{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9977695167286247,
  "eval_steps": 500,
  "global_step": 504,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005947955390334572,
      "grad_norm": 1.824343204498291,
      "learning_rate": 0.0002,
      "loss": 3.1087,
      "step": 1
    },
    {
      "epoch": 0.011895910780669145,
      "grad_norm": 1.374577283859253,
      "learning_rate": 0.0002,
      "loss": 2.9898,
      "step": 2
    },
    {
      "epoch": 0.017843866171003718,
      "grad_norm": 1.5655758380889893,
      "learning_rate": 0.0002,
      "loss": 3.0999,
      "step": 3
    },
    {
      "epoch": 0.02379182156133829,
      "grad_norm": 5.276195049285889,
      "learning_rate": 0.0002,
      "loss": 2.7586,
      "step": 4
    },
    {
      "epoch": 0.02973977695167286,
      "grad_norm": 1.3304948806762695,
      "learning_rate": 0.0002,
      "loss": 2.7295,
      "step": 5
    },
    {
      "epoch": 0.035687732342007436,
      "grad_norm": 1.5444062948226929,
      "learning_rate": 0.0002,
      "loss": 2.5949,
      "step": 6
    },
    {
      "epoch": 0.041635687732342004,
      "grad_norm": 1.2866592407226562,
      "learning_rate": 0.0002,
      "loss": 2.4865,
      "step": 7
    },
    {
      "epoch": 0.04758364312267658,
      "grad_norm": 1.6175459623336792,
      "learning_rate": 0.0002,
      "loss": 2.3327,
      "step": 8
    },
    {
      "epoch": 0.053531598513011154,
      "grad_norm": 1.503796100616455,
      "learning_rate": 0.0002,
      "loss": 2.274,
      "step": 9
    },
    {
      "epoch": 0.05947955390334572,
      "grad_norm": 1.5973471403121948,
      "learning_rate": 0.0002,
      "loss": 2.3809,
      "step": 10
    },
    {
      "epoch": 0.0654275092936803,
      "grad_norm": 1.4748364686965942,
      "learning_rate": 0.0002,
      "loss": 2.1347,
      "step": 11
    },
    {
      "epoch": 0.07137546468401487,
      "grad_norm": 1.7337145805358887,
      "learning_rate": 0.0002,
      "loss": 1.9428,
      "step": 12
    },
    {
      "epoch": 0.07732342007434945,
      "grad_norm": 2.0839593410491943,
      "learning_rate": 0.0002,
      "loss": 2.0188,
      "step": 13
    },
    {
      "epoch": 0.08327137546468401,
      "grad_norm": 2.5018622875213623,
      "learning_rate": 0.0002,
      "loss": 1.7943,
      "step": 14
    },
    {
      "epoch": 0.08921933085501858,
      "grad_norm": 3.416013479232788,
      "learning_rate": 0.0002,
      "loss": 1.6812,
      "step": 15
    },
    {
      "epoch": 0.09516728624535316,
      "grad_norm": 4.241348743438721,
      "learning_rate": 0.0002,
      "loss": 1.6478,
      "step": 16
    },
    {
      "epoch": 0.10111524163568773,
      "grad_norm": 2.6027214527130127,
      "learning_rate": 0.0002,
      "loss": 1.5114,
      "step": 17
    },
    {
      "epoch": 0.10706319702602231,
      "grad_norm": 2.1649773120880127,
      "learning_rate": 0.0002,
      "loss": 1.5063,
      "step": 18
    },
    {
      "epoch": 0.11301115241635688,
      "grad_norm": 1.9704638719558716,
      "learning_rate": 0.0002,
      "loss": 1.3781,
      "step": 19
    },
    {
      "epoch": 0.11895910780669144,
      "grad_norm": 1.7165110111236572,
      "learning_rate": 0.0002,
      "loss": 1.3058,
      "step": 20
    },
    {
      "epoch": 0.12490706319702602,
      "grad_norm": 1.3949488401412964,
      "learning_rate": 0.0002,
      "loss": 1.3533,
      "step": 21
    },
    {
      "epoch": 0.1308550185873606,
      "grad_norm": 1.5300015211105347,
      "learning_rate": 0.0002,
      "loss": 1.2578,
      "step": 22
    },
    {
      "epoch": 0.13680297397769517,
      "grad_norm": 1.9964842796325684,
      "learning_rate": 0.0002,
      "loss": 1.2485,
      "step": 23
    },
    {
      "epoch": 0.14275092936802974,
      "grad_norm": 1.322247862815857,
      "learning_rate": 0.0002,
      "loss": 1.1887,
      "step": 24
    },
    {
      "epoch": 0.14869888475836432,
      "grad_norm": 1.2447245121002197,
      "learning_rate": 0.0002,
      "loss": 1.0807,
      "step": 25
    },
    {
      "epoch": 0.1546468401486989,
      "grad_norm": 1.2943564653396606,
      "learning_rate": 0.0002,
      "loss": 1.1295,
      "step": 26
    },
    {
      "epoch": 0.16059479553903347,
      "grad_norm": 1.2561174631118774,
      "learning_rate": 0.0002,
      "loss": 1.1011,
      "step": 27
    },
    {
      "epoch": 0.16654275092936802,
      "grad_norm": 1.305808663368225,
      "learning_rate": 0.0002,
      "loss": 1.1217,
      "step": 28
    },
    {
      "epoch": 0.1724907063197026,
      "grad_norm": 1.1935081481933594,
      "learning_rate": 0.0002,
      "loss": 1.0971,
      "step": 29
    },
    {
      "epoch": 0.17843866171003717,
      "grad_norm": 0.9849146008491516,
      "learning_rate": 0.0002,
      "loss": 0.9949,
      "step": 30
    },
    {
      "epoch": 0.18438661710037174,
      "grad_norm": 1.236385703086853,
      "learning_rate": 0.0002,
      "loss": 1.0533,
      "step": 31
    },
    {
      "epoch": 0.19033457249070632,
      "grad_norm": 1.091674566268921,
      "learning_rate": 0.0002,
      "loss": 0.9963,
      "step": 32
    },
    {
      "epoch": 0.1962825278810409,
      "grad_norm": 1.4563655853271484,
      "learning_rate": 0.0002,
      "loss": 1.0137,
      "step": 33
    },
    {
      "epoch": 0.20223048327137547,
      "grad_norm": 1.1690599918365479,
      "learning_rate": 0.0002,
      "loss": 0.9163,
      "step": 34
    },
    {
      "epoch": 0.20817843866171004,
      "grad_norm": 1.2094273567199707,
      "learning_rate": 0.0002,
      "loss": 1.0236,
      "step": 35
    },
    {
      "epoch": 0.21412639405204462,
      "grad_norm": 1.388743281364441,
      "learning_rate": 0.0002,
      "loss": 0.9591,
      "step": 36
    },
    {
      "epoch": 0.2200743494423792,
      "grad_norm": 1.2390081882476807,
      "learning_rate": 0.0002,
      "loss": 0.8751,
      "step": 37
    },
    {
      "epoch": 0.22602230483271377,
      "grad_norm": 2.5453145503997803,
      "learning_rate": 0.0002,
      "loss": 0.8459,
      "step": 38
    },
    {
      "epoch": 0.2319702602230483,
      "grad_norm": 1.3461544513702393,
      "learning_rate": 0.0002,
      "loss": 0.9373,
      "step": 39
    },
    {
      "epoch": 0.2379182156133829,
      "grad_norm": 1.2979869842529297,
      "learning_rate": 0.0002,
      "loss": 0.9373,
      "step": 40
    },
    {
      "epoch": 0.24386617100371746,
      "grad_norm": 53.14031219482422,
      "learning_rate": 0.0002,
      "loss": 0.7923,
      "step": 41
    },
    {
      "epoch": 0.24981412639405204,
      "grad_norm": 3.772839307785034,
      "learning_rate": 0.0002,
      "loss": 1.0239,
      "step": 42
    },
    {
      "epoch": 0.25576208178438664,
      "grad_norm": 1.684868574142456,
      "learning_rate": 0.0002,
      "loss": 0.8878,
      "step": 43
    },
    {
      "epoch": 0.2617100371747212,
      "grad_norm": 1.1423863172531128,
      "learning_rate": 0.0002,
      "loss": 0.896,
      "step": 44
    },
    {
      "epoch": 0.26765799256505574,
      "grad_norm": 1.3496270179748535,
      "learning_rate": 0.0002,
      "loss": 1.0198,
      "step": 45
    },
    {
      "epoch": 0.27360594795539034,
      "grad_norm": 1.2799283266067505,
      "learning_rate": 0.0002,
      "loss": 0.88,
      "step": 46
    },
    {
      "epoch": 0.2795539033457249,
      "grad_norm": 1.1613731384277344,
      "learning_rate": 0.0002,
      "loss": 0.9708,
      "step": 47
    },
    {
      "epoch": 0.2855018587360595,
      "grad_norm": 1.1430435180664062,
      "learning_rate": 0.0002,
      "loss": 0.8913,
      "step": 48
    },
    {
      "epoch": 0.29144981412639404,
      "grad_norm": 0.9552589654922485,
      "learning_rate": 0.0002,
      "loss": 0.8091,
      "step": 49
    },
    {
      "epoch": 0.29739776951672864,
      "grad_norm": 1.1052002906799316,
      "learning_rate": 0.0002,
      "loss": 0.9737,
      "step": 50
    },
    {
      "epoch": 0.3033457249070632,
      "grad_norm": 1.3066654205322266,
      "learning_rate": 0.0002,
      "loss": 0.9281,
      "step": 51
    },
    {
      "epoch": 0.3092936802973978,
      "grad_norm": 1.0277180671691895,
      "learning_rate": 0.0002,
      "loss": 0.96,
      "step": 52
    },
    {
      "epoch": 0.31524163568773234,
      "grad_norm": 1.208615779876709,
      "learning_rate": 0.0002,
      "loss": 0.8485,
      "step": 53
    },
    {
      "epoch": 0.32118959107806694,
      "grad_norm": 1.0156666040420532,
      "learning_rate": 0.0002,
      "loss": 0.9602,
      "step": 54
    },
    {
      "epoch": 0.3271375464684015,
      "grad_norm": 1.0822789669036865,
      "learning_rate": 0.0002,
      "loss": 0.891,
      "step": 55
    },
    {
      "epoch": 0.33308550185873603,
      "grad_norm": 1.063072681427002,
      "learning_rate": 0.0002,
      "loss": 0.7898,
      "step": 56
    },
    {
      "epoch": 0.33903345724907064,
      "grad_norm": 1.061710238456726,
      "learning_rate": 0.0002,
      "loss": 0.7903,
      "step": 57
    },
    {
      "epoch": 0.3449814126394052,
      "grad_norm": 1.1998765468597412,
      "learning_rate": 0.0002,
      "loss": 0.8346,
      "step": 58
    },
    {
      "epoch": 0.3509293680297398,
      "grad_norm": 1.083093285560608,
      "learning_rate": 0.0002,
      "loss": 0.8103,
      "step": 59
    },
    {
      "epoch": 0.35687732342007433,
      "grad_norm": 1.0685770511627197,
      "learning_rate": 0.0002,
      "loss": 0.7284,
      "step": 60
    },
    {
      "epoch": 0.36282527881040894,
      "grad_norm": 1.3935203552246094,
      "learning_rate": 0.0002,
      "loss": 0.8349,
      "step": 61
    },
    {
      "epoch": 0.3687732342007435,
      "grad_norm": 1.005191445350647,
      "learning_rate": 0.0002,
      "loss": 0.8611,
      "step": 62
    },
    {
      "epoch": 0.3747211895910781,
      "grad_norm": 1.1198813915252686,
      "learning_rate": 0.0002,
      "loss": 0.7994,
      "step": 63
    },
    {
      "epoch": 0.38066914498141263,
      "grad_norm": 1.454626202583313,
      "learning_rate": 0.0002,
      "loss": 0.7984,
      "step": 64
    },
    {
      "epoch": 0.38661710037174724,
      "grad_norm": 1.1353782415390015,
      "learning_rate": 0.0002,
      "loss": 0.7505,
      "step": 65
    },
    {
      "epoch": 0.3925650557620818,
      "grad_norm": 1.1953253746032715,
      "learning_rate": 0.0002,
      "loss": 0.7754,
      "step": 66
    },
    {
      "epoch": 0.39851301115241633,
      "grad_norm": 1.0996239185333252,
      "learning_rate": 0.0002,
      "loss": 0.8747,
      "step": 67
    },
    {
      "epoch": 0.40446096654275093,
      "grad_norm": 1.5701665878295898,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 68
    },
    {
      "epoch": 0.4104089219330855,
      "grad_norm": 1.29320228099823,
      "learning_rate": 0.0002,
      "loss": 0.7546,
      "step": 69
    },
    {
      "epoch": 0.4163568773234201,
      "grad_norm": 1.284342646598816,
      "learning_rate": 0.0002,
      "loss": 0.7324,
      "step": 70
    },
    {
      "epoch": 0.42230483271375463,
      "grad_norm": 1.0330944061279297,
      "learning_rate": 0.0002,
      "loss": 0.8614,
      "step": 71
    },
    {
      "epoch": 0.42825278810408923,
      "grad_norm": 1.0411959886550903,
      "learning_rate": 0.0002,
      "loss": 0.7431,
      "step": 72
    },
    {
      "epoch": 0.4342007434944238,
      "grad_norm": 1.2095258235931396,
      "learning_rate": 0.0002,
      "loss": 0.7909,
      "step": 73
    },
    {
      "epoch": 0.4401486988847584,
      "grad_norm": 1.3570586442947388,
      "learning_rate": 0.0002,
      "loss": 0.8113,
      "step": 74
    },
    {
      "epoch": 0.44609665427509293,
      "grad_norm": 1.0079586505889893,
      "learning_rate": 0.0002,
      "loss": 0.8457,
      "step": 75
    },
    {
      "epoch": 0.45204460966542753,
      "grad_norm": 0.9446130990982056,
      "learning_rate": 0.0002,
      "loss": 0.7934,
      "step": 76
    },
    {
      "epoch": 0.4579925650557621,
      "grad_norm": 1.0489394664764404,
      "learning_rate": 0.0002,
      "loss": 0.856,
      "step": 77
    },
    {
      "epoch": 0.4639405204460966,
      "grad_norm": 1.0112191438674927,
      "learning_rate": 0.0002,
      "loss": 0.7455,
      "step": 78
    },
    {
      "epoch": 0.46988847583643123,
      "grad_norm": 0.9976668953895569,
      "learning_rate": 0.0002,
      "loss": 0.8143,
      "step": 79
    },
    {
      "epoch": 0.4758364312267658,
      "grad_norm": 1.0991159677505493,
      "learning_rate": 0.0002,
      "loss": 0.6766,
      "step": 80
    },
    {
      "epoch": 0.4817843866171004,
      "grad_norm": 1.1794909238815308,
      "learning_rate": 0.0002,
      "loss": 0.6635,
      "step": 81
    },
    {
      "epoch": 0.4877323420074349,
      "grad_norm": 1.0414669513702393,
      "learning_rate": 0.0002,
      "loss": 0.8228,
      "step": 82
    },
    {
      "epoch": 0.49368029739776953,
      "grad_norm": 1.0767929553985596,
      "learning_rate": 0.0002,
      "loss": 0.8092,
      "step": 83
    },
    {
      "epoch": 0.4996282527881041,
      "grad_norm": 1.3375307321548462,
      "learning_rate": 0.0002,
      "loss": 0.7361,
      "step": 84
    },
    {
      "epoch": 0.5055762081784386,
      "grad_norm": 1.2492313385009766,
      "learning_rate": 0.0002,
      "loss": 0.8343,
      "step": 85
    },
    {
      "epoch": 0.5115241635687733,
      "grad_norm": 1.0948379039764404,
      "learning_rate": 0.0002,
      "loss": 0.7559,
      "step": 86
    },
    {
      "epoch": 0.5174721189591078,
      "grad_norm": 1.1456286907196045,
      "learning_rate": 0.0002,
      "loss": 0.7819,
      "step": 87
    },
    {
      "epoch": 0.5234200743494424,
      "grad_norm": 0.9729915857315063,
      "learning_rate": 0.0002,
      "loss": 0.7507,
      "step": 88
    },
    {
      "epoch": 0.5293680297397769,
      "grad_norm": 1.089845895767212,
      "learning_rate": 0.0002,
      "loss": 0.857,
      "step": 89
    },
    {
      "epoch": 0.5353159851301115,
      "grad_norm": 1.0552901029586792,
      "learning_rate": 0.0002,
      "loss": 0.8387,
      "step": 90
    },
    {
      "epoch": 0.5412639405204461,
      "grad_norm": 1.2134290933609009,
      "learning_rate": 0.0002,
      "loss": 0.845,
      "step": 91
    },
    {
      "epoch": 0.5472118959107807,
      "grad_norm": 1.2725780010223389,
      "learning_rate": 0.0002,
      "loss": 0.8203,
      "step": 92
    },
    {
      "epoch": 0.5531598513011152,
      "grad_norm": 1.3931224346160889,
      "learning_rate": 0.0002,
      "loss": 0.8428,
      "step": 93
    },
    {
      "epoch": 0.5591078066914498,
      "grad_norm": 1.0805130004882812,
      "learning_rate": 0.0002,
      "loss": 0.7855,
      "step": 94
    },
    {
      "epoch": 0.5650557620817844,
      "grad_norm": 1.018471598625183,
      "learning_rate": 0.0002,
      "loss": 0.6318,
      "step": 95
    },
    {
      "epoch": 0.571003717472119,
      "grad_norm": 3.2651963233947754,
      "learning_rate": 0.0002,
      "loss": 0.7039,
      "step": 96
    },
    {
      "epoch": 0.5769516728624535,
      "grad_norm": 0.9978213906288147,
      "learning_rate": 0.0002,
      "loss": 0.6734,
      "step": 97
    },
    {
      "epoch": 0.5828996282527881,
      "grad_norm": 0.8679234385490417,
      "learning_rate": 0.0002,
      "loss": 0.7277,
      "step": 98
    },
    {
      "epoch": 0.5888475836431227,
      "grad_norm": 1.1249589920043945,
      "learning_rate": 0.0002,
      "loss": 0.7392,
      "step": 99
    },
    {
      "epoch": 0.5947955390334573,
      "grad_norm": 0.9032052755355835,
      "learning_rate": 0.0002,
      "loss": 0.6876,
      "step": 100
    },
    {
      "epoch": 0.6007434944237918,
      "grad_norm": 0.9359114170074463,
      "learning_rate": 0.0002,
      "loss": 0.6448,
      "step": 101
    },
    {
      "epoch": 0.6066914498141264,
      "grad_norm": 1.076899528503418,
      "learning_rate": 0.0002,
      "loss": 0.7399,
      "step": 102
    },
    {
      "epoch": 0.6126394052044609,
      "grad_norm": 1.4630522727966309,
      "learning_rate": 0.0002,
      "loss": 0.6902,
      "step": 103
    },
    {
      "epoch": 0.6185873605947956,
      "grad_norm": 1.0862653255462646,
      "learning_rate": 0.0002,
      "loss": 0.854,
      "step": 104
    },
    {
      "epoch": 0.6245353159851301,
      "grad_norm": 1.3135863542556763,
      "learning_rate": 0.0002,
      "loss": 0.8188,
      "step": 105
    },
    {
      "epoch": 0.6304832713754647,
      "grad_norm": 0.9794917106628418,
      "learning_rate": 0.0002,
      "loss": 0.6275,
      "step": 106
    },
    {
      "epoch": 0.6364312267657992,
      "grad_norm": 1.02755868434906,
      "learning_rate": 0.0002,
      "loss": 0.7058,
      "step": 107
    },
    {
      "epoch": 0.6423791821561339,
      "grad_norm": 0.9642486572265625,
      "learning_rate": 0.0002,
      "loss": 0.8053,
      "step": 108
    },
    {
      "epoch": 0.6483271375464684,
      "grad_norm": 1.1192632913589478,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 109
    },
    {
      "epoch": 0.654275092936803,
      "grad_norm": 1.1808356046676636,
      "learning_rate": 0.0002,
      "loss": 0.678,
      "step": 110
    },
    {
      "epoch": 0.6602230483271375,
      "grad_norm": 1.2461023330688477,
      "learning_rate": 0.0002,
      "loss": 0.6892,
      "step": 111
    },
    {
      "epoch": 0.6661710037174721,
      "grad_norm": 1.0632222890853882,
      "learning_rate": 0.0002,
      "loss": 0.698,
      "step": 112
    },
    {
      "epoch": 0.6721189591078067,
      "grad_norm": 1.0353591442108154,
      "learning_rate": 0.0002,
      "loss": 0.7453,
      "step": 113
    },
    {
      "epoch": 0.6780669144981413,
      "grad_norm": 1.124794602394104,
      "learning_rate": 0.0002,
      "loss": 0.815,
      "step": 114
    },
    {
      "epoch": 0.6840148698884758,
      "grad_norm": 1.0341081619262695,
      "learning_rate": 0.0002,
      "loss": 0.7816,
      "step": 115
    },
    {
      "epoch": 0.6899628252788104,
      "grad_norm": 1.082952857017517,
      "learning_rate": 0.0002,
      "loss": 0.6825,
      "step": 116
    },
    {
      "epoch": 0.695910780669145,
      "grad_norm": 0.9126180410385132,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 117
    },
    {
      "epoch": 0.7018587360594796,
      "grad_norm": 1.2339016199111938,
      "learning_rate": 0.0002,
      "loss": 0.7759,
      "step": 118
    },
    {
      "epoch": 0.7078066914498141,
      "grad_norm": 1.5227537155151367,
      "learning_rate": 0.0002,
      "loss": 0.8574,
      "step": 119
    },
    {
      "epoch": 0.7137546468401487,
      "grad_norm": 1.0859841108322144,
      "learning_rate": 0.0002,
      "loss": 0.7241,
      "step": 120
    },
    {
      "epoch": 0.7197026022304833,
      "grad_norm": 1.0609203577041626,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 121
    },
    {
      "epoch": 0.7256505576208179,
      "grad_norm": 0.9025185704231262,
      "learning_rate": 0.0002,
      "loss": 0.7538,
      "step": 122
    },
    {
      "epoch": 0.7315985130111524,
      "grad_norm": 0.9280850291252136,
      "learning_rate": 0.0002,
      "loss": 0.8023,
      "step": 123
    },
    {
      "epoch": 0.737546468401487,
      "grad_norm": 1.0120896100997925,
      "learning_rate": 0.0002,
      "loss": 0.797,
      "step": 124
    },
    {
      "epoch": 0.7434944237918215,
      "grad_norm": 0.9294270277023315,
      "learning_rate": 0.0002,
      "loss": 0.7939,
      "step": 125
    },
    {
      "epoch": 0.7494423791821562,
      "grad_norm": 1.001685380935669,
      "learning_rate": 0.0002,
      "loss": 0.7297,
      "step": 126
    },
    {
      "epoch": 0.7553903345724907,
      "grad_norm": 1.0650714635849,
      "learning_rate": 0.0002,
      "loss": 0.7955,
      "step": 127
    },
    {
      "epoch": 0.7613382899628253,
      "grad_norm": 0.9343367218971252,
      "learning_rate": 0.0002,
      "loss": 0.7008,
      "step": 128
    },
    {
      "epoch": 0.7672862453531598,
      "grad_norm": 1.0042743682861328,
      "learning_rate": 0.0002,
      "loss": 0.8086,
      "step": 129
    },
    {
      "epoch": 0.7732342007434945,
      "grad_norm": 0.9538952708244324,
      "learning_rate": 0.0002,
      "loss": 0.6839,
      "step": 130
    },
    {
      "epoch": 0.779182156133829,
      "grad_norm": 1.0010913610458374,
      "learning_rate": 0.0002,
      "loss": 0.6247,
      "step": 131
    },
    {
      "epoch": 0.7851301115241636,
      "grad_norm": 0.8673060536384583,
      "learning_rate": 0.0002,
      "loss": 0.7232,
      "step": 132
    },
    {
      "epoch": 0.7910780669144981,
      "grad_norm": 1.070591688156128,
      "learning_rate": 0.0002,
      "loss": 0.785,
      "step": 133
    },
    {
      "epoch": 0.7970260223048327,
      "grad_norm": 1.0302468538284302,
      "learning_rate": 0.0002,
      "loss": 0.8044,
      "step": 134
    },
    {
      "epoch": 0.8029739776951673,
      "grad_norm": 1.0886098146438599,
      "learning_rate": 0.0002,
      "loss": 0.6984,
      "step": 135
    },
    {
      "epoch": 0.8089219330855019,
      "grad_norm": 0.9349246025085449,
      "learning_rate": 0.0002,
      "loss": 0.6711,
      "step": 136
    },
    {
      "epoch": 0.8148698884758364,
      "grad_norm": 1.2482446432113647,
      "learning_rate": 0.0002,
      "loss": 0.6468,
      "step": 137
    },
    {
      "epoch": 0.820817843866171,
      "grad_norm": 1.184043049812317,
      "learning_rate": 0.0002,
      "loss": 0.6909,
      "step": 138
    },
    {
      "epoch": 0.8267657992565056,
      "grad_norm": 0.8721855878829956,
      "learning_rate": 0.0002,
      "loss": 0.6553,
      "step": 139
    },
    {
      "epoch": 0.8327137546468402,
      "grad_norm": 1.30323326587677,
      "learning_rate": 0.0002,
      "loss": 0.7109,
      "step": 140
    },
    {
      "epoch": 0.8386617100371747,
      "grad_norm": 1.0187689065933228,
      "learning_rate": 0.0002,
      "loss": 0.6682,
      "step": 141
    },
    {
      "epoch": 0.8446096654275093,
      "grad_norm": 2.3475165367126465,
      "learning_rate": 0.0002,
      "loss": 0.7615,
      "step": 142
    },
    {
      "epoch": 0.8505576208178439,
      "grad_norm": 0.9803043603897095,
      "learning_rate": 0.0002,
      "loss": 0.7179,
      "step": 143
    },
    {
      "epoch": 0.8565055762081785,
      "grad_norm": 1.2290213108062744,
      "learning_rate": 0.0002,
      "loss": 0.8696,
      "step": 144
    },
    {
      "epoch": 0.862453531598513,
      "grad_norm": 1.1041066646575928,
      "learning_rate": 0.0002,
      "loss": 0.8196,
      "step": 145
    },
    {
      "epoch": 0.8684014869888476,
      "grad_norm": 0.9638866186141968,
      "learning_rate": 0.0002,
      "loss": 0.655,
      "step": 146
    },
    {
      "epoch": 0.8743494423791821,
      "grad_norm": 0.9777591824531555,
      "learning_rate": 0.0002,
      "loss": 0.8053,
      "step": 147
    },
    {
      "epoch": 0.8802973977695168,
      "grad_norm": 0.8717353343963623,
      "learning_rate": 0.0002,
      "loss": 0.6726,
      "step": 148
    },
    {
      "epoch": 0.8862453531598513,
      "grad_norm": 1.1772398948669434,
      "learning_rate": 0.0002,
      "loss": 0.7188,
      "step": 149
    },
    {
      "epoch": 0.8921933085501859,
      "grad_norm": 1.1113988161087036,
      "learning_rate": 0.0002,
      "loss": 0.7177,
      "step": 150
    },
    {
      "epoch": 0.8981412639405204,
      "grad_norm": 1.6691763401031494,
      "learning_rate": 0.0002,
      "loss": 0.5822,
      "step": 151
    },
    {
      "epoch": 0.9040892193308551,
      "grad_norm": 1.0139896869659424,
      "learning_rate": 0.0002,
      "loss": 0.7416,
      "step": 152
    },
    {
      "epoch": 0.9100371747211896,
      "grad_norm": 1.2538039684295654,
      "learning_rate": 0.0002,
      "loss": 0.7822,
      "step": 153
    },
    {
      "epoch": 0.9159851301115242,
      "grad_norm": 0.833595335483551,
      "learning_rate": 0.0002,
      "loss": 0.6617,
      "step": 154
    },
    {
      "epoch": 0.9219330855018587,
      "grad_norm": 0.869482696056366,
      "learning_rate": 0.0002,
      "loss": 0.6146,
      "step": 155
    },
    {
      "epoch": 0.9278810408921933,
      "grad_norm": 0.973523736000061,
      "learning_rate": 0.0002,
      "loss": 0.6685,
      "step": 156
    },
    {
      "epoch": 0.9338289962825279,
      "grad_norm": 0.982566237449646,
      "learning_rate": 0.0002,
      "loss": 0.6685,
      "step": 157
    },
    {
      "epoch": 0.9397769516728625,
      "grad_norm": 1.0534875392913818,
      "learning_rate": 0.0002,
      "loss": 0.656,
      "step": 158
    },
    {
      "epoch": 0.945724907063197,
      "grad_norm": 1.11860990524292,
      "learning_rate": 0.0002,
      "loss": 0.7261,
      "step": 159
    },
    {
      "epoch": 0.9516728624535316,
      "grad_norm": 1.0286844968795776,
      "learning_rate": 0.0002,
      "loss": 0.7271,
      "step": 160
    },
    {
      "epoch": 0.9576208178438662,
      "grad_norm": 0.8426340818405151,
      "learning_rate": 0.0002,
      "loss": 0.718,
      "step": 161
    },
    {
      "epoch": 0.9635687732342008,
      "grad_norm": 0.990667998790741,
      "learning_rate": 0.0002,
      "loss": 0.7795,
      "step": 162
    },
    {
      "epoch": 0.9695167286245353,
      "grad_norm": 1.1110923290252686,
      "learning_rate": 0.0002,
      "loss": 0.6946,
      "step": 163
    },
    {
      "epoch": 0.9754646840148699,
      "grad_norm": 1.0378597974777222,
      "learning_rate": 0.0002,
      "loss": 0.8106,
      "step": 164
    },
    {
      "epoch": 0.9814126394052045,
      "grad_norm": 0.9507467746734619,
      "learning_rate": 0.0002,
      "loss": 0.7031,
      "step": 165
    },
    {
      "epoch": 0.9873605947955391,
      "grad_norm": 0.8636868596076965,
      "learning_rate": 0.0002,
      "loss": 0.6636,
      "step": 166
    },
    {
      "epoch": 0.9933085501858736,
      "grad_norm": 1.0482003688812256,
      "learning_rate": 0.0002,
      "loss": 0.7204,
      "step": 167
    },
    {
      "epoch": 0.9992565055762082,
      "grad_norm": 0.9490022659301758,
      "learning_rate": 0.0002,
      "loss": 0.5626,
      "step": 168
    },
    {
      "epoch": 1.0052044609665427,
      "grad_norm": 0.8918917179107666,
      "learning_rate": 0.0002,
      "loss": 0.6822,
      "step": 169
    },
    {
      "epoch": 1.0111524163568772,
      "grad_norm": 0.9100430011749268,
      "learning_rate": 0.0002,
      "loss": 0.7147,
      "step": 170
    },
    {
      "epoch": 1.0171003717472118,
      "grad_norm": 0.9007158279418945,
      "learning_rate": 0.0002,
      "loss": 0.6015,
      "step": 171
    },
    {
      "epoch": 1.0230483271375466,
      "grad_norm": 0.9267099499702454,
      "learning_rate": 0.0002,
      "loss": 0.6562,
      "step": 172
    },
    {
      "epoch": 1.0289962825278811,
      "grad_norm": 1.0618972778320312,
      "learning_rate": 0.0002,
      "loss": 0.6601,
      "step": 173
    },
    {
      "epoch": 1.0349442379182157,
      "grad_norm": 1.1782737970352173,
      "learning_rate": 0.0002,
      "loss": 0.6172,
      "step": 174
    },
    {
      "epoch": 1.0408921933085502,
      "grad_norm": 1.141661286354065,
      "learning_rate": 0.0002,
      "loss": 0.5854,
      "step": 175
    },
    {
      "epoch": 1.0468401486988848,
      "grad_norm": 1.1038105487823486,
      "learning_rate": 0.0002,
      "loss": 0.6655,
      "step": 176
    },
    {
      "epoch": 1.0527881040892193,
      "grad_norm": 1.1518810987472534,
      "learning_rate": 0.0002,
      "loss": 0.493,
      "step": 177
    },
    {
      "epoch": 1.0587360594795538,
      "grad_norm": 1.0501494407653809,
      "learning_rate": 0.0002,
      "loss": 0.6256,
      "step": 178
    },
    {
      "epoch": 1.0646840148698884,
      "grad_norm": 0.9064037799835205,
      "learning_rate": 0.0002,
      "loss": 0.5695,
      "step": 179
    },
    {
      "epoch": 1.070631970260223,
      "grad_norm": 1.1978446245193481,
      "learning_rate": 0.0002,
      "loss": 0.617,
      "step": 180
    },
    {
      "epoch": 1.0765799256505577,
      "grad_norm": 0.9782500267028809,
      "learning_rate": 0.0002,
      "loss": 0.6542,
      "step": 181
    },
    {
      "epoch": 1.0825278810408923,
      "grad_norm": 1.6090043783187866,
      "learning_rate": 0.0002,
      "loss": 0.6827,
      "step": 182
    },
    {
      "epoch": 1.0884758364312268,
      "grad_norm": 1.2153990268707275,
      "learning_rate": 0.0002,
      "loss": 0.7469,
      "step": 183
    },
    {
      "epoch": 1.0944237918215614,
      "grad_norm": 0.9915666580200195,
      "learning_rate": 0.0002,
      "loss": 0.6475,
      "step": 184
    },
    {
      "epoch": 1.100371747211896,
      "grad_norm": 1.1319215297698975,
      "learning_rate": 0.0002,
      "loss": 0.7879,
      "step": 185
    },
    {
      "epoch": 1.1063197026022304,
      "grad_norm": 1.0497454404830933,
      "learning_rate": 0.0002,
      "loss": 0.6386,
      "step": 186
    },
    {
      "epoch": 1.112267657992565,
      "grad_norm": 1.1735246181488037,
      "learning_rate": 0.0002,
      "loss": 0.7276,
      "step": 187
    },
    {
      "epoch": 1.1182156133828995,
      "grad_norm": 1.280543327331543,
      "learning_rate": 0.0002,
      "loss": 0.6403,
      "step": 188
    },
    {
      "epoch": 1.124163568773234,
      "grad_norm": 1.0103743076324463,
      "learning_rate": 0.0002,
      "loss": 0.5913,
      "step": 189
    },
    {
      "epoch": 1.1301115241635689,
      "grad_norm": 1.0629348754882812,
      "learning_rate": 0.0002,
      "loss": 0.708,
      "step": 190
    },
    {
      "epoch": 1.1360594795539034,
      "grad_norm": 0.9152292609214783,
      "learning_rate": 0.0002,
      "loss": 0.8295,
      "step": 191
    },
    {
      "epoch": 1.142007434944238,
      "grad_norm": 0.9847885370254517,
      "learning_rate": 0.0002,
      "loss": 0.6569,
      "step": 192
    },
    {
      "epoch": 1.1479553903345725,
      "grad_norm": 1.5211213827133179,
      "learning_rate": 0.0002,
      "loss": 0.676,
      "step": 193
    },
    {
      "epoch": 1.153903345724907,
      "grad_norm": 1.0376240015029907,
      "learning_rate": 0.0002,
      "loss": 0.5558,
      "step": 194
    },
    {
      "epoch": 1.1598513011152416,
      "grad_norm": 0.9746745824813843,
      "learning_rate": 0.0002,
      "loss": 0.6051,
      "step": 195
    },
    {
      "epoch": 1.1657992565055761,
      "grad_norm": 1.0810937881469727,
      "learning_rate": 0.0002,
      "loss": 0.5947,
      "step": 196
    },
    {
      "epoch": 1.1717472118959107,
      "grad_norm": 0.8687509894371033,
      "learning_rate": 0.0002,
      "loss": 0.5834,
      "step": 197
    },
    {
      "epoch": 1.1776951672862452,
      "grad_norm": 1.3437882661819458,
      "learning_rate": 0.0002,
      "loss": 0.7133,
      "step": 198
    },
    {
      "epoch": 1.18364312267658,
      "grad_norm": 0.9247745871543884,
      "learning_rate": 0.0002,
      "loss": 0.668,
      "step": 199
    },
    {
      "epoch": 1.1895910780669146,
      "grad_norm": 1.116870403289795,
      "learning_rate": 0.0002,
      "loss": 0.5763,
      "step": 200
    },
    {
      "epoch": 1.195539033457249,
      "grad_norm": 1.0791046619415283,
      "learning_rate": 0.0002,
      "loss": 0.7535,
      "step": 201
    },
    {
      "epoch": 1.2014869888475836,
      "grad_norm": 1.1156578063964844,
      "learning_rate": 0.0002,
      "loss": 0.623,
      "step": 202
    },
    {
      "epoch": 1.2074349442379182,
      "grad_norm": 1.3306505680084229,
      "learning_rate": 0.0002,
      "loss": 0.6239,
      "step": 203
    },
    {
      "epoch": 1.2133828996282527,
      "grad_norm": 1.1251856088638306,
      "learning_rate": 0.0002,
      "loss": 0.6122,
      "step": 204
    },
    {
      "epoch": 1.2193308550185873,
      "grad_norm": 0.9659932255744934,
      "learning_rate": 0.0002,
      "loss": 0.7163,
      "step": 205
    },
    {
      "epoch": 1.2252788104089218,
      "grad_norm": 1.1080381870269775,
      "learning_rate": 0.0002,
      "loss": 0.5558,
      "step": 206
    },
    {
      "epoch": 1.2312267657992564,
      "grad_norm": 1.1085773706436157,
      "learning_rate": 0.0002,
      "loss": 0.6306,
      "step": 207
    },
    {
      "epoch": 1.2371747211895912,
      "grad_norm": 1.5555293560028076,
      "learning_rate": 0.0002,
      "loss": 0.7266,
      "step": 208
    },
    {
      "epoch": 1.2431226765799257,
      "grad_norm": 1.1494146585464478,
      "learning_rate": 0.0002,
      "loss": 0.6902,
      "step": 209
    },
    {
      "epoch": 1.2490706319702602,
      "grad_norm": 1.0835429430007935,
      "learning_rate": 0.0002,
      "loss": 0.7219,
      "step": 210
    },
    {
      "epoch": 1.2550185873605948,
      "grad_norm": 1.1306850910186768,
      "learning_rate": 0.0002,
      "loss": 0.5931,
      "step": 211
    },
    {
      "epoch": 1.2609665427509293,
      "grad_norm": 1.148278832435608,
      "learning_rate": 0.0002,
      "loss": 0.6452,
      "step": 212
    },
    {
      "epoch": 1.266914498141264,
      "grad_norm": 1.097596526145935,
      "learning_rate": 0.0002,
      "loss": 0.6486,
      "step": 213
    },
    {
      "epoch": 1.2728624535315984,
      "grad_norm": 1.3407150506973267,
      "learning_rate": 0.0002,
      "loss": 0.5874,
      "step": 214
    },
    {
      "epoch": 1.2788104089219332,
      "grad_norm": 0.8781871199607849,
      "learning_rate": 0.0002,
      "loss": 0.6079,
      "step": 215
    },
    {
      "epoch": 1.2847583643122675,
      "grad_norm": 0.998188853263855,
      "learning_rate": 0.0002,
      "loss": 0.7292,
      "step": 216
    },
    {
      "epoch": 1.2907063197026023,
      "grad_norm": 1.0128471851348877,
      "learning_rate": 0.0002,
      "loss": 0.6729,
      "step": 217
    },
    {
      "epoch": 1.2966542750929368,
      "grad_norm": 1.2172327041625977,
      "learning_rate": 0.0002,
      "loss": 0.7529,
      "step": 218
    },
    {
      "epoch": 1.3026022304832714,
      "grad_norm": 0.9904006719589233,
      "learning_rate": 0.0002,
      "loss": 0.639,
      "step": 219
    },
    {
      "epoch": 1.308550185873606,
      "grad_norm": 0.8886059522628784,
      "learning_rate": 0.0002,
      "loss": 0.6128,
      "step": 220
    },
    {
      "epoch": 1.3144981412639405,
      "grad_norm": 1.0350927114486694,
      "learning_rate": 0.0002,
      "loss": 0.6668,
      "step": 221
    },
    {
      "epoch": 1.320446096654275,
      "grad_norm": 1.0321650505065918,
      "learning_rate": 0.0002,
      "loss": 0.6493,
      "step": 222
    },
    {
      "epoch": 1.3263940520446096,
      "grad_norm": 0.8952768445014954,
      "learning_rate": 0.0002,
      "loss": 0.7005,
      "step": 223
    },
    {
      "epoch": 1.3323420074349444,
      "grad_norm": 1.3372063636779785,
      "learning_rate": 0.0002,
      "loss": 0.6347,
      "step": 224
    },
    {
      "epoch": 1.3382899628252787,
      "grad_norm": 0.9312218427658081,
      "learning_rate": 0.0002,
      "loss": 0.5963,
      "step": 225
    },
    {
      "epoch": 1.3442379182156134,
      "grad_norm": 0.8845749497413635,
      "learning_rate": 0.0002,
      "loss": 0.5445,
      "step": 226
    },
    {
      "epoch": 1.350185873605948,
      "grad_norm": 1.292598843574524,
      "learning_rate": 0.0002,
      "loss": 0.6662,
      "step": 227
    },
    {
      "epoch": 1.3561338289962825,
      "grad_norm": 1.0537996292114258,
      "learning_rate": 0.0002,
      "loss": 0.7332,
      "step": 228
    },
    {
      "epoch": 1.362081784386617,
      "grad_norm": 0.9492632150650024,
      "learning_rate": 0.0002,
      "loss": 0.6157,
      "step": 229
    },
    {
      "epoch": 1.3680297397769516,
      "grad_norm": 1.0352752208709717,
      "learning_rate": 0.0002,
      "loss": 0.5248,
      "step": 230
    },
    {
      "epoch": 1.3739776951672862,
      "grad_norm": 1.085534930229187,
      "learning_rate": 0.0002,
      "loss": 0.6358,
      "step": 231
    },
    {
      "epoch": 1.3799256505576207,
      "grad_norm": 1.098999261856079,
      "learning_rate": 0.0002,
      "loss": 0.6181,
      "step": 232
    },
    {
      "epoch": 1.3858736059479555,
      "grad_norm": 1.114450454711914,
      "learning_rate": 0.0002,
      "loss": 0.7067,
      "step": 233
    },
    {
      "epoch": 1.3918215613382898,
      "grad_norm": 1.3746989965438843,
      "learning_rate": 0.0002,
      "loss": 0.6016,
      "step": 234
    },
    {
      "epoch": 1.3977695167286246,
      "grad_norm": 0.916519820690155,
      "learning_rate": 0.0002,
      "loss": 0.6579,
      "step": 235
    },
    {
      "epoch": 1.4037174721189591,
      "grad_norm": 1.0786117315292358,
      "learning_rate": 0.0002,
      "loss": 0.6516,
      "step": 236
    },
    {
      "epoch": 1.4096654275092937,
      "grad_norm": 0.9264970421791077,
      "learning_rate": 0.0002,
      "loss": 0.5927,
      "step": 237
    },
    {
      "epoch": 1.4156133828996282,
      "grad_norm": 1.0969526767730713,
      "learning_rate": 0.0002,
      "loss": 0.6747,
      "step": 238
    },
    {
      "epoch": 1.4215613382899628,
      "grad_norm": 0.9945991635322571,
      "learning_rate": 0.0002,
      "loss": 0.6719,
      "step": 239
    },
    {
      "epoch": 1.4275092936802973,
      "grad_norm": 1.0272929668426514,
      "learning_rate": 0.0002,
      "loss": 0.7015,
      "step": 240
    },
    {
      "epoch": 1.4334572490706319,
      "grad_norm": 1.2321354150772095,
      "learning_rate": 0.0002,
      "loss": 0.6904,
      "step": 241
    },
    {
      "epoch": 1.4394052044609666,
      "grad_norm": 1.1331416368484497,
      "learning_rate": 0.0002,
      "loss": 0.6444,
      "step": 242
    },
    {
      "epoch": 1.4453531598513012,
      "grad_norm": 1.0527664422988892,
      "learning_rate": 0.0002,
      "loss": 0.6135,
      "step": 243
    },
    {
      "epoch": 1.4513011152416357,
      "grad_norm": 1.0586967468261719,
      "learning_rate": 0.0002,
      "loss": 0.6705,
      "step": 244
    },
    {
      "epoch": 1.4572490706319703,
      "grad_norm": 1.0302836894989014,
      "learning_rate": 0.0002,
      "loss": 0.6757,
      "step": 245
    },
    {
      "epoch": 1.4631970260223048,
      "grad_norm": 0.9323686957359314,
      "learning_rate": 0.0002,
      "loss": 0.7458,
      "step": 246
    },
    {
      "epoch": 1.4691449814126394,
      "grad_norm": 1.103028416633606,
      "learning_rate": 0.0002,
      "loss": 0.5832,
      "step": 247
    },
    {
      "epoch": 1.475092936802974,
      "grad_norm": 1.1638356447219849,
      "learning_rate": 0.0002,
      "loss": 0.6286,
      "step": 248
    },
    {
      "epoch": 1.4810408921933085,
      "grad_norm": 1.0685887336730957,
      "learning_rate": 0.0002,
      "loss": 0.6255,
      "step": 249
    },
    {
      "epoch": 1.486988847583643,
      "grad_norm": 0.9854826927185059,
      "learning_rate": 0.0002,
      "loss": 0.7764,
      "step": 250
    },
    {
      "epoch": 1.4929368029739778,
      "grad_norm": 1.1790441274642944,
      "learning_rate": 0.0002,
      "loss": 0.5791,
      "step": 251
    },
    {
      "epoch": 1.4988847583643123,
      "grad_norm": 0.9097880721092224,
      "learning_rate": 0.0002,
      "loss": 0.5517,
      "step": 252
    },
    {
      "epoch": 1.504832713754647,
      "grad_norm": 1.1351274251937866,
      "learning_rate": 0.0002,
      "loss": 0.6725,
      "step": 253
    },
    {
      "epoch": 1.5107806691449814,
      "grad_norm": 1.2710050344467163,
      "learning_rate": 0.0002,
      "loss": 0.7106,
      "step": 254
    },
    {
      "epoch": 1.516728624535316,
      "grad_norm": 1.2035406827926636,
      "learning_rate": 0.0002,
      "loss": 0.623,
      "step": 255
    },
    {
      "epoch": 1.5226765799256505,
      "grad_norm": 1.100200891494751,
      "learning_rate": 0.0002,
      "loss": 0.6418,
      "step": 256
    },
    {
      "epoch": 1.528624535315985,
      "grad_norm": 1.3976622819900513,
      "learning_rate": 0.0002,
      "loss": 0.7007,
      "step": 257
    },
    {
      "epoch": 1.5345724907063198,
      "grad_norm": 1.2113823890686035,
      "learning_rate": 0.0002,
      "loss": 0.7271,
      "step": 258
    },
    {
      "epoch": 1.5405204460966542,
      "grad_norm": 1.1983304023742676,
      "learning_rate": 0.0002,
      "loss": 0.6937,
      "step": 259
    },
    {
      "epoch": 1.546468401486989,
      "grad_norm": 1.2594386339187622,
      "learning_rate": 0.0002,
      "loss": 0.708,
      "step": 260
    },
    {
      "epoch": 1.5524163568773233,
      "grad_norm": 1.1495513916015625,
      "learning_rate": 0.0002,
      "loss": 0.621,
      "step": 261
    },
    {
      "epoch": 1.558364312267658,
      "grad_norm": 1.0474885702133179,
      "learning_rate": 0.0002,
      "loss": 0.5433,
      "step": 262
    },
    {
      "epoch": 1.5643122676579926,
      "grad_norm": 1.1138205528259277,
      "learning_rate": 0.0002,
      "loss": 0.6886,
      "step": 263
    },
    {
      "epoch": 1.5702602230483271,
      "grad_norm": 0.9678700566291809,
      "learning_rate": 0.0002,
      "loss": 0.5945,
      "step": 264
    },
    {
      "epoch": 1.5762081784386617,
      "grad_norm": 0.928419828414917,
      "learning_rate": 0.0002,
      "loss": 0.7483,
      "step": 265
    },
    {
      "epoch": 1.5821561338289962,
      "grad_norm": 0.8806396126747131,
      "learning_rate": 0.0002,
      "loss": 0.6022,
      "step": 266
    },
    {
      "epoch": 1.588104089219331,
      "grad_norm": 0.9389284253120422,
      "learning_rate": 0.0002,
      "loss": 0.6638,
      "step": 267
    },
    {
      "epoch": 1.5940520446096653,
      "grad_norm": 1.0797287225723267,
      "learning_rate": 0.0002,
      "loss": 0.6298,
      "step": 268
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.9545785784721375,
      "learning_rate": 0.0002,
      "loss": 0.6335,
      "step": 269
    },
    {
      "epoch": 1.6059479553903344,
      "grad_norm": 0.9800273776054382,
      "learning_rate": 0.0002,
      "loss": 0.5772,
      "step": 270
    },
    {
      "epoch": 1.6118959107806692,
      "grad_norm": 1.3683196306228638,
      "learning_rate": 0.0002,
      "loss": 0.6181,
      "step": 271
    },
    {
      "epoch": 1.6178438661710037,
      "grad_norm": 1.1559855937957764,
      "learning_rate": 0.0002,
      "loss": 0.62,
      "step": 272
    },
    {
      "epoch": 1.6237918215613383,
      "grad_norm": 1.1240603923797607,
      "learning_rate": 0.0002,
      "loss": 0.682,
      "step": 273
    },
    {
      "epoch": 1.6297397769516728,
      "grad_norm": 0.9673051834106445,
      "learning_rate": 0.0002,
      "loss": 0.6835,
      "step": 274
    },
    {
      "epoch": 1.6356877323420074,
      "grad_norm": 1.1218955516815186,
      "learning_rate": 0.0002,
      "loss": 0.6591,
      "step": 275
    },
    {
      "epoch": 1.6416356877323421,
      "grad_norm": 1.2360399961471558,
      "learning_rate": 0.0002,
      "loss": 0.7957,
      "step": 276
    },
    {
      "epoch": 1.6475836431226765,
      "grad_norm": 1.2180172204971313,
      "learning_rate": 0.0002,
      "loss": 0.6951,
      "step": 277
    },
    {
      "epoch": 1.6535315985130112,
      "grad_norm": 1.2104121446609497,
      "learning_rate": 0.0002,
      "loss": 0.6549,
      "step": 278
    },
    {
      "epoch": 1.6594795539033456,
      "grad_norm": 0.9836241006851196,
      "learning_rate": 0.0002,
      "loss": 0.6077,
      "step": 279
    },
    {
      "epoch": 1.6654275092936803,
      "grad_norm": 0.8980191349983215,
      "learning_rate": 0.0002,
      "loss": 0.7533,
      "step": 280
    },
    {
      "epoch": 1.6713754646840149,
      "grad_norm": 1.056117296218872,
      "learning_rate": 0.0002,
      "loss": 0.6283,
      "step": 281
    },
    {
      "epoch": 1.6773234200743494,
      "grad_norm": 1.0315310955047607,
      "learning_rate": 0.0002,
      "loss": 0.7099,
      "step": 282
    },
    {
      "epoch": 1.683271375464684,
      "grad_norm": 1.1293710470199585,
      "learning_rate": 0.0002,
      "loss": 0.7628,
      "step": 283
    },
    {
      "epoch": 1.6892193308550185,
      "grad_norm": 0.8841990232467651,
      "learning_rate": 0.0002,
      "loss": 0.6076,
      "step": 284
    },
    {
      "epoch": 1.6951672862453533,
      "grad_norm": 1.0221779346466064,
      "learning_rate": 0.0002,
      "loss": 0.6003,
      "step": 285
    },
    {
      "epoch": 1.7011152416356876,
      "grad_norm": 0.9923282861709595,
      "learning_rate": 0.0002,
      "loss": 0.5743,
      "step": 286
    },
    {
      "epoch": 1.7070631970260224,
      "grad_norm": 1.1585432291030884,
      "learning_rate": 0.0002,
      "loss": 0.6121,
      "step": 287
    },
    {
      "epoch": 1.713011152416357,
      "grad_norm": 0.9201356172561646,
      "learning_rate": 0.0002,
      "loss": 0.6343,
      "step": 288
    },
    {
      "epoch": 1.7189591078066915,
      "grad_norm": 1.164581298828125,
      "learning_rate": 0.0002,
      "loss": 0.615,
      "step": 289
    },
    {
      "epoch": 1.724907063197026,
      "grad_norm": 0.9991989135742188,
      "learning_rate": 0.0002,
      "loss": 0.62,
      "step": 290
    },
    {
      "epoch": 1.7308550185873606,
      "grad_norm": 1.0976234674453735,
      "learning_rate": 0.0002,
      "loss": 0.687,
      "step": 291
    },
    {
      "epoch": 1.7368029739776951,
      "grad_norm": 1.1581001281738281,
      "learning_rate": 0.0002,
      "loss": 0.6227,
      "step": 292
    },
    {
      "epoch": 1.7427509293680297,
      "grad_norm": 1.0079922676086426,
      "learning_rate": 0.0002,
      "loss": 0.675,
      "step": 293
    },
    {
      "epoch": 1.7486988847583644,
      "grad_norm": 1.0962276458740234,
      "learning_rate": 0.0002,
      "loss": 0.6694,
      "step": 294
    },
    {
      "epoch": 1.7546468401486988,
      "grad_norm": 1.0988850593566895,
      "learning_rate": 0.0002,
      "loss": 0.6114,
      "step": 295
    },
    {
      "epoch": 1.7605947955390335,
      "grad_norm": 1.4446635246276855,
      "learning_rate": 0.0002,
      "loss": 0.5885,
      "step": 296
    },
    {
      "epoch": 1.766542750929368,
      "grad_norm": 1.2141138315200806,
      "learning_rate": 0.0002,
      "loss": 0.7856,
      "step": 297
    },
    {
      "epoch": 1.7724907063197026,
      "grad_norm": 1.1908177137374878,
      "learning_rate": 0.0002,
      "loss": 0.7033,
      "step": 298
    },
    {
      "epoch": 1.7784386617100372,
      "grad_norm": 1.019839882850647,
      "learning_rate": 0.0002,
      "loss": 0.6763,
      "step": 299
    },
    {
      "epoch": 1.7843866171003717,
      "grad_norm": 1.039696216583252,
      "learning_rate": 0.0002,
      "loss": 0.6279,
      "step": 300
    },
    {
      "epoch": 1.7903345724907063,
      "grad_norm": 0.974805474281311,
      "learning_rate": 0.0002,
      "loss": 0.5777,
      "step": 301
    },
    {
      "epoch": 1.7962825278810408,
      "grad_norm": 1.1052793264389038,
      "learning_rate": 0.0002,
      "loss": 0.7294,
      "step": 302
    },
    {
      "epoch": 1.8022304832713756,
      "grad_norm": 1.4657918214797974,
      "learning_rate": 0.0002,
      "loss": 0.5142,
      "step": 303
    },
    {
      "epoch": 1.80817843866171,
      "grad_norm": 1.0391294956207275,
      "learning_rate": 0.0002,
      "loss": 0.6388,
      "step": 304
    },
    {
      "epoch": 1.8141263940520447,
      "grad_norm": 1.0521687269210815,
      "learning_rate": 0.0002,
      "loss": 0.6944,
      "step": 305
    },
    {
      "epoch": 1.8200743494423792,
      "grad_norm": 1.0755914449691772,
      "learning_rate": 0.0002,
      "loss": 0.6974,
      "step": 306
    },
    {
      "epoch": 1.8260223048327138,
      "grad_norm": 1.128304123878479,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 307
    },
    {
      "epoch": 1.8319702602230483,
      "grad_norm": 1.0178970098495483,
      "learning_rate": 0.0002,
      "loss": 0.6074,
      "step": 308
    },
    {
      "epoch": 1.8379182156133829,
      "grad_norm": 0.9115421772003174,
      "learning_rate": 0.0002,
      "loss": 0.725,
      "step": 309
    },
    {
      "epoch": 1.8438661710037176,
      "grad_norm": 1.0200258493423462,
      "learning_rate": 0.0002,
      "loss": 0.6866,
      "step": 310
    },
    {
      "epoch": 1.849814126394052,
      "grad_norm": 1.286431908607483,
      "learning_rate": 0.0002,
      "loss": 0.6618,
      "step": 311
    },
    {
      "epoch": 1.8557620817843867,
      "grad_norm": 1.0576943159103394,
      "learning_rate": 0.0002,
      "loss": 0.6217,
      "step": 312
    },
    {
      "epoch": 1.861710037174721,
      "grad_norm": 0.9450961351394653,
      "learning_rate": 0.0002,
      "loss": 0.7285,
      "step": 313
    },
    {
      "epoch": 1.8676579925650558,
      "grad_norm": 1.2659786939620972,
      "learning_rate": 0.0002,
      "loss": 0.6004,
      "step": 314
    },
    {
      "epoch": 1.8736059479553904,
      "grad_norm": 1.0950329303741455,
      "learning_rate": 0.0002,
      "loss": 0.6792,
      "step": 315
    },
    {
      "epoch": 1.879553903345725,
      "grad_norm": 1.0434305667877197,
      "learning_rate": 0.0002,
      "loss": 0.6651,
      "step": 316
    },
    {
      "epoch": 1.8855018587360595,
      "grad_norm": 2.390085458755493,
      "learning_rate": 0.0002,
      "loss": 0.6394,
      "step": 317
    },
    {
      "epoch": 1.891449814126394,
      "grad_norm": 1.1428786516189575,
      "learning_rate": 0.0002,
      "loss": 0.6519,
      "step": 318
    },
    {
      "epoch": 1.8973977695167288,
      "grad_norm": 1.1516354084014893,
      "learning_rate": 0.0002,
      "loss": 0.5967,
      "step": 319
    },
    {
      "epoch": 1.903345724907063,
      "grad_norm": 0.9553952813148499,
      "learning_rate": 0.0002,
      "loss": 0.6626,
      "step": 320
    },
    {
      "epoch": 1.9092936802973979,
      "grad_norm": 1.1295243501663208,
      "learning_rate": 0.0002,
      "loss": 0.6151,
      "step": 321
    },
    {
      "epoch": 1.9152416356877322,
      "grad_norm": 1.1593585014343262,
      "learning_rate": 0.0002,
      "loss": 0.6412,
      "step": 322
    },
    {
      "epoch": 1.921189591078067,
      "grad_norm": 1.830063819885254,
      "learning_rate": 0.0002,
      "loss": 0.6746,
      "step": 323
    },
    {
      "epoch": 1.9271375464684015,
      "grad_norm": 1.3906419277191162,
      "learning_rate": 0.0002,
      "loss": 0.6351,
      "step": 324
    },
    {
      "epoch": 1.933085501858736,
      "grad_norm": 1.4869827032089233,
      "learning_rate": 0.0002,
      "loss": 0.6806,
      "step": 325
    },
    {
      "epoch": 1.9390334572490706,
      "grad_norm": 1.110323429107666,
      "learning_rate": 0.0002,
      "loss": 0.5748,
      "step": 326
    },
    {
      "epoch": 1.9449814126394052,
      "grad_norm": 1.4225271940231323,
      "learning_rate": 0.0002,
      "loss": 0.6572,
      "step": 327
    },
    {
      "epoch": 1.95092936802974,
      "grad_norm": 1.0343074798583984,
      "learning_rate": 0.0002,
      "loss": 0.5376,
      "step": 328
    },
    {
      "epoch": 1.9568773234200743,
      "grad_norm": 0.9949336647987366,
      "learning_rate": 0.0002,
      "loss": 0.69,
      "step": 329
    },
    {
      "epoch": 1.962825278810409,
      "grad_norm": 1.0380656719207764,
      "learning_rate": 0.0002,
      "loss": 0.63,
      "step": 330
    },
    {
      "epoch": 1.9687732342007433,
      "grad_norm": 1.390371561050415,
      "learning_rate": 0.0002,
      "loss": 0.7011,
      "step": 331
    },
    {
      "epoch": 1.9747211895910781,
      "grad_norm": 1.4939589500427246,
      "learning_rate": 0.0002,
      "loss": 0.5689,
      "step": 332
    },
    {
      "epoch": 1.9806691449814127,
      "grad_norm": 1.2601418495178223,
      "learning_rate": 0.0002,
      "loss": 0.5838,
      "step": 333
    },
    {
      "epoch": 1.9866171003717472,
      "grad_norm": 2.679206132888794,
      "learning_rate": 0.0002,
      "loss": 0.7981,
      "step": 334
    },
    {
      "epoch": 1.9925650557620818,
      "grad_norm": 1.1042869091033936,
      "learning_rate": 0.0002,
      "loss": 0.6212,
      "step": 335
    },
    {
      "epoch": 1.9985130111524163,
      "grad_norm": 1.4491620063781738,
      "learning_rate": 0.0002,
      "loss": 0.6194,
      "step": 336
    },
    {
      "epoch": 2.004460966542751,
      "grad_norm": 0.9622808694839478,
      "learning_rate": 0.0002,
      "loss": 0.6566,
      "step": 337
    },
    {
      "epoch": 2.0104089219330854,
      "grad_norm": 1.5044182538986206,
      "learning_rate": 0.0002,
      "loss": 0.6093,
      "step": 338
    },
    {
      "epoch": 2.01635687732342,
      "grad_norm": 1.699040174484253,
      "learning_rate": 0.0002,
      "loss": 0.5237,
      "step": 339
    },
    {
      "epoch": 2.0223048327137545,
      "grad_norm": 1.1767878532409668,
      "learning_rate": 0.0002,
      "loss": 0.6145,
      "step": 340
    },
    {
      "epoch": 2.0282527881040893,
      "grad_norm": 1.2151747941970825,
      "learning_rate": 0.0002,
      "loss": 0.6171,
      "step": 341
    },
    {
      "epoch": 2.0342007434944236,
      "grad_norm": 1.2429864406585693,
      "learning_rate": 0.0002,
      "loss": 0.6489,
      "step": 342
    },
    {
      "epoch": 2.0401486988847584,
      "grad_norm": 1.164552092552185,
      "learning_rate": 0.0002,
      "loss": 0.509,
      "step": 343
    },
    {
      "epoch": 2.046096654275093,
      "grad_norm": 1.1822024583816528,
      "learning_rate": 0.0002,
      "loss": 0.6568,
      "step": 344
    },
    {
      "epoch": 2.0520446096654275,
      "grad_norm": 7.130686283111572,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 345
    },
    {
      "epoch": 2.0579925650557622,
      "grad_norm": 1.1000553369522095,
      "learning_rate": 0.0002,
      "loss": 0.5202,
      "step": 346
    },
    {
      "epoch": 2.0639405204460965,
      "grad_norm": 1.0652920007705688,
      "learning_rate": 0.0002,
      "loss": 0.5797,
      "step": 347
    },
    {
      "epoch": 2.0698884758364313,
      "grad_norm": 1.3442667722702026,
      "learning_rate": 0.0002,
      "loss": 0.6164,
      "step": 348
    },
    {
      "epoch": 2.0758364312267656,
      "grad_norm": 1.1383881568908691,
      "learning_rate": 0.0002,
      "loss": 0.5462,
      "step": 349
    },
    {
      "epoch": 2.0817843866171004,
      "grad_norm": 0.9077207446098328,
      "learning_rate": 0.0002,
      "loss": 0.4876,
      "step": 350
    },
    {
      "epoch": 2.0877323420074347,
      "grad_norm": 1.0893263816833496,
      "learning_rate": 0.0002,
      "loss": 0.6335,
      "step": 351
    },
    {
      "epoch": 2.0936802973977695,
      "grad_norm": 0.9917628169059753,
      "learning_rate": 0.0002,
      "loss": 0.5745,
      "step": 352
    },
    {
      "epoch": 2.0996282527881043,
      "grad_norm": 1.0131194591522217,
      "learning_rate": 0.0002,
      "loss": 0.5427,
      "step": 353
    },
    {
      "epoch": 2.1055762081784386,
      "grad_norm": 2.081542730331421,
      "learning_rate": 0.0002,
      "loss": 0.5692,
      "step": 354
    },
    {
      "epoch": 2.1115241635687734,
      "grad_norm": 1.659941554069519,
      "learning_rate": 0.0002,
      "loss": 0.4985,
      "step": 355
    },
    {
      "epoch": 2.1174721189591077,
      "grad_norm": 1.309465765953064,
      "learning_rate": 0.0002,
      "loss": 0.5439,
      "step": 356
    },
    {
      "epoch": 2.1234200743494425,
      "grad_norm": 1.2212107181549072,
      "learning_rate": 0.0002,
      "loss": 0.5133,
      "step": 357
    },
    {
      "epoch": 2.129368029739777,
      "grad_norm": 1.0576850175857544,
      "learning_rate": 0.0002,
      "loss": 0.6627,
      "step": 358
    },
    {
      "epoch": 2.1353159851301116,
      "grad_norm": 1.2587406635284424,
      "learning_rate": 0.0002,
      "loss": 0.5235,
      "step": 359
    },
    {
      "epoch": 2.141263940520446,
      "grad_norm": 1.05579674243927,
      "learning_rate": 0.0002,
      "loss": 0.5994,
      "step": 360
    },
    {
      "epoch": 2.1472118959107807,
      "grad_norm": 1.2232978343963623,
      "learning_rate": 0.0002,
      "loss": 0.586,
      "step": 361
    },
    {
      "epoch": 2.1531598513011154,
      "grad_norm": 1.0725406408309937,
      "learning_rate": 0.0002,
      "loss": 0.4801,
      "step": 362
    },
    {
      "epoch": 2.1591078066914497,
      "grad_norm": 1.0593106746673584,
      "learning_rate": 0.0002,
      "loss": 0.4877,
      "step": 363
    },
    {
      "epoch": 2.1650557620817845,
      "grad_norm": 2.2175445556640625,
      "learning_rate": 0.0002,
      "loss": 0.57,
      "step": 364
    },
    {
      "epoch": 2.171003717472119,
      "grad_norm": 1.1013628244400024,
      "learning_rate": 0.0002,
      "loss": 0.5797,
      "step": 365
    },
    {
      "epoch": 2.1769516728624536,
      "grad_norm": 1.186463713645935,
      "learning_rate": 0.0002,
      "loss": 0.5938,
      "step": 366
    },
    {
      "epoch": 2.182899628252788,
      "grad_norm": 1.1608301401138306,
      "learning_rate": 0.0002,
      "loss": 0.4446,
      "step": 367
    },
    {
      "epoch": 2.1888475836431227,
      "grad_norm": 1.2304465770721436,
      "learning_rate": 0.0002,
      "loss": 0.5154,
      "step": 368
    },
    {
      "epoch": 2.194795539033457,
      "grad_norm": 1.1233623027801514,
      "learning_rate": 0.0002,
      "loss": 0.6199,
      "step": 369
    },
    {
      "epoch": 2.200743494423792,
      "grad_norm": 1.2339355945587158,
      "learning_rate": 0.0002,
      "loss": 0.584,
      "step": 370
    },
    {
      "epoch": 2.2066914498141266,
      "grad_norm": 1.6794264316558838,
      "learning_rate": 0.0002,
      "loss": 0.5609,
      "step": 371
    },
    {
      "epoch": 2.212639405204461,
      "grad_norm": 1.1440285444259644,
      "learning_rate": 0.0002,
      "loss": 0.5306,
      "step": 372
    },
    {
      "epoch": 2.2185873605947957,
      "grad_norm": 1.6425179243087769,
      "learning_rate": 0.0002,
      "loss": 0.6556,
      "step": 373
    },
    {
      "epoch": 2.22453531598513,
      "grad_norm": 1.1068412065505981,
      "learning_rate": 0.0002,
      "loss": 0.6442,
      "step": 374
    },
    {
      "epoch": 2.2304832713754648,
      "grad_norm": 1.1996163129806519,
      "learning_rate": 0.0002,
      "loss": 0.51,
      "step": 375
    },
    {
      "epoch": 2.236431226765799,
      "grad_norm": 1.193741798400879,
      "learning_rate": 0.0002,
      "loss": 0.5323,
      "step": 376
    },
    {
      "epoch": 2.242379182156134,
      "grad_norm": 1.3267923593521118,
      "learning_rate": 0.0002,
      "loss": 0.5576,
      "step": 377
    },
    {
      "epoch": 2.248327137546468,
      "grad_norm": 1.1256170272827148,
      "learning_rate": 0.0002,
      "loss": 0.6218,
      "step": 378
    },
    {
      "epoch": 2.254275092936803,
      "grad_norm": 1.1282093524932861,
      "learning_rate": 0.0002,
      "loss": 0.5095,
      "step": 379
    },
    {
      "epoch": 2.2602230483271377,
      "grad_norm": 1.156480073928833,
      "learning_rate": 0.0002,
      "loss": 0.5423,
      "step": 380
    },
    {
      "epoch": 2.266171003717472,
      "grad_norm": 1.2083227634429932,
      "learning_rate": 0.0002,
      "loss": 0.5714,
      "step": 381
    },
    {
      "epoch": 2.272118959107807,
      "grad_norm": 1.5389211177825928,
      "learning_rate": 0.0002,
      "loss": 0.4797,
      "step": 382
    },
    {
      "epoch": 2.278066914498141,
      "grad_norm": 1.1418561935424805,
      "learning_rate": 0.0002,
      "loss": 0.585,
      "step": 383
    },
    {
      "epoch": 2.284014869888476,
      "grad_norm": 1.0242007970809937,
      "learning_rate": 0.0002,
      "loss": 0.562,
      "step": 384
    },
    {
      "epoch": 2.2899628252788102,
      "grad_norm": 1.4750384092330933,
      "learning_rate": 0.0002,
      "loss": 0.558,
      "step": 385
    },
    {
      "epoch": 2.295910780669145,
      "grad_norm": 1.176080346107483,
      "learning_rate": 0.0002,
      "loss": 0.4977,
      "step": 386
    },
    {
      "epoch": 2.3018587360594793,
      "grad_norm": 1.1733489036560059,
      "learning_rate": 0.0002,
      "loss": 0.5984,
      "step": 387
    },
    {
      "epoch": 2.307806691449814,
      "grad_norm": 1.0431591272354126,
      "learning_rate": 0.0002,
      "loss": 0.5805,
      "step": 388
    },
    {
      "epoch": 2.313754646840149,
      "grad_norm": 1.1595654487609863,
      "learning_rate": 0.0002,
      "loss": 0.5639,
      "step": 389
    },
    {
      "epoch": 2.319702602230483,
      "grad_norm": 1.2077865600585938,
      "learning_rate": 0.0002,
      "loss": 0.5021,
      "step": 390
    },
    {
      "epoch": 2.325650557620818,
      "grad_norm": 1.4747991561889648,
      "learning_rate": 0.0002,
      "loss": 0.5809,
      "step": 391
    },
    {
      "epoch": 2.3315985130111523,
      "grad_norm": 1.0486669540405273,
      "learning_rate": 0.0002,
      "loss": 0.6334,
      "step": 392
    },
    {
      "epoch": 2.337546468401487,
      "grad_norm": 1.8145817518234253,
      "learning_rate": 0.0002,
      "loss": 0.5223,
      "step": 393
    },
    {
      "epoch": 2.3434944237918214,
      "grad_norm": 1.393776535987854,
      "learning_rate": 0.0002,
      "loss": 0.4034,
      "step": 394
    },
    {
      "epoch": 2.349442379182156,
      "grad_norm": 1.208957314491272,
      "learning_rate": 0.0002,
      "loss": 0.5929,
      "step": 395
    },
    {
      "epoch": 2.3553903345724905,
      "grad_norm": 1.3021739721298218,
      "learning_rate": 0.0002,
      "loss": 0.6176,
      "step": 396
    },
    {
      "epoch": 2.3613382899628252,
      "grad_norm": 1.4258112907409668,
      "learning_rate": 0.0002,
      "loss": 0.6765,
      "step": 397
    },
    {
      "epoch": 2.36728624535316,
      "grad_norm": 1.2623789310455322,
      "learning_rate": 0.0002,
      "loss": 0.5973,
      "step": 398
    },
    {
      "epoch": 2.3732342007434943,
      "grad_norm": 1.1510920524597168,
      "learning_rate": 0.0002,
      "loss": 0.6167,
      "step": 399
    },
    {
      "epoch": 2.379182156133829,
      "grad_norm": 1.0992542505264282,
      "learning_rate": 0.0002,
      "loss": 0.4328,
      "step": 400
    },
    {
      "epoch": 2.3851301115241634,
      "grad_norm": 1.3745630979537964,
      "learning_rate": 0.0002,
      "loss": 0.6485,
      "step": 401
    },
    {
      "epoch": 2.391078066914498,
      "grad_norm": 1.2044932842254639,
      "learning_rate": 0.0002,
      "loss": 0.6345,
      "step": 402
    },
    {
      "epoch": 2.3970260223048325,
      "grad_norm": 1.4290401935577393,
      "learning_rate": 0.0002,
      "loss": 0.5706,
      "step": 403
    },
    {
      "epoch": 2.4029739776951673,
      "grad_norm": 1.3580994606018066,
      "learning_rate": 0.0002,
      "loss": 0.5373,
      "step": 404
    },
    {
      "epoch": 2.4089219330855016,
      "grad_norm": 1.0747463703155518,
      "learning_rate": 0.0002,
      "loss": 0.5597,
      "step": 405
    },
    {
      "epoch": 2.4148698884758364,
      "grad_norm": 1.288228154182434,
      "learning_rate": 0.0002,
      "loss": 0.6366,
      "step": 406
    },
    {
      "epoch": 2.420817843866171,
      "grad_norm": 1.2379798889160156,
      "learning_rate": 0.0002,
      "loss": 0.474,
      "step": 407
    },
    {
      "epoch": 2.4267657992565055,
      "grad_norm": 1.234220266342163,
      "learning_rate": 0.0002,
      "loss": 0.5363,
      "step": 408
    },
    {
      "epoch": 2.4327137546468403,
      "grad_norm": 1.2338114976882935,
      "learning_rate": 0.0002,
      "loss": 0.4992,
      "step": 409
    },
    {
      "epoch": 2.4386617100371746,
      "grad_norm": 1.3846346139907837,
      "learning_rate": 0.0002,
      "loss": 0.6412,
      "step": 410
    },
    {
      "epoch": 2.4446096654275093,
      "grad_norm": 1.2423279285430908,
      "learning_rate": 0.0002,
      "loss": 0.528,
      "step": 411
    },
    {
      "epoch": 2.4505576208178437,
      "grad_norm": 1.235088586807251,
      "learning_rate": 0.0002,
      "loss": 0.5763,
      "step": 412
    },
    {
      "epoch": 2.4565055762081784,
      "grad_norm": 1.3832026720046997,
      "learning_rate": 0.0002,
      "loss": 0.6355,
      "step": 413
    },
    {
      "epoch": 2.4624535315985128,
      "grad_norm": 1.214076280593872,
      "learning_rate": 0.0002,
      "loss": 0.6345,
      "step": 414
    },
    {
      "epoch": 2.4684014869888475,
      "grad_norm": 1.463728904724121,
      "learning_rate": 0.0002,
      "loss": 0.6186,
      "step": 415
    },
    {
      "epoch": 2.4743494423791823,
      "grad_norm": 1.0485203266143799,
      "learning_rate": 0.0002,
      "loss": 0.4723,
      "step": 416
    },
    {
      "epoch": 2.4802973977695166,
      "grad_norm": 1.5590802431106567,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 417
    },
    {
      "epoch": 2.4862453531598514,
      "grad_norm": 1.1843955516815186,
      "learning_rate": 0.0002,
      "loss": 0.5043,
      "step": 418
    },
    {
      "epoch": 2.4921933085501857,
      "grad_norm": 1.579487919807434,
      "learning_rate": 0.0002,
      "loss": 0.5257,
      "step": 419
    },
    {
      "epoch": 2.4981412639405205,
      "grad_norm": 1.2086743116378784,
      "learning_rate": 0.0002,
      "loss": 0.6798,
      "step": 420
    },
    {
      "epoch": 2.5040892193308553,
      "grad_norm": 1.195058822631836,
      "learning_rate": 0.0002,
      "loss": 0.6485,
      "step": 421
    },
    {
      "epoch": 2.5100371747211896,
      "grad_norm": 1.283530354499817,
      "learning_rate": 0.0002,
      "loss": 0.5873,
      "step": 422
    },
    {
      "epoch": 2.515985130111524,
      "grad_norm": 1.2394181489944458,
      "learning_rate": 0.0002,
      "loss": 0.6391,
      "step": 423
    },
    {
      "epoch": 2.5219330855018587,
      "grad_norm": 1.2939765453338623,
      "learning_rate": 0.0002,
      "loss": 0.6709,
      "step": 424
    },
    {
      "epoch": 2.5278810408921935,
      "grad_norm": 1.1421490907669067,
      "learning_rate": 0.0002,
      "loss": 0.563,
      "step": 425
    },
    {
      "epoch": 2.533828996282528,
      "grad_norm": 1.1861711740493774,
      "learning_rate": 0.0002,
      "loss": 0.5176,
      "step": 426
    },
    {
      "epoch": 2.5397769516728625,
      "grad_norm": 1.3675614595413208,
      "learning_rate": 0.0002,
      "loss": 0.529,
      "step": 427
    },
    {
      "epoch": 2.545724907063197,
      "grad_norm": 1.5121257305145264,
      "learning_rate": 0.0002,
      "loss": 0.7019,
      "step": 428
    },
    {
      "epoch": 2.5516728624535316,
      "grad_norm": 1.169859766960144,
      "learning_rate": 0.0002,
      "loss": 0.5985,
      "step": 429
    },
    {
      "epoch": 2.5576208178438664,
      "grad_norm": 1.3540085554122925,
      "learning_rate": 0.0002,
      "loss": 0.5499,
      "step": 430
    },
    {
      "epoch": 2.5635687732342007,
      "grad_norm": 1.0933575630187988,
      "learning_rate": 0.0002,
      "loss": 0.5208,
      "step": 431
    },
    {
      "epoch": 2.569516728624535,
      "grad_norm": 1.1712511777877808,
      "learning_rate": 0.0002,
      "loss": 0.536,
      "step": 432
    },
    {
      "epoch": 2.57546468401487,
      "grad_norm": 1.4346905946731567,
      "learning_rate": 0.0002,
      "loss": 0.6115,
      "step": 433
    },
    {
      "epoch": 2.5814126394052046,
      "grad_norm": 2.1174967288970947,
      "learning_rate": 0.0002,
      "loss": 0.5341,
      "step": 434
    },
    {
      "epoch": 2.587360594795539,
      "grad_norm": 1.3911654949188232,
      "learning_rate": 0.0002,
      "loss": 0.5707,
      "step": 435
    },
    {
      "epoch": 2.5933085501858737,
      "grad_norm": 1.3274894952774048,
      "learning_rate": 0.0002,
      "loss": 0.6272,
      "step": 436
    },
    {
      "epoch": 2.599256505576208,
      "grad_norm": 0.9820629358291626,
      "learning_rate": 0.0002,
      "loss": 0.4664,
      "step": 437
    },
    {
      "epoch": 2.605204460966543,
      "grad_norm": 1.1450122594833374,
      "learning_rate": 0.0002,
      "loss": 0.4652,
      "step": 438
    },
    {
      "epoch": 2.6111524163568776,
      "grad_norm": 2.7575511932373047,
      "learning_rate": 0.0002,
      "loss": 0.5652,
      "step": 439
    },
    {
      "epoch": 2.617100371747212,
      "grad_norm": 1.4760148525238037,
      "learning_rate": 0.0002,
      "loss": 0.6331,
      "step": 440
    },
    {
      "epoch": 2.623048327137546,
      "grad_norm": 1.2463843822479248,
      "learning_rate": 0.0002,
      "loss": 0.5721,
      "step": 441
    },
    {
      "epoch": 2.628996282527881,
      "grad_norm": 1.2081701755523682,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 442
    },
    {
      "epoch": 2.6349442379182157,
      "grad_norm": 1.1693692207336426,
      "learning_rate": 0.0002,
      "loss": 0.5505,
      "step": 443
    },
    {
      "epoch": 2.64089219330855,
      "grad_norm": 1.3918544054031372,
      "learning_rate": 0.0002,
      "loss": 0.5387,
      "step": 444
    },
    {
      "epoch": 2.646840148698885,
      "grad_norm": 1.3081449270248413,
      "learning_rate": 0.0002,
      "loss": 0.4156,
      "step": 445
    },
    {
      "epoch": 2.652788104089219,
      "grad_norm": 1.1178191900253296,
      "learning_rate": 0.0002,
      "loss": 0.6099,
      "step": 446
    },
    {
      "epoch": 2.658736059479554,
      "grad_norm": 1.172034740447998,
      "learning_rate": 0.0002,
      "loss": 0.6112,
      "step": 447
    },
    {
      "epoch": 2.6646840148698887,
      "grad_norm": 1.3142459392547607,
      "learning_rate": 0.0002,
      "loss": 0.6123,
      "step": 448
    },
    {
      "epoch": 2.670631970260223,
      "grad_norm": 1.2363723516464233,
      "learning_rate": 0.0002,
      "loss": 0.5472,
      "step": 449
    },
    {
      "epoch": 2.6765799256505574,
      "grad_norm": 1.2881202697753906,
      "learning_rate": 0.0002,
      "loss": 0.4572,
      "step": 450
    },
    {
      "epoch": 2.682527881040892,
      "grad_norm": 1.0761253833770752,
      "learning_rate": 0.0002,
      "loss": 0.5366,
      "step": 451
    },
    {
      "epoch": 2.688475836431227,
      "grad_norm": 1.0405654907226562,
      "learning_rate": 0.0002,
      "loss": 0.6858,
      "step": 452
    },
    {
      "epoch": 2.694423791821561,
      "grad_norm": 1.3384194374084473,
      "learning_rate": 0.0002,
      "loss": 0.616,
      "step": 453
    },
    {
      "epoch": 2.700371747211896,
      "grad_norm": 8.933956146240234,
      "learning_rate": 0.0002,
      "loss": 0.6136,
      "step": 454
    },
    {
      "epoch": 2.7063197026022303,
      "grad_norm": 1.1435190439224243,
      "learning_rate": 0.0002,
      "loss": 0.5598,
      "step": 455
    },
    {
      "epoch": 2.712267657992565,
      "grad_norm": 1.2891956567764282,
      "learning_rate": 0.0002,
      "loss": 0.5945,
      "step": 456
    },
    {
      "epoch": 2.7182156133829,
      "grad_norm": 1.3077706098556519,
      "learning_rate": 0.0002,
      "loss": 0.5793,
      "step": 457
    },
    {
      "epoch": 2.724163568773234,
      "grad_norm": 1.1445353031158447,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 458
    },
    {
      "epoch": 2.7301115241635685,
      "grad_norm": 1.1466567516326904,
      "learning_rate": 0.0002,
      "loss": 0.5748,
      "step": 459
    },
    {
      "epoch": 2.7360594795539033,
      "grad_norm": 1.0083645582199097,
      "learning_rate": 0.0002,
      "loss": 0.5676,
      "step": 460
    },
    {
      "epoch": 2.742007434944238,
      "grad_norm": 0.9980899691581726,
      "learning_rate": 0.0002,
      "loss": 0.5975,
      "step": 461
    },
    {
      "epoch": 2.7479553903345724,
      "grad_norm": 1.0702303647994995,
      "learning_rate": 0.0002,
      "loss": 0.5235,
      "step": 462
    },
    {
      "epoch": 2.753903345724907,
      "grad_norm": 1.3305853605270386,
      "learning_rate": 0.0002,
      "loss": 0.6493,
      "step": 463
    },
    {
      "epoch": 2.7598513011152415,
      "grad_norm": 1.4583408832550049,
      "learning_rate": 0.0002,
      "loss": 0.5948,
      "step": 464
    },
    {
      "epoch": 2.7657992565055762,
      "grad_norm": 1.1704531908035278,
      "learning_rate": 0.0002,
      "loss": 0.5631,
      "step": 465
    },
    {
      "epoch": 2.771747211895911,
      "grad_norm": 1.1165651082992554,
      "learning_rate": 0.0002,
      "loss": 0.6674,
      "step": 466
    },
    {
      "epoch": 2.7776951672862453,
      "grad_norm": 1.2043639421463013,
      "learning_rate": 0.0002,
      "loss": 0.649,
      "step": 467
    },
    {
      "epoch": 2.7836431226765797,
      "grad_norm": 1.0930832624435425,
      "learning_rate": 0.0002,
      "loss": 0.541,
      "step": 468
    },
    {
      "epoch": 2.7895910780669144,
      "grad_norm": 1.1153466701507568,
      "learning_rate": 0.0002,
      "loss": 0.4666,
      "step": 469
    },
    {
      "epoch": 2.795539033457249,
      "grad_norm": 3.27708101272583,
      "learning_rate": 0.0002,
      "loss": 0.6981,
      "step": 470
    },
    {
      "epoch": 2.8014869888475835,
      "grad_norm": 1.200003981590271,
      "learning_rate": 0.0002,
      "loss": 0.5579,
      "step": 471
    },
    {
      "epoch": 2.8074349442379183,
      "grad_norm": 1.2021151781082153,
      "learning_rate": 0.0002,
      "loss": 0.5383,
      "step": 472
    },
    {
      "epoch": 2.8133828996282526,
      "grad_norm": 1.0844088792800903,
      "learning_rate": 0.0002,
      "loss": 0.6027,
      "step": 473
    },
    {
      "epoch": 2.8193308550185874,
      "grad_norm": 1.1981035470962524,
      "learning_rate": 0.0002,
      "loss": 0.5589,
      "step": 474
    },
    {
      "epoch": 2.825278810408922,
      "grad_norm": 1.23332679271698,
      "learning_rate": 0.0002,
      "loss": 0.6733,
      "step": 475
    },
    {
      "epoch": 2.8312267657992565,
      "grad_norm": 1.2242364883422852,
      "learning_rate": 0.0002,
      "loss": 0.6392,
      "step": 476
    },
    {
      "epoch": 2.8371747211895912,
      "grad_norm": 1.2482764720916748,
      "learning_rate": 0.0002,
      "loss": 0.6185,
      "step": 477
    },
    {
      "epoch": 2.8431226765799256,
      "grad_norm": 1.3755487203598022,
      "learning_rate": 0.0002,
      "loss": 0.5761,
      "step": 478
    },
    {
      "epoch": 2.8490706319702603,
      "grad_norm": 1.2065231800079346,
      "learning_rate": 0.0002,
      "loss": 0.5185,
      "step": 479
    },
    {
      "epoch": 2.8550185873605947,
      "grad_norm": 1.1161603927612305,
      "learning_rate": 0.0002,
      "loss": 0.5697,
      "step": 480
    },
    {
      "epoch": 2.8609665427509294,
      "grad_norm": 1.7466390132904053,
      "learning_rate": 0.0002,
      "loss": 0.5835,
      "step": 481
    },
    {
      "epoch": 2.8669144981412638,
      "grad_norm": 1.371319055557251,
      "learning_rate": 0.0002,
      "loss": 0.6031,
      "step": 482
    },
    {
      "epoch": 2.8728624535315985,
      "grad_norm": 1.4363592863082886,
      "learning_rate": 0.0002,
      "loss": 0.6028,
      "step": 483
    },
    {
      "epoch": 2.8788104089219333,
      "grad_norm": 1.1025314331054688,
      "learning_rate": 0.0002,
      "loss": 0.492,
      "step": 484
    },
    {
      "epoch": 2.8847583643122676,
      "grad_norm": 1.04302978515625,
      "learning_rate": 0.0002,
      "loss": 0.6032,
      "step": 485
    },
    {
      "epoch": 2.8907063197026024,
      "grad_norm": 1.5093481540679932,
      "learning_rate": 0.0002,
      "loss": 0.6832,
      "step": 486
    },
    {
      "epoch": 2.8966542750929367,
      "grad_norm": 1.068484902381897,
      "learning_rate": 0.0002,
      "loss": 0.4942,
      "step": 487
    },
    {
      "epoch": 2.9026022304832715,
      "grad_norm": 2.1092681884765625,
      "learning_rate": 0.0002,
      "loss": 0.4909,
      "step": 488
    },
    {
      "epoch": 2.908550185873606,
      "grad_norm": 1.22842276096344,
      "learning_rate": 0.0002,
      "loss": 0.6226,
      "step": 489
    },
    {
      "epoch": 2.9144981412639406,
      "grad_norm": 1.1664717197418213,
      "learning_rate": 0.0002,
      "loss": 0.6093,
      "step": 490
    },
    {
      "epoch": 2.920446096654275,
      "grad_norm": 1.2886883020401,
      "learning_rate": 0.0002,
      "loss": 0.5866,
      "step": 491
    },
    {
      "epoch": 2.9263940520446097,
      "grad_norm": 1.1186504364013672,
      "learning_rate": 0.0002,
      "loss": 0.5942,
      "step": 492
    },
    {
      "epoch": 2.9323420074349444,
      "grad_norm": 1.2734028100967407,
      "learning_rate": 0.0002,
      "loss": 0.5871,
      "step": 493
    },
    {
      "epoch": 2.9382899628252788,
      "grad_norm": 1.1976778507232666,
      "learning_rate": 0.0002,
      "loss": 0.6766,
      "step": 494
    },
    {
      "epoch": 2.9442379182156135,
      "grad_norm": 1.524681806564331,
      "learning_rate": 0.0002,
      "loss": 0.602,
      "step": 495
    },
    {
      "epoch": 2.950185873605948,
      "grad_norm": 1.4174754619598389,
      "learning_rate": 0.0002,
      "loss": 0.6079,
      "step": 496
    },
    {
      "epoch": 2.9561338289962826,
      "grad_norm": 1.1006587743759155,
      "learning_rate": 0.0002,
      "loss": 0.6393,
      "step": 497
    },
    {
      "epoch": 2.962081784386617,
      "grad_norm": 1.3037843704223633,
      "learning_rate": 0.0002,
      "loss": 0.6065,
      "step": 498
    },
    {
      "epoch": 2.9680297397769517,
      "grad_norm": 1.5767035484313965,
      "learning_rate": 0.0002,
      "loss": 0.6444,
      "step": 499
    },
    {
      "epoch": 2.973977695167286,
      "grad_norm": 1.2918823957443237,
      "learning_rate": 0.0002,
      "loss": 0.5221,
      "step": 500
    },
    {
      "epoch": 2.979925650557621,
      "grad_norm": 1.0898538827896118,
      "learning_rate": 0.0002,
      "loss": 0.687,
      "step": 501
    },
    {
      "epoch": 2.9858736059479556,
      "grad_norm": 1.6398361921310425,
      "learning_rate": 0.0002,
      "loss": 0.5781,
      "step": 502
    },
    {
      "epoch": 2.99182156133829,
      "grad_norm": 1.2504217624664307,
      "learning_rate": 0.0002,
      "loss": 0.5916,
      "step": 503
    },
    {
      "epoch": 2.9977695167286247,
      "grad_norm": 1.8298507928848267,
      "learning_rate": 0.0002,
      "loss": 0.5777,
      "step": 504
    }
  ],
  "logging_steps": 1,
  "max_steps": 504,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 5,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.641281849472852e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}