{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9730290456431536,
  "eval_steps": 500,
  "global_step": 240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008298755186721992,
      "grad_norm": 0.80859375,
      "learning_rate": 8.695652173913044e-06,
      "loss": 2.3455,
      "step": 1
    },
    {
      "epoch": 0.016597510373443983,
      "grad_norm": 0.78125,
      "learning_rate": 1.739130434782609e-05,
      "loss": 2.2601,
      "step": 2
    },
    {
      "epoch": 0.024896265560165973,
      "grad_norm": 0.75390625,
      "learning_rate": 2.608695652173913e-05,
      "loss": 2.2494,
      "step": 3
    },
    {
      "epoch": 0.03319502074688797,
      "grad_norm": 0.8046875,
      "learning_rate": 3.478260869565218e-05,
      "loss": 2.2988,
      "step": 4
    },
    {
      "epoch": 0.04149377593360996,
      "grad_norm": 0.95703125,
      "learning_rate": 4.347826086956522e-05,
      "loss": 3.0938,
      "step": 5
    },
    {
      "epoch": 0.04979253112033195,
      "grad_norm": 0.7578125,
      "learning_rate": 5.217391304347826e-05,
      "loss": 2.1633,
      "step": 6
    },
    {
      "epoch": 0.058091286307053944,
      "grad_norm": 0.7421875,
      "learning_rate": 6.086956521739131e-05,
      "loss": 2.1488,
      "step": 7
    },
    {
      "epoch": 0.06639004149377593,
      "grad_norm": 0.62109375,
      "learning_rate": 6.956521739130436e-05,
      "loss": 2.0031,
      "step": 8
    },
    {
      "epoch": 0.07468879668049792,
      "grad_norm": 0.52734375,
      "learning_rate": 7.82608695652174e-05,
      "loss": 1.9347,
      "step": 9
    },
    {
      "epoch": 0.08298755186721991,
      "grad_norm": 0.369140625,
      "learning_rate": 8.695652173913044e-05,
      "loss": 1.8457,
      "step": 10
    },
    {
      "epoch": 0.0912863070539419,
      "grad_norm": 0.357421875,
      "learning_rate": 9.565217391304348e-05,
      "loss": 1.7273,
      "step": 11
    },
    {
      "epoch": 0.0995850622406639,
      "grad_norm": 0.376953125,
      "learning_rate": 0.00010434782608695653,
      "loss": 1.7584,
      "step": 12
    },
    {
      "epoch": 0.1078838174273859,
      "grad_norm": 0.39453125,
      "learning_rate": 0.00011304347826086956,
      "loss": 1.6387,
      "step": 13
    },
    {
      "epoch": 0.11618257261410789,
      "grad_norm": 0.458984375,
      "learning_rate": 0.00012173913043478263,
      "loss": 1.6614,
      "step": 14
    },
    {
      "epoch": 0.12448132780082988,
      "grad_norm": 0.482421875,
      "learning_rate": 0.00013043478260869567,
      "loss": 1.648,
      "step": 15
    },
    {
      "epoch": 0.13278008298755187,
      "grad_norm": 0.447265625,
      "learning_rate": 0.0001391304347826087,
      "loss": 1.6337,
      "step": 16
    },
    {
      "epoch": 0.14107883817427386,
      "grad_norm": 0.37109375,
      "learning_rate": 0.00014782608695652173,
      "loss": 1.6004,
      "step": 17
    },
    {
      "epoch": 0.14937759336099585,
      "grad_norm": 0.41796875,
      "learning_rate": 0.0001565217391304348,
      "loss": 1.6146,
      "step": 18
    },
    {
      "epoch": 0.15767634854771784,
      "grad_norm": 0.38671875,
      "learning_rate": 0.00016521739130434784,
      "loss": 1.4754,
      "step": 19
    },
    {
      "epoch": 0.16597510373443983,
      "grad_norm": 1.890625,
      "learning_rate": 0.00017391304347826088,
      "loss": 2.4256,
      "step": 20
    },
    {
      "epoch": 0.17427385892116182,
      "grad_norm": 0.296875,
      "learning_rate": 0.00018260869565217392,
      "loss": 1.4601,
      "step": 21
    },
    {
      "epoch": 0.1825726141078838,
      "grad_norm": 0.291015625,
      "learning_rate": 0.00019130434782608697,
      "loss": 1.4737,
      "step": 22
    },
    {
      "epoch": 0.1908713692946058,
      "grad_norm": 0.27734375,
      "learning_rate": 0.0002,
      "loss": 1.4942,
      "step": 23
    },
    {
      "epoch": 0.1991701244813278,
      "grad_norm": 0.259765625,
      "learning_rate": 0.00019998952044849376,
      "loss": 1.42,
      "step": 24
    },
    {
      "epoch": 0.2074688796680498,
      "grad_norm": 0.255859375,
      "learning_rate": 0.00019995808399039496,
      "loss": 1.4399,
      "step": 25
    },
    {
      "epoch": 0.2157676348547718,
      "grad_norm": 0.41796875,
      "learning_rate": 0.00019990569721450326,
      "loss": 2.1457,
      "step": 26
    },
    {
      "epoch": 0.22406639004149378,
      "grad_norm": 0.302734375,
      "learning_rate": 0.00019983237110061697,
      "loss": 1.4462,
      "step": 27
    },
    {
      "epoch": 0.23236514522821577,
      "grad_norm": 0.2890625,
      "learning_rate": 0.00019973812101723188,
      "loss": 1.3847,
      "step": 28
    },
    {
      "epoch": 0.24066390041493776,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.00019962296671832003,
      "loss": 1.4013,
      "step": 29
    },
    {
      "epoch": 0.24896265560165975,
      "grad_norm": 0.2578125,
      "learning_rate": 0.00019948693233918952,
      "loss": 1.3528,
      "step": 30
    },
    {
      "epoch": 0.2572614107883817,
      "grad_norm": 0.255859375,
      "learning_rate": 0.00019933004639142605,
      "loss": 1.3523,
      "step": 31
    },
    {
      "epoch": 0.26556016597510373,
      "grad_norm": 0.265625,
      "learning_rate": 0.000199152341756917,
      "loss": 1.36,
      "step": 32
    },
    {
      "epoch": 0.27385892116182575,
      "grad_norm": 0.26953125,
      "learning_rate": 0.00019895385568095982,
      "loss": 1.3316,
      "step": 33
    },
    {
      "epoch": 0.2821576763485477,
      "grad_norm": 0.265625,
      "learning_rate": 0.00019873462976445553,
      "loss": 1.371,
      "step": 34
    },
    {
      "epoch": 0.29045643153526973,
      "grad_norm": 0.2578125,
      "learning_rate": 0.00019849470995518992,
      "loss": 1.4089,
      "step": 35
    },
    {
      "epoch": 0.2987551867219917,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0001982341465382029,
      "loss": 1.3663,
      "step": 36
    },
    {
      "epoch": 0.3070539419087137,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00019795299412524945,
      "loss": 1.3573,
      "step": 37
    },
    {
      "epoch": 0.3153526970954357,
      "grad_norm": 0.24609375,
      "learning_rate": 0.00019765131164335345,
      "loss": 1.3048,
      "step": 38
    },
    {
      "epoch": 0.3236514522821577,
      "grad_norm": 0.7734375,
      "learning_rate": 0.000197329162322457,
      "loss": 2.0409,
      "step": 39
    },
    {
      "epoch": 0.33195020746887965,
      "grad_norm": 0.25390625,
      "learning_rate": 0.00019698661368216817,
      "loss": 1.3583,
      "step": 40
    },
    {
      "epoch": 0.34024896265560167,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.00019662373751760934,
      "loss": 1.2915,
      "step": 41
    },
    {
      "epoch": 0.34854771784232363,
      "grad_norm": 0.2578125,
      "learning_rate": 0.00019624060988436966,
      "loss": 1.3238,
      "step": 42
    },
    {
      "epoch": 0.35684647302904565,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.0001958373110825644,
      "loss": 1.3412,
      "step": 43
    },
    {
      "epoch": 0.3651452282157676,
      "grad_norm": 0.236328125,
      "learning_rate": 0.00019541392564000488,
      "loss": 1.2947,
      "step": 44
    },
    {
      "epoch": 0.37344398340248963,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.00019497054229448223,
      "loss": 1.3453,
      "step": 45
    },
    {
      "epoch": 0.3817427385892116,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0001945072539751685,
      "loss": 1.3186,
      "step": 46
    },
    {
      "epoch": 0.3900414937759336,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.00019402415778313977,
      "loss": 1.3524,
      "step": 47
    },
    {
      "epoch": 0.3983402489626556,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00019352135497102463,
      "loss": 1.3403,
      "step": 48
    },
    {
      "epoch": 0.4066390041493776,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0001929989509217824,
      "loss": 1.3266,
      "step": 49
    },
    {
      "epoch": 0.4149377593360996,
      "grad_norm": 0.251953125,
      "learning_rate": 0.0001924570551266159,
      "loss": 1.3373,
      "step": 50
    },
    {
      "epoch": 0.42323651452282157,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.00019189578116202307,
      "loss": 1.268,
      "step": 51
    },
    {
      "epoch": 0.4315352697095436,
      "grad_norm": 0.23828125,
      "learning_rate": 0.00019131524666599233,
      "loss": 1.2878,
      "step": 52
    },
    {
      "epoch": 0.43983402489626555,
      "grad_norm": 0.236328125,
      "learning_rate": 0.00019071557331334669,
      "loss": 1.2968,
      "step": 53
    },
    {
      "epoch": 0.44813278008298757,
      "grad_norm": 0.2353515625,
      "learning_rate": 0.0001900968867902419,
      "loss": 1.3174,
      "step": 54
    },
    {
      "epoch": 0.45643153526970953,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.00018945931676782373,
      "loss": 1.283,
      "step": 55
    },
    {
      "epoch": 0.46473029045643155,
      "grad_norm": 0.248046875,
      "learning_rate": 0.0001888029968750498,
      "loss": 1.279,
      "step": 56
    },
    {
      "epoch": 0.4730290456431535,
      "grad_norm": 0.240234375,
      "learning_rate": 0.00018812806467068268,
      "loss": 1.3151,
      "step": 57
    },
    {
      "epoch": 0.48132780082987553,
      "grad_norm": 0.244140625,
      "learning_rate": 0.00018743466161445823,
      "loss": 1.2763,
      "step": 58
    },
    {
      "epoch": 0.4896265560165975,
      "grad_norm": 0.25390625,
      "learning_rate": 0.00018672293303743738,
      "loss": 1.3259,
      "step": 59
    },
    {
      "epoch": 0.4979253112033195,
      "grad_norm": 0.255859375,
      "learning_rate": 0.00018599302811154572,
      "loss": 1.287,
      "step": 60
    },
    {
      "epoch": 0.5062240663900415,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00018524509981830852,
      "loss": 1.2794,
      "step": 61
    },
    {
      "epoch": 0.5145228215767634,
      "grad_norm": 0.255859375,
      "learning_rate": 0.00018447930491678733,
      "loss": 1.3154,
      "step": 62
    },
    {
      "epoch": 0.5228215767634855,
      "grad_norm": 0.248046875,
      "learning_rate": 0.00018369580391072433,
      "loss": 1.2979,
      "step": 63
    },
    {
      "epoch": 0.5311203319502075,
      "grad_norm": 0.2421875,
      "learning_rate": 0.00018289476101490256,
      "loss": 1.326,
      "step": 64
    },
    {
      "epoch": 0.5394190871369294,
      "grad_norm": 0.24609375,
      "learning_rate": 0.00018207634412072764,
      "loss": 1.2843,
      "step": 65
    },
    {
      "epoch": 0.5477178423236515,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.00018124072476103956,
      "loss": 1.3233,
      "step": 66
    },
    {
      "epoch": 0.5560165975103735,
      "grad_norm": 0.248046875,
      "learning_rate": 0.00018038807807416068,
      "loss": 1.3208,
      "step": 67
    },
    {
      "epoch": 0.5643153526970954,
      "grad_norm": 0.2431640625,
      "learning_rate": 0.00017951858276718844,
      "loss": 1.2949,
      "step": 68
    },
    {
      "epoch": 0.5726141078838174,
      "grad_norm": 0.412109375,
      "learning_rate": 0.00017863242107853995,
      "loss": 1.9205,
      "step": 69
    },
    {
      "epoch": 0.5809128630705395,
      "grad_norm": 0.2578125,
      "learning_rate": 0.0001777297787397563,
      "loss": 1.2847,
      "step": 70
    },
    {
      "epoch": 0.5892116182572614,
      "grad_norm": 0.2578125,
      "learning_rate": 0.00017681084493657525,
      "loss": 1.298,
      "step": 71
    },
    {
      "epoch": 0.5975103734439834,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0001758758122692791,
      "loss": 1.2805,
      "step": 72
    },
    {
      "epoch": 0.6058091286307054,
      "grad_norm": 0.25390625,
      "learning_rate": 0.00017492487671232784,
      "loss": 1.2829,
      "step": 73
    },
    {
      "epoch": 0.6141078838174274,
      "grad_norm": 0.2470703125,
      "learning_rate": 0.00017395823757328444,
      "loss": 1.2523,
      "step": 74
    },
    {
      "epoch": 0.6224066390041494,
      "grad_norm": 0.25,
      "learning_rate": 0.00017297609745104184,
      "loss": 1.2977,
      "step": 75
    },
    {
      "epoch": 0.6307053941908713,
      "grad_norm": 0.25,
      "learning_rate": 0.0001719786621933599,
      "loss": 1.3103,
      "step": 76
    },
    {
      "epoch": 0.6390041493775933,
      "grad_norm": 0.25390625,
      "learning_rate": 0.00017096614085372185,
      "loss": 1.3006,
      "step": 77
    },
    {
      "epoch": 0.6473029045643154,
      "grad_norm": 0.435546875,
      "learning_rate": 0.00016993874564751822,
      "loss": 1.8113,
      "step": 78
    },
    {
      "epoch": 0.6556016597510373,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00016889669190756868,
      "loss": 2.3268,
      "step": 79
    },
    {
      "epoch": 0.6639004149377593,
      "grad_norm": 0.2734375,
      "learning_rate": 0.00016784019803899,
      "loss": 1.2574,
      "step": 80
    },
    {
      "epoch": 0.6721991701244814,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0001667694854734204,
      "loss": 1.2365,
      "step": 81
    },
    {
      "epoch": 0.6804979253112033,
      "grad_norm": 0.26953125,
      "learning_rate": 0.0001656847786226095,
      "loss": 1.2974,
      "step": 82
    },
    {
      "epoch": 0.6887966804979253,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00016458630483138356,
      "loss": 1.2225,
      "step": 83
    },
    {
      "epoch": 0.6970954356846473,
      "grad_norm": 0.2734375,
      "learning_rate": 0.00016347429432999602,
      "loss": 1.2477,
      "step": 84
    },
    {
      "epoch": 0.7053941908713693,
      "grad_norm": 0.28125,
      "learning_rate": 0.00016234898018587337,
      "loss": 1.2872,
      "step": 85
    },
    {
      "epoch": 0.7136929460580913,
      "grad_norm": 0.271484375,
      "learning_rate": 0.0001612105982547663,
      "loss": 1.2824,
      "step": 86
    },
    {
      "epoch": 0.7219917012448133,
      "grad_norm": 0.26171875,
      "learning_rate": 0.00016005938713131642,
      "loss": 1.3056,
      "step": 87
    },
    {
      "epoch": 0.7302904564315352,
      "grad_norm": 0.265625,
      "learning_rate": 0.00015889558809904902,
      "loss": 1.244,
      "step": 88
    },
    {
      "epoch": 0.7385892116182573,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00015771944507980207,
      "loss": 1.2742,
      "step": 89
    },
    {
      "epoch": 0.7468879668049793,
      "grad_norm": 0.259765625,
      "learning_rate": 0.00015653120458260263,
      "loss": 1.3265,
      "step": 90
    },
    {
      "epoch": 0.7551867219917012,
      "grad_norm": 0.25,
      "learning_rate": 0.00015533111565200044,
      "loss": 1.2373,
      "step": 91
    },
    {
      "epoch": 0.7634854771784232,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0001541194298158708,
      "loss": 1.3201,
      "step": 92
    },
    {
      "epoch": 0.7717842323651453,
      "grad_norm": 0.27734375,
      "learning_rate": 0.00015289640103269625,
      "loss": 1.2052,
      "step": 93
    },
    {
      "epoch": 0.7800829875518672,
      "grad_norm": 0.283203125,
      "learning_rate": 0.00015166228563833934,
      "loss": 1.2638,
      "step": 94
    },
    {
      "epoch": 0.7883817427385892,
      "grad_norm": 0.2578125,
      "learning_rate": 0.00015041734229231688,
      "loss": 1.2334,
      "step": 95
    },
    {
      "epoch": 0.7966804979253111,
      "grad_norm": 0.263671875,
      "learning_rate": 0.00014916183192358718,
      "loss": 1.2396,
      "step": 96
    },
    {
      "epoch": 0.8049792531120332,
      "grad_norm": 0.2578125,
      "learning_rate": 0.00014789601767586173,
      "loss": 1.2272,
      "step": 97
    },
    {
      "epoch": 0.8132780082987552,
      "grad_norm": 0.28125,
      "learning_rate": 0.00014662016485245274,
      "loss": 1.2784,
      "step": 98
    },
    {
      "epoch": 0.8215767634854771,
      "grad_norm": 0.275390625,
      "learning_rate": 0.00014533454086066772,
      "loss": 1.2512,
      "step": 99
    },
    {
      "epoch": 0.8298755186721992,
      "grad_norm": 0.275390625,
      "learning_rate": 0.00014403941515576344,
      "loss": 1.2562,
      "step": 100
    },
    {
      "epoch": 0.8381742738589212,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00014273505918447054,
      "loss": 1.3054,
      "step": 101
    },
    {
      "epoch": 0.8464730290456431,
      "grad_norm": 0.263671875,
      "learning_rate": 0.00014142174632810072,
      "loss": 1.2831,
      "step": 102
    },
    {
      "epoch": 0.8547717842323651,
      "grad_norm": 0.279296875,
      "learning_rate": 0.0001400997518452484,
      "loss": 1.31,
      "step": 103
    },
    {
      "epoch": 0.8630705394190872,
      "grad_norm": 0.25,
      "learning_rate": 0.00013876935281409907,
      "loss": 1.2472,
      "step": 104
    },
    {
      "epoch": 0.8713692946058091,
      "grad_norm": 0.28515625,
      "learning_rate": 0.00013743082807435615,
      "loss": 1.2792,
      "step": 105
    },
    {
      "epoch": 0.8796680497925311,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00013608445816879866,
      "loss": 1.2853,
      "step": 106
    },
    {
      "epoch": 0.8879668049792531,
      "grad_norm": 0.265625,
      "learning_rate": 0.00013473052528448201,
      "loss": 1.2923,
      "step": 107
    },
    {
      "epoch": 0.8962655601659751,
      "grad_norm": 0.275390625,
      "learning_rate": 0.00013336931319359426,
      "loss": 1.2768,
      "step": 108
    },
    {
      "epoch": 0.9045643153526971,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00013200110719397968,
      "loss": 1.2701,
      "step": 109
    },
    {
      "epoch": 0.9128630705394191,
      "grad_norm": 0.25390625,
      "learning_rate": 0.00013062619404934317,
      "loss": 1.21,
      "step": 110
    },
    {
      "epoch": 0.921161825726141,
      "grad_norm": 0.255859375,
      "learning_rate": 0.00012924486192914705,
      "loss": 1.2808,
      "step": 111
    },
    {
      "epoch": 0.9294605809128631,
      "grad_norm": 0.26171875,
      "learning_rate": 0.00012785740034821329,
      "loss": 1.2616,
      "step": 112
    },
    {
      "epoch": 0.9377593360995851,
      "grad_norm": 0.265625,
      "learning_rate": 0.00012646410010604397,
      "loss": 1.2094,
      "step": 113
    },
    {
      "epoch": 0.946058091286307,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00012506525322587207,
      "loss": 1.2128,
      "step": 114
    },
    {
      "epoch": 0.9543568464730291,
      "grad_norm": 0.267578125,
      "learning_rate": 0.0001236611528934562,
      "loss": 1.2573,
      "step": 115
    },
    {
      "epoch": 0.9626556016597511,
      "grad_norm": 0.259765625,
      "learning_rate": 0.00012225209339563145,
      "loss": 1.2367,
      "step": 116
    },
    {
      "epoch": 0.970954356846473,
      "grad_norm": 0.263671875,
      "learning_rate": 0.00012083837005862946,
      "loss": 1.2313,
      "step": 117
    },
    {
      "epoch": 0.979253112033195,
      "grad_norm": 0.265625,
      "learning_rate": 0.00011942027918618074,
      "loss": 1.2495,
      "step": 118
    },
    {
      "epoch": 0.9875518672199171,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0001179981179974121,
      "loss": 1.2225,
      "step": 119
    },
    {
      "epoch": 0.995850622406639,
      "grad_norm": 0.28125,
      "learning_rate": 0.00011657218456455206,
      "loss": 1.2869,
      "step": 120
    },
    {
      "epoch": 1.004149377593361,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00011514277775045768,
      "loss": 1.2421,
      "step": 121
    },
    {
      "epoch": 1.012448132780083,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00011371019714597562,
      "loss": 1.2361,
      "step": 122
    },
    {
      "epoch": 1.0020746887966805,
      "grad_norm": 0.26953125,
      "learning_rate": 0.00011227474300715055,
      "loss": 1.2266,
      "step": 123
    },
    {
      "epoch": 1.0103734439834025,
      "grad_norm": 0.26953125,
      "learning_rate": 0.00011083671619229408,
      "loss": 1.1711,
      "step": 124
    },
    {
      "epoch": 1.0186721991701244,
      "grad_norm": 0.26171875,
      "learning_rate": 0.00010939641809892767,
      "loss": 1.1194,
      "step": 125
    },
    {
      "epoch": 1.0269709543568464,
      "grad_norm": 0.267578125,
      "learning_rate": 0.00010795415060061243,
      "loss": 1.2046,
      "step": 126
    },
    {
      "epoch": 1.0352697095435686,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00010651021598367906,
      "loss": 1.7307,
      "step": 127
    },
    {
      "epoch": 1.0435684647302905,
      "grad_norm": 0.26953125,
      "learning_rate": 0.00010506491688387127,
      "loss": 1.1314,
      "step": 128
    },
    {
      "epoch": 1.0518672199170125,
      "grad_norm": 0.26953125,
      "learning_rate": 0.00010361855622291637,
      "loss": 1.1103,
      "step": 129
    },
    {
      "epoch": 1.0601659751037344,
      "grad_norm": 0.287109375,
      "learning_rate": 0.00010217143714503508,
      "loss": 1.1383,
      "step": 130
    },
    {
      "epoch": 1.0684647302904564,
      "grad_norm": 0.306640625,
      "learning_rate": 0.00010072386295340572,
      "loss": 1.1588,
      "step": 131
    },
    {
      "epoch": 1.0767634854771784,
      "grad_norm": 0.291015625,
      "learning_rate": 9.927613704659429e-05,
      "loss": 1.1423,
      "step": 132
    },
    {
      "epoch": 1.0850622406639003,
      "grad_norm": 0.298828125,
      "learning_rate": 9.782856285496495e-05,
      "loss": 1.103,
      "step": 133
    },
    {
      "epoch": 1.0933609958506225,
      "grad_norm": 0.302734375,
      "learning_rate": 9.638144377708367e-05,
      "loss": 1.1682,
      "step": 134
    },
    {
      "epoch": 1.1016597510373445,
      "grad_norm": 0.296875,
      "learning_rate": 9.493508311612874e-05,
      "loss": 1.1089,
      "step": 135
    },
    {
      "epoch": 1.1099585062240664,
      "grad_norm": 0.291015625,
      "learning_rate": 9.348978401632101e-05,
      "loss": 1.072,
      "step": 136
    },
    {
      "epoch": 1.1182572614107884,
      "grad_norm": 0.296875,
      "learning_rate": 9.204584939938762e-05,
      "loss": 1.1051,
      "step": 137
    },
    {
      "epoch": 1.1265560165975104,
      "grad_norm": 0.29296875,
      "learning_rate": 9.060358190107234e-05,
      "loss": 1.1506,
      "step": 138
    },
    {
      "epoch": 1.1348547717842323,
      "grad_norm": 0.298828125,
      "learning_rate": 8.916328380770595e-05,
      "loss": 1.0714,
      "step": 139
    },
    {
      "epoch": 1.1431535269709543,
      "grad_norm": 0.294921875,
      "learning_rate": 8.772525699284946e-05,
      "loss": 1.1131,
      "step": 140
    },
    {
      "epoch": 1.1514522821576763,
      "grad_norm": 0.30078125,
      "learning_rate": 8.628980285402439e-05,
      "loss": 1.1183,
      "step": 141
    },
    {
      "epoch": 1.1597510373443982,
      "grad_norm": 0.3046875,
      "learning_rate": 8.485722224954237e-05,
      "loss": 1.1111,
      "step": 142
    },
    {
      "epoch": 1.1680497925311204,
      "grad_norm": 0.3125,
      "learning_rate": 8.342781543544798e-05,
      "loss": 1.0872,
      "step": 143
    },
    {
      "epoch": 1.1763485477178424,
      "grad_norm": 0.31640625,
      "learning_rate": 8.200188200258791e-05,
      "loss": 1.1321,
      "step": 144
    },
    {
      "epoch": 1.1846473029045643,
      "grad_norm": 0.322265625,
      "learning_rate": 8.057972081381927e-05,
      "loss": 1.1097,
      "step": 145
    },
    {
      "epoch": 1.1929460580912863,
      "grad_norm": 0.318359375,
      "learning_rate": 7.916162994137056e-05,
      "loss": 1.0885,
      "step": 146
    },
    {
      "epoch": 1.2012448132780082,
      "grad_norm": 0.31640625,
      "learning_rate": 7.774790660436858e-05,
      "loss": 1.105,
      "step": 147
    },
    {
      "epoch": 1.2095435684647302,
      "grad_norm": 0.314453125,
      "learning_rate": 7.633884710654383e-05,
      "loss": 1.0775,
      "step": 148
    },
    {
      "epoch": 1.2178423236514524,
      "grad_norm": 0.322265625,
      "learning_rate": 7.493474677412794e-05,
      "loss": 1.1075,
      "step": 149
    },
    {
      "epoch": 1.2261410788381744,
      "grad_norm": 0.328125,
      "learning_rate": 7.353589989395604e-05,
      "loss": 1.1398,
      "step": 150
    },
    {
      "epoch": 1.2344398340248963,
      "grad_norm": 0.333984375,
      "learning_rate": 7.214259965178674e-05,
      "loss": 1.1224,
      "step": 151
    },
    {
      "epoch": 1.2427385892116183,
      "grad_norm": 0.314453125,
      "learning_rate": 7.075513807085299e-05,
      "loss": 1.0898,
      "step": 152
    },
    {
      "epoch": 1.2510373443983402,
      "grad_norm": 0.33984375,
      "learning_rate": 6.937380595065685e-05,
      "loss": 1.0957,
      "step": 153
    },
    {
      "epoch": 1.2593360995850622,
      "grad_norm": 0.32421875,
      "learning_rate": 6.799889280602031e-05,
      "loss": 1.134,
      "step": 154
    },
    {
      "epoch": 1.2676348547717842,
      "grad_norm": 0.32421875,
      "learning_rate": 6.663068680640574e-05,
      "loss": 1.1226,
      "step": 155
    },
    {
      "epoch": 1.2759336099585061,
      "grad_norm": 0.318359375,
      "learning_rate": 6.526947471551798e-05,
      "loss": 1.1155,
      "step": 156
    },
    {
      "epoch": 1.284232365145228,
      "grad_norm": 0.3203125,
      "learning_rate": 6.391554183120138e-05,
      "loss": 1.1167,
      "step": 157
    },
    {
      "epoch": 1.2925311203319503,
      "grad_norm": 0.32421875,
      "learning_rate": 6.25691719256439e-05,
      "loss": 1.0482,
      "step": 158
    },
    {
      "epoch": 1.3008298755186722,
      "grad_norm": 0.326171875,
      "learning_rate": 6.123064718590099e-05,
      "loss": 1.0664,
      "step": 159
    },
    {
      "epoch": 1.3091286307053942,
      "grad_norm": 0.333984375,
      "learning_rate": 5.9900248154751616e-05,
      "loss": 1.1055,
      "step": 160
    },
    {
      "epoch": 1.3174273858921162,
      "grad_norm": 0.318359375,
      "learning_rate": 5.857825367189931e-05,
      "loss": 1.0993,
      "step": 161
    },
    {
      "epoch": 1.3257261410788381,
      "grad_norm": 0.333984375,
      "learning_rate": 5.7264940815529485e-05,
      "loss": 1.086,
      "step": 162
    },
    {
      "epoch": 1.33402489626556,
      "grad_norm": 0.328125,
      "learning_rate": 5.596058484423656e-05,
      "loss": 1.0708,
      "step": 163
    },
    {
      "epoch": 1.3423236514522823,
      "grad_norm": 1.0234375,
      "learning_rate": 5.46654591393323e-05,
      "loss": 1.7514,
      "step": 164
    },
    {
      "epoch": 1.3506224066390042,
      "grad_norm": 0.3359375,
      "learning_rate": 5.337983514754723e-05,
      "loss": 1.0925,
      "step": 165
    },
    {
      "epoch": 1.3589211618257262,
      "grad_norm": 0.333984375,
      "learning_rate": 5.2103982324138244e-05,
      "loss": 1.122,
      "step": 166
    },
    {
      "epoch": 1.3672199170124482,
      "grad_norm": 0.345703125,
      "learning_rate": 5.083816807641284e-05,
      "loss": 1.1124,
      "step": 167
    },
    {
      "epoch": 1.3755186721991701,
      "grad_norm": 0.330078125,
      "learning_rate": 4.958265770768316e-05,
      "loss": 1.0847,
      "step": 168
    },
    {
      "epoch": 1.383817427385892,
      "grad_norm": 0.33984375,
      "learning_rate": 4.833771436166069e-05,
      "loss": 1.1018,
      "step": 169
    },
    {
      "epoch": 1.392116182572614,
      "grad_norm": 0.34375,
      "learning_rate": 4.710359896730379e-05,
      "loss": 1.0666,
      "step": 170
    },
    {
      "epoch": 1.400414937759336,
      "grad_norm": 0.337890625,
      "learning_rate": 4.5880570184129215e-05,
      "loss": 1.0682,
      "step": 171
    },
    {
      "epoch": 1.408713692946058,
      "grad_norm": 0.33984375,
      "learning_rate": 4.466888434799958e-05,
      "loss": 1.0828,
      "step": 172
    },
    {
      "epoch": 1.4170124481327802,
      "grad_norm": 0.349609375,
      "learning_rate": 4.34687954173974e-05,
      "loss": 1.118,
      "step": 173
    },
    {
      "epoch": 1.4253112033195021,
      "grad_norm": 0.337890625,
      "learning_rate": 4.2280554920197936e-05,
      "loss": 1.0306,
      "step": 174
    },
    {
      "epoch": 1.433609958506224,
      "grad_norm": 0.3359375,
      "learning_rate": 4.1104411900951015e-05,
      "loss": 1.085,
      "step": 175
    },
    {
      "epoch": 1.441908713692946,
      "grad_norm": 0.349609375,
      "learning_rate": 3.994061286868361e-05,
      "loss": 1.1037,
      "step": 176
    },
    {
      "epoch": 1.450207468879668,
      "grad_norm": 0.67578125,
      "learning_rate": 3.878940174523371e-05,
      "loss": 1.5781,
      "step": 177
    },
    {
      "epoch": 1.45850622406639,
      "grad_norm": 0.3359375,
      "learning_rate": 3.7651019814126654e-05,
      "loss": 1.0691,
      "step": 178
    },
    {
      "epoch": 1.4668049792531122,
      "grad_norm": 0.345703125,
      "learning_rate": 3.652570567000402e-05,
      "loss": 1.0805,
      "step": 179
    },
    {
      "epoch": 1.4751037344398341,
      "grad_norm": 0.341796875,
      "learning_rate": 3.541369516861648e-05,
      "loss": 1.0789,
      "step": 180
    },
    {
      "epoch": 1.483402489626556,
      "grad_norm": 0.34765625,
      "learning_rate": 3.431522137739049e-05,
      "loss": 1.1362,
      "step": 181
    },
    {
      "epoch": 1.491701244813278,
      "grad_norm": 0.34375,
      "learning_rate": 3.323051452657961e-05,
      "loss": 1.0381,
      "step": 182
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.3359375,
      "learning_rate": 3.215980196101002e-05,
      "loss": 1.0476,
      "step": 183
    },
    {
      "epoch": 1.508298755186722,
      "grad_norm": 0.333984375,
      "learning_rate": 3.110330809243134e-05,
      "loss": 1.0587,
      "step": 184
    },
    {
      "epoch": 1.516597510373444,
      "grad_norm": 0.349609375,
      "learning_rate": 3.0061254352481804e-05,
      "loss": 1.0961,
      "step": 185
    },
    {
      "epoch": 1.5248962655601659,
      "grad_norm": 0.353515625,
      "learning_rate": 2.9033859146278197e-05,
      "loss": 1.0974,
      "step": 186
    },
    {
      "epoch": 1.5331950207468878,
      "grad_norm": 0.3515625,
      "learning_rate": 2.8021337806640135e-05,
      "loss": 1.0793,
      "step": 187
    },
    {
      "epoch": 1.5414937759336098,
      "grad_norm": 0.34765625,
      "learning_rate": 2.702390254895819e-05,
      "loss": 1.1351,
      "step": 188
    },
    {
      "epoch": 1.549792531120332,
      "grad_norm": 0.33984375,
      "learning_rate": 2.6041762426715566e-05,
      "loss": 1.0445,
      "step": 189
    },
    {
      "epoch": 1.558091286307054,
      "grad_norm": 0.341796875,
      "learning_rate": 2.5075123287672175e-05,
      "loss": 1.0722,
      "step": 190
    },
    {
      "epoch": 1.566390041493776,
      "grad_norm": 0.33984375,
      "learning_rate": 2.4124187730720917e-05,
      "loss": 1.1048,
      "step": 191
    },
    {
      "epoch": 1.5746887966804979,
      "grad_norm": 0.33984375,
      "learning_rate": 2.3189155063424782e-05,
      "loss": 1.0925,
      "step": 192
    },
    {
      "epoch": 1.58298755186722,
      "grad_norm": 0.3515625,
      "learning_rate": 2.2270221260243673e-05,
      "loss": 1.1181,
      "step": 193
    },
    {
      "epoch": 1.591286307053942,
      "grad_norm": 0.34765625,
      "learning_rate": 2.1367578921460074e-05,
      "loss": 1.0998,
      "step": 194
    },
    {
      "epoch": 1.599585062240664,
      "grad_norm": 0.34765625,
      "learning_rate": 2.0481417232811573e-05,
      "loss": 1.1088,
      "step": 195
    },
    {
      "epoch": 1.607883817427386,
      "grad_norm": 0.34375,
      "learning_rate": 1.961192192583934e-05,
      "loss": 1.0283,
      "step": 196
    },
    {
      "epoch": 1.616182572614108,
      "grad_norm": 0.349609375,
      "learning_rate": 1.8759275238960473e-05,
      "loss": 1.0841,
      "step": 197
    },
    {
      "epoch": 1.6244813278008299,
      "grad_norm": 0.330078125,
      "learning_rate": 1.7923655879272393e-05,
      "loss": 1.0461,
      "step": 198
    },
    {
      "epoch": 1.6327800829875518,
      "grad_norm": 0.6328125,
      "learning_rate": 1.7105238985097472e-05,
      "loss": 1.5407,
      "step": 199
    },
    {
      "epoch": 1.6410788381742738,
      "grad_norm": 0.34375,
      "learning_rate": 1.6304196089275658e-05,
      "loss": 1.0827,
      "step": 200
    },
    {
      "epoch": 1.6493775933609958,
      "grad_norm": 0.353515625,
      "learning_rate": 1.5520695083212678e-05,
      "loss": 1.1022,
      "step": 201
    },
    {
      "epoch": 1.6576763485477177,
      "grad_norm": 0.34375,
      "learning_rate": 1.4754900181691467e-05,
      "loss": 1.0796,
      "step": 202
    },
    {
      "epoch": 1.6659751037344397,
      "grad_norm": 0.345703125,
      "learning_rate": 1.4006971888454323e-05,
      "loss": 1.0797,
      "step": 203
    },
    {
      "epoch": 1.6742738589211619,
      "grad_norm": 0.345703125,
      "learning_rate": 1.3277066962562645e-05,
      "loss": 1.0898,
      "step": 204
    },
    {
      "epoch": 1.6825726141078838,
      "grad_norm": 0.35546875,
      "learning_rate": 1.2565338385541792e-05,
      "loss": 1.0774,
      "step": 205
    },
    {
      "epoch": 1.6908713692946058,
      "grad_norm": 0.345703125,
      "learning_rate": 1.1871935329317363e-05,
      "loss": 1.0741,
      "step": 206
    },
    {
      "epoch": 1.6991701244813278,
      "grad_norm": 0.361328125,
      "learning_rate": 1.1197003124950222e-05,
      "loss": 1.0568,
      "step": 207
    },
    {
      "epoch": 1.70746887966805,
      "grad_norm": 0.361328125,
      "learning_rate": 1.0540683232176307e-05,
      "loss": 1.125,
      "step": 208
    },
    {
      "epoch": 1.715767634854772,
      "grad_norm": 0.34765625,
      "learning_rate": 9.903113209758096e-06,
      "loss": 1.0842,
      "step": 209
    },
    {
      "epoch": 1.7240663900414939,
      "grad_norm": 0.3359375,
      "learning_rate": 9.284426686653303e-06,
      "loss": 1.086,
      "step": 210
    },
    {
      "epoch": 1.7323651452282158,
      "grad_norm": 0.3515625,
      "learning_rate": 8.68475333400769e-06,
      "loss": 1.0743,
      "step": 211
    },
    {
      "epoch": 1.7406639004149378,
      "grad_norm": 0.345703125,
      "learning_rate": 8.10421883797694e-06,
      "loss": 1.1064,
      "step": 212
    },
    {
      "epoch": 1.7489626556016598,
      "grad_norm": 0.5546875,
      "learning_rate": 7.542944873384106e-06,
      "loss": 1.4331,
      "step": 213
    },
    {
      "epoch": 1.7572614107883817,
      "grad_norm": 0.349609375,
      "learning_rate": 7.0010490782176145e-06,
      "loss": 1.0565,
      "step": 214
    },
    {
      "epoch": 1.7655601659751037,
      "grad_norm": 0.34765625,
      "learning_rate": 6.4786450289753715e-06,
      "loss": 1.1227,
      "step": 215
    },
    {
      "epoch": 1.7738589211618256,
      "grad_norm": 0.345703125,
      "learning_rate": 5.975842216860239e-06,
      "loss": 1.0866,
      "step": 216
    },
    {
      "epoch": 1.7821576763485476,
      "grad_norm": 0.345703125,
      "learning_rate": 5.492746024831541e-06,
      "loss": 1.0861,
      "step": 217
    },
    {
      "epoch": 1.7904564315352696,
      "grad_norm": 0.337890625,
      "learning_rate": 5.029457705517793e-06,
      "loss": 1.0773,
      "step": 218
    },
    {
      "epoch": 1.7987551867219918,
      "grad_norm": 0.359375,
      "learning_rate": 4.586074359995119e-06,
      "loss": 1.0457,
      "step": 219
    },
    {
      "epoch": 1.8070539419087137,
      "grad_norm": 0.345703125,
      "learning_rate": 4.162688917435631e-06,
      "loss": 1.0953,
      "step": 220
    },
    {
      "epoch": 1.8153526970954357,
      "grad_norm": 0.353515625,
      "learning_rate": 3.7593901156303566e-06,
      "loss": 1.0966,
      "step": 221
    },
    {
      "epoch": 1.8236514522821576,
      "grad_norm": 0.341796875,
      "learning_rate": 3.3762624823906573e-06,
      "loss": 1.0517,
      "step": 222
    },
    {
      "epoch": 1.8319502074688798,
      "grad_norm": 0.33984375,
      "learning_rate": 3.0133863178318232e-06,
      "loss": 1.0669,
      "step": 223
    },
    {
      "epoch": 1.8402489626556018,
      "grad_norm": 0.349609375,
      "learning_rate": 2.6708376775430033e-06,
      "loss": 1.091,
      "step": 224
    },
    {
      "epoch": 1.8485477178423237,
      "grad_norm": 0.34765625,
      "learning_rate": 2.3486883566465777e-06,
      "loss": 1.1095,
      "step": 225
    },
    {
      "epoch": 1.8568464730290457,
      "grad_norm": 0.60546875,
      "learning_rate": 2.0470058747505516e-06,
      "loss": 1.4381,
      "step": 226
    },
    {
      "epoch": 1.8651452282157677,
      "grad_norm": 0.3359375,
      "learning_rate": 1.7658534617971067e-06,
      "loss": 1.0979,
      "step": 227
    },
    {
      "epoch": 1.8734439834024896,
      "grad_norm": 0.35546875,
      "learning_rate": 1.5052900448100815e-06,
      "loss": 1.0746,
      "step": 228
    },
    {
      "epoch": 1.8817427385892116,
      "grad_norm": 0.34765625,
      "learning_rate": 1.2653702355444608e-06,
      "loss": 1.0931,
      "step": 229
    },
    {
      "epoch": 1.8900414937759336,
      "grad_norm": 0.357421875,
      "learning_rate": 1.0461443190402099e-06,
      "loss": 1.1229,
      "step": 230
    },
    {
      "epoch": 1.8983402489626555,
      "grad_norm": 0.345703125,
      "learning_rate": 8.476582430830049e-07,
      "loss": 1.1039,
      "step": 231
    },
    {
      "epoch": 1.9066390041493775,
      "grad_norm": 0.333984375,
      "learning_rate": 6.699536085739588e-07,
      "loss": 1.07,
      "step": 232
    },
    {
      "epoch": 1.9149377593360994,
      "grad_norm": 0.34375,
      "learning_rate": 5.130676608104845e-07,
      "loss": 1.1166,
      "step": 233
    },
    {
      "epoch": 1.9232365145228216,
      "grad_norm": 0.34375,
      "learning_rate": 3.7703328167999485e-07,
      "loss": 1.0579,
      "step": 234
    },
    {
      "epoch": 1.9315352697095436,
      "grad_norm": 0.33984375,
      "learning_rate": 2.6187898276813784e-07,
      "loss": 1.0934,
      "step": 235
    },
    {
      "epoch": 1.9398340248962656,
      "grad_norm": 0.37109375,
      "learning_rate": 1.6762889938303217e-07,
      "loss": 1.0792,
      "step": 236
    },
    {
      "epoch": 1.9481327800829875,
      "grad_norm": 0.337890625,
      "learning_rate": 9.430278549675819e-08,
      "loss": 1.1108,
      "step": 237
    },
    {
      "epoch": 1.9564315352697097,
      "grad_norm": 0.353515625,
      "learning_rate": 4.191600960505859e-08,
      "loss": 1.1207,
      "step": 238
    },
    {
      "epoch": 1.9647302904564317,
      "grad_norm": 0.349609375,
      "learning_rate": 1.0479551506259456e-08,
      "loss": 1.0674,
      "step": 239
    },
    {
      "epoch": 1.9730290456431536,
      "grad_norm": 0.357421875,
      "learning_rate": 0.0,
      "loss": 1.0868,
      "step": 240
    }
  ],
  "logging_steps": 1,
  "max_steps": 240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 120,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.94707278659584e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}