{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.710144927536232,
  "eval_steps": 500,
  "global_step": 1024,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 46.72853930172847,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.6225,
      "step": 1
    },
    {
      "epoch": 0.01,
      "grad_norm": 52.07096639492344,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.649,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 53.48678929212073,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.6198,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 40.93037505319626,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.5608,
      "step": 4
    },
    {
      "epoch": 0.02,
      "grad_norm": 52.14947742859088,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.5664,
      "step": 5
    },
    {
      "epoch": 0.02,
      "grad_norm": 39.68797135248386,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.5766,
      "step": 6
    },
    {
      "epoch": 0.03,
      "grad_norm": 27.608786716750426,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.3961,
      "step": 7
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.843193562709772,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.2716,
      "step": 8
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.204064082882707,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.2625,
      "step": 9
    },
    {
      "epoch": 0.04,
      "grad_norm": 7.24061080880167,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.2859,
      "step": 10
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.127471947114424,
      "learning_rate": 4.4e-06,
      "loss": 0.2321,
      "step": 11
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.14969705214205,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.223,
      "step": 12
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.332899725870617,
      "learning_rate": 5.2e-06,
      "loss": 0.25,
      "step": 13
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.231171781798085,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.2154,
      "step": 14
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.5854830947431044,
      "learning_rate": 6e-06,
      "loss": 0.2381,
      "step": 15
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.840138857999184,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.1981,
      "step": 16
    },
    {
      "epoch": 0.06,
      "grad_norm": 8.377814173650734,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.2692,
      "step": 17
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.205565266596127,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.2197,
      "step": 18
    },
    {
      "epoch": 0.07,
      "grad_norm": 7.0540544766233415,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.2349,
      "step": 19
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.65295971356695,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.3208,
      "step": 20
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.52181319610902,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.2307,
      "step": 21
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.482819938202367,
      "learning_rate": 8.8e-06,
      "loss": 0.2024,
      "step": 22
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.609864061676336,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.2463,
      "step": 23
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.245726835362501,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.2328,
      "step": 24
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.138546706803432,
      "learning_rate": 1e-05,
      "loss": 0.2496,
      "step": 25
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.781197388695438,
      "learning_rate": 1.04e-05,
      "loss": 0.2714,
      "step": 26
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.108329488753126,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 0.2507,
      "step": 27
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.800437254625411,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.2105,
      "step": 28
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.649524899716385,
      "learning_rate": 1.16e-05,
      "loss": 0.2352,
      "step": 29
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.3212480454181823,
      "learning_rate": 1.2e-05,
      "loss": 0.2121,
      "step": 30
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.8259124298359497,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 0.2102,
      "step": 31
    },
    {
      "epoch": 0.12,
      "grad_norm": 8.370375637151021,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.2299,
      "step": 32
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.8976045378474216,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 0.1925,
      "step": 33
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.546522390784352,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.221,
      "step": 34
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.0450418382891864,
      "learning_rate": 1.4e-05,
      "loss": 0.2493,
      "step": 35
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.689732934577616,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.2286,
      "step": 36
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.0181283532549883,
      "learning_rate": 1.48e-05,
      "loss": 0.2058,
      "step": 37
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.0290096617307323,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.2133,
      "step": 38
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.3763932465021425,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 0.2596,
      "step": 39
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4603343566227587,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.2122,
      "step": 40
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.332369451595958,
      "learning_rate": 1.64e-05,
      "loss": 0.2208,
      "step": 41
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.7861266281997907,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.2456,
      "step": 42
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.7225805400820935,
      "learning_rate": 1.72e-05,
      "loss": 0.2465,
      "step": 43
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.9056402543182682,
      "learning_rate": 1.76e-05,
      "loss": 0.22,
      "step": 44
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8766590187035856,
      "learning_rate": 1.8e-05,
      "loss": 0.1729,
      "step": 45
    },
    {
      "epoch": 0.17,
      "grad_norm": 7.788754664844443,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.2463,
      "step": 46
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.44553452940045,
      "learning_rate": 1.88e-05,
      "loss": 0.1861,
      "step": 47
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.0748717088951483,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.2392,
      "step": 48
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.1429857670987684,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.196,
      "step": 49
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.6694741818907692,
      "learning_rate": 2e-05,
      "loss": 0.1938,
      "step": 50
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.075853192374163,
      "learning_rate": 1.9999980867200107e-05,
      "loss": 0.2176,
      "step": 51
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.5881096938435726,
      "learning_rate": 1.9999923468873635e-05,
      "loss": 0.2713,
      "step": 52
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.041830913696081,
      "learning_rate": 1.9999827805240225e-05,
      "loss": 0.2162,
      "step": 53
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.5182608103998274,
      "learning_rate": 1.999969387666594e-05,
      "loss": 0.1877,
      "step": 54
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.300761589056313,
      "learning_rate": 1.9999521683663265e-05,
      "loss": 0.2189,
      "step": 55
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.8419792972213038,
      "learning_rate": 1.9999311226891104e-05,
      "loss": 0.2094,
      "step": 56
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.219928784202957,
      "learning_rate": 1.9999062507154786e-05,
      "loss": 0.2236,
      "step": 57
    },
    {
      "epoch": 0.21,
      "grad_norm": 4.091917760353657,
      "learning_rate": 1.999877552540605e-05,
      "loss": 0.1977,
      "step": 58
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.3233992848215195,
      "learning_rate": 1.9998450282743055e-05,
      "loss": 0.2333,
      "step": 59
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.964219225360724,
      "learning_rate": 1.9998086780410353e-05,
      "loss": 0.2208,
      "step": 60
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.219456222024613,
      "learning_rate": 1.9997685019798913e-05,
      "loss": 0.2002,
      "step": 61
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.084361648388864,
      "learning_rate": 1.999724500244609e-05,
      "loss": 0.2443,
      "step": 62
    },
    {
      "epoch": 0.23,
      "grad_norm": 5.875576520826541,
      "learning_rate": 1.9996766730035642e-05,
      "loss": 0.2263,
      "step": 63
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.803152179024348,
      "learning_rate": 1.999625020439771e-05,
      "loss": 0.1829,
      "step": 64
    },
    {
      "epoch": 0.24,
      "grad_norm": 5.487803901369769,
      "learning_rate": 1.99956954275088e-05,
      "loss": 0.247,
      "step": 65
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.330507485677968,
      "learning_rate": 1.999510240149181e-05,
      "loss": 0.1726,
      "step": 66
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.750879407303498,
      "learning_rate": 1.9994471128615984e-05,
      "loss": 0.2313,
      "step": 67
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.390482378065639,
      "learning_rate": 1.9993801611296923e-05,
      "loss": 0.193,
      "step": 68
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.192715936136871,
      "learning_rate": 1.9993093852096582e-05,
      "loss": 0.1794,
      "step": 69
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.367128095928201,
      "learning_rate": 1.999234785372324e-05,
      "loss": 0.2131,
      "step": 70
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.050893259478687,
      "learning_rate": 1.9991563619031508e-05,
      "loss": 0.1899,
      "step": 71
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.120204599576964,
      "learning_rate": 1.9990741151022302e-05,
      "loss": 0.2087,
      "step": 72
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.29023535186417,
      "learning_rate": 1.9989880452842847e-05,
      "loss": 0.1858,
      "step": 73
    },
    {
      "epoch": 0.27,
      "grad_norm": 4.564103746693168,
      "learning_rate": 1.9988981527786656e-05,
      "loss": 0.1837,
      "step": 74
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.468247385963039,
      "learning_rate": 1.9988044379293523e-05,
      "loss": 0.2097,
      "step": 75
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.339814904044548,
      "learning_rate": 1.99870690109495e-05,
      "loss": 0.1997,
      "step": 76
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.753844502932187,
      "learning_rate": 1.9986055426486888e-05,
      "loss": 0.199,
      "step": 77
    },
    {
      "epoch": 0.28,
      "grad_norm": 8.614473813172529,
      "learning_rate": 1.9985003629784237e-05,
      "loss": 0.2653,
      "step": 78
    },
    {
      "epoch": 0.29,
      "grad_norm": 6.272374952434112,
      "learning_rate": 1.9983913624866304e-05,
      "loss": 0.1559,
      "step": 79
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.4341615932729357,
      "learning_rate": 1.9982785415904063e-05,
      "loss": 0.1937,
      "step": 80
    },
    {
      "epoch": 0.29,
      "grad_norm": 5.231118772027732,
      "learning_rate": 1.9981619007214675e-05,
      "loss": 0.2214,
      "step": 81
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.460575468938803,
      "learning_rate": 1.998041440326146e-05,
      "loss": 0.2901,
      "step": 82
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.3535481254926514,
      "learning_rate": 1.9979171608653926e-05,
      "loss": 0.1895,
      "step": 83
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.095485056227556,
      "learning_rate": 1.9977890628147684e-05,
      "loss": 0.2148,
      "step": 84
    },
    {
      "epoch": 0.31,
      "grad_norm": 11.074389998820713,
      "learning_rate": 1.9976571466644493e-05,
      "loss": 0.2083,
      "step": 85
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.453386007839937,
      "learning_rate": 1.99752141291922e-05,
      "loss": 0.2142,
      "step": 86
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.1609452514142085,
      "learning_rate": 1.9973818620984737e-05,
      "loss": 0.1934,
      "step": 87
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.1487715446979685,
      "learning_rate": 1.99723849473621e-05,
      "loss": 0.2345,
      "step": 88
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.2634858893508945,
      "learning_rate": 1.9970913113810336e-05,
      "loss": 0.1982,
      "step": 89
    },
    {
      "epoch": 0.33,
      "grad_norm": 6.523213533930424,
      "learning_rate": 1.996940312596149e-05,
      "loss": 0.2248,
      "step": 90
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.773562252049304,
      "learning_rate": 1.9967854989593634e-05,
      "loss": 0.213,
      "step": 91
    },
    {
      "epoch": 0.33,
      "grad_norm": 7.830843979970007,
      "learning_rate": 1.9966268710630795e-05,
      "loss": 0.2478,
      "step": 92
    },
    {
      "epoch": 0.34,
      "grad_norm": 3.636412312338061,
      "learning_rate": 1.996464429514297e-05,
      "loss": 0.2259,
      "step": 93
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.644965516136683,
      "learning_rate": 1.996298174934608e-05,
      "loss": 0.1852,
      "step": 94
    },
    {
      "epoch": 0.34,
      "grad_norm": 8.186163466631733,
      "learning_rate": 1.9961281079601957e-05,
      "loss": 0.2509,
      "step": 95
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.540579065302634,
      "learning_rate": 1.9959542292418317e-05,
      "loss": 0.199,
      "step": 96
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.8497364574995268,
      "learning_rate": 1.995776539444873e-05,
      "loss": 0.2237,
      "step": 97
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.3100586831605106,
      "learning_rate": 1.9955950392492604e-05,
      "loss": 0.2275,
      "step": 98
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.974566132033935,
      "learning_rate": 1.9954097293495155e-05,
      "loss": 0.2426,
      "step": 99
    },
    {
      "epoch": 0.36,
      "grad_norm": 4.945067016519828,
      "learning_rate": 1.9952206104547378e-05,
      "loss": 0.212,
      "step": 100
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.347489899552784,
      "learning_rate": 1.995027683288602e-05,
      "loss": 0.1971,
      "step": 101
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.1269065904933235,
      "learning_rate": 1.994830948589355e-05,
      "loss": 0.2092,
      "step": 102
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.4384730018546996,
      "learning_rate": 1.9946304071098143e-05,
      "loss": 0.203,
      "step": 103
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3363092433164303,
      "learning_rate": 1.9944260596173642e-05,
      "loss": 0.2092,
      "step": 104
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.4273522949310387,
      "learning_rate": 1.994217906893952e-05,
      "loss": 0.1938,
      "step": 105
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.938357565849595,
      "learning_rate": 1.9940059497360874e-05,
      "loss": 0.1935,
      "step": 106
    },
    {
      "epoch": 0.39,
      "grad_norm": 5.912638673049391,
      "learning_rate": 1.993790188954836e-05,
      "loss": 0.2313,
      "step": 107
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4222878204576312,
      "learning_rate": 1.9935706253758206e-05,
      "loss": 0.2192,
      "step": 108
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4706157047642812,
      "learning_rate": 1.993347259839214e-05,
      "loss": 0.2038,
      "step": 109
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.552023589176243,
      "learning_rate": 1.9931200931997372e-05,
      "loss": 0.1944,
      "step": 110
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.6398901043384204,
      "learning_rate": 1.9928891263266578e-05,
      "loss": 0.1801,
      "step": 111
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.5704568250317315,
      "learning_rate": 1.9926543601037843e-05,
      "loss": 0.2268,
      "step": 112
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.1956913784627607,
      "learning_rate": 1.992415795429463e-05,
      "loss": 0.2252,
      "step": 113
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.5341556196077897,
      "learning_rate": 1.992173433216577e-05,
      "loss": 0.1716,
      "step": 114
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.9626316933204295,
      "learning_rate": 1.9919272743925386e-05,
      "loss": 0.187,
      "step": 115
    },
    {
      "epoch": 0.42,
      "grad_norm": 3.7049021117634253,
      "learning_rate": 1.99167731989929e-05,
      "loss": 0.188,
      "step": 116
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.613297684884965,
      "learning_rate": 1.9914235706932975e-05,
      "loss": 0.1915,
      "step": 117
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.538398454776101,
      "learning_rate": 1.9911660277455473e-05,
      "loss": 0.1815,
      "step": 118
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.6517117154934629,
      "learning_rate": 1.990904692041542e-05,
      "loss": 0.2048,
      "step": 119
    },
    {
      "epoch": 0.43,
      "grad_norm": 5.519157487836765,
      "learning_rate": 1.9906395645813e-05,
      "loss": 0.2435,
      "step": 120
    },
    {
      "epoch": 0.44,
      "grad_norm": 3.551984088038686,
      "learning_rate": 1.9903706463793463e-05,
      "loss": 0.2155,
      "step": 121
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.2040044547253967,
      "learning_rate": 1.990097938464713e-05,
      "loss": 0.2238,
      "step": 122
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.410778671062537,
      "learning_rate": 1.989821441880933e-05,
      "loss": 0.1958,
      "step": 123
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.19869128373236,
      "learning_rate": 1.989541157686037e-05,
      "loss": 0.2028,
      "step": 124
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.536087636301242,
      "learning_rate": 1.9892570869525495e-05,
      "loss": 0.2,
      "step": 125
    },
    {
      "epoch": 0.46,
      "grad_norm": 3.2379922220159085,
      "learning_rate": 1.9889692307674847e-05,
      "loss": 0.1823,
      "step": 126
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.7508465935428106,
      "learning_rate": 1.9886775902323405e-05,
      "loss": 0.2064,
      "step": 127
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.6835865398958427,
      "learning_rate": 1.9883821664630977e-05,
      "loss": 0.2229,
      "step": 128
    },
    {
      "epoch": 0.47,
      "grad_norm": 3.0100393765483524,
      "learning_rate": 1.988082960590213e-05,
      "loss": 0.1546,
      "step": 129
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.3477760338260234,
      "learning_rate": 1.987779973758615e-05,
      "loss": 0.1789,
      "step": 130
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.458849610334373,
      "learning_rate": 1.9874732071277015e-05,
      "loss": 0.2424,
      "step": 131
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3606318777601056,
      "learning_rate": 1.987162661871333e-05,
      "loss": 0.1741,
      "step": 132
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.909771599382989,
      "learning_rate": 1.9868483391778303e-05,
      "loss": 0.1713,
      "step": 133
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.139911494220399,
      "learning_rate": 1.986530240249968e-05,
      "loss": 0.1794,
      "step": 134
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.9818599307655007,
      "learning_rate": 1.9862083663049698e-05,
      "loss": 0.2265,
      "step": 135
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.8642280226326846,
      "learning_rate": 1.985882718574506e-05,
      "loss": 0.2317,
      "step": 136
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4019059864946621,
      "learning_rate": 1.9855532983046876e-05,
      "loss": 0.1804,
      "step": 137
    },
    {
      "epoch": 0.5,
      "grad_norm": 3.937781886771039,
      "learning_rate": 1.9852201067560607e-05,
      "loss": 0.1904,
      "step": 138
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.3537933164261284,
      "learning_rate": 1.984883145203603e-05,
      "loss": 0.1719,
      "step": 139
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.953870998813679,
      "learning_rate": 1.984542414936718e-05,
      "loss": 0.2321,
      "step": 140
    },
    {
      "epoch": 0.51,
      "grad_norm": 3.778644025099447,
      "learning_rate": 1.98419791725923e-05,
      "loss": 0.2287,
      "step": 141
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.0978810240658883,
      "learning_rate": 1.9838496534893807e-05,
      "loss": 0.1946,
      "step": 142
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.4833931529998075,
      "learning_rate": 1.983497624959822e-05,
      "loss": 0.2447,
      "step": 143
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9962594108181742,
      "learning_rate": 1.9831418330176127e-05,
      "loss": 0.1613,
      "step": 144
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.1099326800106075,
      "learning_rate": 1.982782279024211e-05,
      "loss": 0.1758,
      "step": 145
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9917248634127067,
      "learning_rate": 1.9824189643554724e-05,
      "loss": 0.1832,
      "step": 146
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.9093376030904632,
      "learning_rate": 1.9820518904016425e-05,
      "loss": 0.2125,
      "step": 147
    },
    {
      "epoch": 0.54,
      "grad_norm": 3.8284205946412198,
      "learning_rate": 1.9816810585673515e-05,
      "loss": 0.2274,
      "step": 148
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.3817574265830486,
      "learning_rate": 1.9813064702716094e-05,
      "loss": 0.1811,
      "step": 149
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.411906280970247,
      "learning_rate": 1.9809281269478015e-05,
      "loss": 0.1909,
      "step": 150
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.074349965257892,
      "learning_rate": 1.9805460300436805e-05,
      "loss": 0.2112,
      "step": 151
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.3421132125177087,
      "learning_rate": 1.9801601810213634e-05,
      "loss": 0.2584,
      "step": 152
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4208905656674993,
      "learning_rate": 1.9797705813573247e-05,
      "loss": 0.1921,
      "step": 153
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.1939035088667658,
      "learning_rate": 1.979377232542391e-05,
      "loss": 0.1541,
      "step": 154
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.2384977607692,
      "learning_rate": 1.9789801360817348e-05,
      "loss": 0.1978,
      "step": 155
    },
    {
      "epoch": 0.57,
      "grad_norm": 5.4439444275988365,
      "learning_rate": 1.9785792934948697e-05,
      "loss": 0.2153,
      "step": 156
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.2451798232206324,
      "learning_rate": 1.9781747063156436e-05,
      "loss": 0.177,
      "step": 157
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.967272591808535,
      "learning_rate": 1.9777663760922342e-05,
      "loss": 0.192,
      "step": 158
    },
    {
      "epoch": 0.58,
      "grad_norm": 6.104903104955908,
      "learning_rate": 1.977354304387141e-05,
      "loss": 0.2066,
      "step": 159
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.8069115347552653,
      "learning_rate": 1.976938492777182e-05,
      "loss": 0.1982,
      "step": 160
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.7788359549191153,
      "learning_rate": 1.976518942853484e-05,
      "loss": 0.189,
      "step": 161
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.211094265354669,
      "learning_rate": 1.9760956562214808e-05,
      "loss": 0.1832,
      "step": 162
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5547979175056448,
      "learning_rate": 1.975668634500904e-05,
      "loss": 0.1643,
      "step": 163
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.3883926579039336,
      "learning_rate": 1.9752378793257777e-05,
      "loss": 0.2382,
      "step": 164
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7224510071144223,
      "learning_rate": 1.9748033923444123e-05,
      "loss": 0.1834,
      "step": 165
    },
    {
      "epoch": 0.6,
      "grad_norm": 4.561364149201268,
      "learning_rate": 1.9743651752193983e-05,
      "loss": 0.2058,
      "step": 166
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.3829318882111994,
      "learning_rate": 1.9739232296276004e-05,
      "loss": 0.2104,
      "step": 167
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.4674065600010064,
      "learning_rate": 1.9734775572601487e-05,
      "loss": 0.1665,
      "step": 168
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.8431949486041035,
      "learning_rate": 1.9730281598224366e-05,
      "loss": 0.1685,
      "step": 169
    },
    {
      "epoch": 0.62,
      "grad_norm": 3.7012831942032336,
      "learning_rate": 1.9725750390341093e-05,
      "loss": 0.2276,
      "step": 170
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.824165248665086,
      "learning_rate": 1.9721181966290614e-05,
      "loss": 0.1914,
      "step": 171
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.0423081016544904,
      "learning_rate": 1.9716576343554274e-05,
      "loss": 0.1748,
      "step": 172
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.879666187913567,
      "learning_rate": 1.9711933539755764e-05,
      "loss": 0.1746,
      "step": 173
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.6209642404840574,
      "learning_rate": 1.9707253572661057e-05,
      "loss": 0.1475,
      "step": 174
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.8343240393881313,
      "learning_rate": 1.9702536460178318e-05,
      "loss": 0.1903,
      "step": 175
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.513338761986029,
      "learning_rate": 1.969778222035787e-05,
      "loss": 0.1733,
      "step": 176
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.681664720505642,
      "learning_rate": 1.969299087139209e-05,
      "loss": 0.1774,
      "step": 177
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.371858975566763,
      "learning_rate": 1.9688162431615367e-05,
      "loss": 0.2187,
      "step": 178
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.1607291066632204,
      "learning_rate": 1.9683296919504013e-05,
      "loss": 0.2059,
      "step": 179
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.2653264740906485,
      "learning_rate": 1.9678394353676203e-05,
      "loss": 0.1524,
      "step": 180
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.3029091047043004,
      "learning_rate": 1.9673454752891898e-05,
      "loss": 0.159,
      "step": 181
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6003395131617928,
      "learning_rate": 1.9668478136052776e-05,
      "loss": 0.1665,
      "step": 182
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.366100268481903,
      "learning_rate": 1.9663464522202162e-05,
      "loss": 0.162,
      "step": 183
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.787342810380197,
      "learning_rate": 1.9658413930524955e-05,
      "loss": 0.1807,
      "step": 184
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.349030904122291,
      "learning_rate": 1.9653326380347532e-05,
      "loss": 0.2119,
      "step": 185
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.6889625324614177,
      "learning_rate": 1.9648201891137725e-05,
      "loss": 0.1599,
      "step": 186
    },
    {
      "epoch": 0.68,
      "grad_norm": 3.1263858368560586,
      "learning_rate": 1.964304048250469e-05,
      "loss": 0.1907,
      "step": 187
    },
    {
      "epoch": 0.68,
      "grad_norm": 5.597521679201683,
      "learning_rate": 1.963784217419887e-05,
      "loss": 0.1747,
      "step": 188
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.3137721457903386,
      "learning_rate": 1.96326069861119e-05,
      "loss": 0.2071,
      "step": 189
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.4519505366650374,
      "learning_rate": 1.9627334938276547e-05,
      "loss": 0.1674,
      "step": 190
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.1148683625073919,
      "learning_rate": 1.9622026050866613e-05,
      "loss": 0.1467,
      "step": 191
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.2690170842798527,
      "learning_rate": 1.961668034419688e-05,
      "loss": 0.2129,
      "step": 192
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.619889210397982,
      "learning_rate": 1.961129783872301e-05,
      "loss": 0.1622,
      "step": 193
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.9136904827036365,
      "learning_rate": 1.9605878555041484e-05,
      "loss": 0.2006,
      "step": 194
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.20031747350585,
      "learning_rate": 1.9600422513889515e-05,
      "loss": 0.1566,
      "step": 195
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.7166231249244475,
      "learning_rate": 1.9594929736144978e-05,
      "loss": 0.2185,
      "step": 196
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.795589774468367,
      "learning_rate": 1.9589400242826307e-05,
      "loss": 0.1703,
      "step": 197
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.5740329013856154,
      "learning_rate": 1.9583834055092446e-05,
      "loss": 0.2231,
      "step": 198
    },
    {
      "epoch": 0.72,
      "grad_norm": 3.5969245391305824,
      "learning_rate": 1.9578231194242744e-05,
      "loss": 0.183,
      "step": 199
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.0963714016675181,
      "learning_rate": 1.9572591681716888e-05,
      "loss": 0.175,
      "step": 200
    },
    {
      "epoch": 0.73,
      "grad_norm": 3.1527309757053024,
      "learning_rate": 1.9566915539094805e-05,
      "loss": 0.1612,
      "step": 201
    },
    {
      "epoch": 0.73,
      "grad_norm": 6.11802400882258,
      "learning_rate": 1.95612027880966e-05,
      "loss": 0.1739,
      "step": 202
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.1003663250353997,
      "learning_rate": 1.9555453450582453e-05,
      "loss": 0.24,
      "step": 203
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.1350893178704067,
      "learning_rate": 1.9549667548552557e-05,
      "loss": 0.2067,
      "step": 204
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.4276457497833874,
      "learning_rate": 1.9543845104147e-05,
      "loss": 0.1765,
      "step": 205
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.165213308146818,
      "learning_rate": 1.9537986139645724e-05,
      "loss": 0.223,
      "step": 206
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.441589230972257,
      "learning_rate": 1.9532090677468408e-05,
      "loss": 0.2,
      "step": 207
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.18549945843444,
      "learning_rate": 1.9526158740174392e-05,
      "loss": 0.1921,
      "step": 208
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.9825782386080286,
      "learning_rate": 1.9520190350462585e-05,
      "loss": 0.2034,
      "step": 209
    },
    {
      "epoch": 0.76,
      "grad_norm": 3.3463873892037093,
      "learning_rate": 1.951418553117139e-05,
      "loss": 0.1935,
      "step": 210
    },
    {
      "epoch": 0.76,
      "grad_norm": 5.385084120106734,
      "learning_rate": 1.950814430527861e-05,
      "loss": 0.2351,
      "step": 211
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.8919707500813345,
      "learning_rate": 1.950206669590136e-05,
      "loss": 0.1415,
      "step": 212
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.3002484179864218,
      "learning_rate": 1.949595272629597e-05,
      "loss": 0.2166,
      "step": 213
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.1625160748683374,
      "learning_rate": 1.9489802419857918e-05,
      "loss": 0.1766,
      "step": 214
    },
    {
      "epoch": 0.78,
      "grad_norm": 3.619701192742547,
      "learning_rate": 1.9483615800121717e-05,
      "loss": 0.1698,
      "step": 215
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.1829379252643495,
      "learning_rate": 1.947739289076084e-05,
      "loss": 0.142,
      "step": 216
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.766727320600325,
      "learning_rate": 1.947113371558762e-05,
      "loss": 0.1819,
      "step": 217
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.399502285949815,
      "learning_rate": 1.9464838298553172e-05,
      "loss": 0.2344,
      "step": 218
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.110766225428027,
      "learning_rate": 1.9458506663747286e-05,
      "loss": 0.2011,
      "step": 219
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.051592837631265,
      "learning_rate": 1.9452138835398333e-05,
      "loss": 0.1914,
      "step": 220
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.3725438569751955,
      "learning_rate": 1.9445734837873203e-05,
      "loss": 0.1765,
      "step": 221
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.1705247736939817,
      "learning_rate": 1.9439294695677168e-05,
      "loss": 0.1807,
      "step": 222
    },
    {
      "epoch": 0.81,
      "grad_norm": 3.3017606951651333,
      "learning_rate": 1.943281843345382e-05,
      "loss": 0.1905,
      "step": 223
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.4309126557975618,
      "learning_rate": 1.9426306075984968e-05,
      "loss": 0.1944,
      "step": 224
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.9415676897308591,
      "learning_rate": 1.9419757648190532e-05,
      "loss": 0.1838,
      "step": 225
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.3799577712400077,
      "learning_rate": 1.9413173175128472e-05,
      "loss": 0.1506,
      "step": 226
    },
    {
      "epoch": 0.82,
      "grad_norm": 3.270902216094506,
      "learning_rate": 1.9406552681994663e-05,
      "loss": 0.1633,
      "step": 227
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.684043789647969,
      "learning_rate": 1.9399896194122824e-05,
      "loss": 0.2286,
      "step": 228
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.708958534514703,
      "learning_rate": 1.93932037369844e-05,
      "loss": 0.1855,
      "step": 229
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7125886875419585,
      "learning_rate": 1.9386475336188484e-05,
      "loss": 0.1873,
      "step": 230
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8059962309471966,
      "learning_rate": 1.9379711017481703e-05,
      "loss": 0.1835,
      "step": 231
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.9096318404555614,
      "learning_rate": 1.9372910806748124e-05,
      "loss": 0.1928,
      "step": 232
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6880354115265441,
      "learning_rate": 1.936607473000917e-05,
      "loss": 0.2217,
      "step": 233
    },
    {
      "epoch": 0.85,
      "grad_norm": 4.590097931857785,
      "learning_rate": 1.935920281342349e-05,
      "loss": 0.1751,
      "step": 234
    },
    {
      "epoch": 0.85,
      "grad_norm": 3.3170870647428825,
      "learning_rate": 1.9352295083286897e-05,
      "loss": 0.1923,
      "step": 235
    },
    {
      "epoch": 0.86,
      "grad_norm": 3.399731385689797,
      "learning_rate": 1.934535156603222e-05,
      "loss": 0.1687,
      "step": 236
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.879018288953777,
      "learning_rate": 1.9338372288229253e-05,
      "loss": 0.1754,
      "step": 237
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7500404375244818,
      "learning_rate": 1.933135727658462e-05,
      "loss": 0.1724,
      "step": 238
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.4685486773525307,
      "learning_rate": 1.9324306557941684e-05,
      "loss": 0.2298,
      "step": 239
    },
    {
      "epoch": 0.87,
      "grad_norm": 3.93522263930692,
      "learning_rate": 1.931722015928044e-05,
      "loss": 0.1679,
      "step": 240
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.618587281781777,
      "learning_rate": 1.9310098107717418e-05,
      "loss": 0.2414,
      "step": 241
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.6014521447134005,
      "learning_rate": 1.930294043050558e-05,
      "loss": 0.2271,
      "step": 242
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7444357606743905,
      "learning_rate": 1.9295747155034203e-05,
      "loss": 0.1611,
      "step": 243
    },
    {
      "epoch": 0.88,
      "grad_norm": 3.90614869544953,
      "learning_rate": 1.928851830882879e-05,
      "loss": 0.1621,
      "step": 244
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.2866019796864734,
      "learning_rate": 1.928125391955095e-05,
      "loss": 0.2127,
      "step": 245
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.4492994823583496,
      "learning_rate": 1.9273954014998307e-05,
      "loss": 0.1873,
      "step": 246
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.4875080022090086,
      "learning_rate": 1.9266618623104384e-05,
      "loss": 0.1898,
      "step": 247
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.028262963276311,
      "learning_rate": 1.92592477719385e-05,
      "loss": 0.228,
      "step": 248
    },
    {
      "epoch": 0.9,
      "grad_norm": 3.682312273260727,
      "learning_rate": 1.9251841489705657e-05,
      "loss": 0.2226,
      "step": 249
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.1022001680261764,
      "learning_rate": 1.9244399804746436e-05,
      "loss": 0.1901,
      "step": 250
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.329486372814356,
      "learning_rate": 1.923692274553689e-05,
      "loss": 0.2105,
      "step": 251
    },
    {
      "epoch": 0.91,
      "grad_norm": 4.806818514585948,
      "learning_rate": 1.9229410340688442e-05,
      "loss": 0.2075,
      "step": 252
    },
    {
      "epoch": 0.92,
      "grad_norm": 4.495243683594372,
      "learning_rate": 1.922186261894775e-05,
      "loss": 0.2092,
      "step": 253
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.075041098782442,
      "learning_rate": 1.9214279609196632e-05,
      "loss": 0.1941,
      "step": 254
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.0341228117597683,
      "learning_rate": 1.9206661340451927e-05,
      "loss": 0.1555,
      "step": 255
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.1119610276967546,
      "learning_rate": 1.9199007841865395e-05,
      "loss": 0.1792,
      "step": 256
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.930768434457091,
      "learning_rate": 1.919131914272361e-05,
      "loss": 0.1916,
      "step": 257
    },
    {
      "epoch": 0.93,
      "grad_norm": 6.188828842690194,
      "learning_rate": 1.9183595272447843e-05,
      "loss": 0.2002,
      "step": 258
    },
    {
      "epoch": 0.94,
      "grad_norm": 5.415222515014791,
      "learning_rate": 1.9175836260593937e-05,
      "loss": 0.2208,
      "step": 259
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.4738850204790706,
      "learning_rate": 1.9168042136852228e-05,
      "loss": 0.1406,
      "step": 260
    },
    {
      "epoch": 0.95,
      "grad_norm": 5.329131640007567,
      "learning_rate": 1.916021293104739e-05,
      "loss": 0.2008,
      "step": 261
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.3257765001120143,
      "learning_rate": 1.9152348673138355e-05,
      "loss": 0.1554,
      "step": 262
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.693530182633664,
      "learning_rate": 1.914444939321817e-05,
      "loss": 0.1566,
      "step": 263
    },
    {
      "epoch": 0.96,
      "grad_norm": 3.245725077616388,
      "learning_rate": 1.913651512151391e-05,
      "loss": 0.1848,
      "step": 264
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.9134572204267684,
      "learning_rate": 1.9128545888386537e-05,
      "loss": 0.1812,
      "step": 265
    },
    {
      "epoch": 0.96,
      "grad_norm": 3.4660094621483957,
      "learning_rate": 1.9120541724330802e-05,
      "loss": 0.2319,
      "step": 266
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.2377157513624635,
      "learning_rate": 1.9112502659975122e-05,
      "loss": 0.2421,
      "step": 267
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.2389817823232567,
      "learning_rate": 1.910442872608145e-05,
      "loss": 0.1975,
      "step": 268
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.7644043566808019,
      "learning_rate": 1.9096319953545186e-05,
      "loss": 0.143,
      "step": 269
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.6256754670976832,
      "learning_rate": 1.908817637339503e-05,
      "loss": 0.1733,
      "step": 270
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.8315145141270555,
      "learning_rate": 1.9079998016792884e-05,
      "loss": 0.1925,
      "step": 271
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.404672759197368,
      "learning_rate": 1.9071784915033717e-05,
      "loss": 0.1981,
      "step": 272
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.141620203041048,
      "learning_rate": 1.9063537099545456e-05,
      "loss": 0.2028,
      "step": 273
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.3122940568496484,
      "learning_rate": 1.9055254601888867e-05,
      "loss": 0.1891,
      "step": 274
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.8860834704818434,
      "learning_rate": 1.9046937453757413e-05,
      "loss": 0.2066,
      "step": 275
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.474444058656227,
      "learning_rate": 1.9038585686977168e-05,
      "loss": 0.1903,
      "step": 276
    },
    {
      "epoch": 1.0,
      "grad_norm": 6.220834314967025,
      "learning_rate": 1.9030199333506667e-05,
      "loss": 0.2024,
      "step": 277
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.4267945413252041,
      "learning_rate": 1.9021778425436797e-05,
      "loss": 0.1615,
      "step": 278
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.7879083169277576,
      "learning_rate": 1.9013322994990663e-05,
      "loss": 0.1873,
      "step": 279
    },
    {
      "epoch": 1.01,
      "grad_norm": 3.1195876950952917,
      "learning_rate": 1.9004833074523478e-05,
      "loss": 0.2165,
      "step": 280
    },
    {
      "epoch": 1.02,
      "grad_norm": 4.854144089703857,
      "learning_rate": 1.8996308696522435e-05,
      "loss": 0.1893,
      "step": 281
    },
    {
      "epoch": 1.02,
      "grad_norm": 3.70882944826412,
      "learning_rate": 1.8987749893606575e-05,
      "loss": 0.2176,
      "step": 282
    },
    {
      "epoch": 1.03,
      "grad_norm": 3.7796610032532536,
      "learning_rate": 1.897915669852667e-05,
      "loss": 0.2176,
      "step": 283
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.9855057847958348,
      "learning_rate": 1.8970529144165103e-05,
      "loss": 0.2027,
      "step": 284
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.9758605320008039,
      "learning_rate": 1.8961867263535716e-05,
      "loss": 0.1694,
      "step": 285
    },
    {
      "epoch": 1.04,
      "grad_norm": 5.102642644944852,
      "learning_rate": 1.8953171089783725e-05,
      "loss": 0.1985,
      "step": 286
    },
    {
      "epoch": 1.04,
      "grad_norm": 3.2543005950166566,
      "learning_rate": 1.8944440656185556e-05,
      "loss": 0.1696,
      "step": 287
    },
    {
      "epoch": 1.04,
      "grad_norm": 3.5584636258947406,
      "learning_rate": 1.8935675996148738e-05,
      "loss": 0.183,
      "step": 288
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.407017198203034,
      "learning_rate": 1.892687714321177e-05,
      "loss": 0.1962,
      "step": 289
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.043682911740225,
      "learning_rate": 1.8918044131043987e-05,
      "loss": 0.1702,
      "step": 290
    },
    {
      "epoch": 1.05,
      "grad_norm": 3.6780900744494582,
      "learning_rate": 1.890917699344544e-05,
      "loss": 0.2027,
      "step": 291
    },
    {
      "epoch": 1.06,
      "grad_norm": 5.164665837636065,
      "learning_rate": 1.890027576434677e-05,
      "loss": 0.1587,
      "step": 292
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.1398407090298532,
      "learning_rate": 1.8891340477809055e-05,
      "loss": 0.1847,
      "step": 293
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.852078257346115,
      "learning_rate": 1.8882371168023708e-05,
      "loss": 0.143,
      "step": 294
    },
    {
      "epoch": 1.07,
      "grad_norm": 3.6919711611156467,
      "learning_rate": 1.887336786931233e-05,
      "loss": 0.2032,
      "step": 295
    },
    {
      "epoch": 1.07,
      "grad_norm": 4.556863118840815,
      "learning_rate": 1.8864330616126586e-05,
      "loss": 0.217,
      "step": 296
    },
    {
      "epoch": 1.08,
      "grad_norm": 3.3765282389059856,
      "learning_rate": 1.8855259443048067e-05,
      "loss": 0.1626,
      "step": 297
    },
    {
      "epoch": 1.08,
      "grad_norm": 2.5829969852710155,
      "learning_rate": 1.8846154384788162e-05,
      "loss": 0.1788,
      "step": 298
    },
    {
      "epoch": 1.08,
      "grad_norm": 5.075805270867687,
      "learning_rate": 1.8837015476187917e-05,
      "loss": 0.2225,
      "step": 299
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.8880177157758162,
      "learning_rate": 1.8827842752217917e-05,
      "loss": 0.1716,
      "step": 300
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.1521649968736736,
      "learning_rate": 1.8818636247978146e-05,
      "loss": 0.1715,
      "step": 301
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.9626798729665158,
      "learning_rate": 1.8809395998697835e-05,
      "loss": 0.2013,
      "step": 302
    },
    {
      "epoch": 1.1,
      "grad_norm": 3.058369631517267,
      "learning_rate": 1.880012203973536e-05,
      "loss": 0.1943,
      "step": 303
    },
    {
      "epoch": 1.1,
      "grad_norm": 5.488270353933477,
      "learning_rate": 1.8790814406578073e-05,
      "loss": 0.1833,
      "step": 304
    },
    {
      "epoch": 1.11,
      "grad_norm": 4.685311512196503,
      "learning_rate": 1.87814731348422e-05,
      "loss": 0.2316,
      "step": 305
    },
    {
      "epoch": 1.11,
      "grad_norm": 3.7320384197766616,
      "learning_rate": 1.877209826027267e-05,
      "loss": 0.1603,
      "step": 306
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.6481891382723979,
      "learning_rate": 1.8762689818743008e-05,
      "loss": 0.1622,
      "step": 307
    },
    {
      "epoch": 1.12,
      "grad_norm": 5.644998190465421,
      "learning_rate": 1.8753247846255175e-05,
      "loss": 0.1855,
      "step": 308
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.2321393392076536,
      "learning_rate": 1.874377237893945e-05,
      "loss": 0.1754,
      "step": 309
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.8834280167969224,
      "learning_rate": 1.8734263453054274e-05,
      "loss": 0.1835,
      "step": 310
    },
    {
      "epoch": 1.13,
      "grad_norm": 5.492233853467942,
      "learning_rate": 1.8724721104986123e-05,
      "loss": 0.2232,
      "step": 311
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.4242250524390119,
      "learning_rate": 1.871514537124936e-05,
      "loss": 0.2018,
      "step": 312
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.1249221000549425,
      "learning_rate": 1.8705536288486118e-05,
      "loss": 0.1849,
      "step": 313
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.0843956925311917,
      "learning_rate": 1.869589389346611e-05,
      "loss": 0.1728,
      "step": 314
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.785800054399146,
      "learning_rate": 1.868621822308655e-05,
      "loss": 0.2017,
      "step": 315
    },
    {
      "epoch": 1.14,
      "grad_norm": 4.249640134708495,
      "learning_rate": 1.8676509314371977e-05,
      "loss": 0.2105,
      "step": 316
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.511072289018074,
      "learning_rate": 1.8666767204474094e-05,
      "loss": 0.1906,
      "step": 317
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.8096550921548866,
      "learning_rate": 1.8656991930671687e-05,
      "loss": 0.1985,
      "step": 318
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.7248204418661588,
      "learning_rate": 1.8647183530370415e-05,
      "loss": 0.1934,
      "step": 319
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.1999611680627995,
      "learning_rate": 1.863734204110272e-05,
      "loss": 0.1963,
      "step": 320
    },
    {
      "epoch": 1.16,
      "grad_norm": 3.883573530691994,
      "learning_rate": 1.862746750052764e-05,
      "loss": 0.1542,
      "step": 321
    },
    {
      "epoch": 1.17,
      "grad_norm": 3.3431228063402654,
      "learning_rate": 1.861755994643071e-05,
      "loss": 0.1942,
      "step": 322
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.4330036856731683,
      "learning_rate": 1.8607619416723767e-05,
      "loss": 0.1424,
      "step": 323
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.5436726427593486,
      "learning_rate": 1.859764594944485e-05,
      "loss": 0.1894,
      "step": 324
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.9100137445854823,
      "learning_rate": 1.8587639582758032e-05,
      "loss": 0.1626,
      "step": 325
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.7635046945827246,
      "learning_rate": 1.8577600354953273e-05,
      "loss": 0.1338,
      "step": 326
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.901137989162943,
      "learning_rate": 1.856752830444628e-05,
      "loss": 0.1627,
      "step": 327
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.907571641908837,
      "learning_rate": 1.8557423469778356e-05,
      "loss": 0.2286,
      "step": 328
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.6357641602746575,
      "learning_rate": 1.854728588961626e-05,
      "loss": 0.1862,
      "step": 329
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.2297582716959188,
      "learning_rate": 1.8537115602752054e-05,
      "loss": 0.1598,
      "step": 330
    },
    {
      "epoch": 1.2,
      "grad_norm": 2.756509393456029,
      "learning_rate": 1.8526912648102943e-05,
      "loss": 0.1943,
      "step": 331
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.9929307269781298,
      "learning_rate": 1.851667706471115e-05,
      "loss": 0.1716,
      "step": 332
    },
    {
      "epoch": 1.21,
      "grad_norm": 2.4494045336334973,
      "learning_rate": 1.850640889174375e-05,
      "loss": 0.1869,
      "step": 333
    },
    {
      "epoch": 1.21,
      "grad_norm": 2.3399930651589775,
      "learning_rate": 1.8496108168492518e-05,
      "loss": 0.148,
      "step": 334
    },
    {
      "epoch": 1.21,
      "grad_norm": 3.2799543214110143,
      "learning_rate": 1.848577493437379e-05,
      "loss": 0.2033,
      "step": 335
    },
    {
      "epoch": 1.22,
      "grad_norm": 2.1138964961618587,
      "learning_rate": 1.8475409228928314e-05,
      "loss": 0.1749,
      "step": 336
    },
    {
      "epoch": 1.22,
      "grad_norm": 4.242446831879502,
      "learning_rate": 1.8465011091821072e-05,
      "loss": 0.2122,
      "step": 337
    },
    {
      "epoch": 1.22,
      "grad_norm": 4.695309052803132,
      "learning_rate": 1.8454580562841165e-05,
      "loss": 0.2316,
      "step": 338
    },
    {
      "epoch": 1.23,
      "grad_norm": 2.435856865520849,
      "learning_rate": 1.844411768190164e-05,
      "loss": 0.203,
      "step": 339
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.9001456672745267,
      "learning_rate": 1.8433622489039333e-05,
      "loss": 0.1451,
      "step": 340
    },
    {
      "epoch": 1.24,
      "grad_norm": 4.549439234979374,
      "learning_rate": 1.842309502441473e-05,
      "loss": 0.2038,
      "step": 341
    },
    {
      "epoch": 1.24,
      "grad_norm": 4.385192686383965,
      "learning_rate": 1.8412535328311813e-05,
      "loss": 0.1586,
      "step": 342
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.7513922789924818,
      "learning_rate": 1.8401943441137885e-05,
      "loss": 0.168,
      "step": 343
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.9285666796307979,
      "learning_rate": 1.839131940342344e-05,
      "loss": 0.1634,
      "step": 344
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.224049868781548,
      "learning_rate": 1.8380663255821996e-05,
      "loss": 0.1359,
      "step": 345
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.4503345293569696,
      "learning_rate": 1.8369975039109937e-05,
      "loss": 0.163,
      "step": 346
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.3812989577289319,
      "learning_rate": 1.8359254794186368e-05,
      "loss": 0.1529,
      "step": 347
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.5773982243086926,
      "learning_rate": 1.8348502562072955e-05,
      "loss": 0.1815,
      "step": 348
    },
    {
      "epoch": 1.26,
      "grad_norm": 2.6216961252274693,
      "learning_rate": 1.833771838391375e-05,
      "loss": 0.1544,
      "step": 349
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.6715459139127302,
      "learning_rate": 1.8326902300975063e-05,
      "loss": 0.1321,
      "step": 350
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.9867901775532317,
      "learning_rate": 1.831605435464528e-05,
      "loss": 0.1791,
      "step": 351
    },
    {
      "epoch": 1.28,
      "grad_norm": 4.5435987253093,
      "learning_rate": 1.8305174586434724e-05,
      "loss": 0.1515,
      "step": 352
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.591922944422485,
      "learning_rate": 1.8294263037975476e-05,
      "loss": 0.1705,
      "step": 353
    },
    {
      "epoch": 1.28,
      "grad_norm": 6.804612650593382,
      "learning_rate": 1.828331975102123e-05,
      "loss": 0.2053,
      "step": 354
    },
    {
      "epoch": 1.29,
      "grad_norm": 4.489415156525158,
      "learning_rate": 1.8272344767447136e-05,
      "loss": 0.1113,
      "step": 355
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.980091389672242,
      "learning_rate": 1.8261338129249623e-05,
      "loss": 0.1738,
      "step": 356
    },
    {
      "epoch": 1.29,
      "grad_norm": 3.5329647835174973,
      "learning_rate": 1.8250299878546247e-05,
      "loss": 0.1633,
      "step": 357
    },
    {
      "epoch": 1.3,
      "grad_norm": 9.669322574297373,
      "learning_rate": 1.8239230057575542e-05,
      "loss": 0.2105,
      "step": 358
    },
    {
      "epoch": 1.3,
      "grad_norm": 4.629218320442219,
      "learning_rate": 1.8228128708696843e-05,
      "loss": 0.1475,
      "step": 359
    },
    {
      "epoch": 1.3,
      "grad_norm": 4.2683344830305066,
      "learning_rate": 1.8216995874390128e-05,
      "loss": 0.1766,
      "step": 360
    },
    {
      "epoch": 1.31,
      "grad_norm": 5.259612939815896,
      "learning_rate": 1.820583159725585e-05,
      "loss": 0.1653,
      "step": 361
    },
    {
      "epoch": 1.31,
      "grad_norm": 7.874360229733074,
      "learning_rate": 1.819463592001479e-05,
      "loss": 0.2051,
      "step": 362
    },
    {
      "epoch": 1.32,
      "grad_norm": 12.290094203185955,
      "learning_rate": 1.8183408885507875e-05,
      "loss": 0.2252,
      "step": 363
    },
    {
      "epoch": 1.32,
      "grad_norm": 4.16333993532805,
      "learning_rate": 1.817215053669603e-05,
      "loss": 0.2108,
      "step": 364
    },
    {
      "epoch": 1.32,
      "grad_norm": 3.0903366930930707,
      "learning_rate": 1.8160860916659992e-05,
      "loss": 0.1633,
      "step": 365
    },
    {
      "epoch": 1.33,
      "grad_norm": 3.7314448836454175,
      "learning_rate": 1.814954006860018e-05,
      "loss": 0.1782,
      "step": 366
    },
    {
      "epoch": 1.33,
      "grad_norm": 4.894045709612284,
      "learning_rate": 1.81381880358365e-05,
      "loss": 0.1879,
      "step": 367
    },
    {
      "epoch": 1.33,
      "grad_norm": 8.950871898351894,
      "learning_rate": 1.8126804861808175e-05,
      "loss": 0.244,
      "step": 368
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.935243161337477,
      "learning_rate": 1.8115390590073612e-05,
      "loss": 0.2373,
      "step": 369
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.0892858965142254,
      "learning_rate": 1.81039452643102e-05,
      "loss": 0.1375,
      "step": 370
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.515794039091583,
      "learning_rate": 1.8092468928314174e-05,
      "loss": 0.1906,
      "step": 371
    },
    {
      "epoch": 1.35,
      "grad_norm": 6.421281762051045,
      "learning_rate": 1.808096162600041e-05,
      "loss": 0.1788,
      "step": 372
    },
    {
      "epoch": 1.35,
      "grad_norm": 5.868435345616694,
      "learning_rate": 1.80694234014023e-05,
      "loss": 0.2106,
      "step": 373
    },
    {
      "epoch": 1.36,
      "grad_norm": 6.43009286592627,
      "learning_rate": 1.8057854298671545e-05,
      "loss": 0.2174,
      "step": 374
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.8999377814562108,
      "learning_rate": 1.804625436207802e-05,
      "loss": 0.2107,
      "step": 375
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.4891862257518893,
      "learning_rate": 1.803462363600957e-05,
      "loss": 0.1765,
      "step": 376
    },
    {
      "epoch": 1.37,
      "grad_norm": 2.8148290153185176,
      "learning_rate": 1.8022962164971867e-05,
      "loss": 0.1718,
      "step": 377
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.283593621364449,
      "learning_rate": 1.8011269993588234e-05,
      "loss": 0.2016,
      "step": 378
    },
    {
      "epoch": 1.37,
      "grad_norm": 2.7240607903837484,
      "learning_rate": 1.799954716659946e-05,
      "loss": 0.1616,
      "step": 379
    },
    {
      "epoch": 1.38,
      "grad_norm": 7.249514152232641,
      "learning_rate": 1.798779372886365e-05,
      "loss": 0.2167,
      "step": 380
    },
    {
      "epoch": 1.38,
      "grad_norm": 4.823177269746868,
      "learning_rate": 1.7976009725356038e-05,
      "loss": 0.1705,
      "step": 381
    },
    {
      "epoch": 1.38,
      "grad_norm": 2.0737152931895473,
      "learning_rate": 1.796419520116882e-05,
      "loss": 0.1362,
      "step": 382
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.799994738461166,
      "learning_rate": 1.795235020151098e-05,
      "loss": 0.1902,
      "step": 383
    },
    {
      "epoch": 1.39,
      "grad_norm": 3.154863855560098,
      "learning_rate": 1.7940474771708118e-05,
      "loss": 0.1785,
      "step": 384
    },
    {
      "epoch": 1.39,
      "grad_norm": 7.260410750937351,
      "learning_rate": 1.792856895720228e-05,
      "loss": 0.1899,
      "step": 385
    },
    {
      "epoch": 1.4,
      "grad_norm": 3.789354975114773,
      "learning_rate": 1.791663280355178e-05,
      "loss": 0.1398,
      "step": 386
    },
    {
      "epoch": 1.4,
      "grad_norm": 2.471019931110548,
      "learning_rate": 1.790466635643103e-05,
      "loss": 0.1435,
      "step": 387
    },
    {
      "epoch": 1.41,
      "grad_norm": 2.7578364315313038,
      "learning_rate": 1.789266966163035e-05,
      "loss": 0.2226,
      "step": 388
    },
    {
      "epoch": 1.41,
      "grad_norm": 5.486957706332273,
      "learning_rate": 1.7880642765055816e-05,
      "loss": 0.2131,
      "step": 389
    },
    {
      "epoch": 1.41,
      "grad_norm": 5.818537789023371,
      "learning_rate": 1.786858571272907e-05,
      "loss": 0.1607,
      "step": 390
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.2136227115170777,
      "learning_rate": 1.7856498550787144e-05,
      "loss": 0.1343,
      "step": 391
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.5229015464461961,
      "learning_rate": 1.7844381325482293e-05,
      "loss": 0.1439,
      "step": 392
    },
    {
      "epoch": 1.42,
      "grad_norm": 3.7097316114681247,
      "learning_rate": 1.7832234083181798e-05,
      "loss": 0.1705,
      "step": 393
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.142154874900152,
      "learning_rate": 1.7820056870367813e-05,
      "loss": 0.1675,
      "step": 394
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.856745555144273,
      "learning_rate": 1.7807849733637177e-05,
      "loss": 0.1743,
      "step": 395
    },
    {
      "epoch": 1.43,
      "grad_norm": 2.2781705347105574,
      "learning_rate": 1.7795612719701228e-05,
      "loss": 0.2098,
      "step": 396
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.6131174517225353,
      "learning_rate": 1.7783345875385636e-05,
      "loss": 0.1858,
      "step": 397
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.3350288472401863,
      "learning_rate": 1.7771049247630215e-05,
      "loss": 0.1797,
      "step": 398
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.2632289769561353,
      "learning_rate": 1.7758722883488747e-05,
      "loss": 0.1883,
      "step": 399
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.7542308866562655,
      "learning_rate": 1.7746366830128803e-05,
      "loss": 0.183,
      "step": 400
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.7123868563789086,
      "learning_rate": 1.773398113483157e-05,
      "loss": 0.1631,
      "step": 401
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.1534890362072965,
      "learning_rate": 1.7721565844991643e-05,
      "loss": 0.1599,
      "step": 402
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.164814912872314,
      "learning_rate": 1.7709121008116883e-05,
      "loss": 0.1851,
      "step": 403
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.0030157197464,
      "learning_rate": 1.76966466718282e-05,
      "loss": 0.1942,
      "step": 404
    },
    {
      "epoch": 1.47,
      "grad_norm": 3.488435796832804,
      "learning_rate": 1.768414288385939e-05,
      "loss": 0.1327,
      "step": 405
    },
    {
      "epoch": 1.47,
      "grad_norm": 3.5011582713063167,
      "learning_rate": 1.7671609692056946e-05,
      "loss": 0.1652,
      "step": 406
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.2725214247403236,
      "learning_rate": 1.765904714437988e-05,
      "loss": 0.1891,
      "step": 407
    },
    {
      "epoch": 1.48,
      "grad_norm": 2.827813302901743,
      "learning_rate": 1.7646455288899535e-05,
      "loss": 0.1432,
      "step": 408
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.371825459719137,
      "learning_rate": 1.76338341737994e-05,
      "loss": 0.1414,
      "step": 409
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.979991911142439,
      "learning_rate": 1.7621183847374935e-05,
      "loss": 0.1582,
      "step": 410
    },
    {
      "epoch": 1.49,
      "grad_norm": 2.6953718527839974,
      "learning_rate": 1.7608504358033363e-05,
      "loss": 0.1791,
      "step": 411
    },
    {
      "epoch": 1.49,
      "grad_norm": 2.372842866632226,
      "learning_rate": 1.7595795754293514e-05,
      "loss": 0.183,
      "step": 412
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.1462489900704242,
      "learning_rate": 1.7583058084785626e-05,
      "loss": 0.1525,
      "step": 413
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.794684903380181,
      "learning_rate": 1.7570291398251153e-05,
      "loss": 0.1563,
      "step": 414
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.5536847761281407,
      "learning_rate": 1.7557495743542586e-05,
      "loss": 0.1498,
      "step": 415
    },
    {
      "epoch": 1.51,
      "grad_norm": 5.033146747767831,
      "learning_rate": 1.7544671169623263e-05,
      "loss": 0.2027,
      "step": 416
    },
    {
      "epoch": 1.51,
      "grad_norm": 4.627482726440702,
      "learning_rate": 1.753181772556719e-05,
      "loss": 0.2272,
      "step": 417
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.8280182578141009,
      "learning_rate": 1.751893546055884e-05,
      "loss": 0.1536,
      "step": 418
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.8824484299005066,
      "learning_rate": 1.750602442389297e-05,
      "loss": 0.2159,
      "step": 419
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.8759220400623319,
      "learning_rate": 1.749308466497444e-05,
      "loss": 0.1189,
      "step": 420
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.9833046214559626,
      "learning_rate": 1.7480116233318013e-05,
      "loss": 0.1366,
      "step": 421
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.7395056435824792,
      "learning_rate": 1.746711917854817e-05,
      "loss": 0.1538,
      "step": 422
    },
    {
      "epoch": 1.53,
      "grad_norm": 3.5305473680481865,
      "learning_rate": 1.745409355039892e-05,
      "loss": 0.1892,
      "step": 423
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.258477367145113,
      "learning_rate": 1.744103939871361e-05,
      "loss": 0.2052,
      "step": 424
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.390083899989616,
      "learning_rate": 1.7427956773444733e-05,
      "loss": 0.1416,
      "step": 425
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.530447990557084,
      "learning_rate": 1.7414845724653743e-05,
      "loss": 0.1506,
      "step": 426
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.3749722590696383,
      "learning_rate": 1.7401706302510854e-05,
      "loss": 0.17,
      "step": 427
    },
    {
      "epoch": 1.55,
      "grad_norm": 2.9766333202640345,
      "learning_rate": 1.738853855729485e-05,
      "loss": 0.2007,
      "step": 428
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.9436893098276,
      "learning_rate": 1.7375342539392903e-05,
      "loss": 0.147,
      "step": 429
    },
    {
      "epoch": 1.56,
      "grad_norm": 2.258826675200108,
      "learning_rate": 1.7362118299300363e-05,
      "loss": 0.1288,
      "step": 430
    },
    {
      "epoch": 1.56,
      "grad_norm": 2.691431274072751,
      "learning_rate": 1.7348865887620574e-05,
      "loss": 0.1721,
      "step": 431
    },
    {
      "epoch": 1.57,
      "grad_norm": 2.446923649161082,
      "learning_rate": 1.733558535506469e-05,
      "loss": 0.1558,
      "step": 432
    },
    {
      "epoch": 1.57,
      "grad_norm": 2.341545822778966,
      "learning_rate": 1.732227675245147e-05,
      "loss": 0.1407,
      "step": 433
    },
    {
      "epoch": 1.57,
      "grad_norm": 2.3744345065715042,
      "learning_rate": 1.730894013070707e-05,
      "loss": 0.1684,
      "step": 434
    },
    {
      "epoch": 1.58,
      "grad_norm": 4.788243952486983,
      "learning_rate": 1.7295575540864878e-05,
      "loss": 0.1598,
      "step": 435
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.120925937437422,
      "learning_rate": 1.7282183034065296e-05,
      "loss": 0.1729,
      "step": 436
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.993159907132356,
      "learning_rate": 1.7268762661555557e-05,
      "loss": 0.1589,
      "step": 437
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.7651575966235176,
      "learning_rate": 1.7255314474689524e-05,
      "loss": 0.1595,
      "step": 438
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.7469664941889762,
      "learning_rate": 1.7241838524927485e-05,
      "loss": 0.14,
      "step": 439
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.58000014180683,
      "learning_rate": 1.7228334863835972e-05,
      "loss": 0.2079,
      "step": 440
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.392841581472421,
      "learning_rate": 1.7214803543087556e-05,
      "loss": 0.1557,
      "step": 441
    },
    {
      "epoch": 1.6,
      "grad_norm": 6.692545196051933,
      "learning_rate": 1.7201244614460645e-05,
      "loss": 0.1946,
      "step": 442
    },
    {
      "epoch": 1.61,
      "grad_norm": 4.9639824306218445,
      "learning_rate": 1.7187658129839295e-05,
      "loss": 0.1934,
      "step": 443
    },
    {
      "epoch": 1.61,
      "grad_norm": 4.600507727479392,
      "learning_rate": 1.7174044141213e-05,
      "loss": 0.1554,
      "step": 444
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.831446485269553,
      "learning_rate": 1.716040270067651e-05,
      "loss": 0.1493,
      "step": 445
    },
    {
      "epoch": 1.62,
      "grad_norm": 3.6460841437042335,
      "learning_rate": 1.7146733860429614e-05,
      "loss": 0.2242,
      "step": 446
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.7349528609804215,
      "learning_rate": 1.7133037672776943e-05,
      "loss": 0.1657,
      "step": 447
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.909659149899981,
      "learning_rate": 1.7119314190127786e-05,
      "loss": 0.1416,
      "step": 448
    },
    {
      "epoch": 1.63,
      "grad_norm": 3.800824146188335,
      "learning_rate": 1.7105563464995873e-05,
      "loss": 0.1596,
      "step": 449
    },
    {
      "epoch": 1.63,
      "grad_norm": 2.6123862898271386,
      "learning_rate": 1.7091785549999177e-05,
      "loss": 0.1164,
      "step": 450
    },
    {
      "epoch": 1.63,
      "grad_norm": 2.5702633045133734,
      "learning_rate": 1.7077980497859715e-05,
      "loss": 0.1657,
      "step": 451
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.8590199199348927,
      "learning_rate": 1.7064148361403347e-05,
      "loss": 0.1544,
      "step": 452
    },
    {
      "epoch": 1.64,
      "grad_norm": 2.2010676694838485,
      "learning_rate": 1.7050289193559576e-05,
      "loss": 0.1552,
      "step": 453
    },
    {
      "epoch": 1.64,
      "grad_norm": 3.4735777802822874,
      "learning_rate": 1.7036403047361336e-05,
      "loss": 0.1845,
      "step": 454
    },
    {
      "epoch": 1.65,
      "grad_norm": 3.525410489743807,
      "learning_rate": 1.70224899759448e-05,
      "loss": 0.1757,
      "step": 455
    },
    {
      "epoch": 1.65,
      "grad_norm": 5.245801333039583,
      "learning_rate": 1.7008550032549167e-05,
      "loss": 0.1303,
      "step": 456
    },
    {
      "epoch": 1.66,
      "grad_norm": 2.7102854239542022,
      "learning_rate": 1.699458327051647e-05,
      "loss": 0.1493,
      "step": 457
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.7401112413904483,
      "learning_rate": 1.6980589743291362e-05,
      "loss": 0.145,
      "step": 458
    },
    {
      "epoch": 1.66,
      "grad_norm": 4.34019844447212,
      "learning_rate": 1.6966569504420913e-05,
      "loss": 0.1506,
      "step": 459
    },
    {
      "epoch": 1.67,
      "grad_norm": 5.020753282702713,
      "learning_rate": 1.695252260755441e-05,
      "loss": 0.1665,
      "step": 460
    },
    {
      "epoch": 1.67,
      "grad_norm": 5.692611051423799,
      "learning_rate": 1.693844910644314e-05,
      "loss": 0.202,
      "step": 461
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.3944083840141086,
      "learning_rate": 1.6924349054940204e-05,
      "loss": 0.129,
      "step": 462
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.248559809005285,
      "learning_rate": 1.6910222507000295e-05,
      "loss": 0.1518,
      "step": 463
    },
    {
      "epoch": 1.68,
      "grad_norm": 5.043263710574603,
      "learning_rate": 1.6896069516679494e-05,
      "loss": 0.147,
      "step": 464
    },
    {
      "epoch": 1.68,
      "grad_norm": 4.744308803957159,
      "learning_rate": 1.688189013813507e-05,
      "loss": 0.1708,
      "step": 465
    },
    {
      "epoch": 1.69,
      "grad_norm": 3.7010206909367422,
      "learning_rate": 1.6867684425625265e-05,
      "loss": 0.1399,
      "step": 466
    },
    {
      "epoch": 1.69,
      "grad_norm": 4.972435061041302,
      "learning_rate": 1.6853452433509087e-05,
      "loss": 0.1423,
      "step": 467
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.827326748489026,
      "learning_rate": 1.683919421624611e-05,
      "loss": 0.1689,
      "step": 468
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.8240952506218318,
      "learning_rate": 1.6824909828396254e-05,
      "loss": 0.1335,
      "step": 469
    },
    {
      "epoch": 1.7,
      "grad_norm": 3.8849118259684547,
      "learning_rate": 1.681059932461959e-05,
      "loss": 0.1741,
      "step": 470
    },
    {
      "epoch": 1.71,
      "grad_norm": 5.192115777745488,
      "learning_rate": 1.679626275967612e-05,
      "loss": 0.1519,
      "step": 471
    },
    {
      "epoch": 1.71,
      "grad_norm": 6.675697646151228,
      "learning_rate": 1.6781900188425565e-05,
      "loss": 0.2234,
      "step": 472
    },
    {
      "epoch": 1.71,
      "grad_norm": 2.378328321963269,
      "learning_rate": 1.6767511665827167e-05,
      "loss": 0.1446,
      "step": 473
    },
    {
      "epoch": 1.72,
      "grad_norm": 2.152827717749817,
      "learning_rate": 1.6753097246939475e-05,
      "loss": 0.1939,
      "step": 474
    },
    {
      "epoch": 1.72,
      "grad_norm": 4.530064408209738,
      "learning_rate": 1.6738656986920123e-05,
      "loss": 0.174,
      "step": 475
    },
    {
      "epoch": 1.72,
      "grad_norm": 4.147277341203001,
      "learning_rate": 1.672419094102563e-05,
      "loss": 0.1599,
      "step": 476
    },
    {
      "epoch": 1.73,
      "grad_norm": 4.06626962419669,
      "learning_rate": 1.670969916461119e-05,
      "loss": 0.1363,
      "step": 477
    },
    {
      "epoch": 1.73,
      "grad_norm": 2.2285257210803655,
      "learning_rate": 1.6695181713130462e-05,
      "loss": 0.1243,
      "step": 478
    },
    {
      "epoch": 1.74,
      "grad_norm": 2.11800924466845,
      "learning_rate": 1.6680638642135337e-05,
      "loss": 0.1815,
      "step": 479
    },
    {
      "epoch": 1.74,
      "grad_norm": 3.5710416042434225,
      "learning_rate": 1.6666070007275746e-05,
      "loss": 0.1618,
      "step": 480
    },
    {
      "epoch": 1.74,
      "grad_norm": 3.4925284732723556,
      "learning_rate": 1.6651475864299453e-05,
      "loss": 0.1431,
      "step": 481
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.1151296227345755,
      "learning_rate": 1.6636856269051813e-05,
      "loss": 0.1711,
      "step": 482
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.211774961679525,
      "learning_rate": 1.662221127747559e-05,
      "loss": 0.1556,
      "step": 483
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.5191237192219527,
      "learning_rate": 1.6607540945610722e-05,
      "loss": 0.133,
      "step": 484
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.7691955437105935,
      "learning_rate": 1.659284532959411e-05,
      "loss": 0.171,
      "step": 485
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.7670255226792704,
      "learning_rate": 1.6578124485659414e-05,
      "loss": 0.1575,
      "step": 486
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.8490943163056435,
      "learning_rate": 1.6563378470136823e-05,
      "loss": 0.1991,
      "step": 487
    },
    {
      "epoch": 1.77,
      "grad_norm": 4.3648865869808615,
      "learning_rate": 1.6548607339452853e-05,
      "loss": 0.1308,
      "step": 488
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.997485749829696,
      "learning_rate": 1.653381115013012e-05,
      "loss": 0.1599,
      "step": 489
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.240270800677183,
      "learning_rate": 1.6518989958787126e-05,
      "loss": 0.1334,
      "step": 490
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.4535988541945755,
      "learning_rate": 1.6504143822138057e-05,
      "loss": 0.1354,
      "step": 491
    },
    {
      "epoch": 1.78,
      "grad_norm": 3.6949356337532597,
      "learning_rate": 1.6489272796992536e-05,
      "loss": 0.1218,
      "step": 492
    },
    {
      "epoch": 1.79,
      "grad_norm": 2.6278736186790126,
      "learning_rate": 1.6474376940255442e-05,
      "loss": 0.1348,
      "step": 493
    },
    {
      "epoch": 1.79,
      "grad_norm": 6.545343087933527,
      "learning_rate": 1.6459456308926662e-05,
      "loss": 0.1979,
      "step": 494
    },
    {
      "epoch": 1.79,
      "grad_norm": 3.8448673196036114,
      "learning_rate": 1.644451096010088e-05,
      "loss": 0.1453,
      "step": 495
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.222708589993964,
      "learning_rate": 1.642954095096737e-05,
      "loss": 0.1607,
      "step": 496
    },
    {
      "epoch": 1.8,
      "grad_norm": 2.0932849514016136,
      "learning_rate": 1.6414546338809783e-05,
      "loss": 0.1323,
      "step": 497
    },
    {
      "epoch": 1.8,
      "grad_norm": 2.090129133667646,
      "learning_rate": 1.639952718100589e-05,
      "loss": 0.1374,
      "step": 498
    },
    {
      "epoch": 1.81,
      "grad_norm": 3.9907789165574146,
      "learning_rate": 1.63844835350274e-05,
      "loss": 0.1673,
      "step": 499
    },
    {
      "epoch": 1.81,
      "grad_norm": 3.476284562079525,
      "learning_rate": 1.636941545843973e-05,
      "loss": 0.1558,
      "step": 500
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.9177845569523475,
      "learning_rate": 1.6354323008901774e-05,
      "loss": 0.15,
      "step": 501
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4575612480313431,
      "learning_rate": 1.6339206244165705e-05,
      "loss": 0.1188,
      "step": 502
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.717706895583365,
      "learning_rate": 1.632406522207672e-05,
      "loss": 0.1334,
      "step": 503
    },
    {
      "epoch": 1.83,
      "grad_norm": 5.659401345162485,
      "learning_rate": 1.630890000057285e-05,
      "loss": 0.1779,
      "step": 504
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.491520253691608,
      "learning_rate": 1.6293710637684733e-05,
      "loss": 0.1338,
      "step": 505
    },
    {
      "epoch": 1.83,
      "grad_norm": 3.501568193131719,
      "learning_rate": 1.6278497191535364e-05,
      "loss": 0.1632,
      "step": 506
    },
    {
      "epoch": 1.84,
      "grad_norm": 4.368617842420329,
      "learning_rate": 1.6263259720339917e-05,
      "loss": 0.1522,
      "step": 507
    },
    {
      "epoch": 1.84,
      "grad_norm": 2.9101425724438186,
      "learning_rate": 1.6247998282405486e-05,
      "loss": 0.1488,
      "step": 508
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.808283260496058,
      "learning_rate": 1.623271293613088e-05,
      "loss": 0.179,
      "step": 509
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.3428833027807503,
      "learning_rate": 1.621740374000639e-05,
      "loss": 0.1387,
      "step": 510
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.7989422431398234,
      "learning_rate": 1.6202070752613578e-05,
      "loss": 0.1447,
      "step": 511
    },
    {
      "epoch": 1.86,
      "grad_norm": 5.55243784188933,
      "learning_rate": 1.6186714032625036e-05,
      "loss": 0.1494,
      "step": 512
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.5876168671231325,
      "learning_rate": 1.6171333638804177e-05,
      "loss": 0.1445,
      "step": 513
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.792462005235641,
      "learning_rate": 1.6155929630004995e-05,
      "loss": 0.1399,
      "step": 514
    },
    {
      "epoch": 1.87,
      "grad_norm": 2.6737163218064905,
      "learning_rate": 1.6140502065171865e-05,
      "loss": 0.1911,
      "step": 515
    },
    {
      "epoch": 1.87,
      "grad_norm": 3.288839494781435,
      "learning_rate": 1.6125051003339277e-05,
      "loss": 0.1426,
      "step": 516
    },
    {
      "epoch": 1.87,
      "grad_norm": 3.488830712267379,
      "learning_rate": 1.610957650363165e-05,
      "loss": 0.1948,
      "step": 517
    },
    {
      "epoch": 1.88,
      "grad_norm": 5.190300035091314,
      "learning_rate": 1.6094078625263085e-05,
      "loss": 0.202,
      "step": 518
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.8355095526941088,
      "learning_rate": 1.6078557427537144e-05,
      "loss": 0.1177,
      "step": 519
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.085477268858648,
      "learning_rate": 1.6063012969846624e-05,
      "loss": 0.1267,
      "step": 520
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.4529248520291653,
      "learning_rate": 1.604744531167332e-05,
      "loss": 0.1693,
      "step": 521
    },
    {
      "epoch": 1.89,
      "grad_norm": 3.0178268735143807,
      "learning_rate": 1.603185451258781e-05,
      "loss": 0.1645,
      "step": 522
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.9260850388201756,
      "learning_rate": 1.6016240632249224e-05,
      "loss": 0.162,
      "step": 523
    },
    {
      "epoch": 1.9,
      "grad_norm": 3.1705810235022667,
      "learning_rate": 1.6000603730405013e-05,
      "loss": 0.1814,
      "step": 524
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.083015779524835,
      "learning_rate": 1.5984943866890718e-05,
      "loss": 0.1824,
      "step": 525
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.8130349177336165,
      "learning_rate": 1.5969261101629744e-05,
      "loss": 0.1718,
      "step": 526
    },
    {
      "epoch": 1.91,
      "grad_norm": 3.932867184382188,
      "learning_rate": 1.5953555494633136e-05,
      "loss": 0.2069,
      "step": 527
    },
    {
      "epoch": 1.91,
      "grad_norm": 3.4904804321157124,
      "learning_rate": 1.593782710599934e-05,
      "loss": 0.1849,
      "step": 528
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.66805945239631,
      "learning_rate": 1.5922075995913976e-05,
      "loss": 0.2011,
      "step": 529
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.3140490522354895,
      "learning_rate": 1.5906302224649613e-05,
      "loss": 0.1611,
      "step": 530
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.6973615025326412,
      "learning_rate": 1.589050585256554e-05,
      "loss": 0.1263,
      "step": 531
    },
    {
      "epoch": 1.93,
      "grad_norm": 2.0370354462908686,
      "learning_rate": 1.5874686940107507e-05,
      "loss": 0.169,
      "step": 532
    },
    {
      "epoch": 1.93,
      "grad_norm": 2.288230600809746,
      "learning_rate": 1.5858845547807545e-05,
      "loss": 0.1618,
      "step": 533
    },
    {
      "epoch": 1.93,
      "grad_norm": 5.0758124402179,
      "learning_rate": 1.5842981736283686e-05,
      "loss": 0.1651,
      "step": 534
    },
    {
      "epoch": 1.94,
      "grad_norm": 4.119404936544872,
      "learning_rate": 1.582709556623976e-05,
      "loss": 0.1861,
      "step": 535
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.811183559427656,
      "learning_rate": 1.581118709846514e-05,
      "loss": 0.1233,
      "step": 536
    },
    {
      "epoch": 1.95,
      "grad_norm": 2.4729194568089277,
      "learning_rate": 1.5795256393834546e-05,
      "loss": 0.1612,
      "step": 537
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.7394691177808947,
      "learning_rate": 1.5779303513307765e-05,
      "loss": 0.1328,
      "step": 538
    },
    {
      "epoch": 1.95,
      "grad_norm": 2.053299955833243,
      "learning_rate": 1.576332851792945e-05,
      "loss": 0.1236,
      "step": 539
    },
    {
      "epoch": 1.96,
      "grad_norm": 3.210936578739264,
      "learning_rate": 1.574733146882889e-05,
      "loss": 0.1507,
      "step": 540
    },
    {
      "epoch": 1.96,
      "grad_norm": 2.5058993835342718,
      "learning_rate": 1.5731312427219738e-05,
      "loss": 0.1546,
      "step": 541
    },
    {
      "epoch": 1.96,
      "grad_norm": 2.2199777398424327,
      "learning_rate": 1.571527145439983e-05,
      "loss": 0.2009,
      "step": 542
    },
    {
      "epoch": 1.97,
      "grad_norm": 3.310778322843867,
      "learning_rate": 1.5699208611750904e-05,
      "loss": 0.187,
      "step": 543
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.424165488653433,
      "learning_rate": 1.5683123960738395e-05,
      "loss": 0.1595,
      "step": 544
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.5557389502573584,
      "learning_rate": 1.566701756291118e-05,
      "loss": 0.1161,
      "step": 545
    },
    {
      "epoch": 1.98,
      "grad_norm": 2.0891723510913467,
      "learning_rate": 1.5650889479901356e-05,
      "loss": 0.1316,
      "step": 546
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.5710257315573026,
      "learning_rate": 1.5634739773424007e-05,
      "loss": 0.1625,
      "step": 547
    },
    {
      "epoch": 1.99,
      "grad_norm": 2.588230137333489,
      "learning_rate": 1.5618568505276948e-05,
      "loss": 0.1624,
      "step": 548
    },
    {
      "epoch": 1.99,
      "grad_norm": 3.6544371356800887,
      "learning_rate": 1.5602375737340507e-05,
      "loss": 0.1609,
      "step": 549
    },
    {
      "epoch": 1.99,
      "grad_norm": 2.423586451954487,
      "learning_rate": 1.558616153157728e-05,
      "loss": 0.1657,
      "step": 550
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.5339695017155988,
      "learning_rate": 1.5569925950031906e-05,
      "loss": 0.1668,
      "step": 551
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.9781615139327298,
      "learning_rate": 1.5553669054830806e-05,
      "loss": 0.1493,
      "step": 552
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.6930188711451026,
      "learning_rate": 1.553739090818196e-05,
      "loss": 0.1542,
      "step": 553
    },
    {
      "epoch": 2.01,
      "grad_norm": 2.7797364025493785,
      "learning_rate": 1.552109157237468e-05,
      "loss": 0.1464,
      "step": 554
    },
    {
      "epoch": 2.01,
      "grad_norm": 2.696510654578581,
      "learning_rate": 1.5504771109779348e-05,
      "loss": 0.1525,
      "step": 555
    },
    {
      "epoch": 2.01,
      "grad_norm": 6.6918895076554685,
      "learning_rate": 1.5488429582847194e-05,
      "loss": 0.2056,
      "step": 556
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.019023563423498,
      "learning_rate": 1.547206705411005e-05,
      "loss": 0.1386,
      "step": 557
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.9728321730671903,
      "learning_rate": 1.5455683586180117e-05,
      "loss": 0.1631,
      "step": 558
    },
    {
      "epoch": 2.03,
      "grad_norm": 3.911889269659483,
      "learning_rate": 1.5439279241749714e-05,
      "loss": 0.1803,
      "step": 559
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.8394667273054477,
      "learning_rate": 1.542285408359105e-05,
      "loss": 0.1675,
      "step": 560
    },
    {
      "epoch": 2.03,
      "grad_norm": 4.093274128064072,
      "learning_rate": 1.5406408174555978e-05,
      "loss": 0.156,
      "step": 561
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.848164435113447,
      "learning_rate": 1.5389941577575753e-05,
      "loss": 0.1569,
      "step": 562
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.6564858409846661,
      "learning_rate": 1.5373454355660805e-05,
      "loss": 0.1415,
      "step": 563
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.554610631406401,
      "learning_rate": 1.5356946571900465e-05,
      "loss": 0.1513,
      "step": 564
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.9867639979699985,
      "learning_rate": 1.5340418289462765e-05,
      "loss": 0.1582,
      "step": 565
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.643399853726559,
      "learning_rate": 1.5323869571594166e-05,
      "loss": 0.1421,
      "step": 566
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.1911260149017444,
      "learning_rate": 1.5307300481619334e-05,
      "loss": 0.1806,
      "step": 567
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.9052229931897104,
      "learning_rate": 1.5290711082940883e-05,
      "loss": 0.1276,
      "step": 568
    },
    {
      "epoch": 2.06,
      "grad_norm": 3.178075384690503,
      "learning_rate": 1.527410143903914e-05,
      "loss": 0.1619,
      "step": 569
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.3933412605550552,
      "learning_rate": 1.5257471613471908e-05,
      "loss": 0.1258,
      "step": 570
    },
    {
      "epoch": 2.07,
      "grad_norm": 3.929781120919883,
      "learning_rate": 1.5240821669874204e-05,
      "loss": 0.1738,
      "step": 571
    },
    {
      "epoch": 2.07,
      "grad_norm": 2.368831991887992,
      "learning_rate": 1.5224151671958045e-05,
      "loss": 0.1855,
      "step": 572
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.5633002534137816,
      "learning_rate": 1.5207461683512174e-05,
      "loss": 0.1473,
      "step": 573
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.0564880232824487,
      "learning_rate": 1.5190751768401835e-05,
      "loss": 0.1546,
      "step": 574
    },
    {
      "epoch": 2.08,
      "grad_norm": 3.3813174994385737,
      "learning_rate": 1.5174021990568518e-05,
      "loss": 0.1723,
      "step": 575
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.7458502122771002,
      "learning_rate": 1.515727241402972e-05,
      "loss": 0.146,
      "step": 576
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.3792766298864636,
      "learning_rate": 1.51405031028787e-05,
      "loss": 0.1452,
      "step": 577
    },
    {
      "epoch": 2.09,
      "grad_norm": 5.154008898661378,
      "learning_rate": 1.512371412128424e-05,
      "loss": 0.189,
      "step": 578
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.2158356135773447,
      "learning_rate": 1.5106905533490373e-05,
      "loss": 0.1701,
      "step": 579
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.2802305847588675,
      "learning_rate": 1.509007740381618e-05,
      "loss": 0.1287,
      "step": 580
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.802517989281865,
      "learning_rate": 1.5073229796655504e-05,
      "loss": 0.1723,
      "step": 581
    },
    {
      "epoch": 2.11,
      "grad_norm": 5.043367884459983,
      "learning_rate": 1.505636277647672e-05,
      "loss": 0.132,
      "step": 582
    },
    {
      "epoch": 2.11,
      "grad_norm": 3.9364570447266467,
      "learning_rate": 1.5039476407822501e-05,
      "loss": 0.1316,
      "step": 583
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.790422402968305,
      "learning_rate": 1.5022570755309542e-05,
      "loss": 0.1438,
      "step": 584
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.139570124252977,
      "learning_rate": 1.5005645883628342e-05,
      "loss": 0.1249,
      "step": 585
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.7297604359039545,
      "learning_rate": 1.4988701857542932e-05,
      "loss": 0.1557,
      "step": 586
    },
    {
      "epoch": 2.13,
      "grad_norm": 4.609583051574302,
      "learning_rate": 1.497173874189065e-05,
      "loss": 0.2113,
      "step": 587
    },
    {
      "epoch": 2.13,
      "grad_norm": 2.065397185639641,
      "learning_rate": 1.495475660158187e-05,
      "loss": 0.1789,
      "step": 588
    },
    {
      "epoch": 2.13,
      "grad_norm": 3.4534182242890794,
      "learning_rate": 1.4937755501599774e-05,
      "loss": 0.1966,
      "step": 589
    },
    {
      "epoch": 2.14,
      "grad_norm": 3.7137088675703933,
      "learning_rate": 1.492073550700009e-05,
      "loss": 0.1558,
      "step": 590
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.905846213256674,
      "learning_rate": 1.4903696682910847e-05,
      "loss": 0.1723,
      "step": 591
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.47784663516246,
      "learning_rate": 1.4886639094532129e-05,
      "loss": 0.1869,
      "step": 592
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.0044679459734596,
      "learning_rate": 1.486956280713582e-05,
      "loss": 0.1655,
      "step": 593
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.4131117849348476,
      "learning_rate": 1.4852467886065357e-05,
      "loss": 0.1618,
      "step": 594
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.671333119059078,
      "learning_rate": 1.4835354396735483e-05,
      "loss": 0.1863,
      "step": 595
    },
    {
      "epoch": 2.16,
      "grad_norm": 3.8424507020526653,
      "learning_rate": 1.4818222404631993e-05,
      "loss": 0.1798,
      "step": 596
    },
    {
      "epoch": 2.16,
      "grad_norm": 4.951912325631532,
      "learning_rate": 1.4801071975311478e-05,
      "loss": 0.1444,
      "step": 597
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.1295985694856765,
      "learning_rate": 1.4783903174401086e-05,
      "loss": 0.1685,
      "step": 598
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.874237145559384,
      "learning_rate": 1.4766716067598263e-05,
      "loss": 0.1352,
      "step": 599
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.242083664103047,
      "learning_rate": 1.4749510720670506e-05,
      "loss": 0.1374,
      "step": 600
    },
    {
      "epoch": 2.18,
      "grad_norm": 2.3038584984777737,
      "learning_rate": 1.4732287199455103e-05,
      "loss": 0.1463,
      "step": 601
    },
    {
      "epoch": 2.18,
      "grad_norm": 5.0262258343937045,
      "learning_rate": 1.4715045569858895e-05,
      "loss": 0.1328,
      "step": 602
    },
    {
      "epoch": 2.18,
      "grad_norm": 4.043534512425941,
      "learning_rate": 1.4697785897858013e-05,
      "loss": 0.1754,
      "step": 603
    },
    {
      "epoch": 2.19,
      "grad_norm": 2.4226711679753263,
      "learning_rate": 1.4680508249497622e-05,
      "loss": 0.2144,
      "step": 604
    },
    {
      "epoch": 2.19,
      "grad_norm": 2.6075633120754405,
      "learning_rate": 1.4663212690891681e-05,
      "loss": 0.149,
      "step": 605
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.9483058032740208,
      "learning_rate": 1.4645899288222686e-05,
      "loss": 0.1277,
      "step": 606
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.8634381599518843,
      "learning_rate": 1.4628568107741408e-05,
      "loss": 0.1398,
      "step": 607
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.8575537132659474,
      "learning_rate": 1.461121921576665e-05,
      "loss": 0.1436,
      "step": 608
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.9620752891015423,
      "learning_rate": 1.4593852678684986e-05,
      "loss": 0.1411,
      "step": 609
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.838103433532202,
      "learning_rate": 1.457646856295051e-05,
      "loss": 0.1204,
      "step": 610
    },
    {
      "epoch": 2.21,
      "grad_norm": 3.260278555454537,
      "learning_rate": 1.455906693508459e-05,
      "loss": 0.1339,
      "step": 611
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.7172095087010821,
      "learning_rate": 1.4541647861675592e-05,
      "loss": 0.115,
      "step": 612
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.991597648856854,
      "learning_rate": 1.452421140937865e-05,
      "loss": 0.1325,
      "step": 613
    },
    {
      "epoch": 2.22,
      "grad_norm": 3.5948078337979164,
      "learning_rate": 1.4506757644915393e-05,
      "loss": 0.1799,
      "step": 614
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.1725595869845393,
      "learning_rate": 1.4489286635073695e-05,
      "loss": 0.2037,
      "step": 615
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.3633370205327395,
      "learning_rate": 1.4471798446707426e-05,
      "loss": 0.1377,
      "step": 616
    },
    {
      "epoch": 2.24,
      "grad_norm": 3.6515647608687654,
      "learning_rate": 1.4454293146736187e-05,
      "loss": 0.1853,
      "step": 617
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.1163407494968527,
      "learning_rate": 1.443677080214506e-05,
      "loss": 0.107,
      "step": 618
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.8246831591572374,
      "learning_rate": 1.441923147998434e-05,
      "loss": 0.1621,
      "step": 619
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.9466425861832106,
      "learning_rate": 1.4401675247369307e-05,
      "loss": 0.1303,
      "step": 620
    },
    {
      "epoch": 2.25,
      "grad_norm": 3.5393100208262855,
      "learning_rate": 1.4384102171479932e-05,
      "loss": 0.1164,
      "step": 621
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.9109710089013174,
      "learning_rate": 1.4366512319560642e-05,
      "loss": 0.1333,
      "step": 622
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.9752913219647925,
      "learning_rate": 1.434890575892006e-05,
      "loss": 0.1439,
      "step": 623
    },
    {
      "epoch": 2.26,
      "grad_norm": 3.230634064111403,
      "learning_rate": 1.4331282556930753e-05,
      "loss": 0.1642,
      "step": 624
    },
    {
      "epoch": 2.26,
      "grad_norm": 2.246924034268435,
      "learning_rate": 1.4313642781028953e-05,
      "loss": 0.1201,
      "step": 625
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.6322864446435443,
      "learning_rate": 1.4295986498714326e-05,
      "loss": 0.1001,
      "step": 626
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.8890631930316624,
      "learning_rate": 1.427831377754969e-05,
      "loss": 0.1478,
      "step": 627
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.83222544176522,
      "learning_rate": 1.4260624685160778e-05,
      "loss": 0.1113,
      "step": 628
    },
    {
      "epoch": 2.28,
      "grad_norm": 3.281562685100209,
      "learning_rate": 1.4242919289235958e-05,
      "loss": 0.1365,
      "step": 629
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.5058391781451417,
      "learning_rate": 1.4225197657525996e-05,
      "loss": 0.1388,
      "step": 630
    },
    {
      "epoch": 2.29,
      "grad_norm": 3.198653024717054,
      "learning_rate": 1.4207459857843771e-05,
      "loss": 0.0818,
      "step": 631
    },
    {
      "epoch": 2.29,
      "grad_norm": 4.026932943074049,
      "learning_rate": 1.4189705958064041e-05,
      "loss": 0.1551,
      "step": 632
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.5337607362274954,
      "learning_rate": 1.417193602612317e-05,
      "loss": 0.1136,
      "step": 633
    },
    {
      "epoch": 2.3,
      "grad_norm": 4.125149342194021,
      "learning_rate": 1.4154150130018867e-05,
      "loss": 0.1267,
      "step": 634
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.6807972057361718,
      "learning_rate": 1.413634833780993e-05,
      "loss": 0.1162,
      "step": 635
    },
    {
      "epoch": 2.3,
      "grad_norm": 6.5015463609549915,
      "learning_rate": 1.4118530717615982e-05,
      "loss": 0.1712,
      "step": 636
    },
    {
      "epoch": 2.31,
      "grad_norm": 4.433843930215299,
      "learning_rate": 1.4100697337617218e-05,
      "loss": 0.1483,
      "step": 637
    },
    {
      "epoch": 2.31,
      "grad_norm": 3.023664438844803,
      "learning_rate": 1.4082848266054136e-05,
      "loss": 0.144,
      "step": 638
    },
    {
      "epoch": 2.32,
      "grad_norm": 3.8997338181314816,
      "learning_rate": 1.4064983571227282e-05,
      "loss": 0.1578,
      "step": 639
    },
    {
      "epoch": 2.32,
      "grad_norm": 5.196782343567986,
      "learning_rate": 1.4047103321496977e-05,
      "loss": 0.2304,
      "step": 640
    },
    {
      "epoch": 2.32,
      "grad_norm": 6.966578918531625,
      "learning_rate": 1.4029207585283072e-05,
      "loss": 0.1701,
      "step": 641
    },
    {
      "epoch": 2.33,
      "grad_norm": 5.437415435237373,
      "learning_rate": 1.4011296431064675e-05,
      "loss": 0.1554,
      "step": 642
    },
    {
      "epoch": 2.33,
      "grad_norm": 3.5287376283141723,
      "learning_rate": 1.3993369927379891e-05,
      "loss": 0.14,
      "step": 643
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.634767923381933,
      "learning_rate": 1.3975428142825562e-05,
      "loss": 0.1685,
      "step": 644
    },
    {
      "epoch": 2.34,
      "grad_norm": 3.8673725956552065,
      "learning_rate": 1.3957471146057e-05,
      "loss": 0.1594,
      "step": 645
    },
    {
      "epoch": 2.34,
      "grad_norm": 2.0633317372185687,
      "learning_rate": 1.3939499005787735e-05,
      "loss": 0.1252,
      "step": 646
    },
    {
      "epoch": 2.34,
      "grad_norm": 4.104068349515131,
      "learning_rate": 1.3921511790789235e-05,
      "loss": 0.1892,
      "step": 647
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.4003136635922295,
      "learning_rate": 1.3903509569890663e-05,
      "loss": 0.1454,
      "step": 648
    },
    {
      "epoch": 2.35,
      "grad_norm": 3.0059666624586203,
      "learning_rate": 1.388549241197859e-05,
      "loss": 0.1944,
      "step": 649
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.7332874811258845,
      "learning_rate": 1.3867460385996756e-05,
      "loss": 0.1556,
      "step": 650
    },
    {
      "epoch": 2.36,
      "grad_norm": 2.4249766244944784,
      "learning_rate": 1.3849413560945788e-05,
      "loss": 0.1684,
      "step": 651
    },
    {
      "epoch": 2.36,
      "grad_norm": 2.7624132464699303,
      "learning_rate": 1.3831352005882947e-05,
      "loss": 0.1691,
      "step": 652
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.6462363499749473,
      "learning_rate": 1.3813275789921856e-05,
      "loss": 0.1531,
      "step": 653
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.9098892115932387,
      "learning_rate": 1.3795184982232234e-05,
      "loss": 0.1559,
      "step": 654
    },
    {
      "epoch": 2.37,
      "grad_norm": 4.737845980438906,
      "learning_rate": 1.3777079652039649e-05,
      "loss": 0.1493,
      "step": 655
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.96373225232142,
      "learning_rate": 1.3758959868625233e-05,
      "loss": 0.1547,
      "step": 656
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.6573947051585216,
      "learning_rate": 1.3740825701325417e-05,
      "loss": 0.1407,
      "step": 657
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.2430731743842967,
      "learning_rate": 1.3722677219531684e-05,
      "loss": 0.1199,
      "step": 658
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.9306302053944655,
      "learning_rate": 1.370451449269029e-05,
      "loss": 0.1629,
      "step": 659
    },
    {
      "epoch": 2.39,
      "grad_norm": 2.166845655575119,
      "learning_rate": 1.3686337590301997e-05,
      "loss": 0.1672,
      "step": 660
    },
    {
      "epoch": 2.39,
      "grad_norm": 2.7289112859521425,
      "learning_rate": 1.366814658192181e-05,
      "loss": 0.1494,
      "step": 661
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.5117108734297187,
      "learning_rate": 1.364994153715872e-05,
      "loss": 0.1134,
      "step": 662
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.016864300473532,
      "learning_rate": 1.3631722525675412e-05,
      "loss": 0.1292,
      "step": 663
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.9808047084535987,
      "learning_rate": 1.361348961718804e-05,
      "loss": 0.1404,
      "step": 664
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.2189165012578695,
      "learning_rate": 1.359524288146591e-05,
      "loss": 0.1594,
      "step": 665
    },
    {
      "epoch": 2.41,
      "grad_norm": 3.587067010599109,
      "learning_rate": 1.3576982388331258e-05,
      "loss": 0.1048,
      "step": 666
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.872852383025508,
      "learning_rate": 1.3558708207658948e-05,
      "loss": 0.1292,
      "step": 667
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.1099352315229647,
      "learning_rate": 1.3540420409376237e-05,
      "loss": 0.0937,
      "step": 668
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.072876069571932,
      "learning_rate": 1.3522119063462482e-05,
      "loss": 0.1318,
      "step": 669
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.680823510834941,
      "learning_rate": 1.3503804239948874e-05,
      "loss": 0.1397,
      "step": 670
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.8436852709115086,
      "learning_rate": 1.3485476008918186e-05,
      "loss": 0.1376,
      "step": 671
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.483810788507097,
      "learning_rate": 1.3467134440504497e-05,
      "loss": 0.2099,
      "step": 672
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.024202118173722,
      "learning_rate": 1.3448779604892916e-05,
      "loss": 0.1631,
      "step": 673
    },
    {
      "epoch": 2.44,
      "grad_norm": 3.3874042816937453,
      "learning_rate": 1.3430411572319323e-05,
      "loss": 0.1549,
      "step": 674
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.0445973830792226,
      "learning_rate": 1.3412030413070096e-05,
      "loss": 0.1459,
      "step": 675
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.2567358837105926,
      "learning_rate": 1.3393636197481842e-05,
      "loss": 0.1509,
      "step": 676
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.2499068792437322,
      "learning_rate": 1.3375228995941135e-05,
      "loss": 0.1435,
      "step": 677
    },
    {
      "epoch": 2.46,
      "grad_norm": 2.6864374390893127,
      "learning_rate": 1.335680887888423e-05,
      "loss": 0.1579,
      "step": 678
    },
    {
      "epoch": 2.46,
      "grad_norm": 2.11068958107654,
      "learning_rate": 1.3338375916796813e-05,
      "loss": 0.1688,
      "step": 679
    },
    {
      "epoch": 2.46,
      "grad_norm": 3.2643755262296548,
      "learning_rate": 1.3319930180213713e-05,
      "loss": 0.1757,
      "step": 680
    },
    {
      "epoch": 2.47,
      "grad_norm": 2.1849128315834347,
      "learning_rate": 1.3301471739718659e-05,
      "loss": 0.0955,
      "step": 681
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.395893452706116,
      "learning_rate": 1.3283000665943972e-05,
      "loss": 0.1269,
      "step": 682
    },
    {
      "epoch": 2.47,
      "grad_norm": 3.8435765299606652,
      "learning_rate": 1.3264517029570325e-05,
      "loss": 0.1759,
      "step": 683
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.344144583853651,
      "learning_rate": 1.3246020901326465e-05,
      "loss": 0.1203,
      "step": 684
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.18453194642565,
      "learning_rate": 1.3227512351988927e-05,
      "loss": 0.1507,
      "step": 685
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.273708225796424,
      "learning_rate": 1.3208991452381798e-05,
      "loss": 0.1451,
      "step": 686
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.4991174581863214,
      "learning_rate": 1.3190458273376403e-05,
      "loss": 0.1629,
      "step": 687
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.2252484353713458,
      "learning_rate": 1.3171912885891063e-05,
      "loss": 0.1625,
      "step": 688
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.736884623960957,
      "learning_rate": 1.3153355360890815e-05,
      "loss": 0.1282,
      "step": 689
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.15785290638828,
      "learning_rate": 1.3134785769387147e-05,
      "loss": 0.1391,
      "step": 690
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.9811024522964154,
      "learning_rate": 1.3116204182437711e-05,
      "loss": 0.142,
      "step": 691
    },
    {
      "epoch": 2.51,
      "grad_norm": 3.662387340221045,
      "learning_rate": 1.3097610671146063e-05,
      "loss": 0.1779,
      "step": 692
    },
    {
      "epoch": 2.51,
      "grad_norm": 3.3860466495572146,
      "learning_rate": 1.307900530666139e-05,
      "loss": 0.1509,
      "step": 693
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.28060851336785,
      "learning_rate": 1.3060388160178237e-05,
      "loss": 0.149,
      "step": 694
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.8513128405544685,
      "learning_rate": 1.304175930293623e-05,
      "loss": 0.1709,
      "step": 695
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.4462263633502537,
      "learning_rate": 1.302311880621981e-05,
      "loss": 0.1194,
      "step": 696
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.9548570447274782,
      "learning_rate": 1.3004466741357953e-05,
      "loss": 0.1237,
      "step": 697
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.4806136831831311,
      "learning_rate": 1.2985803179723903e-05,
      "loss": 0.1269,
      "step": 698
    },
    {
      "epoch": 2.53,
      "grad_norm": 2.9715702991593127,
      "learning_rate": 1.2967128192734903e-05,
      "loss": 0.1349,
      "step": 699
    },
    {
      "epoch": 2.54,
      "grad_norm": 3.466274211369672,
      "learning_rate": 1.294844185185191e-05,
      "loss": 0.1688,
      "step": 700
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.7553058535020245,
      "learning_rate": 1.2929744228579325e-05,
      "loss": 0.1097,
      "step": 701
    },
    {
      "epoch": 2.54,
      "grad_norm": 3.0216851671552094,
      "learning_rate": 1.2911035394464724e-05,
      "loss": 0.1119,
      "step": 702
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.0733992908118593,
      "learning_rate": 1.2892315421098587e-05,
      "loss": 0.1498,
      "step": 703
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.9082569188886227,
      "learning_rate": 1.2873584380114012e-05,
      "loss": 0.1466,
      "step": 704
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.1693782252874465,
      "learning_rate": 1.2854842343186455e-05,
      "loss": 0.1188,
      "step": 705
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.8457394355525738,
      "learning_rate": 1.283608938203344e-05,
      "loss": 0.0983,
      "step": 706
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.9391659305171975,
      "learning_rate": 1.2817325568414299e-05,
      "loss": 0.1319,
      "step": 707
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.9424785468417647,
      "learning_rate": 1.2798550974129888e-05,
      "loss": 0.1374,
      "step": 708
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.620705577547139,
      "learning_rate": 1.2779765671022327e-05,
      "loss": 0.1311,
      "step": 709
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.7842863143216023,
      "learning_rate": 1.2760969730974692e-05,
      "loss": 0.1369,
      "step": 710
    },
    {
      "epoch": 2.58,
      "grad_norm": 3.4504806028599027,
      "learning_rate": 1.274216322591078e-05,
      "loss": 0.1343,
      "step": 711
    },
    {
      "epoch": 2.58,
      "grad_norm": 2.548126380604583,
      "learning_rate": 1.2723346227794817e-05,
      "loss": 0.1321,
      "step": 712
    },
    {
      "epoch": 2.58,
      "grad_norm": 4.370878901486762,
      "learning_rate": 1.2704518808631168e-05,
      "loss": 0.1326,
      "step": 713
    },
    {
      "epoch": 2.59,
      "grad_norm": 3.774946256872503,
      "learning_rate": 1.2685681040464081e-05,
      "loss": 0.1352,
      "step": 714
    },
    {
      "epoch": 2.59,
      "grad_norm": 3.1517657140996853,
      "learning_rate": 1.2666832995377412e-05,
      "loss": 0.1001,
      "step": 715
    },
    {
      "epoch": 2.59,
      "grad_norm": 3.0860083371990448,
      "learning_rate": 1.264797474549433e-05,
      "loss": 0.1631,
      "step": 716
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.9332984003257843,
      "learning_rate": 1.2629106362977064e-05,
      "loss": 0.1014,
      "step": 717
    },
    {
      "epoch": 2.6,
      "grad_norm": 3.1029076169572742,
      "learning_rate": 1.2610227920026608e-05,
      "loss": 0.1436,
      "step": 718
    },
    {
      "epoch": 2.61,
      "grad_norm": 3.7906759875527136,
      "learning_rate": 1.2591339488882458e-05,
      "loss": 0.1615,
      "step": 719
    },
    {
      "epoch": 2.61,
      "grad_norm": 4.934283918080738,
      "learning_rate": 1.2572441141822322e-05,
      "loss": 0.136,
      "step": 720
    },
    {
      "epoch": 2.61,
      "grad_norm": 5.096116343618296,
      "learning_rate": 1.255353295116187e-05,
      "loss": 0.1435,
      "step": 721
    },
    {
      "epoch": 2.62,
      "grad_norm": 4.146936550181582,
      "learning_rate": 1.2534614989254423e-05,
      "loss": 0.2009,
      "step": 722
    },
    {
      "epoch": 2.62,
      "grad_norm": 4.085178260856516,
      "learning_rate": 1.2515687328490696e-05,
      "loss": 0.1425,
      "step": 723
    },
    {
      "epoch": 2.62,
      "grad_norm": 2.4774821694630833,
      "learning_rate": 1.2496750041298515e-05,
      "loss": 0.1202,
      "step": 724
    },
    {
      "epoch": 2.63,
      "grad_norm": 4.005381342341569,
      "learning_rate": 1.2477803200142551e-05,
      "loss": 0.1252,
      "step": 725
    },
    {
      "epoch": 2.63,
      "grad_norm": 3.117326777003886,
      "learning_rate": 1.2458846877524025e-05,
      "loss": 0.1052,
      "step": 726
    },
    {
      "epoch": 2.63,
      "grad_norm": 5.011961221758716,
      "learning_rate": 1.2439881145980444e-05,
      "loss": 0.1592,
      "step": 727
    },
    {
      "epoch": 2.64,
      "grad_norm": 6.402429747161904,
      "learning_rate": 1.2420906078085316e-05,
      "loss": 0.1506,
      "step": 728
    },
    {
      "epoch": 2.64,
      "grad_norm": 5.911505976531602,
      "learning_rate": 1.2401921746447874e-05,
      "loss": 0.1356,
      "step": 729
    },
    {
      "epoch": 2.64,
      "grad_norm": 4.085941837226761,
      "learning_rate": 1.2382928223712807e-05,
      "loss": 0.1561,
      "step": 730
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.8930462870586915,
      "learning_rate": 1.2363925582559968e-05,
      "loss": 0.1513,
      "step": 731
    },
    {
      "epoch": 2.65,
      "grad_norm": 5.765133151777372,
      "learning_rate": 1.2344913895704099e-05,
      "loss": 0.149,
      "step": 732
    },
    {
      "epoch": 2.66,
      "grad_norm": 5.964131756390052,
      "learning_rate": 1.2325893235894563e-05,
      "loss": 0.1371,
      "step": 733
    },
    {
      "epoch": 2.66,
      "grad_norm": 5.656003687233519,
      "learning_rate": 1.2306863675915058e-05,
      "loss": 0.1365,
      "step": 734
    },
    {
      "epoch": 2.66,
      "grad_norm": 3.126958801018975,
      "learning_rate": 1.228782528858333e-05,
      "loss": 0.1202,
      "step": 735
    },
    {
      "epoch": 2.67,
      "grad_norm": 2.4984355268123632,
      "learning_rate": 1.2268778146750914e-05,
      "loss": 0.1414,
      "step": 736
    },
    {
      "epoch": 2.67,
      "grad_norm": 3.6962151946346475,
      "learning_rate": 1.2249722323302841e-05,
      "loss": 0.1453,
      "step": 737
    },
    {
      "epoch": 2.67,
      "grad_norm": 3.313461638240804,
      "learning_rate": 1.2230657891157365e-05,
      "loss": 0.1123,
      "step": 738
    },
    {
      "epoch": 2.68,
      "grad_norm": 4.601078169226705,
      "learning_rate": 1.2211584923265673e-05,
      "loss": 0.1497,
      "step": 739
    },
    {
      "epoch": 2.68,
      "grad_norm": 3.7073417285450403,
      "learning_rate": 1.2192503492611625e-05,
      "loss": 0.1199,
      "step": 740
    },
    {
      "epoch": 2.68,
      "grad_norm": 3.1824208325942913,
      "learning_rate": 1.2173413672211457e-05,
      "loss": 0.1697,
      "step": 741
    },
    {
      "epoch": 2.69,
      "grad_norm": 2.471535241485352,
      "learning_rate": 1.2154315535113513e-05,
      "loss": 0.1334,
      "step": 742
    },
    {
      "epoch": 2.69,
      "grad_norm": 3.4155726172820446,
      "learning_rate": 1.2135209154397962e-05,
      "loss": 0.1278,
      "step": 743
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.010032705672301,
      "learning_rate": 1.2116094603176513e-05,
      "loss": 0.1412,
      "step": 744
    },
    {
      "epoch": 2.7,
      "grad_norm": 3.0906837228871273,
      "learning_rate": 1.2096971954592146e-05,
      "loss": 0.1441,
      "step": 745
    },
    {
      "epoch": 2.7,
      "grad_norm": 2.9486512312912088,
      "learning_rate": 1.2077841281818816e-05,
      "loss": 0.1516,
      "step": 746
    },
    {
      "epoch": 2.71,
      "grad_norm": 2.266804011074804,
      "learning_rate": 1.2058702658061197e-05,
      "loss": 0.107,
      "step": 747
    },
    {
      "epoch": 2.71,
      "grad_norm": 2.5883749931630056,
      "learning_rate": 1.203955615655438e-05,
      "loss": 0.1841,
      "step": 748
    },
    {
      "epoch": 2.71,
      "grad_norm": 2.4797481006413724,
      "learning_rate": 1.2020401850563596e-05,
      "loss": 0.1044,
      "step": 749
    },
    {
      "epoch": 2.72,
      "grad_norm": 2.4908442077053508,
      "learning_rate": 1.2001239813383951e-05,
      "loss": 0.1816,
      "step": 750
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.5916571109906084,
      "learning_rate": 1.1982070118340128e-05,
      "loss": 0.1458,
      "step": 751
    },
    {
      "epoch": 2.72,
      "grad_norm": 2.8372465633779353,
      "learning_rate": 1.1962892838786116e-05,
      "loss": 0.1146,
      "step": 752
    },
    {
      "epoch": 2.73,
      "grad_norm": 2.6671483653198065,
      "learning_rate": 1.1943708048104923e-05,
      "loss": 0.1137,
      "step": 753
    },
    {
      "epoch": 2.73,
      "grad_norm": 4.055069878690088,
      "learning_rate": 1.19245158197083e-05,
      "loss": 0.1104,
      "step": 754
    },
    {
      "epoch": 2.74,
      "grad_norm": 2.486327947054191,
      "learning_rate": 1.1905316227036465e-05,
      "loss": 0.1429,
      "step": 755
    },
    {
      "epoch": 2.74,
      "grad_norm": 3.2617966124439186,
      "learning_rate": 1.1886109343557808e-05,
      "loss": 0.1379,
      "step": 756
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.9313464650423409,
      "learning_rate": 1.186689524276862e-05,
      "loss": 0.1198,
      "step": 757
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.729544641553348,
      "learning_rate": 1.1847673998192815e-05,
      "loss": 0.1379,
      "step": 758
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.7821147579366157,
      "learning_rate": 1.1828445683381628e-05,
      "loss": 0.1413,
      "step": 759
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.2212472895092787,
      "learning_rate": 1.180921037191337e-05,
      "loss": 0.1128,
      "step": 760
    },
    {
      "epoch": 2.76,
      "grad_norm": 3.8905217179467093,
      "learning_rate": 1.1789968137393108e-05,
      "loss": 0.1412,
      "step": 761
    },
    {
      "epoch": 2.76,
      "grad_norm": 2.84781205596861,
      "learning_rate": 1.1770719053452408e-05,
      "loss": 0.135,
      "step": 762
    },
    {
      "epoch": 2.76,
      "grad_norm": 2.4542496505430895,
      "learning_rate": 1.1751463193749044e-05,
      "loss": 0.1683,
      "step": 763
    },
    {
      "epoch": 2.77,
      "grad_norm": 2.879779310063717,
      "learning_rate": 1.1732200631966717e-05,
      "loss": 0.1012,
      "step": 764
    },
    {
      "epoch": 2.77,
      "grad_norm": 2.879306798231344,
      "learning_rate": 1.1712931441814776e-05,
      "loss": 0.1282,
      "step": 765
    },
    {
      "epoch": 2.78,
      "grad_norm": 2.836413783186633,
      "learning_rate": 1.1693655697027935e-05,
      "loss": 0.09,
      "step": 766
    },
    {
      "epoch": 2.78,
      "grad_norm": 3.436793644315394,
      "learning_rate": 1.1674373471365987e-05,
      "loss": 0.0996,
      "step": 767
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.9949362166058173,
      "learning_rate": 1.165508483861352e-05,
      "loss": 0.0854,
      "step": 768
    },
    {
      "epoch": 2.79,
      "grad_norm": 5.489936954970715,
      "learning_rate": 1.1635789872579649e-05,
      "loss": 0.1484,
      "step": 769
    },
    {
      "epoch": 2.79,
      "grad_norm": 6.2559843796412595,
      "learning_rate": 1.1616488647097718e-05,
      "loss": 0.1283,
      "step": 770
    },
    {
      "epoch": 2.79,
      "grad_norm": 3.2854828105346408,
      "learning_rate": 1.1597181236025024e-05,
      "loss": 0.1118,
      "step": 771
    },
    {
      "epoch": 2.8,
      "grad_norm": 2.1551228922320806,
      "learning_rate": 1.1577867713242532e-05,
      "loss": 0.1048,
      "step": 772
    },
    {
      "epoch": 2.8,
      "grad_norm": 3.1898874907443555,
      "learning_rate": 1.1558548152654596e-05,
      "loss": 0.0907,
      "step": 773
    },
    {
      "epoch": 2.8,
      "grad_norm": 2.5624001707919226,
      "learning_rate": 1.1539222628188675e-05,
      "loss": 0.1064,
      "step": 774
    },
    {
      "epoch": 2.81,
      "grad_norm": 3.449986468785642,
      "learning_rate": 1.151989121379505e-05,
      "loss": 0.1341,
      "step": 775
    },
    {
      "epoch": 2.81,
      "grad_norm": 4.025583340892815,
      "learning_rate": 1.1500553983446527e-05,
      "loss": 0.1442,
      "step": 776
    },
    {
      "epoch": 2.82,
      "grad_norm": 4.1909284054957805,
      "learning_rate": 1.148121101113819e-05,
      "loss": 0.1241,
      "step": 777
    },
    {
      "epoch": 2.82,
      "grad_norm": 3.8649243191633613,
      "learning_rate": 1.1461862370887076e-05,
      "loss": 0.141,
      "step": 778
    },
    {
      "epoch": 2.82,
      "grad_norm": 6.6164700789226325,
      "learning_rate": 1.1442508136731919e-05,
      "loss": 0.1329,
      "step": 779
    },
    {
      "epoch": 2.83,
      "grad_norm": 2.266029845209213,
      "learning_rate": 1.1423148382732854e-05,
      "loss": 0.1081,
      "step": 780
    },
    {
      "epoch": 2.83,
      "grad_norm": 3.3044700369516824,
      "learning_rate": 1.1403783182971144e-05,
      "loss": 0.0904,
      "step": 781
    },
    {
      "epoch": 2.83,
      "grad_norm": 2.88988382240665,
      "learning_rate": 1.1384412611548887e-05,
      "loss": 0.1107,
      "step": 782
    },
    {
      "epoch": 2.84,
      "grad_norm": 5.240950603595819,
      "learning_rate": 1.136503674258874e-05,
      "loss": 0.1299,
      "step": 783
    },
    {
      "epoch": 2.84,
      "grad_norm": 5.643337451467647,
      "learning_rate": 1.134565565023362e-05,
      "loss": 0.1525,
      "step": 784
    },
    {
      "epoch": 2.84,
      "grad_norm": 5.3873764814070455,
      "learning_rate": 1.1326269408646445e-05,
      "loss": 0.1751,
      "step": 785
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.968627074400663,
      "learning_rate": 1.1306878092009828e-05,
      "loss": 0.0931,
      "step": 786
    },
    {
      "epoch": 2.85,
      "grad_norm": 6.546444780384288,
      "learning_rate": 1.128748177452581e-05,
      "loss": 0.1543,
      "step": 787
    },
    {
      "epoch": 2.86,
      "grad_norm": 3.173360357653791,
      "learning_rate": 1.1268080530415557e-05,
      "loss": 0.1093,
      "step": 788
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.7129676786561316,
      "learning_rate": 1.1248674433919097e-05,
      "loss": 0.1051,
      "step": 789
    },
    {
      "epoch": 2.86,
      "grad_norm": 5.005043159410687,
      "learning_rate": 1.122926355929502e-05,
      "loss": 0.1245,
      "step": 790
    },
    {
      "epoch": 2.87,
      "grad_norm": 3.616282965265042,
      "learning_rate": 1.1209847980820208e-05,
      "loss": 0.1571,
      "step": 791
    },
    {
      "epoch": 2.87,
      "grad_norm": 6.867872868581158,
      "learning_rate": 1.119042777278953e-05,
      "loss": 0.1368,
      "step": 792
    },
    {
      "epoch": 2.87,
      "grad_norm": 3.9446600508604286,
      "learning_rate": 1.1171003009515577e-05,
      "loss": 0.1509,
      "step": 793
    },
    {
      "epoch": 2.88,
      "grad_norm": 3.0200121905654864,
      "learning_rate": 1.1151573765328374e-05,
      "loss": 0.1545,
      "step": 794
    },
    {
      "epoch": 2.88,
      "grad_norm": 2.588590429607821,
      "learning_rate": 1.1132140114575085e-05,
      "loss": 0.0959,
      "step": 795
    },
    {
      "epoch": 2.88,
      "grad_norm": 3.31566385103503,
      "learning_rate": 1.1112702131619747e-05,
      "loss": 0.1096,
      "step": 796
    },
    {
      "epoch": 2.89,
      "grad_norm": 2.8316523924308608,
      "learning_rate": 1.1093259890842962e-05,
      "loss": 0.153,
      "step": 797
    },
    {
      "epoch": 2.89,
      "grad_norm": 2.98969306150191,
      "learning_rate": 1.1073813466641633e-05,
      "loss": 0.1479,
      "step": 798
    },
    {
      "epoch": 2.89,
      "grad_norm": 3.7372446047393515,
      "learning_rate": 1.1054362933428666e-05,
      "loss": 0.1468,
      "step": 799
    },
    {
      "epoch": 2.9,
      "grad_norm": 6.5695166197444905,
      "learning_rate": 1.1034908365632695e-05,
      "loss": 0.2094,
      "step": 800
    },
    {
      "epoch": 2.9,
      "grad_norm": 6.383190307410641,
      "learning_rate": 1.101544983769779e-05,
      "loss": 0.1734,
      "step": 801
    },
    {
      "epoch": 2.91,
      "grad_norm": 4.9063394108919445,
      "learning_rate": 1.0995987424083178e-05,
      "loss": 0.1627,
      "step": 802
    },
    {
      "epoch": 2.91,
      "grad_norm": 2.657197763917094,
      "learning_rate": 1.0976521199262945e-05,
      "loss": 0.1815,
      "step": 803
    },
    {
      "epoch": 2.91,
      "grad_norm": 2.2684003765940095,
      "learning_rate": 1.0957051237725775e-05,
      "loss": 0.1371,
      "step": 804
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.2154914226854294,
      "learning_rate": 1.0937577613974641e-05,
      "loss": 0.1858,
      "step": 805
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.7328927887784134,
      "learning_rate": 1.0918100402526533e-05,
      "loss": 0.1304,
      "step": 806
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.5938542193737084,
      "learning_rate": 1.0898619677912165e-05,
      "loss": 0.13,
      "step": 807
    },
    {
      "epoch": 2.93,
      "grad_norm": 3.4966470407927943,
      "learning_rate": 1.0879135514675706e-05,
      "loss": 0.1625,
      "step": 808
    },
    {
      "epoch": 2.93,
      "grad_norm": 3.7617991298221796,
      "learning_rate": 1.0859647987374467e-05,
      "loss": 0.1358,
      "step": 809
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.741516706211674,
      "learning_rate": 1.0840157170578645e-05,
      "loss": 0.1189,
      "step": 810
    },
    {
      "epoch": 2.94,
      "grad_norm": 2.5475441861052572,
      "learning_rate": 1.0820663138871018e-05,
      "loss": 0.1573,
      "step": 811
    },
    {
      "epoch": 2.94,
      "grad_norm": 2.4128398950709427,
      "learning_rate": 1.0801165966846662e-05,
      "loss": 0.1067,
      "step": 812
    },
    {
      "epoch": 2.95,
      "grad_norm": 2.5320429584826663,
      "learning_rate": 1.0781665729112688e-05,
      "loss": 0.1459,
      "step": 813
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.493873281090723,
      "learning_rate": 1.0762162500287916e-05,
      "loss": 0.1191,
      "step": 814
    },
    {
      "epoch": 2.95,
      "grad_norm": 3.952044884083061,
      "learning_rate": 1.0742656355002622e-05,
      "loss": 0.1048,
      "step": 815
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.5201057122988098,
      "learning_rate": 1.0723147367898243e-05,
      "loss": 0.1158,
      "step": 816
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.090876232653966,
      "learning_rate": 1.0703635613627084e-05,
      "loss": 0.1115,
      "step": 817
    },
    {
      "epoch": 2.96,
      "grad_norm": 3.7294381501440776,
      "learning_rate": 1.068412116685205e-05,
      "loss": 0.1681,
      "step": 818
    },
    {
      "epoch": 2.97,
      "grad_norm": 3.844594828018935,
      "learning_rate": 1.0664604102246337e-05,
      "loss": 0.1511,
      "step": 819
    },
    {
      "epoch": 2.97,
      "grad_norm": 2.043390310904499,
      "learning_rate": 1.0645084494493166e-05,
      "loss": 0.125,
      "step": 820
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.6488372225760455,
      "learning_rate": 1.0625562418285483e-05,
      "loss": 0.0856,
      "step": 821
    },
    {
      "epoch": 2.98,
      "grad_norm": 2.204496057082104,
      "learning_rate": 1.0606037948325686e-05,
      "loss": 0.1177,
      "step": 822
    },
    {
      "epoch": 2.98,
      "grad_norm": 3.018303589734971,
      "learning_rate": 1.0586511159325332e-05,
      "loss": 0.1294,
      "step": 823
    },
    {
      "epoch": 2.99,
      "grad_norm": 3.9483671709663186,
      "learning_rate": 1.0566982126004848e-05,
      "loss": 0.1558,
      "step": 824
    },
    {
      "epoch": 2.99,
      "grad_norm": 4.049023207699091,
      "learning_rate": 1.0547450923093248e-05,
      "loss": 0.1277,
      "step": 825
    },
    {
      "epoch": 2.99,
      "grad_norm": 2.6010779766826935,
      "learning_rate": 1.052791762532786e-05,
      "loss": 0.1483,
      "step": 826
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.2969099287950385,
      "learning_rate": 1.0508382307454013e-05,
      "loss": 0.1291,
      "step": 827
    },
    {
      "epoch": 3.0,
      "grad_norm": 7.521016410890929,
      "learning_rate": 1.0488845044224774e-05,
      "loss": 0.1446,
      "step": 828
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.7746271335369643,
      "learning_rate": 1.0469305910400651e-05,
      "loss": 0.1152,
      "step": 829
    },
    {
      "epoch": 3.01,
      "grad_norm": 2.503472911411605,
      "learning_rate": 1.0449764980749317e-05,
      "loss": 0.1209,
      "step": 830
    },
    {
      "epoch": 3.01,
      "grad_norm": 4.989852384432169,
      "learning_rate": 1.0430222330045306e-05,
      "loss": 0.1118,
      "step": 831
    },
    {
      "epoch": 3.01,
      "grad_norm": 5.101365842693291,
      "learning_rate": 1.0410678033069745e-05,
      "loss": 0.1537,
      "step": 832
    },
    {
      "epoch": 3.02,
      "grad_norm": 2.1016500861413885,
      "learning_rate": 1.039113216461006e-05,
      "loss": 0.0934,
      "step": 833
    },
    {
      "epoch": 3.02,
      "grad_norm": 3.769379114344539,
      "learning_rate": 1.0371584799459684e-05,
      "loss": 0.1566,
      "step": 834
    },
    {
      "epoch": 3.03,
      "grad_norm": 2.422947810670274,
      "learning_rate": 1.0352036012417788e-05,
      "loss": 0.1568,
      "step": 835
    },
    {
      "epoch": 3.03,
      "grad_norm": 2.627030152793082,
      "learning_rate": 1.0332485878288977e-05,
      "loss": 0.1504,
      "step": 836
    },
    {
      "epoch": 3.03,
      "grad_norm": 2.0242132903425296,
      "learning_rate": 1.0312934471883007e-05,
      "loss": 0.1321,
      "step": 837
    },
    {
      "epoch": 3.04,
      "grad_norm": 2.706534950512403,
      "learning_rate": 1.029338186801451e-05,
      "loss": 0.1275,
      "step": 838
    },
    {
      "epoch": 3.04,
      "grad_norm": 2.480492841544911,
      "learning_rate": 1.0273828141502701e-05,
      "loss": 0.1446,
      "step": 839
    },
    {
      "epoch": 3.04,
      "grad_norm": 2.796571352376509,
      "learning_rate": 1.0254273367171085e-05,
      "loss": 0.1145,
      "step": 840
    },
    {
      "epoch": 3.05,
      "grad_norm": 2.4558515055056893,
      "learning_rate": 1.0234717619847181e-05,
      "loss": 0.1289,
      "step": 841
    },
    {
      "epoch": 3.05,
      "grad_norm": 3.0435993390412546,
      "learning_rate": 1.0215160974362224e-05,
      "loss": 0.1194,
      "step": 842
    },
    {
      "epoch": 3.05,
      "grad_norm": 2.0750968956245606,
      "learning_rate": 1.0195603505550892e-05,
      "loss": 0.137,
      "step": 843
    },
    {
      "epoch": 3.06,
      "grad_norm": 2.251552421329542,
      "learning_rate": 1.0176045288251014e-05,
      "loss": 0.0879,
      "step": 844
    },
    {
      "epoch": 3.06,
      "grad_norm": 2.7520051243350743,
      "learning_rate": 1.0156486397303284e-05,
      "loss": 0.1469,
      "step": 845
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.5364115076968212,
      "learning_rate": 1.0136926907550968e-05,
      "loss": 0.0941,
      "step": 846
    },
    {
      "epoch": 3.07,
      "grad_norm": 2.9525892000929086,
      "learning_rate": 1.0117366893839626e-05,
      "loss": 0.1302,
      "step": 847
    },
    {
      "epoch": 3.07,
      "grad_norm": 2.8403962261072326,
      "learning_rate": 1.0097806431016825e-05,
      "loss": 0.1463,
      "step": 848
    },
    {
      "epoch": 3.08,
      "grad_norm": 2.9328411197586712,
      "learning_rate": 1.0078245593931853e-05,
      "loss": 0.1399,
      "step": 849
    },
    {
      "epoch": 3.08,
      "grad_norm": 2.1509439724109476,
      "learning_rate": 1.0058684457435419e-05,
      "loss": 0.1191,
      "step": 850
    },
    {
      "epoch": 3.08,
      "grad_norm": 4.226341686273808,
      "learning_rate": 1.0039123096379388e-05,
      "loss": 0.1727,
      "step": 851
    },
    {
      "epoch": 3.09,
      "grad_norm": 2.5346322318450034,
      "learning_rate": 1.0019561585616485e-05,
      "loss": 0.1087,
      "step": 852
    },
    {
      "epoch": 3.09,
      "grad_norm": 2.285971644224022,
      "learning_rate": 1e-05,
      "loss": 0.1271,
      "step": 853
    },
    {
      "epoch": 3.09,
      "grad_norm": 3.762171370216196,
      "learning_rate": 9.980438414383518e-06,
      "loss": 0.1438,
      "step": 854
    },
    {
      "epoch": 3.1,
      "grad_norm": 2.592486428756341,
      "learning_rate": 9.960876903620614e-06,
      "loss": 0.1404,
      "step": 855
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.7624009252093233,
      "learning_rate": 9.941315542564583e-06,
      "loss": 0.1122,
      "step": 856
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.9192938052793418,
      "learning_rate": 9.92175440606815e-06,
      "loss": 0.1297,
      "step": 857
    },
    {
      "epoch": 3.11,
      "grad_norm": 2.425000066412495,
      "learning_rate": 9.902193568983177e-06,
      "loss": 0.1192,
      "step": 858
    },
    {
      "epoch": 3.11,
      "grad_norm": 3.5557215609232777,
      "learning_rate": 9.882633106160377e-06,
      "loss": 0.1262,
      "step": 859
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.6285117231596429,
      "learning_rate": 9.863073092449033e-06,
      "loss": 0.0993,
      "step": 860
    },
    {
      "epoch": 3.12,
      "grad_norm": 2.9836089506100527,
      "learning_rate": 9.843513602696717e-06,
      "loss": 0.0971,
      "step": 861
    },
    {
      "epoch": 3.12,
      "grad_norm": 3.4998856846183584,
      "learning_rate": 9.823954711748987e-06,
      "loss": 0.1138,
      "step": 862
    },
    {
      "epoch": 3.13,
      "grad_norm": 2.6057450722798823,
      "learning_rate": 9.80439649444911e-06,
      "loss": 0.159,
      "step": 863
    },
    {
      "epoch": 3.13,
      "grad_norm": 2.7988449670156896,
      "learning_rate": 9.78483902563778e-06,
      "loss": 0.1232,
      "step": 864
    },
    {
      "epoch": 3.13,
      "grad_norm": 2.159254429419325,
      "learning_rate": 9.76528238015282e-06,
      "loss": 0.1404,
      "step": 865
    },
    {
      "epoch": 3.14,
      "grad_norm": 2.72441087365472,
      "learning_rate": 9.745726632828913e-06,
      "loss": 0.1157,
      "step": 866
    },
    {
      "epoch": 3.14,
      "grad_norm": 5.251457256016179,
      "learning_rate": 9.726171858497297e-06,
      "loss": 0.1316,
      "step": 867
    },
    {
      "epoch": 3.14,
      "grad_norm": 3.358081729074561,
      "learning_rate": 9.706618131985489e-06,
      "loss": 0.1528,
      "step": 868
    },
    {
      "epoch": 3.15,
      "grad_norm": 3.739333861254032,
      "learning_rate": 9.687065528116996e-06,
      "loss": 0.1427,
      "step": 869
    },
    {
      "epoch": 3.15,
      "grad_norm": 3.3341234083630154,
      "learning_rate": 9.667514121711025e-06,
      "loss": 0.1434,
      "step": 870
    },
    {
      "epoch": 3.16,
      "grad_norm": 3.9910243031387322,
      "learning_rate": 9.647963987582212e-06,
      "loss": 0.1734,
      "step": 871
    },
    {
      "epoch": 3.16,
      "grad_norm": 2.4751044620978386,
      "learning_rate": 9.628415200540317e-06,
      "loss": 0.1638,
      "step": 872
    },
    {
      "epoch": 3.16,
      "grad_norm": 3.2151625865660622,
      "learning_rate": 9.608867835389943e-06,
      "loss": 0.0991,
      "step": 873
    },
    {
      "epoch": 3.17,
      "grad_norm": 2.050505065138004,
      "learning_rate": 9.589321966930255e-06,
      "loss": 0.1425,
      "step": 874
    },
    {
      "epoch": 3.17,
      "grad_norm": 5.714400327262755,
      "learning_rate": 9.569777669954694e-06,
      "loss": 0.1296,
      "step": 875
    },
    {
      "epoch": 3.17,
      "grad_norm": 4.068545396635205,
      "learning_rate": 9.550235019250688e-06,
      "loss": 0.1246,
      "step": 876
    },
    {
      "epoch": 3.18,
      "grad_norm": 4.985889593405149,
      "learning_rate": 9.530694089599352e-06,
      "loss": 0.1523,
      "step": 877
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.6430696320161076,
      "learning_rate": 9.51115495577523e-06,
      "loss": 0.1001,
      "step": 878
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.954482721811444,
      "learning_rate": 9.491617692545992e-06,
      "loss": 0.1406,
      "step": 879
    },
    {
      "epoch": 3.19,
      "grad_norm": 2.7044752892563295,
      "learning_rate": 9.472082374672145e-06,
      "loss": 0.1586,
      "step": 880
    },
    {
      "epoch": 3.19,
      "grad_norm": 2.231076930402335,
      "learning_rate": 9.452549076906755e-06,
      "loss": 0.1238,
      "step": 881
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.9953863011580084,
      "learning_rate": 9.433017873995159e-06,
      "loss": 0.1152,
      "step": 882
    },
    {
      "epoch": 3.2,
      "grad_norm": 3.19721787254886,
      "learning_rate": 9.413488840674674e-06,
      "loss": 0.1232,
      "step": 883
    },
    {
      "epoch": 3.2,
      "grad_norm": 2.556877803087002,
      "learning_rate": 9.393962051674319e-06,
      "loss": 0.1219,
      "step": 884
    },
    {
      "epoch": 3.21,
      "grad_norm": 3.087032793077124,
      "learning_rate": 9.374437581714523e-06,
      "loss": 0.121,
      "step": 885
    },
    {
      "epoch": 3.21,
      "grad_norm": 4.557098944942871,
      "learning_rate": 9.354915505506839e-06,
      "loss": 0.1115,
      "step": 886
    },
    {
      "epoch": 3.21,
      "grad_norm": 4.378434203901308,
      "learning_rate": 9.335395897753666e-06,
      "loss": 0.1437,
      "step": 887
    },
    {
      "epoch": 3.22,
      "grad_norm": 4.066606657793286,
      "learning_rate": 9.315878833147953e-06,
      "loss": 0.1152,
      "step": 888
    },
    {
      "epoch": 3.22,
      "grad_norm": 2.3245834907089438,
      "learning_rate": 9.296364386372918e-06,
      "loss": 0.1084,
      "step": 889
    },
    {
      "epoch": 3.22,
      "grad_norm": 3.170010357615001,
      "learning_rate": 9.27685263210176e-06,
      "loss": 0.148,
      "step": 890
    },
    {
      "epoch": 3.23,
      "grad_norm": 2.9743249642039773,
      "learning_rate": 9.257343644997381e-06,
      "loss": 0.1657,
      "step": 891
    },
    {
      "epoch": 3.23,
      "grad_norm": 5.4428836550892346,
      "learning_rate": 9.237837499712088e-06,
      "loss": 0.1298,
      "step": 892
    },
    {
      "epoch": 3.24,
      "grad_norm": 3.1902337963099145,
      "learning_rate": 9.218334270887316e-06,
      "loss": 0.1243,
      "step": 893
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.6894082283835923,
      "learning_rate": 9.19883403315334e-06,
      "loss": 0.0869,
      "step": 894
    },
    {
      "epoch": 3.24,
      "grad_norm": 3.9967215576278963,
      "learning_rate": 9.179336861128987e-06,
      "loss": 0.1099,
      "step": 895
    },
    {
      "epoch": 3.25,
      "grad_norm": 2.494250107497,
      "learning_rate": 9.159842829421358e-06,
      "loss": 0.0927,
      "step": 896
    },
    {
      "epoch": 3.25,
      "grad_norm": 2.220453898319641,
      "learning_rate": 9.140352012625538e-06,
      "loss": 0.0954,
      "step": 897
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.8585494898212889,
      "learning_rate": 9.1208644853243e-06,
      "loss": 0.0819,
      "step": 898
    },
    {
      "epoch": 3.26,
      "grad_norm": 2.2204249039070514,
      "learning_rate": 9.101380322087837e-06,
      "loss": 0.0749,
      "step": 899
    },
    {
      "epoch": 3.26,
      "grad_norm": 4.291342318127167,
      "learning_rate": 9.081899597473469e-06,
      "loss": 0.1208,
      "step": 900
    },
    {
      "epoch": 3.26,
      "grad_norm": 2.2154186516648853,
      "learning_rate": 9.06242238602536e-06,
      "loss": 0.1011,
      "step": 901
    },
    {
      "epoch": 3.27,
      "grad_norm": 2.8799603623281826,
      "learning_rate": 9.042948762274227e-06,
      "loss": 0.1121,
      "step": 902
    },
    {
      "epoch": 3.27,
      "grad_norm": 3.141717635263762,
      "learning_rate": 9.023478800737056e-06,
      "loss": 0.1083,
      "step": 903
    },
    {
      "epoch": 3.28,
      "grad_norm": 2.8267039279104207,
      "learning_rate": 9.004012575916825e-06,
      "loss": 0.0846,
      "step": 904
    },
    {
      "epoch": 3.28,
      "grad_norm": 3.7032607732393883,
      "learning_rate": 8.984550162302211e-06,
      "loss": 0.1043,
      "step": 905
    },
    {
      "epoch": 3.28,
      "grad_norm": 2.9401418991396793,
      "learning_rate": 8.965091634367306e-06,
      "loss": 0.0996,
      "step": 906
    },
    {
      "epoch": 3.29,
      "grad_norm": 3.270486472986497,
      "learning_rate": 8.945637066571337e-06,
      "loss": 0.0688,
      "step": 907
    },
    {
      "epoch": 3.29,
      "grad_norm": 4.689647025961772,
      "learning_rate": 8.92618653335837e-06,
      "loss": 0.0861,
      "step": 908
    },
    {
      "epoch": 3.29,
      "grad_norm": 2.759321778364716,
      "learning_rate": 8.90674010915704e-06,
      "loss": 0.0763,
      "step": 909
    },
    {
      "epoch": 3.3,
      "grad_norm": 3.623678804797402,
      "learning_rate": 8.887297868380255e-06,
      "loss": 0.1029,
      "step": 910
    },
    {
      "epoch": 3.3,
      "grad_norm": 3.222924533432378,
      "learning_rate": 8.867859885424915e-06,
      "loss": 0.0955,
      "step": 911
    },
    {
      "epoch": 3.3,
      "grad_norm": 5.9486353523383375,
      "learning_rate": 8.84842623467163e-06,
      "loss": 0.0838,
      "step": 912
    },
    {
      "epoch": 3.31,
      "grad_norm": 3.5182081825185385,
      "learning_rate": 8.828996990484425e-06,
      "loss": 0.1039,
      "step": 913
    },
    {
      "epoch": 3.31,
      "grad_norm": 4.536580119862204,
      "learning_rate": 8.809572227210472e-06,
      "loss": 0.0969,
      "step": 914
    },
    {
      "epoch": 3.32,
      "grad_norm": 2.277344712508898,
      "learning_rate": 8.790152019179794e-06,
      "loss": 0.0895,
      "step": 915
    },
    {
      "epoch": 3.32,
      "grad_norm": 5.668867235184441,
      "learning_rate": 8.770736440704979e-06,
      "loss": 0.1596,
      "step": 916
    },
    {
      "epoch": 3.32,
      "grad_norm": 3.6328993424164415,
      "learning_rate": 8.751325566080905e-06,
      "loss": 0.1252,
      "step": 917
    },
    {
      "epoch": 3.33,
      "grad_norm": 3.1744558717883855,
      "learning_rate": 8.731919469584443e-06,
      "loss": 0.0791,
      "step": 918
    },
    {
      "epoch": 3.33,
      "grad_norm": 3.7579305119348754,
      "learning_rate": 8.712518225474191e-06,
      "loss": 0.1331,
      "step": 919
    },
    {
      "epoch": 3.33,
      "grad_norm": 4.125964433544078,
      "learning_rate": 8.693121907990177e-06,
      "loss": 0.1353,
      "step": 920
    },
    {
      "epoch": 3.34,
      "grad_norm": 3.370157612596519,
      "learning_rate": 8.673730591353559e-06,
      "loss": 0.1312,
      "step": 921
    },
    {
      "epoch": 3.34,
      "grad_norm": 2.8376369760631097,
      "learning_rate": 8.654344349766384e-06,
      "loss": 0.0966,
      "step": 922
    },
    {
      "epoch": 3.34,
      "grad_norm": 3.8923297715397966,
      "learning_rate": 8.634963257411266e-06,
      "loss": 0.14,
      "step": 923
    },
    {
      "epoch": 3.35,
      "grad_norm": 2.778872787089105,
      "learning_rate": 8.615587388451116e-06,
      "loss": 0.153,
      "step": 924
    },
    {
      "epoch": 3.35,
      "grad_norm": 3.6919755017122267,
      "learning_rate": 8.59621681702886e-06,
      "loss": 0.1919,
      "step": 925
    },
    {
      "epoch": 3.36,
      "grad_norm": 2.5356430790154016,
      "learning_rate": 8.576851617267151e-06,
      "loss": 0.1215,
      "step": 926
    },
    {
      "epoch": 3.36,
      "grad_norm": 3.352088530955654,
      "learning_rate": 8.557491863268088e-06,
      "loss": 0.1371,
      "step": 927
    },
    {
      "epoch": 3.36,
      "grad_norm": 2.026956267831763,
      "learning_rate": 8.53813762911293e-06,
      "loss": 0.1086,
      "step": 928
    },
    {
      "epoch": 3.37,
      "grad_norm": 3.087795837752131,
      "learning_rate": 8.518788988861815e-06,
      "loss": 0.1163,
      "step": 929
    },
    {
      "epoch": 3.37,
      "grad_norm": 2.734278815496621,
      "learning_rate": 8.499446016553475e-06,
      "loss": 0.1262,
      "step": 930
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.4827892129376603,
      "learning_rate": 8.480108786204956e-06,
      "loss": 0.1029,
      "step": 931
    },
    {
      "epoch": 3.38,
      "grad_norm": 3.7153729607054373,
      "learning_rate": 8.460777371811327e-06,
      "loss": 0.1248,
      "step": 932
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.9463462730942114,
      "learning_rate": 8.441451847345407e-06,
      "loss": 0.1276,
      "step": 933
    },
    {
      "epoch": 3.38,
      "grad_norm": 2.39746437004284,
      "learning_rate": 8.42213228675747e-06,
      "loss": 0.104,
      "step": 934
    },
    {
      "epoch": 3.39,
      "grad_norm": 2.9019853243357323,
      "learning_rate": 8.402818763974978e-06,
      "loss": 0.1404,
      "step": 935
    },
    {
      "epoch": 3.39,
      "grad_norm": 3.316510882049375,
      "learning_rate": 8.383511352902285e-06,
      "loss": 0.123,
      "step": 936
    },
    {
      "epoch": 3.39,
      "grad_norm": 4.45987374379237,
      "learning_rate": 8.364210127420353e-06,
      "loss": 0.1448,
      "step": 937
    },
    {
      "epoch": 3.4,
      "grad_norm": 3.3388940872208988,
      "learning_rate": 8.344915161386485e-06,
      "loss": 0.0981,
      "step": 938
    },
    {
      "epoch": 3.4,
      "grad_norm": 3.507666488522441,
      "learning_rate": 8.325626528634016e-06,
      "loss": 0.1128,
      "step": 939
    },
    {
      "epoch": 3.41,
      "grad_norm": 2.5575931258105378,
      "learning_rate": 8.306344302972066e-06,
      "loss": 0.1281,
      "step": 940
    },
    {
      "epoch": 3.41,
      "grad_norm": 2.4145418113500163,
      "learning_rate": 8.287068558185225e-06,
      "loss": 0.1386,
      "step": 941
    },
    {
      "epoch": 3.41,
      "grad_norm": 2.1059955109379964,
      "learning_rate": 8.267799368033288e-06,
      "loss": 0.0773,
      "step": 942
    },
    {
      "epoch": 3.42,
      "grad_norm": 2.9284128398528977,
      "learning_rate": 8.248536806250958e-06,
      "loss": 0.125,
      "step": 943
    },
    {
      "epoch": 3.42,
      "grad_norm": 3.798647088081658,
      "learning_rate": 8.229280946547595e-06,
      "loss": 0.0952,
      "step": 944
    },
    {
      "epoch": 3.42,
      "grad_norm": 3.7971420986887936,
      "learning_rate": 8.210031862606896e-06,
      "loss": 0.1174,
      "step": 945
    },
    {
      "epoch": 3.43,
      "grad_norm": 4.4350462780310975,
      "learning_rate": 8.190789628086632e-06,
      "loss": 0.1513,
      "step": 946
    },
    {
      "epoch": 3.43,
      "grad_norm": 3.0237165584661247,
      "learning_rate": 8.171554316618374e-06,
      "loss": 0.1024,
      "step": 947
    },
    {
      "epoch": 3.43,
      "grad_norm": 3.7262130451250446,
      "learning_rate": 8.15232600180719e-06,
      "loss": 0.1436,
      "step": 948
    },
    {
      "epoch": 3.44,
      "grad_norm": 2.8803876073941224,
      "learning_rate": 8.133104757231381e-06,
      "loss": 0.1137,
      "step": 949
    },
    {
      "epoch": 3.44,
      "grad_norm": 2.4558420470179225,
      "learning_rate": 8.113890656442194e-06,
      "loss": 0.1126,
      "step": 950
    },
    {
      "epoch": 3.45,
      "grad_norm": 2.392366218086255,
      "learning_rate": 8.094683772963537e-06,
      "loss": 0.1201,
      "step": 951
    },
    {
      "epoch": 3.45,
      "grad_norm": 2.9831404455847785,
      "learning_rate": 8.075484180291702e-06,
      "loss": 0.1063,
      "step": 952
    },
    {
      "epoch": 3.45,
      "grad_norm": 2.826548310448461,
      "learning_rate": 8.056291951895079e-06,
      "loss": 0.1187,
      "step": 953
    },
    {
      "epoch": 3.46,
      "grad_norm": 4.27161161174605,
      "learning_rate": 8.037107161213886e-06,
      "loss": 0.1192,
      "step": 954
    },
    {
      "epoch": 3.46,
      "grad_norm": 4.976367947641533,
      "learning_rate": 8.017929881659874e-06,
      "loss": 0.1512,
      "step": 955
    },
    {
      "epoch": 3.46,
      "grad_norm": 3.039712986640314,
      "learning_rate": 7.99876018661605e-06,
      "loss": 0.1401,
      "step": 956
    },
    {
      "epoch": 3.47,
      "grad_norm": 3.211419500673735,
      "learning_rate": 7.979598149436404e-06,
      "loss": 0.0895,
      "step": 957
    },
    {
      "epoch": 3.47,
      "grad_norm": 3.9885902165400418,
      "learning_rate": 7.960443843445622e-06,
      "loss": 0.1327,
      "step": 958
    },
    {
      "epoch": 3.47,
      "grad_norm": 2.6607525975325776,
      "learning_rate": 7.941297341938803e-06,
      "loss": 0.1471,
      "step": 959
    },
    {
      "epoch": 3.48,
      "grad_norm": 2.107695072062342,
      "learning_rate": 7.922158718181184e-06,
      "loss": 0.1243,
      "step": 960
    },
    {
      "epoch": 3.48,
      "grad_norm": 3.3554052153258462,
      "learning_rate": 7.903028045407858e-06,
      "loss": 0.1187,
      "step": 961
    },
    {
      "epoch": 3.49,
      "grad_norm": 3.057712546839428,
      "learning_rate": 7.883905396823487e-06,
      "loss": 0.1382,
      "step": 962
    },
    {
      "epoch": 3.49,
      "grad_norm": 5.305491453527781,
      "learning_rate": 7.86479084560204e-06,
      "loss": 0.1308,
      "step": 963
    },
    {
      "epoch": 3.49,
      "grad_norm": 3.716016528734415,
      "learning_rate": 7.845684464886487e-06,
      "loss": 0.1244,
      "step": 964
    },
    {
      "epoch": 3.5,
      "grad_norm": 2.9596409781438693,
      "learning_rate": 7.826586327788546e-06,
      "loss": 0.1009,
      "step": 965
    },
    {
      "epoch": 3.5,
      "grad_norm": 3.918947251962521,
      "learning_rate": 7.80749650738838e-06,
      "loss": 0.1137,
      "step": 966
    },
    {
      "epoch": 3.5,
      "grad_norm": 2.799796610731959,
      "learning_rate": 7.788415076734332e-06,
      "loss": 0.1044,
      "step": 967
    },
    {
      "epoch": 3.51,
      "grad_norm": 3.109717304524271,
      "learning_rate": 7.769342108842641e-06,
      "loss": 0.1632,
      "step": 968
    },
    {
      "epoch": 3.51,
      "grad_norm": 2.9002460157326264,
      "learning_rate": 7.75027767669716e-06,
      "loss": 0.1258,
      "step": 969
    },
    {
      "epoch": 3.51,
      "grad_norm": 2.8964520967979692,
      "learning_rate": 7.731221853249089e-06,
      "loss": 0.1289,
      "step": 970
    },
    {
      "epoch": 3.52,
      "grad_norm": 3.2852566235883196,
      "learning_rate": 7.712174711416675e-06,
      "loss": 0.1559,
      "step": 971
    },
    {
      "epoch": 3.52,
      "grad_norm": 2.808680050434026,
      "learning_rate": 7.693136324084949e-06,
      "loss": 0.1106,
      "step": 972
    },
    {
      "epoch": 3.53,
      "grad_norm": 2.1171654735689076,
      "learning_rate": 7.674106764105442e-06,
      "loss": 0.0848,
      "step": 973
    },
    {
      "epoch": 3.53,
      "grad_norm": 2.430199608103024,
      "learning_rate": 7.655086104295904e-06,
      "loss": 0.0993,
      "step": 974
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.8828805141022391,
      "learning_rate": 7.636074417440036e-06,
      "loss": 0.0905,
      "step": 975
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.7733655894828702,
      "learning_rate": 7.617071776287196e-06,
      "loss": 0.1263,
      "step": 976
    },
    {
      "epoch": 3.54,
      "grad_norm": 2.7250039443873773,
      "learning_rate": 7.598078253552129e-06,
      "loss": 0.0973,
      "step": 977
    },
    {
      "epoch": 3.54,
      "grad_norm": 2.7007235642921934,
      "learning_rate": 7.5790939219146874e-06,
      "loss": 0.0961,
      "step": 978
    },
    {
      "epoch": 3.55,
      "grad_norm": 2.1500766229328083,
      "learning_rate": 7.5601188540195595e-06,
      "loss": 0.1183,
      "step": 979
    },
    {
      "epoch": 3.55,
      "grad_norm": 3.8280992040940123,
      "learning_rate": 7.541153122475978e-06,
      "loss": 0.1369,
      "step": 980
    },
    {
      "epoch": 3.55,
      "grad_norm": 2.499330376660129,
      "learning_rate": 7.522196799857453e-06,
      "loss": 0.1126,
      "step": 981
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.739850187704323,
      "learning_rate": 7.503249958701489e-06,
      "loss": 0.0658,
      "step": 982
    },
    {
      "epoch": 3.56,
      "grad_norm": 2.7369026119660806,
      "learning_rate": 7.484312671509307e-06,
      "loss": 0.0871,
      "step": 983
    },
    {
      "epoch": 3.57,
      "grad_norm": 2.6596378118313466,
      "learning_rate": 7.46538501074558e-06,
      "loss": 0.0965,
      "step": 984
    },
    {
      "epoch": 3.57,
      "grad_norm": 1.6417651435900409,
      "learning_rate": 7.446467048838131e-06,
      "loss": 0.0663,
      "step": 985
    },
    {
      "epoch": 3.57,
      "grad_norm": 2.775971598647237,
      "learning_rate": 7.427558858177679e-06,
      "loss": 0.1038,
      "step": 986
    },
    {
      "epoch": 3.58,
      "grad_norm": 2.4252559638844713,
      "learning_rate": 7.408660511117547e-06,
      "loss": 0.0925,
      "step": 987
    },
    {
      "epoch": 3.58,
      "grad_norm": 3.503010606223773,
      "learning_rate": 7.389772079973397e-06,
      "loss": 0.1048,
      "step": 988
    },
    {
      "epoch": 3.58,
      "grad_norm": 4.857436489741838,
      "learning_rate": 7.37089363702294e-06,
      "loss": 0.0894,
      "step": 989
    },
    {
      "epoch": 3.59,
      "grad_norm": 2.1916188059776274,
      "learning_rate": 7.352025254505672e-06,
      "loss": 0.0624,
      "step": 990
    },
    {
      "epoch": 3.59,
      "grad_norm": 2.7807935646365083,
      "learning_rate": 7.333167004622593e-06,
      "loss": 0.0847,
      "step": 991
    },
    {
      "epoch": 3.59,
      "grad_norm": 4.040001908804561,
      "learning_rate": 7.31431895953592e-06,
      "loss": 0.1527,
      "step": 992
    },
    {
      "epoch": 3.6,
      "grad_norm": 3.926769476612875,
      "learning_rate": 7.295481191368836e-06,
      "loss": 0.0891,
      "step": 993
    },
    {
      "epoch": 3.6,
      "grad_norm": 3.254662559682786,
      "learning_rate": 7.276653772205187e-06,
      "loss": 0.0833,
      "step": 994
    },
    {
      "epoch": 3.61,
      "grad_norm": 4.705257635176926,
      "learning_rate": 7.257836774089222e-06,
      "loss": 0.1193,
      "step": 995
    },
    {
      "epoch": 3.61,
      "grad_norm": 2.893193298171004,
      "learning_rate": 7.239030269025311e-06,
      "loss": 0.1171,
      "step": 996
    },
    {
      "epoch": 3.61,
      "grad_norm": 2.6770430406514327,
      "learning_rate": 7.2202343289776775e-06,
      "loss": 0.105,
      "step": 997
    },
    {
      "epoch": 3.62,
      "grad_norm": 3.903790164902213,
      "learning_rate": 7.201449025870113e-06,
      "loss": 0.148,
      "step": 998
    },
    {
      "epoch": 3.62,
      "grad_norm": 5.530297464086568,
      "learning_rate": 7.182674431585703e-06,
      "loss": 0.0866,
      "step": 999
    },
    {
      "epoch": 3.62,
      "grad_norm": 4.869988050781526,
      "learning_rate": 7.163910617966563e-06,
      "loss": 0.1006,
      "step": 1000
    },
    {
      "epoch": 3.63,
      "grad_norm": 3.579073892813218,
      "learning_rate": 7.145157656813546e-06,
      "loss": 0.0912,
      "step": 1001
    },
    {
      "epoch": 3.63,
      "grad_norm": 2.35628309330303,
      "learning_rate": 7.126415619885987e-06,
      "loss": 0.0528,
      "step": 1002
    },
    {
      "epoch": 3.63,
      "grad_norm": 4.049861882203114,
      "learning_rate": 7.107684578901414e-06,
      "loss": 0.1446,
      "step": 1003
    },
    {
      "epoch": 3.64,
      "grad_norm": 3.2253320575037296,
      "learning_rate": 7.088964605535278e-06,
      "loss": 0.0971,
      "step": 1004
    },
    {
      "epoch": 3.64,
      "grad_norm": 4.435165553701907,
      "learning_rate": 7.0702557714206786e-06,
      "loss": 0.1013,
      "step": 1005
    },
    {
      "epoch": 3.64,
      "grad_norm": 4.651994900498167,
      "learning_rate": 7.0515581481480925e-06,
      "loss": 0.1181,
      "step": 1006
    },
    {
      "epoch": 3.65,
      "grad_norm": 4.968879604252837,
      "learning_rate": 7.032871807265097e-06,
      "loss": 0.1307,
      "step": 1007
    },
    {
      "epoch": 3.65,
      "grad_norm": 4.606076160778083,
      "learning_rate": 7.014196820276098e-06,
      "loss": 0.1174,
      "step": 1008
    },
    {
      "epoch": 3.66,
      "grad_norm": 3.255916908644234,
      "learning_rate": 6.99553325864205e-06,
      "loss": 0.1248,
      "step": 1009
    },
    {
      "epoch": 3.66,
      "grad_norm": 2.45046540729099,
      "learning_rate": 6.976881193780196e-06,
      "loss": 0.087,
      "step": 1010
    },
    {
      "epoch": 3.66,
      "grad_norm": 2.3217962969736443,
      "learning_rate": 6.958240697063774e-06,
      "loss": 0.071,
      "step": 1011
    },
    {
      "epoch": 3.67,
      "grad_norm": 2.4568876117053033,
      "learning_rate": 6.9396118398217675e-06,
      "loss": 0.0962,
      "step": 1012
    },
    {
      "epoch": 3.67,
      "grad_norm": 2.2280111104466642,
      "learning_rate": 6.920994693338612e-06,
      "loss": 0.1103,
      "step": 1013
    },
    {
      "epoch": 3.67,
      "grad_norm": 3.1209935926030083,
      "learning_rate": 6.90238932885394e-06,
      "loss": 0.1,
      "step": 1014
    },
    {
      "epoch": 3.68,
      "grad_norm": 2.37659138818212,
      "learning_rate": 6.883795817562293e-06,
      "loss": 0.0838,
      "step": 1015
    },
    {
      "epoch": 3.68,
      "grad_norm": 4.292882340244696,
      "learning_rate": 6.865214230612858e-06,
      "loss": 0.1078,
      "step": 1016
    },
    {
      "epoch": 3.68,
      "grad_norm": 3.079448839621652,
      "learning_rate": 6.846644639109186e-06,
      "loss": 0.1048,
      "step": 1017
    },
    {
      "epoch": 3.69,
      "grad_norm": 2.100105736333372,
      "learning_rate": 6.8280871141089415e-06,
      "loss": 0.0917,
      "step": 1018
    },
    {
      "epoch": 3.69,
      "grad_norm": 3.498056667988704,
      "learning_rate": 6.809541726623602e-06,
      "loss": 0.0976,
      "step": 1019
    },
    {
      "epoch": 3.7,
      "grad_norm": 3.102683380380855,
      "learning_rate": 6.791008547618207e-06,
      "loss": 0.1066,
      "step": 1020
    },
    {
      "epoch": 3.7,
      "grad_norm": 2.9892418404379795,
      "learning_rate": 6.772487648011075e-06,
      "loss": 0.1073,
      "step": 1021
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.8806743914425466,
      "learning_rate": 6.753979098673539e-06,
      "loss": 0.1038,
      "step": 1022
    },
    {
      "epoch": 3.71,
      "grad_norm": 3.122278074964517,
      "learning_rate": 6.735482970429676e-06,
      "loss": 0.0774,
      "step": 1023
    },
    {
      "epoch": 3.71,
      "grad_norm": 3.9134486678550173,
      "learning_rate": 6.716999334056031e-06,
      "loss": 0.1493,
      "step": 1024
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 1656,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 512,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}