": 4.142857142857143,
+      "grad_norm": 0.21115175404144435,
+      "learning_rate": 1.8674715993119842e-06,
+      "loss": 0.3245,
+      "step": 551
+    },
+    {
+      "epoch": 4.150375939849624,
+      "grad_norm": 0.2011539008263431,
+      "learning_rate": 1.866852693640086e-06,
+      "loss": 0.3078,
+      "step": 552
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.23018410202238473,
+      "learning_rate": 1.866232449244792e-06,
+      "loss": 0.3261,
+      "step": 553
+    },
+    {
+      "epoch": 4.165413533834586,
+      "grad_norm": 0.20988125958274145,
+      "learning_rate": 1.8656108670839764e-06,
+      "loss": 0.3256,
+      "step": 554
+    },
+    {
+      "epoch": 4.172932330827067,
+      "grad_norm": 0.2100829504993273,
+      "learning_rate": 1.8649879481175788e-06,
+      "loss": 0.3303,
+      "step": 555
+    },
+    {
+      "epoch": 4.180451127819548,
+      "grad_norm": 0.21686818222789023,
+      "learning_rate": 1.8643636933076036e-06,
+      "loss": 0.3382,
+      "step": 556
+    },
+    {
+      "epoch": 4.18796992481203,
+      "grad_norm": 0.22971238936938962,
+      "learning_rate": 1.8637381036181188e-06,
+      "loss": 0.322,
+      "step": 557
+    },
+    {
+      "epoch": 4.195488721804511,
+      "grad_norm": 0.2521295872783236,
+      "learning_rate": 1.863111180015253e-06,
+      "loss": 0.3267,
+      "step": 558
+    },
+    {
+      "epoch": 4.203007518796992,
+      "grad_norm": 0.20168449244778702,
+      "learning_rate": 1.8624829234671956e-06,
+      "loss": 0.3067,
+      "step": 559
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.21395299256511263,
+      "learning_rate": 1.8618533349441936e-06,
+      "loss": 0.3262,
+      "step": 560
+    },
+    {
+      "epoch": 4.2180451127819545,
+      "grad_norm": 0.233290534692693,
+      "learning_rate": 1.8612224154185524e-06,
+      "loss": 0.3288,
+      "step": 561
+    },
+    {
+      "epoch": 4.225563909774436,
+      "grad_norm": 0.22835254332927807,
+      "learning_rate": 1.8605901658646316e-06,
+      "loss": 0.3389,
+      "step": 562
+    },
+    {
+      "epoch": 4.2330827067669174,
+      "grad_norm": 0.22009448553273145,
+      "learning_rate": 1.8599565872588454e-06,
+      "loss": 0.3336,
+      "step": 563
+    },
+    {
+      "epoch": 4.2406015037593985,
+      "grad_norm": 0.20349127004485415,
+      "learning_rate": 1.859321680579661e-06,
+      "loss": 0.3264,
+      "step": 564
+    },
+    {
+      "epoch": 4.2481203007518795,
+      "grad_norm": 0.20577915297872051,
+      "learning_rate": 1.8586854468075955e-06,
+      "loss": 0.325,
+      "step": 565
+    },
+    {
+      "epoch": 4.2556390977443606,
+      "grad_norm": 0.20342142294948548,
+      "learning_rate": 1.8580478869252167e-06,
+      "loss": 0.3326,
+      "step": 566
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.21462067824412764,
+      "learning_rate": 1.8574090019171393e-06,
+      "loss": 0.3283,
+      "step": 567
+    },
+    {
+      "epoch": 4.2706766917293235,
+      "grad_norm": 0.2096072400914041,
+      "learning_rate": 1.8567687927700252e-06,
+      "loss": 0.3149,
+      "step": 568
+    },
+    {
+      "epoch": 4.2781954887218046,
+      "grad_norm": 0.20956817753165383,
+      "learning_rate": 1.856127260472581e-06,
+      "loss": 0.3293,
+      "step": 569
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.22892166767350813,
+      "learning_rate": 1.8554844060155569e-06,
+      "loss": 0.3203,
+      "step": 570
+    },
+    {
+      "epoch": 4.293233082706767,
+      "grad_norm": 0.22280843986964585,
+      "learning_rate": 1.854840230391744e-06,
+      "loss": 0.3257,
+      "step": 571
+    },
+    {
+      "epoch": 4.3007518796992485,
+      "grad_norm": 0.2094933778702391,
+      "learning_rate": 1.8541947345959753e-06,
+      "loss": 0.339,
+      "step": 572
+    },
+    {
+      "epoch": 4.30827067669173,
+      "grad_norm": 0.22183798663521667,
+      "learning_rate": 1.8535479196251215e-06,
+      "loss": 0.3305,
+      "step": 573
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.21040540014648454,
+      "learning_rate": 1.852899786478091e-06,
+      "loss": 0.3259,
+      "step": 574
+    },
+    {
+      "epoch": 4.323308270676692,
+      "grad_norm": 0.20020508433274356,
+      "learning_rate": 1.8522503361558273e-06,
+      "loss": 0.3244,
+      "step": 575
+    },
+    {
+      "epoch": 4.330827067669173,
+      "grad_norm": 0.19400477596987956,
+      "learning_rate": 1.8515995696613093e-06,
+      "loss": 0.3225,
+      "step": 576
+    },
+    {
+      "epoch": 4.338345864661654,
+      "grad_norm": 0.2029098675950304,
+      "learning_rate": 1.8509474879995475e-06,
+      "loss": 0.3157,
+      "step": 577
+    },
+    {
+      "epoch": 4.345864661654136,
+      "grad_norm": 0.20478090612746774,
+      "learning_rate": 1.8502940921775837e-06,
+      "loss": 0.3149,
+      "step": 578
+    },
+    {
+      "epoch": 4.353383458646617,
+      "grad_norm": 0.21409121008483814,
+      "learning_rate": 1.8496393832044893e-06,
+      "loss": 0.3293,
+      "step": 579
+    },
+    {
+      "epoch": 4.360902255639098,
+      "grad_norm": 0.2201100976583189,
+      "learning_rate": 1.848983362091364e-06,
+      "loss": 0.3361,
+      "step": 580
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.20986838623303375,
+      "learning_rate": 1.848326029851333e-06,
+      "loss": 0.3164,
+      "step": 581
+    },
+    {
+      "epoch": 4.37593984962406,
+      "grad_norm": 0.20331762188437297,
+      "learning_rate": 1.8476673874995477e-06,
+      "loss": 0.331,
+      "step": 582
+    },
+    {
+      "epoch": 4.383458646616542,
+      "grad_norm": 0.21203724060408413,
+      "learning_rate": 1.8470074360531813e-06,
+      "loss": 0.33,
+      "step": 583
+    },
+    {
+      "epoch": 4.390977443609023,
+      "grad_norm": 0.21466063536589736,
+      "learning_rate": 1.84634617653143e-06,
+      "loss": 0.3262,
+      "step": 584
+    },
+    {
+      "epoch": 4.398496240601504,
+      "grad_norm": 0.19980651328534854,
+      "learning_rate": 1.8456836099555085e-06,
+      "loss": 0.3305,
+      "step": 585
+    },
+    {
+      "epoch": 4.406015037593985,
+      "grad_norm": 0.2073405403918328,
+      "learning_rate": 1.8450197373486526e-06,
+      "loss": 0.3292,
+      "step": 586
+    },
+    {
+      "epoch": 4.413533834586466,
+      "grad_norm": 0.23276016730959861,
+      "learning_rate": 1.8443545597361122e-06,
+      "loss": 0.3262,
+      "step": 587
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.2087976777991402,
+      "learning_rate": 1.8436880781451543e-06,
+      "loss": 0.3267,
+      "step": 588
+    },
+    {
+      "epoch": 4.428571428571429,
+      "grad_norm": 0.21319769227847604,
+      "learning_rate": 1.8430202936050594e-06,
+      "loss": 0.3206,
+      "step": 589
+    },
+    {
+      "epoch": 4.43609022556391,
+      "grad_norm": 0.22329669347376965,
+      "learning_rate": 1.8423512071471204e-06,
+      "loss": 0.3197,
+      "step": 590
+    },
+    {
+      "epoch": 4.443609022556391,
+      "grad_norm": 0.20179009057868758,
+      "learning_rate": 1.84168081980464e-06,
+      "loss": 0.3214,
+      "step": 591
+    },
+    {
+      "epoch": 4.451127819548872,
+      "grad_norm": 0.21218711398634865,
+      "learning_rate": 1.841009132612931e-06,
+      "loss": 0.3281,
+      "step": 592
+    },
+    {
+      "epoch": 4.458646616541353,
+      "grad_norm": 0.24346079652859673,
+      "learning_rate": 1.8403361466093123e-06,
+      "loss": 0.3274,
+      "step": 593
+    },
+    {
+      "epoch": 4.466165413533835,
+      "grad_norm": 0.23942563245336926,
+      "learning_rate": 1.8396618628331101e-06,
+      "loss": 0.3306,
+      "step": 594
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.21109188524439826,
+      "learning_rate": 1.8389862823256542e-06,
+      "loss": 0.327,
+      "step": 595
+    },
+    {
+      "epoch": 4.481203007518797,
+      "grad_norm": 0.2141861728018816,
+      "learning_rate": 1.8383094061302765e-06,
+      "loss": 0.3159,
+      "step": 596
+    },
+    {
+      "epoch": 4.488721804511278,
+      "grad_norm": 0.20319860539025952,
+      "learning_rate": 1.8376312352923105e-06,
+      "loss": 0.3294,
+      "step": 597
+    },
+    {
+      "epoch": 4.496240601503759,
+      "grad_norm": 0.21783274602762182,
+      "learning_rate": 1.8369517708590885e-06,
+      "loss": 0.3368,
+      "step": 598
+    },
+    {
+      "epoch": 4.503759398496241,
+      "grad_norm": 0.2270729707501999,
+      "learning_rate": 1.8362710138799415e-06,
+      "loss": 0.3281,
+      "step": 599
+    },
+    {
+      "epoch": 4.511278195488722,
+      "grad_norm": 0.2333445128026975,
+      "learning_rate": 1.8355889654061959e-06,
+      "loss": 0.3421,
+      "step": 600
+    },
+    {
+      "epoch": 4.518796992481203,
+      "grad_norm": 0.21554522669500226,
+      "learning_rate": 1.8349056264911729e-06,
+      "loss": 0.3192,
+      "step": 601
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.20821536041887787,
+      "learning_rate": 1.834220998190186e-06,
+      "loss": 0.3146,
+      "step": 602
+    },
+    {
+      "epoch": 4.533834586466165,
+      "grad_norm": 0.21783267902928094,
+      "learning_rate": 1.8335350815605414e-06,
+      "loss": 0.3297,
+      "step": 603
+    },
+    {
+      "epoch": 4.541353383458647,
+      "grad_norm": 0.22366919305990812,
+      "learning_rate": 1.8328478776615333e-06,
+      "loss": 0.3313,
+      "step": 604
+    },
+    {
+      "epoch": 4.548872180451128,
+      "grad_norm": 0.23002685292454794,
+      "learning_rate": 1.8321593875544449e-06,
+      "loss": 0.3267,
+      "step": 605
+    },
+    {
+      "epoch": 4.556390977443609,
+      "grad_norm": 0.23332720009942567,
+      "learning_rate": 1.8314696123025452e-06,
+      "loss": 0.331,
+      "step": 606
+    },
+    {
+      "epoch": 4.56390977443609,
+      "grad_norm": 0.21092062447114457,
+      "learning_rate": 1.8307785529710884e-06,
+      "loss": 0.3127,
+      "step": 607
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.23376631509642226,
+      "learning_rate": 1.8300862106273111e-06,
+      "loss": 0.3357,
+      "step": 608
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.21302148488421058,
+      "learning_rate": 1.8293925863404325e-06,
+      "loss": 0.3284,
+      "step": 609
+    },
+    {
+      "epoch": 4.586466165413534,
+      "grad_norm": 0.2107085442070409,
+      "learning_rate": 1.8286976811816504e-06,
+      "loss": 0.3237,
+      "step": 610
+    },
+    {
+      "epoch": 4.593984962406015,
+      "grad_norm": 0.21529936164667476,
+      "learning_rate": 1.8280014962241408e-06,
+      "loss": 0.3072,
+      "step": 611
+    },
+    {
+      "epoch": 4.601503759398496,
+      "grad_norm": 0.2062313181751271,
+      "learning_rate": 1.8273040325430573e-06,
+      "loss": 0.3242,
+      "step": 612
+    },
+    {
+      "epoch": 4.609022556390977,
+      "grad_norm": 0.22006829634177041,
+      "learning_rate": 1.8266052912155265e-06,
+      "loss": 0.3258,
+      "step": 613
+    },
+    {
+      "epoch": 4.616541353383458,
+      "grad_norm": 0.20290593734554904,
+      "learning_rate": 1.8259052733206502e-06,
+      "loss": 0.3251,
+      "step": 614
+    },
+    {
+      "epoch": 4.62406015037594,
+      "grad_norm": 0.21248953910730098,
+      "learning_rate": 1.8252039799394993e-06,
+      "loss": 0.3213,
+      "step": 615
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.2212153604804533,
+      "learning_rate": 1.8245014121551172e-06,
+      "loss": 0.3331,
+      "step": 616
+    },
+    {
+      "epoch": 4.639097744360902,
+      "grad_norm": 0.20597942589850046,
+      "learning_rate": 1.8237975710525129e-06,
+      "loss": 0.3356,
+      "step": 617
+    },
+    {
+      "epoch": 4.646616541353383,
+      "grad_norm": 0.2140383470605456,
+      "learning_rate": 1.8230924577186632e-06,
+      "loss": 0.3295,
+      "step": 618
+    },
+    {
+      "epoch": 4.654135338345864,
+      "grad_norm": 0.22086451821351638,
+      "learning_rate": 1.82238607324251e-06,
+      "loss": 0.331,
+      "step": 619
+    },
+    {
+      "epoch": 4.661654135338345,
+      "grad_norm": 0.2193120945812253,
+      "learning_rate": 1.8216784187149567e-06,
+      "loss": 0.328,
+      "step": 620
+    },
+    {
+      "epoch": 4.669172932330827,
+      "grad_norm": 0.21269153442945485,
+      "learning_rate": 1.8209694952288702e-06,
+      "loss": 0.3197,
+      "step": 621
+    },
+    {
+      "epoch": 4.676691729323308,
+      "grad_norm": 0.23273630726283714,
+      "learning_rate": 1.8202593038790752e-06,
+      "loss": 0.3306,
+      "step": 622
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.20240845866192853,
+      "learning_rate": 1.8195478457623556e-06,
+      "loss": 0.3195,
+      "step": 623
+    },
+    {
+      "epoch": 4.69172932330827,
+      "grad_norm": 0.20178488031777755,
+      "learning_rate": 1.8188351219774515e-06,
+      "loss": 0.3126,
+      "step": 624
+    },
+    {
+      "epoch": 4.6992481203007515,
+      "grad_norm": 0.2140404914856119,
+      "learning_rate": 1.8181211336250569e-06,
+      "loss": 0.3246,
+      "step": 625
+    },
+    {
+      "epoch": 4.706766917293233,
+      "grad_norm": 0.19983278660759954,
+      "learning_rate": 1.8174058818078198e-06,
+      "loss": 0.3173,
+      "step": 626
+    },
+    {
+      "epoch": 4.714285714285714,
+      "grad_norm": 0.21865934756717245,
+      "learning_rate": 1.8166893676303384e-06,
+      "loss": 0.3172,
+      "step": 627
+    },
+    {
+      "epoch": 4.7218045112781954,
+      "grad_norm": 0.1971024910157171,
+      "learning_rate": 1.8159715921991609e-06,
+      "loss": 0.3074,
+      "step": 628
+    },
+    {
+      "epoch": 4.7293233082706765,
+      "grad_norm": 0.20503574613176542,
+      "learning_rate": 1.8152525566227838e-06,
+      "loss": 0.3091,
+      "step": 629
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.19998114104414777,
+      "learning_rate": 1.8145322620116487e-06,
+      "loss": 0.3197,
+      "step": 630
+    },
+    {
+      "epoch": 4.7443609022556394,
+      "grad_norm": 0.22581732560481396,
+      "learning_rate": 1.8138107094781426e-06,
+      "loss": 0.3327,
+      "step": 631
+    },
+    {
+      "epoch": 4.7518796992481205,
+      "grad_norm": 0.21566670833432028,
+      "learning_rate": 1.8130879001365942e-06,
+      "loss": 0.3032,
+      "step": 632
+    },
+    {
+      "epoch": 4.7593984962406015,
+      "grad_norm": 0.2257497799311242,
+      "learning_rate": 1.8123638351032739e-06,
+      "loss": 0.3296,
+      "step": 633
+    },
+    {
+      "epoch": 4.7669172932330826,
+      "grad_norm": 0.21994131956192217,
+      "learning_rate": 1.8116385154963912e-06,
+      "loss": 0.3231,
+      "step": 634
+    },
+    {
+      "epoch": 4.774436090225564,
+      "grad_norm": 0.21817518037957692,
+      "learning_rate": 1.8109119424360928e-06,
+      "loss": 0.3229,
+      "step": 635
+    },
+    {
+      "epoch": 4.7819548872180455,
+      "grad_norm": 0.21259796555017774,
+      "learning_rate": 1.8101841170444613e-06,
+      "loss": 0.3262,
+      "step": 636
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.22246999603476378,
+      "learning_rate": 1.8094550404455132e-06,
+      "loss": 0.3295,
+      "step": 637
+    },
+    {
+      "epoch": 4.796992481203008,
+      "grad_norm": 0.20201658108547707,
+      "learning_rate": 1.8087247137651982e-06,
+      "loss": 0.3187,
+      "step": 638
+    },
+    {
+      "epoch": 4.804511278195489,
+      "grad_norm": 0.22374673898127856,
+      "learning_rate": 1.8079931381313951e-06,
+      "loss": 0.3273,
+      "step": 639
+    },
+    {
+      "epoch": 4.81203007518797,
+      "grad_norm": 0.22671902039196323,
+      "learning_rate": 1.8072603146739124e-06,
+      "loss": 0.3252,
+      "step": 640
+    },
+    {
+      "epoch": 4.819548872180452,
+      "grad_norm": 0.20986977933811138,
+      "learning_rate": 1.8065262445244859e-06,
+      "loss": 0.324,
+      "step": 641
+    },
+    {
+      "epoch": 4.827067669172933,
+      "grad_norm": 0.2200512932074022,
+      "learning_rate": 1.8057909288167757e-06,
+      "loss": 0.3352,
+      "step": 642
+    },
+    {
+      "epoch": 4.834586466165414,
+      "grad_norm": 0.21390390790382163,
+      "learning_rate": 1.8050543686863666e-06,
+      "loss": 0.323,
+      "step": 643
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.21159571872683847,
+      "learning_rate": 1.8043165652707648e-06,
+      "loss": 0.3237,
+      "step": 644
+    },
+    {
+      "epoch": 4.849624060150376,
+      "grad_norm": 0.20628886503175456,
+      "learning_rate": 1.8035775197093963e-06,
+      "loss": 0.3192,
+      "step": 645
+    },
+    {
+      "epoch": 4.857142857142857,
+      "grad_norm": 0.22474368943086362,
+      "learning_rate": 1.8028372331436057e-06,
+      "loss": 0.3328,
+      "step": 646
+    },
+    {
+      "epoch": 4.864661654135339,
+      "grad_norm": 0.21459540910322406,
+      "learning_rate": 1.8020957067166542e-06,
+      "loss": 0.3201,
+      "step": 647
+    },
+    {
+      "epoch": 4.87218045112782,
+      "grad_norm": 0.2222395832313217,
+      "learning_rate": 1.8013529415737175e-06,
+      "loss": 0.332,
+      "step": 648
+    },
+    {
+      "epoch": 4.879699248120301,
+      "grad_norm": 0.2101263663769711,
+      "learning_rate": 1.8006089388618846e-06,
+      "loss": 0.32,
+      "step": 649
+    },
+    {
+      "epoch": 4.887218045112782,
+      "grad_norm": 0.21948895714308847,
+      "learning_rate": 1.7998636997301558e-06,
+      "loss": 0.3138,
+      "step": 650
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.21295298358059733,
+      "learning_rate": 1.7991172253294397e-06,
+      "loss": 0.3189,
+      "step": 651
+    },
+    {
+      "epoch": 4.902255639097744,
+      "grad_norm": 0.22574378397743394,
+      "learning_rate": 1.798369516812555e-06,
+      "loss": 0.3351,
+      "step": 652
+    },
+    {
+      "epoch": 4.909774436090226,
+      "grad_norm": 0.2034064589654031,
+      "learning_rate": 1.797620575334224e-06,
+      "loss": 0.3163,
+      "step": 653
+    },
+    {
+      "epoch": 4.917293233082707,
+      "grad_norm": 0.2432477917419446,
+      "learning_rate": 1.7968704020510739e-06,
+      "loss": 0.326,
+      "step": 654
+    },
+    {
+      "epoch": 4.924812030075188,
+      "grad_norm": 0.21475354412650804,
+      "learning_rate": 1.7961189981216345e-06,
+      "loss": 0.3218,
+      "step": 655
+    },
+    {
+      "epoch": 4.932330827067669,
+      "grad_norm": 0.20015463407837608,
+      "learning_rate": 1.7953663647063363e-06,
+      "loss": 0.3196,
+      "step": 656
+    },
+    {
+      "epoch": 4.93984962406015,
+      "grad_norm": 0.20463635727466584,
+      "learning_rate": 1.794612502967508e-06,
+      "loss": 0.3108,
+      "step": 657
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.21700254216093687,
+      "learning_rate": 1.793857414069375e-06,
+      "loss": 0.324,
+      "step": 658
+    },
+    {
+      "epoch": 4.954887218045113,
+      "grad_norm": 0.2263391523913246,
+      "learning_rate": 1.7931010991780591e-06,
+      "loss": 0.3242,
+      "step": 659
+    },
+    {
+      "epoch": 4.962406015037594,
+      "grad_norm": 0.20976589235009296,
+      "learning_rate": 1.7923435594615742e-06,
+      "loss": 0.3232,
+      "step": 660
+    },
+    {
+      "epoch": 4.969924812030075,
+      "grad_norm": 0.21562623452666757,
+      "learning_rate": 1.7915847960898266e-06,
+      "loss": 0.3307,
+      "step": 661
+    },
+    {
+      "epoch": 4.977443609022556,
+      "grad_norm": 0.22549177325574526,
+      "learning_rate": 1.790824810234612e-06,
+      "loss": 0.3278,
+      "step": 662
+    },
+    {
+      "epoch": 4.984962406015038,
+      "grad_norm": 0.21506501265119463,
+      "learning_rate": 1.7900636030696136e-06,
+      "loss": 0.3159,
+      "step": 663
+    },
+    {
+      "epoch": 4.992481203007519,
+      "grad_norm": 0.21648997672110742,
+      "learning_rate": 1.789301175770402e-06,
+      "loss": 0.3178,
+      "step": 664
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.22702375657759305,
+      "learning_rate": 1.7885375295144304e-06,
+      "loss": 0.3345,
+      "step": 665
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.40189385414123535,
+      "eval_runtime": 1.9502,
+      "eval_samples_per_second": 167.674,
+      "eval_steps_per_second": 3.077,
+      "step": 665
+    },
+    {
+      "epoch": 5.007518796992481,
+      "grad_norm": 0.23892498086084846,
+      "learning_rate": 1.7877726654810363e-06,
+      "loss": 0.3141,
+      "step": 666
+    },
+    {
+      "epoch": 5.015037593984962,
+      "grad_norm": 0.21387431326276987,
+      "learning_rate": 1.7870065848514364e-06,
+      "loss": 0.3147,
+      "step": 667
+    },
+    {
+      "epoch": 5.022556390977444,
+      "grad_norm": 0.21819644539994343,
+      "learning_rate": 1.7862392888087267e-06,
+      "loss": 0.3177,
+      "step": 668
+    },
+    {
+      "epoch": 5.030075187969925,
+      "grad_norm": 0.21631342389569838,
+      "learning_rate": 1.785470778537881e-06,
+      "loss": 0.3176,
+      "step": 669
+    },
+    {
+      "epoch": 5.037593984962406,
+      "grad_norm": 0.24458115896396057,
+      "learning_rate": 1.7847010552257467e-06,
+      "loss": 0.3282,
+      "step": 670
+    },
+    {
+      "epoch": 5.045112781954887,
+      "grad_norm": 0.23737966417738837,
+      "learning_rate": 1.7839301200610463e-06,
+      "loss": 0.3142,
+      "step": 671
+    },
+    {
+      "epoch": 5.052631578947368,
+      "grad_norm": 0.22340796926863182,
+      "learning_rate": 1.7831579742343727e-06,
+      "loss": 0.3188,
+      "step": 672
+    },
+    {
+      "epoch": 5.06015037593985,
+      "grad_norm": 0.2266683859119338,
+      "learning_rate": 1.7823846189381891e-06,
+      "loss": 0.3257,
+      "step": 673
+    },
+    {
+      "epoch": 5.067669172932331,
+      "grad_norm": 0.20544274095802417,
+      "learning_rate": 1.7816100553668258e-06,
+      "loss": 0.3219,
+      "step": 674
+    },
+    {
+      "epoch": 5.075187969924812,
+      "grad_norm": 0.21403398736382864,
+      "learning_rate": 1.7808342847164796e-06,
+      "loss": 0.3116,
+      "step": 675
+    },
+    {
+      "epoch": 5.082706766917293,
+      "grad_norm": 0.2311285775397372,
+      "learning_rate": 1.780057308185212e-06,
+      "loss": 0.3064,
+      "step": 676
+    },
+    {
+      "epoch": 5.090225563909774,
+      "grad_norm": 0.23172251425438767,
+      "learning_rate": 1.7792791269729456e-06,
+      "loss": 0.3099,
+      "step": 677
+    },
+    {
+      "epoch": 5.097744360902255,
+      "grad_norm": 0.2310375381210173,
+      "learning_rate": 1.7784997422814643e-06,
+      "loss": 0.3204,
+      "step": 678
+    },
+    {
+      "epoch": 5.105263157894737,
+      "grad_norm": 0.21863463898490895,
+      "learning_rate": 1.77771915531441e-06,
+      "loss": 0.323,
+      "step": 679
+    },
+    {
+      "epoch": 5.112781954887218,
+      "grad_norm": 0.2201920567724236,
+      "learning_rate": 1.776937367277282e-06,
+      "loss": 0.3236,
+      "step": 680
+    },
+    {
+      "epoch": 5.120300751879699,
+      "grad_norm": 0.21927260076284794,
+      "learning_rate": 1.7761543793774343e-06,
+      "loss": 0.3136,
+      "step": 681
+    },
+    {
+      "epoch": 5.12781954887218,
+      "grad_norm": 0.20807471027359115,
+      "learning_rate": 1.7753701928240733e-06,
+      "loss": 0.3085,
+      "step": 682
+    },
+    {
+      "epoch": 5.135338345864661,
+      "grad_norm": 0.2031050056336962,
+      "learning_rate": 1.7745848088282575e-06,
+      "loss": 0.3393,
+      "step": 683
+    },
+    {
+      "epoch": 5.142857142857143,
+      "grad_norm": 0.22329285372196572,
+      "learning_rate": 1.7737982286028937e-06,
+      "loss": 0.3058,
+      "step": 684
+    },
+    {
+      "epoch": 5.150375939849624,
+      "grad_norm": 0.21949790457938556,
+      "learning_rate": 1.773010453362737e-06,
+      "loss": 0.3205,
+      "step": 685
+    },
+    {
+      "epoch": 5.157894736842105,
+      "grad_norm": 0.20677774217736627,
+      "learning_rate": 1.7722214843243873e-06,
+      "loss": 0.3208,
+      "step": 686
+    },
+    {
+      "epoch": 5.165413533834586,
+      "grad_norm": 0.22895437879279737,
+      "learning_rate": 1.771431322706288e-06,
+      "loss": 0.3281,
+      "step": 687
+    },
+    {
+      "epoch": 5.172932330827067,
+      "grad_norm": 0.21160033167383413,
+      "learning_rate": 1.7706399697287258e-06,
+      "loss": 0.3189,
+      "step": 688
+    },
+    {
+      "epoch": 5.180451127819548,
+      "grad_norm": 0.22687847587045343,
+      "learning_rate": 1.769847426613825e-06,
+      "loss": 0.3325,
+      "step": 689
+    },
+    {
+      "epoch": 5.18796992481203,
+      "grad_norm": 0.21556988035815947,
+      "learning_rate": 1.76905369458555e-06,
+      "loss": 0.3154,
+      "step": 690
+    },
+    {
+      "epoch": 5.195488721804511,
+      "grad_norm": 0.20818968304553498,
+      "learning_rate": 1.7682587748696996e-06,
+      "loss": 0.3202,
+      "step": 691
+    },
+    {
+      "epoch": 5.203007518796992,
+      "grad_norm": 0.21278188745155407,
+      "learning_rate": 1.7674626686939077e-06,
+      "loss": 0.3098,
+      "step": 692
+    },
+    {
+      "epoch": 5.2105263157894735,
+      "grad_norm": 0.225081065950259,
+      "learning_rate": 1.766665377287641e-06,
+      "loss": 0.3307,
+      "step": 693
+    },
+    {
+      "epoch": 5.2180451127819545,
+      "grad_norm": 0.21930239554532296,
+      "learning_rate": 1.7658669018821952e-06,
+      "loss": 0.3249,
+      "step": 694
+    },
+    {
+      "epoch": 5.225563909774436,
+      "grad_norm": 0.22322503866247112,
+      "learning_rate": 1.7650672437106957e-06,
+      "loss": 0.304,
+      "step": 695
+    },
+    {
+      "epoch": 5.2330827067669174,
+      "grad_norm": 0.21453239681430245,
+      "learning_rate": 1.7642664040080937e-06,
+      "loss": 0.3206,
+      "step": 696
+    },
+    {
+      "epoch": 5.2406015037593985,
+      "grad_norm": 0.20644616319998588,
+      "learning_rate": 1.763464384011166e-06,
+      "loss": 0.3095,
+      "step": 697
+    },
+    {
+      "epoch": 5.2481203007518795,
+      "grad_norm": 0.20973578562848572,
+      "learning_rate": 1.762661184958511e-06,
+      "loss": 0.3189,
+      "step": 698
+    },
+    {
+      "epoch": 5.2556390977443606,
+      "grad_norm": 0.2183337241315401,
+      "learning_rate": 1.7618568080905491e-06,
+      "loss": 0.3201,
+      "step": 699
+    },
+    {
+      "epoch": 5.2631578947368425,
+      "grad_norm": 0.20360430389330117,
+      "learning_rate": 1.7610512546495192e-06,
+      "loss": 0.307,
+      "step": 700
+    },
+    {
+      "epoch": 5.2706766917293235,
+      "grad_norm": 0.20225367651152637,
+      "learning_rate": 1.7602445258794772e-06,
+      "loss": 0.3037,
+      "step": 701
+    },
+    {
+      "epoch": 5.2781954887218046,
+      "grad_norm": 0.20745155558894363,
+      "learning_rate": 1.759436623026294e-06,
+      "loss": 0.2981,
+      "step": 702
+    },
+    {
+      "epoch": 5.285714285714286,
+      "grad_norm": 0.21390919300477093,
+      "learning_rate": 1.7586275473376539e-06,
+      "loss": 0.331,
+      "step": 703
+    },
+    {
+      "epoch": 5.293233082706767,
+      "grad_norm": 0.20877875528251802,
+      "learning_rate": 1.7578173000630525e-06,
+      "loss": 0.3071,
+      "step": 704
+    },
+    {
+      "epoch": 5.3007518796992485,
+      "grad_norm": 0.2044291108102548,
+      "learning_rate": 1.7570058824537948e-06,
+      "loss": 0.3091,
+      "step": 705
+    },
+    {
+      "epoch": 5.30827067669173,
+      "grad_norm": 0.22931786278593194,
+      "learning_rate": 1.7561932957629926e-06,
+      "loss": 0.3213,
+      "step": 706
+    },
+    {
+      "epoch": 5.315789473684211,
+      "grad_norm": 0.23567075631682882,
+      "learning_rate": 1.755379541245564e-06,
+      "loss": 0.3134,
+      "step": 707
+    },
+    {
+      "epoch": 5.323308270676692,
+      "grad_norm": 0.21840573408031794,
+      "learning_rate": 1.75456462015823e-06,
+      "loss": 0.3186,
+      "step": 708
+    },
+    {
+      "epoch": 5.330827067669173,
+      "grad_norm": 0.20002166312784125,
+      "learning_rate": 1.7537485337595137e-06,
+      "loss": 0.3098,
+      "step": 709
+    },
+    {
+      "epoch": 5.338345864661654,
+      "grad_norm": 0.2268571589910804,
+      "learning_rate": 1.7529312833097376e-06,
+      "loss": 0.3171,
+      "step": 710
+    },
+    {
+      "epoch": 5.345864661654136,
+      "grad_norm": 0.20811944768298693,
+      "learning_rate": 1.7521128700710216e-06,
+      "loss": 0.3088,
+      "step": 711
+    },
+    {
+      "epoch": 5.353383458646617,
+      "grad_norm": 0.19588271267158855,
+      "learning_rate": 1.7512932953072824e-06,
+      "loss": 0.3163,
+      "step": 712
+    },
+    {
+      "epoch": 5.360902255639098,
+      "grad_norm": 0.2125647454620394,
+      "learning_rate": 1.7504725602842287e-06,
+      "loss": 0.3118,
+      "step": 713
+    },
+    {
+      "epoch": 5.368421052631579,
+      "grad_norm": 0.2287017940771759,
+      "learning_rate": 1.7496506662693628e-06,
+      "loss": 0.3226,
+      "step": 714
+    },
+    {
+      "epoch": 5.37593984962406,
+      "grad_norm": 0.21184316643427245,
+      "learning_rate": 1.748827614531976e-06,
+      "loss": 0.3262,
+      "step": 715
+    },
+    {
+      "epoch": 5.383458646616542,
+      "grad_norm": 0.20966178164700763,
+      "learning_rate": 1.7480034063431478e-06,
+      "loss": 0.312,
+      "step": 716
+    },
+    {
+      "epoch": 5.390977443609023,
+      "grad_norm": 0.21021746634333796,
+      "learning_rate": 1.7471780429757434e-06,
+      "loss": 0.3178,
+      "step": 717
+    },
+    {
+      "epoch": 5.398496240601504,
+      "grad_norm": 0.2194510015853351,
+      "learning_rate": 1.7463515257044127e-06,
+      "loss": 0.3185,
+      "step": 718
+    },
+    {
+      "epoch": 5.406015037593985,
+      "grad_norm": 0.21418470750511118,
+      "learning_rate": 1.7455238558055862e-06,
+      "loss": 0.3123,
+      "step": 719
+    },
+    {
+      "epoch": 5.413533834586466,
+      "grad_norm": 0.2089387265520661,
+      "learning_rate": 1.744695034557476e-06,
+      "loss": 0.3128,
+      "step": 720
+    },
+    {
+      "epoch": 5.421052631578947,
+      "grad_norm": 0.20998448347287793,
+      "learning_rate": 1.7438650632400717e-06,
+      "loss": 0.322,
+      "step": 721
+    },
+    {
+      "epoch": 5.428571428571429,
+      "grad_norm": 0.20889373466674302,
+      "learning_rate": 1.7430339431351387e-06,
+      "loss": 0.3212,
+      "step": 722
+    },
+    {
+      "epoch": 5.43609022556391,
+      "grad_norm": 0.21755189146343046,
+      "learning_rate": 1.7422016755262167e-06,
+      "loss": 0.3191,
+      "step": 723
+    },
+    {
+      "epoch": 5.443609022556391,
+      "grad_norm": 0.2188595970986605,
+      "learning_rate": 1.7413682616986183e-06,
+      "loss": 0.3291,
+      "step": 724
+    },
+    {
+      "epoch": 5.451127819548872,
+      "grad_norm": 0.2186570910330145,
+      "learning_rate": 1.7405337029394247e-06,
+      "loss": 0.3075,
+      "step": 725
+    },
+    {
+      "epoch": 5.458646616541353,
+      "grad_norm": 0.1999017675041625,
+      "learning_rate": 1.7396980005374869e-06,
+      "loss": 0.3058,
+      "step": 726
+    },
+    {
+      "epoch": 5.466165413533835,
+      "grad_norm": 0.20610712313493904,
+      "learning_rate": 1.738861155783421e-06,
+      "loss": 0.3218,
+      "step": 727
+    },
+    {
+      "epoch": 5.473684210526316,
+      "grad_norm": 0.22310046251230534,
+      "learning_rate": 1.7380231699696077e-06,
+      "loss": 0.3255,
+      "step": 728
+    },
+    {
+      "epoch": 5.481203007518797,
+      "grad_norm": 0.20551963840527027,
+      "learning_rate": 1.73718404439019e-06,
+      "loss": 0.3234,
+      "step": 729
+    },
+    {
+      "epoch": 5.488721804511278,
+      "grad_norm": 0.20493366320847423,
+      "learning_rate": 1.7363437803410707e-06,
+      "loss": 0.3188,
+      "step": 730
+    },
+    {
+      "epoch": 5.496240601503759,
+      "grad_norm": 0.2102714498186948,
+      "learning_rate": 1.7355023791199113e-06,
+      "loss": 0.3268,
+      "step": 731
+    },
+    {
+      "epoch": 5.503759398496241,
+      "grad_norm": 0.20670362925066613,
+      "learning_rate": 1.7346598420261294e-06,
+      "loss": 0.3264,
+      "step": 732
+    },
+    {
+      "epoch": 5.511278195488722,
+      "grad_norm": 0.20343649736420952,
+      "learning_rate": 1.7338161703608958e-06,
+      "loss": 0.3093,
+      "step": 733
+    },
+    {
+      "epoch": 5.518796992481203,
+      "grad_norm": 0.2078074861865767,
+      "learning_rate": 1.7329713654271352e-06,
+      "loss": 0.3186,
+      "step": 734
+    },
+    {
+      "epoch": 5.526315789473684,
+      "grad_norm": 0.2158488195100477,
+      "learning_rate": 1.732125428529521e-06,
+      "loss": 0.3123,
+      "step": 735
+    },
+    {
+      "epoch": 5.533834586466165,
+      "grad_norm": 0.2067204986116603,
+      "learning_rate": 1.7312783609744753e-06,
+      "loss": 0.3001,
+      "step": 736
+    },
+    {
+      "epoch": 5.541353383458647,
+      "grad_norm": 0.24342475026134403,
+      "learning_rate": 1.7304301640701669e-06,
+      "loss": 0.3278,
+      "step": 737
+    },
+    {
+      "epoch": 5.548872180451128,
+      "grad_norm": 0.22867754964476195,
+      "learning_rate": 1.729580839126507e-06,
+      "loss": 0.3211,
+      "step": 738
+    },
+    {
+      "epoch": 5.556390977443609,
+      "grad_norm": 0.20643711350658164,
+      "learning_rate": 1.7287303874551515e-06,
+      "loss": 0.3196,
+      "step": 739
+    },
+    {
+      "epoch": 5.56390977443609,
+      "grad_norm": 0.21581908081000192,
+      "learning_rate": 1.7278788103694942e-06,
+      "loss": 0.3108,
+      "step": 740
+    },
+    {
+      "epoch": 5.571428571428571,
+      "grad_norm": 0.24093453604934334,
+      "learning_rate": 1.7270261091846673e-06,
+      "loss": 0.3191,
+      "step": 741
+    },
+    {
+      "epoch": 5.578947368421053,
+      "grad_norm": 0.21533945160977155,
+      "learning_rate": 1.7261722852175393e-06,
+      "loss": 0.3084,
+      "step": 742
+    },
+    {
+      "epoch": 5.586466165413534,
+      "grad_norm": 0.20410182943989716,
+      "learning_rate": 1.7253173397867133e-06,
+      "loss": 0.3172,
+      "step": 743
+    },
+    {
+      "epoch": 5.593984962406015,
+      "grad_norm": 0.2235292780585372,
+      "learning_rate": 1.7244612742125236e-06,
+      "loss": 0.3244,
+      "step": 744
+    },
+    {
+      "epoch": 5.601503759398496,
+      "grad_norm": 0.21960530681392318,
+      "learning_rate": 1.723604089817034e-06,
+      "loss": 0.3186,
+      "step": 745
+    },
+    {
+      "epoch": 5.609022556390977,
+      "grad_norm": 0.21075267529469283,
+      "learning_rate": 1.7227457879240371e-06,
+      "loss": 0.3105,
+      "step": 746
+    },
+    {
+      "epoch": 5.616541353383458,
+      "grad_norm": 0.22436076476447328,
+      "learning_rate": 1.7218863698590508e-06,
+      "loss": 0.3152,
+      "step": 747
+    },
+    {
+      "epoch": 5.62406015037594,
+      "grad_norm": 0.23377197959904097,
+      "learning_rate": 1.7210258369493169e-06,
+      "loss": 0.3237,
+      "step": 748
+    },
+    {
+      "epoch": 5.631578947368421,
+      "grad_norm": 0.2075473972794605,
+      "learning_rate": 1.7201641905237984e-06,
+      "loss": 0.318,
+      "step": 749
+    },
+    {
+      "epoch": 5.639097744360902,
+      "grad_norm": 0.21460838341912697,
+      "learning_rate": 1.719301431913179e-06,
+      "loss": 0.315,
+      "step": 750
+    },
+    {
+      "epoch": 5.646616541353383,
+      "grad_norm": 0.22867562696593072,
+      "learning_rate": 1.718437562449859e-06,
+      "loss": 0.3154,
+      "step": 751
+    },
+    {
+      "epoch": 5.654135338345864,
+      "grad_norm": 0.21990982161149428,
+      "learning_rate": 1.7175725834679548e-06,
+      "loss": 0.3053,
+      "step": 752
+    },
+    {
+      "epoch": 5.661654135338345,
+      "grad_norm": 0.2205798691425424,
+      "learning_rate": 1.7167064963032963e-06,
+      "loss": 0.3241,
+      "step": 753
+    },
+    {
+      "epoch": 5.669172932330827,
+      "grad_norm": 0.20428823359694795,
+      "learning_rate": 1.7158393022934243e-06,
+      "loss": 0.3107,
+      "step": 754
+    },
+    {
+      "epoch": 5.676691729323308,
+      "grad_norm": 0.21156317248740222,
+      "learning_rate": 1.7149710027775895e-06,
+      "loss": 0.3197,
+      "step": 755
+    },
+    {
+      "epoch": 5.684210526315789,
+      "grad_norm": 0.212126647300641,
+      "learning_rate": 1.7141015990967498e-06,
+      "loss": 0.3212,
+      "step": 756
+    },
+    {
+      "epoch": 5.69172932330827,
+      "grad_norm": 0.20689053658797066,
+      "learning_rate": 1.7132310925935677e-06,
+      "loss": 0.3137,
+      "step": 757
+    },
+    {
+      "epoch": 5.6992481203007515,
+      "grad_norm": 0.2068412816055437,
+      "learning_rate": 1.71235948461241e-06,
+      "loss": 0.3164,
+      "step": 758
+    },
+    {
+      "epoch": 5.706766917293233,
+      "grad_norm": 0.21320368767433007,
+      "learning_rate": 1.7114867764993436e-06,
+      "loss": 0.315,
+      "step": 759
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.2036246579929244,
+      "learning_rate": 1.7106129696021349e-06,
+      "loss": 0.3084,
+      "step": 760
+    },
+    {
+      "epoch": 5.7218045112781954,
+      "grad_norm": 0.22034755482082968,
+      "learning_rate": 1.7097380652702467e-06,
+      "loss": 0.3145,
+      "step": 761
+    },
+    {
+      "epoch": 5.7293233082706765,
+      "grad_norm": 0.20358984799933827,
+      "learning_rate": 1.7088620648548374e-06,
+      "loss": 0.3084,
+      "step": 762
+    },
+    {
+      "epoch": 5.7368421052631575,
+      "grad_norm": 0.2240601427244857,
+      "learning_rate": 1.707984969708757e-06,
+      "loss": 0.3352,
+      "step": 763
+    },
+    {
+      "epoch": 5.7443609022556394,
+      "grad_norm": 0.21960494749123055,
+      "learning_rate": 1.7071067811865474e-06,
+      "loss": 0.3189,
+      "step": 764
+    },
+    {
+      "epoch": 5.7518796992481205,
+      "grad_norm": 0.2286194626642664,
+      "learning_rate": 1.7062275006444384e-06,
+      "loss": 0.3391,
+      "step": 765
+    },
+    {
+      "epoch": 5.7593984962406015,
+      "grad_norm": 0.2161867828726593,
+      "learning_rate": 1.7053471294403461e-06,
+      "loss": 0.3101,
+      "step": 766
+    },
+    {
+      "epoch": 5.7669172932330826,
+      "grad_norm": 0.21967492181433476,
+      "learning_rate": 1.7044656689338713e-06,
+      "loss": 0.3101,
+      "step": 767
+    },
+    {
+      "epoch": 5.774436090225564,
+      "grad_norm": 0.21534070270516892,
+      "learning_rate": 1.703583120486297e-06,
+      "loss": 0.313,
+      "step": 768
+    },
+    {
+      "epoch": 5.7819548872180455,
+      "grad_norm": 0.20279650624521833,
+      "learning_rate": 1.7026994854605862e-06,
+      "loss": 0.3023,
+      "step": 769
+    },
+    {
+      "epoch": 5.7894736842105265,
+      "grad_norm": 0.21001541065607962,
+      "learning_rate": 1.7018147652213804e-06,
+      "loss": 0.3299,
+      "step": 770
+    },
+    {
+      "epoch": 5.796992481203008,
+      "grad_norm": 0.22732907674378983,
+      "learning_rate": 1.7009289611349963e-06,
+      "loss": 0.3291,
+      "step": 771
+    },
+    {
+      "epoch": 5.804511278195489,
+      "grad_norm": 0.23806438322228457,
+      "learning_rate": 1.7000420745694253e-06,
+      "loss": 0.3291,
+      "step": 772
+    },
+    {
+      "epoch": 5.81203007518797,
+      "grad_norm": 0.21475714964419018,
+      "learning_rate": 1.6991541068943297e-06,
+      "loss": 0.3223,
+      "step": 773
+    },
+    {
+      "epoch": 5.819548872180452,
+      "grad_norm": 0.213634715524603,
+      "learning_rate": 1.6982650594810422e-06,
+      "loss": 0.3194,
+      "step": 774
+    },
+    {
+      "epoch": 5.827067669172933,
+      "grad_norm": 0.21171898480266835,
+      "learning_rate": 1.6973749337025622e-06,
+      "loss": 0.3082,
+      "step": 775
+    },
+    {
+      "epoch": 5.834586466165414,
+      "grad_norm": 0.22171352541024097,
+      "learning_rate": 1.696483730933555e-06,
+      "loss": 0.3116,
+      "step": 776
+    },
+    {
+      "epoch": 5.842105263157895,
+      "grad_norm": 0.2248230594099052,
+      "learning_rate": 1.695591452550349e-06,
+      "loss": 0.315,
+      "step": 777
+    },
+    {
+      "epoch": 5.849624060150376,
+      "grad_norm": 0.2003611399062983,
+      "learning_rate": 1.6946980999309341e-06,
+      "loss": 0.3235,
+      "step": 778
+    },
+    {
+      "epoch": 5.857142857142857,
+      "grad_norm": 0.20269497019328855,
+      "learning_rate": 1.6938036744549585e-06,
+      "loss": 0.3167,
+      "step": 779
+    },
+    {
+      "epoch": 5.864661654135339,
+      "grad_norm": 0.21530240089678346,
+      "learning_rate": 1.6929081775037276e-06,
+      "loss": 0.3212,
+      "step": 780
+    },
+    {
+      "epoch": 5.87218045112782,
+      "grad_norm": 0.22005514169597593,
+      "learning_rate": 1.6920116104602013e-06,
+      "loss": 0.3146,
+      "step": 781
+    },
+    {
+      "epoch": 5.879699248120301,
+      "grad_norm": 0.21657610192960647,
+      "learning_rate": 1.6911139747089931e-06,
+      "loss": 0.3209,
+      "step": 782
+    },
+    {
+      "epoch": 5.887218045112782,
+      "grad_norm": 0.21497629485640513,
+      "learning_rate": 1.6902152716363654e-06,
+      "loss": 0.3148,
+      "step": 783
+    },
+    {
+      "epoch": 5.894736842105263,
+      "grad_norm": 0.21196849546339327,
+      "learning_rate": 1.68931550263023e-06,
+      "loss": 0.314,
+      "step": 784
+    },
+    {
+      "epoch": 5.902255639097744,
+      "grad_norm": 0.1989231112618944,
+      "learning_rate": 1.688414669080145e-06,
+      "loss": 0.309,
+      "step": 785
+    },
+    {
+      "epoch": 5.909774436090226,
+      "grad_norm": 0.2112430398244574,
+      "learning_rate": 1.6875127723773114e-06,
+      "loss": 0.3091,
+      "step": 786
+    },
+    {
+      "epoch": 5.917293233082707,
+      "grad_norm": 0.2141991235029919,
+      "learning_rate": 1.6866098139145728e-06,
+      "loss": 0.2966,
+      "step": 787
+    },
+    {
+      "epoch": 5.924812030075188,
+      "grad_norm": 0.21041641608190922,
+      "learning_rate": 1.685705795086413e-06,
+      "loss": 0.3178,
+      "step": 788
+    },
+    {
+      "epoch": 5.932330827067669,
+      "grad_norm": 0.199849696583008,
+      "learning_rate": 1.684800717288953e-06,
+      "loss": 0.3031,
+      "step": 789
+    },
+    {
+      "epoch": 5.93984962406015,
+      "grad_norm": 0.19966008856088602,
+      "learning_rate": 1.6838945819199485e-06,
+      "loss": 0.3075,
+      "step": 790
+    },
+    {
+      "epoch": 5.947368421052632,
+      "grad_norm": 0.22254285639071142,
+      "learning_rate": 1.6829873903787898e-06,
+      "loss": 0.3166,
+      "step": 791
+    },
+    {
+      "epoch": 5.954887218045113,
+      "grad_norm": 0.2122052340512222,
+      "learning_rate": 1.6820791440664969e-06,
+      "loss": 0.3292,
+      "step": 792
+    },
+    {
+      "epoch": 5.962406015037594,
+      "grad_norm": 0.2122333668234056,
+      "learning_rate": 1.6811698443857197e-06,
+      "loss": 0.3223,
+      "step": 793
+    },
+    {
+      "epoch": 5.969924812030075,
+      "grad_norm": 0.22090470127846504,
+      "learning_rate": 1.6802594927407344e-06,
+      "loss": 0.3317,
+      "step": 794
+    },
+    {
+      "epoch": 5.977443609022556,
+      "grad_norm": 0.21208291055821393,
+      "learning_rate": 1.679348090537442e-06,
+      "loss": 0.3217,
+      "step": 795
+    },
+    {
+      "epoch": 5.984962406015038,
+      "grad_norm": 0.20594612733097703,
+      "learning_rate": 1.6784356391833662e-06,
+      "loss": 0.3095,
+      "step": 796
+    },
+    {
+      "epoch": 5.992481203007519,
+      "grad_norm": 0.2080846333623414,
+      "learning_rate": 1.6775221400876504e-06,
+      "loss": 0.3136,
+      "step": 797
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.2070820383853516,
+      "learning_rate": 1.6766075946610565e-06,
+      "loss": 0.3126,
+      "step": 798
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 0.40656396746635437,
+      "eval_runtime": 1.7292,
+      "eval_samples_per_second": 189.102,
+      "eval_steps_per_second": 3.47,
+      "step": 798
+    },
+    {
+      "epoch": 6.007518796992481,
+      "grad_norm": 0.2153973994928368,
+      "learning_rate": 1.675692004315962e-06,
+      "loss": 0.312,
+      "step": 799
+    },
+    {
+      "epoch": 6.015037593984962,
+      "grad_norm": 0.22288340019399558,
+      "learning_rate": 1.6747753704663584e-06,
+      "loss": 0.3091,
+      "step": 800
+    },
+    {
+      "epoch": 6.022556390977444,
+      "grad_norm": 0.23838837159345758,
+      "learning_rate": 1.6738576945278485e-06,
+      "loss": 0.3152,
+      "step": 801
+    },
+    {
+      "epoch": 6.030075187969925,
+      "grad_norm": 0.2417974026100096,
+      "learning_rate": 1.6729389779176443e-06,
+      "loss": 0.3199,
+      "step": 802
+    },
+    {
+      "epoch": 6.037593984962406,
+      "grad_norm": 0.20528233624304118,
+      "learning_rate": 1.6720192220545658e-06,
+      "loss": 0.3112,
+      "step": 803
+    },
+    {
+      "epoch": 6.045112781954887,
+      "grad_norm": 0.20960925595209362,
+      "learning_rate": 1.6710984283590367e-06,
+      "loss": 0.3177,
+      "step": 804
+    },
+    {
+      "epoch": 6.052631578947368,
+      "grad_norm": 0.20993638247076332,
+      "learning_rate": 1.6701765982530845e-06,
+      "loss": 0.3068,
+      "step": 805
+    },
+    {
+      "epoch": 6.06015037593985,
+      "grad_norm": 0.21142156129301282,
+      "learning_rate": 1.6692537331603372e-06,
+      "loss": 0.2964,
+      "step": 806
+    },
+    {
+      "epoch": 6.067669172932331,
+      "grad_norm": 0.22257208850413235,
+      "learning_rate": 1.6683298345060202e-06,
+      "loss": 0.3225,
+      "step": 807
+    },
+    {
+      "epoch": 6.075187969924812,
+      "grad_norm": 0.21190472300432445,
+      "learning_rate": 1.6674049037169562e-06,
+      "loss": 0.3076,
+      "step": 808
+    },
+    {
+      "epoch": 6.082706766917293,
+      "grad_norm": 0.22284313113891321,
+      "learning_rate": 1.6664789422215615e-06,
+      "loss": 0.3159,
+      "step": 809
+    },
+    {
+      "epoch": 6.090225563909774,
+      "grad_norm": 0.21186499890863486,
+      "learning_rate": 1.665551951449844e-06,
+      "loss": 0.3254,
+      "step": 810
+    },
+    {
+      "epoch": 6.097744360902255,
+      "grad_norm": 0.212673983270181,
+      "learning_rate": 1.6646239328334018e-06,
+      "loss": 0.2934,
+      "step": 811
+    },
+    {
+      "epoch": 6.105263157894737,
+      "grad_norm": 0.2198952701667914,
+      "learning_rate": 1.6636948878054189e-06,
+      "loss": 0.3087,
+      "step": 812
+    },
+    {
+      "epoch": 6.112781954887218,
+      "grad_norm": 0.20390537375836518,
+      "learning_rate": 1.662764817800666e-06,
+      "loss": 0.3096,
+      "step": 813
+    },
+    {
+      "epoch": 6.120300751879699,
+      "grad_norm": 0.20023688996668293,
+      "learning_rate": 1.6618337242554961e-06,
+      "loss": 0.3001,
+      "step": 814
+    },
+    {
+      "epoch": 6.12781954887218,
+      "grad_norm": 0.21826483749407138,
+      "learning_rate": 1.660901608607843e-06,
+      "loss": 0.3202,
+      "step": 815
+    },
+    {
+      "epoch": 6.135338345864661,
+      "grad_norm": 0.21088380823636216,
+      "learning_rate": 1.6599684722972187e-06,
+      "loss": 0.3132,
+      "step": 816
+    },
+    {
+      "epoch": 6.142857142857143,
+      "grad_norm": 0.1941466999158382,
+      "learning_rate": 1.6590343167647114e-06,
+      "loss": 0.3052,
+      "step": 817
+    },
+    {
+      "epoch": 6.150375939849624,
+      "grad_norm": 0.21991689721772353,
+      "learning_rate": 1.6580991434529841e-06,
+      "loss": 0.3141,
+      "step": 818
+    },
+    {
+      "epoch": 6.157894736842105,
+      "grad_norm": 0.21753222297018376,
+      "learning_rate": 1.6571629538062707e-06,
+      "loss": 0.3342,
+      "step": 819
+    },
+    {
+      "epoch": 6.165413533834586,
+      "grad_norm": 0.20405868527298704,
+      "learning_rate": 1.6562257492703755e-06,
+      "loss": 0.3042,
+      "step": 820
+    },
+    {
+      "epoch": 6.172932330827067,
+      "grad_norm": 0.1988275766681763,
+      "learning_rate": 1.6552875312926692e-06,
+      "loss": 0.3036,
+      "step": 821
+    },
+    {
+      "epoch": 6.180451127819548,
+      "grad_norm": 0.21180676365840176,
+      "learning_rate": 1.6543483013220887e-06,
+      "loss": 0.3148,
+      "step": 822
+    },
+    {
+      "epoch": 6.18796992481203,
+      "grad_norm": 0.21301918741056566,
+      "learning_rate": 1.653408060809133e-06,
+      "loss": 0.317,
+      "step": 823
+    },
+    {
+      "epoch": 6.195488721804511,
+      "grad_norm": 0.22671722466573713,
+      "learning_rate": 1.6524668112058615e-06,
+      "loss": 0.3194,
+      "step": 824
+    },
+    {
+      "epoch": 6.203007518796992,
+      "grad_norm": 0.19964905626627832,
+      "learning_rate": 1.6515245539658929e-06,
+      "loss": 0.3068,
+      "step": 825
+    },
+    {
+      "epoch": 6.2105263157894735,
+      "grad_norm": 0.2075477884877615,
+      "learning_rate": 1.6505812905444012e-06,
+      "loss": 0.2979,
+      "step": 826
+    },
+    {
+      "epoch": 6.2180451127819545,
+      "grad_norm": 0.22647376299345656,
+      "learning_rate": 1.649637022398115e-06,
+      "loss": 0.3166,
+      "step": 827
+    },
+    {
+      "epoch": 6.225563909774436,
+      "grad_norm": 0.22201761657677216,
+      "learning_rate": 1.6486917509853137e-06,
+      "loss": 0.3075,
+      "step": 828
+    },
+    {
+      "epoch": 6.2330827067669174,
+      "grad_norm": 0.20673888962838763,
+      "learning_rate": 1.6477454777658273e-06,
+      "loss": 0.3188,
+      "step": 829
+    },
+    {
+      "epoch": 6.2406015037593985,
+      "grad_norm": 0.2060932164690671,
+      "learning_rate": 1.646798204201032e-06,
+      "loss": 0.3061,
+      "step": 830
+    },
+    {
+      "epoch": 6.2481203007518795,
+      "grad_norm": 0.2231255680000611,
+      "learning_rate": 1.6458499317538487e-06,
+      "loss": 0.3033,
+      "step": 831
+    },
+    {
+      "epoch": 6.2556390977443606,
+      "grad_norm": 0.2116354201283869,
+      "learning_rate": 1.6449006618887418e-06,
+      "loss": 0.3099,
+      "step": 832
+    },
+    {
+      "epoch": 6.2631578947368425,
+      "grad_norm": 0.22045159771322886,
+      "learning_rate": 1.6439503960717154e-06,
+      "loss": 0.3179,
+      "step": 833
+    },
+    {
+      "epoch": 6.2706766917293235,
+      "grad_norm": 0.22074981763864907,
+      "learning_rate": 1.642999135770312e-06,
+      "loss": 0.3126,
+      "step": 834
+    },
+    {
+      "epoch": 6.2781954887218046,
+      "grad_norm": 0.22934530854673307,
+      "learning_rate": 1.6420468824536094e-06,
+      "loss": 0.3134,
+      "step": 835
+    },
+    {
+      "epoch": 6.285714285714286,
+      "grad_norm": 0.2074706028084191,
+      "learning_rate": 1.64109363759222e-06,
+      "loss": 0.3027,
+      "step": 836
+    },
+    {
+      "epoch": 6.293233082706767,
+      "grad_norm": 0.22170645036574993,
+      "learning_rate": 1.6401394026582867e-06,
+      "loss": 0.3286,
+      "step": 837
+    },
+    {
+      "epoch": 6.3007518796992485,
+      "grad_norm": 0.20288793239454608,
+      "learning_rate": 1.6391841791254816e-06,
+      "loss": 0.3058,
+      "step": 838
+    },
+    {
+      "epoch": 6.30827067669173,
+      "grad_norm": 0.22559821683563852,
+      "learning_rate": 1.6382279684690033e-06,
+      "loss": 0.3216,
+      "step": 839
+    },
+    {
+      "epoch": 6.315789473684211,
+      "grad_norm": 0.2072350498932216,
+      "learning_rate": 1.6372707721655755e-06,
+      "loss": 0.3055,
+      "step": 840
+    },
+    {
+      "epoch": 6.323308270676692,
+      "grad_norm": 0.21385809431537767,
+      "learning_rate": 1.6363125916934434e-06,
+      "loss": 0.3092,
+      "step": 841
+    },
+    {
+      "epoch": 6.330827067669173,
+      "grad_norm": 0.2095515195159735,
+      "learning_rate": 1.6353534285323722e-06,
+      "loss": 0.308,
+      "step": 842
+    },
+    {
+      "epoch": 6.338345864661654,
+      "grad_norm": 0.20712031695211866,
+      "learning_rate": 1.6343932841636455e-06,
+      "loss": 0.3153,
+      "step": 843
+    },
+    {
+      "epoch": 6.345864661654136,
+      "grad_norm": 0.2274316488691548,
+      "learning_rate": 1.6334321600700611e-06,
+      "loss": 0.3051,
+      "step": 844
+    },
+    {
+      "epoch": 6.353383458646617,
+      "grad_norm": 0.21628420123235265,
+      "learning_rate": 1.6324700577359308e-06,
+      "loss": 0.2933,
+      "step": 845
+    },
+    {
+      "epoch": 6.360902255639098,
+      "grad_norm": 0.23433463434930477,
+      "learning_rate": 1.6315069786470765e-06,
+      "loss": 0.3056,
+      "step": 846
+    },
+    {
+      "epoch": 6.368421052631579,
+      "grad_norm": 0.21193466819658774,
+      "learning_rate": 1.6305429242908287e-06,
+      "loss": 0.2948,
+      "step": 847
+    },
+    {
+      "epoch": 6.37593984962406,
+      "grad_norm": 0.2120033206450346,
+      "learning_rate": 1.629577896156024e-06,
+      "loss": 0.3218,
+      "step": 848
+    },
+    {
+      "epoch": 6.383458646616542,
+      "grad_norm": 0.20445572936519565,
+      "learning_rate": 1.6286118957330035e-06,
+      "loss": 0.3095,
+      "step": 849
+    },
+    {
+      "epoch": 6.390977443609023,
+      "grad_norm": 0.2173319899252014,
+      "learning_rate": 1.6276449245136088e-06,
+      "loss": 0.3141,
+      "step": 850
+    },
+    {
+      "epoch": 6.398496240601504,
+      "grad_norm": 0.22071869905376287,
+      "learning_rate": 1.6266769839911815e-06,
+      "loss": 0.3014,
+      "step": 851
+    },
+    {
+      "epoch": 6.406015037593985,
+      "grad_norm": 0.1980967394942533,
+      "learning_rate": 1.6257080756605598e-06,
+      "loss": 0.3147,
+      "step": 852
+    },
+    {
+      "epoch": 6.413533834586466,
+      "grad_norm": 0.20927678532538777,
+      "learning_rate": 1.624738201018077e-06,
+      "loss": 0.3031,
+      "step": 853
+    },
+    {
+      "epoch": 6.421052631578947,
+      "grad_norm": 0.21796133575161092,
+      "learning_rate": 1.623767361561558e-06,
+      "loss": 0.3131,
+      "step": 854
+    },
+    {
+      "epoch": 6.428571428571429,
+      "grad_norm": 0.22992866579855728,
+      "learning_rate": 1.6227955587903179e-06,
+      "loss": 0.3148,
+      "step": 855
+    },
+    {
+      "epoch": 6.43609022556391,
+      "grad_norm": 0.2517835370747452,
+      "learning_rate": 1.6218227942051602e-06,
+      "loss": 0.3136,
+      "step": 856
+    },
+    {
+      "epoch": 6.443609022556391,
+      "grad_norm": 0.22390903888199015,
+      "learning_rate": 1.6208490693083734e-06,
+      "loss": 0.3104,
+      "step": 857
+    },
+    {
+      "epoch": 6.451127819548872,
+      "grad_norm": 0.21931774851195146,
+      "learning_rate": 1.6198743856037283e-06,
+      "loss": 0.3081,
+      "step": 858
+    },
+    {
+      "epoch": 6.458646616541353,
+      "grad_norm": 0.22923269990052064,
+      "learning_rate": 1.618898744596477e-06,
+      "loss": 0.3192,
+      "step": 859
+    },
+    {
+      "epoch": 6.466165413533835,
+      "grad_norm": 0.22135433661911214,
+      "learning_rate": 1.6179221477933507e-06,
+      "loss": 0.3161,
+      "step": 860
+    },
+    {
+      "epoch": 6.473684210526316,
+      "grad_norm": 0.21994370489358217,
+      "learning_rate": 1.6169445967025555e-06,
+      "loss": 0.3094,
+      "step": 861
+    },
+    {
+      "epoch": 6.481203007518797,
+      "grad_norm": 0.21381513896050433,
+      "learning_rate": 1.6159660928337721e-06,
+      "loss": 0.3112,
+      "step": 862
+    },
+    {
+      "epoch": 6.488721804511278,
+      "grad_norm": 0.22581238246645574,
+      "learning_rate": 1.6149866376981524e-06,
+      "loss": 0.3101,
+      "step": 863
+    },
+    {
+      "epoch": 6.496240601503759,
+      "grad_norm": 0.21662830474997113,
+      "learning_rate": 1.6140062328083168e-06,
+      "loss": 0.3092,
+      "step": 864
+    },
+    {
+      "epoch": 6.503759398496241,
+      "grad_norm": 0.21095947699934636,
+      "learning_rate": 1.6130248796783535e-06,
+      "loss": 0.3038,
+      "step": 865
+    },
+    {
+      "epoch": 6.511278195488722,
+      "grad_norm": 0.20668761652929815,
+      "learning_rate": 1.6120425798238143e-06,
+      "loss": 0.3308,
+      "step": 866
+    },
+    {
+      "epoch": 6.518796992481203,
+      "grad_norm": 0.2179807978678519,
+      "learning_rate": 1.6110593347617132e-06,
+      "loss": 0.3114,
+      "step": 867
+    },
+    {
+      "epoch": 6.526315789473684,
+      "grad_norm": 0.213255294390063,
+      "learning_rate": 1.6100751460105243e-06,
+      "loss": 0.3023,
+      "step": 868
+    },
+    {
+      "epoch": 6.533834586466165,
+      "grad_norm": 0.24915605849507963,
+      "learning_rate": 1.609090015090179e-06,
+      "loss": 0.3099,
+      "step": 869
+    },
+    {
+      "epoch": 6.541353383458647,
+      "grad_norm": 0.22737864730336083,
+      "learning_rate": 1.6081039435220634e-06,
+      "loss": 0.3083,
+      "step": 870
+    },
+    {
+      "epoch": 6.548872180451128,
+      "grad_norm": 0.21649243611999594,
+      "learning_rate": 1.6071169328290162e-06,
+      "loss": 0.3088,
+      "step": 871
+    },
+    {
+      "epoch": 6.556390977443609,
+      "grad_norm": 0.20821151923173215,
+      "learning_rate": 1.6061289845353274e-06,
+      "loss": 0.307,
+      "step": 872
+    },
+    {
+      "epoch": 6.56390977443609,
+      "grad_norm": 0.19606242750533562,
+      "learning_rate": 1.6051401001667336e-06,
+      "loss": 0.2984,
+      "step": 873
+    },
+    {
+      "epoch": 6.571428571428571,
+      "grad_norm": 0.2009099358483978,
+      "learning_rate": 1.6041502812504185e-06,
+      "loss": 0.3208,
+      "step": 874
+    },
+    {
+      "epoch": 6.578947368421053,
+      "grad_norm": 0.2219569941974826,
+      "learning_rate": 1.6031595293150075e-06,
+      "loss": 0.3189,
+      "step": 875
+    },
+    {
+      "epoch": 6.586466165413534,
+      "grad_norm": 0.2088412792166697,
+      "learning_rate": 1.6021678458905683e-06,
+      "loss": 0.3064,
+      "step": 876
+    },
+    {
+      "epoch": 6.593984962406015,
+      "grad_norm": 0.20149333179686743,
+      "learning_rate": 1.6011752325086064e-06,
+      "loss": 0.3129,
+      "step": 877
+    },
+    {
+      "epoch": 6.601503759398496,
+      "grad_norm": 0.23673557624356328,
+      "learning_rate": 1.6001816907020633e-06,
+      "loss": 0.3104,
+      "step": 878
+    },
+    {
+      "epoch": 6.609022556390977,
+      "grad_norm": 0.21053575279967765,
+      "learning_rate": 1.599187222005315e-06,
+      "loss": 0.3008,
+      "step": 879
+    },
+    {
+      "epoch": 6.616541353383458,
+      "grad_norm": 0.21167403031420431,
+      "learning_rate": 1.5981918279541685e-06,
+      "loss": 0.3134,
+      "step": 880
+    },
+    {
+      "epoch": 6.62406015037594,
+      "grad_norm": 0.2301898684309179,
+      "learning_rate": 1.5971955100858603e-06,
+      "loss": 0.3195,
+      "step": 881
+    },
+    {
+      "epoch": 6.631578947368421,
+      "grad_norm": 0.2237369014347961,
+      "learning_rate": 1.5961982699390525e-06,
+      "loss": 0.317,
+      "step": 882
+    },
+    {
+      "epoch": 6.639097744360902,
+      "grad_norm": 0.21961371633470778,
+      "learning_rate": 1.5952001090538332e-06,
+      "loss": 0.3108,
+      "step": 883
+    },
+    {
+      "epoch": 6.646616541353383,
+      "grad_norm": 0.20956718110170902,
+      "learning_rate": 1.5942010289717105e-06,
+      "loss": 0.3253,
+      "step": 884
+    },
+    {
+      "epoch": 6.654135338345864,
+      "grad_norm": 0.21027877835334544,
+      "learning_rate": 1.5932010312356137e-06,
+      "loss": 0.3137,
+      "step": 885
+    },
+    {
+      "epoch": 6.661654135338345,
+      "grad_norm": 0.22110953719268514,
+      "learning_rate": 1.5922001173898887e-06,
+      "loss": 0.3131,
+      "step": 886
+    },
+    {
+      "epoch": 6.669172932330827,
+      "grad_norm": 0.21208757471141795,
+      "learning_rate": 1.591198288980296e-06,
+      "loss": 0.3124,
+      "step": 887
+    },
+    {
+      "epoch": 6.676691729323308,
+      "grad_norm": 0.22405054886894052,
+      "learning_rate": 1.5901955475540083e-06,
+      "loss": 0.3225,
+      "step": 888
+    },
+    {
+      "epoch": 6.684210526315789,
+      "grad_norm": 0.19500287410341544,
+      "learning_rate": 1.5891918946596095e-06,
+      "loss": 0.3046,
+      "step": 889
+    },
+    {
+      "epoch": 6.69172932330827,
+      "grad_norm": 0.22310418667918672,
+      "learning_rate": 1.5881873318470893e-06,
+      "loss": 0.3137,
+      "step": 890
+    },
+    {
+      "epoch": 6.6992481203007515,
+      "grad_norm": 0.2100360542627228,
+      "learning_rate": 1.5871818606678447e-06,
+      "loss": 0.3014,
+      "step": 891
+    },
+    {
+      "epoch": 6.706766917293233,
+      "grad_norm": 0.20123545592426573,
+      "learning_rate": 1.5861754826746733e-06,
+      "loss": 0.3051,
+      "step": 892
+    },
+    {
+      "epoch": 6.714285714285714,
+      "grad_norm": 0.23432082269300825,
+      "learning_rate": 1.5851681994217754e-06,
+      "loss": 0.3114,
+      "step": 893
+    },
+    {
+      "epoch": 6.7218045112781954,
+      "grad_norm": 0.21045626227678593,
+      "learning_rate": 1.5841600124647477e-06,
+      "loss": 0.3043,
+      "step": 894
+    },
+    {
+      "epoch": 6.7293233082706765,
+      "grad_norm": 0.2148028353809671,
+      "learning_rate": 1.5831509233605829e-06,
+      "loss": 0.2975,
+      "step": 895
+    },
+    {
+      "epoch": 6.7368421052631575,
+      "grad_norm": 0.23372582029586272,
+      "learning_rate": 1.5821409336676674e-06,
+      "loss": 0.307,
+      "step": 896
+    },
+    {
+      "epoch": 6.7443609022556394,
+      "grad_norm": 0.2109509190201137,
+      "learning_rate": 1.581130044945778e-06,
+      "loss": 0.3073,
+      "step": 897
+    },
+    {
+      "epoch": 6.7518796992481205,
+      "grad_norm": 0.20677044532204628,
+      "learning_rate": 1.5801182587560803e-06,
+      "loss": 0.3082,
+      "step": 898
+    },
+    {
+      "epoch": 6.7593984962406015,
+      "grad_norm": 0.2118502832702876,
+      "learning_rate": 1.5791055766611255e-06,
+      "loss": 0.2946,
+      "step": 899
+    },
+    {
+      "epoch": 6.7669172932330826,
+      "grad_norm": 0.20772859233133106,
+      "learning_rate": 1.5780920002248483e-06,
+      "loss": 0.2992,
+      "step": 900
+    },
+    {
+      "epoch": 6.774436090225564,
+      "grad_norm": 0.2050288728226551,
+      "learning_rate": 1.5770775310125651e-06,
+      "loss": 0.3055,
+      "step": 901
+    },
+    {
+      "epoch": 6.7819548872180455,
+      "grad_norm": 0.221304146064523,
+      "learning_rate": 1.5760621705909705e-06,
+      "loss": 0.3053,
+      "step": 902
+    },
+    {
+      "epoch": 6.7894736842105265,
+      "grad_norm": 0.21211921918947635,
+      "learning_rate": 1.5750459205281361e-06,
+      "loss": 0.3009,
+      "step": 903
+    },
+    {
+      "epoch": 6.796992481203008,
+      "grad_norm": 0.2202801866110241,
+      "learning_rate": 1.5740287823935066e-06,
+      "loss": 0.3084,
+      "step": 904
+    },
+    {
+      "epoch": 6.804511278195489,
+      "grad_norm": 0.20422892839272033,
+      "learning_rate": 1.573010757757899e-06,
+      "loss": 0.3108,
+      "step": 905
+    },
+    {
+      "epoch": 6.81203007518797,
+      "grad_norm": 0.23655870600950782,
+      "learning_rate": 1.5719918481934986e-06,
+      "loss": 0.3139,
+      "step": 906
+    },
+    {
+      "epoch": 6.819548872180452,
+      "grad_norm": 0.22882164719545064,
+      "learning_rate": 1.570972055273858e-06,
+      "loss": 0.3037,
+      "step": 907
+    },
+    {
+      "epoch": 6.827067669172933,
+      "grad_norm": 0.21304218430615807,
+      "learning_rate": 1.569951380573894e-06,
+      "loss": 0.3192,
+      "step": 908
+    },
+    {
+      "epoch": 6.834586466165414,
+      "grad_norm": 0.22450218123998616,
+      "learning_rate": 1.5689298256698845e-06,
+      "loss": 0.3054,
+      "step": 909
+    },
+    {
+      "epoch": 6.842105263157895,
+      "grad_norm": 0.20690311330716382,
+      "learning_rate": 1.567907392139467e-06,
+      "loss": 0.3001,
+      "step": 910
+    },
+    {
+      "epoch": 6.849624060150376,
+      "grad_norm": 0.21072713948268548,
+      "learning_rate": 1.5668840815616364e-06,
+      "loss": 0.2985,
+      "step": 911
+    },
+    {
+      "epoch": 6.857142857142857,
+      "grad_norm": 0.21501946132368438,
+      "learning_rate": 1.5658598955167418e-06,
+      "loss": 0.3128,
+      "step": 912
+    },
+    {
+      "epoch": 6.864661654135339,
+      "grad_norm": 0.22015224709420586,
+      "learning_rate": 1.5648348355864838e-06,
+      "loss": 0.3182,
+      "step": 913
+    },
+    {
+      "epoch": 6.87218045112782,
+      "grad_norm": 0.1968315810641621,
+      "learning_rate": 1.5638089033539132e-06,
+      "loss": 0.2817,
+      "step": 914
+    },
+    {
+      "epoch": 6.879699248120301,
+      "grad_norm": 0.20823462877713128,
+      "learning_rate": 1.562782100403428e-06,
+      "loss": 0.3235,
+      "step": 915
+    },
+    {
+      "epoch": 6.887218045112782,
+      "grad_norm": 0.21180982434706477,
+      "learning_rate": 1.5617544283207708e-06,
+      "loss": 0.3161,
+      "step": 916
+    },
+    {
+      "epoch": 6.894736842105263,
+      "grad_norm": 0.21695466709882028,
+      "learning_rate": 1.5607258886930259e-06,
+      "loss": 0.3146,
+      "step": 917
+    },
+    {
+      "epoch": 6.902255639097744,
+      "grad_norm": 0.2275417645060086,
+      "learning_rate": 1.5596964831086181e-06,
+      "loss": 0.3091,
+      "step": 918
+    },
+    {
+      "epoch": 6.909774436090226,
+      "grad_norm": 0.21555439103843685,
+      "learning_rate": 1.5586662131573092e-06,
+      "loss": 0.3083,
+      "step": 919
+    },
+    {
+      "epoch": 6.917293233082707,
+      "grad_norm": 0.2302673166792445,
+      "learning_rate": 1.5576350804301957e-06,
+      "loss": 0.3041,
+      "step": 920
+    },
+    {
+      "epoch": 6.924812030075188,
+      "grad_norm": 0.2091803721558529,
+      "learning_rate": 1.556603086519707e-06,
+      "loss": 0.3163,
+      "step": 921
+    },
+    {
+      "epoch": 6.932330827067669,
+      "grad_norm": 0.23412905015056917,
+      "learning_rate": 1.5555702330196021e-06,
+      "loss": 0.3025,
+      "step": 922
+    },
+    {
+      "epoch": 6.93984962406015,
+      "grad_norm": 0.2169400873747117,
+      "learning_rate": 1.5545365215249676e-06,
+      "loss": 0.3085,
+      "step": 923
+    },
+    {
+      "epoch": 6.947368421052632,
+      "grad_norm": 0.23022064484925123,
+      "learning_rate": 1.5535019536322157e-06,
+      "loss": 0.3152,
+      "step": 924
+    },
+    {
+      "epoch": 6.954887218045113,
+      "grad_norm": 0.21297341956678353,
+      "learning_rate": 1.5524665309390801e-06,
+      "loss": 0.3126,
+      "step": 925
+    },
+    {
+      "epoch": 6.962406015037594,
+      "grad_norm": 0.23486269986837136,
+      "learning_rate": 1.551430255044615e-06,
+      "loss": 0.3153,
+      "step": 926
+    },
+    {
+      "epoch": 6.969924812030075,
+      "grad_norm": 0.23577954570491066,
+      "learning_rate": 1.5503931275491928e-06,
+      "loss": 0.3058,
+      "step": 927
+    },
+    {
+      "epoch": 6.977443609022556,
+      "grad_norm": 0.20990260919483159,
+      "learning_rate": 1.5493551500545005e-06,
+      "loss": 0.3131,
+      "step": 928
+    },
+    {
+      "epoch": 6.984962406015038,
+      "grad_norm": 0.2123860736204426,
+      "learning_rate": 1.5483163241635383e-06,
+      "loss": 0.2999,
+      "step": 929
+    },
+    {
+      "epoch": 6.992481203007519,
+      "grad_norm": 0.20188552031358695,
+      "learning_rate": 1.547276651480616e-06,
+      "loss": 0.3005,
+      "step": 930
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.20257044747946332,
+      "learning_rate": 1.5462361336113511e-06,
+      "loss": 0.3093,
+      "step": 931
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 0.4083765149116516,
+      "eval_runtime": 1.7484,
+      "eval_samples_per_second": 187.033,
+      "eval_steps_per_second": 3.432,
+      "step": 931
+    },
+    {
+      "epoch": 7.007518796992481,
+      "grad_norm": 0.22729812118612971,
+      "learning_rate": 1.5451947721626675e-06,
+      "loss": 0.2985,
+      "step": 932
+    },
+    {
+      "epoch": 7.015037593984962,
+      "grad_norm": 0.20322232338420682,
+      "learning_rate": 1.5441525687427906e-06,
+      "loss": 0.3018,
+      "step": 933
+    },
+    {
+      "epoch": 7.022556390977444,
+      "grad_norm": 0.23120403569639872,
+      "learning_rate": 1.5431095249612464e-06,
+      "loss": 0.3111,
+      "step": 934
+    },
+    {
+      "epoch": 7.030075187969925,
+      "grad_norm": 0.23473844154932877,
+      "learning_rate": 1.5420656424288595e-06,
+      "loss": 0.299,
+      "step": 935
+    },
+    {
+      "epoch": 7.037593984962406,
+      "grad_norm": 0.22756638422934444,
+      "learning_rate": 1.5410209227577485e-06,
+      "loss": 0.3059,
+      "step": 936
+    },
+    {
+      "epoch": 7.045112781954887,
+      "grad_norm": 0.22212347093422286,
+      "learning_rate": 1.5399753675613257e-06,
+      "loss": 0.3204,
+      "step": 937
+    },
+    {
+      "epoch": 7.052631578947368,
+      "grad_norm": 0.20577795129103008,
+      "learning_rate": 1.5389289784542943e-06,
+      "loss": 0.3068,
+      "step": 938
+    },
+    {
+      "epoch": 7.06015037593985,
+      "grad_norm": 0.2573372629344865,
+      "learning_rate": 1.5378817570526437e-06,
+      "loss": 0.2959,
+      "step": 939
+    },
+    {
+      "epoch": 7.067669172932331,
+      "grad_norm": 0.22940940786895628,
+      "learning_rate": 1.53683370497365e-06,
+      "loss": 0.3126,
+      "step": 940
+    },
+    {
+      "epoch": 7.075187969924812,
+      "grad_norm": 0.2014593554085678,
+      "learning_rate": 1.5357848238358719e-06,
+      "loss": 0.313,
+      "step": 941
+    },
+    {
+      "epoch": 7.082706766917293,
+      "grad_norm": 0.2140277828264318,
+      "learning_rate": 1.5347351152591484e-06,
+      "loss": 0.3003,
+      "step": 942
+    },
+    {
+      "epoch": 7.090225563909774,
+      "grad_norm": 0.20079794469497084,
+      "learning_rate": 1.5336845808645955e-06,
+      "loss": 0.2994,
+      "step": 943
+    },
+    {
+      "epoch": 7.097744360902255,
+      "grad_norm": 0.24294585604648733,
+      "learning_rate": 1.532633222274606e-06,
+      "loss": 0.2913,
+      "step": 944
+    },
+    {
+      "epoch": 7.105263157894737,
+      "grad_norm": 0.22638100437400335,
+      "learning_rate": 1.5315810411128447e-06,
+      "loss": 0.3031,
+      "step": 945
+    },
+    {
+      "epoch": 7.112781954887218,
+      "grad_norm": 0.2408685934405196,
+      "learning_rate": 1.5305280390042468e-06,
+      "loss": 0.3038,
+      "step": 946
+    },
+    {
+      "epoch": 7.120300751879699,
+      "grad_norm": 0.22625923863408523,
+      "learning_rate": 1.5294742175750156e-06,
+      "loss": 0.3149,
+      "step": 947
+    },
+    {
+      "epoch": 7.12781954887218,
+      "grad_norm": 0.2064450978019771,
+      "learning_rate": 1.5284195784526194e-06,
+      "loss": 0.3069,
+      "step": 948
+    },
+    {
+      "epoch": 7.135338345864661,
+      "grad_norm": 0.20295083395452115,
+      "learning_rate": 1.5273641232657894e-06,
+      "loss": 0.2901,
+      "step": 949
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.23222144675273126,
+      "learning_rate": 1.5263078536445172e-06,
+      "loss": 0.2991,
+      "step": 950
+    },
+    {
+      "epoch": 7.150375939849624,
+      "grad_norm": 0.2240823876034011,
+      "learning_rate": 1.5252507712200525e-06,
+      "loss": 0.3125,
+      "step": 951
+    },
+    {
+      "epoch": 7.157894736842105,
+      "grad_norm": 0.21441600164258307,
+      "learning_rate": 1.524192877624899e-06,
+      "loss": 0.303,
+      "step": 952
+    },
+    {
+      "epoch": 7.165413533834586,
+      "grad_norm": 0.20216779191916676,
+      "learning_rate": 1.523134174492815e-06,
+      "loss": 0.3136,
+      "step": 953
+    },
+    {
+      "epoch": 7.172932330827067,
+      "grad_norm": 0.2176155643065972,
+      "learning_rate": 1.5220746634588074e-06,
+      "loss": 0.3056,
+      "step": 954
+    },
+    {
+      "epoch": 7.180451127819548,
+      "grad_norm": 0.21911864288143557,
+      "learning_rate": 1.521014346159131e-06,
+      "loss": 0.3096,
+      "step": 955
+    },
+    {
+      "epoch": 7.18796992481203,
+      "grad_norm": 0.20102212546112333,
+      "learning_rate": 1.519953224231287e-06,
+      "loss": 0.3007,
+      "step": 956
+    },
+    {
+      "epoch": 7.195488721804511,
+      "grad_norm": 0.20332071638125665,
+      "learning_rate": 1.5188912993140174e-06,
+      "loss": 0.3142,
+      "step": 957
+    },
+    {
+      "epoch": 7.203007518796992,
+      "grad_norm": 0.21921819239141443,
+      "learning_rate": 1.5178285730473067e-06,
+      "loss": 0.2983,
+      "step": 958
+    },
+    {
+      "epoch": 7.2105263157894735,
+      "grad_norm": 0.21026649933501168,
+      "learning_rate": 1.5167650470723739e-06,
+      "loss": 0.306,
+      "step": 959
+    },
+    {
+      "epoch": 7.2180451127819545,
+      "grad_norm": 0.21128354472374633,
+      "learning_rate": 1.5157007230316756e-06,
+      "loss": 0.3217,
+      "step": 960
+    },
+    {
+      "epoch": 7.225563909774436,
+      "grad_norm": 0.19953043386870342,
+      "learning_rate": 1.5146356025688998e-06,
+      "loss": 0.3025,
+      "step": 961
+    },
+    {
+      "epoch": 7.2330827067669174,
+      "grad_norm": 0.2093299243664622,
+      "learning_rate": 1.5135696873289646e-06,
+      "loss": 0.3104,
+      "step": 962
+    },
+    {
+      "epoch": 7.2406015037593985,
+      "grad_norm": 0.2084757830231614,
+      "learning_rate": 1.512502978958015e-06,
+      "loss": 0.2972,
+      "step": 963
+    },
+    {
+      "epoch": 7.2481203007518795,
+      "grad_norm": 0.21677819942539958,
+      "learning_rate": 1.5114354791034222e-06,
+      "loss": 0.312,
+      "step": 964
+    },
+    {
+      "epoch": 7.2556390977443606,
+      "grad_norm": 0.20254373592223793,
+      "learning_rate": 1.5103671894137784e-06,
+      "loss": 0.2906,
+      "step": 965
+    },
+    {
+      "epoch": 7.2631578947368425,
+      "grad_norm": 0.1966975008771876,
+      "learning_rate": 1.509298111538896e-06,
+      "loss": 0.3018,
+      "step": 966
+    },
+    {
+      "epoch": 7.2706766917293235,
+      "grad_norm": 0.20780744921761188,
+      "learning_rate": 1.5082282471298054e-06,
+      "loss": 0.3089,
+      "step": 967
+    },
+    {
+      "epoch": 7.2781954887218046,
+      "grad_norm": 0.2231578003232769,
+      "learning_rate": 1.5071575978387502e-06,
+      "loss": 0.3055,
+      "step": 968
+    },
+    {
+      "epoch": 7.285714285714286,
+      "grad_norm": 0.21067566951824146,
+      "learning_rate": 1.5060861653191874e-06,
+      "loss": 0.3067,
+      "step": 969
+    },
+    {
+      "epoch": 7.293233082706767,
+      "grad_norm": 0.20792884123256017,
+      "learning_rate": 1.5050139512257829e-06,
+      "loss": 0.2993,
+      "step": 970
+    },
+    {
+      "epoch": 7.3007518796992485,
+      "grad_norm": 0.20639769388918786,
+      "learning_rate": 1.50394095721441e-06,
+      "loss": 0.3116,
+      "step": 971
+    },
+    {
+      "epoch": 7.30827067669173,
+      "grad_norm": 0.21862393386472542,
+      "learning_rate": 1.502867184942146e-06,
+      "loss": 0.3089,
+      "step": 972
+    },
+    {
+      "epoch": 7.315789473684211,
+      "grad_norm": 0.2141709119718362,
+      "learning_rate": 1.5017926360672709e-06,
+      "loss": 0.3051,
+      "step": 973
+    },
+    {
+      "epoch": 7.323308270676692,
+      "grad_norm": 0.20520200878333011,
+      "learning_rate": 1.5007173122492634e-06,
+      "loss": 0.3034,
+      "step": 974
+    },
+    {
+      "epoch": 7.330827067669173,
+      "grad_norm": 0.2200473633798064,
+      "learning_rate": 1.4996412151487986e-06,
+      "loss": 0.2945,
+      "step": 975
+    },
+    {
+      "epoch": 7.338345864661654,
+      "grad_norm": 0.21354050934894486,
+      "learning_rate": 1.4985643464277474e-06,
+      "loss": 0.3052,
+      "step": 976
+    },
+    {
+      "epoch": 7.345864661654136,
+      "grad_norm": 0.2092927158717497,
+      "learning_rate": 1.4974867077491704e-06,
+      "loss": 0.3143,
+      "step": 977
+    },
+    {
+      "epoch": 7.353383458646617,
+      "grad_norm": 0.21642475005451853,
+      "learning_rate": 1.4964083007773188e-06,
+      "loss": 0.31,
+      "step": 978
+    },
+    {
+      "epoch": 7.360902255639098,
+      "grad_norm": 0.21098037488594398,
+      "learning_rate": 1.4953291271776292e-06,
+      "loss": 0.3021,
+      "step": 979
+    },
+    {
+      "epoch": 7.368421052631579,
+      "grad_norm": 0.2083290416229606,
+      "learning_rate": 1.4942491886167227e-06,
+      "loss": 0.2972,
+      "step": 980
+    },
+    {
+      "epoch": 7.37593984962406,
+      "grad_norm": 0.19987220247424053,
+      "learning_rate": 1.493168486762402e-06,
+      "loss": 0.2998,
+      "step": 981
+    },
+    {
+      "epoch": 7.383458646616542,
+      "grad_norm": 0.22255905333996887,
+      "learning_rate": 1.4920870232836484e-06,
+      "loss": 0.3079,
+      "step": 982
+    },
+    {
+      "epoch": 7.390977443609023,
+      "grad_norm": 0.20739333838336624,
+      "learning_rate": 1.4910047998506192e-06,
+      "loss": 0.3054,
+      "step": 983
+    },
+    {
+      "epoch": 7.398496240601504,
+      "grad_norm": 0.21370067604709275,
+      "learning_rate": 1.489921818134645e-06,
+      "loss": 0.2988,
+      "step": 984
+    },
+    {
+      "epoch": 7.406015037593985,
+      "grad_norm": 0.21119460235620954,
+      "learning_rate": 1.4888380798082287e-06,
+      "loss": 0.2968,
+      "step": 985
+    },
+    {
+      "epoch": 7.413533834586466,
+      "grad_norm": 0.2070057590815469,
+      "learning_rate": 1.4877535865450405e-06,
+      "loss": 0.2965,
+      "step": 986
+    },
+    {
+      "epoch": 7.421052631578947,
+      "grad_norm": 0.2307851748047924,
+      "learning_rate": 1.4866683400199165e-06,
+      "loss": 0.3051,
+      "step": 987
+    },
+    {
+      "epoch": 7.428571428571429,
+      "grad_norm": 0.2048867188764631,
+      "learning_rate": 1.4855823419088573e-06,
+      "loss": 0.3051,
+      "step": 988
+    },
+    {
+      "epoch": 7.43609022556391,
+      "grad_norm": 0.21061037537287386,
+      "learning_rate": 1.4844955938890226e-06,
+      "loss": 0.3053,
+      "step": 989
+    },
+    {
+      "epoch": 7.443609022556391,
+      "grad_norm": 0.20720189243365467,
+      "learning_rate": 1.4834080976387313e-06,
+      "loss": 0.3077,
+      "step": 990
+    },
+    {
+      "epoch": 7.451127819548872,
+      "grad_norm": 0.23482421169719778,
+      "learning_rate": 1.4823198548374577e-06,
+      "loss": 0.3144,
+      "step": 991
+    },
+    {
+      "epoch": 7.458646616541353,
+      "grad_norm": 0.19704107059721282,
+      "learning_rate": 1.4812308671658283e-06,
+      "loss": 0.2949,
+      "step": 992
+    },
+    {
+      "epoch": 7.466165413533835,
+      "grad_norm": 0.2182280233214576,
+      "learning_rate": 1.480141136305621e-06,
+      "loss": 0.3085,
+      "step": 993
+    },
+    {
+      "epoch": 7.473684210526316,
+      "grad_norm": 0.20652155546491494,
+      "learning_rate": 1.479050663939761e-06,
+      "loss": 0.3087,
+      "step": 994
+    },
+    {
+      "epoch": 7.481203007518797,
+      "grad_norm": 0.1936541522603524,
+      "learning_rate": 1.4779594517523184e-06,
+      "loss": 0.3086,
+      "step": 995
+    },
+    {
+      "epoch": 7.488721804511278,
+      "grad_norm": 0.20895621908959106,
+      "learning_rate": 1.476867501428506e-06,
+      "loss": 0.2999,
+      "step": 996
+    },
+    {
+      "epoch": 7.496240601503759,
+      "grad_norm": 0.22463025801404887,
+      "learning_rate": 1.4757748146546769e-06,
+      "loss": 0.317,
+      "step": 997
+    },
+    {
+      "epoch": 7.503759398496241,
+      "grad_norm": 0.20659673654060515,
+      "learning_rate": 1.4746813931183205e-06,
+      "loss": 0.2986,
+      "step": 998
+    },
+    {
+      "epoch": 7.511278195488722,
+      "grad_norm": 0.21382028891664315,
+      "learning_rate": 1.4735872385080625e-06,
+      "loss": 0.3022,
+      "step": 999
+    },
+    {
+      "epoch": 7.518796992481203,
+      "grad_norm": 0.22475393761457704,
+      "learning_rate": 1.4724923525136595e-06,
+      "loss": 0.301,
+      "step": 1000
+    },
+    {
+      "epoch": 7.526315789473684,
+      "grad_norm": 0.22213236302074785,
+      "learning_rate": 1.4713967368259978e-06,
+      "loss": 0.3042,
+      "step": 1001
+    },
+    {
+      "epoch": 7.533834586466165,
+      "grad_norm": 0.22986174422016992,
+      "learning_rate": 1.4703003931370908e-06,
+      "loss": 0.3028,
+      "step": 1002
+    },
+    {
+      "epoch": 7.541353383458647,
+      "grad_norm": 0.22336173279661478,
+      "learning_rate": 1.4692033231400763e-06,
+      "loss": 0.297,
+      "step": 1003
+    },
+    {
+      "epoch": 7.548872180451128,
+      "grad_norm": 0.2321161752995302,
+      "learning_rate": 1.4681055285292136e-06,
+      "loss": 0.3056,
+      "step": 1004
+    },
+    {
+      "epoch": 7.556390977443609,
+      "grad_norm": 0.20049708954733497,
+      "learning_rate": 1.4670070109998814e-06,
+      "loss": 0.296,
+      "step": 1005
+    },
+    {
+      "epoch": 7.56390977443609,
+      "grad_norm": 0.20776639538546973,
+      "learning_rate": 1.465907772248574e-06,
+      "loss": 0.2982,
+      "step": 1006
+    },
+    {
+      "epoch": 7.571428571428571,
+      "grad_norm": 0.2078827915774836,
+      "learning_rate": 1.4648078139729004e-06,
+      "loss": 0.2972,
+      "step": 1007
+    },
+    {
+      "epoch": 7.578947368421053,
+      "grad_norm": 0.19700550399284045,
+      "learning_rate": 1.4637071378715804e-06,
+      "loss": 0.2904,
+      "step": 1008
+    },
+    {
+      "epoch": 7.586466165413534,
+      "grad_norm": 0.21970225827804013,
+      "learning_rate": 1.4626057456444423e-06,
+      "loss": 0.3158,
+      "step": 1009
+    },
+    {
+      "epoch": 7.593984962406015,
+      "grad_norm": 0.2242660011836564,
+      "learning_rate": 1.4615036389924206e-06,
+      "loss": 0.308,
+      "step": 1010
+    },
+    {
+      "epoch": 7.601503759398496,
+      "grad_norm": 0.22619836818132635,
+      "learning_rate": 1.460400819617553e-06,
+      "loss": 0.2961,
+      "step": 1011
+    },
+    {
+      "epoch": 7.609022556390977,
+      "grad_norm": 0.2265593191921108,
+      "learning_rate": 1.4592972892229778e-06,
+      "loss": 0.3076,
+      "step": 1012
+    },
+    {
+      "epoch": 7.616541353383458,
+      "grad_norm": 0.21911678441632354,
+      "learning_rate": 1.4581930495129316e-06,
+      "loss": 0.2994,
+      "step": 1013
+    },
+    {
+      "epoch": 7.62406015037594,
+      "grad_norm": 0.20733964973494715,
+      "learning_rate": 1.457088102192746e-06,
+      "loss": 0.2908,
+      "step": 1014
+    },
+    {
+      "epoch": 7.631578947368421,
+      "grad_norm": 0.20049342753872562,
+      "learning_rate": 1.455982448968846e-06,
+      "loss": 0.2989,
+      "step": 1015
+    },
+    {
+      "epoch": 7.639097744360902,
+      "grad_norm": 0.19857362703087655,
+      "learning_rate": 1.454876091548746e-06,
+      "loss": 0.3104,
+      "step": 1016
+    },
+    {
+      "epoch": 7.646616541353383,
+      "grad_norm": 0.22007506692976259,
+      "learning_rate": 1.4537690316410489e-06,
+      "loss": 0.3138,
+      "step": 1017
+    },
+    {
+      "epoch": 7.654135338345864,
+      "grad_norm": 0.20187864198361868,
+      "learning_rate": 1.4526612709554417e-06,
+      "loss": 0.3067,
+      "step": 1018
+    },
+    {
+      "epoch": 7.661654135338345,
+      "grad_norm": 0.2127251270084208,
+      "learning_rate": 1.4515528112026937e-06,
+      "loss": 0.3057,
+      "step": 1019
+    },
+    {
+      "epoch": 7.669172932330827,
+      "grad_norm": 0.22507769046149131,
+      "learning_rate": 1.4504436540946545e-06,
+      "loss": 0.3011,
+      "step": 1020
+    },
+    {
+      "epoch": 7.676691729323308,
+      "grad_norm": 0.20501365019200407,
+      "learning_rate": 1.4493338013442498e-06,
+      "loss": 0.3199,
+      "step": 1021
+    },
+    {
+      "epoch": 7.684210526315789,
+      "grad_norm": 0.22411539566165253,
+      "learning_rate": 1.4482232546654797e-06,
+      "loss": 0.3095,
+      "step": 1022
+    },
+    {
+      "epoch": 7.69172932330827,
+      "grad_norm": 0.22004724615393004,
+      "learning_rate": 1.4471120157734168e-06,
+      "loss": 0.3009,
+      "step": 1023
+    },
+    {
+      "epoch": 7.6992481203007515,
+      "grad_norm": 0.20357440259133786,
+      "learning_rate": 1.4460000863842022e-06,
+      "loss": 0.2988,
+      "step": 1024
+    },
+    {
+      "epoch": 7.706766917293233,
+      "grad_norm": 0.21835229698278655,
+      "learning_rate": 1.4448874682150428e-06,
+      "loss": 0.3056,
+      "step": 1025
+    },
+    {
+      "epoch": 7.714285714285714,
+      "grad_norm": 0.21988983990206087,
+      "learning_rate": 1.4437741629842103e-06,
+      "loss": 0.3059,
+      "step": 1026
+    },
+    {
+      "epoch": 7.7218045112781954,
+      "grad_norm": 0.21744015641532968,
+      "learning_rate": 1.4426601724110362e-06,
+      "loss": 0.2926,
+      "step": 1027
+    },
+    {
+      "epoch": 7.7293233082706765,
+      "grad_norm": 0.20834033009932193,
+      "learning_rate": 1.4415454982159118e-06,
+      "loss": 0.3027,
+      "step": 1028
+    },
+    {
+      "epoch": 7.7368421052631575,
+      "grad_norm": 0.22797597999851474,
+      "learning_rate": 1.4404301421202832e-06,
+      "loss": 0.3063,
+      "step": 1029
+    },
+    {
+      "epoch": 7.7443609022556394,
+      "grad_norm": 0.20732662827142517,
+      "learning_rate": 1.43931410584665e-06,
+      "loss": 0.3011,
+      "step": 1030
+    },
+    {
+      "epoch": 7.7518796992481205,
+      "grad_norm": 0.20562270223338222,
+      "learning_rate": 1.438197391118562e-06,
+      "loss": 0.304,
+      "step": 1031
+    },
+    {
+      "epoch": 7.7593984962406015,
+      "grad_norm": 0.2184101326298157,
+      "learning_rate": 1.4370799996606166e-06,
+      "loss": 0.3008,
+      "step": 1032
+    },
+    {
+      "epoch": 7.7669172932330826,
+      "grad_norm": 0.19980933352273597,
+      "learning_rate": 1.4359619331984568e-06,
+      "loss": 0.2989,
+      "step": 1033
+    },
+    {
+      "epoch": 7.774436090225564,
+      "grad_norm": 0.20069612717474367,
+      "learning_rate": 1.4348431934587684e-06,
+      "loss": 0.312,
+      "step": 1034
+    },
+    {
+      "epoch": 7.7819548872180455,
+      "grad_norm": 0.22806051108990316,
+      "learning_rate": 1.4337237821692753e-06,
+      "loss": 0.3225,
+      "step": 1035
+    },
+    {
+      "epoch": 7.7894736842105265,
+      "grad_norm": 0.20146206831412372,
+      "learning_rate": 1.4326037010587405e-06,
+      "loss": 0.2992,
+      "step": 1036
+    },
+    {
+      "epoch": 7.796992481203008,
+      "grad_norm": 0.20254406889050466,
+      "learning_rate": 1.4314829518569598e-06,
+      "loss": 0.3016,
+      "step": 1037
+    },
+    {
+      "epoch": 7.804511278195489,
+      "grad_norm": 0.2182358592250706,
+      "learning_rate": 1.430361536294762e-06,
+      "loss": 0.3121,
+      "step": 1038
+    },
+    {
+      "epoch": 7.81203007518797,
+      "grad_norm": 0.2158915413150538,
+      "learning_rate": 1.4292394561040046e-06,
+      "loss": 0.2888,
+      "step": 1039
+    },
+    {
+      "epoch": 7.819548872180452,
+      "grad_norm": 0.2068369676354273,
+      "learning_rate": 1.4281167130175712e-06,
+      "loss": 0.3055,
+      "step": 1040
+    },
+    {
+      "epoch": 7.827067669172933,
+      "grad_norm": 0.21443676790369376,
+      "learning_rate": 1.4269933087693693e-06,
+      "loss": 0.3054,
+      "step": 1041
+    },
+    {
+      "epoch": 7.834586466165414,
+      "grad_norm": 0.2231564324339094,
+      "learning_rate": 1.4258692450943274e-06,
+      "loss": 0.3064,
+      "step": 1042
+    },
+    {
+      "epoch": 7.842105263157895,
+      "grad_norm": 0.203130659888196,
+      "learning_rate": 1.4247445237283928e-06,
+      "loss": 0.3079,
+      "step": 1043
+    },
+    {
+      "epoch": 7.849624060150376,
+      "grad_norm": 0.21677065088992487,
+      "learning_rate": 1.4236191464085282e-06,
+      "loss": 0.2989,
+      "step": 1044
+    },
+    {
+      "epoch": 7.857142857142857,
+      "grad_norm": 0.20678039461880873,
+      "learning_rate": 1.422493114872709e-06,
+      "loss": 0.3082,
+      "step": 1045
+    },
+    {
+      "epoch": 7.864661654135339,
+      "grad_norm": 0.2321345531523906,
+      "learning_rate": 1.4213664308599219e-06,
+      "loss": 0.3019,
+      "step": 1046
+    },
+    {
+      "epoch": 7.87218045112782,
+      "grad_norm": 0.21557816804397253,
+      "learning_rate": 1.4202390961101597e-06,
+      "loss": 0.2912,
+      "step": 1047
+    },
+    {
+      "epoch": 7.879699248120301,
+      "grad_norm": 0.21040765261440572,
+      "learning_rate": 1.419111112364422e-06,
+      "loss": 0.3079,
+      "step": 1048
+    },
+    {
+      "epoch": 7.887218045112782,
+      "grad_norm": 0.19977221230393577,
+      "learning_rate": 1.4179824813647092e-06,
+      "loss": 0.3,
+      "step": 1049
+    },
+    {
+      "epoch": 7.894736842105263,
+      "grad_norm": 0.2096252956538686,
+      "learning_rate": 1.4168532048540223e-06,
+      "loss": 0.2959,
+      "step": 1050
+    },
+    {
+      "epoch": 7.902255639097744,
+      "grad_norm": 0.21384926208970823,
+      "learning_rate": 1.4157232845763583e-06,
+      "loss": 0.3014,
+      "step": 1051
+    },
+    {
+      "epoch": 7.909774436090226,
+      "grad_norm": 0.2172676455195607,
+      "learning_rate": 1.414592722276709e-06,
+      "loss": 0.2927,
+      "step": 1052
+    },
+    {
+      "epoch": 7.917293233082707,
+      "grad_norm": 0.20770294265686418,
+      "learning_rate": 1.4134615197010576e-06,
+      "loss": 0.31,
+      "step": 1053
+    },
+    {
+      "epoch": 7.924812030075188,
+      "grad_norm": 0.2140710575815151,
+      "learning_rate": 1.4123296785963759e-06,
+      "loss": 0.3068,
+      "step": 1054
+    },
+    {
+      "epoch": 7.932330827067669,
+      "grad_norm": 0.22575911188520523,
+      "learning_rate": 1.4111972007106223e-06,
+      "loss": 0.3044,
+      "step": 1055
+    },
+    {
+      "epoch": 7.93984962406015,
+      "grad_norm": 0.21134117410192374,
+      "learning_rate": 1.410064087792738e-06,
+      "loss": 0.3144,
+      "step": 1056
+    },
+    {
+      "epoch": 7.947368421052632,
+      "grad_norm": 0.2046761554991222,
+      "learning_rate": 1.4089303415926457e-06,
+      "loss": 0.3118,
+      "step": 1057
+    },
+    {
+      "epoch": 7.954887218045113,
+      "grad_norm": 0.20953573245243937,
+      "learning_rate": 1.4077959638612448e-06,
+      "loss": 0.3097,
+      "step": 1058
+    },
+    {
+      "epoch": 7.962406015037594,
+      "grad_norm": 0.2130723334377698,
+      "learning_rate": 1.4066609563504117e-06,
+      "loss": 0.2967,
+      "step": 1059
+    },
+    {
+      "epoch": 7.969924812030075,
+      "grad_norm": 0.2093643737473869,
+      "learning_rate": 1.4055253208129937e-06,
+      "loss": 0.2939,
+      "step": 1060
+    },
+    {
+      "epoch": 7.977443609022556,
+      "grad_norm": 0.20941927614526787,
+      "learning_rate": 1.4043890590028093e-06,
+      "loss": 0.2908,
+      "step": 1061
+    },
+    {
+      "epoch": 7.984962406015038,
+      "grad_norm": 0.2029502676305885,
+      "learning_rate": 1.4032521726746437e-06,
+      "loss": 0.2876,
+      "step": 1062
+    },
+    {
+      "epoch": 7.992481203007519,
+      "grad_norm": 0.23497736147830545,
+      "learning_rate": 1.4021146635842463e-06,
+      "loss": 0.3031,
+      "step": 1063
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.21241643028196838,
+      "learning_rate": 1.4009765334883286e-06,
+      "loss": 0.2953,
+      "step": 1064
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 0.4088137149810791,
+      "eval_runtime": 1.69,
+      "eval_samples_per_second": 193.49,
+      "eval_steps_per_second": 3.55,
+      "step": 1064
+    },
+    {
+      "epoch": 8.007518796992482,
+      "grad_norm": 0.20817782030654958,
+      "learning_rate": 1.3998377841445612e-06,
+      "loss": 0.3014,
+      "step": 1065
+    },
+    {
+      "epoch": 8.015037593984962,
+      "grad_norm": 0.2133394029921935,
+      "learning_rate": 1.3986984173115708e-06,
+      "loss": 0.3033,
+      "step": 1066
+    },
+    {
+      "epoch": 8.022556390977444,
+      "grad_norm": 0.2057995280708651,
+      "learning_rate": 1.3975584347489382e-06,
+      "loss": 0.2935,
+      "step": 1067
+    },
+    {
+      "epoch": 8.030075187969924,
+      "grad_norm": 0.22449262737720213,
+      "learning_rate": 1.396417838217194e-06,
+      "loss": 0.3129,
+      "step": 1068
+    },
+    {
+      "epoch": 8.037593984962406,
+      "grad_norm": 0.21265805071465954,
+      "learning_rate": 1.3952766294778183e-06,
+      "loss": 0.3111,
+      "step": 1069
+    },
+    {
+      "epoch": 8.045112781954888,
+      "grad_norm": 0.21767201134067785,
+      "learning_rate": 1.3941348102932358e-06,
+      "loss": 0.299,
+      "step": 1070
+    },
+    {
+      "epoch": 8.052631578947368,
+      "grad_norm": 0.2179851028638274,
+      "learning_rate": 1.3929923824268143e-06,
+      "loss": 0.2938,
+      "step": 1071
+    },
+    {
+      "epoch": 8.06015037593985,
+      "grad_norm": 0.2263997200150036,
+      "learning_rate": 1.3918493476428617e-06,
+      "loss": 0.2908,
+      "step": 1072
+    },
+    {
+      "epoch": 8.06766917293233,
+      "grad_norm": 0.20600352116454262,
+      "learning_rate": 1.3907057077066226e-06,
+      "loss": 0.3057,
+      "step": 1073
+    },
+    {
+      "epoch": 8.075187969924812,
+      "grad_norm": 0.22146919346976246,
+      "learning_rate": 1.3895614643842772e-06,
+      "loss": 0.3116,
+      "step": 1074
+    },
+    {
+      "epoch": 8.082706766917294,
+      "grad_norm": 0.21707124442278705,
+      "learning_rate": 1.3884166194429364e-06,
+      "loss": 0.2954,
+      "step": 1075
+    },
+    {
+      "epoch": 8.090225563909774,
+      "grad_norm": 0.204641507472267,
+      "learning_rate": 1.3872711746506412e-06,
+      "loss": 0.288,
+      "step": 1076
+    },
+    {
+      "epoch": 8.097744360902256,
+      "grad_norm": 0.20192188089122523,
+      "learning_rate": 1.386125131776358e-06,
+      "loss": 0.286,
+      "step": 1077
+    },
+    {
+      "epoch": 8.105263157894736,
+      "grad_norm": 0.20274838768130682,
+      "learning_rate": 1.3849784925899777e-06,
+      "loss": 0.3023,
+      "step": 1078
+    },
+    {
+      "epoch": 8.112781954887218,
+      "grad_norm": 0.19618842925793323,
+      "learning_rate": 1.3838312588623118e-06,
+      "loss": 0.3072,
+      "step": 1079
+    },
+    {
+      "epoch": 8.1203007518797,
+      "grad_norm": 0.21979111077561195,
+      "learning_rate": 1.3826834323650898e-06,
+      "loss": 0.2946,
+      "step": 1080
+    },
+    {
+      "epoch": 8.12781954887218,
+      "grad_norm": 0.22142542827205883,
+      "learning_rate": 1.3815350148709567e-06,
+      "loss": 0.2967,
+      "step": 1081
+    },
+    {
+      "epoch": 8.135338345864662,
+      "grad_norm": 0.22699770768760275,
+      "learning_rate": 1.3803860081534707e-06,
+      "loss": 0.3039,
+      "step": 1082
+    },
+    {
+      "epoch": 8.142857142857142,
+      "grad_norm": 0.22179781456491268,
+      "learning_rate": 1.3792364139870997e-06,
+      "loss": 0.3037,
+      "step": 1083
+    },
+    {
+      "epoch": 8.150375939849624,
+      "grad_norm": 0.20148208464605669,
+      "learning_rate": 1.3780862341472182e-06,
+      "loss": 0.2902,
+      "step": 1084
+    },
+    {
+      "epoch": 8.157894736842104,
+      "grad_norm": 0.2017796053275557,
+      "learning_rate": 1.3769354704101058e-06,
+      "loss": 0.2971,
+      "step": 1085
+    },
+    {
+      "epoch": 8.165413533834586,
+      "grad_norm": 0.22406532488262934,
+      "learning_rate": 1.375784124552944e-06,
+      "loss": 0.3016,
+      "step": 1086
+    },
+    {
+      "epoch": 8.172932330827068,
+      "grad_norm": 0.20981598555350164,
+      "learning_rate": 1.3746321983538127e-06,
+      "loss": 0.3008,
+      "step": 1087
+    },
+    {
+      "epoch": 8.180451127819548,
+      "grad_norm": 0.22666808129405372,
+      "learning_rate": 1.3734796935916885e-06,
+      "loss": 0.3012,
+      "step": 1088
+    },
+    {
+      "epoch": 8.18796992481203,
+      "grad_norm": 0.21411165425710393,
+      "learning_rate": 1.3723266120464417e-06,
+      "loss": 0.3058,
+      "step": 1089
+    },
+    {
+      "epoch": 8.19548872180451,
+      "grad_norm": 0.21901804774876135,
+      "learning_rate": 1.3711729554988322e-06,
+      "loss": 0.2969,
+      "step": 1090
+    },
+    {
+      "epoch": 8.203007518796992,
+      "grad_norm": 0.26075287234745315,
+      "learning_rate": 1.3700187257305099e-06,
+      "loss": 0.3009,
+      "step": 1091
+    },
+    {
+      "epoch": 8.210526315789474,
+      "grad_norm": 0.21519712464746138,
+      "learning_rate": 1.3688639245240078e-06,
+      "loss": 0.3096,
+      "step": 1092
+    },
+    {
+      "epoch": 8.218045112781954,
+      "grad_norm": 0.20989026647466416,
+      "learning_rate": 1.3677085536627428e-06,
+      "loss": 0.2996,
+      "step": 1093
+    },
+    {
+      "epoch": 8.225563909774436,
+      "grad_norm": 0.21474660995598494,
+      "learning_rate": 1.3665526149310114e-06,
+      "loss": 0.2979,
+      "step": 1094
+    },
+    {
+      "epoch": 8.233082706766917,
+      "grad_norm": 0.2134225685790549,
+      "learning_rate": 1.3653961101139864e-06,
+      "loss": 0.3028,
+      "step": 1095
+    },
+    {
+      "epoch": 8.240601503759398,
+      "grad_norm": 0.2097631556590145,
+      "learning_rate": 1.3642390409977154e-06,
+      "loss": 0.3,
+      "step": 1096
+    },
+    {
+      "epoch": 8.24812030075188,
+      "grad_norm": 0.21086640220125608,
+      "learning_rate": 1.3630814093691174e-06,
+      "loss": 0.3059,
+      "step": 1097
+    },
+    {
+      "epoch": 8.25563909774436,
+      "grad_norm": 0.2131316317944869,
+      "learning_rate": 1.36192321701598e-06,
+      "loss": 0.2923,
+      "step": 1098
+    },
+    {
+      "epoch": 8.263157894736842,
+      "grad_norm": 0.22096413786421468,
+      "learning_rate": 1.3607644657269568e-06,
+      "loss": 0.2879,
+      "step": 1099
+    },
+    {
+      "epoch": 8.270676691729323,
+      "grad_norm": 0.21212872497711976,
+      "learning_rate": 1.3596051572915649e-06,
+      "loss": 0.2999,
+      "step": 1100
+    },
+    {
+      "epoch": 8.278195488721805,
+      "grad_norm": 0.21190359856323573,
+      "learning_rate": 1.3584452935001809e-06,
+      "loss": 0.3003,
+      "step": 1101
+    },
+    {
+      "epoch": 8.285714285714286,
+      "grad_norm": 0.2137879581119127,
+      "learning_rate": 1.3572848761440402e-06,
+      "loss": 0.295,
+      "step": 1102
+    },
+    {
+      "epoch": 8.293233082706767,
+      "grad_norm": 0.21537641678487898,
+      "learning_rate": 1.3561239070152324e-06,
+      "loss": 0.2794,
+      "step": 1103
+    },
+    {
+      "epoch": 8.300751879699249,
+      "grad_norm": 0.21264373031612438,
+      "learning_rate": 1.3549623879066994e-06,
+      "loss": 0.29,
+      "step": 1104
+    },
+    {
+      "epoch": 8.308270676691729,
+      "grad_norm": 0.2127152862822725,
+      "learning_rate": 1.3538003206122326e-06,
+      "loss": 0.3069,
+      "step": 1105
+    },
+    {
+      "epoch": 8.31578947368421,
+      "grad_norm": 0.21436150784949187,
+      "learning_rate": 1.3526377069264698e-06,
+      "loss": 0.3052,
+      "step": 1106
+    },
+    {
+      "epoch": 8.323308270676693,
+      "grad_norm": 0.2332249678868736,
+      "learning_rate": 1.3514745486448927e-06,
+      "loss": 0.3022,
+      "step": 1107
+    },
+    {
+      "epoch": 8.330827067669173,
+      "grad_norm": 0.2056306874023312,
+      "learning_rate": 1.3503108475638244e-06,
+      "loss": 0.2929,
+      "step": 1108
+    },
+    {
+      "epoch": 8.338345864661655,
+      "grad_norm": 0.2261400809067208,
+      "learning_rate": 1.3491466054804251e-06,
+      "loss": 0.3072,
+      "step": 1109
+    },
+    {
+      "epoch": 8.345864661654135,
+      "grad_norm": 0.1983912838920007,
+      "learning_rate": 1.347981824192692e-06,
+      "loss": 0.292,
+      "step": 1110
+    },
+    {
+      "epoch": 8.353383458646617,
+      "grad_norm": 0.2146796148037333,
+      "learning_rate": 1.346816505499454e-06,
+      "loss": 0.3056,
+      "step": 1111
+    },
+    {
+      "epoch": 8.360902255639097,
+      "grad_norm": 0.2073315294146159,
+      "learning_rate": 1.3456506512003704e-06,
+      "loss": 0.287,
+      "step": 1112
+    },
+    {
+      "epoch": 8.368421052631579,
+      "grad_norm": 0.20553747334341252,
+      "learning_rate": 1.3444842630959277e-06,
+      "loss": 0.2901,
+      "step": 1113
+    },
+    {
+      "epoch": 8.37593984962406,
+      "grad_norm": 0.20673717869413888,
+      "learning_rate": 1.3433173429874364e-06,
+      "loss": 0.2957,
+      "step": 1114
+    },
+    {
+      "epoch": 8.38345864661654,
+      "grad_norm": 0.20405048843234588,
+      "learning_rate": 1.3421498926770287e-06,
+      "loss": 0.292,
+      "step": 1115
+    },
+    {
+      "epoch": 8.390977443609023,
+      "grad_norm": 0.21521418367273532,
+      "learning_rate": 1.3409819139676558e-06,
+      "loss": 0.3046,
+      "step": 1116
+    },
+    {
+      "epoch": 8.398496240601503,
+      "grad_norm": 0.21937138801383918,
+      "learning_rate": 1.3398134086630851e-06,
+      "loss": 0.292,
+      "step": 1117
+    },
+    {
+      "epoch": 8.406015037593985,
+      "grad_norm": 0.21300669796091187,
+      "learning_rate": 1.3386443785678969e-06,
+      "loss": 0.3037,
+      "step": 1118
+    },
+    {
+      "epoch": 8.413533834586467,
+      "grad_norm": 0.21637213563737281,
+      "learning_rate": 1.3374748254874816e-06,
+      "loss": 0.2986,
+      "step": 1119
+    },
+    {
+      "epoch": 8.421052631578947,
+      "grad_norm": 0.20116002445441078,
+      "learning_rate": 1.336304751228039e-06,
+      "loss": 0.305,
+      "step": 1120
+    },
+    {
+      "epoch": 8.428571428571429,
+      "grad_norm": 0.20504364624083374,
+      "learning_rate": 1.3351341575965709e-06,
+      "loss": 0.2978,
+      "step": 1121
+    },
+    {
+      "epoch": 8.436090225563909,
+      "grad_norm": 0.2001467631305983,
+      "learning_rate": 1.3339630464008838e-06,
+      "loss": 0.3013,
+      "step": 1122
+    },
+    {
+      "epoch": 8.443609022556391,
+      "grad_norm": 0.20467776008020364,
+      "learning_rate": 1.3327914194495823e-06,
+      "loss": 0.3055,
+      "step": 1123
+    },
+    {
+      "epoch": 8.451127819548873,
+      "grad_norm": 0.19740431670562725,
+      "learning_rate": 1.3316192785520678e-06,
+      "loss": 0.2764,
+      "step": 1124
+    },
+    {
+      "epoch": 8.458646616541353,
+      "grad_norm": 0.21556644950753934,
+      "learning_rate": 1.3304466255185352e-06,
+      "loss": 0.3009,
+      "step": 1125
+    },
+    {
+      "epoch": 8.466165413533835,
+      "grad_norm": 0.20210811402318257,
+      "learning_rate": 1.3292734621599706e-06,
+      "loss": 0.2994,
+      "step": 1126
+    },
+    {
+      "epoch": 8.473684210526315,
+      "grad_norm": 0.22808688500139443,
+      "learning_rate": 1.3280997902881478e-06,
+      "loss": 0.3046,
+      "step": 1127
+    },
+    {
+      "epoch": 8.481203007518797,
+      "grad_norm": 0.21942249710886044,
+      "learning_rate": 1.3269256117156266e-06,
+      "loss": 0.2962,
+      "step": 1128
+    },
+    {
+      "epoch": 8.488721804511279,
+      "grad_norm": 0.2132686773882206,
+      "learning_rate": 1.3257509282557486e-06,
+      "loss": 0.3092,
+      "step": 1129
+    },
+    {
+      "epoch": 8.496240601503759,
+      "grad_norm": 0.2196457897231672,
+      "learning_rate": 1.3245757417226355e-06,
+      "loss": 0.309,
+      "step": 1130
+    },
+    {
+      "epoch": 8.503759398496241,
+      "grad_norm": 0.21433109933411607,
+      "learning_rate": 1.323400053931186e-06,
+      "loss": 0.2825,
+      "step": 1131
+    },
+    {
+      "epoch": 8.511278195488721,
+      "grad_norm": 0.22180547884755783,
+      "learning_rate": 1.3222238666970727e-06,
+      "loss": 0.3138,
+      "step": 1132
+    },
+    {
+      "epoch": 8.518796992481203,
+      "grad_norm": 0.22244768469985368,
+      "learning_rate": 1.3210471818367395e-06,
+      "loss": 0.2979,
+      "step": 1133
+    },
+    {
+      "epoch": 8.526315789473685,
+      "grad_norm": 0.2242663471056699,
+      "learning_rate": 1.3198700011673989e-06,
+      "loss": 0.306,
+      "step": 1134
+    },
+    {
+      "epoch": 8.533834586466165,
+      "grad_norm": 0.19659603839969708,
+      "learning_rate": 1.3186923265070293e-06,
+      "loss": 0.2825,
+      "step": 1135
+    },
+    {
+      "epoch": 8.541353383458647,
+      "grad_norm": 0.2024585933488925,
+      "learning_rate": 1.3175141596743717e-06,
+      "loss": 0.2987,
+      "step": 1136
+    },
+    {
+      "epoch": 8.548872180451127,
+      "grad_norm": 0.19898440046689905,
+      "learning_rate": 1.3163355024889274e-06,
+      "loss": 0.2964,
+      "step": 1137
+    },
+    {
+      "epoch": 8.556390977443609,
+      "grad_norm": 0.21503585611798087,
+      "learning_rate": 1.3151563567709546e-06,
+      "loss": 0.3047,
+      "step": 1138
+    },
+    {
+      "epoch": 8.563909774436091,
+      "grad_norm": 0.20822712265527593,
+      "learning_rate": 1.3139767243414662e-06,
+      "loss": 0.2914,
+      "step": 1139
+    },
+    {
+      "epoch": 8.571428571428571,
+      "grad_norm": 0.2199354427995726,
+      "learning_rate": 1.3127966070222272e-06,
+      "loss": 0.296,
+      "step": 1140
+    },
+    {
+      "epoch": 8.578947368421053,
+      "grad_norm": 0.20587713487731102,
+      "learning_rate": 1.3116160066357504e-06,
+      "loss": 0.3053,
+      "step": 1141
+    },
+    {
+      "epoch": 8.586466165413533,
+      "grad_norm": 0.22471222216721196,
+      "learning_rate": 1.310434925005296e-06,
+      "loss": 0.2956,
+      "step": 1142
+    },
+    {
+      "epoch": 8.593984962406015,
+      "grad_norm": 0.2156778454849994,
+      "learning_rate": 1.309253363954866e-06,
+      "loss": 0.3059,
+      "step": 1143
+    },
+    {
+      "epoch": 8.601503759398497,
+      "grad_norm": 0.21266323469780982,
+      "learning_rate": 1.3080713253092037e-06,
+      "loss": 0.3029,
+      "step": 1144
+    },
+    {
+      "epoch": 8.609022556390977,
+      "grad_norm": 0.23671277652318629,
+      "learning_rate": 1.3068888108937898e-06,
+      "loss": 0.3047,
+      "step": 1145
+    },
+    {
+      "epoch": 8.61654135338346,
+      "grad_norm": 0.22157443692780895,
+      "learning_rate": 1.3057058225348399e-06,
+      "loss": 0.3009,
+      "step": 1146
+    },
+    {
+      "epoch": 8.62406015037594,
+      "grad_norm": 0.20912620337891097,
+      "learning_rate": 1.3045223620593005e-06,
+      "loss": 0.3064,
+      "step": 1147
+    },
+    {
+      "epoch": 8.631578947368421,
+      "grad_norm": 0.20131673917497922,
+      "learning_rate": 1.3033384312948486e-06,
+      "loss": 0.2962,
+      "step": 1148
+    },
+    {
+      "epoch": 8.639097744360903,
+      "grad_norm": 0.19497047373734824,
+      "learning_rate": 1.302154032069887e-06,
+      "loss": 0.2789,
+      "step": 1149
+    },
+    {
+      "epoch": 8.646616541353383,
+      "grad_norm": 0.20531390005094816,
+      "learning_rate": 1.3009691662135413e-06,
+      "loss": 0.2987,
+      "step": 1150
+    },
+    {
+      "epoch": 8.654135338345865,
+      "grad_norm": 0.21493133797420882,
+      "learning_rate": 1.299783835555659e-06,
+      "loss": 0.2928,
+      "step": 1151
+    },
+    {
+      "epoch": 8.661654135338345,
+      "grad_norm": 0.2154943258171156,
+      "learning_rate": 1.2985980419268043e-06,
+      "loss": 0.2946,
+      "step": 1152
+    },
+    {
+      "epoch": 8.669172932330827,
+      "grad_norm": 0.2210772267822021,
+      "learning_rate": 1.297411787158257e-06,
+      "loss": 0.2992,
+      "step": 1153
+    },
+    {
+      "epoch": 8.676691729323307,
+      "grad_norm": 0.2138583928671344,
+      "learning_rate": 1.296225073082009e-06,
+      "loss": 0.2999,
+      "step": 1154
+    },
+    {
+      "epoch": 8.68421052631579,
+      "grad_norm": 0.19756610215818468,
+      "learning_rate": 1.295037901530761e-06,
+      "loss": 0.2999,
+      "step": 1155
+    },
+    {
+      "epoch": 8.691729323308271,
+      "grad_norm": 0.23625528626730638,
+      "learning_rate": 1.2938502743379209e-06,
+      "loss": 0.3191,
+      "step": 1156
+    },
+    {
+      "epoch": 8.699248120300751,
+      "grad_norm": 0.19916550250788262,
+      "learning_rate": 1.2926621933376001e-06,
+      "loss": 0.2959,
+      "step": 1157
+    },
+    {
+      "epoch": 8.706766917293233,
+      "grad_norm": 0.19974905609943985,
+      "learning_rate": 1.2914736603646106e-06,
+      "loss": 0.3034,
+      "step": 1158
+    },
+    {
+      "epoch": 8.714285714285714,
+      "grad_norm": 0.21792378011493949,
+      "learning_rate": 1.2902846772544622e-06,
+      "loss": 0.3045,
+      "step": 1159
+    },
+    {
+      "epoch": 8.721804511278195,
+      "grad_norm": 0.21445777747365574,
+      "learning_rate": 1.2890952458433607e-06,
+      "loss": 0.2861,
+      "step": 1160
+    },
+    {
+      "epoch": 8.729323308270677,
+      "grad_norm": 0.20759334214702657,
+      "learning_rate": 1.2879053679682036e-06,
+      "loss": 0.2996,
+      "step": 1161
+    },
+    {
+      "epoch": 8.736842105263158,
+      "grad_norm": 0.2090741657105473,
+      "learning_rate": 1.286715045466578e-06,
+      "loss": 0.3019,
+      "step": 1162
+    },
+    {
+      "epoch": 8.74436090225564,
+      "grad_norm": 0.21694531421912436,
+      "learning_rate": 1.2855242801767576e-06,
+      "loss": 0.3023,
+      "step": 1163
+    },
+    {
+      "epoch": 8.75187969924812,
+      "grad_norm": 0.2093495351011162,
+      "learning_rate": 1.2843330739377001e-06,
+      "loss": 0.2965,
+      "step": 1164
+    },
+    {
+      "epoch": 8.759398496240602,
+      "grad_norm": 0.20034868619340887,
+      "learning_rate": 1.283141428589044e-06,
+      "loss": 0.2986,
+      "step": 1165
+    },
+    {
+      "epoch": 8.766917293233083,
+      "grad_norm": 0.20324783216979675,
+      "learning_rate": 1.281949345971106e-06,
+      "loss": 0.302,
+      "step": 1166
+    },
+    {
+      "epoch": 8.774436090225564,
+      "grad_norm": 0.21733426864966626,
+      "learning_rate": 1.280756827924878e-06,
+      "loss": 0.3012,
+      "step": 1167
+    },
+    {
+      "epoch": 8.781954887218046,
+      "grad_norm": 0.20530501961205552,
+      "learning_rate": 1.279563876292025e-06,
+      "loss": 0.2853,
+      "step": 1168
+    },
+    {
+      "epoch": 8.789473684210526,
+      "grad_norm": 0.20673516416103366,
+      "learning_rate": 1.2783704929148807e-06,
+      "loss": 0.2899,
+      "step": 1169
+    },
+    {
+      "epoch": 8.796992481203008,
+      "grad_norm": 0.20621212126129707,
+      "learning_rate": 1.277176679636446e-06,
+      "loss": 0.2975,
+      "step": 1170
+    },
+    {
+      "epoch": 8.80451127819549,
+      "grad_norm": 0.21398159808037504,
+      "learning_rate": 1.2759824383003854e-06,
+      "loss": 0.2951,
+      "step": 1171
+    },
+    {
+      "epoch": 8.81203007518797,
+      "grad_norm": 0.2378456968790718,
+      "learning_rate": 1.274787770751025e-06,
+      "loss": 0.304,
+      "step": 1172
+    },
+    {
+      "epoch": 8.819548872180452,
+      "grad_norm": 0.2135960705720791,
+      "learning_rate": 1.2735926788333492e-06,
+      "loss": 0.2986,
+      "step": 1173
+    },
+    {
+      "epoch": 8.827067669172932,
+      "grad_norm": 0.21581681059335892,
+      "learning_rate": 1.272397164392997e-06,
+      "loss": 0.2986,
+      "step": 1174
+    },
+    {
+      "epoch": 8.834586466165414,
+      "grad_norm": 0.20173206751910175,
+      "learning_rate": 1.2712012292762601e-06,
+      "loss": 0.2844,
+      "step": 1175
+    },
+    {
+      "epoch": 8.842105263157894,
+      "grad_norm": 0.2079216741243103,
+      "learning_rate": 1.2700048753300804e-06,
+      "loss": 0.2927,
+      "step": 1176
+    },
+    {
+      "epoch": 8.849624060150376,
+      "grad_norm": 0.21375519130699175,
+      "learning_rate": 1.2688081044020465e-06,
+      "loss": 0.2978,
+      "step": 1177
+    },
+    {
+      "epoch": 8.857142857142858,
+      "grad_norm": 0.21340344496722288,
+      "learning_rate": 1.2676109183403907e-06,
+      "loss": 0.297,
+      "step": 1178
+    },
+    {
+      "epoch": 8.864661654135338,
+      "grad_norm": 0.21289902612407594,
+      "learning_rate": 1.2664133189939865e-06,
+      "loss": 0.3039,
+      "step": 1179
+    },
+    {
+      "epoch": 8.87218045112782,
+      "grad_norm": 0.2074879733391532,
+      "learning_rate": 1.2652153082123455e-06,
+      "loss": 0.2938,
+      "step": 1180
+    },
+    {
+      "epoch": 8.8796992481203,
+      "grad_norm": 0.20599575825982888,
+      "learning_rate": 1.2640168878456155e-06,
+      "loss": 0.2897,
+      "step": 1181
+    },
+    {
+      "epoch": 8.887218045112782,
+      "grad_norm": 0.21037552190713538,
+      "learning_rate": 1.2628180597445752e-06,
+      "loss": 0.2956,
+      "step": 1182
+    },
+    {
+      "epoch": 8.894736842105264,
+      "grad_norm": 0.21828876260650584,
+      "learning_rate": 1.2616188257606349e-06,
+      "loss": 0.3015,
+      "step": 1183
+    },
+    {
+      "epoch": 8.902255639097744,
+      "grad_norm": 0.22880900827996184,
+      "learning_rate": 1.2604191877458306e-06,
+      "loss": 0.2925,
+      "step": 1184
+    },
+    {
+      "epoch": 8.909774436090226,
+      "grad_norm": 0.21305291030812987,
+      "learning_rate": 1.259219147552822e-06,
+      "loss": 0.2988,
+      "step": 1185
+    },
+    {
+      "epoch": 8.917293233082706,
+      "grad_norm": 0.20723486236019464,
+      "learning_rate": 1.258018707034891e-06,
+      "loss": 0.3003,
+      "step": 1186
+    },
+    {
+      "epoch": 8.924812030075188,
+      "grad_norm": 0.20613075490814792,
+      "learning_rate": 1.256817868045937e-06,
+      "loss": 0.2924,
+      "step": 1187
+    },
+    {
+      "epoch": 8.93233082706767,
+      "grad_norm": 0.20870234243315994,
+      "learning_rate": 1.2556166324404746e-06,
+      "loss": 0.3092,
+      "step": 1188
+    },
+    {
+      "epoch": 8.93984962406015,
+      "grad_norm": 0.2025529465671229,
+      "learning_rate": 1.2544150020736317e-06,
+      "loss": 0.2917,
+      "step": 1189
+    },
+    {
+      "epoch": 8.947368421052632,
+      "grad_norm": 0.20581996734778843,
+      "learning_rate": 1.253212978801145e-06,
+      "loss": 0.2914,
+      "step": 1190
+    },
+    {
+      "epoch": 8.954887218045112,
+      "grad_norm": 0.20165590090270216,
+      "learning_rate": 1.2520105644793586e-06,
+      "loss": 0.2904,
+      "step": 1191
+    },
+    {
+      "epoch": 8.962406015037594,
+      "grad_norm": 0.215632818082792,
+      "learning_rate": 1.25080776096522e-06,
+      "loss": 0.3075,
+      "step": 1192
+    },
+    {
+      "epoch": 8.969924812030076,
+      "grad_norm": 0.24924847106525386,
+      "learning_rate": 1.2496045701162783e-06,
+      "loss": 0.2944,
+      "step": 1193
+    },
+    {
+      "epoch": 8.977443609022556,
+      "grad_norm": 0.22661415929132303,
+      "learning_rate": 1.2484009937906806e-06,
+      "loss": 0.3058,
+      "step": 1194
+    },
+    {
+      "epoch": 8.984962406015038,
+      "grad_norm": 0.2025793548846957,
+      "learning_rate": 1.247197033847169e-06,
+      "loss": 0.2881,
+      "step": 1195
+    },
+    {
+      "epoch": 8.992481203007518,
+      "grad_norm": 0.21511673175671492,
+      "learning_rate": 1.2459926921450779e-06,
+      "loss": 0.2931,
+      "step": 1196
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.2131216257839441,
+      "learning_rate": 1.2447879705443325e-06,
+      "loss": 0.3008,
+      "step": 1197
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 0.41143351793289185,
+      "eval_runtime": 1.6666,
+      "eval_samples_per_second": 196.204,
+      "eval_steps_per_second": 3.6,
+      "step": 1197
+    },
+    {
+      "epoch": 9.007518796992482,
+      "grad_norm": 0.224424638577128,
+      "learning_rate": 1.243582870905443e-06,
+      "loss": 0.2985,
+      "step": 1198
+    },
+    {
+      "epoch": 9.015037593984962,
+      "grad_norm": 0.20269452103934013,
+      "learning_rate": 1.242377395089505e-06,
+      "loss": 0.292,
+      "step": 1199
+    },
+    {
+      "epoch": 9.022556390977444,
+      "grad_norm": 0.19873154603484666,
+      "learning_rate": 1.2411715449581937e-06,
+      "loss": 0.3114,
+      "step": 1200
+    },
+    {
+      "epoch": 9.030075187969924,
+      "grad_norm": 0.2080085731980578,
+      "learning_rate": 1.239965322373763e-06,
+      "loss": 0.297,
+      "step": 1201
+    },
+    {
+      "epoch": 9.037593984962406,
+      "grad_norm": 0.20571836125610782,
+      "learning_rate": 1.2387587291990422e-06,
+      "loss": 0.2827,
+      "step": 1202
+    },
+    {
+      "epoch": 9.045112781954888,
+      "grad_norm": 0.216819886100415,
+      "learning_rate": 1.2375517672974325e-06,
+      "loss": 0.288,
+      "step": 1203
+    },
+    {
+      "epoch": 9.052631578947368,
+      "grad_norm": 0.20306933324570078,
+      "learning_rate": 1.236344438532905e-06,
+      "loss": 0.2937,
+      "step": 1204
+    },
+    {
+      "epoch": 9.06015037593985,
+      "grad_norm": 0.2153501806070825,
+      "learning_rate": 1.235136744769997e-06,
+      "loss": 0.2871,
+      "step": 1205
+    },
+    {
+      "epoch": 9.06766917293233,
+      "grad_norm": 0.2160024677501555,
+      "learning_rate": 1.2339286878738093e-06,
+      "loss": 0.2915,
+      "step": 1206
+    },
+    {
+      "epoch": 9.075187969924812,
+      "grad_norm": 0.20595481758928932,
+      "learning_rate": 1.232720269710004e-06,
+      "loss": 0.3011,
+      "step": 1207
+    },
+    {
+      "epoch": 9.082706766917294,
+      "grad_norm": 0.2058725880670359,
+      "learning_rate": 1.231511492144801e-06,
+      "loss": 0.3018,
+      "step": 1208
+    },
+    {
+      "epoch": 9.090225563909774,
+      "grad_norm": 0.20743416665240907,
+      "learning_rate": 1.2303023570449754e-06,
+      "loss": 0.2903,
+      "step": 1209
+    },
+    {
+      "epoch": 9.097744360902256,
+      "grad_norm": 0.21570675343367635,
+      "learning_rate": 1.2290928662778535e-06,
+      "loss": 0.3039,
+      "step": 1210
+    },
+    {
+      "epoch": 9.105263157894736,
+      "grad_norm": 0.19818653990361934,
+      "learning_rate": 1.227883021711312e-06,
+      "loss": 0.2963,
+      "step": 1211
+    },
+    {
+      "epoch": 9.112781954887218,
+      "grad_norm": 0.20226192333488752,
+      "learning_rate": 1.2266728252137732e-06,
+      "loss": 0.2968,
+      "step": 1212
+    },
+    {
+      "epoch": 9.1203007518797,
+      "grad_norm": 0.20216148854981156,
+      "learning_rate": 1.225462278654204e-06,
+      "loss": 0.2714,
+      "step": 1213
+    },
+    {
+      "epoch": 9.12781954887218,
+      "grad_norm": 0.20674685892029684,
+      "learning_rate": 1.2242513839021106e-06,
+      "loss": 0.2883,
+      "step": 1214
+    },
+    {
+      "epoch": 9.135338345864662,
+      "grad_norm": 0.204857914096769,
+      "learning_rate": 1.2230401428275382e-06,
+      "loss": 0.3028,
+      "step": 1215
+    },
+    {
+      "epoch": 9.142857142857142,
+      "grad_norm": 0.20009717392979123,
+      "learning_rate": 1.2218285573010652e-06,
+      "loss": 0.2906,
+      "step": 1216
+    },
+    {
+      "epoch": 9.150375939849624,
+      "grad_norm": 0.20526416303978723,
+      "learning_rate": 1.2206166291938036e-06,
+      "loss": 0.2948,
+      "step": 1217
+    },
+    {
+      "epoch": 9.157894736842104,
+      "grad_norm": 0.19601070459001904,
+      "learning_rate": 1.2194043603773935e-06,
+      "loss": 0.2976,
+      "step": 1218
+    },
+    {
+      "epoch": 9.165413533834586,
+      "grad_norm": 0.2264563932935685,
+      "learning_rate": 1.2181917527240018e-06,
+      "loss": 0.2966,
+      "step": 1219
+    },
+    {
+      "epoch": 9.172932330827068,
+      "grad_norm": 0.19113324019802203,
+      "learning_rate": 1.2169788081063178e-06,
+      "loss": 0.2844,
+      "step": 1220
+    },
+    {
+      "epoch": 9.180451127819548,
+      "grad_norm": 0.211509677278123,
+      "learning_rate": 1.2157655283975523e-06,
+      "loss": 0.3049,
+      "step": 1221
+    },
+    {
+      "epoch": 9.18796992481203,
+      "grad_norm": 0.2110290965417544,
+      "learning_rate": 1.2145519154714329e-06,
+      "loss": 0.2978,
+      "step": 1222
+    },
+    {
+      "epoch": 9.19548872180451,
+      "grad_norm": 0.20404434656216502,
+      "learning_rate": 1.2133379712022015e-06,
+      "loss": 0.2764,
+      "step": 1223
+    },
+    {
+      "epoch": 9.203007518796992,
+      "grad_norm": 0.2004238869425289,
+      "learning_rate": 1.2121236974646125e-06,
+      "loss": 0.2991,
+      "step": 1224
+    },
+    {
+      "epoch": 9.210526315789474,
+      "grad_norm": 0.19773886752575123,
+      "learning_rate": 1.210909096133929e-06,
+      "loss": 0.2869,
+      "step": 1225
+    },
+    {
+      "epoch": 9.218045112781954,
+      "grad_norm": 0.22479462784056872,
+      "learning_rate": 1.2096941690859192e-06,
+      "loss": 0.291,
+      "step": 1226
+    },
+    {
+      "epoch": 9.225563909774436,
+      "grad_norm": 0.20320432304062072,
+      "learning_rate": 1.2084789181968552e-06,
+      "loss": 0.2866,
+      "step": 1227
+    },
+    {
+      "epoch": 9.233082706766917,
+      "grad_norm": 0.20784724255417666,
+      "learning_rate": 1.2072633453435091e-06,
+      "loss": 0.2871,
+      "step": 1228
+    },
+    {
+      "epoch": 9.240601503759398,
+      "grad_norm": 0.200307655764377,
+      "learning_rate": 1.2060474524031497e-06,
+      "loss": 0.2872,
+      "step": 1229
+    },
+    {
+      "epoch": 9.24812030075188,
+      "grad_norm": 0.2213303175719757,
+      "learning_rate": 1.2048312412535407e-06,
+      "loss": 0.3036,
+      "step": 1230
+    },
+    {
+      "epoch": 9.25563909774436,
+      "grad_norm": 0.20873902594851224,
+      "learning_rate": 1.203614713772937e-06,
+      "loss": 0.2973,
+      "step": 1231
+    },
+    {
+      "epoch": 9.263157894736842,
+      "grad_norm": 0.2093641930085856,
+      "learning_rate": 1.2023978718400817e-06,
+      "loss": 0.2977,
+      "step": 1232
+    },
+    {
+      "epoch": 9.270676691729323,
+      "grad_norm": 0.21111118004097562,
+      "learning_rate": 1.2011807173342045e-06,
+      "loss": 0.2953,
+      "step": 1233
+    },
+    {
+      "epoch": 9.278195488721805,
+      "grad_norm": 0.2068744837836518,
+      "learning_rate": 1.1999632521350167e-06,
+      "loss": 0.2866,
+      "step": 1234
+    },
+    {
+      "epoch": 9.285714285714286,
+      "grad_norm": 0.2135794796373281,
+      "learning_rate": 1.19874547812271e-06,
+      "loss": 0.2913,
+      "step": 1235
+    },
+    {
+      "epoch": 9.293233082706767,
+      "grad_norm": 0.21057547191173726,
+      "learning_rate": 1.1975273971779527e-06,
+      "loss": 0.3049,
+      "step": 1236
+    },
+    {
+      "epoch": 9.300751879699249,
+      "grad_norm": 0.1943573864751774,
+      "learning_rate": 1.1963090111818877e-06,
+      "loss": 0.2837,
+      "step": 1237
+    },
+    {
+      "epoch": 9.308270676691729,
+      "grad_norm": 0.20520016798849328,
+      "learning_rate": 1.1950903220161284e-06,
+      "loss": 0.2854,
+      "step": 1238
+    },
+    {
+      "epoch": 9.31578947368421,
+      "grad_norm": 0.21645282705167362,
+      "learning_rate": 1.1938713315627564e-06,
+      "loss": 0.2926,
+      "step": 1239
+    },
+    {
+      "epoch": 9.323308270676693,
+      "grad_norm": 0.23182850654488063,
+      "learning_rate": 1.1926520417043194e-06,
+      "loss": 0.2988,
+      "step": 1240
+    },
+    {
+      "epoch": 9.330827067669173,
+      "grad_norm": 0.21155615895694405,
+      "learning_rate": 1.1914324543238265e-06,
+      "loss": 0.2883,
+      "step": 1241
+    },
+    {
+      "epoch": 9.338345864661655,
+      "grad_norm": 0.21222069479921965,
+      "learning_rate": 1.1902125713047466e-06,
+      "loss": 0.2927,
+      "step": 1242
+    },
+    {
+      "epoch": 9.345864661654135,
+      "grad_norm": 0.20006062281045214,
+      "learning_rate": 1.1889923945310057e-06,
+      "loss": 0.2953,
+      "step": 1243
+    },
+    {
+      "epoch": 9.353383458646617,
+      "grad_norm": 0.2079624950849432,
+      "learning_rate": 1.1877719258869824e-06,
+      "loss": 0.3003,
+      "step": 1244
+    },
+    {
+      "epoch": 9.360902255639097,
+      "grad_norm": 0.21296328642733298,
+      "learning_rate": 1.1865511672575073e-06,
+      "loss": 0.2885,
+      "step": 1245
+    },
+    {
+      "epoch": 9.368421052631579,
+      "grad_norm": 0.20694348938009086,
+      "learning_rate": 1.1853301205278577e-06,
+      "loss": 0.2912,
+      "step": 1246
+    },
+    {
+      "epoch": 9.37593984962406,
+      "grad_norm": 0.20329473357933436,
+      "learning_rate": 1.1841087875837565e-06,
+      "loss": 0.29,
+      "step": 1247
+    },
+    {
+      "epoch": 9.38345864661654,
+      "grad_norm": 0.1993785147821778,
+      "learning_rate": 1.1828871703113684e-06,
+      "loss": 0.2905,
+      "step": 1248
+    },
+    {
+      "epoch": 9.390977443609023,
+      "grad_norm": 0.2122347700982876,
+      "learning_rate": 1.1816652705972976e-06,
+      "loss": 0.3071,
+      "step": 1249
+    },
+    {
+      "epoch": 9.398496240601503,
+      "grad_norm": 0.2048009664598089,
+      "learning_rate": 1.1804430903285835e-06,
+      "loss": 0.282,
+      "step": 1250
+    },
+    {
+      "epoch": 9.406015037593985,
+      "grad_norm": 0.21044280882727723,
+      "learning_rate": 1.1792206313926998e-06,
+      "loss": 0.2967,
+      "step": 1251
+    },
+    {
+      "epoch": 9.413533834586467,
+      "grad_norm": 0.20679825127211796,
+      "learning_rate": 1.1779978956775504e-06,
+      "loss": 0.2896,
+      "step": 1252
+    },
+    {
+      "epoch": 9.421052631578947,
+      "grad_norm": 0.2481649978114574,
+      "learning_rate": 1.1767748850714658e-06,
+      "loss": 0.2854,
+      "step": 1253
+    },
+    {
+      "epoch": 9.428571428571429,
+      "grad_norm": 0.2034634567978017,
+      "learning_rate": 1.1755516014632022e-06,
+      "loss": 0.2896,
+      "step": 1254
+    },
+    {
+      "epoch": 9.436090225563909,
+      "grad_norm": 0.20860711648699784,
+      "learning_rate": 1.174328046741936e-06,
+      "loss": 0.2861,
+      "step": 1255
+    },
+    {
+      "epoch": 9.443609022556391,
+      "grad_norm": 0.21119100522887582,
+      "learning_rate": 1.1731042227972644e-06,
+      "loss": 0.2837,
+      "step": 1256
+    },
+    {
+      "epoch": 9.451127819548873,
+      "grad_norm": 0.21218647244421882,
+      "learning_rate": 1.171880131519198e-06,
+      "loss": 0.3026,
+      "step": 1257
+    },
+    {
+      "epoch": 9.458646616541353,
+      "grad_norm": 0.21401328950555387,
+      "learning_rate": 1.170655774798162e-06,
+      "loss": 0.3067,
+      "step": 1258
+    },
+    {
+      "epoch": 9.466165413533835,
+      "grad_norm": 0.2105399310018128,
+      "learning_rate": 1.1694311545249907e-06,
+      "loss": 0.305,
+      "step": 1259
+    },
+    {
+      "epoch": 9.473684210526315,
+      "grad_norm": 0.222298526771204,
+      "learning_rate": 1.1682062725909257e-06,
+      "loss": 0.3008,
+      "step": 1260
+    },
+    {
+      "epoch": 9.481203007518797,
+      "grad_norm": 0.19523626506260727,
+      "learning_rate": 1.1669811308876126e-06,
+      "loss": 0.2861,
+      "step": 1261
+    },
+    {
+      "epoch": 9.488721804511279,
+      "grad_norm": 0.2247998388218289,
+      "learning_rate": 1.1657557313070979e-06,
+      "loss": 0.2842,
+      "step": 1262
+    },
+    {
+      "epoch": 9.496240601503759,
+      "grad_norm": 0.22220096395001612,
+      "learning_rate": 1.164530075741827e-06,
+      "loss": 0.2947,
+      "step": 1263
+    },
+    {
+      "epoch": 9.503759398496241,
+      "grad_norm": 0.20950889510167556,
+      "learning_rate": 1.1633041660846404e-06,
+      "loss": 0.3019,
+      "step": 1264
+    },
+    {
+      "epoch": 9.511278195488721,
+      "grad_norm": 0.21405565144827812,
+      "learning_rate": 1.1620780042287704e-06,
+      "loss": 0.299,
+      "step": 1265
+    },
+    {
+      "epoch": 9.518796992481203,
+      "grad_norm": 0.2011973062699592,
+      "learning_rate": 1.1608515920678396e-06,
+      "loss": 0.2781,
+      "step": 1266
+    },
+    {
+      "epoch": 9.526315789473685,
+      "grad_norm": 0.22152135688870034,
+      "learning_rate": 1.1596249314958571e-06,
+      "loss": 0.3034,
+      "step": 1267
+    },
+    {
+      "epoch": 9.533834586466165,
+      "grad_norm": 0.21322051125075853,
+      "learning_rate": 1.158398024407215e-06,
+      "loss": 0.2954,
+      "step": 1268
+    },
+    {
+      "epoch": 9.541353383458647,
+      "grad_norm": 0.20131023360643577,
+      "learning_rate": 1.1571708726966862e-06,
+      "loss": 0.3038,
+      "step": 1269
+    },
+    {
+      "epoch": 9.548872180451127,
+      "grad_norm": 0.2177970335955727,
+      "learning_rate": 1.1559434782594222e-06,
+      "loss": 0.2898,
+      "step": 1270
+    },
+    {
+      "epoch": 9.556390977443609,
+      "grad_norm": 0.20711929843861174,
+      "learning_rate": 1.1547158429909485e-06,
+      "loss": 0.2899,
+      "step": 1271
+    },
+    {
+      "epoch": 9.563909774436091,
+      "grad_norm": 0.2146218056483543,
+      "learning_rate": 1.1534879687871628e-06,
+      "loss": 0.2952,
+      "step": 1272
+    },
+    {
+      "epoch": 9.571428571428571,
+      "grad_norm": 0.2147647389506873,
+      "learning_rate": 1.152259857544332e-06,
+      "loss": 0.3091,
+      "step": 1273
+    },
+    {
+      "epoch": 9.578947368421053,
+      "grad_norm": 0.21870099453247513,
+      "learning_rate": 1.151031511159089e-06,
+      "loss": 0.3009,
+      "step": 1274
+    },
+    {
+      "epoch": 9.586466165413533,
+      "grad_norm": 0.21182703150935986,
+      "learning_rate": 1.1498029315284293e-06,
+      "loss": 0.2805,
+      "step": 1275
+    },
+    {
+      "epoch": 9.593984962406015,
+      "grad_norm": 0.20799873871137411,
+      "learning_rate": 1.1485741205497092e-06,
+      "loss": 0.2829,
+      "step": 1276
+    },
+    {
+      "epoch": 9.601503759398497,
+      "grad_norm": 0.19870013857909624,
+      "learning_rate": 1.1473450801206425e-06,
+      "loss": 0.2868,
+      "step": 1277
+    },
+    {
+      "epoch": 9.609022556390977,
+      "grad_norm": 0.2080442489140835,
+      "learning_rate": 1.146115812139297e-06,
+      "loss": 0.284,
+      "step": 1278
+    },
+    {
+      "epoch": 9.61654135338346,
+      "grad_norm": 0.222528555556549,
+      "learning_rate": 1.1448863185040915e-06,
+      "loss": 0.3032,
+      "step": 1279
+    },
+    {
+      "epoch": 9.62406015037594,
+      "grad_norm": 0.21305204333455835,
+      "learning_rate": 1.1436566011137938e-06,
+      "loss": 0.2914,
+      "step": 1280
+    },
+    {
+      "epoch": 9.631578947368421,
+      "grad_norm": 0.20443274884099732,
+      "learning_rate": 1.142426661867517e-06,
+      "loss": 0.2925,
+      "step": 1281
+    },
+    {
+      "epoch": 9.639097744360903,
+      "grad_norm": 0.21954855338035528,
+      "learning_rate": 1.1411965026647174e-06,
+      "loss": 0.2922,
+      "step": 1282
+    },
+    {
+      "epoch": 9.646616541353383,
+      "grad_norm": 0.20152175879482276,
+      "learning_rate": 1.1399661254051904e-06,
+      "loss": 0.2984,
+      "step": 1283
+    },
+    {
+      "epoch": 9.654135338345865,
+      "grad_norm": 0.2262785038302715,
+      "learning_rate": 1.1387355319890683e-06,
+      "loss": 0.3162,
+      "step": 1284
+    },
+    {
+      "epoch": 9.661654135338345,
+      "grad_norm": 0.21538979145757237,
+      "learning_rate": 1.1375047243168171e-06,
+      "loss": 0.3002,
+      "step": 1285
+    },
+    {
+      "epoch": 9.669172932330827,
+      "grad_norm": 0.20722399717716364,
+      "learning_rate": 1.1362737042892342e-06,
+      "loss": 0.2876,
+      "step": 1286
+    },
+    {
+      "epoch": 9.676691729323307,
+      "grad_norm": 0.20584697482813563,
+      "learning_rate": 1.135042473807444e-06,
+      "loss": 0.2946,
+      "step": 1287
+    },
+    {
+      "epoch": 9.68421052631579,
+      "grad_norm": 0.20169238671856968,
+      "learning_rate": 1.133811034772897e-06,
+      "loss": 0.2736,
+      "step": 1288
+    },
+    {
+      "epoch": 9.691729323308271,
+      "grad_norm": 0.2154546252948222,
+      "learning_rate": 1.1325793890873652e-06,
+      "loss": 0.303,
+      "step": 1289
+    },
+    {
+      "epoch": 9.699248120300751,
+      "grad_norm": 0.2023084403153331,
+      "learning_rate": 1.13134753865294e-06,
+      "loss": 0.2919,
+      "step": 1290
+    },
+    {
+      "epoch": 9.706766917293233,
+      "grad_norm": 0.20566328379347343,
+      "learning_rate": 1.130115485372028e-06,
+      "loss": 0.2928,
+      "step": 1291
+    },
+    {
+      "epoch": 9.714285714285714,
+      "grad_norm": 0.2159634360656206,
+      "learning_rate": 1.1288832311473506e-06,
+      "loss": 0.2946,
+      "step": 1292
+    },
+    {
+      "epoch": 9.721804511278195,
+      "grad_norm": 0.20514336395116092,
+      "learning_rate": 1.1276507778819388e-06,
+      "loss": 0.2979,
+      "step": 1293
+    },
+    {
+      "epoch": 9.729323308270677,
+      "grad_norm": 0.19954061656769959,
+      "learning_rate": 1.1264181274791309e-06,
+      "loss": 0.2884,
+      "step": 1294
+    },
+    {
+      "epoch": 9.736842105263158,
+      "grad_norm": 0.20182414462038462,
+      "learning_rate": 1.1251852818425696e-06,
+      "loss": 0.292,
+      "step": 1295
+    },
+    {
+      "epoch": 9.74436090225564,
+      "grad_norm": 0.20552503910723252,
+      "learning_rate": 1.1239522428761994e-06,
+      "loss": 0.3169,
+      "step": 1296
+    },
+    {
+      "epoch": 9.75187969924812,
+      "grad_norm": 0.22362299588159676,
+      "learning_rate": 1.1227190124842631e-06,
+      "loss": 0.2982,
+      "step": 1297
+    },
+    {
+      "epoch": 9.759398496240602,
+      "grad_norm": 0.19318679265868963,
+      "learning_rate": 1.1214855925712996e-06,
+      "loss": 0.2953,
+      "step": 1298
+    },
+    {
+      "epoch": 9.766917293233083,
+      "grad_norm": 0.2122816557067997,
+      "learning_rate": 1.1202519850421398e-06,
+      "loss": 0.2868,
+      "step": 1299
+    },
+    {
+      "epoch": 9.774436090225564,
+      "grad_norm": 0.207770375568432,
+      "learning_rate": 1.1190181918019048e-06,
+      "loss": 0.2937,
+      "step": 1300
+    },
+    {
+      "epoch": 9.781954887218046,
+      "grad_norm": 0.21288960837948287,
+      "learning_rate": 1.1177842147560024e-06,
+      "loss": 0.2881,
+      "step": 1301
+    },
+    {
+      "epoch": 9.789473684210526,
+      "grad_norm": 0.21149188467579766,
+      "learning_rate": 1.116550055810124e-06,
+      "loss": 0.2937,
+      "step": 1302
+    },
+    {
+      "epoch": 9.796992481203008,
+      "grad_norm": 0.20562112179659087,
+      "learning_rate": 1.1153157168702427e-06,
+      "loss": 0.2941,
+      "step": 1303
+    },
+    {
+      "epoch": 9.80451127819549,
+      "grad_norm": 0.2130025093887851,
+      "learning_rate": 1.1140811998426088e-06,
+      "loss": 0.3109,
+      "step": 1304
+    },
+    {
+      "epoch": 9.81203007518797,
+      "grad_norm": 0.1974239185202293,
+      "learning_rate": 1.1128465066337476e-06,
+      "loss": 0.275,
+      "step": 1305
+    },
+    {
+      "epoch": 9.819548872180452,
+      "grad_norm": 0.23241056268204668,
+      "learning_rate": 1.111611639150457e-06,
+      "loss": 0.303,
+      "step": 1306
+    },
+    {
+      "epoch": 9.827067669172932,
+      "grad_norm": 0.1978364564248463,
+      "learning_rate": 1.1103765992998038e-06,
+      "loss": 0.2967,
+      "step": 1307
+    },
+    {
+      "epoch": 9.834586466165414,
+      "grad_norm": 0.20058171867243593,
+      "learning_rate": 1.109141388989121e-06,
+      "loss": 0.2852,
+      "step": 1308
+    },
+    {
+      "epoch": 9.842105263157894,
+      "grad_norm": 0.19777128762345644,
+      "learning_rate": 1.1079060101260046e-06,
+      "loss": 0.2749,
+      "step": 1309
+    },
+    {
+      "epoch": 9.849624060150376,
+      "grad_norm": 0.20517047210415504,
+      "learning_rate": 1.1066704646183115e-06,
+      "loss": 0.2945,
+      "step": 1310
+    },
+    {
+      "epoch": 9.857142857142858,
+      "grad_norm": 0.20212224538654983,
+      "learning_rate": 1.1054347543741555e-06,
+      "loss": 0.2912,
+      "step": 1311
+    },
+    {
+      "epoch": 9.864661654135338,
+      "grad_norm": 0.20430181454132004,
+      "learning_rate": 1.104198881301905e-06,
+      "loss": 0.2943,
+      "step": 1312
+    },
+    {
+      "epoch": 9.87218045112782,
+      "grad_norm": 0.20147539276129348,
+      "learning_rate": 1.1029628473101795e-06,
+      "loss": 0.2964,
+      "step": 1313
+    },
+    {
+      "epoch": 9.8796992481203,
+      "grad_norm": 0.22022086286976436,
+      "learning_rate": 1.1017266543078474e-06,
+      "loss": 0.2936,
+      "step": 1314
+    },
+    {
+      "epoch": 9.887218045112782,
+      "grad_norm": 0.206256996152521,
+      "learning_rate": 1.1004903042040226e-06,
+      "loss": 0.2815,
+      "step": 1315
+    },
+    {
+      "epoch": 9.894736842105264,
+      "grad_norm": 0.20385926015079842,
+      "learning_rate": 1.0992537989080618e-06,
+      "loss": 0.2968,
+      "step": 1316
+    },
+    {
+      "epoch": 9.902255639097744,
+      "grad_norm": 0.219457697426562,
+      "learning_rate": 1.0980171403295609e-06,
+      "loss": 0.2957,
+      "step": 1317
+    },
+    {
+      "epoch": 9.909774436090226,
+      "grad_norm": 0.2044164828855926,
+      "learning_rate": 1.0967803303783523e-06,
+      "loss": 0.2948,
+      "step": 1318
+    },
+    {
+      "epoch": 9.917293233082706,
+      "grad_norm": 0.20327127812043458,
+      "learning_rate": 1.0955433709645036e-06,
+      "loss": 0.2832,
+      "step": 1319
+    },
+    {
+      "epoch": 9.924812030075188,
+      "grad_norm": 0.20374552894938844,
+      "learning_rate": 1.0943062639983118e-06,
+      "loss": 0.3073,
+      "step": 1320
+    },
+    {
+      "epoch": 9.93233082706767,
+      "grad_norm": 0.20013096678185213,
+      "learning_rate": 1.0930690113903023e-06,
+      "loss": 0.2964,
+      "step": 1321
+    },
+    {
+      "epoch": 9.93984962406015,
+      "grad_norm": 0.2139161112124275,
+      "learning_rate": 1.0918316150512255e-06,
+      "loss": 0.2908,
+      "step": 1322
+    },
+    {
+      "epoch": 9.947368421052632,
+      "grad_norm": 0.21204548261306833,
+      "learning_rate": 1.0905940768920533e-06,
+      "loss": 0.2948,
+      "step": 1323
+    },
+    {
+      "epoch": 9.954887218045112,
+      "grad_norm": 0.20090141205074505,
+      "learning_rate": 1.089356398823977e-06,
+      "loss": 0.2914,
+      "step": 1324
+    },
+    {
+      "epoch": 9.962406015037594,
+      "grad_norm": 0.20523312653674097,
+      "learning_rate": 1.0881185827584044e-06,
+      "loss": 0.292,
+      "step": 1325
+    },
+    {
+      "epoch": 9.969924812030076,
+      "grad_norm": 0.204981893521801,
+      "learning_rate": 1.0868806306069554e-06,
+      "loss": 0.2871,
+      "step": 1326
+    },
+    {
+      "epoch": 9.977443609022556,
+      "grad_norm": 0.2306680853957588,
+      "learning_rate": 1.0856425442814608e-06,
+      "loss": 0.3053,
+      "step": 1327
+    },
+    {
+      "epoch": 9.984962406015038,
+      "grad_norm": 0.19237219372434447,
+      "learning_rate": 1.0844043256939583e-06,
+      "loss": 0.2892,
+      "step": 1328
+    },
+    {
+      "epoch": 9.992481203007518,
+      "grad_norm": 0.21871660332888584,
+      "learning_rate": 1.0831659767566902e-06,
+      "loss": 0.2906,
+      "step": 1329
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.2306888984101785,
+      "learning_rate": 1.0819274993820996e-06,
+      "loss": 0.3019,
+      "step": 1330
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 0.4153304994106293,
+      "eval_runtime": 1.7207,
+      "eval_samples_per_second": 190.038,
+      "eval_steps_per_second": 3.487,
+      "step": 1330
+    },
+    {
+      "epoch": 10.007518796992482,
+      "grad_norm": 0.19681170432842926,
+      "learning_rate": 1.080688895482829e-06,
+      "loss": 0.2941,
+      "step": 1331
+    },
+    {
+      "epoch": 10.015037593984962,
+      "grad_norm": 0.19174892952316389,
+      "learning_rate": 1.0794501669717144e-06,
+      "loss": 0.2905,
+      "step": 1332
+    },
+    {
+      "epoch": 10.022556390977444,
+      "grad_norm": 0.21885064321487846,
+      "learning_rate": 1.078211315761786e-06,
+      "loss": 0.287,
+      "step": 1333
+    },
+    {
+      "epoch": 10.030075187969924,
+      "grad_norm": 0.22259914138842915,
+      "learning_rate": 1.0769723437662628e-06,
+      "loss": 0.2865,
+      "step": 1334
+    },
+    {
+      "epoch": 10.037593984962406,
+      "grad_norm": 0.20904390697074923,
+      "learning_rate": 1.0757332528985504e-06,
+      "loss": 0.2834,
+      "step": 1335
+    },
+    {
+      "epoch": 10.045112781954888,
+      "grad_norm": 0.21368269162948408,
+      "learning_rate": 1.0744940450722377e-06,
+      "loss": 0.2924,
+      "step": 1336
+    },
+    {
+      "epoch": 10.052631578947368,
+      "grad_norm": 0.19896890538067222,
+      "learning_rate": 1.0732547222010948e-06,
+      "loss": 0.2705,
+      "step": 1337
+    },
+    {
+      "epoch": 10.06015037593985,
+      "grad_norm": 0.1986898849007135,
+      "learning_rate": 1.0720152861990693e-06,
+      "loss": 0.2832,
+      "step": 1338
+    },
+    {
+      "epoch": 10.06766917293233,
+      "grad_norm": 0.20607906641604004,
+      "learning_rate": 1.0707757389802831e-06,
+      "loss": 0.286,
+      "step": 1339
+    },
+    {
+      "epoch": 10.075187969924812,
+      "grad_norm": 0.20487883770670406,
+      "learning_rate": 1.0695360824590304e-06,
+      "loss": 0.2782,
+      "step": 1340
+    },
+    {
+      "epoch": 10.082706766917294,
+      "grad_norm": 0.21600836394446693,
+      "learning_rate": 1.0682963185497735e-06,
+      "loss": 0.2947,
+      "step": 1341
+    },
+    {
+      "epoch": 10.090225563909774,
+      "grad_norm": 0.21390386758339744,
+      "learning_rate": 1.0670564491671414e-06,
+      "loss": 0.2897,
+      "step": 1342
+    },
+    {
+      "epoch": 10.097744360902256,
+      "grad_norm": 0.21759058021167488,
+      "learning_rate": 1.0658164762259256e-06,
+      "loss": 0.2963,
+      "step": 1343
+    },
+    {
+      "epoch": 10.105263157894736,
+      "grad_norm": 0.21034778639341328,
+      "learning_rate": 1.0645764016410775e-06,
+      "loss": 0.2825,
+      "step": 1344
+    },
+    {
+      "epoch": 10.112781954887218,
+      "grad_norm": 0.20838687285521432,
+      "learning_rate": 1.0633362273277048e-06,
+      "loss": 0.2892,
+      "step": 1345
+    },
+    {
+      "epoch": 10.1203007518797,
+      "grad_norm": 0.20129502760961798,
+      "learning_rate": 1.0620959552010708e-06,
+      "loss": 0.2901,
+      "step": 1346
+    },
+    {
+      "epoch": 10.12781954887218,
+      "grad_norm": 0.20833903529427383,
+      "learning_rate": 1.0608555871765888e-06,
+      "loss": 0.2917,
+      "step": 1347
+    },
+    {
+      "epoch": 10.135338345864662,
+      "grad_norm": 0.19982329946055044,
+      "learning_rate": 1.0596151251698198e-06,
+      "loss": 0.3042,
+      "step": 1348
+    },
+    {
+      "epoch": 10.142857142857142,
+      "grad_norm": 0.20499132053482846,
+      "learning_rate": 1.0583745710964712e-06,
+      "loss": 0.2904,
+      "step": 1349
+    },
+    {
+      "epoch": 10.150375939849624,
+      "grad_norm": 0.2134425027858943,
+      "learning_rate": 1.0571339268723913e-06,
+      "loss": 0.2804,
+      "step": 1350
+    },
+    {
+      "epoch": 10.157894736842104,
+      "grad_norm": 0.22087182424746496,
+      "learning_rate": 1.0558931944135685e-06,
+      "loss": 0.2975,
+      "step": 1351
+    },
+    {
+      "epoch": 10.165413533834586,
+      "grad_norm": 0.2002737859845604,
+      "learning_rate": 1.054652375636127e-06,
+      "loss": 0.2875,
+      "step": 1352
+    },
+    {
+      "epoch": 10.172932330827068,
+      "grad_norm": 0.19848510506214584,
+      "learning_rate": 1.0534114724563249e-06,
+      "loss": 0.2856,
+      "step": 1353
+    },
+    {
+      "epoch": 10.180451127819548,
+      "grad_norm": 0.21128328338096053,
+      "learning_rate": 1.0521704867905493e-06,
+      "loss": 0.2898,
+      "step": 1354
+    },
+    {
+      "epoch": 10.18796992481203,
+      "grad_norm": 0.20547163169468866,
+      "learning_rate": 1.0509294205553167e-06,
+      "loss": 0.2957,
+      "step": 1355
+    },
+    {
+      "epoch": 10.19548872180451,
+      "grad_norm": 0.22037537178725775,
+      "learning_rate": 1.0496882756672665e-06,
+      "loss": 0.2864,
+      "step": 1356
+    },
+    {
+      "epoch": 10.203007518796992,
+      "grad_norm": 0.20577836834339372,
+      "learning_rate": 1.04844705404316e-06,
+      "loss": 0.284,
+      "step": 1357
+    },
+    {
+      "epoch": 10.210526315789474,
+      "grad_norm": 0.20858289305359118,
+      "learning_rate": 1.047205757599877e-06,
+      "loss": 0.2891,
+      "step": 1358
+    },
+    {
+      "epoch": 10.218045112781954,
+      "grad_norm": 0.21246624884493964,
+      "learning_rate": 1.0459643882544125e-06,
+      "loss": 0.2752,
+      "step": 1359
+    },
+    {
+      "epoch": 10.225563909774436,
+      "grad_norm": 0.20918535997467164,
+      "learning_rate": 1.0447229479238748e-06,
+      "loss": 0.2857,
+      "step": 1360
+    },
+    {
+      "epoch": 10.233082706766917,
+      "grad_norm": 0.2070618802907519,
+      "learning_rate": 1.0434814385254815e-06,
+      "loss": 0.2873,
+      "step": 1361
+    },
+    {
+      "epoch": 10.240601503759398,
+      "grad_norm": 0.20398002396966958,
+      "learning_rate": 1.0422398619765568e-06,
+      "loss": 0.2843,
+      "step": 1362
+    },
+    {
+      "epoch": 10.24812030075188,
+      "grad_norm": 0.19803409048799472,
+      "learning_rate": 1.0409982201945287e-06,
+      "loss": 0.2731,
+      "step": 1363
+    },
+    {
+      "epoch": 10.25563909774436,
+      "grad_norm": 0.20748150239635188,
+      "learning_rate": 1.0397565150969259e-06,
+      "loss": 0.2816,
+      "step": 1364
+    },
+    {
+      "epoch": 10.263157894736842,
+      "grad_norm": 0.21960847800484354,
+      "learning_rate": 1.0385147486013746e-06,
+      "loss": 0.2891,
+      "step": 1365
+    },
+    {
+      "epoch": 10.270676691729323,
+      "grad_norm": 0.20392991684098832,
+      "learning_rate": 1.0372729226255961e-06,
+      "loss": 0.2843,
+      "step": 1366
+    },
+    {
+      "epoch": 10.278195488721805,
+      "grad_norm": 0.20731842158232597,
+      "learning_rate": 1.0360310390874038e-06,
+      "loss": 0.2872,
+      "step": 1367
+    },
+    {
+      "epoch": 10.285714285714286,
+      "grad_norm": 0.20970140856266484,
+      "learning_rate": 1.0347890999046998e-06,
+      "loss": 0.2949,
+      "step": 1368
+    },
+    {
+      "epoch": 10.293233082706767,
+      "grad_norm": 0.22156853983323596,
+      "learning_rate": 1.0335471069954716e-06,
+      "loss": 0.2865,
+      "step": 1369
+    },
+    {
+      "epoch": 10.300751879699249,
+      "grad_norm": 0.205217130412181,
+      "learning_rate": 1.0323050622777904e-06,
+      "loss": 0.2885,
+      "step": 1370
+    },
+    {
+      "epoch": 10.308270676691729,
+      "grad_norm": 0.21695776921209273,
+      "learning_rate": 1.0310629676698072e-06,
+      "loss": 0.3047,
+      "step": 1371
+    },
+    {
+      "epoch": 10.31578947368421,
+      "grad_norm": 0.21294742472192352,
+      "learning_rate": 1.0298208250897503e-06,
+      "loss": 0.2909,
+      "step": 1372
+    },
+    {
+      "epoch": 10.323308270676693,
+      "grad_norm": 0.20846907733258346,
+      "learning_rate": 1.0285786364559214e-06,
+      "loss": 0.2873,
+      "step": 1373
+    },
+    {
+      "epoch": 10.330827067669173,
+      "grad_norm": 0.20705543799034337,
+      "learning_rate": 1.0273364036866938e-06,
+      "loss": 0.2847,
+      "step": 1374
+    },
+    {
+      "epoch": 10.338345864661655,
+      "grad_norm": 0.20648571233242763,
+      "learning_rate": 1.0260941287005086e-06,
+      "loss": 0.2826,
+      "step": 1375
+    },
+    {
+      "epoch": 10.345864661654135,
+      "grad_norm": 0.19457391952489667,
+      "learning_rate": 1.024851813415873e-06,
+      "loss": 0.2882,
+      "step": 1376
+    },
+    {
+      "epoch": 10.353383458646617,
+      "grad_norm": 0.2187006367658191,
+      "learning_rate": 1.0236094597513554e-06,
+      "loss": 0.2884,
+      "step": 1377
+    },
+    {
+      "epoch": 10.360902255639097,
+      "grad_norm": 0.20437389284523438,
+      "learning_rate": 1.022367069625584e-06,
+      "loss": 0.296,
+      "step": 1378
+    },
+    {
+      "epoch": 10.368421052631579,
+      "grad_norm": 0.20551378167466136,
+      "learning_rate": 1.0211246449572428e-06,
+      "loss": 0.2925,
+      "step": 1379
+    },
+    {
+      "epoch": 10.37593984962406,
+      "grad_norm": 0.2075027228756993,
+      "learning_rate": 1.01988218766507e-06,
+      "loss": 0.2975,
+      "step": 1380
+    },
+    {
+      "epoch": 10.38345864661654,
+      "grad_norm": 0.21293346082496797,
+      "learning_rate": 1.0186396996678537e-06,
+      "loss": 0.2968,
+      "step": 1381
+    },
+    {
+      "epoch": 10.390977443609023,
+      "grad_norm": 0.22854691990879997,
+      "learning_rate": 1.017397182884429e-06,
+      "loss": 0.2937,
+      "step": 1382
+    },
+    {
+      "epoch": 10.398496240601503,
+      "grad_norm": 0.21472228803398524,
+      "learning_rate": 1.0161546392336767e-06,
+      "loss": 0.2885,
+      "step": 1383
+    },
+    {
+      "epoch": 10.406015037593985,
+      "grad_norm": 0.2001593663991696,
+      "learning_rate": 1.0149120706345174e-06,
+      "loss": 0.2951,
+      "step": 1384
+    },
+    {
+      "epoch": 10.413533834586467,
+      "grad_norm": 0.20130367861943538,
+      "learning_rate": 1.0136694790059115e-06,
+      "loss": 0.2967,
+      "step": 1385
+    },
+    {
+      "epoch": 10.421052631578947,
+      "grad_norm": 0.20067481811809082,
+      "learning_rate": 1.0124268662668544e-06,
+      "loss": 0.2854,
+      "step": 1386
+    },
+    {
+      "epoch": 10.428571428571429,
+      "grad_norm": 0.21133576848312008,
+      "learning_rate": 1.0111842343363745e-06,
+      "loss": 0.2915,
+      "step": 1387
+    },
+    {
+      "epoch": 10.436090225563909,
+      "grad_norm": 0.2076099438257019,
+      "learning_rate": 1.0099415851335297e-06,
+      "loss": 0.2951,
+      "step": 1388
+    },
+    {
+      "epoch": 10.443609022556391,
+      "grad_norm": 0.22003263499545722,
+      "learning_rate": 1.0086989205774042e-06,
+      "loss": 0.2964,
+      "step": 1389
+    },
+    {
+      "epoch": 10.451127819548873,
+      "grad_norm": 0.2171840193751207,
+      "learning_rate": 1.0074562425871065e-06,
+      "loss": 0.2874,
+      "step": 1390
+    },
+    {
+      "epoch": 10.458646616541353,
+      "grad_norm": 0.2030227812779697,
+      "learning_rate": 1.0062135530817653e-06,
+      "loss": 0.2888,
+      "step": 1391
+    },
+    {
+      "epoch": 10.466165413533835,
+      "grad_norm": 0.21105829774185922,
+      "learning_rate": 1.0049708539805272e-06,
+      "loss": 0.2885,
+      "step": 1392
+    },
+    {
+      "epoch": 10.473684210526315,
+      "grad_norm": 0.20766902033660467,
+      "learning_rate": 1.0037281472025543e-06,
+      "loss": 0.2904,
+      "step": 1393
+    },
+    {
+      "epoch": 10.481203007518797,
+      "grad_norm": 0.20486347880762315,
+      "learning_rate": 1.0024854346670194e-06,
+      "loss": 0.2829,
+      "step": 1394
+    },
+    {
+      "epoch": 10.488721804511279,
+      "grad_norm": 0.2242180960185622,
+      "learning_rate": 1.0012427182931054e-06,
+      "loss": 0.291,
+      "step": 1395
+    },
+    {
+      "epoch": 10.496240601503759,
+      "grad_norm": 0.20386092222137786,
+      "learning_rate": 1e-06,
+      "loss": 0.2849,
+      "step": 1396
+    },
+    {
+      "epoch": 10.503759398496241,
+      "grad_norm": 0.20637692671036934,
+      "learning_rate": 9.98757281706895e-07,
+      "loss": 0.2848,
+      "step": 1397
+    },
+    {
+      "epoch": 10.511278195488721,
+      "grad_norm": 0.21854416894009485,
+      "learning_rate": 9.975145653329805e-07,
+      "loss": 0.2917,
+      "step": 1398
+    },
+    {
+      "epoch": 10.518796992481203,
+      "grad_norm": 0.21493682165175318,
+      "learning_rate": 9.96271852797446e-07,
+      "loss": 0.2895,
+      "step": 1399
+    },
+    {
+      "epoch": 10.526315789473685,
+      "grad_norm": 0.218900479659177,
+      "learning_rate": 9.950291460194727e-07,
+      "loss": 0.2831,
+      "step": 1400
+    },
+    {
+      "epoch": 10.533834586466165,
+      "grad_norm": 0.21056620745751034,
+      "learning_rate": 9.93786446918235e-07,
+      "loss": 0.2899,
+      "step": 1401
+    },
+    {
+      "epoch": 10.541353383458647,
+      "grad_norm": 0.22643333681456726,
+      "learning_rate": 9.925437574128937e-07,
+      "loss": 0.2802,
+      "step": 1402
+    },
+    {
+      "epoch": 10.548872180451127,
+      "grad_norm": 0.21258250361244194,
+      "learning_rate": 9.913010794225959e-07,
+      "loss": 0.295,
+      "step": 1403
+    },
+    {
+      "epoch": 10.556390977443609,
+      "grad_norm": 0.22640185950043243,
+      "learning_rate": 9.900584148664704e-07,
+      "loss": 0.3121,
+      "step": 1404
+    },
+    {
+      "epoch": 10.563909774436091,
+      "grad_norm": 0.21887035697745727,
+      "learning_rate": 9.888157656636256e-07,
+      "loss": 0.2784,
+      "step": 1405
+    },
+    {
+      "epoch": 10.571428571428571,
+      "grad_norm": 0.20396893879005673,
+      "learning_rate": 9.875731337331457e-07,
+      "loss": 0.2887,
+      "step": 1406
+    },
+    {
+      "epoch": 10.578947368421053,
+      "grad_norm": 0.19690159439525456,
+      "learning_rate": 9.863305209940888e-07,
+      "loss": 0.2851,
+      "step": 1407
+    },
+    {
+      "epoch": 10.586466165413533,
+      "grad_norm": 0.215618961054342,
+      "learning_rate": 9.850879293654827e-07,
+      "loss": 0.3117,
+      "step": 1408
+    },
+    {
+      "epoch": 10.593984962406015,
+      "grad_norm": 0.21900360756849227,
+      "learning_rate": 9.838453607663236e-07,
+      "loss": 0.2967,
+      "step": 1409
+    },
+    {
+      "epoch": 10.601503759398497,
+      "grad_norm": 0.20408242300035567,
+      "learning_rate": 9.826028171155707e-07,
+      "loss": 0.2865,
+      "step": 1410
+    },
+    {
+      "epoch": 10.609022556390977,
+      "grad_norm": 0.20372895018573411,
+      "learning_rate": 9.813603003321464e-07,
+      "loss": 0.2896,
+      "step": 1411
+    },
+    {
+      "epoch": 10.61654135338346,
+      "grad_norm": 0.2247102965155153,
+      "learning_rate": 9.801178123349297e-07,
+      "loss": 0.2866,
+      "step": 1412
+    },
+    {
+      "epoch": 10.62406015037594,
+      "grad_norm": 0.20773964666913358,
+      "learning_rate": 9.788753550427573e-07,
+      "loss": 0.286,
+      "step": 1413
+    },
+    {
+      "epoch": 10.631578947368421,
+      "grad_norm": 0.20865663971885026,
+      "learning_rate": 9.776329303744159e-07,
+      "loss": 0.2975,
+      "step": 1414
+    },
+    {
+      "epoch": 10.639097744360903,
+      "grad_norm": 0.21813901578799585,
+      "learning_rate": 9.763905402486447e-07,
+      "loss": 0.2902,
+      "step": 1415
+    },
+    {
+      "epoch": 10.646616541353383,
+      "grad_norm": 0.21374016310154975,
+      "learning_rate": 9.751481865841267e-07,
+      "loss": 0.2878,
+      "step": 1416
+    },
+    {
+      "epoch": 10.654135338345865,
+      "grad_norm": 0.21357617902262344,
+      "learning_rate": 9.739058712994913e-07,
+      "loss": 0.2862,
+      "step": 1417
+    },
+    {
+      "epoch": 10.661654135338345,
+      "grad_norm": 0.20124939604647557,
+      "learning_rate": 9.726635963133062e-07,
+      "loss": 0.2848,
+      "step": 1418
+    },
+    {
+      "epoch": 10.669172932330827,
+      "grad_norm": 0.21067602118753498,
+      "learning_rate": 9.714213635440787e-07,
+      "loss": 0.2856,
+      "step": 1419
+    },
+    {
+      "epoch": 10.676691729323307,
+      "grad_norm": 0.2108741405626798,
+      "learning_rate": 9.701791749102494e-07,
+      "loss": 0.2964,
+      "step": 1420
+    },
+    {
+      "epoch": 10.68421052631579,
+      "grad_norm": 0.20953450724912287,
+      "learning_rate": 9.689370323301927e-07,
+      "loss": 0.2812,
+      "step": 1421
+    },
+    {
+      "epoch": 10.691729323308271,
+      "grad_norm": 0.21105082129493943,
+      "learning_rate": 9.676949377222095e-07,
+      "loss": 0.2874,
+      "step": 1422
+    },
+    {
+      "epoch": 10.699248120300751,
+      "grad_norm": 0.2119410087916358,
+      "learning_rate": 9.664528930045283e-07,
+      "loss": 0.2842,
+      "step": 1423
+    },
+    {
+      "epoch": 10.706766917293233,
+      "grad_norm": 0.21864898133833646,
+      "learning_rate": 9.652109000953006e-07,
+      "loss": 0.2826,
+      "step": 1424
+    },
+    {
+      "epoch": 10.714285714285714,
+      "grad_norm": 0.2068668727173531,
+      "learning_rate": 9.639689609125961e-07,
+      "loss": 0.2933,
+      "step": 1425
+    },
+    {
+      "epoch": 10.721804511278195,
+      "grad_norm": 0.20673359369102676,
+      "learning_rate": 9.627270773744042e-07,
+      "loss": 0.2861,
+      "step": 1426
+    },
+    {
+      "epoch": 10.729323308270677,
+      "grad_norm": 0.20702429593613236,
+      "learning_rate": 9.614852513986256e-07,
+      "loss": 0.2854,
+      "step": 1427
+    },
+    {
+      "epoch": 10.736842105263158,
+      "grad_norm": 0.22622870810483642,
+      "learning_rate": 9.602434849030745e-07,
+      "loss": 0.293,
+      "step": 1428
+    },
+    {
+      "epoch": 10.74436090225564,
+      "grad_norm": 0.2076636869005624,
+      "learning_rate": 9.590017798054712e-07,
+      "loss": 0.2983,
+      "step": 1429
+    },
+    {
+      "epoch": 10.75187969924812,
+      "grad_norm": 0.19654692587552539,
+      "learning_rate": 9.577601380234433e-07,
+      "loss": 0.2941,
+      "step": 1430
+    },
+    {
+      "epoch": 10.759398496240602,
+      "grad_norm": 0.20327339474429681,
+      "learning_rate": 9.565185614745184e-07,
+      "loss": 0.2925,
+      "step": 1431
+    },
+    {
+      "epoch": 10.766917293233083,
+      "grad_norm": 0.20927564452996533,
+      "learning_rate": 9.552770520761254e-07,
+      "loss": 0.2817,
+      "step": 1432
+    },
+    {
+      "epoch": 10.774436090225564,
+      "grad_norm": 0.2134007274720979,
+      "learning_rate": 9.540356117455876e-07,
+      "loss": 0.2924,
+      "step": 1433
+    },
+    {
+      "epoch": 10.781954887218046,
+      "grad_norm": 0.21073384792100897,
+      "learning_rate": 9.527942424001234e-07,
+      "loss": 0.2815,
+      "step": 1434
+    },
+    {
+      "epoch": 10.789473684210526,
+      "grad_norm": 0.21652183351195642,
+      "learning_rate": 9.5155294595684e-07,
+      "loss": 0.2872,
+      "step": 1435
+    },
+    {
+      "epoch": 10.796992481203008,
+      "grad_norm": 0.20030374105213078,
+      "learning_rate": 9.503117243327335e-07,
+      "loss": 0.2716,
+      "step": 1436
+    },
+    {
+      "epoch": 10.80451127819549,
+      "grad_norm": 0.20729748766786749,
+      "learning_rate": 9.490705794446829e-07,
+      "loss": 0.2792,
+      "step": 1437
+    },
+    {
+      "epoch": 10.81203007518797,
+      "grad_norm": 0.20440707531557242,
+      "learning_rate": 9.478295132094505e-07,
+      "loss": 0.289,
+      "step": 1438
+    },
+    {
+      "epoch": 10.819548872180452,
+      "grad_norm": 0.22064579072775056,
+      "learning_rate": 9.465885275436751e-07,
+      "loss": 0.2933,
+      "step": 1439
+    },
+    {
+      "epoch": 10.827067669172932,
+      "grad_norm": 0.2011164651778188,
+      "learning_rate": 9.453476243638729e-07,
+      "loss": 0.2832,
+      "step": 1440
+    },
+    {
+      "epoch": 10.834586466165414,
+      "grad_norm": 0.21781611428469067,
+      "learning_rate": 9.441068055864314e-07,
+      "loss": 0.2877,
+      "step": 1441
+    },
+    {
+      "epoch": 10.842105263157894,
+      "grad_norm": 0.21488010146521766,
+      "learning_rate": 9.428660731276088e-07,
+      "loss": 0.2799,
+      "step": 1442
+    },
+    {
+      "epoch": 10.849624060150376,
+      "grad_norm": 0.2211111712214731,
+      "learning_rate": 9.416254289035287e-07,
+      "loss": 0.2883,
+      "step": 1443
+    },
+    {
+      "epoch": 10.857142857142858,
+      "grad_norm": 0.20950500356468713,
+      "learning_rate": 9.403848748301802e-07,
+      "loss": 0.2845,
+      "step": 1444
+    },
+    {
+      "epoch": 10.864661654135338,
+      "grad_norm": 0.22084925014028886,
+      "learning_rate": 9.391444128234111e-07,
+      "loss": 0.2954,
+      "step": 1445
+    },
+    {
+      "epoch": 10.87218045112782,
+      "grad_norm": 0.2135462161289135,
+      "learning_rate": 9.37904044798929e-07,
+      "loss": 0.2988,
+      "step": 1446
+    },
+    {
+      "epoch": 10.8796992481203,
+      "grad_norm": 0.20229027696488924,
+      "learning_rate": 9.366637726722949e-07,
+      "loss": 0.2952,
+      "step": 1447
+    },
+    {
+      "epoch": 10.887218045112782,
+      "grad_norm": 0.21073221663969607,
+      "learning_rate": 9.354235983589227e-07,
+      "loss": 0.2791,
+      "step": 1448
+    },
+    {
+      "epoch": 10.894736842105264,
+      "grad_norm": 0.21453228509452357,
+      "learning_rate": 9.341835237740747e-07,
+      "loss": 0.2911,
+      "step": 1449
+    },
+    {
+      "epoch": 10.902255639097744,
+      "grad_norm": 0.21833502736175364,
+      "learning_rate": 9.329435508328585e-07,
+      "loss": 0.285,
+      "step": 1450
+    },
+    {
+      "epoch": 10.909774436090226,
+      "grad_norm": 0.21805203173251778,
+      "learning_rate": 9.317036814502267e-07,
+      "loss": 0.2939,
+      "step": 1451
+    },
+    {
+      "epoch": 10.917293233082706,
+      "grad_norm": 0.20790765987187484,
+      "learning_rate": 9.304639175409698e-07,
+      "loss": 0.2935,
+      "step": 1452
+    },
+    {
+      "epoch": 10.924812030075188,
+      "grad_norm": 0.21217414192544815,
+      "learning_rate": 9.292242610197171e-07,
+      "loss": 0.298,
+      "step": 1453
+    },
+    {
+      "epoch": 10.93233082706767,
+      "grad_norm": 0.21249019973381855,
+      "learning_rate": 9.279847138009307e-07,
+      "loss": 0.3,
+      "step": 1454
+    },
+    {
+      "epoch": 10.93984962406015,
+      "grad_norm": 0.20628355262329678,
+      "learning_rate": 9.267452777989054e-07,
+      "loss": 0.2976,
+      "step": 1455
+    },
+    {
+      "epoch": 10.947368421052632,
+      "grad_norm": 0.2192467704289311,
+      "learning_rate": 9.255059549277622e-07,
+      "loss": 0.2994,
+      "step": 1456
+    },
+    {
+      "epoch": 10.954887218045112,
+      "grad_norm": 0.21042163004208095,
+      "learning_rate": 9.2426674710145e-07,
+      "loss": 0.2932,
+      "step": 1457
+    },
+    {
+      "epoch": 10.962406015037594,
+      "grad_norm": 0.2036798039242205,
+      "learning_rate": 9.230276562337372e-07,
+      "loss": 0.2965,
+      "step": 1458
+    },
+    {
+      "epoch": 10.969924812030076,
+      "grad_norm": 0.20205376906695882,
+      "learning_rate": 9.217886842382142e-07,
+      "loss": 0.2898,
+      "step": 1459
+    },
+    {
+      "epoch": 10.977443609022556,
+      "grad_norm": 0.20880809075408568,
+      "learning_rate": 9.205498330282856e-07,
+      "loss": 0.2868,
+      "step": 1460
+    },
+    {
+      "epoch": 10.984962406015038,
+      "grad_norm": 0.204279179631734,
+      "learning_rate": 9.193111045171713e-07,
+      "loss": 0.2938,
+      "step": 1461
+    },
+    {
+      "epoch": 10.992481203007518,
+      "grad_norm": 0.2235524776870455,
+      "learning_rate": 9.180725006179001e-07,
+      "loss": 0.3057,
+      "step": 1462
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.20452108189217916,
+      "learning_rate": 9.168340232433098e-07,
+      "loss": 0.2877,
+      "step": 1463
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 0.41695505380630493,
+      "eval_runtime": 1.6622,
+      "eval_samples_per_second": 196.724,
+      "eval_steps_per_second": 3.61,
+      "step": 1463
+    },
+    {
+      "epoch": 11.007518796992482,
+      "grad_norm": 0.20330177790097773,
+      "learning_rate": 9.155956743060415e-07,
+      "loss": 0.2826,
+      "step": 1464
+    },
+    {
+      "epoch": 11.015037593984962,
+      "grad_norm": 0.22370121833962614,
+      "learning_rate": 9.143574557185393e-07,
+      "loss": 0.2898,
+      "step": 1465
+    },
+    {
+      "epoch": 11.022556390977444,
+      "grad_norm": 0.20610165239016257,
+      "learning_rate": 9.131193693930445e-07,
+      "loss": 0.287,
+      "step": 1466
+    },
+    {
+      "epoch": 11.030075187969924,
+      "grad_norm": 0.20924619623991986,
+      "learning_rate": 9.118814172415958e-07,
+      "loss": 0.2779,
+      "step": 1467
+    },
+    {
+      "epoch": 11.037593984962406,
+      "grad_norm": 0.20241334599962196,
+      "learning_rate": 9.106436011760228e-07,
+      "loss": 0.2827,
+      "step": 1468
+    },
+    {
+      "epoch": 11.045112781954888,
+      "grad_norm": 0.20043825350859878,
+      "learning_rate": 9.094059231079469e-07,
+      "loss": 0.2741,
+      "step": 1469
+    },
+    {
+      "epoch": 11.052631578947368,
+      "grad_norm": 0.19871409256179817,
+      "learning_rate": 9.081683849487744e-07,
+      "loss": 0.2813,
+      "step": 1470
+    },
+    {
+      "epoch": 11.06015037593985,
+      "grad_norm": 0.20367735919304716,
+      "learning_rate": 9.069309886096976e-07,
+      "loss": 0.2908,
+      "step": 1471
+    },
+    {
+      "epoch": 11.06766917293233,
+      "grad_norm": 0.20964076623404249,
+      "learning_rate": 9.056937360016879e-07,
+      "loss": 0.2877,
+      "step": 1472
+    },
+    {
+      "epoch": 11.075187969924812,
+      "grad_norm": 0.2079368551550497,
+      "learning_rate": 9.044566290354965e-07,
+      "loss": 0.2797,
+      "step": 1473
+    },
+    {
+      "epoch": 11.082706766917294,
+      "grad_norm": 0.20293605856873465,
+      "learning_rate": 9.032196696216475e-07,
+      "loss": 0.2952,
+      "step": 1474
+    },
+    {
+      "epoch": 11.090225563909774,
+      "grad_norm": 0.2122736484569499,
+      "learning_rate": 9.019828596704393e-07,
+      "loss": 0.2916,
+      "step": 1475
+    },
+    {
+      "epoch": 11.097744360902256,
+      "grad_norm": 0.21090239440099207,
+      "learning_rate": 9.007462010919385e-07,
+      "loss": 0.2932,
+      "step": 1476
+    },
+    {
+      "epoch": 11.105263157894736,
+      "grad_norm": 0.21376684547379807,
+      "learning_rate": 8.995096957959773e-07,
+      "loss": 0.2859,
+      "step": 1477
+    },
+    {
+      "epoch": 11.112781954887218,
+      "grad_norm": 0.21235190500428544,
+      "learning_rate": 8.982733456921529e-07,
+      "loss": 0.2748,
+      "step": 1478
+    },
+    {
+      "epoch": 11.1203007518797,
+      "grad_norm": 0.21460620566972763,
+      "learning_rate": 8.970371526898206e-07,
+      "loss": 0.2866,
+      "step": 1479
+    },
+    {
+      "epoch": 11.12781954887218,
+      "grad_norm": 0.23888862618146314,
+      "learning_rate": 8.958011186980952e-07,
+      "loss": 0.2933,
+      "step": 1480
+    },
+    {
+      "epoch": 11.135338345864662,
+      "grad_norm": 0.20824117644066678,
+      "learning_rate": 8.945652456258445e-07,
+      "loss": 0.3089,
+      "step": 1481
+    },
+    {
+      "epoch": 11.142857142857142,
+      "grad_norm": 0.20716238534601034,
+      "learning_rate": 8.933295353816887e-07,
+      "loss": 0.2808,
+      "step": 1482
+    },
+    {
+      "epoch": 11.150375939849624,
+      "grad_norm": 0.20005797883625837,
+      "learning_rate": 8.920939898739953e-07,
+      "loss": 0.2834,
+      "step": 1483
+    },
+    {
+      "epoch": 11.157894736842104,
+      "grad_norm": 0.18697954833639382,
+      "learning_rate": 8.908586110108793e-07,
+      "loss": 0.2742,
+      "step": 1484
+    },
+    {
+      "epoch": 11.165413533834586,
+      "grad_norm": 0.22247297972241645,
+      "learning_rate": 8.896234007001963e-07,
+      "loss": 0.2811,
+      "step": 1485
+    },
+    {
+      "epoch": 11.172932330827068,
+      "grad_norm": 0.20042086304375717,
+      "learning_rate": 8.883883608495433e-07,
+      "loss": 0.2882,
+      "step": 1486
+    },
+    {
+      "epoch": 11.180451127819548,
+      "grad_norm": 0.20552695511265298,
+      "learning_rate": 8.871534933662524e-07,
+      "loss": 0.2884,
+      "step": 1487
+    },
+    {
+      "epoch": 11.18796992481203,
+      "grad_norm": 0.20804294511531693,
+      "learning_rate": 8.859188001573915e-07,
+      "loss": 0.289,
+      "step": 1488
+    },
+    {
+      "epoch": 11.19548872180451,
+      "grad_norm": 0.21121290519096694,
+      "learning_rate": 8.846842831297572e-07,
+      "loss": 0.2793,
+      "step": 1489
+    },
+    {
+      "epoch": 11.203007518796992,
+      "grad_norm": 0.21572884627726377,
+      "learning_rate": 8.83449944189876e-07,
+      "loss": 0.2848,
+      "step": 1490
+    },
+    {
+      "epoch": 11.210526315789474,
+      "grad_norm": 0.2042382705667945,
+      "learning_rate": 8.822157852439976e-07,
+      "loss": 0.2787,
+      "step": 1491
+    },
+    {
+      "epoch": 11.218045112781954,
+      "grad_norm": 0.21451042929256958,
+      "learning_rate": 8.809818081980953e-07,
+      "loss": 0.2729,
+      "step": 1492
+    },
+    {
+      "epoch": 11.225563909774436,
+      "grad_norm": 0.20849016214360944,
+      "learning_rate": 8.7974801495786e-07,
+      "loss": 0.2821,
+      "step": 1493
+    },
+    {
+      "epoch": 11.233082706766917,
+      "grad_norm": 0.21261180376552338,
+      "learning_rate": 8.785144074287004e-07,
+      "loss": 0.2811,
+      "step": 1494
+    },
+    {
+      "epoch": 11.240601503759398,
+      "grad_norm": 0.22481423327589029,
+      "learning_rate": 8.772809875157366e-07,
+      "loss": 0.2968,
+      "step": 1495
+    },
+    {
+      "epoch": 11.24812030075188,
+      "grad_norm": 0.20973557311812768,
+      "learning_rate": 8.760477571238005e-07,
+      "loss": 0.2869,
+      "step": 1496
+    },
+    {
+      "epoch": 11.25563909774436,
+      "grad_norm": 0.21407224295152683,
+      "learning_rate": 8.748147181574302e-07,
+      "loss": 0.2884,
+      "step": 1497
+    },
+    {
+      "epoch": 11.263157894736842,
+      "grad_norm": 0.21574538098349355,
+      "learning_rate": 8.735818725208691e-07,
+      "loss": 0.2913,
+      "step": 1498
+    },
+    {
+      "epoch": 11.270676691729323,
+      "grad_norm": 0.20698091012696945,
+      "learning_rate": 8.72349222118061e-07,
+      "loss": 0.2847,
+      "step": 1499
+    },
+    {
+      "epoch": 11.278195488721805,
+      "grad_norm": 0.2020553003858107,
+      "learning_rate": 8.711167688526493e-07,
+      "loss": 0.2796,
+      "step": 1500
+    },
+    {
+      "epoch": 11.285714285714286,
+      "grad_norm": 0.21997740504705213,
+      "learning_rate": 8.698845146279719e-07,
+      "loss": 0.2917,
+      "step": 1501
+    },
+    {
+      "epoch": 11.293233082706767,
+      "grad_norm": 0.20295516905641686,
+      "learning_rate": 8.686524613470603e-07,
+      "loss": 0.2833,
+      "step": 1502
+    },
+    {
+      "epoch": 11.300751879699249,
+      "grad_norm": 0.2051684801268154,
+      "learning_rate": 8.67420610912635e-07,
+      "loss": 0.2816,
+      "step": 1503
+    },
+    {
+      "epoch": 11.308270676691729,
+      "grad_norm": 0.19842915205985956,
+      "learning_rate": 8.661889652271029e-07,
+      "loss": 0.2775,
+      "step": 1504
+    },
+    {
+      "epoch": 11.31578947368421,
+      "grad_norm": 0.2058266032411938,
+      "learning_rate": 8.649575261925562e-07,
+      "loss": 0.2773,
+      "step": 1505
+    },
+    {
+      "epoch": 11.323308270676693,
+      "grad_norm": 0.19352208628931772,
+      "learning_rate": 8.63726295710766e-07,
+      "loss": 0.2813,
+      "step": 1506
+    },
+    {
+      "epoch": 11.330827067669173,
+      "grad_norm": 0.2216826416225841,
+      "learning_rate": 8.624952756831831e-07,
+      "loss": 0.2889,
+      "step": 1507
+    },
+    {
+      "epoch": 11.338345864661655,
+      "grad_norm": 0.2443693492302327,
+      "learning_rate": 8.612644680109318e-07,
+      "loss": 0.2959,
+      "step": 1508
+    },
+    {
+      "epoch": 11.345864661654135,
+      "grad_norm": 0.19628601066558107,
+      "learning_rate": 8.600338745948098e-07,
+      "loss": 0.276,
+      "step": 1509
+    },
+    {
+      "epoch": 11.353383458646617,
+      "grad_norm": 0.20399416559526012,
+      "learning_rate": 8.588034973352826e-07,
+      "loss": 0.2831,
+      "step": 1510
+    },
+    {
+      "epoch": 11.360902255639097,
+      "grad_norm": 0.22924909160045553,
+      "learning_rate": 8.575733381324833e-07,
+      "loss": 0.2934,
+      "step": 1511
+    },
+    {
+      "epoch": 11.368421052631579,
+      "grad_norm": 0.19210013825332756,
+      "learning_rate": 8.563433988862064e-07,
+      "loss": 0.2873,
+      "step": 1512
+    },
+    {
+      "epoch": 11.37593984962406,
+      "grad_norm": 0.22650465904396883,
+      "learning_rate": 8.551136814959088e-07,
+      "loss": 0.2782,
+      "step": 1513
+    },
+    {
+      "epoch": 11.38345864661654,
+      "grad_norm": 0.21335856676255077,
+      "learning_rate": 8.53884187860703e-07,
+      "loss": 0.2786,
+      "step": 1514
+    },
+    {
+      "epoch": 11.390977443609023,
+      "grad_norm": 0.20277029617176404,
+      "learning_rate": 8.526549198793575e-07,
+      "loss": 0.2846,
+      "step": 1515
+    },
+    {
+      "epoch": 11.398496240601503,
+      "grad_norm": 0.22994186915966486,
+      "learning_rate": 8.514258794502904e-07,
+      "loss": 0.2851,
+      "step": 1516
+    },
+    {
+      "epoch": 11.406015037593985,
+      "grad_norm": 0.2006563409208994,
+      "learning_rate": 8.501970684715708e-07,
+      "loss": 0.2812,
+      "step": 1517
+    },
+    {
+      "epoch": 11.413533834586467,
+      "grad_norm": 0.20616430366309293,
+      "learning_rate": 8.48968488840911e-07,
+      "loss": 0.2696,
+      "step": 1518
+    },
+    {
+      "epoch": 11.421052631578947,
+      "grad_norm": 0.20199379833341313,
+      "learning_rate": 8.47740142455668e-07,
+      "loss": 0.2814,
+      "step": 1519
+    },
+    {
+      "epoch": 11.428571428571429,
+      "grad_norm": 0.19857019557614725,
+      "learning_rate": 8.46512031212837e-07,
+      "loss": 0.2913,
+      "step": 1520
+    },
+    {
+      "epoch": 11.436090225563909,
+      "grad_norm": 0.21233889211764867,
+      "learning_rate": 8.452841570090516e-07,
+      "loss": 0.2816,
+      "step": 1521
+    },
+    {
+      "epoch": 11.443609022556391,
+      "grad_norm": 0.21342220229550715,
+      "learning_rate": 8.440565217405778e-07,
+      "loss": 0.2947,
+      "step": 1522
+    },
+    {
+      "epoch": 11.451127819548873,
+      "grad_norm": 0.20378055129335573,
+      "learning_rate": 8.428291273033138e-07,
+      "loss": 0.2886,
+      "step": 1523
+    },
+    {
+      "epoch": 11.458646616541353,
+      "grad_norm": 0.2152378116506678,
+      "learning_rate": 8.41601975592785e-07,
+      "loss": 0.2711,
+      "step": 1524
+    },
+    {
+      "epoch": 11.466165413533835,
+      "grad_norm": 0.20474549408504025,
+      "learning_rate": 8.40375068504143e-07,
+      "loss": 0.2791,
+      "step": 1525
+    },
+    {
+      "epoch": 11.473684210526315,
+      "grad_norm": 0.220887880527946,
+      "learning_rate": 8.391484079321601e-07,
+      "loss": 0.302,
+      "step": 1526
+    },
+    {
+      "epoch": 11.481203007518797,
+      "grad_norm": 0.2148922646449907,
+      "learning_rate": 8.379219957712295e-07,
+      "loss": 0.2816,
+      "step": 1527
+    },
+    {
+      "epoch": 11.488721804511279,
+      "grad_norm": 0.20517892748188335,
+      "learning_rate": 8.366958339153598e-07,
+      "loss": 0.2816,
+      "step": 1528
+    },
+    {
+      "epoch": 11.496240601503759,
+      "grad_norm": 0.2066083543541821,
+      "learning_rate": 8.354699242581728e-07,
+      "loss": 0.288,
+      "step": 1529
+    },
+    {
+      "epoch": 11.503759398496241,
+      "grad_norm": 0.19852751566766777,
+      "learning_rate": 8.342442686929023e-07,
+      "loss": 0.2727,
+      "step": 1530
+    },
+    {
+      "epoch": 11.511278195488721,
+      "grad_norm": 0.20405738391539885,
+      "learning_rate": 8.330188691123876e-07,
+      "loss": 0.2843,
+      "step": 1531
+    },
+    {
+      "epoch": 11.518796992481203,
+      "grad_norm": 0.20744397086348126,
+      "learning_rate": 8.317937274090746e-07,
+      "loss": 0.2925,
+      "step": 1532
+    },
+    {
+      "epoch": 11.526315789473685,
+      "grad_norm": 0.21082546471713512,
+      "learning_rate": 8.305688454750094e-07,
+      "loss": 0.2792,
+      "step": 1533
+    },
+    {
+      "epoch": 11.533834586466165,
+      "grad_norm": 0.21135198043537126,
+      "learning_rate": 8.293442252018384e-07,
+      "loss": 0.293,
+      "step": 1534
+    },
+    {
+      "epoch": 11.541353383458647,
+      "grad_norm": 0.2098965769620484,
+      "learning_rate": 8.28119868480802e-07,
+      "loss": 0.2947,
+      "step": 1535
+    },
+    {
+      "epoch": 11.548872180451127,
+      "grad_norm": 0.21402140919978466,
+      "learning_rate": 8.26895777202736e-07,
+      "loss": 0.2841,
+      "step": 1536
+    },
+    {
+      "epoch": 11.556390977443609,
+      "grad_norm": 0.2030770644329813,
+      "learning_rate": 8.256719532580638e-07,
+      "loss": 0.2735,
+      "step": 1537
+    },
+    {
+      "epoch": 11.563909774436091,
+      "grad_norm": 0.2154114306925539,
+      "learning_rate": 8.244483985367982e-07,
+      "loss": 0.2848,
+      "step": 1538
+    },
+    {
+      "epoch": 11.571428571428571,
+      "grad_norm": 0.2035560763915741,
+      "learning_rate": 8.232251149285342e-07,
+      "loss": 0.2807,
+      "step": 1539
+    },
+    {
+      "epoch": 11.578947368421053,
+      "grad_norm": 0.2004917742996665,
+      "learning_rate": 8.220021043224499e-07,
+      "loss": 0.2781,
+      "step": 1540
+    },
+    {
+      "epoch": 11.586466165413533,
+      "grad_norm": 0.20119440683384618,
+      "learning_rate": 8.207793686072999e-07,
+      "loss": 0.2808,
+      "step": 1541
+    },
+    {
+      "epoch": 11.593984962406015,
+      "grad_norm": 0.21911972491970708,
+      "learning_rate": 8.195569096714166e-07,
+      "loss": 0.2835,
+      "step": 1542
+    },
+    {
+      "epoch": 11.601503759398497,
+      "grad_norm": 0.21386835108401422,
+      "learning_rate": 8.183347294027023e-07,
+      "loss": 0.2887,
+      "step": 1543
+    },
+    {
+      "epoch": 11.609022556390977,
+      "grad_norm": 0.21411304576414483,
+      "learning_rate": 8.171128296886315e-07,
+      "loss": 0.2857,
+      "step": 1544
+    },
+    {
+      "epoch": 11.61654135338346,
+      "grad_norm": 0.21546620542630665,
+      "learning_rate": 8.158912124162433e-07,
+      "loss": 0.2855,
+      "step": 1545
+    },
+    {
+      "epoch": 11.62406015037594,
+      "grad_norm": 0.2113164548999208,
+      "learning_rate": 8.146698794721423e-07,
+      "loss": 0.2944,
+      "step": 1546
+    },
+    {
+      "epoch": 11.631578947368421,
+      "grad_norm": 0.20835381926655233,
+      "learning_rate": 8.134488327424926e-07,
+      "loss": 0.282,
+      "step": 1547
+    },
+    {
+      "epoch": 11.639097744360903,
+      "grad_norm": 0.21233522528248863,
+      "learning_rate": 8.122280741130175e-07,
+      "loss": 0.2936,
+      "step": 1548
+    },
+    {
+      "epoch": 11.646616541353383,
+      "grad_norm": 0.2000278076002988,
+      "learning_rate": 8.110076054689942e-07,
+      "loss": 0.2987,
+      "step": 1549
+    },
+    {
+      "epoch": 11.654135338345865,
+      "grad_norm": 0.21193011661326266,
+      "learning_rate": 8.097874286952533e-07,
+      "loss": 0.2866,
+      "step": 1550
+    },
+    {
+      "epoch": 11.661654135338345,
+      "grad_norm": 0.20967790552507773,
+      "learning_rate": 8.085675456761734e-07,
+      "loss": 0.2895,
+      "step": 1551
+    },
+    {
+      "epoch": 11.669172932330827,
+      "grad_norm": 0.22296692219387382,
+      "learning_rate": 8.073479582956806e-07,
+      "loss": 0.2888,
+      "step": 1552
+    },
+    {
+      "epoch": 11.676691729323307,
+      "grad_norm": 0.2110383254490028,
+      "learning_rate": 8.061286684372434e-07,
+      "loss": 0.2732,
+      "step": 1553
+    },
+    {
+      "epoch": 11.68421052631579,
+      "grad_norm": 0.21218778596608678,
+      "learning_rate": 8.049096779838717e-07,
+      "loss": 0.2925,
+      "step": 1554
+    },
+    {
+      "epoch": 11.691729323308271,
+      "grad_norm": 0.2135804510360402,
+      "learning_rate": 8.036909888181127e-07,
+      "loss": 0.2972,
+      "step": 1555
+    },
+    {
+      "epoch": 11.699248120300751,
+      "grad_norm": 0.20170890259191038,
+      "learning_rate": 8.024726028220474e-07,
+      "loss": 0.2782,
+      "step": 1556
+    },
+    {
+      "epoch": 11.706766917293233,
+      "grad_norm": 0.21746949585362838,
+      "learning_rate": 8.012545218772904e-07,
+      "loss": 0.2901,
+      "step": 1557
+    },
+    {
+      "epoch": 11.714285714285714,
+      "grad_norm": 0.2013184500912784,
+      "learning_rate": 8.000367478649834e-07,
+      "loss": 0.2758,
+      "step": 1558
+    },
+    {
+      "epoch": 11.721804511278195,
+      "grad_norm": 0.2008479279515695,
+      "learning_rate": 7.988192826657958e-07,
+      "loss": 0.2857,
+      "step": 1559
+    },
+    {
+      "epoch": 11.729323308270677,
+      "grad_norm": 0.21546662747227516,
+      "learning_rate": 7.976021281599181e-07,
+      "loss": 0.2955,
+      "step": 1560
+    },
+    {
+      "epoch": 11.736842105263158,
+      "grad_norm": 0.21466030521247562,
+      "learning_rate": 7.963852862270633e-07,
+      "loss": 0.2893,
+      "step": 1561
+    },
+    {
+      "epoch": 11.74436090225564,
+      "grad_norm": 0.21177348183150663,
+      "learning_rate": 7.951687587464593e-07,
+      "loss": 0.291,
+      "step": 1562
+    },
+    {
+      "epoch": 11.75187969924812,
+      "grad_norm": 0.20119078776563148,
+      "learning_rate": 7.939525475968505e-07,
+      "loss": 0.2837,
+      "step": 1563
+    },
+    {
+      "epoch": 11.759398496240602,
+      "grad_norm": 0.21807103112674095,
+      "learning_rate": 7.92736654656491e-07,
+      "loss": 0.2865,
+      "step": 1564
+    },
+    {
+      "epoch": 11.766917293233083,
+      "grad_norm": 0.21682942025967444,
+      "learning_rate": 7.91521081803145e-07,
+      "loss": 0.2997,
+      "step": 1565
+    },
+    {
+      "epoch": 11.774436090225564,
+      "grad_norm": 0.19622705408802765,
+      "learning_rate": 7.903058309140808e-07,
+      "loss": 0.2802,
+      "step": 1566
+    },
+    {
+      "epoch": 11.781954887218046,
+      "grad_norm": 0.21334064014086848,
+      "learning_rate": 7.890909038660713e-07,
+      "loss": 0.285,
+      "step": 1567
+    },
+    {
+      "epoch": 11.789473684210526,
+      "grad_norm": 0.2123893047843262,
+      "learning_rate": 7.878763025353874e-07,
+      "loss": 0.2972,
+      "step": 1568
+    },
+    {
+      "epoch": 11.796992481203008,
+      "grad_norm": 0.2022537629464449,
+      "learning_rate": 7.866620287977987e-07,
+      "loss": 0.2887,
+      "step": 1569
+    },
+    {
+      "epoch": 11.80451127819549,
+      "grad_norm": 0.23259674326317048,
+      "learning_rate": 7.854480845285672e-07,
+      "loss": 0.2954,
+      "step": 1570
+    },
+    {
+      "epoch": 11.81203007518797,
+      "grad_norm": 0.2150437322912329,
+      "learning_rate": 7.842344716024477e-07,
+      "loss": 0.3,
+      "step": 1571
+    },
+    {
+      "epoch": 11.819548872180452,
+      "grad_norm": 0.2070738614968804,
+      "learning_rate": 7.830211918936819e-07,
+      "loss": 0.2867,
+      "step": 1572
+    },
+    {
+      "epoch": 11.827067669172932,
+      "grad_norm": 0.1981054932123341,
+      "learning_rate": 7.818082472759983e-07,
+      "loss": 0.2748,
+      "step": 1573
+    },
+    {
+      "epoch": 11.834586466165414,
+      "grad_norm": 0.20956182912529026,
+      "learning_rate": 7.805956396226062e-07,
+      "loss": 0.2935,
+      "step": 1574
+    },
+    {
+      "epoch": 11.842105263157894,
+      "grad_norm": 0.2094661687296418,
+      "learning_rate": 7.793833708061965e-07,
+      "loss": 0.2831,
+      "step": 1575
+    },
+    {
+      "epoch": 11.849624060150376,
+      "grad_norm": 0.19379975471879904,
+      "learning_rate": 7.781714426989345e-07,
+      "loss": 0.2725,
+      "step": 1576
+    },
+    {
+      "epoch": 11.857142857142858,
+      "grad_norm": 0.2256343445161518,
+      "learning_rate": 7.769598571724619e-07,
+      "loss": 0.2895,
+      "step": 1577
+    },
+    {
+      "epoch": 11.864661654135338,
+      "grad_norm": 0.20881003018358407,
+      "learning_rate": 7.75748616097889e-07,
+      "loss": 0.2858,
+      "step": 1578
+    },
+    {
+      "epoch": 11.87218045112782,
+      "grad_norm": 0.23917406596290777,
+      "learning_rate": 7.74537721345796e-07,
+      "loss": 0.2887,
+      "step": 1579
+    },
+    {
+      "epoch": 11.8796992481203,
+      "grad_norm": 0.1994444464072194,
+      "learning_rate": 7.733271747862264e-07,
+      "loss": 0.2877,
+      "step": 1580
+    },
+    {
+      "epoch": 11.887218045112782,
+      "grad_norm": 0.19518175854479017,
+      "learning_rate": 7.72116978288688e-07,
+      "loss": 0.2729,
+      "step": 1581
+    },
+    {
+      "epoch": 11.894736842105264,
+      "grad_norm": 0.2055302059597974,
+      "learning_rate": 7.709071337221468e-07,
+      "loss": 0.29,
+      "step": 1582
+    },
+    {
+      "epoch": 11.902255639097744,
+      "grad_norm": 0.20635549882054927,
+      "learning_rate": 7.696976429550247e-07,
+      "loss": 0.2933,
+      "step": 1583
+    },
+    {
+      "epoch": 11.909774436090226,
+      "grad_norm": 0.23245531875951023,
+      "learning_rate": 7.68488507855199e-07,
+      "loss": 0.2851,
+      "step": 1584
+    },
+    {
+      "epoch": 11.917293233082706,
+      "grad_norm": 0.20195803998342107,
+      "learning_rate": 7.672797302899958e-07,
+      "loss": 0.2824,
+      "step": 1585
+    },
+    {
+      "epoch": 11.924812030075188,
+      "grad_norm": 0.21189800275013768,
+      "learning_rate": 7.660713121261909e-07,
+      "loss": 0.2847,
+      "step": 1586
+    },
+    {
+      "epoch": 11.93233082706767,
+      "grad_norm": 0.20824653864926013,
+      "learning_rate": 7.648632552300032e-07,
+      "loss": 0.2918,
+      "step": 1587
+    },
+    {
+      "epoch": 11.93984962406015,
+      "grad_norm": 0.23197140231599442,
+      "learning_rate": 7.636555614670952e-07,
+      "loss": 0.2841,
+      "step": 1588
+    },
+    {
+      "epoch": 11.947368421052632,
+      "grad_norm": 0.21110471824606641,
+      "learning_rate": 7.624482327025674e-07,
+      "loss": 0.2906,
+      "step": 1589
+    },
+    {
+      "epoch": 11.954887218045112,
+      "grad_norm": 0.20549421086170175,
+      "learning_rate": 7.612412708009582e-07,
+      "loss": 0.2835,
+      "step": 1590
+    },
+    {
+      "epoch": 11.962406015037594,
+      "grad_norm": 0.2086076870250293,
+      "learning_rate": 7.600346776262371e-07,
+      "loss": 0.2792,
+      "step": 1591
+    },
+    {
+      "epoch": 11.969924812030076,
+      "grad_norm": 0.20278216530120155,
+      "learning_rate": 7.588284550418067e-07,
+      "loss": 0.2861,
+      "step": 1592
+    },
+    {
+      "epoch": 11.977443609022556,
+      "grad_norm": 0.21772343500687005,
+      "learning_rate": 7.57622604910495e-07,
+      "loss": 0.2913,
+      "step": 1593
+    },
+    {
+      "epoch": 11.984962406015038,
+      "grad_norm": 0.21893297032528244,
+      "learning_rate": 7.56417129094557e-07,
+      "loss": 0.2841,
+      "step": 1594
+    },
+    {
+      "epoch": 11.992481203007518,
+      "grad_norm": 0.20337614158130457,
+      "learning_rate": 7.552120294556674e-07,
+      "loss": 0.2837,
+      "step": 1595
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.2140462264307299,
+      "learning_rate": 7.54007307854922e-07,
+      "loss": 0.2836,
+      "step": 1596
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 0.41957220435142517,
+      "eval_runtime": 1.714,
+      "eval_samples_per_second": 190.782,
+      "eval_steps_per_second": 3.501,
+      "step": 1596
+    },
+    {
+      "epoch": 12.007518796992482,
+      "grad_norm": 0.20155507093464683,
+      "learning_rate": 7.52802966152831e-07,
+      "loss": 0.2817,
+      "step": 1597
+    },
+    {
+      "epoch": 12.015037593984962,
+      "grad_norm": 0.21089247748985046,
+      "learning_rate": 7.515990062093194e-07,
+      "loss": 0.2912,
+      "step": 1598
+    },
+    {
+      "epoch": 12.022556390977444,
+      "grad_norm": 0.18975503644272712,
+      "learning_rate": 7.503954298837214e-07,
+      "loss": 0.2725,
+      "step": 1599
+    },
+    {
+      "epoch": 12.030075187969924,
+      "grad_norm": 0.2028946230901638,
+      "learning_rate": 7.4919223903478e-07,
+      "loss": 0.2774,
+      "step": 1600
+    },
+    {
+      "epoch": 12.037593984962406,
+      "grad_norm": 0.21294011813631644,
+      "learning_rate": 7.479894355206413e-07,
+      "loss": 0.2765,
+      "step": 1601
+    },
+    {
+      "epoch": 12.045112781954888,
+      "grad_norm": 0.20949256785335862,
+      "learning_rate": 7.46787021198855e-07,
+      "loss": 0.2749,
+      "step": 1602
+    },
+    {
+      "epoch": 12.052631578947368,
+      "grad_norm": 0.20848027531806115,
+      "learning_rate": 7.455849979263682e-07,
+      "loss": 0.2841,
+      "step": 1603
+    },
+    {
+      "epoch": 12.06015037593985,
+      "grad_norm": 0.19902873163425142,
+      "learning_rate": 7.443833675595253e-07,
+      "loss": 0.2887,
+      "step": 1604
+    },
+    {
+      "epoch": 12.06766917293233,
+      "grad_norm": 0.2154505867231227,
+      "learning_rate": 7.431821319540629e-07,
+      "loss": 0.2828,
+      "step": 1605
+    },
+    {
+      "epoch": 12.075187969924812,
+      "grad_norm": 0.2191032057746781,
+      "learning_rate": 7.419812929651091e-07,
+      "loss": 0.2858,
+      "step": 1606
+    },
+    {
+      "epoch": 12.082706766917294,
+      "grad_norm": 0.20641787723573152,
+      "learning_rate": 7.407808524471781e-07,
+      "loss": 0.2707,
+      "step": 1607
+    },
+    {
+      "epoch": 12.090225563909774,
+      "grad_norm": 0.2004024900881927,
+      "learning_rate": 7.395808122541695e-07,
+      "loss": 0.2848,
+      "step": 1608
+    },
+    {
+      "epoch": 12.097744360902256,
+      "grad_norm": 0.19221533264110013,
+      "learning_rate": 7.383811742393653e-07,
+      "loss": 0.2824,
+      "step": 1609
+    },
+    {
+      "epoch": 12.105263157894736,
+      "grad_norm": 0.20430104911034552,
+      "learning_rate": 7.371819402554247e-07,
+      "loss": 0.2852,
+      "step": 1610
+    },
+    {
+      "epoch": 12.112781954887218,
+      "grad_norm": 0.20334310960764398,
+      "learning_rate": 7.35983112154385e-07,
+      "loss": 0.2664,
+      "step": 1611
+    },
+    {
+      "epoch": 12.1203007518797,
+      "grad_norm": 0.21580607933403126,
+      "learning_rate": 7.347846917876543e-07,
+      "loss": 0.2915,
+      "step": 1612
+    },
+    {
+      "epoch": 12.12781954887218,
+      "grad_norm": 0.205268200859249,
+      "learning_rate": 7.335866810060139e-07,
+      "loss": 0.2924,
+      "step": 1613
+    },
+    {
+      "epoch": 12.135338345864662,
+      "grad_norm": 0.20974145546384027,
+      "learning_rate": 7.323890816596093e-07,
+      "loss": 0.2679,
+      "step": 1614
+    },
+    {
+      "epoch": 12.142857142857142,
+      "grad_norm": 0.20299728145714382,
+      "learning_rate": 7.311918955979537e-07,
+      "loss": 0.2915,
+      "step": 1615
+    },
+    {
+      "epoch": 12.150375939849624,
+      "grad_norm": 0.22549953970199085,
+      "learning_rate": 7.299951246699196e-07,
+      "loss": 0.2815,
+      "step": 1616
+    },
+    {
+      "epoch": 12.157894736842104,
+      "grad_norm": 0.19172754890219298,
+      "learning_rate": 7.287987707237402e-07,
+      "loss": 0.2684,
+      "step": 1617
+    },
+    {
+      "epoch": 12.165413533834586,
+      "grad_norm": 0.21473995724840902,
+      "learning_rate": 7.276028356070032e-07,
+      "loss": 0.2829,
+      "step": 1618
+    },
+    {
+      "epoch": 12.172932330827068,
+      "grad_norm": 0.20777920071365577,
+      "learning_rate": 7.264073211666509e-07,
+      "loss": 0.2861,
+      "step": 1619
+    },
+    {
+      "epoch": 12.180451127819548,
+      "grad_norm": 0.20845960145773262,
+      "learning_rate": 7.252122292489746e-07,
+      "loss": 0.2739,
+      "step": 1620
+    },
+    {
+      "epoch": 12.18796992481203,
+      "grad_norm": 0.20071515126061362,
+      "learning_rate": 7.240175616996146e-07,
+      "loss": 0.2805,
+      "step": 1621
+    },
+    {
+      "epoch": 12.19548872180451,
+      "grad_norm": 0.2185987269684635,
+      "learning_rate": 7.228233203635538e-07,
+      "loss": 0.2866,
+      "step": 1622
+    },
+    {
+      "epoch": 12.203007518796992,
+      "grad_norm": 0.21247656917070554,
+      "learning_rate": 7.216295070851193e-07,
+      "loss": 0.288,
+      "step": 1623
+    },
+    {
+      "epoch": 12.210526315789474,
+      "grad_norm": 0.21443972333115377,
+      "learning_rate": 7.204361237079746e-07,
+      "loss": 0.2813,
+      "step": 1624
+    },
+    {
+      "epoch": 12.218045112781954,
+      "grad_norm": 0.2009173329833493,
+      "learning_rate": 7.192431720751217e-07,
+      "loss": 0.2856,
+      "step": 1625
+    },
+    {
+      "epoch": 12.225563909774436,
+      "grad_norm": 0.20289450957544955,
+      "learning_rate": 7.180506540288938e-07,
+      "loss": 0.2758,
+      "step": 1626
+    },
+    {
+      "epoch": 12.233082706766917,
+      "grad_norm": 0.2051479316444853,
+      "learning_rate": 7.168585714109561e-07,
+      "loss": 0.2945,
+      "step": 1627
+    },
+    {
+      "epoch": 12.240601503759398,
+      "grad_norm": 0.21219199853875725,
+      "learning_rate": 7.156669260622997e-07,
+      "loss": 0.2893,
+      "step": 1628
+    },
+    {
+      "epoch": 12.24812030075188,
+      "grad_norm": 0.19379893397370718,
+      "learning_rate": 7.144757198232422e-07,
+      "loss": 0.2922,
+      "step": 1629
+    },
+    {
+      "epoch": 12.25563909774436,
+      "grad_norm": 0.21859658387028447,
+      "learning_rate": 7.132849545334217e-07,
+      "loss": 0.2825,
+      "step": 1630
+    },
+    {
+      "epoch": 12.263157894736842,
+      "grad_norm": 0.19869110231995074,
+      "learning_rate": 7.120946320317963e-07,
+      "loss": 0.2795,
+      "step": 1631
+    },
+    {
+      "epoch": 12.270676691729323,
+      "grad_norm": 0.2165394520126934,
+      "learning_rate": 7.109047541566391e-07,
+      "loss": 0.289,
+      "step": 1632
+    },
+    {
+      "epoch": 12.278195488721805,
+      "grad_norm": 0.21272825098121914,
+      "learning_rate": 7.097153227455378e-07,
+      "loss": 0.2911,
+      "step": 1633
+    },
+    {
+      "epoch": 12.285714285714286,
+      "grad_norm": 0.21777253804939642,
+      "learning_rate": 7.085263396353895e-07,
+      "loss": 0.2836,
+      "step": 1634
+    },
+    {
+      "epoch": 12.293233082706767,
+      "grad_norm": 0.2283268006884094,
+      "learning_rate": 7.073378066623998e-07,
+      "loss": 0.2798,
+      "step": 1635
+    },
+    {
+      "epoch": 12.300751879699249,
+      "grad_norm": 0.20190902704480512,
+      "learning_rate": 7.061497256620792e-07,
+      "loss": 0.2818,
+      "step": 1636
+    },
+    {
+      "epoch": 12.308270676691729,
+      "grad_norm": 0.20306478373898174,
+      "learning_rate": 7.049620984692391e-07,
+      "loss": 0.2871,
+      "step": 1637
+    },
+    {
+      "epoch": 12.31578947368421,
+      "grad_norm": 0.1988384159625026,
+      "learning_rate": 7.037749269179914e-07,
+      "loss": 0.2723,
+      "step": 1638
+    },
+    {
+      "epoch": 12.323308270676693,
+      "grad_norm": 0.20752158948617067,
+      "learning_rate": 7.02588212841743e-07,
+      "loss": 0.2804,
+      "step": 1639
+    },
+    {
+      "epoch": 12.330827067669173,
+      "grad_norm": 0.19994917905972487,
+      "learning_rate": 7.014019580731959e-07,
+      "loss": 0.2735,
+      "step": 1640
+    },
+    {
+      "epoch": 12.338345864661655,
+      "grad_norm": 0.2081598572924818,
+      "learning_rate": 7.002161644443411e-07,
+      "loss": 0.2812,
+      "step": 1641
+    },
+    {
+      "epoch": 12.345864661654135,
+      "grad_norm": 0.20964002010293234,
+      "learning_rate": 6.990308337864589e-07,
+      "loss": 0.2845,
+      "step": 1642
+    },
+    {
+      "epoch": 12.353383458646617,
+      "grad_norm": 0.20921790006316004,
+      "learning_rate": 6.978459679301132e-07,
+      "loss": 0.276,
+      "step": 1643
+    },
+    {
+      "epoch": 12.360902255639097,
+      "grad_norm": 0.2084835174921707,
+      "learning_rate": 6.966615687051516e-07,
+      "loss": 0.276,
+      "step": 1644
+    },
+    {
+      "epoch": 12.368421052631579,
+      "grad_norm": 0.21887017948208867,
+      "learning_rate": 6.954776379406995e-07,
+      "loss": 0.2912,
+      "step": 1645
+    },
+    {
+      "epoch": 12.37593984962406,
+      "grad_norm": 0.22665165113642693,
+      "learning_rate": 6.942941774651605e-07,
+      "loss": 0.2847,
+      "step": 1646
+    },
+    {
+      "epoch": 12.38345864661654,
+      "grad_norm": 0.20476784321624958,
+      "learning_rate": 6.9311118910621e-07,
+      "loss": 0.2808,
+      "step": 1647
+    },
+    {
+      "epoch": 12.390977443609023,
+      "grad_norm": 0.20083227230048167,
+      "learning_rate": 6.919286746907962e-07,
+      "loss": 0.2759,
+      "step": 1648
+    },
+    {
+      "epoch": 12.398496240601503,
+      "grad_norm": 0.20050752040238665,
+      "learning_rate": 6.907466360451337e-07,
+      "loss": 0.2817,
+      "step": 1649
+    },
+    {
+      "epoch": 12.406015037593985,
+      "grad_norm": 0.20799679846374464,
+      "learning_rate": 6.895650749947041e-07,
+      "loss": 0.2907,
+      "step": 1650
+    },
+    {
+      "epoch": 12.413533834586467,
+      "grad_norm": 0.20878651841981422,
+      "learning_rate": 6.883839933642493e-07,
+      "loss": 0.2889,
+      "step": 1651
+    },
+    {
+      "epoch": 12.421052631578947,
+      "grad_norm": 0.20929546140519786,
+      "learning_rate": 6.87203392977773e-07,
+      "loss": 0.2832,
+      "step": 1652
+    },
+    {
+      "epoch": 12.428571428571429,
+      "grad_norm": 0.20121444930741136,
+      "learning_rate": 6.860232756585336e-07,
+      "loss": 0.2761,
+      "step": 1653
+    },
+    {
+      "epoch": 12.436090225563909,
+      "grad_norm": 0.19891826875037882,
+      "learning_rate": 6.848436432290456e-07,
+      "loss": 0.2891,
+      "step": 1654
+    },
+    {
+      "epoch": 12.443609022556391,
+      "grad_norm": 0.20352007829226487,
+      "learning_rate": 6.836644975110726e-07,
+      "loss": 0.2957,
+      "step": 1655
+    },
+    {
+      "epoch": 12.451127819548873,
+      "grad_norm": 0.2013777470770896,
+      "learning_rate": 6.824858403256283e-07,
+      "loss": 0.2845,
+      "step": 1656
+    },
+    {
+      "epoch": 12.458646616541353,
+      "grad_norm": 0.20591786187297378,
+      "learning_rate": 6.813076734929703e-07,
+      "loss": 0.2809,
+      "step": 1657
+    },
+    {
+      "epoch": 12.466165413533835,
+      "grad_norm": 0.2107485806504614,
+      "learning_rate": 6.80129998832601e-07,
+      "loss": 0.2746,
+      "step": 1658
+    },
+    {
+      "epoch": 12.473684210526315,
+      "grad_norm": 0.21175377744180218,
+      "learning_rate": 6.789528181632603e-07,
+      "loss": 0.2807,
+      "step": 1659
+    },
+    {
+      "epoch": 12.481203007518797,
+      "grad_norm": 0.20748150511494287,
+      "learning_rate": 6.777761333029274e-07,
+      "loss": 0.2728,
+      "step": 1660
+    },
+    {
+      "epoch": 12.488721804511279,
+      "grad_norm": 0.19911841789036155,
+      "learning_rate": 6.765999460688144e-07,
+      "loss": 0.2753,
+      "step": 1661
+    },
+    {
+      "epoch": 12.496240601503759,
+      "grad_norm": 0.22386093392970424,
+      "learning_rate": 6.754242582773645e-07,
+      "loss": 0.2981,
+      "step": 1662
+    },
+    {
+      "epoch": 12.503759398496241,
+      "grad_norm": 0.20984963154087713,
+      "learning_rate": 6.742490717442517e-07,
+      "loss": 0.289,
+      "step": 1663
+    },
+    {
+      "epoch": 12.511278195488721,
+      "grad_norm": 0.20653383170097164,
+      "learning_rate": 6.730743882843734e-07,
+      "loss": 0.2867,
+      "step": 1664
+    },
+    {
+      "epoch": 12.518796992481203,
+      "grad_norm": 0.20965574996358757,
+      "learning_rate": 6.719002097118523e-07,
+      "loss": 0.2868,
+      "step": 1665
+    },
+    {
+      "epoch": 12.526315789473685,
+      "grad_norm": 0.20386313757769786,
+      "learning_rate": 6.707265378400295e-07,
+      "loss": 0.28,
+      "step": 1666
+    },
+    {
+      "epoch": 12.533834586466165,
+      "grad_norm": 0.2267744780973429,
+      "learning_rate": 6.69553374481465e-07,
+      "loss": 0.2993,
+      "step": 1667
+    },
+    {
+      "epoch": 12.541353383458647,
+      "grad_norm": 0.19687725368977047,
+      "learning_rate": 6.683807214479323e-07,
+      "loss": 0.2748,
+      "step": 1668
+    },
+    {
+      "epoch": 12.548872180451127,
+      "grad_norm": 0.2001230148296508,
+      "learning_rate": 6.672085805504178e-07,
+      "loss": 0.2893,
+      "step": 1669
+    },
+    {
+      "epoch": 12.556390977443609,
+      "grad_norm": 0.21019088613480197,
+      "learning_rate": 6.660369535991162e-07,
+      "loss": 0.2837,
+      "step": 1670
+    },
+    {
+      "epoch": 12.563909774436091,
+      "grad_norm": 0.21178061217156188,
+      "learning_rate": 6.648658424034292e-07,
+      "loss": 0.2927,
+      "step": 1671
+    },
+    {
+      "epoch": 12.571428571428571,
+      "grad_norm": 0.209897975817615,
+      "learning_rate": 6.636952487719612e-07,
+      "loss": 0.2808,
+      "step": 1672
+    },
+    {
+      "epoch": 12.578947368421053,
+      "grad_norm": 0.20725101594422754,
+      "learning_rate": 6.625251745125182e-07,
+      "loss": 0.281,
+      "step": 1673
+    },
+    {
+      "epoch": 12.586466165413533,
+      "grad_norm": 0.21586464576447126,
+      "learning_rate": 6.61355621432103e-07,
+      "loss": 0.2872,
+      "step": 1674
+    },
+    {
+      "epoch": 12.593984962406015,
+      "grad_norm": 0.20843621948425897,
+      "learning_rate": 6.601865913369149e-07,
+      "loss": 0.2882,
+      "step": 1675
+    },
+    {
+      "epoch": 12.601503759398497,
+      "grad_norm": 0.21640569457851286,
+      "learning_rate": 6.590180860323439e-07,
+      "loss": 0.2819,
+      "step": 1676
+    },
+    {
+      "epoch": 12.609022556390977,
+      "grad_norm": 0.19220593963285373,
+      "learning_rate": 6.578501073229713e-07,
+      "loss": 0.2776,
+      "step": 1677
+    },
+    {
+      "epoch": 12.61654135338346,
+      "grad_norm": 0.19533950775332323,
+      "learning_rate": 6.566826570125634e-07,
+      "loss": 0.275,
+      "step": 1678
+    },
+    {
+      "epoch": 12.62406015037594,
+      "grad_norm": 0.2132119594391018,
+      "learning_rate": 6.555157369040723e-07,
+      "loss": 0.2834,
+      "step": 1679
+    },
+    {
+      "epoch": 12.631578947368421,
+      "grad_norm": 0.2111142698264728,
+      "learning_rate": 6.543493487996292e-07,
+      "loss": 0.282,
+      "step": 1680
+    },
+    {
+      "epoch": 12.639097744360903,
+      "grad_norm": 0.2029210559761566,
+      "learning_rate": 6.531834945005459e-07,
+      "loss": 0.2743,
+      "step": 1681
+    },
+    {
+      "epoch": 12.646616541353383,
+      "grad_norm": 0.21123878720175188,
+      "learning_rate": 6.520181758073078e-07,
+      "loss": 0.2896,
+      "step": 1682
+    },
+    {
+      "epoch": 12.654135338345865,
+      "grad_norm": 0.21353587378112893,
+      "learning_rate": 6.508533945195749e-07,
+      "loss": 0.2949,
+      "step": 1683
+    },
+    {
+      "epoch": 12.661654135338345,
+      "grad_norm": 0.22073531642862246,
+      "learning_rate": 6.496891524361756e-07,
+      "loss": 0.2847,
+      "step": 1684
+    },
+    {
+      "epoch": 12.669172932330827,
+      "grad_norm": 0.22270628328596737,
+      "learning_rate": 6.485254513551072e-07,
+      "loss": 0.2874,
+      "step": 1685
+    },
+    {
+      "epoch": 12.676691729323307,
+      "grad_norm": 0.20929573537260124,
+      "learning_rate": 6.473622930735303e-07,
+      "loss": 0.2844,
+      "step": 1686
+    },
+    {
+      "epoch": 12.68421052631579,
+      "grad_norm": 0.2064089049872504,
+      "learning_rate": 6.461996793877673e-07,
+      "loss": 0.2909,
+      "step": 1687
+    },
+    {
+      "epoch": 12.691729323308271,
+      "grad_norm": 0.2045014977920784,
+      "learning_rate": 6.450376120933008e-07,
+      "loss": 0.2808,
+      "step": 1688
+    },
+    {
+      "epoch": 12.699248120300751,
+      "grad_norm": 0.20508091473071074,
+      "learning_rate": 6.438760929847678e-07,
+      "loss": 0.2914,
+      "step": 1689
+    },
+    {
+      "epoch": 12.706766917293233,
+      "grad_norm": 0.19424420185889268,
+      "learning_rate": 6.427151238559602e-07,
+      "loss": 0.2849,
+      "step": 1690
+    },
+    {
+      "epoch": 12.714285714285714,
+      "grad_norm": 0.20779530962212145,
+      "learning_rate": 6.415547064998193e-07,
+      "loss": 0.2893,
+      "step": 1691
+    },
+    {
+      "epoch": 12.721804511278195,
+      "grad_norm": 0.2052167841456742,
+      "learning_rate": 6.403948427084356e-07,
+      "loss": 0.2914,
+      "step": 1692
+    },
+    {
+      "epoch": 12.729323308270677,
+      "grad_norm": 0.20553125158312013,
+      "learning_rate": 6.392355342730431e-07,
+      "loss": 0.2739,
+      "step": 1693
+    },
+    {
+      "epoch": 12.736842105263158,
+      "grad_norm": 0.20840206724761057,
+      "learning_rate": 6.380767829840201e-07,
+      "loss": 0.2842,
+      "step": 1694
+    },
+    {
+      "epoch": 12.74436090225564,
+      "grad_norm": 0.21033746847320853,
+      "learning_rate": 6.369185906308825e-07,
+      "loss": 0.2781,
+      "step": 1695
+    },
+    {
+      "epoch": 12.75187969924812,
+      "grad_norm": 0.20662261490523776,
+      "learning_rate": 6.357609590022847e-07,
+      "loss": 0.2865,
+      "step": 1696
+    },
+    {
+      "epoch": 12.759398496240602,
+      "grad_norm": 0.1996876114948403,
+      "learning_rate": 6.346038898860136e-07,
+      "loss": 0.2881,
+      "step": 1697
+    },
+    {
+      "epoch": 12.766917293233083,
+      "grad_norm": 0.2017305963104442,
+      "learning_rate": 6.334473850689888e-07,
+      "loss": 0.2711,
+      "step": 1698
+    },
+    {
+      "epoch": 12.774436090225564,
+      "grad_norm": 0.1997731965233079,
+      "learning_rate": 6.322914463372569e-07,
+      "loss": 0.2699,
+      "step": 1699
+    },
+    {
+      "epoch": 12.781954887218046,
+      "grad_norm": 0.21168443626518807,
+      "learning_rate": 6.311360754759923e-07,
+      "loss": 0.2837,
+      "step": 1700
+    },
+    {
+      "epoch": 12.789473684210526,
+      "grad_norm": 0.21311077554459254,
+      "learning_rate": 6.299812742694901e-07,
+      "loss": 0.2971,
+      "step": 1701
+    },
+    {
+      "epoch": 12.796992481203008,
+      "grad_norm": 0.2029649051524788,
+      "learning_rate": 6.288270445011677e-07,
+      "loss": 0.2755,
+      "step": 1702
+    },
+    {
+      "epoch": 12.80451127819549,
+      "grad_norm": 0.21983156354074512,
+      "learning_rate": 6.276733879535583e-07,
+      "loss": 0.2862,
+      "step": 1703
+    },
+    {
+      "epoch": 12.81203007518797,
+      "grad_norm": 0.19781930228438968,
+      "learning_rate": 6.265203064083115e-07,
+      "loss": 0.2841,
+      "step": 1704
+    },
+    {
+      "epoch": 12.819548872180452,
+      "grad_norm": 0.21187055791970438,
+      "learning_rate": 6.253678016461872e-07,
+      "loss": 0.2734,
+      "step": 1705
+    },
+    {
+      "epoch": 12.827067669172932,
+      "grad_norm": 0.21436705565454725,
+      "learning_rate": 6.242158754470561e-07,
+      "loss": 0.2885,
+      "step": 1706
+    },
+    {
+      "epoch": 12.834586466165414,
+      "grad_norm": 0.19808067773029125,
+      "learning_rate": 6.23064529589894e-07,
+      "loss": 0.2737,
+      "step": 1707
+    },
+    {
+      "epoch": 12.842105263157894,
+      "grad_norm": 0.22814155420976528,
+      "learning_rate": 6.219137658527817e-07,
+      "loss": 0.2876,
+      "step": 1708
+    },
+    {
+      "epoch": 12.849624060150376,
+      "grad_norm": 0.196518511301975,
+      "learning_rate": 6.207635860129001e-07,
+      "loss": 0.2763,
+      "step": 1709
+    },
+    {
+      "epoch": 12.857142857142858,
+      "grad_norm": 0.20347442545222302,
+      "learning_rate": 6.196139918465291e-07,
+      "loss": 0.271,
+      "step": 1710
+    },
+    {
+      "epoch": 12.864661654135338,
+      "grad_norm": 0.20636750524985165,
+      "learning_rate": 6.184649851290428e-07,
+      "loss": 0.2806,
+      "step": 1711
+    },
+    {
+      "epoch": 12.87218045112782,
+      "grad_norm": 0.20277138102346362,
+      "learning_rate": 6.173165676349102e-07,
+      "loss": 0.2869,
+      "step": 1712
+    },
+    {
+      "epoch": 12.8796992481203,
+      "grad_norm": 0.20545969073674578,
+      "learning_rate": 6.161687411376886e-07,
+      "loss": 0.2793,
+      "step": 1713
+    },
+    {
+      "epoch": 12.887218045112782,
+      "grad_norm": 0.2126156280924279,
+      "learning_rate": 6.150215074100224e-07,
+      "loss": 0.2895,
+      "step": 1714
+    },
+    {
+      "epoch": 12.894736842105264,
+      "grad_norm": 0.20775486564982407,
+      "learning_rate": 6.138748682236423e-07,
+      "loss": 0.2719,
+      "step": 1715
+    },
+    {
+      "epoch": 12.902255639097744,
+      "grad_norm": 0.2013376227548472,
+      "learning_rate": 6.12728825349359e-07,
+      "loss": 0.2735,
+      "step": 1716
+    },
+    {
+      "epoch": 12.909774436090226,
+      "grad_norm": 0.20300358281612374,
+      "learning_rate": 6.115833805570638e-07,
+      "loss": 0.277,
+      "step": 1717
+    },
+    {
+      "epoch": 12.917293233082706,
+      "grad_norm": 0.19537276752799804,
+      "learning_rate": 6.104385356157229e-07,
+      "loss": 0.2737,
+      "step": 1718
+    },
+    {
+      "epoch": 12.924812030075188,
+      "grad_norm": 0.19656827220930295,
+      "learning_rate": 6.092942922933775e-07,
+      "loss": 0.2806,
+      "step": 1719
+    },
+    {
+      "epoch": 12.93233082706767,
+      "grad_norm": 0.20782648287820082,
+      "learning_rate": 6.081506523571384e-07,
+      "loss": 0.2825,
+      "step": 1720
+    },
+    {
+      "epoch": 12.93984962406015,
+      "grad_norm": 0.1956689261170356,
+      "learning_rate": 6.070076175731859e-07,
+      "loss": 0.2805,
+      "step": 1721
+    },
+    {
+      "epoch": 12.947368421052632,
+      "grad_norm": 0.2229197337920542,
+      "learning_rate": 6.058651897067641e-07,
+      "loss": 0.277,
+      "step": 1722
+    },
+    {
+      "epoch": 12.954887218045112,
+      "grad_norm": 0.19686015701728082,
+      "learning_rate": 6.047233705221819e-07,
+      "loss": 0.2758,
+      "step": 1723
+    },
+    {
+      "epoch": 12.962406015037594,
+      "grad_norm": 0.20260982194696558,
+      "learning_rate": 6.035821617828059e-07,
+      "loss": 0.2809,
+      "step": 1724
+    },
+    {
+      "epoch": 12.969924812030076,
+      "grad_norm": 0.2104204606694037,
+      "learning_rate": 6.024415652510622e-07,
+      "loss": 0.2822,
+      "step": 1725
+    },
+    {
+      "epoch": 12.977443609022556,
+      "grad_norm": 0.21443076950737805,
+      "learning_rate": 6.01301582688429e-07,
+      "loss": 0.2867,
+      "step": 1726
+    },
+    {
+      "epoch": 12.984962406015038,
+      "grad_norm": 0.2184203061064024,
+      "learning_rate": 6.001622158554388e-07,
+      "loss": 0.284,
+      "step": 1727
+    },
+    {
+      "epoch": 12.992481203007518,
+      "grad_norm": 0.19855523754598478,
+      "learning_rate": 5.990234665116712e-07,
+      "loss": 0.2728,
+      "step": 1728
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.2013098322647892,
+      "learning_rate": 5.978853364157538e-07,
+      "loss": 0.2779,
+      "step": 1729
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 0.4210340678691864,
+      "eval_runtime": 1.7443,
+      "eval_samples_per_second": 187.471,
+      "eval_steps_per_second": 3.44,
+      "step": 1729
+    },
+    {
+      "epoch": 13.007518796992482,
+      "grad_norm": 0.1973480008218584,
+      "learning_rate": 5.967478273253562e-07,
+      "loss": 0.2777,
+      "step": 1730
+    },
+    {
+      "epoch": 13.015037593984962,
+      "grad_norm": 0.2144900396017876,
+      "learning_rate": 5.956109409971907e-07,
+      "loss": 0.2946,
+      "step": 1731
+    },
+    {
+      "epoch": 13.022556390977444,
+      "grad_norm": 0.23713618160440814,
+      "learning_rate": 5.944746791870061e-07,
+      "loss": 0.2803,
+      "step": 1732
+    },
+    {
+      "epoch": 13.030075187969924,
+      "grad_norm": 0.20158191541243006,
+      "learning_rate": 5.933390436495885e-07,
+      "loss": 0.2725,
+      "step": 1733
+    },
+    {
+      "epoch": 13.037593984962406,
+      "grad_norm": 0.2189732568995817,
+      "learning_rate": 5.92204036138755e-07,
+      "loss": 0.272,
+      "step": 1734
+    },
+    {
+      "epoch": 13.045112781954888,
+      "grad_norm": 0.20065500275391585,
+      "learning_rate": 5.910696584073544e-07,
+      "loss": 0.287,
+      "step": 1735
+    },
+    {
+      "epoch": 13.052631578947368,
+      "grad_norm": 0.23039381599042535,
+      "learning_rate": 5.899359122072617e-07,
+      "loss": 0.283,
+      "step": 1736
+    },
+    {
+      "epoch": 13.06015037593985,
+      "grad_norm": 0.19643140919055457,
+      "learning_rate": 5.888027992893779e-07,
+      "loss": 0.2711,
+      "step": 1737
+    },
+    {
+      "epoch": 13.06766917293233,
+      "grad_norm": 0.2628198815737447,
+      "learning_rate": 5.87670321403624e-07,
+      "loss": 0.2919,
+      "step": 1738
+    },
+    {
+      "epoch": 13.075187969924812,
+      "grad_norm": 0.2057794786262758,
+      "learning_rate": 5.865384802989424e-07,
+      "loss": 0.2707,
+      "step": 1739
+    },
+    {
+      "epoch": 13.082706766917294,
+      "grad_norm": 0.2129809980818249,
+      "learning_rate": 5.854072777232914e-07,
+      "loss": 0.2807,
+      "step": 1740
+    },
+    {
+      "epoch": 13.090225563909774,
+      "grad_norm": 0.21578468946076385,
+      "learning_rate": 5.842767154236419e-07,
+      "loss": 0.2846,
+      "step": 1741
+    },
+    {
+      "epoch": 13.097744360902256,
+      "grad_norm": 0.21174876678537902,
+      "learning_rate": 5.831467951459783e-07,
+      "loss": 0.2864,
+      "step": 1742
+    },
+    {
+      "epoch": 13.105263157894736,
+      "grad_norm": 0.20724208520393902,
+      "learning_rate": 5.820175186352909e-07,
+      "loss": 0.2732,
+      "step": 1743
+    },
+    {
+      "epoch": 13.112781954887218,
+      "grad_norm": 0.18993999247392698,
+      "learning_rate": 5.808888876355784e-07,
+      "loss": 0.2678,
+      "step": 1744
+    },
+    {
+      "epoch": 13.1203007518797,
+      "grad_norm": 0.20160635189820433,
+      "learning_rate": 5.797609038898404e-07,
+      "loss": 0.2866,
+      "step": 1745
+    },
+    {
+      "epoch": 13.12781954887218,
+      "grad_norm": 0.22787524881427376,
+      "learning_rate": 5.786335691400788e-07,
+      "loss": 0.2941,
+      "step": 1746
+    },
+    {
+      "epoch": 13.135338345864662,
+      "grad_norm": 0.20403760605173848,
+      "learning_rate": 5.77506885127291e-07,
+      "loss": 0.2821,
+      "step": 1747
+    },
+    {
+      "epoch": 13.142857142857142,
+      "grad_norm": 0.2111236780007078,
+      "learning_rate": 5.763808535914723e-07,
+      "loss": 0.2894,
+      "step": 1748
+    },
+    {
+      "epoch": 13.150375939849624,
+      "grad_norm": 0.20730833888827283,
+      "learning_rate": 5.752554762716073e-07,
+      "loss": 0.2759,
+      "step": 1749
+    },
+    {
+      "epoch": 13.157894736842104,
+      "grad_norm": 0.23031658253330314,
+      "learning_rate": 5.741307549056728e-07,
+      "loss": 0.285,
+      "step": 1750
+    },
+    {
+      "epoch": 13.165413533834586,
+      "grad_norm": 0.2042932094987695,
+      "learning_rate": 5.730066912306309e-07,
+      "loss": 0.2864,
+      "step": 1751
+    },
+    {
+      "epoch": 13.172932330827068,
+      "grad_norm": 0.2182822762330046,
+      "learning_rate": 5.718832869824291e-07,
+      "loss": 0.2687,
+      "step": 1752
+    },
+    {
+      "epoch": 13.180451127819548,
+      "grad_norm": 0.21833757080429636,
+      "learning_rate": 5.707605438959954e-07,
+      "loss": 0.2858,
+      "step": 1753
+    },
+    {
+      "epoch": 13.18796992481203,
+      "grad_norm": 0.20993215432170534,
+      "learning_rate": 5.69638463705238e-07,
+      "loss": 0.2825,
+      "step": 1754
+    },
+    {
+      "epoch": 13.19548872180451,
+      "grad_norm": 0.21096215017811318,
+      "learning_rate": 5.685170481430401e-07,
+      "loss": 0.2704,
+      "step": 1755
+    },
+    {
+      "epoch": 13.203007518796992,
+      "grad_norm": 0.20731183919054988,
+      "learning_rate": 5.673962989412598e-07,
+      "loss": 0.2936,
+      "step": 1756
+    },
+    {
+      "epoch": 13.210526315789474,
+      "grad_norm": 0.19064627499012127,
+      "learning_rate": 5.662762178307248e-07,
+      "loss": 0.2617,
+      "step": 1757
+    },
+    {
+      "epoch": 13.218045112781954,
+      "grad_norm": 0.20438717435419318,
+      "learning_rate": 5.651568065412319e-07,
+      "loss": 0.2657,
+      "step": 1758
+    },
+    {
+      "epoch": 13.225563909774436,
+      "grad_norm": 0.2068620637193469,
+      "learning_rate": 5.64038066801543e-07,
+      "loss": 0.2709,
+      "step": 1759
+    },
+    {
+      "epoch": 13.233082706766917,
+      "grad_norm": 0.21865940373263065,
+      "learning_rate": 5.629200003393837e-07,
+      "loss": 0.2842,
+      "step": 1760
+    },
+    {
+      "epoch": 13.240601503759398,
+      "grad_norm": 0.20563243087289018,
+      "learning_rate": 5.618026088814382e-07,
+      "loss": 0.2745,
+      "step": 1761
+    },
+    {
+      "epoch": 13.24812030075188,
+      "grad_norm": 0.23248899532519754,
+      "learning_rate": 5.606858941533503e-07,
+      "loss": 0.2798,
+      "step": 1762
+    },
+    {
+      "epoch": 13.25563909774436,
+      "grad_norm": 0.2215141329553668,
+      "learning_rate": 5.595698578797168e-07,
+      "loss": 0.2714,
+      "step": 1763
+    },
+    {
+      "epoch": 13.263157894736842,
+      "grad_norm": 0.20762548815617707,
+      "learning_rate": 5.584545017840885e-07,
+      "loss": 0.2766,
+      "step": 1764
+    },
+    {
+      "epoch": 13.270676691729323,
+      "grad_norm": 0.19566864333777864,
+      "learning_rate": 5.573398275889638e-07,
+      "loss": 0.2784,
+      "step": 1765
+    },
+    {
+      "epoch": 13.278195488721805,
+      "grad_norm": 0.20691280909449536,
+      "learning_rate": 5.562258370157897e-07,
+      "loss": 0.2974,
+      "step": 1766
+    },
+    {
+      "epoch": 13.285714285714286,
+      "grad_norm": 0.2072270657078513,
+      "learning_rate": 5.551125317849572e-07,
+      "loss": 0.2813,
+      "step": 1767
+    },
+    {
+      "epoch": 13.293233082706767,
+      "grad_norm": 0.2413942981405282,
+      "learning_rate": 5.539999136157976e-07,
+      "loss": 0.2695,
+      "step": 1768
+    },
+    {
+      "epoch": 13.300751879699249,
+      "grad_norm": 0.2114637243498959,
+      "learning_rate": 5.52887984226583e-07,
+      "loss": 0.2829,
+      "step": 1769
+    },
+    {
+      "epoch": 13.308270676691729,
+      "grad_norm": 0.21388019418411683,
+      "learning_rate": 5.517767453345199e-07,
+      "loss": 0.2773,
+      "step": 1770
+    },
+    {
+      "epoch": 13.31578947368421,
+      "grad_norm": 0.20887199727652786,
+      "learning_rate": 5.506661986557503e-07,
+      "loss": 0.2876,
+      "step": 1771
+    },
+    {
+      "epoch": 13.323308270676693,
+      "grad_norm": 0.2018847117135878,
+      "learning_rate": 5.495563459053454e-07,
+      "loss": 0.2858,
+      "step": 1772
+    },
+    {
+      "epoch": 13.330827067669173,
+      "grad_norm": 0.20642055658469557,
+      "learning_rate": 5.484471887973062e-07,
+      "loss": 0.2891,
+      "step": 1773
+    },
+    {
+      "epoch": 13.338345864661655,
+      "grad_norm": 0.2270671380364826,
+      "learning_rate": 5.473387290445581e-07,
+      "loss": 0.2924,
+      "step": 1774
+    },
+    {
+      "epoch": 13.345864661654135,
+      "grad_norm": 0.21223595365368947,
+      "learning_rate": 5.46230968358951e-07,
+      "loss": 0.2878,
+      "step": 1775
+    },
+    {
+      "epoch": 13.353383458646617,
+      "grad_norm": 0.21706006126988825,
+      "learning_rate": 5.451239084512536e-07,
+      "loss": 0.2712,
+      "step": 1776
+    },
+    {
+      "epoch": 13.360902255639097,
+      "grad_norm": 0.19744709638909977,
+      "learning_rate": 5.44017551031154e-07,
+      "loss": 0.2748,
+      "step": 1777
+    },
+    {
+      "epoch": 13.368421052631579,
+      "grad_norm": 0.21005248623191583,
+      "learning_rate": 5.429118978072537e-07,
+      "loss": 0.2883,
+      "step": 1778
+    },
+    {
+      "epoch": 13.37593984962406,
+      "grad_norm": 0.20443079276473403,
+      "learning_rate": 5.418069504870684e-07,
+      "loss": 0.2808,
+      "step": 1779
+    },
+    {
+      "epoch": 13.38345864661654,
+      "grad_norm": 0.19540513553509112,
+      "learning_rate": 5.407027107770219e-07,
+      "loss": 0.2712,
+      "step": 1780
+    },
+    {
+      "epoch": 13.390977443609023,
+      "grad_norm": 0.20882196958871402,
+      "learning_rate": 5.395991803824469e-07,
+      "loss": 0.2862,
+      "step": 1781
+    },
+    {
+      "epoch": 13.398496240601503,
+      "grad_norm": 0.23158115606949117,
+      "learning_rate": 5.38496361007579e-07,
+      "loss": 0.2854,
+      "step": 1782
+    },
+    {
+      "epoch": 13.406015037593985,
+      "grad_norm": 0.21655303907655388,
+      "learning_rate": 5.373942543555575e-07,
+      "loss": 0.2835,
+      "step": 1783
+    },
+    {
+      "epoch": 13.413533834586467,
+      "grad_norm": 0.21635569123546528,
+      "learning_rate": 5.362928621284193e-07,
+      "loss": 0.2766,
+      "step": 1784
+    },
+    {
+      "epoch": 13.421052631578947,
+      "grad_norm": 0.21509987063157418,
+      "learning_rate": 5.351921860270993e-07,
+      "loss": 0.2903,
+      "step": 1785
+    },
+    {
+      "epoch": 13.428571428571429,
+      "grad_norm": 0.2184827351840628,
+      "learning_rate": 5.340922277514257e-07,
+      "loss": 0.2726,
+      "step": 1786
+    },
+    {
+      "epoch": 13.436090225563909,
+      "grad_norm": 0.20739626010728268,
+      "learning_rate": 5.329929890001186e-07,
+      "loss": 0.2727,
+      "step": 1787
+    },
+    {
+      "epoch": 13.443609022556391,
+      "grad_norm": 0.19463252957953395,
+      "learning_rate": 5.31894471470786e-07,
+      "loss": 0.2724,
+      "step": 1788
+    },
+    {
+      "epoch": 13.451127819548873,
+      "grad_norm": 0.20716386158004302,
+      "learning_rate": 5.307966768599236e-07,
+      "loss": 0.2905,
+      "step": 1789
+    },
+    {
+      "epoch": 13.458646616541353,
+      "grad_norm": 0.20359287552470517,
+      "learning_rate": 5.296996068629089e-07,
+      "loss": 0.2779,
+      "step": 1790
+    },
+    {
+      "epoch": 13.466165413533835,
+      "grad_norm": 0.1964179372032805,
+      "learning_rate": 5.286032631740023e-07,
+      "loss": 0.275,
+      "step": 1791
+    },
+    {
+      "epoch": 13.473684210526315,
+      "grad_norm": 0.21748821716097633,
+      "learning_rate": 5.275076474863408e-07,
+      "loss": 0.2857,
+      "step": 1792
+    },
+    {
+      "epoch": 13.481203007518797,
+      "grad_norm": 0.20214632260574256,
+      "learning_rate": 5.264127614919373e-07,
+      "loss": 0.2795,
+      "step": 1793
+    },
+    {
+      "epoch": 13.488721804511279,
+      "grad_norm": 0.2020575785556097,
+      "learning_rate": 5.253186068816795e-07,
+      "loss": 0.2783,
+      "step": 1794
+    },
+    {
+      "epoch": 13.496240601503759,
+      "grad_norm": 0.20801558547353988,
+      "learning_rate": 5.242251853453232e-07,
+      "loss": 0.2908,
+      "step": 1795
+    },
+    {
+      "epoch": 13.503759398496241,
+      "grad_norm": 0.21929376825296934,
+      "learning_rate": 5.231324985714941e-07,
+      "loss": 0.28,
+      "step": 1796
+    },
+    {
+      "epoch": 13.511278195488721,
+      "grad_norm": 0.21030054256176486,
+      "learning_rate": 5.220405482476815e-07,
+      "loss": 0.2909,
+      "step": 1797
+    },
+    {
+      "epoch": 13.518796992481203,
+      "grad_norm": 0.2111820750145914,
+      "learning_rate": 5.209493360602392e-07,
+      "loss": 0.2811,
+      "step": 1798
+    },
+    {
+      "epoch": 13.526315789473685,
+      "grad_norm": 0.21090604411567532,
+      "learning_rate": 5.198588636943789e-07,
+      "loss": 0.2818,
+      "step": 1799
+    },
+    {
+      "epoch": 13.533834586466165,
+      "grad_norm": 0.21775818381845996,
+      "learning_rate": 5.187691328341719e-07,
+      "loss": 0.2722,
+      "step": 1800
+    },
+    {
+      "epoch": 13.541353383458647,
+      "grad_norm": 0.20913886128065462,
+      "learning_rate": 5.176801451625426e-07,
+      "loss": 0.2958,
+      "step": 1801
+    },
+    {
+      "epoch": 13.548872180451127,
+      "grad_norm": 0.20918125751727082,
+      "learning_rate": 5.16591902361269e-07,
+      "loss": 0.2877,
+      "step": 1802
+    },
+    {
+      "epoch": 13.556390977443609,
+      "grad_norm": 0.1940934790085998,
+      "learning_rate": 5.155044061109775e-07,
+      "loss": 0.2807,
+      "step": 1803
+    },
+    {
+      "epoch": 13.563909774436091,
+      "grad_norm": 0.2085110734057618,
+      "learning_rate": 5.14417658091143e-07,
+      "loss": 0.2775,
+      "step": 1804
+    },
+    {
+      "epoch": 13.571428571428571,
+      "grad_norm": 0.21004882579863804,
+      "learning_rate": 5.133316599800832e-07,
+      "loss": 0.2773,
+      "step": 1805
+    },
+    {
+      "epoch": 13.578947368421053,
+      "grad_norm": 0.19795112213689592,
+      "learning_rate": 5.122464134549596e-07,
+      "loss": 0.2708,
+      "step": 1806
+    },
+    {
+      "epoch": 13.586466165413533,
+      "grad_norm": 0.19909431689212545,
+      "learning_rate": 5.111619201917709e-07,
+      "loss": 0.281,
+      "step": 1807
+    },
+    {
+      "epoch": 13.593984962406015,
+      "grad_norm": 0.2131157017019742,
+      "learning_rate": 5.100781818653547e-07,
+      "loss": 0.2838,
+      "step": 1808
+    },
+    {
+      "epoch": 13.601503759398497,
+      "grad_norm": 0.2054980969324712,
+      "learning_rate": 5.089952001493807e-07,
+      "loss": 0.2601,
+      "step": 1809
+    },
+    {
+      "epoch": 13.609022556390977,
+      "grad_norm": 0.20263186888344417,
+      "learning_rate": 5.079129767163514e-07,
+      "loss": 0.2833,
+      "step": 1810
+    },
+    {
+      "epoch": 13.61654135338346,
+      "grad_norm": 0.22130952497190162,
+      "learning_rate": 5.068315132375975e-07,
+      "loss": 0.2819,
+      "step": 1811
+    },
+    {
+      "epoch": 13.62406015037594,
+      "grad_norm": 0.20446951519935547,
+      "learning_rate": 5.057508113832772e-07,
+      "loss": 0.273,
+      "step": 1812
+    },
+    {
+      "epoch": 13.631578947368421,
+      "grad_norm": 0.20150492171420653,
+      "learning_rate": 5.046708728223708e-07,
+      "loss": 0.2634,
+      "step": 1813
+    },
+    {
+      "epoch": 13.639097744360903,
+      "grad_norm": 0.20600383112815174,
+      "learning_rate": 5.035916992226815e-07,
+      "loss": 0.2847,
+      "step": 1814
+    },
+    {
+      "epoch": 13.646616541353383,
+      "grad_norm": 0.21203524888887862,
+      "learning_rate": 5.025132922508293e-07,
+      "loss": 0.2676,
+      "step": 1815
+    },
+    {
+      "epoch": 13.654135338345865,
+      "grad_norm": 0.20865370468411548,
+      "learning_rate": 5.014356535722526e-07,
+      "loss": 0.2784,
+      "step": 1816
+    },
+    {
+      "epoch": 13.661654135338345,
+      "grad_norm": 0.2155452632018879,
+      "learning_rate": 5.00358784851201e-07,
+      "loss": 0.286,
+      "step": 1817
+    },
+    {
+      "epoch": 13.669172932330827,
+      "grad_norm": 0.21350349818340406,
+      "learning_rate": 4.992826877507366e-07,
+      "loss": 0.2937,
+      "step": 1818
+    },
+    {
+      "epoch": 13.676691729323307,
+      "grad_norm": 0.21645276292765617,
+      "learning_rate": 4.982073639327294e-07,
+      "loss": 0.2796,
+      "step": 1819
+    },
+    {
+      "epoch": 13.68421052631579,
+      "grad_norm": 0.2229963069444867,
+      "learning_rate": 4.971328150578539e-07,
+      "loss": 0.2819,
+      "step": 1820
+    },
+    {
+      "epoch": 13.691729323308271,
+      "grad_norm": 0.21397969818856066,
+      "learning_rate": 4.960590427855903e-07,
+      "loss": 0.2765,
+      "step": 1821
+    },
+    {
+      "epoch": 13.699248120300751,
+      "grad_norm": 0.2006776168708881,
+      "learning_rate": 4.949860487742173e-07,
+      "loss": 0.271,
+      "step": 1822
+    },
+    {
+      "epoch": 13.706766917293233,
+      "grad_norm": 0.2046114119073288,
+      "learning_rate": 4.939138346808129e-07,
+      "loss": 0.2717,
+      "step": 1823
+    },
+    {
+      "epoch": 13.714285714285714,
+      "grad_norm": 0.20351593259701506,
+      "learning_rate": 4.928424021612498e-07,
+      "loss": 0.2928,
+      "step": 1824
+    },
+    {
+      "epoch": 13.721804511278195,
+      "grad_norm": 0.20263738098056705,
+      "learning_rate": 4.917717528701949e-07,
+      "loss": 0.287,
+      "step": 1825
+    },
+    {
+      "epoch": 13.729323308270677,
+      "grad_norm": 0.2060634522924733,
+      "learning_rate": 4.907018884611039e-07,
+      "loss": 0.2797,
+      "step": 1826
+    },
+    {
+      "epoch": 13.736842105263158,
+      "grad_norm": 0.19729687832136994,
+      "learning_rate": 4.896328105862218e-07,
+      "loss": 0.2713,
+      "step": 1827
+    },
+    {
+      "epoch": 13.74436090225564,
+      "grad_norm": 0.21060374280298336,
+      "learning_rate": 4.885645208965778e-07,
+      "loss": 0.2914,
+      "step": 1828
+    },
+    {
+      "epoch": 13.75187969924812,
+      "grad_norm": 0.19246930208717883,
+      "learning_rate": 4.874970210419851e-07,
+      "loss": 0.2704,
+      "step": 1829
+    },
+    {
+      "epoch": 13.759398496240602,
+      "grad_norm": 0.2050507458364134,
+      "learning_rate": 4.864303126710356e-07,
+      "loss": 0.2789,
+      "step": 1830
+    },
+    {
+      "epoch": 13.766917293233083,
+      "grad_norm": 0.20516541795044332,
+      "learning_rate": 4.853643974311003e-07,
+      "loss": 0.2772,
+      "step": 1831
+    },
+    {
+      "epoch": 13.774436090225564,
+      "grad_norm": 0.22026383695608867,
+      "learning_rate": 4.842992769683242e-07,
+      "loss": 0.2918,
+      "step": 1832
+    },
+    {
+      "epoch": 13.781954887218046,
+      "grad_norm": 0.2116892797062628,
+      "learning_rate": 4.832349529276262e-07,
+      "loss": 0.2723,
+      "step": 1833
+    },
+    {
+      "epoch": 13.789473684210526,
+      "grad_norm": 0.2050972645893637,
+      "learning_rate": 4.821714269526934e-07,
+      "loss": 0.2856,
+      "step": 1834
+    },
+    {
+      "epoch": 13.796992481203008,
+      "grad_norm": 0.20999276451858898,
+      "learning_rate": 4.811087006859823e-07,
+      "loss": 0.2767,
+      "step": 1835
+    },
+    {
+      "epoch": 13.80451127819549,
+      "grad_norm": 0.20207784745788485,
+      "learning_rate": 4.80046775768713e-07,
+      "loss": 0.2754,
+      "step": 1836
+    },
+    {
+      "epoch": 13.81203007518797,
+      "grad_norm": 0.20714060278157415,
+      "learning_rate": 4.78985653840869e-07,
+      "loss": 0.2873,
+      "step": 1837
+    },
+    {
+      "epoch": 13.819548872180452,
+      "grad_norm": 0.21002205697547952,
+      "learning_rate": 4.779253365411926e-07,
+      "loss": 0.2694,
+      "step": 1838
+    },
+    {
+      "epoch": 13.827067669172932,
+      "grad_norm": 0.2113933461107075,
+      "learning_rate": 4.768658255071851e-07,
+      "loss": 0.2837,
+      "step": 1839
+    },
+    {
+      "epoch": 13.834586466165414,
+      "grad_norm": 0.20152574402813828,
+      "learning_rate": 4.7580712237510056e-07,
+      "loss": 0.2903,
+      "step": 1840
+    },
+    {
+      "epoch": 13.842105263157894,
+      "grad_norm": 0.19964178186268183,
+      "learning_rate": 4.747492287799475e-07,
+      "loss": 0.2784,
+      "step": 1841
+    },
+    {
+      "epoch": 13.849624060150376,
+      "grad_norm": 0.19947486753578864,
+      "learning_rate": 4.7369214635548237e-07,
+      "loss": 0.2887,
+      "step": 1842
+    },
+    {
+      "epoch": 13.857142857142858,
+      "grad_norm": 0.21213439156563046,
+      "learning_rate": 4.726358767342106e-07,
+      "loss": 0.2861,
+      "step": 1843
+    },
+    {
+      "epoch": 13.864661654135338,
+      "grad_norm": 0.21433896585951104,
+      "learning_rate": 4.715804215473809e-07,
+      "loss": 0.2718,
+      "step": 1844
+    },
+    {
+      "epoch": 13.87218045112782,
+      "grad_norm": 0.2163007643909855,
+      "learning_rate": 4.705257824249845e-07,
+      "loss": 0.2699,
+      "step": 1845
+    },
+    {
+      "epoch": 13.8796992481203,
+      "grad_norm": 0.20388976562185163,
+      "learning_rate": 4.6947196099575345e-07,
+      "loss": 0.2813,
+      "step": 1846
+    },
+    {
+      "epoch": 13.887218045112782,
+      "grad_norm": 0.19830417747470178,
+      "learning_rate": 4.6841895888715546e-07,
+      "loss": 0.2803,
+      "step": 1847
+    },
+    {
+      "epoch": 13.894736842105264,
+      "grad_norm": 0.18695934057350605,
+      "learning_rate": 4.6736677772539435e-07,
+      "loss": 0.2651,
+      "step": 1848
+    },
+    {
+      "epoch": 13.902255639097744,
+      "grad_norm": 0.20176507193602475,
+      "learning_rate": 4.6631541913540474e-07,
+      "loss": 0.2827,
+      "step": 1849
+    },
+    {
+      "epoch": 13.909774436090226,
+      "grad_norm": 0.20501251481116037,
+      "learning_rate": 4.6526488474085224e-07,
+      "loss": 0.2723,
+      "step": 1850
+    },
+    {
+      "epoch": 13.917293233082706,
+      "grad_norm": 0.2029110913473786,
+      "learning_rate": 4.642151761641282e-07,
+      "loss": 0.2742,
+      "step": 1851
+    },
+    {
+      "epoch": 13.924812030075188,
+      "grad_norm": 0.1965061800917423,
+      "learning_rate": 4.631662950263502e-07,
+      "loss": 0.2819,
+      "step": 1852
+    },
+    {
+      "epoch": 13.93233082706767,
+      "grad_norm": 0.20791002671176906,
+      "learning_rate": 4.6211824294735647e-07,
+      "loss": 0.2772,
+      "step": 1853
+    },
+    {
+      "epoch": 13.93984962406015,
+      "grad_norm": 0.20335021462025002,
+      "learning_rate": 4.610710215457061e-07,
+      "loss": 0.2811,
+      "step": 1854
+    },
+    {
+      "epoch": 13.947368421052632,
+      "grad_norm": 0.1983298303892597,
+      "learning_rate": 4.6002463243867416e-07,
+      "loss": 0.2758,
+      "step": 1855
+    },
+    {
+      "epoch": 13.954887218045112,
+      "grad_norm": 0.19792709731051142,
+      "learning_rate": 4.5897907724225183e-07,
+      "loss": 0.2795,
+      "step": 1856
+    },
+    {
+      "epoch": 13.962406015037594,
+      "grad_norm": 0.19416915204508933,
+      "learning_rate": 4.5793435757114076e-07,
+      "loss": 0.2598,
+      "step": 1857
+    },
+    {
+      "epoch": 13.969924812030076,
+      "grad_norm": 0.21720277302367624,
+      "learning_rate": 4.5689047503875376e-07,
+      "loss": 0.2716,
+      "step": 1858
+    },
+    {
+      "epoch": 13.977443609022556,
+      "grad_norm": 0.20443786916213968,
+      "learning_rate": 4.558474312572095e-07,
+      "loss": 0.2937,
+      "step": 1859
+    },
+    {
+      "epoch": 13.984962406015038,
+      "grad_norm": 0.20147238287839345,
+      "learning_rate": 4.5480522783733265e-07,
+      "loss": 0.2797,
+      "step": 1860
+    },
+    {
+      "epoch": 13.992481203007518,
+      "grad_norm": 0.21515062266483992,
+      "learning_rate": 4.5376386638864874e-07,
+      "loss": 0.2767,
+      "step": 1861
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.20554743285312888,
+      "learning_rate": 4.527233485193843e-07,
+      "loss": 0.2779,
+      "step": 1862
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 0.42210811376571655,
+      "eval_runtime": 1.6322,
+      "eval_samples_per_second": 200.346,
+      "eval_steps_per_second": 3.676,
+      "step": 1862
+    },
+    {
+      "epoch": 14.007518796992482,
+      "grad_norm": 0.20050481878869533,
+      "learning_rate": 4.5168367583646173e-07,
+      "loss": 0.2879,
+      "step": 1863
+    },
+    {
+      "epoch": 14.015037593984962,
+      "grad_norm": 0.19009295838122864,
+      "learning_rate": 4.5064484994549955e-07,
+      "loss": 0.2662,
+      "step": 1864
+    },
+    {
+      "epoch": 14.022556390977444,
+      "grad_norm": 0.20582351377695815,
+      "learning_rate": 4.496068724508072e-07,
+      "loss": 0.2889,
+      "step": 1865
+    },
+    {
+      "epoch": 14.030075187969924,
+      "grad_norm": 0.1996559900469183,
+      "learning_rate": 4.4856974495538527e-07,
+      "loss": 0.2825,
+      "step": 1866
+    },
+    {
+      "epoch": 14.037593984962406,
+      "grad_norm": 0.2242278011464106,
+      "learning_rate": 4.4753346906092006e-07,
+      "loss": 0.2809,
+      "step": 1867
+    },
+    {
+      "epoch": 14.045112781954888,
+      "grad_norm": 0.19972285031518056,
+      "learning_rate": 4.4649804636778455e-07,
+      "loss": 0.2692,
+      "step": 1868
+    },
+    {
+      "epoch": 14.052631578947368,
+      "grad_norm": 0.20279598343202543,
+      "learning_rate": 4.454634784750322e-07,
+      "loss": 0.2741,
+      "step": 1869
+    },
+    {
+      "epoch": 14.06015037593985,
+      "grad_norm": 0.2041141551015994,
+      "learning_rate": 4.4442976698039803e-07,
+      "loss": 0.2715,
+      "step": 1870
+    },
+    {
+      "epoch": 14.06766917293233,
+      "grad_norm": 0.20802567549224651,
+      "learning_rate": 4.4339691348029297e-07,
+      "loss": 0.277,
+      "step": 1871
+    },
+    {
+      "epoch": 14.075187969924812,
+      "grad_norm": 0.218156554639469,
+      "learning_rate": 4.4236491956980415e-07,
+      "loss": 0.2822,
+      "step": 1872
+    },
+    {
+      "epoch": 14.082706766917294,
+      "grad_norm": 0.21052593521397164,
+      "learning_rate": 4.4133378684269086e-07,
+      "loss": 0.2853,
+      "step": 1873
+    },
+    {
+      "epoch": 14.090225563909774,
+      "grad_norm": 0.20440360099498187,
+      "learning_rate": 4.403035168913817e-07,
+      "loss": 0.273,
+      "step": 1874
+    },
+    {
+      "epoch": 14.097744360902256,
+      "grad_norm": 0.20656217711879954,
+      "learning_rate": 4.3927411130697403e-07,
+      "loss": 0.2925,
+      "step": 1875
+    },
+    {
+      "epoch": 14.105263157894736,
+      "grad_norm": 0.1958112860697871,
+      "learning_rate": 4.38245571679229e-07,
+      "loss": 0.2825,
+      "step": 1876
+    },
+    {
+      "epoch": 14.112781954887218,
+      "grad_norm": 0.20799036187590217,
+      "learning_rate": 4.3721789959657186e-07,
+      "loss": 0.2864,
+      "step": 1877
+    },
+    {
+      "epoch": 14.1203007518797,
+      "grad_norm": 0.20849046755835549,
+      "learning_rate": 4.3619109664608655e-07,
+      "loss": 0.2905,
+      "step": 1878
+    },
+    {
+      "epoch": 14.12781954887218,
+      "grad_norm": 0.2114672946117387,
+      "learning_rate": 4.351651644135164e-07,
+      "loss": 0.2877,
+      "step": 1879
+    },
+    {
+      "epoch": 14.135338345864662,
+      "grad_norm": 0.20801170581229594,
+      "learning_rate": 4.3414010448325824e-07,
+      "loss": 0.2622,
+      "step": 1880
+    },
+    {
+      "epoch": 14.142857142857142,
+      "grad_norm": 0.20699859469201662,
+      "learning_rate": 4.331159184383636e-07,
+      "loss": 0.286,
+      "step": 1881
+    },
+    {
+      "epoch": 14.150375939849624,
+      "grad_norm": 0.205223277235336,
+      "learning_rate": 4.3209260786053283e-07,
+      "loss": 0.2656,
+      "step": 1882
+    },
+    {
+      "epoch": 14.157894736842104,
+      "grad_norm": 0.20423356206652143,
+      "learning_rate": 4.310701743301156e-07,
+      "loss": 0.2788,
+      "step": 1883
+    },
+    {
+      "epoch": 14.165413533834586,
+      "grad_norm": 0.2255010388441208,
+      "learning_rate": 4.3004861942610573e-07,
+      "loss": 0.2748,
+      "step": 1884
+    },
+    {
+      "epoch": 14.172932330827068,
+      "grad_norm": 0.20769047104938534,
+      "learning_rate": 4.290279447261417e-07,
+      "loss": 0.3037,
+      "step": 1885
+    },
+    {
+      "epoch": 14.180451127819548,
+      "grad_norm": 0.2068118132370387,
+      "learning_rate": 4.28008151806501e-07,
+      "loss": 0.2712,
+      "step": 1886
+    },
+    {
+      "epoch": 14.18796992481203,
+      "grad_norm": 0.19780304408127383,
+      "learning_rate": 4.2698924224210085e-07,
+      "loss": 0.2811,
+      "step": 1887
+    },
+    {
+      "epoch": 14.19548872180451,
+      "grad_norm": 0.20174494053440864,
+      "learning_rate": 4.25971217606493e-07,
+      "loss": 0.2742,
+      "step": 1888
+    },
+    {
+      "epoch": 14.203007518796992,
+      "grad_norm": 0.1969580809190312,
+      "learning_rate": 4.2495407947186377e-07,
+      "loss": 0.2661,
+      "step": 1889
+    },
+    {
+      "epoch": 14.210526315789474,
+      "grad_norm": 0.20347095518201466,
+      "learning_rate": 4.239378294090291e-07,
+      "loss": 0.2675,
+      "step": 1890
+    },
+    {
+      "epoch": 14.218045112781954,
+      "grad_norm": 0.19342190869122228,
+      "learning_rate": 4.229224689874349e-07,
+      "loss": 0.2787,
+      "step": 1891
+    },
+    {
+      "epoch": 14.225563909774436,
+      "grad_norm": 0.2183339304764494,
+      "learning_rate": 4.2190799977515145e-07,
+      "loss": 0.2808,
+      "step": 1892
+    },
+    {
+      "epoch": 14.233082706766917,
+      "grad_norm": 0.2060771720742975,
+      "learning_rate": 4.208944233388745e-07,
+      "loss": 0.2794,
+      "step": 1893
+    },
+    {
+      "epoch": 14.240601503759398,
+      "grad_norm": 0.1992686910104434,
+      "learning_rate": 4.1988174124391927e-07,
+      "loss": 0.2705,
+      "step": 1894
+    },
+    {
+      "epoch": 14.24812030075188,
+      "grad_norm": 0.21304823194051536,
+      "learning_rate": 4.1886995505422174e-07,
+      "loss": 0.2777,
+      "step": 1895
+    },
+    {
+      "epoch": 14.25563909774436,
+      "grad_norm": 0.1922007103599193,
+      "learning_rate": 4.178590663323323e-07,
+      "loss": 0.2808,
+      "step": 1896
+    },
+    {
+      "epoch": 14.263157894736842,
+      "grad_norm": 0.20167695507927771,
+      "learning_rate": 4.1684907663941703e-07,
+      "loss": 0.2746,
+      "step": 1897
+    },
+    {
+      "epoch": 14.270676691729323,
+      "grad_norm": 0.2130394084687997,
+      "learning_rate": 4.158399875352525e-07,
+      "loss": 0.2783,
+      "step": 1898
+    },
+    {
+      "epoch": 14.278195488721805,
+      "grad_norm": 0.19677558649105964,
+      "learning_rate": 4.1483180057822453e-07,
+      "loss": 0.2761,
+      "step": 1899
+    },
+    {
+      "epoch": 14.285714285714286,
+      "grad_norm": 0.19672709372355912,
+      "learning_rate": 4.138245173253266e-07,
+      "loss": 0.2791,
+      "step": 1900
+    },
+    {
+      "epoch": 14.293233082706767,
+      "grad_norm": 0.20773292388654246,
+      "learning_rate": 4.128181393321554e-07,
+      "loss": 0.2798,
+      "step": 1901
+    },
+    {
+      "epoch": 14.300751879699249,
+      "grad_norm": 0.20302917621085323,
+      "learning_rate": 4.118126681529107e-07,
+      "loss": 0.2794,
+      "step": 1902
+    },
+    {
+      "epoch": 14.308270676691729,
+      "grad_norm": 0.19267592153809207,
+      "learning_rate": 4.108081053403906e-07,
+      "loss": 0.2741,
+      "step": 1903
+    },
+    {
+      "epoch": 14.31578947368421,
+      "grad_norm": 0.20204840002331625,
+      "learning_rate": 4.0980445244599173e-07,
+      "loss": 0.2838,
+      "step": 1904
+    },
+    {
+      "epoch": 14.323308270676693,
+      "grad_norm": 0.2074879949957779,
+      "learning_rate": 4.0880171101970407e-07,
+      "loss": 0.2678,
+      "step": 1905
+    },
+    {
+      "epoch": 14.330827067669173,
+      "grad_norm": 0.2048814677037202,
+      "learning_rate": 4.0779988261011146e-07,
+      "loss": 0.2749,
+      "step": 1906
+    },
+    {
+      "epoch": 14.338345864661655,
+      "grad_norm": 0.20273584667111136,
+      "learning_rate": 4.067989687643861e-07,
+      "loss": 0.2855,
+      "step": 1907
+    },
+    {
+      "epoch": 14.345864661654135,
+      "grad_norm": 0.20384883597733727,
+      "learning_rate": 4.0579897102828965e-07,
+      "loss": 0.2928,
+      "step": 1908
+    },
+    {
+      "epoch": 14.353383458646617,
+      "grad_norm": 0.20324167017402833,
+      "learning_rate": 4.047998909461668e-07,
+      "loss": 0.277,
+      "step": 1909
+    },
+    {
+      "epoch": 14.360902255639097,
+      "grad_norm": 0.21057142140658675,
+      "learning_rate": 4.0380173006094744e-07,
+      "loss": 0.2876,
+      "step": 1910
+    },
+    {
+      "epoch": 14.368421052631579,
+      "grad_norm": 0.2073534592120044,
+      "learning_rate": 4.028044899141396e-07,
+      "loss": 0.2751,
+      "step": 1911
+    },
+    {
+      "epoch": 14.37593984962406,
+      "grad_norm": 0.2030945732115819,
+      "learning_rate": 4.0180817204583127e-07,
+      "loss": 0.274,
+      "step": 1912
+    },
+    {
+      "epoch": 14.38345864661654,
+      "grad_norm": 0.2057185929226416,
+      "learning_rate": 4.0081277799468473e-07,
+      "loss": 0.2742,
+      "step": 1913
+    },
+    {
+      "epoch": 14.390977443609023,
+      "grad_norm": 0.2192971360337129,
+      "learning_rate": 3.998183092979367e-07,
+      "loss": 0.2797,
+      "step": 1914
+    },
+    {
+      "epoch": 14.398496240601503,
+      "grad_norm": 0.20086369329171025,
+      "learning_rate": 3.988247674913935e-07,
+      "loss": 0.2836,
+      "step": 1915
+    },
+    {
+      "epoch": 14.406015037593985,
+      "grad_norm": 0.19871017932652363,
+      "learning_rate": 3.978321541094317e-07,
+      "loss": 0.2829,
+      "step": 1916
+    },
+    {
+      "epoch": 14.413533834586467,
+      "grad_norm": 0.19423254416261637,
+      "learning_rate": 3.9684047068499227e-07,
+      "loss": 0.2727,
+      "step": 1917
+    },
+    {
+      "epoch": 14.421052631578947,
+      "grad_norm": 0.20838412863665473,
+      "learning_rate": 3.958497187495815e-07,
+      "loss": 0.2733,
+      "step": 1918
+    },
+    {
+      "epoch": 14.428571428571429,
+      "grad_norm": 0.19942291222506192,
+      "learning_rate": 3.9485989983326605e-07,
+      "loss": 0.2759,
+      "step": 1919
+    },
+    {
+      "epoch": 14.436090225563909,
+      "grad_norm": 0.19563657983581506,
+      "learning_rate": 3.938710154646726e-07,
+      "loss": 0.2689,
+      "step": 1920
+    },
+    {
+      "epoch": 14.443609022556391,
+      "grad_norm": 0.20220093666849073,
+      "learning_rate": 3.928830671709835e-07,
+      "loss": 0.2777,
+      "step": 1921
+    },
+    {
+      "epoch": 14.451127819548873,
+      "grad_norm": 0.20264253110587577,
+      "learning_rate": 3.918960564779368e-07,
+      "loss": 0.2716,
+      "step": 1922
+    },
+    {
+      "epoch": 14.458646616541353,
+      "grad_norm": 0.19178158502119938,
+      "learning_rate": 3.9090998490982116e-07,
+      "loss": 0.2729,
+      "step": 1923
+    },
+    {
+      "epoch": 14.466165413533835,
+      "grad_norm": 0.2077636569417559,
+      "learning_rate": 3.8992485398947563e-07,
+      "loss": 0.2899,
+      "step": 1924
+    },
+    {
+      "epoch": 14.473684210526315,
+      "grad_norm": 0.20985515036891492,
+      "learning_rate": 3.8894066523828706e-07,
+      "loss": 0.2796,
+      "step": 1925
+    },
+    {
+      "epoch": 14.481203007518797,
+      "grad_norm": 0.1965304543003896,
+      "learning_rate": 3.879574201761858e-07,
+      "loss": 0.2731,
+      "step": 1926
+    },
+    {
+      "epoch": 14.488721804511279,
+      "grad_norm": 0.2123805149623651,
+      "learning_rate": 3.869751203216468e-07,
+      "loss": 0.28,
+      "step": 1927
+    },
+    {
+      "epoch": 14.496240601503759,
+      "grad_norm": 0.1988156098643813,
+      "learning_rate": 3.8599376719168317e-07,
+      "loss": 0.2797,
+      "step": 1928
+    },
+    {
+      "epoch": 14.503759398496241,
+      "grad_norm": 0.20472970635472604,
+      "learning_rate": 3.8501336230184786e-07,
+      "loss": 0.2761,
+      "step": 1929
+    },
+    {
+      "epoch": 14.511278195488721,
+      "grad_norm": 0.2086745444483143,
+      "learning_rate": 3.8403390716622785e-07,
+      "loss": 0.2826,
+      "step": 1930
+    },
+    {
+      "epoch": 14.518796992481203,
+      "grad_norm": 0.20564302053050398,
+      "learning_rate": 3.8305540329744456e-07,
+      "loss": 0.2629,
+      "step": 1931
+    },
+    {
+      "epoch": 14.526315789473685,
+      "grad_norm": 0.20438080997041855,
+      "learning_rate": 3.8207785220664934e-07,
+      "loss": 0.2656,
+      "step": 1932
+    },
+    {
+      "epoch": 14.533834586466165,
+      "grad_norm": 0.1953496910564303,
+      "learning_rate": 3.811012554035231e-07,
+      "loss": 0.2712,
+      "step": 1933
+    },
+    {
+      "epoch": 14.541353383458647,
+      "grad_norm": 0.21025247640803466,
+      "learning_rate": 3.801256143962719e-07,
+      "loss": 0.2781,
+      "step": 1934
+    },
+    {
+      "epoch": 14.548872180451127,
+      "grad_norm": 0.2052532573577962,
+      "learning_rate": 3.7915093069162685e-07,
+      "loss": 0.2725,
+      "step": 1935
+    },
+    {
+      "epoch": 14.556390977443609,
+      "grad_norm": 0.21665929039106555,
+      "learning_rate": 3.7817720579483956e-07,
+      "loss": 0.2885,
+      "step": 1936
+    },
+    {
+      "epoch": 14.563909774436091,
+      "grad_norm": 0.19624670860518137,
+      "learning_rate": 3.772044412096821e-07,
+      "loss": 0.2835,
+      "step": 1937
+    },
+    {
+      "epoch": 14.571428571428571,
+      "grad_norm": 0.20227641012957082,
+      "learning_rate": 3.762326384384421e-07,
+      "loss": 0.2753,
+      "step": 1938
+    },
+    {
+      "epoch": 14.578947368421053,
+      "grad_norm": 0.20151119393498465,
+      "learning_rate": 3.752617989819232e-07,
+      "loss": 0.2693,
+      "step": 1939
+    },
+    {
+      "epoch": 14.586466165413533,
+      "grad_norm": 0.20681778595033043,
+      "learning_rate": 3.7429192433944013e-07,
+      "loss": 0.2798,
+      "step": 1940
+    },
+    {
+      "epoch": 14.593984962406015,
+      "grad_norm": 0.19726748524065252,
+      "learning_rate": 3.7332301600881866e-07,
+      "loss": 0.275,
+      "step": 1941
+    },
+    {
+      "epoch": 14.601503759398497,
+      "grad_norm": 0.2057510433308467,
+      "learning_rate": 3.723550754863912e-07,
+      "loss": 0.2731,
+      "step": 1942
+    },
+    {
+      "epoch": 14.609022556390977,
+      "grad_norm": 0.19759691458231676,
+      "learning_rate": 3.7138810426699675e-07,
+      "loss": 0.2772,
+      "step": 1943
+    },
+    {
+      "epoch": 14.61654135338346,
+      "grad_norm": 0.20539777670589912,
+      "learning_rate": 3.7042210384397586e-07,
+      "loss": 0.2951,
+      "step": 1944
+    },
+    {
+      "epoch": 14.62406015037594,
+      "grad_norm": 0.19547530934667925,
+      "learning_rate": 3.694570757091715e-07,
+      "loss": 0.2811,
+      "step": 1945
+    },
+    {
+      "epoch": 14.631578947368421,
+      "grad_norm": 0.2010077327880398,
+      "learning_rate": 3.6849302135292346e-07,
+      "loss": 0.2881,
+      "step": 1946
+    },
+    {
+      "epoch": 14.639097744360903,
+      "grad_norm": 0.20526860735213723,
+      "learning_rate": 3.675299422640693e-07,
+      "loss": 0.275,
+      "step": 1947
+    },
+    {
+      "epoch": 14.646616541353383,
+      "grad_norm": 0.20407428860395674,
+      "learning_rate": 3.6656783992993876e-07,
+      "loss": 0.2879,
+      "step": 1948
+    },
+    {
+      "epoch": 14.654135338345865,
+      "grad_norm": 0.19718244049727113,
+      "learning_rate": 3.656067158363546e-07,
+      "loss": 0.2633,
+      "step": 1949
+    },
+    {
+      "epoch": 14.661654135338345,
+      "grad_norm": 0.2092455958184149,
+      "learning_rate": 3.64646571467628e-07,
+      "loss": 0.2845,
+      "step": 1950
+    },
+    {
+      "epoch": 14.669172932330827,
+      "grad_norm": 0.19548959216191925,
+      "learning_rate": 3.6368740830655686e-07,
+      "loss": 0.2622,
+      "step": 1951
+    },
+    {
+      "epoch": 14.676691729323307,
+      "grad_norm": 0.20001903688767875,
+      "learning_rate": 3.6272922783442494e-07,
+      "loss": 0.275,
+      "step": 1952
+    },
+    {
+      "epoch": 14.68421052631579,
+      "grad_norm": 0.21995593511526879,
+      "learning_rate": 3.617720315309968e-07,
+      "loss": 0.2811,
+      "step": 1953
+    },
+    {
+      "epoch": 14.691729323308271,
+      "grad_norm": 0.19946537463773006,
+      "learning_rate": 3.608158208745187e-07,
+      "loss": 0.2735,
+      "step": 1954
+    },
+    {
+      "epoch": 14.699248120300751,
+      "grad_norm": 0.19389584825692904,
+      "learning_rate": 3.5986059734171336e-07,
+      "loss": 0.2773,
+      "step": 1955
+    },
+    {
+      "epoch": 14.706766917293233,
+      "grad_norm": 0.21839956484701017,
+      "learning_rate": 3.5890636240778015e-07,
+      "loss": 0.2869,
+      "step": 1956
+    },
+    {
+      "epoch": 14.714285714285714,
+      "grad_norm": 0.21969296821676065,
+      "learning_rate": 3.579531175463906e-07,
+      "loss": 0.2743,
+      "step": 1957
+    },
+    {
+      "epoch": 14.721804511278195,
+      "grad_norm": 0.20237737288732519,
+      "learning_rate": 3.5700086422968843e-07,
+      "loss": 0.2863,
+      "step": 1958
+    },
+    {
+      "epoch": 14.729323308270677,
+      "grad_norm": 0.2152951745818716,
+      "learning_rate": 3.5604960392828475e-07,
+      "loss": 0.2832,
+      "step": 1959
+    },
+    {
+      "epoch": 14.736842105263158,
+      "grad_norm": 0.2002300172255161,
+      "learning_rate": 3.550993381112585e-07,
+      "loss": 0.2728,
+      "step": 1960
+    },
+    {
+      "epoch": 14.74436090225564,
+      "grad_norm": 0.20002829410787845,
+      "learning_rate": 3.5415006824615133e-07,
+      "loss": 0.2718,
+      "step": 1961
+    },
+    {
+      "epoch": 14.75187969924812,
+      "grad_norm": 0.20744458870791105,
+      "learning_rate": 3.5320179579896834e-07,
+      "loss": 0.2803,
+      "step": 1962
+    },
+    {
+      "epoch": 14.759398496240602,
+      "grad_norm": 0.19480592212499442,
+      "learning_rate": 3.522545222341726e-07,
+      "loss": 0.2677,
+      "step": 1963
+    },
+    {
+      "epoch": 14.766917293233083,
+      "grad_norm": 0.20610469332817938,
+      "learning_rate": 3.513082490146864e-07,
+      "loss": 0.2777,
+      "step": 1964
+    },
+    {
+      "epoch": 14.774436090225564,
+      "grad_norm": 0.21123145787543793,
+      "learning_rate": 3.5036297760188517e-07,
+      "loss": 0.2744,
+      "step": 1965
+    },
+    {
+      "epoch": 14.781954887218046,
+      "grad_norm": 0.20727848846449803,
+      "learning_rate": 3.4941870945559905e-07,
+      "loss": 0.2772,
+      "step": 1966
+    },
+    {
+      "epoch": 14.789473684210526,
+      "grad_norm": 0.19919268796335388,
+      "learning_rate": 3.4847544603410727e-07,
+      "loss": 0.2754,
+      "step": 1967
+    },
+    {
+      "epoch": 14.796992481203008,
+      "grad_norm": 0.19914592817566648,
+      "learning_rate": 3.475331887941387e-07,
+      "loss": 0.275,
+      "step": 1968
+    },
+    {
+      "epoch": 14.80451127819549,
+      "grad_norm": 0.20801841410876298,
+      "learning_rate": 3.4659193919086715e-07,
+      "loss": 0.277,
+      "step": 1969
+    },
+    {
+      "epoch": 14.81203007518797,
+      "grad_norm": 0.22518860101696211,
+      "learning_rate": 3.4565169867791143e-07,
+      "loss": 0.2867,
+      "step": 1970
+    },
+    {
+      "epoch": 14.819548872180452,
+      "grad_norm": 0.200356568258531,
+      "learning_rate": 3.447124687073306e-07,
+      "loss": 0.275,
+      "step": 1971
+    },
+    {
+      "epoch": 14.827067669172932,
+      "grad_norm": 0.19662308826032893,
+      "learning_rate": 3.437742507296246e-07,
+      "loss": 0.2652,
+      "step": 1972
+    },
+    {
+      "epoch": 14.834586466165414,
+      "grad_norm": 0.209924949202204,
+      "learning_rate": 3.428370461937291e-07,
+      "loss": 0.2816,
+      "step": 1973
+    },
+    {
+      "epoch": 14.842105263157894,
+      "grad_norm": 0.202105885120364,
+      "learning_rate": 3.4190085654701604e-07,
+      "loss": 0.274,
+      "step": 1974
+    },
+    {
+      "epoch": 14.849624060150376,
+      "grad_norm": 0.20158367702273305,
+      "learning_rate": 3.409656832352885e-07,
+      "loss": 0.2767,
+      "step": 1975
+    },
+    {
+      "epoch": 14.857142857142858,
+      "grad_norm": 0.21012134376327457,
+      "learning_rate": 3.400315277027812e-07,
+      "loss": 0.2771,
+      "step": 1976
+    },
+    {
+      "epoch": 14.864661654135338,
+      "grad_norm": 0.20444996695645845,
+      "learning_rate": 3.3909839139215704e-07,
+      "loss": 0.2816,
+      "step": 1977
+    },
+    {
+      "epoch": 14.87218045112782,
+      "grad_norm": 0.19877562199868368,
+      "learning_rate": 3.3816627574450364e-07,
+      "loss": 0.2807,
+      "step": 1978
+    },
+    {
+      "epoch": 14.8796992481203,
+      "grad_norm": 0.21241655373004667,
+      "learning_rate": 3.3723518219933387e-07,
+      "loss": 0.2741,
+      "step": 1979
+    },
+    {
+      "epoch": 14.887218045112782,
+      "grad_norm": 0.19572573023160178,
+      "learning_rate": 3.363051121945809e-07,
+      "loss": 0.2869,
+      "step": 1980
+    },
+    {
+      "epoch": 14.894736842105264,
+      "grad_norm": 0.19551193385096538,
+      "learning_rate": 3.3537606716659836e-07,
+      "loss": 0.2647,
+      "step": 1981
+    },
+    {
+      "epoch": 14.902255639097744,
+      "grad_norm": 0.20623814710473015,
+      "learning_rate": 3.3444804855015573e-07,
+      "loss": 0.2941,
+      "step": 1982
+    },
+    {
+      "epoch": 14.909774436090226,
+      "grad_norm": 0.20092197415656857,
+      "learning_rate": 3.3352105777843853e-07,
+      "loss": 0.2775,
+      "step": 1983
+    },
+    {
+      "epoch": 14.917293233082706,
+      "grad_norm": 0.20952193682651182,
+      "learning_rate": 3.3259509628304363e-07,
+      "loss": 0.284,
+      "step": 1984
+    },
+    {
+      "epoch": 14.924812030075188,
+      "grad_norm": 0.2032470850170072,
+      "learning_rate": 3.3167016549397984e-07,
+      "loss": 0.2756,
+      "step": 1985
+    },
+    {
+      "epoch": 14.93233082706767,
+      "grad_norm": 0.19935535216542571,
+      "learning_rate": 3.307462668396628e-07,
+      "loss": 0.2739,
+      "step": 1986
+    },
+    {
+      "epoch": 14.93984962406015,
+      "grad_norm": 0.2171778333141113,
+      "learning_rate": 3.298234017469154e-07,
+      "loss": 0.2773,
+      "step": 1987
+    },
+    {
+      "epoch": 14.947368421052632,
+      "grad_norm": 0.2085213093903102,
+      "learning_rate": 3.289015716409631e-07,
+      "loss": 0.2833,
+      "step": 1988
+    },
+    {
+      "epoch": 14.954887218045112,
+      "grad_norm": 0.20905761688250082,
+      "learning_rate": 3.279807779454342e-07,
+      "loss": 0.2895,
+      "step": 1989
+    },
+    {
+      "epoch": 14.962406015037594,
+      "grad_norm": 0.21840964014017522,
+      "learning_rate": 3.270610220823553e-07,
+      "loss": 0.2796,
+      "step": 1990
+    },
+    {
+      "epoch": 14.969924812030076,
+      "grad_norm": 0.2059399851152583,
+      "learning_rate": 3.261423054721515e-07,
+      "loss": 0.2746,
+      "step": 1991
+    },
+    {
+      "epoch": 14.977443609022556,
+      "grad_norm": 0.20535079750522622,
+      "learning_rate": 3.2522462953364125e-07,
+      "loss": 0.274,
+      "step": 1992
+    },
+    {
+      "epoch": 14.984962406015038,
+      "grad_norm": 0.19853598081158197,
+      "learning_rate": 3.2430799568403776e-07,
+      "loss": 0.2727,
+      "step": 1993
+    },
+    {
+      "epoch": 14.992481203007518,
+      "grad_norm": 0.23132782933486165,
+      "learning_rate": 3.233924053389432e-07,
+      "loss": 0.2764,
+      "step": 1994
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.1987997963374311,
+      "learning_rate": 3.2247785991234943e-07,
+      "loss": 0.279,
+      "step": 1995
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 0.4242452085018158,
+      "eval_runtime": 1.6872,
+      "eval_samples_per_second": 193.811,
+      "eval_steps_per_second": 3.556,
+      "step": 1995
+    },
+    {
+      "epoch": 15.007518796992482,
+      "grad_norm": 0.1967196349408533,
+      "learning_rate": 3.2156436081663353e-07,
+      "loss": 0.2796,
+      "step": 1996
+    },
+    {
+      "epoch": 15.015037593984962,
+      "grad_norm": 0.20857565111032078,
+      "learning_rate": 3.206519094625578e-07,
+      "loss": 0.2821,
+      "step": 1997
+    },
+    {
+      "epoch": 15.022556390977444,
+      "grad_norm": 0.20376148833641733,
+      "learning_rate": 3.1974050725926547e-07,
+      "loss": 0.2816,
+      "step": 1998
+    },
+    {
+      "epoch": 15.030075187969924,
+      "grad_norm": 0.2008341288614672,
+      "learning_rate": 3.188301556142805e-07,
+      "loss": 0.2743,
+      "step": 1999
+    },
+    {
+      "epoch": 15.037593984962406,
+      "grad_norm": 0.22109294094452472,
+      "learning_rate": 3.1792085593350306e-07,
+      "loss": 0.2763,
+      "step": 2000
+    },
+    {
+      "epoch": 15.045112781954888,
+      "grad_norm": 0.19841577830896204,
+      "learning_rate": 3.1701260962121036e-07,
+      "loss": 0.2801,
+      "step": 2001
+    },
+    {
+      "epoch": 15.052631578947368,
+      "grad_norm": 0.20562576694744183,
+      "learning_rate": 3.1610541808005154e-07,
+      "loss": 0.2731,
+      "step": 2002
+    },
+    {
+      "epoch": 15.06015037593985,
+      "grad_norm": 0.2052735770232253,
+      "learning_rate": 3.15199282711047e-07,
+      "loss": 0.2791,
+      "step": 2003
+    },
+    {
+      "epoch": 15.06766917293233,
+      "grad_norm": 0.20414758504154334,
+      "learning_rate": 3.142942049135869e-07,
+      "loss": 0.2673,
+      "step": 2004
+    },
+    {
+      "epoch": 15.075187969924812,
+      "grad_norm": 0.214965006890517,
+      "learning_rate": 3.133901860854271e-07,
+      "loss": 0.2788,
+      "step": 2005
+    },
+    {
+      "epoch": 15.082706766917294,
+      "grad_norm": 0.2015050470686567,
+      "learning_rate": 3.12487227622689e-07,
+      "loss": 0.27,
+      "step": 2006
+    },
+    {
+      "epoch": 15.090225563909774,
+      "grad_norm": 0.2072813957498029,
+      "learning_rate": 3.115853309198552e-07,
+      "loss": 0.2899,
+      "step": 2007
+    },
+    {
+      "epoch": 15.097744360902256,
+      "grad_norm": 0.22298973403134012,
+      "learning_rate": 3.106844973697701e-07,
+      "loss": 0.2724,
+      "step": 2008
+    },
+    {
+      "epoch": 15.105263157894736,
+      "grad_norm": 0.19568547942054326,
+      "learning_rate": 3.0978472836363454e-07,
+      "loss": 0.2641,
+      "step": 2009
+    },
+    {
+      "epoch": 15.112781954887218,
+      "grad_norm": 0.20803370004527222,
+      "learning_rate": 3.0888602529100705e-07,
+      "loss": 0.2772,
+      "step": 2010
+    },
+    {
+      "epoch": 15.1203007518797,
+      "grad_norm": 0.19684002831717112,
+      "learning_rate": 3.0798838953979845e-07,
+      "loss": 0.276,
+      "step": 2011
+    },
+    {
+      "epoch": 15.12781954887218,
+      "grad_norm": 0.2024729073700384,
+      "learning_rate": 3.070918224962725e-07,
+      "loss": 0.2785,
+      "step": 2012
+    },
+    {
+      "epoch": 15.135338345864662,
+      "grad_norm": 0.203828640010915,
+      "learning_rate": 3.061963255450415e-07,
+      "loss": 0.2836,
+      "step": 2013
+    },
+    {
+      "epoch": 15.142857142857142,
+      "grad_norm": 0.2176083279157845,
+      "learning_rate": 3.0530190006906587e-07,
+      "loss": 0.2795,
+      "step": 2014
+    },
+    {
+      "epoch": 15.150375939849624,
+      "grad_norm": 0.19855935325540455,
+      "learning_rate": 3.044085474496507e-07,
+      "loss": 0.2616,
+      "step": 2015
+    },
+    {
+      "epoch": 15.157894736842104,
+      "grad_norm": 0.21457742691321022,
+      "learning_rate": 3.03516269066445e-07,
+      "loss": 0.2949,
+      "step": 2016
+    },
+    {
+      "epoch": 15.165413533834586,
+      "grad_norm": 0.21323618098518354,
+      "learning_rate": 3.026250662974377e-07,
+      "loss": 0.2919,
+      "step": 2017
+    },
+    {
+      "epoch": 15.172932330827068,
+      "grad_norm": 0.2085596555484541,
+      "learning_rate": 3.017349405189579e-07,
+      "loss": 0.2884,
+      "step": 2018
+    },
+    {
+      "epoch": 15.180451127819548,
+      "grad_norm": 0.1993128561633088,
+      "learning_rate": 3.008458931056701e-07,
+      "loss": 0.2716,
+      "step": 2019
+    },
+    {
+      "epoch": 15.18796992481203,
+      "grad_norm": 0.2007344426986475,
+      "learning_rate": 2.9995792543057473e-07,
+      "loss": 0.2809,
+      "step": 2020
+    },
+    {
+      "epoch": 15.19548872180451,
+      "grad_norm": 0.19581042337410498,
+      "learning_rate": 2.990710388650034e-07,
+      "loss": 0.2702,
+      "step": 2021
+    },
+    {
+      "epoch": 15.203007518796992,
+      "grad_norm": 0.18906227561174047,
+      "learning_rate": 2.9818523477861955e-07,
+      "loss": 0.265,
+      "step": 2022
+    },
+    {
+      "epoch": 15.210526315789474,
+      "grad_norm": 0.21178940761048162,
+      "learning_rate": 2.973005145394135e-07,
+      "loss": 0.2711,
+      "step": 2023
+    },
+    {
+      "epoch": 15.218045112781954,
+      "grad_norm": 0.19852684011807434,
+      "learning_rate": 2.96416879513703e-07,
+      "loss": 0.2684,
+      "step": 2024
+    },
+    {
+      "epoch": 15.225563909774436,
+      "grad_norm": 0.20610350575138794,
+      "learning_rate": 2.955343310661286e-07,
+      "loss": 0.2717,
+      "step": 2025
+    },
+    {
+      "epoch": 15.233082706766917,
+      "grad_norm": 0.2215500392913798,
+      "learning_rate": 2.9465287055965393e-07,
+      "loss": 0.2934,
+      "step": 2026
+    },
+    {
+      "epoch": 15.240601503759398,
+      "grad_norm": 0.2136686632159001,
+      "learning_rate": 2.937724993555615e-07,
+      "loss": 0.2937,
+      "step": 2027
+    },
+    {
+      "epoch": 15.24812030075188,
+      "grad_norm": 0.19934816906651984,
+      "learning_rate": 2.9289321881345254e-07,
+      "loss": 0.2773,
+      "step": 2028
+    },
+    {
+      "epoch": 15.25563909774436,
+      "grad_norm": 0.20527531379544026,
+      "learning_rate": 2.920150302912431e-07,
+      "loss": 0.2836,
+      "step": 2029
+    },
+    {
+      "epoch": 15.263157894736842,
+      "grad_norm": 0.22083393019506523,
+      "learning_rate": 2.9113793514516273e-07,
+      "loss": 0.2821,
+      "step": 2030
+    },
+    {
+      "epoch": 15.270676691729323,
+      "grad_norm": 0.21061225073730577,
+      "learning_rate": 2.9026193472975333e-07,
+      "loss": 0.2701,
+      "step": 2031
+    },
+    {
+      "epoch": 15.278195488721805,
+      "grad_norm": 0.18977273493322552,
+      "learning_rate": 2.8938703039786503e-07,
+      "loss": 0.2664,
+      "step": 2032
+    },
+    {
+      "epoch": 15.285714285714286,
+      "grad_norm": 0.21070063578292922,
+      "learning_rate": 2.885132235006564e-07,
+      "loss": 0.2807,
+      "step": 2033
+    },
+    {
+      "epoch": 15.293233082706767,
+      "grad_norm": 0.20215302416094835,
+      "learning_rate": 2.876405153875898e-07,
+      "loss": 0.2752,
+      "step": 2034
+    },
+    {
+      "epoch": 15.300751879699249,
+      "grad_norm": 0.18808046810032925,
+      "learning_rate": 2.867689074064323e-07,
+      "loss": 0.2626,
+      "step": 2035
+    },
+    {
+      "epoch": 15.308270676691729,
+      "grad_norm": 0.2095875734819006,
+      "learning_rate": 2.858984009032502e-07,
+      "loss": 0.2808,
+      "step": 2036
+    },
+    {
+      "epoch": 15.31578947368421,
+      "grad_norm": 0.19452839675582367,
+      "learning_rate": 2.850289972224106e-07,
+      "loss": 0.2738,
+      "step": 2037
+    },
+    {
+      "epoch": 15.323308270676693,
+      "grad_norm": 0.20316691300697795,
+      "learning_rate": 2.841606977065757e-07,
+      "loss": 0.2703,
+      "step": 2038
+    },
+    {
+      "epoch": 15.330827067669173,
+      "grad_norm": 0.1982377132816269,
+      "learning_rate": 2.832935036967038e-07,
+      "loss": 0.2728,
+      "step": 2039
+    },
+    {
+      "epoch": 15.338345864661655,
+      "grad_norm": 0.21319016554947534,
+      "learning_rate": 2.82427416532045e-07,
+      "loss": 0.2676,
+      "step": 2040
+    },
+    {
+      "epoch": 15.345864661654135,
+      "grad_norm": 0.2190548757559591,
+      "learning_rate": 2.815624375501411e-07,
+      "loss": 0.2765,
+      "step": 2041
+    },
+    {
+      "epoch": 15.353383458646617,
+      "grad_norm": 0.2094666625189289,
+      "learning_rate": 2.806985680868209e-07,
+      "loss": 0.2873,
+      "step": 2042
+    },
+    {
+      "epoch": 15.360902255639097,
+      "grad_norm": 0.20834262855355112,
+      "learning_rate": 2.7983580947620165e-07,
+      "loss": 0.2812,
+      "step": 2043
+    },
+    {
+      "epoch": 15.368421052631579,
+      "grad_norm": 0.20442122191750084,
+      "learning_rate": 2.789741630506832e-07,
+      "loss": 0.2741,
+      "step": 2044
+    },
+    {
+      "epoch": 15.37593984962406,
+      "grad_norm": 0.19728700874035063,
+      "learning_rate": 2.781136301409492e-07,
+      "loss": 0.2854,
+      "step": 2045
+    },
+    {
+      "epoch": 15.38345864661654,
+      "grad_norm": 0.21469186366527013,
+      "learning_rate": 2.7725421207596277e-07,
+      "loss": 0.2866,
+      "step": 2046
+    },
+    {
+      "epoch": 15.390977443609023,
+      "grad_norm": 0.20360447579086394,
+      "learning_rate": 2.7639591018296605e-07,
+      "loss": 0.2817,
+      "step": 2047
+    },
+    {
+      "epoch": 15.398496240601503,
+      "grad_norm": 0.204337665194563,
+      "learning_rate": 2.755387257874764e-07,
+      "loss": 0.2745,
+      "step": 2048
+    },
+    {
+      "epoch": 15.406015037593985,
+      "grad_norm": 0.1973643915224638,
+      "learning_rate": 2.746826602132867e-07,
+      "loss": 0.2677,
+      "step": 2049
+    },
+    {
+      "epoch": 15.413533834586467,
+      "grad_norm": 0.194185545668033,
+      "learning_rate": 2.738277147824605e-07,
+      "loss": 0.2665,
+      "step": 2050
+    },
+    {
+      "epoch": 15.421052631578947,
+      "grad_norm": 0.19469587987540068,
+      "learning_rate": 2.7297389081533297e-07,
+      "loss": 0.2657,
+      "step": 2051
+    },
+    {
+      "epoch": 15.428571428571429,
+      "grad_norm": 0.2081283882654499,
+      "learning_rate": 2.721211896305059e-07,
+      "loss": 0.2783,
+      "step": 2052
+    },
+    {
+      "epoch": 15.436090225563909,
+      "grad_norm": 0.22797880536275691,
+      "learning_rate": 2.712696125448485e-07,
+      "loss": 0.2762,
+      "step": 2053
+    },
+    {
+      "epoch": 15.443609022556391,
+      "grad_norm": 0.1985226900230115,
+      "learning_rate": 2.704191608734926e-07,
+      "loss": 0.2762,
+      "step": 2054
+    },
+    {
+      "epoch": 15.451127819548873,
+      "grad_norm": 0.19838565413838996,
+      "learning_rate": 2.695698359298334e-07,
+      "loss": 0.2787,
+      "step": 2055
+    },
+    {
+      "epoch": 15.458646616541353,
+      "grad_norm": 0.19718615224851507,
+      "learning_rate": 2.687216390255249e-07,
+      "loss": 0.281,
+      "step": 2056
+    },
+    {
+      "epoch": 15.466165413533835,
+      "grad_norm": 0.19393530571515896,
+      "learning_rate": 2.678745714704792e-07,
+      "loss": 0.2763,
+      "step": 2057
+    },
+    {
+      "epoch": 15.473684210526315,
+      "grad_norm": 0.20888674094729937,
+      "learning_rate": 2.6702863457286516e-07,
+      "loss": 0.2686,
+      "step": 2058
+    },
+    {
+      "epoch": 15.481203007518797,
+      "grad_norm": 0.2161406909863793,
+      "learning_rate": 2.6618382963910424e-07,
+      "loss": 0.2766,
+      "step": 2059
+    },
+    {
+      "epoch": 15.488721804511279,
+      "grad_norm": 0.19955180835035108,
+      "learning_rate": 2.65340157973871e-07,
+      "loss": 0.277,
+      "step": 2060
+    },
+    {
+      "epoch": 15.496240601503759,
+      "grad_norm": 0.20611473022394117,
+      "learning_rate": 2.6449762088008863e-07,
+      "loss": 0.2805,
+      "step": 2061
+    },
+    {
+      "epoch": 15.503759398496241,
+      "grad_norm": 0.19737782091882605,
+      "learning_rate": 2.636562196589294e-07,
+      "loss": 0.2745,
+      "step": 2062
+    },
+    {
+      "epoch": 15.511278195488721,
+      "grad_norm": 0.19480666129211152,
+      "learning_rate": 2.6281595560981005e-07,
+      "loss": 0.2603,
+      "step": 2063
+    },
+    {
+      "epoch": 15.518796992481203,
+      "grad_norm": 0.22605458050838736,
+      "learning_rate": 2.619768300303925e-07,
+      "loss": 0.2984,
+      "step": 2064
+    },
+    {
+      "epoch": 15.526315789473685,
+      "grad_norm": 0.1920723687249796,
+      "learning_rate": 2.611388442165791e-07,
+      "loss": 0.2686,
+      "step": 2065
+    },
+    {
+      "epoch": 15.533834586466165,
+      "grad_norm": 0.2012149648010654,
+      "learning_rate": 2.603019994625133e-07,
+      "loss": 0.266,
+      "step": 2066
+    },
+    {
+      "epoch": 15.541353383458647,
+      "grad_norm": 0.19538612089773114,
+      "learning_rate": 2.5946629706057534e-07,
+      "loss": 0.2729,
+      "step": 2067
+    },
+    {
+      "epoch": 15.548872180451127,
+      "grad_norm": 0.21195974355504657,
+      "learning_rate": 2.586317383013821e-07,
+      "loss": 0.2766,
+      "step": 2068
+    },
+    {
+      "epoch": 15.556390977443609,
+      "grad_norm": 0.19994892529655686,
+      "learning_rate": 2.577983244737832e-07,
+      "loss": 0.2712,
+      "step": 2069
+    },
+    {
+      "epoch": 15.563909774436091,
+      "grad_norm": 0.20266220135451155,
+      "learning_rate": 2.569660568648616e-07,
+      "loss": 0.278,
+      "step": 2070
+    },
+    {
+      "epoch": 15.571428571428571,
+      "grad_norm": 0.20869238383425567,
+      "learning_rate": 2.561349367599285e-07,
+      "loss": 0.2682,
+      "step": 2071
+    },
+    {
+      "epoch": 15.578947368421053,
+      "grad_norm": 0.20257709269118124,
+      "learning_rate": 2.5530496544252424e-07,
+      "loss": 0.2799,
+      "step": 2072
+    },
+    {
+      "epoch": 15.586466165413533,
+      "grad_norm": 0.21223185133265576,
+      "learning_rate": 2.544761441944139e-07,
+      "loss": 0.2999,
+      "step": 2073
+    },
+    {
+      "epoch": 15.593984962406015,
+      "grad_norm": 0.20154762340852844,
+      "learning_rate": 2.536484742955878e-07,
+      "loss": 0.2717,
+      "step": 2074
+    },
+    {
+      "epoch": 15.601503759398497,
+      "grad_norm": 0.20205348349339933,
+      "learning_rate": 2.5282195702425655e-07,
+      "loss": 0.2747,
+      "step": 2075
+    },
+    {
+      "epoch": 15.609022556390977,
+      "grad_norm": 0.21467423546309633,
+      "learning_rate": 2.5199659365685235e-07,
+      "loss": 0.2744,
+      "step": 2076
+    },
+    {
+      "epoch": 15.61654135338346,
+      "grad_norm": 0.2013717376777752,
+      "learning_rate": 2.511723854680239e-07,
+      "loss": 0.2846,
+      "step": 2077
+    },
+    {
+      "epoch": 15.62406015037594,
+      "grad_norm": 0.19988958803873705,
+      "learning_rate": 2.5034933373063726e-07,
+      "loss": 0.2622,
+      "step": 2078
+    },
+    {
+      "epoch": 15.631578947368421,
+      "grad_norm": 0.20540579042207208,
+      "learning_rate": 2.495274397157713e-07,
+      "loss": 0.2719,
+      "step": 2079
+    },
+    {
+      "epoch": 15.639097744360903,
+      "grad_norm": 0.2051985577789569,
+      "learning_rate": 2.487067046927178e-07,
+      "loss": 0.2771,
+      "step": 2080
+    },
+    {
+      "epoch": 15.646616541353383,
+      "grad_norm": 0.22186562559806977,
+      "learning_rate": 2.478871299289781e-07,
+      "loss": 0.2708,
+      "step": 2081
+    },
+    {
+      "epoch": 15.654135338345865,
+      "grad_norm": 0.21492042424017285,
+      "learning_rate": 2.470687166902622e-07,
+      "loss": 0.2779,
+      "step": 2082
+    },
+    {
+      "epoch": 15.661654135338345,
+      "grad_norm": 0.2033827448781186,
+      "learning_rate": 2.462514662404862e-07,
+      "loss": 0.2728,
+      "step": 2083
+    },
+    {
+      "epoch": 15.669172932330827,
+      "grad_norm": 0.20199475075361722,
+      "learning_rate": 2.454353798417698e-07,
+      "loss": 0.2777,
+      "step": 2084
+    },
+    {
+      "epoch": 15.676691729323307,
+      "grad_norm": 0.20132935784674022,
+      "learning_rate": 2.4462045875443604e-07,
+      "loss": 0.2708,
+      "step": 2085
+    },
+    {
+      "epoch": 15.68421052631579,
+      "grad_norm": 0.2101815175973166,
+      "learning_rate": 2.438067042370072e-07,
+      "loss": 0.2749,
+      "step": 2086
+    },
+    {
+      "epoch": 15.691729323308271,
+      "grad_norm": 0.20043303445734903,
+      "learning_rate": 2.4299411754620526e-07,
+      "loss": 0.2708,
+      "step": 2087
+    },
+    {
+      "epoch": 15.699248120300751,
+      "grad_norm": 0.19050671165588434,
+      "learning_rate": 2.421826999369473e-07,
+      "loss": 0.2605,
+      "step": 2088
+    },
+    {
+      "epoch": 15.706766917293233,
+      "grad_norm": 0.2286006681235388,
+      "learning_rate": 2.4137245266234593e-07,
+      "loss": 0.2727,
+      "step": 2089
+    },
+    {
+      "epoch": 15.714285714285714,
+      "grad_norm": 0.22068311402242474,
+      "learning_rate": 2.4056337697370587e-07,
+      "loss": 0.2852,
+      "step": 2090
+    },
+    {
+      "epoch": 15.721804511278195,
+      "grad_norm": 0.20308757846287054,
+      "learning_rate": 2.3975547412052275e-07,
+      "loss": 0.2656,
+      "step": 2091
+    },
+    {
+      "epoch": 15.729323308270677,
+      "grad_norm": 0.2091700970223572,
+      "learning_rate": 2.389487453504806e-07,
+      "loss": 0.2703,
+      "step": 2092
+    },
+    {
+      "epoch": 15.736842105263158,
+      "grad_norm": 0.20135501484504123,
+      "learning_rate": 2.3814319190945075e-07,
+      "loss": 0.2718,
+      "step": 2093
+    },
+    {
+      "epoch": 15.74436090225564,
+      "grad_norm": 0.19174489876109593,
+      "learning_rate": 2.373388150414889e-07,
+      "loss": 0.2665,
+      "step": 2094
+    },
+    {
+      "epoch": 15.75187969924812,
+      "grad_norm": 0.20527383807311156,
+      "learning_rate": 2.365356159888342e-07,
+      "loss": 0.2602,
+      "step": 2095
+    },
+    {
+      "epoch": 15.759398496240602,
+      "grad_norm": 0.2093451308573309,
+      "learning_rate": 2.3573359599190613e-07,
+      "loss": 0.283,
+      "step": 2096
+    },
+    {
+      "epoch": 15.766917293233083,
+      "grad_norm": 0.22100313801226482,
+      "learning_rate": 2.349327562893044e-07,
+      "loss": 0.2764,
+      "step": 2097
+    },
+    {
+      "epoch": 15.774436090225564,
+      "grad_norm": 0.1955245179364764,
+      "learning_rate": 2.3413309811780458e-07,
+      "loss": 0.2766,
+      "step": 2098
+    },
+    {
+      "epoch": 15.781954887218046,
+      "grad_norm": 0.20399580338878415,
+      "learning_rate": 2.3333462271235905e-07,
+      "loss": 0.2894,
+      "step": 2099
+    },
+    {
+      "epoch": 15.789473684210526,
+      "grad_norm": 0.21487723522442959,
+      "learning_rate": 2.3253733130609187e-07,
+      "loss": 0.2789,
+      "step": 2100
+    },
+    {
+      "epoch": 15.796992481203008,
+      "grad_norm": 0.21543372057407745,
+      "learning_rate": 2.3174122513030035e-07,
+      "loss": 0.2805,
+      "step": 2101
+    },
+    {
+      "epoch": 15.80451127819549,
+      "grad_norm": 0.21247133265252044,
+      "learning_rate": 2.3094630541444992e-07,
+      "loss": 0.2756,
+      "step": 2102
+    },
+    {
+      "epoch": 15.81203007518797,
+      "grad_norm": 0.21193703122318563,
+      "learning_rate": 2.301525733861749e-07,
+      "loss": 0.2846,
+      "step": 2103
+    },
+    {
+      "epoch": 15.819548872180452,
+      "grad_norm": 0.21164142259924398,
+      "learning_rate": 2.2936003027127415e-07,
+      "loss": 0.2772,
+      "step": 2104
+    },
+    {
+      "epoch": 15.827067669172932,
+      "grad_norm": 0.20504874797700212,
+      "learning_rate": 2.2856867729371178e-07,
+      "loss": 0.2799,
+      "step": 2105
+    },
+    {
+      "epoch": 15.834586466165414,
+      "grad_norm": 0.20753289218323281,
+      "learning_rate": 2.2777851567561267e-07,
+      "loss": 0.2809,
+      "step": 2106
+    },
+    {
+      "epoch": 15.842105263157894,
+      "grad_norm": 0.20579915589809744,
+      "learning_rate": 2.26989546637263e-07,
+      "loss": 0.2842,
+      "step": 2107
+    },
+    {
+      "epoch": 15.849624060150376,
+      "grad_norm": 0.2064977888176818,
+      "learning_rate": 2.2620177139710627e-07,
+      "loss": 0.2713,
+      "step": 2108
+    },
+    {
+      "epoch": 15.857142857142858,
+      "grad_norm": 0.19585990174019655,
+      "learning_rate": 2.2541519117174246e-07,
+      "loss": 0.2693,
+      "step": 2109
+    },
+    {
+      "epoch": 15.864661654135338,
+      "grad_norm": 0.21968280857975248,
+      "learning_rate": 2.246298071759266e-07,
+      "loss": 0.2827,
+      "step": 2110
+    },
+    {
+      "epoch": 15.87218045112782,
+      "grad_norm": 0.20600868320996946,
+      "learning_rate": 2.2384562062256562e-07,
+      "loss": 0.274,
+      "step": 2111
+    },
+    {
+      "epoch": 15.8796992481203,
+      "grad_norm": 0.21166291258349695,
+      "learning_rate": 2.2306263272271787e-07,
+      "loss": 0.2747,
+      "step": 2112
+    },
+    {
+      "epoch": 15.887218045112782,
+      "grad_norm": 0.21286326204763406,
+      "learning_rate": 2.2228084468558984e-07,
+      "loss": 0.283,
+      "step": 2113
+    },
+    {
+      "epoch": 15.894736842105264,
+      "grad_norm": 0.20812279911177736,
+      "learning_rate": 2.2150025771853588e-07,
+      "loss": 0.2776,
+      "step": 2114
+    },
+    {
+      "epoch": 15.902255639097744,
+      "grad_norm": 0.20320287846751706,
+      "learning_rate": 2.2072087302705423e-07,
+      "loss": 0.264,
+      "step": 2115
+    },
+    {
+      "epoch": 15.909774436090226,
+      "grad_norm": 0.20407067801261464,
+      "learning_rate": 2.1994269181478798e-07,
+      "loss": 0.275,
+      "step": 2116
+    },
+    {
+      "epoch": 15.917293233082706,
+      "grad_norm": 0.20678476397672502,
+      "learning_rate": 2.1916571528352002e-07,
+      "loss": 0.2769,
+      "step": 2117
+    },
+    {
+      "epoch": 15.924812030075188,
+      "grad_norm": 0.2009050382699085,
+      "learning_rate": 2.1838994463317417e-07,
+      "loss": 0.283,
+      "step": 2118
+    },
+    {
+      "epoch": 15.93233082706767,
+      "grad_norm": 0.19299181735454204,
+      "learning_rate": 2.1761538106181076e-07,
+      "loss": 0.2794,
+      "step": 2119
+    },
+    {
+      "epoch": 15.93984962406015,
+      "grad_norm": 0.20243143216323323,
+      "learning_rate": 2.1684202576562717e-07,
+      "loss": 0.2752,
+      "step": 2120
+    },
+    {
+      "epoch": 15.947368421052632,
+      "grad_norm": 0.20371562251690437,
+      "learning_rate": 2.1606987993895353e-07,
+      "loss": 0.2778,
+      "step": 2121
+    },
+    {
+      "epoch": 15.954887218045112,
+      "grad_norm": 0.2124718871327285,
+      "learning_rate": 2.1529894477425327e-07,
+      "loss": 0.2791,
+      "step": 2122
+    },
+    {
+      "epoch": 15.962406015037594,
+      "grad_norm": 0.2078235350177689,
+      "learning_rate": 2.1452922146211916e-07,
+      "loss": 0.27,
+      "step": 2123
+    },
+    {
+      "epoch": 15.969924812030076,
+      "grad_norm": 0.21778179332447622,
+      "learning_rate": 2.1376071119127337e-07,
+      "loss": 0.2815,
+      "step": 2124
+    },
+    {
+      "epoch": 15.977443609022556,
+      "grad_norm": 0.19939782334682232,
+      "learning_rate": 2.1299341514856363e-07,
+      "loss": 0.2749,
+      "step": 2125
+    },
+    {
+      "epoch": 15.984962406015038,
+      "grad_norm": 0.20390885443896184,
+      "learning_rate": 2.122273345189638e-07,
+      "loss": 0.2765,
+      "step": 2126
+    },
+    {
+      "epoch": 15.992481203007518,
+      "grad_norm": 0.20149234677017613,
+      "learning_rate": 2.1146247048556932e-07,
+      "loss": 0.2721,
+      "step": 2127
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 0.22362178099285013,
+      "learning_rate": 2.1069882422959807e-07,
+      "loss": 0.2808,
+      "step": 2128
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 0.4246521294116974,
+      "eval_runtime": 1.5233,
+      "eval_samples_per_second": 214.661,
+      "eval_steps_per_second": 3.939,
+      "step": 2128
+    },
+    {
+      "epoch": 16.007518796992482,
+      "grad_norm": 0.20076860390902365,
+      "learning_rate": 2.099363969303861e-07,
+      "loss": 0.2784,
+      "step": 2129
+    },
+    {
+      "epoch": 16.015037593984964,
+      "grad_norm": 0.2018204379657976,
+      "learning_rate": 2.0917518976538807e-07,
+      "loss": 0.2707,
+      "step": 2130
+    },
+    {
+      "epoch": 16.022556390977442,
+      "grad_norm": 0.2037207114672595,
+      "learning_rate": 2.084152039101732e-07,
+      "loss": 0.2767,
+      "step": 2131
+    },
+    {
+      "epoch": 16.030075187969924,
+      "grad_norm": 0.19293110851244105,
+      "learning_rate": 2.0765644053842578e-07,
+      "loss": 0.2727,
+      "step": 2132
+    },
+    {
+      "epoch": 16.037593984962406,
+      "grad_norm": 0.2071106886856524,
+      "learning_rate": 2.0689890082194083e-07,
+      "loss": 0.273,
+      "step": 2133
+    },
+    {
+      "epoch": 16.045112781954888,
+      "grad_norm": 0.1961530858650501,
+      "learning_rate": 2.0614258593062493e-07,
+      "loss": 0.272,
+      "step": 2134
+    },
+    {
+      "epoch": 16.05263157894737,
+      "grad_norm": 0.19872766344652157,
+      "learning_rate": 2.0538749703249236e-07,
+      "loss": 0.277,
+      "step": 2135
+    },
+    {
+      "epoch": 16.06015037593985,
+      "grad_norm": 0.20231932314527115,
+      "learning_rate": 2.0463363529366373e-07,
+      "loss": 0.2716,
+      "step": 2136
+    },
+    {
+      "epoch": 16.06766917293233,
+      "grad_norm": 0.19641064315287074,
+      "learning_rate": 2.0388100187836554e-07,
+      "loss": 0.2757,
+      "step": 2137
+    },
+    {
+      "epoch": 16.075187969924812,
+      "grad_norm": 0.2079203994360354,
+      "learning_rate": 2.0312959794892615e-07,
+      "loss": 0.2779,
+      "step": 2138
+    },
+    {
+      "epoch": 16.082706766917294,
+      "grad_norm": 0.19753603642243103,
+      "learning_rate": 2.0237942466577617e-07,
+      "loss": 0.2783,
+      "step": 2139
+    },
+    {
+      "epoch": 16.090225563909776,
+      "grad_norm": 0.20191637811941376,
+      "learning_rate": 2.0163048318744492e-07,
+      "loss": 0.2633,
+      "step": 2140
+    },
+    {
+      "epoch": 16.097744360902254,
+      "grad_norm": 0.20409559589715218,
+      "learning_rate": 2.0088277467056013e-07,
+      "loss": 0.2762,
+      "step": 2141
+    },
+    {
+      "epoch": 16.105263157894736,
+      "grad_norm": 0.20789157507230843,
+      "learning_rate": 2.001363002698443e-07,
+      "loss": 0.2705,
+      "step": 2142
+    },
+    {
+      "epoch": 16.112781954887218,
+      "grad_norm": 0.19763310877472642,
+      "learning_rate": 1.9939106113811544e-07,
+      "loss": 0.2786,
+      "step": 2143
+    },
+    {
+      "epoch": 16.1203007518797,
+      "grad_norm": 0.2099001170010665,
+      "learning_rate": 1.9864705842628237e-07,
+      "loss": 0.2787,
+      "step": 2144
+    },
+    {
+      "epoch": 16.127819548872182,
+      "grad_norm": 0.19357286854035843,
+      "learning_rate": 1.9790429328334592e-07,
+      "loss": 0.2729,
+      "step": 2145
+    },
+    {
+      "epoch": 16.13533834586466,
+      "grad_norm": 0.21079440182307568,
+      "learning_rate": 1.9716276685639422e-07,
+      "loss": 0.2673,
+      "step": 2146
+    },
+    {
+      "epoch": 16.142857142857142,
+      "grad_norm": 0.21416871191846915,
+      "learning_rate": 1.9642248029060383e-07,
+      "loss": 0.2791,
+      "step": 2147
+    },
+    {
+      "epoch": 16.150375939849624,
+      "grad_norm": 0.22014757273327806,
+      "learning_rate": 1.956834347292352e-07,
+      "loss": 0.286,
+      "step": 2148
+    },
+    {
+      "epoch": 16.157894736842106,
+      "grad_norm": 0.21610864757397255,
+      "learning_rate": 1.949456313136335e-07,
+      "loss": 0.2662,
+      "step": 2149
+    },
+    {
+      "epoch": 16.165413533834588,
+      "grad_norm": 0.18640182643641376,
+      "learning_rate": 1.9420907118322427e-07,
+      "loss": 0.2646,
+      "step": 2150
+    },
+    {
+      "epoch": 16.172932330827066,
+      "grad_norm": 0.22513771900664675,
+      "learning_rate": 1.9347375547551436e-07,
+      "loss": 0.2821,
+      "step": 2151
+    },
+    {
+      "epoch": 16.18045112781955,
+      "grad_norm": 0.19756151927794705,
+      "learning_rate": 1.9273968532608753e-07,
+      "loss": 0.2682,
+      "step": 2152
+    },
+    {
+      "epoch": 16.18796992481203,
+      "grad_norm": 0.2089709548983576,
+      "learning_rate": 1.9200686186860492e-07,
+      "loss": 0.2658,
+      "step": 2153
+    },
+    {
+      "epoch": 16.195488721804512,
+      "grad_norm": 0.19745619836783437,
+      "learning_rate": 1.9127528623480172e-07,
+      "loss": 0.2644,
+      "step": 2154
+    },
+    {
+      "epoch": 16.203007518796994,
+      "grad_norm": 0.2011548298095291,
+      "learning_rate": 1.9054495955448656e-07,
+      "loss": 0.2748,
+      "step": 2155
+    },
+    {
+      "epoch": 16.210526315789473,
+      "grad_norm": 0.20760525191643175,
+      "learning_rate": 1.898158829555385e-07,
+      "loss": 0.2805,
+      "step": 2156
+    },
+    {
+      "epoch": 16.218045112781954,
+      "grad_norm": 0.20935765644977122,
+      "learning_rate": 1.890880575639072e-07,
+      "loss": 0.2848,
+      "step": 2157
+    },
+    {
+      "epoch": 16.225563909774436,
+      "grad_norm": 0.19793658697930785,
+      "learning_rate": 1.8836148450360866e-07,
+      "loss": 0.2809,
+      "step": 2158
+    },
+    {
+      "epoch": 16.23308270676692,
+      "grad_norm": 0.20429306087140903,
+      "learning_rate": 1.8763616489672608e-07,
+      "loss": 0.2633,
+      "step": 2159
+    },
+    {
+      "epoch": 16.2406015037594,
+      "grad_norm": 0.19889073802597737,
+      "learning_rate": 1.8691209986340595e-07,
+      "loss": 0.2635,
+      "step": 2160
+    },
+    {
+      "epoch": 16.24812030075188,
+      "grad_norm": 0.20316878425182733,
+      "learning_rate": 1.861892905218575e-07,
+      "loss": 0.279,
+      "step": 2161
+    },
+    {
+      "epoch": 16.25563909774436,
+      "grad_norm": 0.20369990876332597,
+      "learning_rate": 1.8546773798835148e-07,
+      "loss": 0.2747,
+      "step": 2162
+    },
+    {
+      "epoch": 16.263157894736842,
+      "grad_norm": 0.20715648624086466,
+      "learning_rate": 1.8474744337721638e-07,
+      "loss": 0.2754,
+      "step": 2163
+    },
+    {
+      "epoch": 16.270676691729324,
+      "grad_norm": 0.2023092604302329,
+      "learning_rate": 1.8402840780083927e-07,
+      "loss": 0.279,
+      "step": 2164
+    },
+    {
+      "epoch": 16.278195488721803,
+      "grad_norm": 0.205672562903412,
+      "learning_rate": 1.833106323696617e-07,
+      "loss": 0.2802,
+      "step": 2165
+    },
+    {
+      "epoch": 16.285714285714285,
+      "grad_norm": 0.19493532925443421,
+      "learning_rate": 1.825941181921805e-07,
+      "loss": 0.2706,
+      "step": 2166
+    },
+    {
+      "epoch": 16.293233082706767,
+      "grad_norm": 0.2065717993371595,
+      "learning_rate": 1.8187886637494297e-07,
+      "loss": 0.2793,
+      "step": 2167
+    },
+    {
+      "epoch": 16.30075187969925,
+      "grad_norm": 0.19764381061196176,
+      "learning_rate": 1.8116487802254865e-07,
+      "loss": 0.2742,
+      "step": 2168
+    },
+    {
+      "epoch": 16.30827067669173,
+      "grad_norm": 0.1992628697940465,
+      "learning_rate": 1.8045215423764426e-07,
+      "loss": 0.2754,
+      "step": 2169
+    },
+    {
+      "epoch": 16.31578947368421,
+      "grad_norm": 0.18577541480587884,
+      "learning_rate": 1.7974069612092478e-07,
+      "loss": 0.2738,
+      "step": 2170
+    },
+    {
+      "epoch": 16.32330827067669,
+      "grad_norm": 0.197463006450585,
+      "learning_rate": 1.790305047711298e-07,
+      "loss": 0.2719,
+      "step": 2171
+    },
+    {
+      "epoch": 16.330827067669173,
+      "grad_norm": 0.22056559793485764,
+      "learning_rate": 1.7832158128504328e-07,
+      "loss": 0.2805,
+      "step": 2172
+    },
+    {
+      "epoch": 16.338345864661655,
+      "grad_norm": 0.19720033093920675,
+      "learning_rate": 1.776139267574901e-07,
+      "loss": 0.2731,
+      "step": 2173
+    },
+    {
+      "epoch": 16.345864661654137,
+      "grad_norm": 0.19869231596922665,
+      "learning_rate": 1.7690754228133688e-07,
+      "loss": 0.2762,
+      "step": 2174
+    },
+    {
+      "epoch": 16.353383458646615,
+      "grad_norm": 0.2247126870378317,
+      "learning_rate": 1.7620242894748716e-07,
+      "loss": 0.2842,
+      "step": 2175
+    },
+    {
+      "epoch": 16.360902255639097,
+      "grad_norm": 0.19650125647262476,
+      "learning_rate": 1.7549858784488314e-07,
+      "loss": 0.2752,
+      "step": 2176
+    },
+    {
+      "epoch": 16.36842105263158,
+      "grad_norm": 0.20995861048190104,
+      "learning_rate": 1.7479602006050054e-07,
+      "loss": 0.2759,
+      "step": 2177
+    },
+    {
+      "epoch": 16.37593984962406,
+      "grad_norm": 0.19552702528932359,
+      "learning_rate": 1.740947266793501e-07,
+      "loss": 0.2731,
+      "step": 2178
+    },
+    {
+      "epoch": 16.383458646616543,
+      "grad_norm": 0.21112464790645205,
+      "learning_rate": 1.7339470878447337e-07,
+      "loss": 0.2869,
+      "step": 2179
+    },
+    {
+      "epoch": 16.39097744360902,
+      "grad_norm": 0.2008883336298437,
+      "learning_rate": 1.7269596745694292e-07,
+      "loss": 0.2723,
+      "step": 2180
+    },
+    {
+      "epoch": 16.398496240601503,
+      "grad_norm": 0.20635069180727705,
+      "learning_rate": 1.71998503775859e-07,
+      "loss": 0.2796,
+      "step": 2181
+    },
+    {
+      "epoch": 16.406015037593985,
+      "grad_norm": 0.20266183662857984,
+      "learning_rate": 1.713023188183498e-07,
+      "loss": 0.2682,
+      "step": 2182
+    },
+    {
+      "epoch": 16.413533834586467,
+      "grad_norm": 0.19864760535720768,
+      "learning_rate": 1.7060741365956743e-07,
+      "loss": 0.2651,
+      "step": 2183
+    },
+    {
+      "epoch": 16.42105263157895,
+      "grad_norm": 0.1890706888484189,
+      "learning_rate": 1.6991378937268886e-07,
+      "loss": 0.268,
+      "step": 2184
+    },
+    {
+      "epoch": 16.428571428571427,
+      "grad_norm": 0.21104818615864102,
+      "learning_rate": 1.6922144702891173e-07,
+      "loss": 0.2724,
+      "step": 2185
+    },
+    {
+      "epoch": 16.43609022556391,
+      "grad_norm": 0.20568816441378973,
+      "learning_rate": 1.6853038769745465e-07,
+      "loss": 0.2827,
+      "step": 2186
+    },
+    {
+      "epoch": 16.44360902255639,
+      "grad_norm": 0.19689203440186048,
+      "learning_rate": 1.6784061244555513e-07,
+      "loss": 0.2705,
+      "step": 2187
+    },
+    {
+      "epoch": 16.451127819548873,
+      "grad_norm": 0.19049079670289132,
+      "learning_rate": 1.6715212233846654e-07,
+      "loss": 0.2646,
+      "step": 2188
+    },
+    {
+      "epoch": 16.458646616541355,
+      "grad_norm": 0.19395203997181473,
+      "learning_rate": 1.6646491843945853e-07,
+      "loss": 0.2692,
+      "step": 2189
+    },
+    {
+      "epoch": 16.466165413533833,
+      "grad_norm": 0.20981707616744413,
+      "learning_rate": 1.6577900180981363e-07,
+      "loss": 0.2871,
+      "step": 2190
+    },
+    {
+      "epoch": 16.473684210526315,
+      "grad_norm": 0.2074943659134334,
+      "learning_rate": 1.6509437350882716e-07,
+      "loss": 0.2701,
+      "step": 2191
+    },
+    {
+      "epoch": 16.481203007518797,
+      "grad_norm": 0.2049312682099538,
+      "learning_rate": 1.644110345938039e-07,
+      "loss": 0.2862,
+      "step": 2192
+    },
+    {
+      "epoch": 16.48872180451128,
+      "grad_norm": 0.19644871628632343,
+      "learning_rate": 1.6372898612005837e-07,
+      "loss": 0.2882,
+      "step": 2193
+    },
+    {
+      "epoch": 16.49624060150376,
+      "grad_norm": 0.2000672433974904,
+      "learning_rate": 1.6304822914091132e-07,
+      "loss": 0.2799,
+      "step": 2194
+    },
+    {
+      "epoch": 16.50375939849624,
+      "grad_norm": 0.2094870720242447,
+      "learning_rate": 1.6236876470768958e-07,
+      "loss": 0.2757,
+      "step": 2195
+    },
+    {
+      "epoch": 16.51127819548872,
+      "grad_norm": 0.19596350296628412,
+      "learning_rate": 1.616905938697234e-07,
+      "loss": 0.2675,
+      "step": 2196
+    },
+    {
+      "epoch": 16.518796992481203,
+      "grad_norm": 0.2012357351840834,
+      "learning_rate": 1.610137176743457e-07,
+      "loss": 0.2786,
+      "step": 2197
+    },
+    {
+      "epoch": 16.526315789473685,
+      "grad_norm": 0.19794726645204183,
+      "learning_rate": 1.6033813716688948e-07,
+      "loss": 0.2784,
+      "step": 2198
+    },
+    {
+      "epoch": 16.533834586466167,
+      "grad_norm": 0.19655602832047978,
+      "learning_rate": 1.5966385339068756e-07,
+      "loss": 0.2829,
+      "step": 2199
+    },
+    {
+      "epoch": 16.541353383458645,
+      "grad_norm": 0.19818663982909818,
+      "learning_rate": 1.58990867387069e-07,
+      "loss": 0.2836,
+      "step": 2200
+    },
+    {
+      "epoch": 16.548872180451127,
+      "grad_norm": 0.2049153533592169,
+      "learning_rate": 1.5831918019535994e-07,
+      "loss": 0.2801,
+      "step": 2201
+    },
+    {
+      "epoch": 16.55639097744361,
+      "grad_norm": 0.20294600066754848,
+      "learning_rate": 1.5764879285287946e-07,
+      "loss": 0.2721,
+      "step": 2202
+    },
+    {
+      "epoch": 16.56390977443609,
+      "grad_norm": 0.21178260616362152,
+      "learning_rate": 1.569797063949404e-07,
+      "loss": 0.2759,
+      "step": 2203
+    },
+    {
+      "epoch": 16.571428571428573,
+      "grad_norm": 0.19714916812866012,
+      "learning_rate": 1.5631192185484554e-07,
+      "loss": 0.2691,
+      "step": 2204
+    },
+    {
+      "epoch": 16.57894736842105,
+      "grad_norm": 0.21007810822808234,
+      "learning_rate": 1.5564544026388792e-07,
+      "loss": 0.2768,
+      "step": 2205
+    },
+    {
+      "epoch": 16.586466165413533,
+      "grad_norm": 0.21764613684932604,
+      "learning_rate": 1.5498026265134745e-07,
+      "loss": 0.2805,
+      "step": 2206
+    },
+    {
+      "epoch": 16.593984962406015,
+      "grad_norm": 0.20062951607549598,
+      "learning_rate": 1.5431639004449125e-07,
+      "loss": 0.2761,
+      "step": 2207
+    },
+    {
+      "epoch": 16.601503759398497,
+      "grad_norm": 0.19646506887556914,
+      "learning_rate": 1.5365382346857002e-07,
+      "loss": 0.2619,
+      "step": 2208
+    },
+    {
+      "epoch": 16.60902255639098,
+      "grad_norm": 0.19196815047090468,
+      "learning_rate": 1.529925639468186e-07,
+      "loss": 0.2674,
+      "step": 2209
+    },
+    {
+      "epoch": 16.616541353383457,
+      "grad_norm": 0.20872749263088042,
+      "learning_rate": 1.5233261250045215e-07,
+      "loss": 0.2712,
+      "step": 2210
+    },
+    {
+      "epoch": 16.62406015037594,
+      "grad_norm": 0.209205225035663,
+      "learning_rate": 1.5167397014866679e-07,
+      "loss": 0.2652,
+      "step": 2211
+    },
+    {
+      "epoch": 16.63157894736842,
+      "grad_norm": 0.2036470403138517,
+      "learning_rate": 1.5101663790863595e-07,
+      "loss": 0.2755,
+      "step": 2212
+    },
+    {
+      "epoch": 16.639097744360903,
+      "grad_norm": 0.2440834319225481,
+      "learning_rate": 1.503606167955107e-07,
+      "loss": 0.2653,
+      "step": 2213
+    },
+    {
+      "epoch": 16.646616541353385,
+      "grad_norm": 0.21332145362780014,
+      "learning_rate": 1.4970590782241643e-07,
+      "loss": 0.2806,
+      "step": 2214
+    },
+    {
+      "epoch": 16.654135338345863,
+      "grad_norm": 0.21046973624767812,
+      "learning_rate": 1.4905251200045254e-07,
+      "loss": 0.2808,
+      "step": 2215
+    },
+    {
+      "epoch": 16.661654135338345,
+      "grad_norm": 0.21335306254058128,
+      "learning_rate": 1.4840043033869076e-07,
+      "loss": 0.2828,
+      "step": 2216
+    },
+    {
+      "epoch": 16.669172932330827,
+      "grad_norm": 0.19750978984797607,
+      "learning_rate": 1.4774966384417252e-07,
+      "loss": 0.2767,
+      "step": 2217
+    },
+    {
+      "epoch": 16.67669172932331,
+      "grad_norm": 0.20395271227762615,
+      "learning_rate": 1.4710021352190916e-07,
+      "loss": 0.2778,
+      "step": 2218
+    },
+    {
+      "epoch": 16.68421052631579,
+      "grad_norm": 0.2043233354686783,
+      "learning_rate": 1.4645208037487843e-07,
+      "loss": 0.2941,
+      "step": 2219
+    },
+    {
+      "epoch": 16.69172932330827,
+      "grad_norm": 0.2016616172597907,
+      "learning_rate": 1.4580526540402461e-07,
+      "loss": 0.2824,
+      "step": 2220
+    },
+    {
+      "epoch": 16.69924812030075,
+      "grad_norm": 0.20008787746411977,
+      "learning_rate": 1.451597696082557e-07,
+      "loss": 0.2765,
+      "step": 2221
+    },
+    {
+      "epoch": 16.706766917293233,
+      "grad_norm": 0.19387960865086667,
+      "learning_rate": 1.4451559398444313e-07,
+      "loss": 0.2704,
+      "step": 2222
+    },
+    {
+      "epoch": 16.714285714285715,
+      "grad_norm": 0.19418102394282574,
+      "learning_rate": 1.4387273952741863e-07,
+      "loss": 0.278,
+      "step": 2223
+    },
+    {
+      "epoch": 16.721804511278194,
+      "grad_norm": 0.21248816217810673,
+      "learning_rate": 1.432312072299746e-07,
+      "loss": 0.2862,
+      "step": 2224
+    },
+    {
+      "epoch": 16.729323308270676,
+      "grad_norm": 0.20225620373680828,
+      "learning_rate": 1.4259099808286047e-07,
+      "loss": 0.278,
+      "step": 2225
+    },
+    {
+      "epoch": 16.736842105263158,
+      "grad_norm": 0.20070081395434508,
+      "learning_rate": 1.4195211307478328e-07,
+      "loss": 0.2741,
+      "step": 2226
+    },
+    {
+      "epoch": 16.74436090225564,
+      "grad_norm": 0.19975921334311175,
+      "learning_rate": 1.4131455319240426e-07,
+      "loss": 0.2788,
+      "step": 2227
+    },
+    {
+      "epoch": 16.75187969924812,
+      "grad_norm": 0.20568711466994424,
+      "learning_rate": 1.4067831942033902e-07,
+      "loss": 0.2737,
+      "step": 2228
+    },
+    {
+      "epoch": 16.7593984962406,
+      "grad_norm": 0.19109956021612,
+      "learning_rate": 1.4004341274115438e-07,
+      "loss": 0.2655,
+      "step": 2229
+    },
+    {
+      "epoch": 16.76691729323308,
+      "grad_norm": 0.21669051881230525,
+      "learning_rate": 1.3940983413536845e-07,
+      "loss": 0.2678,
+      "step": 2230
+    },
+    {
+      "epoch": 16.774436090225564,
+      "grad_norm": 0.20602149953936,
+      "learning_rate": 1.3877758458144762e-07,
+      "loss": 0.2768,
+      "step": 2231
+    },
+    {
+      "epoch": 16.781954887218046,
+      "grad_norm": 0.20713521746300007,
+      "learning_rate": 1.381466650558063e-07,
+      "loss": 0.2778,
+      "step": 2232
+    },
+    {
+      "epoch": 16.789473684210527,
+      "grad_norm": 0.20351784475728218,
+      "learning_rate": 1.3751707653280443e-07,
+      "loss": 0.2769,
+      "step": 2233
+    },
+    {
+      "epoch": 16.796992481203006,
+      "grad_norm": 0.1987446711701425,
+      "learning_rate": 1.3688881998474699e-07,
+      "loss": 0.277,
+      "step": 2234
+    },
+    {
+      "epoch": 16.804511278195488,
+      "grad_norm": 0.19699081594526194,
+      "learning_rate": 1.3626189638188102e-07,
+      "loss": 0.282,
+      "step": 2235
+    },
+    {
+      "epoch": 16.81203007518797,
+      "grad_norm": 0.19791400563559391,
+      "learning_rate": 1.3563630669239624e-07,
+      "loss": 0.2631,
+      "step": 2236
+    },
+    {
+      "epoch": 16.81954887218045,
+      "grad_norm": 0.20362031627458552,
+      "learning_rate": 1.3501205188242105e-07,
+      "loss": 0.2663,
+      "step": 2237
+    },
+    {
+      "epoch": 16.827067669172934,
+      "grad_norm": 0.20867083854449123,
+      "learning_rate": 1.343891329160235e-07,
+      "loss": 0.275,
+      "step": 2238
+    },
+    {
+      "epoch": 16.834586466165412,
+      "grad_norm": 0.1975810661390982,
+      "learning_rate": 1.3376755075520785e-07,
+      "loss": 0.2822,
+      "step": 2239
+    },
+    {
+      "epoch": 16.842105263157894,
+      "grad_norm": 0.22066652788139085,
+      "learning_rate": 1.331473063599139e-07,
+      "loss": 0.2724,
+      "step": 2240
+    },
+    {
+      "epoch": 16.849624060150376,
+      "grad_norm": 0.19678646837431,
+      "learning_rate": 1.3252840068801607e-07,
+      "loss": 0.282,
+      "step": 2241
+    },
+    {
+      "epoch": 16.857142857142858,
+      "grad_norm": 0.20899674700981935,
+      "learning_rate": 1.3191083469532061e-07,
+      "loss": 0.2714,
+      "step": 2242
+    },
+    {
+      "epoch": 16.86466165413534,
+      "grad_norm": 0.1968501750049605,
+      "learning_rate": 1.3129460933556547e-07,
+      "loss": 0.2626,
+      "step": 2243
+    },
+    {
+      "epoch": 16.872180451127818,
+      "grad_norm": 0.2044810478606153,
+      "learning_rate": 1.306797255604175e-07,
+      "loss": 0.2769,
+      "step": 2244
+    },
+    {
+      "epoch": 16.8796992481203,
+      "grad_norm": 0.192262662252086,
+      "learning_rate": 1.3006618431947248e-07,
+      "loss": 0.2683,
+      "step": 2245
+    },
+    {
+      "epoch": 16.887218045112782,
+      "grad_norm": 0.2063014791082304,
+      "learning_rate": 1.294539865602521e-07,
+      "loss": 0.2749,
+      "step": 2246
+    },
+    {
+      "epoch": 16.894736842105264,
+      "grad_norm": 0.2012008836762921,
+      "learning_rate": 1.2884313322820385e-07,
+      "loss": 0.2762,
+      "step": 2247
+    },
+    {
+      "epoch": 16.902255639097746,
+      "grad_norm": 0.20463838217452232,
+      "learning_rate": 1.2823362526669822e-07,
+      "loss": 0.283,
+      "step": 2248
+    },
+    {
+      "epoch": 16.909774436090224,
+      "grad_norm": 0.20297051900845292,
+      "learning_rate": 1.2762546361702908e-07,
+      "loss": 0.275,
+      "step": 2249
+    },
+    {
+      "epoch": 16.917293233082706,
+      "grad_norm": 0.20073127231449536,
+      "learning_rate": 1.2701864921840989e-07,
+      "loss": 0.2703,
+      "step": 2250
+    },
+    {
+      "epoch": 16.924812030075188,
+      "grad_norm": 0.20080959836868048,
+      "learning_rate": 1.2641318300797453e-07,
+      "loss": 0.2794,
+      "step": 2251
+    },
+    {
+      "epoch": 16.93233082706767,
+      "grad_norm": 0.20208302940490316,
+      "learning_rate": 1.25809065920774e-07,
+      "loss": 0.2732,
+      "step": 2252
+    },
+    {
+      "epoch": 16.93984962406015,
+      "grad_norm": 0.20233228926813535,
+      "learning_rate": 1.252062988897764e-07,
+      "loss": 0.2771,
+      "step": 2253
+    },
+    {
+      "epoch": 16.94736842105263,
+      "grad_norm": 0.20732189052614086,
+      "learning_rate": 1.2460488284586435e-07,
+      "loss": 0.272,
+      "step": 2254
+    },
+    {
+      "epoch": 16.954887218045112,
+      "grad_norm": 0.2171206404112525,
+      "learning_rate": 1.2400481871783465e-07,
+      "loss": 0.2827,
+      "step": 2255
+    },
+    {
+      "epoch": 16.962406015037594,
+      "grad_norm": 0.22059442103381288,
+      "learning_rate": 1.2340610743239542e-07,
+      "loss": 0.2682,
+      "step": 2256
+    },
+    {
+      "epoch": 16.969924812030076,
+      "grad_norm": 0.21274524644866136,
+      "learning_rate": 1.2280874991416668e-07,
+      "loss": 0.286,
+      "step": 2257
+    },
+    {
+      "epoch": 16.977443609022558,
+      "grad_norm": 0.21037946302269153,
+      "learning_rate": 1.2221274708567663e-07,
+      "loss": 0.2809,
+      "step": 2258
+    },
+    {
+      "epoch": 16.984962406015036,
+      "grad_norm": 0.19511399479831715,
+      "learning_rate": 1.2161809986736228e-07,
+      "loss": 0.2633,
+      "step": 2259
+    },
+    {
+      "epoch": 16.992481203007518,
+      "grad_norm": 0.2801616150723427,
+      "learning_rate": 1.210248091775663e-07,
+      "loss": 0.2701,
+      "step": 2260
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 0.1973721128214385,
+      "learning_rate": 1.2043287593253703e-07,
+      "loss": 0.2726,
+      "step": 2261
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 0.42482516169548035,
+      "eval_runtime": 1.662,
+      "eval_samples_per_second": 196.75,
+      "eval_steps_per_second": 3.61,
+      "step": 2261
+    },
+    {
+      "epoch": 17.007518796992482,
+      "grad_norm": 0.2008598640188869,
+      "learning_rate": 1.198423010464259e-07,
+      "loss": 0.2821,
+      "step": 2262
+    },
+    {
+      "epoch": 17.015037593984964,
+      "grad_norm": 0.19767822084115713,
+      "learning_rate": 1.1925308543128732e-07,
+      "loss": 0.2781,
+      "step": 2263
+    },
+    {
+      "epoch": 17.022556390977442,
+      "grad_norm": 0.22377985174997972,
+      "learning_rate": 1.1866522999707551e-07,
+      "loss": 0.2776,
+      "step": 2264
+    },
+    {
+      "epoch": 17.030075187969924,
+      "grad_norm": 0.20956361032649604,
+      "learning_rate": 1.1807873565164505e-07,
+      "loss": 0.2786,
+      "step": 2265
+    },
+    {
+      "epoch": 17.037593984962406,
+      "grad_norm": 0.2073801361501605,
+      "learning_rate": 1.1749360330074798e-07,
+      "loss": 0.2838,
+      "step": 2266
+    },
+    {
+      "epoch": 17.045112781954888,
+      "grad_norm": 0.19426895374849745,
+      "learning_rate": 1.1690983384803288e-07,
+      "loss": 0.2743,
+      "step": 2267
+    },
+    {
+      "epoch": 17.05263157894737,
+      "grad_norm": 0.20818396218102064,
+      "learning_rate": 1.1632742819504404e-07,
+      "loss": 0.2596,
+      "step": 2268
+    },
+    {
+      "epoch": 17.06015037593985,
+      "grad_norm": 0.20243462463514858,
+      "learning_rate": 1.1574638724121887e-07,
+      "loss": 0.2718,
+      "step": 2269
+    },
+    {
+      "epoch": 17.06766917293233,
+      "grad_norm": 0.20784825935343088,
+      "learning_rate": 1.1516671188388805e-07,
+      "loss": 0.2885,
+      "step": 2270
+    },
+    {
+      "epoch": 17.075187969924812,
+      "grad_norm": 0.20262175882717495,
+      "learning_rate": 1.1458840301827233e-07,
+      "loss": 0.2753,
+      "step": 2271
+    },
+    {
+      "epoch": 17.082706766917294,
+      "grad_norm": 0.21262817610423576,
+      "learning_rate": 1.140114615374831e-07,
+      "loss": 0.2801,
+      "step": 2272
+    },
+    {
+      "epoch": 17.090225563909776,
+      "grad_norm": 0.20156459028835183,
+      "learning_rate": 1.1343588833251928e-07,
+      "loss": 0.2687,
+      "step": 2273
+    },
+    {
+      "epoch": 17.097744360902254,
+      "grad_norm": 0.19507618331106927,
+      "learning_rate": 1.1286168429226717e-07,
+      "loss": 0.277,
+      "step": 2274
+    },
+    {
+      "epoch": 17.105263157894736,
+      "grad_norm": 0.19937731375745224,
+      "learning_rate": 1.122888503034981e-07,
+      "loss": 0.2686,
+      "step": 2275
+    },
+    {
+      "epoch": 17.112781954887218,
+      "grad_norm": 0.20643893166769367,
+      "learning_rate": 1.1171738725086832e-07,
+      "loss": 0.268,
+      "step": 2276
+    },
+    {
+      "epoch": 17.1203007518797,
+      "grad_norm": 0.21012387580301267,
+      "learning_rate": 1.1114729601691585e-07,
+      "loss": 0.2705,
+      "step": 2277
+    },
+    {
+      "epoch": 17.127819548872182,
+      "grad_norm": 0.19740556612449853,
+      "learning_rate": 1.1057857748206145e-07,
+      "loss": 0.2898,
+      "step": 2278
+    },
+    {
+      "epoch": 17.13533834586466,
+      "grad_norm": 0.19958889442977282,
+      "learning_rate": 1.1001123252460443e-07,
+      "loss": 0.2791,
+      "step": 2279
+    },
+    {
+      "epoch": 17.142857142857142,
+      "grad_norm": 0.1981220994158996,
+      "learning_rate": 1.0944526202072423e-07,
+      "loss": 0.2654,
+      "step": 2280
+    },
+    {
+      "epoch": 17.150375939849624,
+      "grad_norm": 0.20322241231713334,
+      "learning_rate": 1.0888066684447662e-07,
+      "loss": 0.2757,
+      "step": 2281
+    },
+    {
+      "epoch": 17.157894736842106,
+      "grad_norm": 0.2047832117274103,
+      "learning_rate": 1.0831744786779417e-07,
+      "loss": 0.2662,
+      "step": 2282
+    },
+    {
+      "epoch": 17.165413533834588,
+      "grad_norm": 0.20077395195949221,
+      "learning_rate": 1.0775560596048339e-07,
+      "loss": 0.2677,
+      "step": 2283
+    },
+    {
+      "epoch": 17.172932330827066,
+      "grad_norm": 0.20469910147805986,
+      "learning_rate": 1.0719514199022472e-07,
+      "loss": 0.2697,
+      "step": 2284
+    },
+    {
+      "epoch": 17.18045112781955,
+      "grad_norm": 0.20891649988307615,
+      "learning_rate": 1.0663605682257005e-07,
+      "loss": 0.2644,
+      "step": 2285
+    },
+    {
+      "epoch": 17.18796992481203,
+      "grad_norm": 0.21555487191612216,
+      "learning_rate": 1.0607835132094257e-07,
+      "loss": 0.2898,
+      "step": 2286
+    },
+    {
+      "epoch": 17.195488721804512,
+      "grad_norm": 0.1962133812821318,
+      "learning_rate": 1.055220263466341e-07,
+      "loss": 0.2758,
+      "step": 2287
+    },
+    {
+      "epoch": 17.203007518796994,
+      "grad_norm": 0.19502531989400285,
+      "learning_rate": 1.0496708275880495e-07,
+      "loss": 0.2738,
+      "step": 2288
+    },
+    {
+      "epoch": 17.210526315789473,
+      "grad_norm": 0.20665050062887622,
+      "learning_rate": 1.0441352141448156e-07,
+      "loss": 0.2885,
+      "step": 2289
+    },
+    {
+      "epoch": 17.218045112781954,
+      "grad_norm": 0.23411770090712603,
+      "learning_rate": 1.0386134316855666e-07,
+      "loss": 0.282,
+      "step": 2290
+    },
+    {
+      "epoch": 17.225563909774436,
+      "grad_norm": 0.1990179884188933,
+      "learning_rate": 1.0331054887378566e-07,
+      "loss": 0.2614,
+      "step": 2291
+    },
+    {
+      "epoch": 17.23308270676692,
+      "grad_norm": 0.20573014705382928,
+      "learning_rate": 1.0276113938078768e-07,
+      "loss": 0.2733,
+      "step": 2292
+    },
+    {
+      "epoch": 17.2406015037594,
+      "grad_norm": 0.22666378955574792,
+      "learning_rate": 1.0221311553804312e-07,
+      "loss": 0.2701,
+      "step": 2293
+    },
+    {
+      "epoch": 17.24812030075188,
+      "grad_norm": 0.2015355235550191,
+      "learning_rate": 1.01666478191892e-07,
+      "loss": 0.2776,
+      "step": 2294
+    },
+    {
+      "epoch": 17.25563909774436,
+      "grad_norm": 0.2048888267937596,
+      "learning_rate": 1.0112122818653345e-07,
+      "loss": 0.2801,
+      "step": 2295
+    },
+    {
+      "epoch": 17.263157894736842,
+      "grad_norm": 0.21505806513886844,
+      "learning_rate": 1.0057736636402381e-07,
+      "loss": 0.2818,
+      "step": 2296
+    },
+    {
+      "epoch": 17.270676691729324,
+      "grad_norm": 0.19496085149080833,
+      "learning_rate": 1.0003489356427596e-07,
+      "loss": 0.2694,
+      "step": 2297
+    },
+    {
+      "epoch": 17.278195488721803,
+      "grad_norm": 0.2134470788977197,
+      "learning_rate": 9.949381062505723e-08,
+      "loss": 0.2723,
+      "step": 2298
+    },
+    {
+      "epoch": 17.285714285714285,
+      "grad_norm": 0.1996530189619854,
+      "learning_rate": 9.895411838198886e-08,
+      "loss": 0.2709,
+      "step": 2299
+    },
+    {
+      "epoch": 17.293233082706767,
+      "grad_norm": 0.20714967850349475,
+      "learning_rate": 9.8415817668544e-08,
+      "loss": 0.2779,
+      "step": 2300
+    },
+    {
+      "epoch": 17.30075187969925,
+      "grad_norm": 0.2002011090754509,
+      "learning_rate": 9.787890931604737e-08,
+      "loss": 0.2778,
+      "step": 2301
+    },
+    {
+      "epoch": 17.30827067669173,
+      "grad_norm": 0.20216481204960823,
+      "learning_rate": 9.734339415367254e-08,
+      "loss": 0.2727,
+      "step": 2302
+    },
+    {
+      "epoch": 17.31578947368421,
+      "grad_norm": 0.19934869403268607,
+      "learning_rate": 9.680927300844243e-08,
+      "loss": 0.2799,
+      "step": 2303
+    },
+    {
+      "epoch": 17.32330827067669,
+      "grad_norm": 0.2268112695192392,
+      "learning_rate": 9.627654670522645e-08,
+      "loss": 0.2758,
+      "step": 2304
+    },
+    {
+      "epoch": 17.330827067669173,
+      "grad_norm": 0.21556767269437033,
+      "learning_rate": 9.574521606674035e-08,
+      "loss": 0.2769,
+      "step": 2305
+    },
+    {
+      "epoch": 17.338345864661655,
+      "grad_norm": 0.20100254797586933,
+      "learning_rate": 9.521528191354389e-08,
+      "loss": 0.2713,
+      "step": 2306
+    },
+    {
+      "epoch": 17.345864661654137,
+      "grad_norm": 0.20563096113768378,
+      "learning_rate": 9.468674506404095e-08,
+      "loss": 0.2815,
+      "step": 2307
+    },
+    {
+      "epoch": 17.353383458646615,
+      "grad_norm": 0.19408509939220972,
+      "learning_rate": 9.415960633447673e-08,
+      "loss": 0.2596,
+      "step": 2308
+    },
+    {
+      "epoch": 17.360902255639097,
+      "grad_norm": 0.19800652631386056,
+      "learning_rate": 9.36338665389379e-08,
+      "loss": 0.2675,
+      "step": 2309
+    },
+    {
+      "epoch": 17.36842105263158,
+      "grad_norm": 0.21387912891422403,
+      "learning_rate": 9.310952648935e-08,
+      "loss": 0.2753,
+      "step": 2310
+    },
+    {
+      "epoch": 17.37593984962406,
+      "grad_norm": 0.2020714366184962,
+      "learning_rate": 9.258658699547762e-08,
+      "loss": 0.267,
+      "step": 2311
+    },
+    {
+      "epoch": 17.383458646616543,
+      "grad_norm": 0.209270246259497,
+      "learning_rate": 9.206504886492161e-08,
+      "loss": 0.2747,
+      "step": 2312
+    },
+    {
+      "epoch": 17.39097744360902,
+      "grad_norm": 0.19177263852699147,
+      "learning_rate": 9.15449129031196e-08,
+      "loss": 0.2719,
+      "step": 2313
+    },
+    {
+      "epoch": 17.398496240601503,
+      "grad_norm": 0.21617227718516768,
+      "learning_rate": 9.102617991334272e-08,
+      "loss": 0.271,
+      "step": 2314
+    },
+    {
+      "epoch": 17.406015037593985,
+      "grad_norm": 0.19293633563768237,
+      "learning_rate": 9.050885069669622e-08,
+      "loss": 0.276,
+      "step": 2315
+    },
+    {
+      "epoch": 17.413533834586467,
+      "grad_norm": 0.20570260850438268,
+      "learning_rate": 8.999292605211694e-08,
+      "loss": 0.2921,
+      "step": 2316
+    },
+    {
+      "epoch": 17.42105263157895,
+      "grad_norm": 0.19930250992154275,
+      "learning_rate": 8.947840677637298e-08,
+      "loss": 0.2801,
+      "step": 2317
+    },
+    {
+      "epoch": 17.428571428571427,
+      "grad_norm": 0.18840527458373962,
+      "learning_rate": 8.896529366406181e-08,
+      "loss": 0.269,
+      "step": 2318
+    },
+    {
+      "epoch": 17.43609022556391,
+      "grad_norm": 0.19535061741360238,
+      "learning_rate": 8.845358750760901e-08,
+      "loss": 0.2761,
+      "step": 2319
+    },
+    {
+      "epoch": 17.44360902255639,
+      "grad_norm": 0.20308499516891618,
+      "learning_rate": 8.794328909726822e-08,
+      "loss": 0.2763,
+      "step": 2320
+    },
+    {
+      "epoch": 17.451127819548873,
+      "grad_norm": 0.21284732141935617,
+      "learning_rate": 8.743439922111784e-08,
+      "loss": 0.2779,
+      "step": 2321
+    },
+    {
+      "epoch": 17.458646616541355,
+      "grad_norm": 0.20711678382226886,
+      "learning_rate": 8.692691866506219e-08,
+      "loss": 0.2769,
+      "step": 2322
+    },
+    {
+      "epoch": 17.466165413533833,
+      "grad_norm": 0.21610774953618714,
+      "learning_rate": 8.642084821282802e-08,
+      "loss": 0.2768,
+      "step": 2323
+    },
+    {
+      "epoch": 17.473684210526315,
+      "grad_norm": 0.2087638119164717,
+      "learning_rate": 8.59161886459654e-08,
+      "loss": 0.2789,
+      "step": 2324
+    },
+    {
+      "epoch": 17.481203007518797,
+      "grad_norm": 0.1937501568678847,
+      "learning_rate": 8.541294074384465e-08,
+      "loss": 0.2837,
+      "step": 2325
+    },
+    {
+      "epoch": 17.48872180451128,
+      "grad_norm": 0.2065095162156074,
+      "learning_rate": 8.491110528365652e-08,
+      "loss": 0.2737,
+      "step": 2326
+    },
+    {
+      "epoch": 17.49624060150376,
+      "grad_norm": 0.20375888250462723,
+      "learning_rate": 8.44106830404101e-08,
+      "loss": 0.2778,
+      "step": 2327
+    },
+    {
+      "epoch": 17.50375939849624,
+      "grad_norm": 0.2033514762564809,
+      "learning_rate": 8.39116747869324e-08,
+      "loss": 0.2863,
+      "step": 2328
+    },
+    {
+      "epoch": 17.51127819548872,
+      "grad_norm": 0.1922665786049692,
+      "learning_rate": 8.341408129386629e-08,
+      "loss": 0.2717,
+      "step": 2329
+    },
+    {
+      "epoch": 17.518796992481203,
+      "grad_norm": 0.2034565574762575,
+      "learning_rate": 8.291790332967007e-08,
+      "loss": 0.2758,
+      "step": 2330
+    },
+    {
+      "epoch": 17.526315789473685,
+      "grad_norm": 0.19583382661243598,
+      "learning_rate": 8.242314166061581e-08,
+      "loss": 0.2688,
+      "step": 2331
+    },
+    {
+      "epoch": 17.533834586466167,
+      "grad_norm": 0.19538963716946608,
+      "learning_rate": 8.19297970507885e-08,
+      "loss": 0.2794,
+      "step": 2332
+    },
+    {
+      "epoch": 17.541353383458645,
+      "grad_norm": 0.24776339264048367,
+      "learning_rate": 8.143787026208426e-08,
+      "loss": 0.2775,
+      "step": 2333
+    },
+    {
+      "epoch": 17.548872180451127,
+      "grad_norm": 0.20850668920978474,
+      "learning_rate": 8.094736205421026e-08,
+      "loss": 0.2843,
+      "step": 2334
+    },
+    {
+      "epoch": 17.55639097744361,
+      "grad_norm": 0.20858305767486104,
+      "learning_rate": 8.045827318468224e-08,
+      "loss": 0.2777,
+      "step": 2335
+    },
+    {
+      "epoch": 17.56390977443609,
+      "grad_norm": 0.20956633210380773,
+      "learning_rate": 7.997060440882453e-08,
+      "loss": 0.2679,
+      "step": 2336
+    },
+    {
+      "epoch": 17.571428571428573,
+      "grad_norm": 0.1952524226686647,
+      "learning_rate": 7.94843564797678e-08,
+      "loss": 0.2749,
+      "step": 2337
+    },
+    {
+      "epoch": 17.57894736842105,
+      "grad_norm": 0.2012603479295719,
+      "learning_rate": 7.899953014844918e-08,
+      "loss": 0.2767,
+      "step": 2338
+    },
+    {
+      "epoch": 17.586466165413533,
+      "grad_norm": 0.20745289120797283,
+      "learning_rate": 7.851612616360937e-08,
+      "loss": 0.2878,
+      "step": 2339
+    },
+    {
+      "epoch": 17.593984962406015,
+      "grad_norm": 0.20502858784011407,
+      "learning_rate": 7.803414527179342e-08,
+      "loss": 0.2795,
+      "step": 2340
+    },
+    {
+      "epoch": 17.601503759398497,
+      "grad_norm": 0.20003104481768733,
+      "learning_rate": 7.755358821734782e-08,
+      "loss": 0.2692,
+      "step": 2341
+    },
+    {
+      "epoch": 17.60902255639098,
+      "grad_norm": 0.19181375591293337,
+      "learning_rate": 7.707445574242099e-08,
+      "loss": 0.2656,
+      "step": 2342
+    },
+    {
+      "epoch": 17.616541353383457,
+      "grad_norm": 0.20004114174157944,
+      "learning_rate": 7.659674858696041e-08,
+      "loss": 0.256,
+      "step": 2343
+    },
+    {
+      "epoch": 17.62406015037594,
+      "grad_norm": 0.2043065670276422,
+      "learning_rate": 7.612046748871326e-08,
+      "loss": 0.2721,
+      "step": 2344
+    },
+    {
+      "epoch": 17.63157894736842,
+      "grad_norm": 0.1957910822588089,
+      "learning_rate": 7.564561318322371e-08,
+      "loss": 0.2801,
+      "step": 2345
+    },
+    {
+      "epoch": 17.639097744360903,
+      "grad_norm": 0.20893928390245745,
+      "learning_rate": 7.51721864038326e-08,
+      "loss": 0.2739,
+      "step": 2346
+    },
+    {
+      "epoch": 17.646616541353385,
+      "grad_norm": 0.20262729871424662,
+      "learning_rate": 7.470018788167643e-08,
+      "loss": 0.2662,
+      "step": 2347
+    },
+    {
+      "epoch": 17.654135338345863,
+      "grad_norm": 0.1980097509735441,
+      "learning_rate": 7.422961834568563e-08,
+      "loss": 0.2719,
+      "step": 2348
+    },
+    {
+      "epoch": 17.661654135338345,
+      "grad_norm": 0.20181978758558208,
+      "learning_rate": 7.376047852258426e-08,
+      "loss": 0.281,
+      "step": 2349
+    },
+    {
+      "epoch": 17.669172932330827,
+      "grad_norm": 0.22105086451487474,
+      "learning_rate": 7.329276913688787e-08,
+      "loss": 0.2847,
+      "step": 2350
+    },
+    {
+      "epoch": 17.67669172932331,
+      "grad_norm": 0.20646278540053742,
+      "learning_rate": 7.282649091090332e-08,
+      "loss": 0.2731,
+      "step": 2351
+    },
+    {
+      "epoch": 17.68421052631579,
+      "grad_norm": 0.19924646407679364,
+      "learning_rate": 7.236164456472671e-08,
+      "loss": 0.2727,
+      "step": 2352
+    },
+    {
+      "epoch": 17.69172932330827,
+      "grad_norm": 0.18889420608592222,
+      "learning_rate": 7.189823081624368e-08,
+      "loss": 0.2683,
+      "step": 2353
+    },
+    {
+      "epoch": 17.69924812030075,
+      "grad_norm": 0.20109514980520748,
+      "learning_rate": 7.143625038112666e-08,
+      "loss": 0.275,
+      "step": 2354
+    },
+    {
+      "epoch": 17.706766917293233,
+      "grad_norm": 0.21493580512590818,
+      "learning_rate": 7.097570397283492e-08,
+      "loss": 0.2689,
+      "step": 2355
+    },
+    {
+      "epoch": 17.714285714285715,
+      "grad_norm": 0.2023387190025928,
+      "learning_rate": 7.051659230261297e-08,
+      "loss": 0.2781,
+      "step": 2356
+    },
+    {
+      "epoch": 17.721804511278194,
+      "grad_norm": 0.2133640720949985,
+      "learning_rate": 7.005891607948977e-08,
+      "loss": 0.2841,
+      "step": 2357
+    },
+    {
+      "epoch": 17.729323308270676,
+      "grad_norm": 0.20402743008560092,
+      "learning_rate": 6.960267601027691e-08,
+      "loss": 0.2794,
+      "step": 2358
+    },
+    {
+      "epoch": 17.736842105263158,
+      "grad_norm": 0.19342050169952102,
+      "learning_rate": 6.914787279956902e-08,
+      "loss": 0.2682,
+      "step": 2359
+    },
+    {
+      "epoch": 17.74436090225564,
+      "grad_norm": 0.20346068499967718,
+      "learning_rate": 6.869450714974057e-08,
+      "loss": 0.2691,
+      "step": 2360
+    },
+    {
+      "epoch": 17.75187969924812,
+      "grad_norm": 0.1903735037923294,
+      "learning_rate": 6.824257976094694e-08,
+      "loss": 0.2696,
+      "step": 2361
+    },
+    {
+      "epoch": 17.7593984962406,
+      "grad_norm": 0.2089483570949615,
+      "learning_rate": 6.779209133112163e-08,
+      "loss": 0.2708,
+      "step": 2362
+    },
+    {
+      "epoch": 17.76691729323308,
+      "grad_norm": 0.24579107996536126,
+      "learning_rate": 6.734304255597634e-08,
+      "loss": 0.2706,
+      "step": 2363
+    },
+    {
+      "epoch": 17.774436090225564,
+      "grad_norm": 0.20540781772630715,
+      "learning_rate": 6.689543412899911e-08,
+      "loss": 0.2735,
+      "step": 2364
+    },
+    {
+      "epoch": 17.781954887218046,
+      "grad_norm": 0.20525980921389692,
+      "learning_rate": 6.64492667414539e-08,
+      "loss": 0.2827,
+      "step": 2365
+    },
+    {
+      "epoch": 17.789473684210527,
+      "grad_norm": 0.18570848644719412,
+      "learning_rate": 6.600454108237874e-08,
+      "loss": 0.2663,
+      "step": 2366
+    },
+    {
+      "epoch": 17.796992481203006,
+      "grad_norm": 0.21446185465933224,
+      "learning_rate": 6.556125783858568e-08,
+      "loss": 0.268,
+      "step": 2367
+    },
+    {
+      "epoch": 17.804511278195488,
+      "grad_norm": 0.21171345518929716,
+      "learning_rate": 6.511941769465878e-08,
+      "loss": 0.2855,
+      "step": 2368
+    },
+    {
+      "epoch": 17.81203007518797,
+      "grad_norm": 0.19457680407728223,
+      "learning_rate": 6.467902133295366e-08,
+      "loss": 0.2753,
+      "step": 2369
+    },
+    {
+      "epoch": 17.81954887218045,
+      "grad_norm": 0.21448138900260136,
+      "learning_rate": 6.424006943359606e-08,
+      "loss": 0.2613,
+      "step": 2370
+    },
+    {
+      "epoch": 17.827067669172934,
+      "grad_norm": 0.19683110546630428,
+      "learning_rate": 6.380256267448114e-08,
+      "loss": 0.2678,
+      "step": 2371
+    },
+    {
+      "epoch": 17.834586466165412,
+      "grad_norm": 0.21541454833193813,
+      "learning_rate": 6.336650173127223e-08,
+      "loss": 0.2747,
+      "step": 2372
+    },
+    {
+      "epoch": 17.842105263157894,
+      "grad_norm": 0.20581574131186767,
+      "learning_rate": 6.293188727739962e-08,
+      "loss": 0.2719,
+      "step": 2373
+    },
+    {
+      "epoch": 17.849624060150376,
+      "grad_norm": 0.21101256310449842,
+      "learning_rate": 6.249871998405998e-08,
+      "loss": 0.275,
+      "step": 2374
+    },
+    {
+      "epoch": 17.857142857142858,
+      "grad_norm": 0.20255314144213482,
+      "learning_rate": 6.206700052021474e-08,
+      "loss": 0.2739,
+      "step": 2375
+    },
+    {
+      "epoch": 17.86466165413534,
+      "grad_norm": 0.2022244080931734,
+      "learning_rate": 6.163672955258981e-08,
+      "loss": 0.2752,
+      "step": 2376
+    },
+    {
+      "epoch": 17.872180451127818,
+      "grad_norm": 0.21107846268503025,
+      "learning_rate": 6.120790774567375e-08,
+      "loss": 0.2659,
+      "step": 2377
+    },
+    {
+      "epoch": 17.8796992481203,
+      "grad_norm": 0.19794763349057498,
+      "learning_rate": 6.078053576171738e-08,
+      "loss": 0.282,
+      "step": 2378
+    },
+    {
+      "epoch": 17.887218045112782,
+      "grad_norm": 0.20702307207190832,
+      "learning_rate": 6.035461426073219e-08,
+      "loss": 0.2834,
+      "step": 2379
+    },
+    {
+      "epoch": 17.894736842105264,
+      "grad_norm": 0.2046541768582397,
+      "learning_rate": 5.99301439004899e-08,
+      "loss": 0.2776,
+      "step": 2380
+    },
+    {
+      "epoch": 17.902255639097746,
+      "grad_norm": 0.19190177245300588,
+      "learning_rate": 5.9507125336520805e-08,
+      "loss": 0.272,
+      "step": 2381
+    },
+    {
+      "epoch": 17.909774436090224,
+      "grad_norm": 0.21151486233870267,
+      "learning_rate": 5.908555922211367e-08,
+      "loss": 0.2707,
+      "step": 2382
+    },
+    {
+      "epoch": 17.917293233082706,
+      "grad_norm": 0.20017131555669404,
+      "learning_rate": 5.8665446208313486e-08,
+      "loss": 0.276,
+      "step": 2383
+    },
+    {
+      "epoch": 17.924812030075188,
+      "grad_norm": 0.1955598750874458,
+      "learning_rate": 5.824678694392193e-08,
+      "loss": 0.2693,
+      "step": 2384
+    },
+    {
+      "epoch": 17.93233082706767,
+      "grad_norm": 0.1942521441371996,
+      "learning_rate": 5.782958207549482e-08,
+      "loss": 0.2762,
+      "step": 2385
+    },
+    {
+      "epoch": 17.93984962406015,
+      "grad_norm": 0.18656583625982323,
+      "learning_rate": 5.741383224734253e-08,
+      "loss": 0.2622,
+      "step": 2386
+    },
+    {
+      "epoch": 17.94736842105263,
+      "grad_norm": 0.19604912675345565,
+      "learning_rate": 5.699953810152769e-08,
+      "loss": 0.2658,
+      "step": 2387
+    },
+    {
+      "epoch": 17.954887218045112,
+      "grad_norm": 0.20157119469808593,
+      "learning_rate": 5.6586700277865604e-08,
+      "loss": 0.2824,
+      "step": 2388
+    },
+    {
+      "epoch": 17.962406015037594,
+      "grad_norm": 0.19657461974235416,
+      "learning_rate": 5.617531941392162e-08,
+      "loss": 0.2724,
+      "step": 2389
+    },
+    {
+      "epoch": 17.969924812030076,
+      "grad_norm": 0.20713120023138162,
+      "learning_rate": 5.5765396145011965e-08,
+      "loss": 0.2777,
+      "step": 2390
+    },
+    {
+      "epoch": 17.977443609022558,
+      "grad_norm": 0.19500435323214593,
+      "learning_rate": 5.535693110420092e-08,
+      "loss": 0.2738,
+      "step": 2391
+    },
+    {
+      "epoch": 17.984962406015036,
+      "grad_norm": 0.20335779102355422,
+      "learning_rate": 5.494992492230166e-08,
+      "loss": 0.2714,
+      "step": 2392
+    },
+    {
+      "epoch": 17.992481203007518,
+      "grad_norm": 0.1916252197083233,
+      "learning_rate": 5.454437822787361e-08,
+      "loss": 0.2693,
+      "step": 2393
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 0.1936163714086462,
+      "learning_rate": 5.414029164722278e-08,
+      "loss": 0.2645,
+      "step": 2394
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 0.42579537630081177,
+      "eval_runtime": 1.5834,
+      "eval_samples_per_second": 206.522,
+      "eval_steps_per_second": 3.789,
+      "step": 2394
+    },
+    {
+      "epoch": 18.007518796992482,
+      "grad_norm": 0.20285066757889292,
+      "learning_rate": 5.373766580439976e-08,
+      "loss": 0.2684,
+      "step": 2395
+    },
+    {
+      "epoch": 18.015037593984964,
+      "grad_norm": 0.2065876476975389,
+      "learning_rate": 5.333650132119971e-08,
+      "loss": 0.2782,
+      "step": 2396
+    },
+    {
+      "epoch": 18.022556390977442,
+      "grad_norm": 0.20059714125740474,
+      "learning_rate": 5.293679881716051e-08,
+      "loss": 0.2667,
+      "step": 2397
+    },
+    {
+      "epoch": 18.030075187969924,
+      "grad_norm": 0.21079729668828331,
+      "learning_rate": 5.2538558909562716e-08,
+      "loss": 0.2825,
+      "step": 2398
+    },
+    {
+      "epoch": 18.037593984962406,
+      "grad_norm": 0.19993009203449952,
+      "learning_rate": 5.21417822134278e-08,
+      "loss": 0.276,
+      "step": 2399
+    },
+    {
+      "epoch": 18.045112781954888,
+      "grad_norm": 0.2138213111938324,
+      "learning_rate": 5.1746469341517497e-08,
+      "loss": 0.2749,
+      "step": 2400
+    },
+    {
+      "epoch": 18.05263157894737,
+      "grad_norm": 0.19958292584104848,
+      "learning_rate": 5.135262090433323e-08,
+      "loss": 0.265,
+      "step": 2401
+    },
+    {
+      "epoch": 18.06015037593985,
+      "grad_norm": 0.20857864695183267,
+      "learning_rate": 5.096023751011413e-08,
+      "loss": 0.2768,
+      "step": 2402
+    },
+    {
+      "epoch": 18.06766917293233,
+      "grad_norm": 0.20835087210775477,
+      "learning_rate": 5.05693197648378e-08,
+      "loss": 0.2701,
+      "step": 2403
+    },
+    {
+      "epoch": 18.075187969924812,
+      "grad_norm": 0.20252273040114002,
+      "learning_rate": 5.017986827221732e-08,
+      "loss": 0.2688,
+      "step": 2404
+    },
+    {
+      "epoch": 18.082706766917294,
+      "grad_norm": 0.1966511251626283,
+      "learning_rate": 4.979188363370213e-08,
+      "loss": 0.2754,
+      "step": 2405
+    },
+    {
+      "epoch": 18.090225563909776,
+      "grad_norm": 0.20613051021016723,
+      "learning_rate": 4.940536644847593e-08,
+      "loss": 0.2789,
+      "step": 2406
+    },
+    {
+      "epoch": 18.097744360902254,
+      "grad_norm": 0.2043881674387074,
+      "learning_rate": 4.9020317313456463e-08,
+      "loss": 0.2844,
+      "step": 2407
+    },
+    {
+      "epoch": 18.105263157894736,
+      "grad_norm": 0.19715236872820424,
+      "learning_rate": 4.863673682329372e-08,
+      "loss": 0.2661,
+      "step": 2408
+    },
+    {
+      "epoch": 18.112781954887218,
+      "grad_norm": 0.20649045274083636,
+      "learning_rate": 4.825462557037052e-08,
+      "loss": 0.2761,
+      "step": 2409
+    },
+    {
+      "epoch": 18.1203007518797,
+      "grad_norm": 0.20773688586389935,
+      "learning_rate": 4.78739841447996e-08,
+      "loss": 0.2602,
+      "step": 2410
+    },
+    {
+      "epoch": 18.127819548872182,
+      "grad_norm": 0.20450291864040215,
+      "learning_rate": 4.749481313442483e-08,
+      "loss": 0.2647,
+      "step": 2411
+    },
+    {
+      "epoch": 18.13533834586466,
+      "grad_norm": 0.19032801308534633,
+      "learning_rate": 4.7117113124818144e-08,
+      "loss": 0.2664,
+      "step": 2412
+    },
+    {
+      "epoch": 18.142857142857142,
+      "grad_norm": 0.19949385021975707,
+      "learning_rate": 4.674088469928084e-08,
+      "loss": 0.2737,
+      "step": 2413
+    },
+    {
+      "epoch": 18.150375939849624,
+      "grad_norm": 0.20389879282827722,
+      "learning_rate": 4.636612843884058e-08,
+      "loss": 0.2672,
+      "step": 2414
+    },
+    {
+      "epoch": 18.157894736842106,
+      "grad_norm": 0.2069124249172187,
+      "learning_rate": 4.59928449222523e-08,
+      "loss": 0.2848,
+      "step": 2415
+    },
+    {
+      "epoch": 18.165413533834588,
+      "grad_norm": 0.2022886864213043,
+      "learning_rate": 4.562103472599599e-08,
+      "loss": 0.2814,
+      "step": 2416
+    },
+    {
+      "epoch": 18.172932330827066,
+      "grad_norm": 0.2030246631607149,
+      "learning_rate": 4.5250698424276536e-08,
+      "loss": 0.274,
+      "step": 2417
+    },
+    {
+      "epoch": 18.18045112781955,
+      "grad_norm": 0.2051725642675359,
+      "learning_rate": 4.488183658902256e-08,
+      "loss": 0.2775,
+      "step": 2418
+    },
+    {
+      "epoch": 18.18796992481203,
+      "grad_norm": 0.20030696297908465,
+      "learning_rate": 4.451444978988561e-08,
+      "loss": 0.2738,
+      "step": 2419
+    },
+    {
+      "epoch": 18.195488721804512,
+      "grad_norm": 0.19930397206569325,
+      "learning_rate": 4.414853859423917e-08,
+      "loss": 0.2769,
+      "step": 2420
+    },
+    {
+      "epoch": 18.203007518796994,
+      "grad_norm": 0.19571746792564654,
+      "learning_rate": 4.37841035671781e-08,
+      "loss": 0.2796,
+      "step": 2421
+    },
+    {
+      "epoch": 18.210526315789473,
+      "grad_norm": 0.19888790500396752,
+      "learning_rate": 4.342114527151719e-08,
+      "loss": 0.2735,
+      "step": 2422
+    },
+    {
+      "epoch": 18.218045112781954,
+      "grad_norm": 0.19009050993529703,
+      "learning_rate": 4.3059664267791175e-08,
+      "loss": 0.2637,
+      "step": 2423
+    },
+    {
+      "epoch": 18.225563909774436,
+      "grad_norm": 0.19949105126356467,
+      "learning_rate": 4.2699661114252714e-08,
+      "loss": 0.2709,
+      "step": 2424
+    },
+    {
+      "epoch": 18.23308270676692,
+      "grad_norm": 0.19924714707725164,
+      "learning_rate": 4.234113636687242e-08,
+      "loss": 0.2802,
+      "step": 2425
+    },
+    {
+      "epoch": 18.2406015037594,
+      "grad_norm": 0.2002041163392511,
+      "learning_rate": 4.198409057933805e-08,
+      "loss": 0.2719,
+      "step": 2426
+    },
+    {
+      "epoch": 18.24812030075188,
+      "grad_norm": 0.19517729846539195,
+      "learning_rate": 4.162852430305275e-08,
+      "loss": 0.2747,
+      "step": 2427
+    },
+    {
+      "epoch": 18.25563909774436,
+      "grad_norm": 0.2025137759096834,
+      "learning_rate": 4.127443808713527e-08,
+      "loss": 0.2862,
+      "step": 2428
+    },
+    {
+      "epoch": 18.263157894736842,
+      "grad_norm": 0.20230868661443174,
+      "learning_rate": 4.09218324784183e-08,
+      "loss": 0.2687,
+      "step": 2429
+    },
+    {
+      "epoch": 18.270676691729324,
+      "grad_norm": 0.20795355806008597,
+      "learning_rate": 4.057070802144813e-08,
+      "loss": 0.268,
+      "step": 2430
+    },
+    {
+      "epoch": 18.278195488721803,
+      "grad_norm": 0.19930978228133328,
+      "learning_rate": 4.022106525848346e-08,
+      "loss": 0.262,
+      "step": 2431
+    },
+    {
+      "epoch": 18.285714285714285,
+      "grad_norm": 0.215016201594581,
+      "learning_rate": 3.9872904729495113e-08,
+      "loss": 0.275,
+      "step": 2432
+    },
+    {
+      "epoch": 18.293233082706767,
+      "grad_norm": 0.19831584685371992,
+      "learning_rate": 3.9526226972164455e-08,
+      "loss": 0.2893,
+      "step": 2433
+    },
+    {
+      "epoch": 18.30075187969925,
+      "grad_norm": 0.20562300838449207,
+      "learning_rate": 3.918103252188298e-08,
+      "loss": 0.2762,
+      "step": 2434
+    },
+    {
+      "epoch": 18.30827067669173,
+      "grad_norm": 0.2034008210280336,
+      "learning_rate": 3.88373219117516e-08,
+      "loss": 0.2733,
+      "step": 2435
+    },
+    {
+      "epoch": 18.31578947368421,
+      "grad_norm": 0.1945256675156771,
+      "learning_rate": 3.849509567257958e-08,
+      "loss": 0.2754,
+      "step": 2436
+    },
+    {
+      "epoch": 18.32330827067669,
+      "grad_norm": 0.19760905699681866,
+      "learning_rate": 3.815435433288372e-08,
+      "loss": 0.2817,
+      "step": 2437
+    },
+    {
+      "epoch": 18.330827067669173,
+      "grad_norm": 0.19885926434694431,
+      "learning_rate": 3.7815098418887746e-08,
+      "loss": 0.2754,
+      "step": 2438
+    },
+    {
+      "epoch": 18.338345864661655,
+      "grad_norm": 0.20558159352328514,
+      "learning_rate": 3.747732845452134e-08,
+      "loss": 0.2701,
+      "step": 2439
+    },
+    {
+      "epoch": 18.345864661654137,
+      "grad_norm": 0.19523648243758418,
+      "learning_rate": 3.714104496141923e-08,
+      "loss": 0.2772,
+      "step": 2440
+    },
+    {
+      "epoch": 18.353383458646615,
+      "grad_norm": 0.2110003378498023,
+      "learning_rate": 3.680624845892066e-08,
+      "loss": 0.294,
+      "step": 2441
+    },
+    {
+      "epoch": 18.360902255639097,
+      "grad_norm": 0.20898335630020154,
+      "learning_rate": 3.647293946406849e-08,
+      "loss": 0.2789,
+      "step": 2442
+    },
+    {
+      "epoch": 18.36842105263158,
+      "grad_norm": 0.20112776161821147,
+      "learning_rate": 3.614111849160795e-08,
+      "loss": 0.2595,
+      "step": 2443
+    },
+    {
+      "epoch": 18.37593984962406,
+      "grad_norm": 0.1960516948633975,
+      "learning_rate": 3.581078605398702e-08,
+      "loss": 0.2745,
+      "step": 2444
+    },
+    {
+      "epoch": 18.383458646616543,
+      "grad_norm": 0.20300016048317748,
+      "learning_rate": 3.548194266135385e-08,
+      "loss": 0.2726,
+      "step": 2445
+    },
+    {
+      "epoch": 18.39097744360902,
+      "grad_norm": 0.2012258226544976,
+      "learning_rate": 3.5154588821557975e-08,
+      "loss": 0.2702,
+      "step": 2446
+    },
+    {
+      "epoch": 18.398496240601503,
+      "grad_norm": 0.19496420157206135,
+      "learning_rate": 3.4828725040147776e-08,
+      "loss": 0.2773,
+      "step": 2447
+    },
+    {
+      "epoch": 18.406015037593985,
+      "grad_norm": 0.19691792720597748,
+      "learning_rate": 3.4504351820371035e-08,
+      "loss": 0.2782,
+      "step": 2448
+    },
+    {
+      "epoch": 18.413533834586467,
+      "grad_norm": 0.2052286181404773,
+      "learning_rate": 3.418146966317303e-08,
+      "loss": 0.2825,
+      "step": 2449
+    },
+    {
+      "epoch": 18.42105263157895,
+      "grad_norm": 0.20856693782122787,
+      "learning_rate": 3.38600790671969e-08,
+      "loss": 0.2842,
+      "step": 2450
+    },
+    {
+      "epoch": 18.428571428571427,
+      "grad_norm": 0.2099262730778256,
+      "learning_rate": 3.354018052878182e-08,
+      "loss": 0.2659,
+      "step": 2451
+    },
+    {
+      "epoch": 18.43609022556391,
+      "grad_norm": 0.20212792044738506,
+      "learning_rate": 3.3221774541962847e-08,
+      "loss": 0.2715,
+      "step": 2452
+    },
+    {
+      "epoch": 18.44360902255639,
+      "grad_norm": 0.1992939855946206,
+      "learning_rate": 3.2904861598470276e-08,
+      "loss": 0.2689,
+      "step": 2453
+    },
+    {
+      "epoch": 18.451127819548873,
+      "grad_norm": 0.20721998357231855,
+      "learning_rate": 3.258944218772819e-08,
+      "loss": 0.2768,
+      "step": 2454
+    },
+    {
+      "epoch": 18.458646616541355,
+      "grad_norm": 0.19926713906738022,
+      "learning_rate": 3.2275516796854585e-08,
+      "loss": 0.2733,
+      "step": 2455
+    },
+    {
+      "epoch": 18.466165413533833,
+      "grad_norm": 0.19759256040731113,
+      "learning_rate": 3.196308591065966e-08,
+      "loss": 0.2816,
+      "step": 2456
+    },
+    {
+      "epoch": 18.473684210526315,
+      "grad_norm": 0.218420094631341,
+      "learning_rate": 3.165215001164601e-08,
+      "loss": 0.2909,
+      "step": 2457
+    },
+    {
+      "epoch": 18.481203007518797,
+      "grad_norm": 0.19633939897058217,
+      "learning_rate": 3.1342709580007175e-08,
+      "loss": 0.2749,
+      "step": 2458
+    },
+    {
+      "epoch": 18.48872180451128,
+      "grad_norm": 0.2027072156301188,
+      "learning_rate": 3.103476509362757e-08,
+      "loss": 0.2807,
+      "step": 2459
+    },
+    {
+      "epoch": 18.49624060150376,
+      "grad_norm": 0.19347679302053447,
+      "learning_rate": 3.072831702808065e-08,
+      "loss": 0.269,
+      "step": 2460
+    },
+    {
+      "epoch": 18.50375939849624,
+      "grad_norm": 0.20153359706027985,
+      "learning_rate": 3.0423365856629746e-08,
+      "loss": 0.2838,
+      "step": 2461
+    },
+    {
+      "epoch": 18.51127819548872,
+      "grad_norm": 0.20582895270180354,
+      "learning_rate": 3.011991205022557e-08,
+      "loss": 0.2749,
+      "step": 2462
+    },
+    {
+      "epoch": 18.518796992481203,
+      "grad_norm": 0.20916398378938375,
+      "learning_rate": 2.981795607750704e-08,
+      "loss": 0.2655,
+      "step": 2463
+    },
+    {
+      "epoch": 18.526315789473685,
+      "grad_norm": 0.19831026962331225,
+      "learning_rate": 2.9517498404799668e-08,
+      "loss": 0.2709,
+      "step": 2464
+    },
+    {
+      "epoch": 18.533834586466167,
+      "grad_norm": 0.1964526518672354,
+      "learning_rate": 2.921853949611508e-08,
+      "loss": 0.2669,
+      "step": 2465
+    },
+    {
+      "epoch": 18.541353383458645,
+      "grad_norm": 0.21374992580362967,
+      "learning_rate": 2.892107981315006e-08,
+      "loss": 0.2783,
+      "step": 2466
+    },
+    {
+      "epoch": 18.548872180451127,
+      "grad_norm": 0.19099880919321632,
+      "learning_rate": 2.862511981528659e-08,
+      "loss": 0.2742,
+      "step": 2467
+    },
+    {
+      "epoch": 18.55639097744361,
+      "grad_norm": 0.2107512633482909,
+      "learning_rate": 2.8330659959589942e-08,
+      "loss": 0.2698,
+      "step": 2468
+    },
+    {
+      "epoch": 18.56390977443609,
+      "grad_norm": 0.1995522380310079,
+      "learning_rate": 2.8037700700809464e-08,
+      "loss": 0.2776,
+      "step": 2469
+    },
+    {
+      "epoch": 18.571428571428573,
+      "grad_norm": 0.2007380525097938,
+      "learning_rate": 2.7746242491376138e-08,
+      "loss": 0.2733,
+      "step": 2470
+    },
+    {
+      "epoch": 18.57894736842105,
+      "grad_norm": 0.20488192270154418,
+      "learning_rate": 2.7456285781403577e-08,
+      "loss": 0.2707,
+      "step": 2471
+    },
+    {
+      "epoch": 18.586466165413533,
+      "grad_norm": 0.20323463355451393,
+      "learning_rate": 2.7167831018686137e-08,
+      "loss": 0.2766,
+      "step": 2472
+    },
+    {
+      "epoch": 18.593984962406015,
+      "grad_norm": 0.206963834636859,
+      "learning_rate": 2.6880878648698702e-08,
+      "loss": 0.2771,
+      "step": 2473
+    },
+    {
+      "epoch": 18.601503759398497,
+      "grad_norm": 0.20550200114553913,
+      "learning_rate": 2.659542911459589e-08,
+      "loss": 0.2727,
+      "step": 2474
+    },
+    {
+      "epoch": 18.60902255639098,
+      "grad_norm": 0.21400221505090572,
+      "learning_rate": 2.6311482857211853e-08,
+      "loss": 0.2733,
+      "step": 2475
+    },
+    {
+      "epoch": 18.616541353383457,
+      "grad_norm": 0.19666181265014934,
+      "learning_rate": 2.602904031505848e-08,
+      "loss": 0.2833,
+      "step": 2476
+    },
+    {
+      "epoch": 18.62406015037594,
+      "grad_norm": 0.20180126668786041,
+      "learning_rate": 2.574810192432575e-08,
+      "loss": 0.2628,
+      "step": 2477
+    },
+    {
+      "epoch": 18.63157894736842,
+      "grad_norm": 0.20005919226497337,
+      "learning_rate": 2.5468668118880933e-08,
+      "loss": 0.2743,
+      "step": 2478
+    },
+    {
+      "epoch": 18.639097744360903,
+      "grad_norm": 0.20498914605262167,
+      "learning_rate": 2.5190739330267053e-08,
+      "loss": 0.2753,
+      "step": 2479
+    },
+    {
+      "epoch": 18.646616541353385,
+      "grad_norm": 0.19838900845813362,
+      "learning_rate": 2.491431598770366e-08,
+      "loss": 0.2759,
+      "step": 2480
+    },
+    {
+      "epoch": 18.654135338345863,
+      "grad_norm": 0.1970246951774164,
+      "learning_rate": 2.463939851808472e-08,
+      "loss": 0.273,
+      "step": 2481
+    },
+    {
+      "epoch": 18.661654135338345,
+      "grad_norm": 0.2046681168631904,
+      "learning_rate": 2.4365987345978946e-08,
+      "loss": 0.2717,
+      "step": 2482
+    },
+    {
+      "epoch": 18.669172932330827,
+      "grad_norm": 0.21776596806312165,
+      "learning_rate": 2.4094082893628574e-08,
+      "loss": 0.2689,
+      "step": 2483
+    },
+    {
+      "epoch": 18.67669172932331,
+      "grad_norm": 0.2175682553278816,
+      "learning_rate": 2.382368558094927e-08,
+      "loss": 0.279,
+      "step": 2484
+    },
+    {
+      "epoch": 18.68421052631579,
+      "grad_norm": 0.21515253436172768,
+      "learning_rate": 2.355479582552877e-08,
+      "loss": 0.2767,
+      "step": 2485
+    },
+    {
+      "epoch": 18.69172932330827,
+      "grad_norm": 0.19913490046676602,
+      "learning_rate": 2.3287414042626908e-08,
+      "loss": 0.2685,
+      "step": 2486
+    },
+    {
+      "epoch": 18.69924812030075,
+      "grad_norm": 0.2069989101050845,
+      "learning_rate": 2.3021540645174476e-08,
+      "loss": 0.2733,
+      "step": 2487
+    },
+    {
+      "epoch": 18.706766917293233,
+      "grad_norm": 0.19804941321986833,
+      "learning_rate": 2.275717604377292e-08,
+      "loss": 0.2783,
+      "step": 2488
+    },
+    {
+      "epoch": 18.714285714285715,
+      "grad_norm": 0.1988898724014173,
+      "learning_rate": 2.2494320646693544e-08,
+      "loss": 0.2756,
+      "step": 2489
+    },
+    {
+      "epoch": 18.721804511278194,
+      "grad_norm": 0.20171327270975944,
+      "learning_rate": 2.2232974859877073e-08,
+      "loss": 0.2767,
+      "step": 2490
+    },
+    {
+      "epoch": 18.729323308270676,
+      "grad_norm": 0.20079288882634613,
+      "learning_rate": 2.1973139086932436e-08,
+      "loss": 0.2786,
+      "step": 2491
+    },
+    {
+      "epoch": 18.736842105263158,
+      "grad_norm": 0.20134115299069485,
+      "learning_rate": 2.1714813729136972e-08,
+      "loss": 0.2781,
+      "step": 2492
+    },
+    {
+      "epoch": 18.74436090225564,
+      "grad_norm": 0.20398318288958295,
+      "learning_rate": 2.1457999185435228e-08,
+      "loss": 0.2718,
+      "step": 2493
+    },
+    {
+      "epoch": 18.75187969924812,
+      "grad_norm": 0.20114943228037976,
+      "learning_rate": 2.1202695852438725e-08,
+      "loss": 0.2688,
+      "step": 2494
+    },
+    {
+      "epoch": 18.7593984962406,
+      "grad_norm": 0.2092197678640038,
+      "learning_rate": 2.0948904124424736e-08,
+      "loss": 0.2876,
+      "step": 2495
+    },
+    {
+      "epoch": 18.76691729323308,
+      "grad_norm": 0.2122436861668091,
+      "learning_rate": 2.0696624393336636e-08,
+      "loss": 0.2775,
+      "step": 2496
+    },
+    {
+      "epoch": 18.774436090225564,
+      "grad_norm": 0.19708606542519388,
+      "learning_rate": 2.044585704878221e-08,
+      "loss": 0.2845,
+      "step": 2497
+    },
+    {
+      "epoch": 18.781954887218046,
+      "grad_norm": 0.19261011027612088,
+      "learning_rate": 2.019660247803401e-08,
+      "loss": 0.2678,
+      "step": 2498
+    },
+    {
+      "epoch": 18.789473684210527,
+      "grad_norm": 0.2048784287989896,
+      "learning_rate": 1.9948861066028112e-08,
+      "loss": 0.2657,
+      "step": 2499
+    },
+    {
+      "epoch": 18.796992481203006,
+      "grad_norm": 0.20235711179073998,
+      "learning_rate": 1.9702633195363917e-08,
+      "loss": 0.2701,
+      "step": 2500
+    },
+    {
+      "epoch": 18.804511278195488,
+      "grad_norm": 0.19667758942326377,
+      "learning_rate": 1.9457919246303134e-08,
+      "loss": 0.2729,
+      "step": 2501
+    },
+    {
+      "epoch": 18.81203007518797,
+      "grad_norm": 0.19364771911891476,
+      "learning_rate": 1.921471959676957e-08,
+      "loss": 0.2714,
+      "step": 2502
+    },
+    {
+      "epoch": 18.81954887218045,
+      "grad_norm": 0.20229555677367927,
+      "learning_rate": 1.897303462234856e-08,
+      "loss": 0.2781,
+      "step": 2503
+    },
+    {
+      "epoch": 18.827067669172934,
+      "grad_norm": 0.1941141329834162,
+      "learning_rate": 1.87328646962861e-08,
+      "loss": 0.2678,
+      "step": 2504
+    },
+    {
+      "epoch": 18.834586466165412,
+      "grad_norm": 0.1957877357884975,
+      "learning_rate": 1.849421018948849e-08,
+      "loss": 0.2624,
+      "step": 2505
+    },
+    {
+      "epoch": 18.842105263157894,
+      "grad_norm": 0.192738594416725,
+      "learning_rate": 1.8257071470521467e-08,
+      "loss": 0.2816,
+      "step": 2506
+    },
+    {
+      "epoch": 18.849624060150376,
+      "grad_norm": 0.1964595406227596,
+      "learning_rate": 1.8021448905610414e-08,
+      "loss": 0.2816,
+      "step": 2507
+    },
+    {
+      "epoch": 18.857142857142858,
+      "grad_norm": 0.19976152830350452,
+      "learning_rate": 1.7787342858638588e-08,
+      "loss": 0.2716,
+      "step": 2508
+    },
+    {
+      "epoch": 18.86466165413534,
+      "grad_norm": 0.2050728280788527,
+      "learning_rate": 1.7554753691147672e-08,
+      "loss": 0.2622,
+      "step": 2509
+    },
+    {
+      "epoch": 18.872180451127818,
+      "grad_norm": 0.1921251160468644,
+      "learning_rate": 1.732368176233645e-08,
+      "loss": 0.271,
+      "step": 2510
+    },
+    {
+      "epoch": 18.8796992481203,
+      "grad_norm": 0.19557006659430776,
+      "learning_rate": 1.709412742906091e-08,
+      "loss": 0.2641,
+      "step": 2511
+    },
+    {
+      "epoch": 18.887218045112782,
+      "grad_norm": 0.19807334570273957,
+      "learning_rate": 1.686609104583292e-08,
+      "loss": 0.2808,
+      "step": 2512
+    },
+    {
+      "epoch": 18.894736842105264,
+      "grad_norm": 0.1972100591294326,
+      "learning_rate": 1.6639572964820437e-08,
+      "loss": 0.2701,
+      "step": 2513
+    },
+    {
+      "epoch": 18.902255639097746,
+      "grad_norm": 0.2032989809357687,
+      "learning_rate": 1.641457353584652e-08,
+      "loss": 0.2806,
+      "step": 2514
+    },
+    {
+      "epoch": 18.909774436090224,
+      "grad_norm": 0.2001153828291348,
+      "learning_rate": 1.6191093106388886e-08,
+      "loss": 0.2663,
+      "step": 2515
+    },
+    {
+      "epoch": 18.917293233082706,
+      "grad_norm": 0.21561810463179804,
+      "learning_rate": 1.5969132021579347e-08,
+      "loss": 0.2725,
+      "step": 2516
+    },
+    {
+      "epoch": 18.924812030075188,
+      "grad_norm": 0.19742432953983072,
+      "learning_rate": 1.5748690624203366e-08,
+      "loss": 0.2721,
+      "step": 2517
+    },
+    {
+      "epoch": 18.93233082706767,
+      "grad_norm": 0.20160236369602685,
+      "learning_rate": 1.552976925469951e-08,
+      "loss": 0.2783,
+      "step": 2518
+    },
+    {
+      "epoch": 18.93984962406015,
+      "grad_norm": 0.19077119146252416,
+      "learning_rate": 1.531236825115889e-08,
+      "loss": 0.2689,
+      "step": 2519
+    },
+    {
+      "epoch": 18.94736842105263,
+      "grad_norm": 0.19481424502471467,
+      "learning_rate": 1.50964879493245e-08,
+      "loss": 0.2841,
+      "step": 2520
+    },
+    {
+      "epoch": 18.954887218045112,
+      "grad_norm": 0.19686888894919752,
+      "learning_rate": 1.4882128682590978e-08,
+      "loss": 0.2735,
+      "step": 2521
+    },
+    {
+      "epoch": 18.962406015037594,
+      "grad_norm": 0.21884061974285432,
+      "learning_rate": 1.4669290782003962e-08,
+      "loss": 0.2826,
+      "step": 2522
+    },
+    {
+      "epoch": 18.969924812030076,
+      "grad_norm": 0.2068053948476769,
+      "learning_rate": 1.4457974576259524e-08,
+      "loss": 0.2798,
+      "step": 2523
+    },
+    {
+      "epoch": 18.977443609022558,
+      "grad_norm": 0.20358357929997847,
+      "learning_rate": 1.4248180391703613e-08,
+      "loss": 0.2733,
+      "step": 2524
+    },
+    {
+      "epoch": 18.984962406015036,
+      "grad_norm": 0.20312523273131594,
+      "learning_rate": 1.4039908552331836e-08,
+      "loss": 0.2867,
+      "step": 2525
+    },
+    {
+      "epoch": 18.992481203007518,
+      "grad_norm": 0.20532624329663682,
+      "learning_rate": 1.3833159379788684e-08,
+      "loss": 0.262,
+      "step": 2526
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 0.198860173440584,
+      "learning_rate": 1.362793319336708e-08,
+      "loss": 0.2702,
+      "step": 2527
+    },
+    {
+      "epoch": 19.0,
+      "eval_loss": 0.4260338246822357,
+      "eval_runtime": 1.7195,
+      "eval_samples_per_second": 190.172,
+      "eval_steps_per_second": 3.489,
+      "step": 2527
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2660,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 133,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2649581967900672.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}