{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2884,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00034674063800277393,
      "grad_norm": 3.4529595375061035,
      "learning_rate": 0.0,
      "loss": 0.2571,
      "step": 1
    },
    {
      "epoch": 0.0006934812760055479,
      "grad_norm": 1.7922414541244507,
      "learning_rate": 1e-06,
      "loss": 0.2304,
      "step": 2
    },
    {
      "epoch": 0.0010402219140083217,
      "grad_norm": 2.7978198528289795,
      "learning_rate": 1e-06,
      "loss": 0.2421,
      "step": 3
    },
    {
      "epoch": 0.0013869625520110957,
      "grad_norm": 1.7440896034240723,
      "learning_rate": 1e-06,
      "loss": 0.2437,
      "step": 4
    },
    {
      "epoch": 0.0017337031900138697,
      "grad_norm": 2.0253705978393555,
      "learning_rate": 1e-06,
      "loss": 0.2456,
      "step": 5
    },
    {
      "epoch": 0.0020804438280166435,
      "grad_norm": 7.390465259552002,
      "learning_rate": 1e-06,
      "loss": 0.285,
      "step": 6
    },
    {
      "epoch": 0.0024271844660194173,
      "grad_norm": 2.6305766105651855,
      "learning_rate": 1e-06,
      "loss": 0.2477,
      "step": 7
    },
    {
      "epoch": 0.0027739251040221915,
      "grad_norm": 3.094043254852295,
      "learning_rate": 1e-06,
      "loss": 0.2229,
      "step": 8
    },
    {
      "epoch": 0.0031206657420249652,
      "grad_norm": 2.3640098571777344,
      "learning_rate": 1e-06,
      "loss": 0.2233,
      "step": 9
    },
    {
      "epoch": 0.0034674063800277394,
      "grad_norm": 2.151405096054077,
      "learning_rate": 1e-06,
      "loss": 0.2211,
      "step": 10
    },
    {
      "epoch": 0.0038141470180305132,
      "grad_norm": 5.151231288909912,
      "learning_rate": 1e-06,
      "loss": 0.2354,
      "step": 11
    },
    {
      "epoch": 0.004160887656033287,
      "grad_norm": 2.285297155380249,
      "learning_rate": 1e-06,
      "loss": 0.2483,
      "step": 12
    },
    {
      "epoch": 0.004507628294036061,
      "grad_norm": 2.348036050796509,
      "learning_rate": 1e-06,
      "loss": 0.2419,
      "step": 13
    },
    {
      "epoch": 0.0048543689320388345,
      "grad_norm": 2.321516275405884,
      "learning_rate": 1e-06,
      "loss": 0.2659,
      "step": 14
    },
    {
      "epoch": 0.005201109570041609,
      "grad_norm": 2.291966438293457,
      "learning_rate": 1e-06,
      "loss": 0.2315,
      "step": 15
    },
    {
      "epoch": 0.005547850208044383,
      "grad_norm": 2.938875198364258,
      "learning_rate": 1e-06,
      "loss": 0.2479,
      "step": 16
    },
    {
      "epoch": 0.005894590846047157,
      "grad_norm": 1.7737529277801514,
      "learning_rate": 1e-06,
      "loss": 0.2242,
      "step": 17
    },
    {
      "epoch": 0.0062413314840499305,
      "grad_norm": 2.331916332244873,
      "learning_rate": 1e-06,
      "loss": 0.2386,
      "step": 18
    },
    {
      "epoch": 0.006588072122052704,
      "grad_norm": 2.8218319416046143,
      "learning_rate": 1e-06,
      "loss": 0.2655,
      "step": 19
    },
    {
      "epoch": 0.006934812760055479,
      "grad_norm": 1.9200022220611572,
      "learning_rate": 1e-06,
      "loss": 0.2384,
      "step": 20
    },
    {
      "epoch": 0.007281553398058253,
      "grad_norm": 2.6691722869873047,
      "learning_rate": 1e-06,
      "loss": 0.2594,
      "step": 21
    },
    {
      "epoch": 0.0076282940360610264,
      "grad_norm": 3.2480461597442627,
      "learning_rate": 1e-06,
      "loss": 0.1936,
      "step": 22
    },
    {
      "epoch": 0.0079750346740638,
      "grad_norm": 4.234226226806641,
      "learning_rate": 1e-06,
      "loss": 0.2672,
      "step": 23
    },
    {
      "epoch": 0.008321775312066574,
      "grad_norm": 4.796227931976318,
      "learning_rate": 1e-06,
      "loss": 0.2711,
      "step": 24
    },
    {
      "epoch": 0.008668515950069348,
      "grad_norm": 4.371148109436035,
      "learning_rate": 1e-06,
      "loss": 0.2256,
      "step": 25
    },
    {
      "epoch": 0.009015256588072122,
      "grad_norm": 2.540862798690796,
      "learning_rate": 1e-06,
      "loss": 0.2418,
      "step": 26
    },
    {
      "epoch": 0.009361997226074895,
      "grad_norm": 1.8659073114395142,
      "learning_rate": 1e-06,
      "loss": 0.2439,
      "step": 27
    },
    {
      "epoch": 0.009708737864077669,
      "grad_norm": 2.3605165481567383,
      "learning_rate": 1e-06,
      "loss": 0.2177,
      "step": 28
    },
    {
      "epoch": 0.010055478502080445,
      "grad_norm": 2.2104761600494385,
      "learning_rate": 1e-06,
      "loss": 0.2504,
      "step": 29
    },
    {
      "epoch": 0.010402219140083218,
      "grad_norm": 2.1602864265441895,
      "learning_rate": 1e-06,
      "loss": 0.2372,
      "step": 30
    },
    {
      "epoch": 0.010748959778085992,
      "grad_norm": 2.1729562282562256,
      "learning_rate": 1e-06,
      "loss": 0.2562,
      "step": 31
    },
    {
      "epoch": 0.011095700416088766,
      "grad_norm": 2.284585475921631,
      "learning_rate": 1e-06,
      "loss": 0.2355,
      "step": 32
    },
    {
      "epoch": 0.01144244105409154,
      "grad_norm": 2.268206834793091,
      "learning_rate": 1e-06,
      "loss": 0.2454,
      "step": 33
    },
    {
      "epoch": 0.011789181692094313,
      "grad_norm": 2.5299553871154785,
      "learning_rate": 1e-06,
      "loss": 0.2471,
      "step": 34
    },
    {
      "epoch": 0.012135922330097087,
      "grad_norm": 3.3298118114471436,
      "learning_rate": 1e-06,
      "loss": 0.2728,
      "step": 35
    },
    {
      "epoch": 0.012482662968099861,
      "grad_norm": 4.213827133178711,
      "learning_rate": 1e-06,
      "loss": 0.2271,
      "step": 36
    },
    {
      "epoch": 0.012829403606102635,
      "grad_norm": 4.147632122039795,
      "learning_rate": 1e-06,
      "loss": 0.2474,
      "step": 37
    },
    {
      "epoch": 0.013176144244105409,
      "grad_norm": 3.017467737197876,
      "learning_rate": 1e-06,
      "loss": 0.2443,
      "step": 38
    },
    {
      "epoch": 0.013522884882108182,
      "grad_norm": 3.92279314994812,
      "learning_rate": 1e-06,
      "loss": 0.2309,
      "step": 39
    },
    {
      "epoch": 0.013869625520110958,
      "grad_norm": 2.6705775260925293,
      "learning_rate": 1e-06,
      "loss": 0.2122,
      "step": 40
    },
    {
      "epoch": 0.014216366158113732,
      "grad_norm": 2.2579407691955566,
      "learning_rate": 1e-06,
      "loss": 0.2189,
      "step": 41
    },
    {
      "epoch": 0.014563106796116505,
      "grad_norm": 3.3366434574127197,
      "learning_rate": 1e-06,
      "loss": 0.2268,
      "step": 42
    },
    {
      "epoch": 0.014909847434119279,
      "grad_norm": 1.8895633220672607,
      "learning_rate": 1e-06,
      "loss": 0.2219,
      "step": 43
    },
    {
      "epoch": 0.015256588072122053,
      "grad_norm": 4.174144268035889,
      "learning_rate": 1e-06,
      "loss": 0.209,
      "step": 44
    },
    {
      "epoch": 0.015603328710124827,
      "grad_norm": 2.0545594692230225,
      "learning_rate": 1e-06,
      "loss": 0.2464,
      "step": 45
    },
    {
      "epoch": 0.0159500693481276,
      "grad_norm": 6.63257360458374,
      "learning_rate": 1e-06,
      "loss": 0.2905,
      "step": 46
    },
    {
      "epoch": 0.016296809986130374,
      "grad_norm": 3.2842488288879395,
      "learning_rate": 1e-06,
      "loss": 0.2674,
      "step": 47
    },
    {
      "epoch": 0.016643550624133148,
      "grad_norm": 2.0572311878204346,
      "learning_rate": 1e-06,
      "loss": 0.237,
      "step": 48
    },
    {
      "epoch": 0.01699029126213592,
      "grad_norm": 1.9547162055969238,
      "learning_rate": 1e-06,
      "loss": 0.2268,
      "step": 49
    },
    {
      "epoch": 0.017337031900138695,
      "grad_norm": 2.775322914123535,
      "learning_rate": 1e-06,
      "loss": 0.2315,
      "step": 50
    },
    {
      "epoch": 0.01768377253814147,
      "grad_norm": 1.8670415878295898,
      "learning_rate": 1e-06,
      "loss": 0.21,
      "step": 51
    },
    {
      "epoch": 0.018030513176144243,
      "grad_norm": 2.3691036701202393,
      "learning_rate": 1e-06,
      "loss": 0.2465,
      "step": 52
    },
    {
      "epoch": 0.018377253814147017,
      "grad_norm": 4.806734561920166,
      "learning_rate": 1e-06,
      "loss": 0.2437,
      "step": 53
    },
    {
      "epoch": 0.01872399445214979,
      "grad_norm": 4.114093780517578,
      "learning_rate": 1e-06,
      "loss": 0.2402,
      "step": 54
    },
    {
      "epoch": 0.019070735090152564,
      "grad_norm": 2.5198612213134766,
      "learning_rate": 1e-06,
      "loss": 0.2409,
      "step": 55
    },
    {
      "epoch": 0.019417475728155338,
      "grad_norm": 2.066634178161621,
      "learning_rate": 1e-06,
      "loss": 0.2435,
      "step": 56
    },
    {
      "epoch": 0.019764216366158115,
      "grad_norm": 2.1365854740142822,
      "learning_rate": 1e-06,
      "loss": 0.2321,
      "step": 57
    },
    {
      "epoch": 0.02011095700416089,
      "grad_norm": 3.2070484161376953,
      "learning_rate": 1e-06,
      "loss": 0.2347,
      "step": 58
    },
    {
      "epoch": 0.020457697642163663,
      "grad_norm": 2.717198133468628,
      "learning_rate": 1e-06,
      "loss": 0.1966,
      "step": 59
    },
    {
      "epoch": 0.020804438280166437,
      "grad_norm": 3.56074857711792,
      "learning_rate": 1e-06,
      "loss": 0.2505,
      "step": 60
    },
    {
      "epoch": 0.02115117891816921,
      "grad_norm": 3.655073642730713,
      "learning_rate": 1e-06,
      "loss": 0.2467,
      "step": 61
    },
    {
      "epoch": 0.021497919556171984,
      "grad_norm": 3.183065414428711,
      "learning_rate": 1e-06,
      "loss": 0.2201,
      "step": 62
    },
    {
      "epoch": 0.021844660194174758,
      "grad_norm": 3.160886764526367,
      "learning_rate": 1e-06,
      "loss": 0.2187,
      "step": 63
    },
    {
      "epoch": 0.022191400832177532,
      "grad_norm": 1.6657497882843018,
      "learning_rate": 1e-06,
      "loss": 0.2096,
      "step": 64
    },
    {
      "epoch": 0.022538141470180306,
      "grad_norm": 1.8679834604263306,
      "learning_rate": 1e-06,
      "loss": 0.2478,
      "step": 65
    },
    {
      "epoch": 0.02288488210818308,
      "grad_norm": 1.8762993812561035,
      "learning_rate": 1e-06,
      "loss": 0.218,
      "step": 66
    },
    {
      "epoch": 0.023231622746185853,
      "grad_norm": 2.1899123191833496,
      "learning_rate": 1e-06,
      "loss": 0.2422,
      "step": 67
    },
    {
      "epoch": 0.023578363384188627,
      "grad_norm": 2.1151392459869385,
      "learning_rate": 1e-06,
      "loss": 0.267,
      "step": 68
    },
    {
      "epoch": 0.0239251040221914,
      "grad_norm": 3.3461737632751465,
      "learning_rate": 1e-06,
      "loss": 0.2615,
      "step": 69
    },
    {
      "epoch": 0.024271844660194174,
      "grad_norm": 1.9115928411483765,
      "learning_rate": 1e-06,
      "loss": 0.2385,
      "step": 70
    },
    {
      "epoch": 0.024618585298196948,
      "grad_norm": 1.9966187477111816,
      "learning_rate": 1e-06,
      "loss": 0.2341,
      "step": 71
    },
    {
      "epoch": 0.024965325936199722,
      "grad_norm": 4.439002990722656,
      "learning_rate": 1e-06,
      "loss": 0.2683,
      "step": 72
    },
    {
      "epoch": 0.025312066574202496,
      "grad_norm": 3.926988124847412,
      "learning_rate": 1e-06,
      "loss": 0.2489,
      "step": 73
    },
    {
      "epoch": 0.02565880721220527,
      "grad_norm": 2.2309978008270264,
      "learning_rate": 1e-06,
      "loss": 0.2471,
      "step": 74
    },
    {
      "epoch": 0.026005547850208043,
      "grad_norm": 3.369230031967163,
      "learning_rate": 1e-06,
      "loss": 0.2282,
      "step": 75
    },
    {
      "epoch": 0.026352288488210817,
      "grad_norm": 4.4944047927856445,
      "learning_rate": 1e-06,
      "loss": 0.227,
      "step": 76
    },
    {
      "epoch": 0.02669902912621359,
      "grad_norm": 5.281230449676514,
      "learning_rate": 1e-06,
      "loss": 0.2358,
      "step": 77
    },
    {
      "epoch": 0.027045769764216365,
      "grad_norm": 2.726041555404663,
      "learning_rate": 1e-06,
      "loss": 0.2569,
      "step": 78
    },
    {
      "epoch": 0.027392510402219142,
      "grad_norm": 2.2781665325164795,
      "learning_rate": 1e-06,
      "loss": 0.2436,
      "step": 79
    },
    {
      "epoch": 0.027739251040221916,
      "grad_norm": 5.479579925537109,
      "learning_rate": 1e-06,
      "loss": 0.2588,
      "step": 80
    },
    {
      "epoch": 0.02808599167822469,
      "grad_norm": 5.336531162261963,
      "learning_rate": 1e-06,
      "loss": 0.2315,
      "step": 81
    },
    {
      "epoch": 0.028432732316227463,
      "grad_norm": 1.9904639720916748,
      "learning_rate": 1e-06,
      "loss": 0.249,
      "step": 82
    },
    {
      "epoch": 0.028779472954230237,
      "grad_norm": 2.493849277496338,
      "learning_rate": 1e-06,
      "loss": 0.2586,
      "step": 83
    },
    {
      "epoch": 0.02912621359223301,
      "grad_norm": 2.0761678218841553,
      "learning_rate": 1e-06,
      "loss": 0.2256,
      "step": 84
    },
    {
      "epoch": 0.029472954230235784,
      "grad_norm": 3.9057183265686035,
      "learning_rate": 1e-06,
      "loss": 0.2605,
      "step": 85
    },
    {
      "epoch": 0.029819694868238558,
      "grad_norm": 1.8108359575271606,
      "learning_rate": 1e-06,
      "loss": 0.209,
      "step": 86
    },
    {
      "epoch": 0.030166435506241332,
      "grad_norm": 3.9006757736206055,
      "learning_rate": 1e-06,
      "loss": 0.2302,
      "step": 87
    },
    {
      "epoch": 0.030513176144244106,
      "grad_norm": 3.529353141784668,
      "learning_rate": 1e-06,
      "loss": 0.2497,
      "step": 88
    },
    {
      "epoch": 0.03085991678224688,
      "grad_norm": 1.9021555185317993,
      "learning_rate": 1e-06,
      "loss": 0.2343,
      "step": 89
    },
    {
      "epoch": 0.031206657420249653,
      "grad_norm": 2.0818121433258057,
      "learning_rate": 1e-06,
      "loss": 0.2296,
      "step": 90
    },
    {
      "epoch": 0.03155339805825243,
      "grad_norm": 2.9069199562072754,
      "learning_rate": 1e-06,
      "loss": 0.2336,
      "step": 91
    },
    {
      "epoch": 0.0319001386962552,
      "grad_norm": 2.5499114990234375,
      "learning_rate": 1e-06,
      "loss": 0.2326,
      "step": 92
    },
    {
      "epoch": 0.032246879334257975,
      "grad_norm": 4.565133571624756,
      "learning_rate": 1e-06,
      "loss": 0.2325,
      "step": 93
    },
    {
      "epoch": 0.03259361997226075,
      "grad_norm": 1.965366005897522,
      "learning_rate": 1e-06,
      "loss": 0.2436,
      "step": 94
    },
    {
      "epoch": 0.03294036061026352,
      "grad_norm": 3.4335970878601074,
      "learning_rate": 1e-06,
      "loss": 0.2517,
      "step": 95
    },
    {
      "epoch": 0.033287101248266296,
      "grad_norm": 2.7447807788848877,
      "learning_rate": 1e-06,
      "loss": 0.219,
      "step": 96
    },
    {
      "epoch": 0.03363384188626907,
      "grad_norm": 1.9556831121444702,
      "learning_rate": 1e-06,
      "loss": 0.2071,
      "step": 97
    },
    {
      "epoch": 0.03398058252427184,
      "grad_norm": 2.175020694732666,
      "learning_rate": 1e-06,
      "loss": 0.2481,
      "step": 98
    },
    {
      "epoch": 0.03432732316227462,
      "grad_norm": 3.7034220695495605,
      "learning_rate": 1e-06,
      "loss": 0.2525,
      "step": 99
    },
    {
      "epoch": 0.03467406380027739,
      "grad_norm": 4.264119625091553,
      "learning_rate": 1e-06,
      "loss": 0.21,
      "step": 100
    },
    {
      "epoch": 0.035020804438280165,
      "grad_norm": 4.019951820373535,
      "learning_rate": 1e-06,
      "loss": 0.2495,
      "step": 101
    },
    {
      "epoch": 0.03536754507628294,
      "grad_norm": 2.4060921669006348,
      "learning_rate": 1e-06,
      "loss": 0.2295,
      "step": 102
    },
    {
      "epoch": 0.03571428571428571,
      "grad_norm": 2.3502883911132812,
      "learning_rate": 1e-06,
      "loss": 0.232,
      "step": 103
    },
    {
      "epoch": 0.036061026352288486,
      "grad_norm": 4.202142715454102,
      "learning_rate": 1e-06,
      "loss": 0.2159,
      "step": 104
    },
    {
      "epoch": 0.03640776699029126,
      "grad_norm": 6.274311542510986,
      "learning_rate": 1e-06,
      "loss": 0.2825,
      "step": 105
    },
    {
      "epoch": 0.036754507628294034,
      "grad_norm": 2.756082057952881,
      "learning_rate": 1e-06,
      "loss": 0.251,
      "step": 106
    },
    {
      "epoch": 0.03710124826629681,
      "grad_norm": 1.8495687246322632,
      "learning_rate": 1e-06,
      "loss": 0.2371,
      "step": 107
    },
    {
      "epoch": 0.03744798890429958,
      "grad_norm": 2.5674338340759277,
      "learning_rate": 1e-06,
      "loss": 0.2157,
      "step": 108
    },
    {
      "epoch": 0.037794729542302355,
      "grad_norm": 3.304002285003662,
      "learning_rate": 1e-06,
      "loss": 0.2552,
      "step": 109
    },
    {
      "epoch": 0.03814147018030513,
      "grad_norm": 4.199469566345215,
      "learning_rate": 1e-06,
      "loss": 0.2165,
      "step": 110
    },
    {
      "epoch": 0.0384882108183079,
      "grad_norm": 2.1667439937591553,
      "learning_rate": 1e-06,
      "loss": 0.2639,
      "step": 111
    },
    {
      "epoch": 0.038834951456310676,
      "grad_norm": 2.8193678855895996,
      "learning_rate": 1e-06,
      "loss": 0.2174,
      "step": 112
    },
    {
      "epoch": 0.03918169209431346,
      "grad_norm": 1.8696953058242798,
      "learning_rate": 1e-06,
      "loss": 0.2057,
      "step": 113
    },
    {
      "epoch": 0.03952843273231623,
      "grad_norm": 2.3659722805023193,
      "learning_rate": 1e-06,
      "loss": 0.2731,
      "step": 114
    },
    {
      "epoch": 0.039875173370319004,
      "grad_norm": 1.782591700553894,
      "learning_rate": 1e-06,
      "loss": 0.2287,
      "step": 115
    },
    {
      "epoch": 0.04022191400832178,
      "grad_norm": 1.8214987516403198,
      "learning_rate": 1e-06,
      "loss": 0.2225,
      "step": 116
    },
    {
      "epoch": 0.04056865464632455,
      "grad_norm": 1.8799686431884766,
      "learning_rate": 1e-06,
      "loss": 0.2216,
      "step": 117
    },
    {
      "epoch": 0.040915395284327326,
      "grad_norm": 2.0166141986846924,
      "learning_rate": 1e-06,
      "loss": 0.2429,
      "step": 118
    },
    {
      "epoch": 0.0412621359223301,
      "grad_norm": 5.748719215393066,
      "learning_rate": 1e-06,
      "loss": 0.2297,
      "step": 119
    },
    {
      "epoch": 0.04160887656033287,
      "grad_norm": 2.980369806289673,
      "learning_rate": 1e-06,
      "loss": 0.2321,
      "step": 120
    },
    {
      "epoch": 0.04195561719833565,
      "grad_norm": 2.343538999557495,
      "learning_rate": 1e-06,
      "loss": 0.2129,
      "step": 121
    },
    {
      "epoch": 0.04230235783633842,
      "grad_norm": 1.8937276601791382,
      "learning_rate": 1e-06,
      "loss": 0.2297,
      "step": 122
    },
    {
      "epoch": 0.042649098474341195,
      "grad_norm": 7.0609354972839355,
      "learning_rate": 1e-06,
      "loss": 0.2407,
      "step": 123
    },
    {
      "epoch": 0.04299583911234397,
      "grad_norm": 1.7186024188995361,
      "learning_rate": 1e-06,
      "loss": 0.2042,
      "step": 124
    },
    {
      "epoch": 0.04334257975034674,
      "grad_norm": 2.7151777744293213,
      "learning_rate": 1e-06,
      "loss": 0.2323,
      "step": 125
    },
    {
      "epoch": 0.043689320388349516,
      "grad_norm": 2.631126642227173,
      "learning_rate": 1e-06,
      "loss": 0.258,
      "step": 126
    },
    {
      "epoch": 0.04403606102635229,
      "grad_norm": 2.087282657623291,
      "learning_rate": 1e-06,
      "loss": 0.2013,
      "step": 127
    },
    {
      "epoch": 0.044382801664355064,
      "grad_norm": 3.90163254737854,
      "learning_rate": 1e-06,
      "loss": 0.21,
      "step": 128
    },
    {
      "epoch": 0.04472954230235784,
      "grad_norm": 5.360821723937988,
      "learning_rate": 1e-06,
      "loss": 0.2395,
      "step": 129
    },
    {
      "epoch": 0.04507628294036061,
      "grad_norm": 2.5944528579711914,
      "learning_rate": 1e-06,
      "loss": 0.2365,
      "step": 130
    },
    {
      "epoch": 0.045423023578363385,
      "grad_norm": 2.6796722412109375,
      "learning_rate": 1e-06,
      "loss": 0.2361,
      "step": 131
    },
    {
      "epoch": 0.04576976421636616,
      "grad_norm": 2.1970760822296143,
      "learning_rate": 1e-06,
      "loss": 0.2208,
      "step": 132
    },
    {
      "epoch": 0.04611650485436893,
      "grad_norm": 2.5466551780700684,
      "learning_rate": 1e-06,
      "loss": 0.2339,
      "step": 133
    },
    {
      "epoch": 0.046463245492371706,
      "grad_norm": 4.48819637298584,
      "learning_rate": 1e-06,
      "loss": 0.2848,
      "step": 134
    },
    {
      "epoch": 0.04680998613037448,
      "grad_norm": 1.9005647897720337,
      "learning_rate": 1e-06,
      "loss": 0.21,
      "step": 135
    },
    {
      "epoch": 0.047156726768377254,
      "grad_norm": 9.14037036895752,
      "learning_rate": 1e-06,
      "loss": 0.2805,
      "step": 136
    },
    {
      "epoch": 0.04750346740638003,
      "grad_norm": 1.840706467628479,
      "learning_rate": 1e-06,
      "loss": 0.2,
      "step": 137
    },
    {
      "epoch": 0.0478502080443828,
      "grad_norm": 5.4056315422058105,
      "learning_rate": 1e-06,
      "loss": 0.2754,
      "step": 138
    },
    {
      "epoch": 0.048196948682385575,
      "grad_norm": 1.9696849584579468,
      "learning_rate": 1e-06,
      "loss": 0.1991,
      "step": 139
    },
    {
      "epoch": 0.04854368932038835,
      "grad_norm": 1.8743115663528442,
      "learning_rate": 1e-06,
      "loss": 0.2051,
      "step": 140
    },
    {
      "epoch": 0.04889042995839112,
      "grad_norm": 2.6441774368286133,
      "learning_rate": 1e-06,
      "loss": 0.2093,
      "step": 141
    },
    {
      "epoch": 0.049237170596393896,
      "grad_norm": 1.965954303741455,
      "learning_rate": 1e-06,
      "loss": 0.2259,
      "step": 142
    },
    {
      "epoch": 0.04958391123439667,
      "grad_norm": 3.0601775646209717,
      "learning_rate": 1e-06,
      "loss": 0.2273,
      "step": 143
    },
    {
      "epoch": 0.049930651872399444,
      "grad_norm": 2.259979009628296,
      "learning_rate": 1e-06,
      "loss": 0.2355,
      "step": 144
    },
    {
      "epoch": 0.05027739251040222,
      "grad_norm": 3.249948024749756,
      "learning_rate": 1e-06,
      "loss": 0.2634,
      "step": 145
    },
    {
      "epoch": 0.05062413314840499,
      "grad_norm": 4.351866245269775,
      "learning_rate": 1e-06,
      "loss": 0.2353,
      "step": 146
    },
    {
      "epoch": 0.050970873786407765,
      "grad_norm": 1.9071911573410034,
      "learning_rate": 1e-06,
      "loss": 0.2508,
      "step": 147
    },
    {
      "epoch": 0.05131761442441054,
      "grad_norm": 4.363613128662109,
      "learning_rate": 1e-06,
      "loss": 0.2179,
      "step": 148
    },
    {
      "epoch": 0.05166435506241331,
      "grad_norm": 3.450490713119507,
      "learning_rate": 1e-06,
      "loss": 0.2519,
      "step": 149
    },
    {
      "epoch": 0.052011095700416086,
      "grad_norm": 4.706002712249756,
      "learning_rate": 1e-06,
      "loss": 0.2135,
      "step": 150
    },
    {
      "epoch": 0.05235783633841886,
      "grad_norm": 2.632718324661255,
      "learning_rate": 1e-06,
      "loss": 0.2218,
      "step": 151
    },
    {
      "epoch": 0.052704576976421634,
      "grad_norm": 3.9972386360168457,
      "learning_rate": 1e-06,
      "loss": 0.2227,
      "step": 152
    },
    {
      "epoch": 0.05305131761442441,
      "grad_norm": 1.96170175075531,
      "learning_rate": 1e-06,
      "loss": 0.2189,
      "step": 153
    },
    {
      "epoch": 0.05339805825242718,
      "grad_norm": 4.996349811553955,
      "learning_rate": 1e-06,
      "loss": 0.2225,
      "step": 154
    },
    {
      "epoch": 0.053744798890429955,
      "grad_norm": 5.351596832275391,
      "learning_rate": 1e-06,
      "loss": 0.2043,
      "step": 155
    },
    {
      "epoch": 0.05409153952843273,
      "grad_norm": 2.1169140338897705,
      "learning_rate": 1e-06,
      "loss": 0.2483,
      "step": 156
    },
    {
      "epoch": 0.0544382801664355,
      "grad_norm": 2.441293478012085,
      "learning_rate": 1e-06,
      "loss": 0.216,
      "step": 157
    },
    {
      "epoch": 0.054785020804438284,
      "grad_norm": 2.3679656982421875,
      "learning_rate": 1e-06,
      "loss": 0.2292,
      "step": 158
    },
    {
      "epoch": 0.05513176144244106,
      "grad_norm": 5.575881004333496,
      "learning_rate": 1e-06,
      "loss": 0.2512,
      "step": 159
    },
    {
      "epoch": 0.05547850208044383,
      "grad_norm": 2.48142671585083,
      "learning_rate": 1e-06,
      "loss": 0.2249,
      "step": 160
    },
    {
      "epoch": 0.055825242718446605,
      "grad_norm": 3.078421115875244,
      "learning_rate": 1e-06,
      "loss": 0.2205,
      "step": 161
    },
    {
      "epoch": 0.05617198335644938,
      "grad_norm": 2.7186150550842285,
      "learning_rate": 1e-06,
      "loss": 0.2208,
      "step": 162
    },
    {
      "epoch": 0.05651872399445215,
      "grad_norm": 2.42346453666687,
      "learning_rate": 1e-06,
      "loss": 0.2404,
      "step": 163
    },
    {
      "epoch": 0.056865464632454926,
      "grad_norm": 2.095087766647339,
      "learning_rate": 1e-06,
      "loss": 0.2354,
      "step": 164
    },
    {
      "epoch": 0.0572122052704577,
      "grad_norm": 2.2342190742492676,
      "learning_rate": 1e-06,
      "loss": 0.2002,
      "step": 165
    },
    {
      "epoch": 0.057558945908460474,
      "grad_norm": 1.9590411186218262,
      "learning_rate": 1e-06,
      "loss": 0.2073,
      "step": 166
    },
    {
      "epoch": 0.05790568654646325,
      "grad_norm": 4.556865692138672,
      "learning_rate": 1e-06,
      "loss": 0.2742,
      "step": 167
    },
    {
      "epoch": 0.05825242718446602,
      "grad_norm": 2.7081010341644287,
      "learning_rate": 1e-06,
      "loss": 0.2455,
      "step": 168
    },
    {
      "epoch": 0.058599167822468795,
      "grad_norm": 2.4702224731445312,
      "learning_rate": 1e-06,
      "loss": 0.2209,
      "step": 169
    },
    {
      "epoch": 0.05894590846047157,
      "grad_norm": 2.41284441947937,
      "learning_rate": 1e-06,
      "loss": 0.259,
      "step": 170
    },
    {
      "epoch": 0.05929264909847434,
      "grad_norm": 3.244028091430664,
      "learning_rate": 1e-06,
      "loss": 0.2358,
      "step": 171
    },
    {
      "epoch": 0.059639389736477116,
      "grad_norm": 2.7794923782348633,
      "learning_rate": 1e-06,
      "loss": 0.1973,
      "step": 172
    },
    {
      "epoch": 0.05998613037447989,
      "grad_norm": 3.4682586193084717,
      "learning_rate": 1e-06,
      "loss": 0.2523,
      "step": 173
    },
    {
      "epoch": 0.060332871012482664,
      "grad_norm": 2.3559231758117676,
      "learning_rate": 1e-06,
      "loss": 0.2529,
      "step": 174
    },
    {
      "epoch": 0.06067961165048544,
      "grad_norm": 3.0808186531066895,
      "learning_rate": 1e-06,
      "loss": 0.1952,
      "step": 175
    },
    {
      "epoch": 0.06102635228848821,
      "grad_norm": 2.975972890853882,
      "learning_rate": 1e-06,
      "loss": 0.2226,
      "step": 176
    },
    {
      "epoch": 0.061373092926490985,
      "grad_norm": 2.133094072341919,
      "learning_rate": 1e-06,
      "loss": 0.2202,
      "step": 177
    },
    {
      "epoch": 0.06171983356449376,
      "grad_norm": 1.9771744012832642,
      "learning_rate": 1e-06,
      "loss": 0.2025,
      "step": 178
    },
    {
      "epoch": 0.06206657420249653,
      "grad_norm": 1.7458380460739136,
      "learning_rate": 1e-06,
      "loss": 0.1922,
      "step": 179
    },
    {
      "epoch": 0.06241331484049931,
      "grad_norm": 2.3998160362243652,
      "learning_rate": 1e-06,
      "loss": 0.1807,
      "step": 180
    },
    {
      "epoch": 0.06276005547850208,
      "grad_norm": 2.3493542671203613,
      "learning_rate": 1e-06,
      "loss": 0.2146,
      "step": 181
    },
    {
      "epoch": 0.06310679611650485,
      "grad_norm": 6.937662124633789,
      "learning_rate": 1e-06,
      "loss": 0.2579,
      "step": 182
    },
    {
      "epoch": 0.06345353675450763,
      "grad_norm": 1.6203336715698242,
      "learning_rate": 1e-06,
      "loss": 0.198,
      "step": 183
    },
    {
      "epoch": 0.0638002773925104,
      "grad_norm": 1.5239827632904053,
      "learning_rate": 1e-06,
      "loss": 0.189,
      "step": 184
    },
    {
      "epoch": 0.06414701803051318,
      "grad_norm": 3.796680450439453,
      "learning_rate": 1e-06,
      "loss": 0.2651,
      "step": 185
    },
    {
      "epoch": 0.06449375866851595,
      "grad_norm": 2.5366694927215576,
      "learning_rate": 1e-06,
      "loss": 0.2121,
      "step": 186
    },
    {
      "epoch": 0.06484049930651872,
      "grad_norm": 1.8307462930679321,
      "learning_rate": 1e-06,
      "loss": 0.2303,
      "step": 187
    },
    {
      "epoch": 0.0651872399445215,
      "grad_norm": 4.409234523773193,
      "learning_rate": 1e-06,
      "loss": 0.2159,
      "step": 188
    },
    {
      "epoch": 0.06553398058252427,
      "grad_norm": 4.105565547943115,
      "learning_rate": 1e-06,
      "loss": 0.2293,
      "step": 189
    },
    {
      "epoch": 0.06588072122052704,
      "grad_norm": 4.292323589324951,
      "learning_rate": 1e-06,
      "loss": 0.2063,
      "step": 190
    },
    {
      "epoch": 0.06622746185852982,
      "grad_norm": 2.141298294067383,
      "learning_rate": 1e-06,
      "loss": 0.1966,
      "step": 191
    },
    {
      "epoch": 0.06657420249653259,
      "grad_norm": 2.1823067665100098,
      "learning_rate": 1e-06,
      "loss": 0.2019,
      "step": 192
    },
    {
      "epoch": 0.06692094313453537,
      "grad_norm": 4.099167823791504,
      "learning_rate": 1e-06,
      "loss": 0.237,
      "step": 193
    },
    {
      "epoch": 0.06726768377253814,
      "grad_norm": 2.6066272258758545,
      "learning_rate": 1e-06,
      "loss": 0.2275,
      "step": 194
    },
    {
      "epoch": 0.06761442441054091,
      "grad_norm": 2.4590089321136475,
      "learning_rate": 1e-06,
      "loss": 0.2467,
      "step": 195
    },
    {
      "epoch": 0.06796116504854369,
      "grad_norm": 3.860490083694458,
      "learning_rate": 1e-06,
      "loss": 0.2552,
      "step": 196
    },
    {
      "epoch": 0.06830790568654646,
      "grad_norm": 2.6223294734954834,
      "learning_rate": 1e-06,
      "loss": 0.2669,
      "step": 197
    },
    {
      "epoch": 0.06865464632454923,
      "grad_norm": 3.189718723297119,
      "learning_rate": 1e-06,
      "loss": 0.2096,
      "step": 198
    },
    {
      "epoch": 0.06900138696255201,
      "grad_norm": 5.834605693817139,
      "learning_rate": 1e-06,
      "loss": 0.2272,
      "step": 199
    },
    {
      "epoch": 0.06934812760055478,
      "grad_norm": 4.342336177825928,
      "learning_rate": 1e-06,
      "loss": 0.205,
      "step": 200
    },
    {
      "epoch": 0.06969486823855756,
      "grad_norm": 5.050157070159912,
      "learning_rate": 1e-06,
      "loss": 0.2252,
      "step": 201
    },
    {
      "epoch": 0.07004160887656033,
      "grad_norm": 1.8504279851913452,
      "learning_rate": 1e-06,
      "loss": 0.1912,
      "step": 202
    },
    {
      "epoch": 0.0703883495145631,
      "grad_norm": 2.362457275390625,
      "learning_rate": 1e-06,
      "loss": 0.2248,
      "step": 203
    },
    {
      "epoch": 0.07073509015256588,
      "grad_norm": 4.938218116760254,
      "learning_rate": 1e-06,
      "loss": 0.2119,
      "step": 204
    },
    {
      "epoch": 0.07108183079056865,
      "grad_norm": 4.182404041290283,
      "learning_rate": 1e-06,
      "loss": 0.2224,
      "step": 205
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 1.9187567234039307,
      "learning_rate": 1e-06,
      "loss": 0.2193,
      "step": 206
    },
    {
      "epoch": 0.0717753120665742,
      "grad_norm": 4.429518699645996,
      "learning_rate": 1e-06,
      "loss": 0.2174,
      "step": 207
    },
    {
      "epoch": 0.07212205270457697,
      "grad_norm": 5.2877044677734375,
      "learning_rate": 1e-06,
      "loss": 0.2216,
      "step": 208
    },
    {
      "epoch": 0.07246879334257975,
      "grad_norm": 4.316680431365967,
      "learning_rate": 1e-06,
      "loss": 0.254,
      "step": 209
    },
    {
      "epoch": 0.07281553398058252,
      "grad_norm": 2.9150466918945312,
      "learning_rate": 1e-06,
      "loss": 0.2083,
      "step": 210
    },
    {
      "epoch": 0.0731622746185853,
      "grad_norm": 1.7052415609359741,
      "learning_rate": 1e-06,
      "loss": 0.2126,
      "step": 211
    },
    {
      "epoch": 0.07350901525658807,
      "grad_norm": 3.053804636001587,
      "learning_rate": 1e-06,
      "loss": 0.2239,
      "step": 212
    },
    {
      "epoch": 0.07385575589459084,
      "grad_norm": 2.949108123779297,
      "learning_rate": 1e-06,
      "loss": 0.2101,
      "step": 213
    },
    {
      "epoch": 0.07420249653259361,
      "grad_norm": 1.8862674236297607,
      "learning_rate": 1e-06,
      "loss": 0.2066,
      "step": 214
    },
    {
      "epoch": 0.07454923717059639,
      "grad_norm": 2.371164321899414,
      "learning_rate": 1e-06,
      "loss": 0.2532,
      "step": 215
    },
    {
      "epoch": 0.07489597780859916,
      "grad_norm": 2.709270477294922,
      "learning_rate": 1e-06,
      "loss": 0.1968,
      "step": 216
    },
    {
      "epoch": 0.07524271844660194,
      "grad_norm": 5.096902847290039,
      "learning_rate": 1e-06,
      "loss": 0.2111,
      "step": 217
    },
    {
      "epoch": 0.07558945908460471,
      "grad_norm": 2.7359046936035156,
      "learning_rate": 1e-06,
      "loss": 0.1927,
      "step": 218
    },
    {
      "epoch": 0.07593619972260748,
      "grad_norm": 1.800238847732544,
      "learning_rate": 1e-06,
      "loss": 0.1898,
      "step": 219
    },
    {
      "epoch": 0.07628294036061026,
      "grad_norm": 4.802062511444092,
      "learning_rate": 1e-06,
      "loss": 0.221,
      "step": 220
    },
    {
      "epoch": 0.07662968099861303,
      "grad_norm": 2.447291374206543,
      "learning_rate": 1e-06,
      "loss": 0.2287,
      "step": 221
    },
    {
      "epoch": 0.0769764216366158,
      "grad_norm": 7.738630771636963,
      "learning_rate": 1e-06,
      "loss": 0.2055,
      "step": 222
    },
    {
      "epoch": 0.07732316227461858,
      "grad_norm": 2.3890295028686523,
      "learning_rate": 1e-06,
      "loss": 0.2246,
      "step": 223
    },
    {
      "epoch": 0.07766990291262135,
      "grad_norm": 4.384284496307373,
      "learning_rate": 1e-06,
      "loss": 0.2348,
      "step": 224
    },
    {
      "epoch": 0.07801664355062413,
      "grad_norm": 2.8400914669036865,
      "learning_rate": 1e-06,
      "loss": 0.2089,
      "step": 225
    },
    {
      "epoch": 0.07836338418862691,
      "grad_norm": 2.7480292320251465,
      "learning_rate": 1e-06,
      "loss": 0.2736,
      "step": 226
    },
    {
      "epoch": 0.07871012482662969,
      "grad_norm": 5.927616596221924,
      "learning_rate": 1e-06,
      "loss": 0.2613,
      "step": 227
    },
    {
      "epoch": 0.07905686546463246,
      "grad_norm": 5.0669846534729,
      "learning_rate": 1e-06,
      "loss": 0.24,
      "step": 228
    },
    {
      "epoch": 0.07940360610263524,
      "grad_norm": 5.679924011230469,
      "learning_rate": 1e-06,
      "loss": 0.2188,
      "step": 229
    },
    {
      "epoch": 0.07975034674063801,
      "grad_norm": 3.200345277786255,
      "learning_rate": 1e-06,
      "loss": 0.2018,
      "step": 230
    },
    {
      "epoch": 0.08009708737864078,
      "grad_norm": 3.164886474609375,
      "learning_rate": 1e-06,
      "loss": 0.2325,
      "step": 231
    },
    {
      "epoch": 0.08044382801664356,
      "grad_norm": 3.412248373031616,
      "learning_rate": 1e-06,
      "loss": 0.2029,
      "step": 232
    },
    {
      "epoch": 0.08079056865464633,
      "grad_norm": 4.144454479217529,
      "learning_rate": 1e-06,
      "loss": 0.1915,
      "step": 233
    },
    {
      "epoch": 0.0811373092926491,
      "grad_norm": 2.3590505123138428,
      "learning_rate": 1e-06,
      "loss": 0.2341,
      "step": 234
    },
    {
      "epoch": 0.08148404993065188,
      "grad_norm": 1.8131296634674072,
      "learning_rate": 1e-06,
      "loss": 0.2044,
      "step": 235
    },
    {
      "epoch": 0.08183079056865465,
      "grad_norm": 3.05785870552063,
      "learning_rate": 1e-06,
      "loss": 0.2044,
      "step": 236
    },
    {
      "epoch": 0.08217753120665743,
      "grad_norm": 3.1392011642456055,
      "learning_rate": 1e-06,
      "loss": 0.203,
      "step": 237
    },
    {
      "epoch": 0.0825242718446602,
      "grad_norm": 2.5274972915649414,
      "learning_rate": 1e-06,
      "loss": 0.2357,
      "step": 238
    },
    {
      "epoch": 0.08287101248266297,
      "grad_norm": 1.6625158786773682,
      "learning_rate": 1e-06,
      "loss": 0.2181,
      "step": 239
    },
    {
      "epoch": 0.08321775312066575,
      "grad_norm": 1.920361876487732,
      "learning_rate": 1e-06,
      "loss": 0.2186,
      "step": 240
    },
    {
      "epoch": 0.08356449375866852,
      "grad_norm": 3.1764955520629883,
      "learning_rate": 1e-06,
      "loss": 0.1844,
      "step": 241
    },
    {
      "epoch": 0.0839112343966713,
      "grad_norm": 3.478422164916992,
      "learning_rate": 1e-06,
      "loss": 0.2248,
      "step": 242
    },
    {
      "epoch": 0.08425797503467407,
      "grad_norm": 2.3174679279327393,
      "learning_rate": 1e-06,
      "loss": 0.2345,
      "step": 243
    },
    {
      "epoch": 0.08460471567267684,
      "grad_norm": 3.9944794178009033,
      "learning_rate": 1e-06,
      "loss": 0.1961,
      "step": 244
    },
    {
      "epoch": 0.08495145631067962,
      "grad_norm": 4.681294918060303,
      "learning_rate": 1e-06,
      "loss": 0.2107,
      "step": 245
    },
    {
      "epoch": 0.08529819694868239,
      "grad_norm": 1.9410604238510132,
      "learning_rate": 1e-06,
      "loss": 0.2098,
      "step": 246
    },
    {
      "epoch": 0.08564493758668516,
      "grad_norm": 6.453781604766846,
      "learning_rate": 1e-06,
      "loss": 0.2285,
      "step": 247
    },
    {
      "epoch": 0.08599167822468794,
      "grad_norm": 1.9601918458938599,
      "learning_rate": 1e-06,
      "loss": 0.2293,
      "step": 248
    },
    {
      "epoch": 0.08633841886269071,
      "grad_norm": 3.6088569164276123,
      "learning_rate": 1e-06,
      "loss": 0.2397,
      "step": 249
    },
    {
      "epoch": 0.08668515950069348,
      "grad_norm": 6.8316826820373535,
      "learning_rate": 1e-06,
      "loss": 0.2361,
      "step": 250
    },
    {
      "epoch": 0.08703190013869626,
      "grad_norm": 4.1882147789001465,
      "learning_rate": 1e-06,
      "loss": 0.2352,
      "step": 251
    },
    {
      "epoch": 0.08737864077669903,
      "grad_norm": 2.6391122341156006,
      "learning_rate": 1e-06,
      "loss": 0.2207,
      "step": 252
    },
    {
      "epoch": 0.0877253814147018,
      "grad_norm": 1.6957664489746094,
      "learning_rate": 1e-06,
      "loss": 0.1932,
      "step": 253
    },
    {
      "epoch": 0.08807212205270458,
      "grad_norm": 2.1076745986938477,
      "learning_rate": 1e-06,
      "loss": 0.2082,
      "step": 254
    },
    {
      "epoch": 0.08841886269070735,
      "grad_norm": 2.3015823364257812,
      "learning_rate": 1e-06,
      "loss": 0.2228,
      "step": 255
    },
    {
      "epoch": 0.08876560332871013,
      "grad_norm": 4.031891822814941,
      "learning_rate": 1e-06,
      "loss": 0.2055,
      "step": 256
    },
    {
      "epoch": 0.0891123439667129,
      "grad_norm": 2.306821346282959,
      "learning_rate": 1e-06,
      "loss": 0.2286,
      "step": 257
    },
    {
      "epoch": 0.08945908460471567,
      "grad_norm": 2.1814613342285156,
      "learning_rate": 1e-06,
      "loss": 0.2046,
      "step": 258
    },
    {
      "epoch": 0.08980582524271845,
      "grad_norm": 5.232451438903809,
      "learning_rate": 1e-06,
      "loss": 0.216,
      "step": 259
    },
    {
      "epoch": 0.09015256588072122,
      "grad_norm": 7.794690132141113,
      "learning_rate": 1e-06,
      "loss": 0.1838,
      "step": 260
    },
    {
      "epoch": 0.090499306518724,
      "grad_norm": 3.107283115386963,
      "learning_rate": 1e-06,
      "loss": 0.2261,
      "step": 261
    },
    {
      "epoch": 0.09084604715672677,
      "grad_norm": 1.7422479391098022,
      "learning_rate": 1e-06,
      "loss": 0.2222,
      "step": 262
    },
    {
      "epoch": 0.09119278779472954,
      "grad_norm": 2.579590320587158,
      "learning_rate": 1e-06,
      "loss": 0.2015,
      "step": 263
    },
    {
      "epoch": 0.09153952843273232,
      "grad_norm": 3.788679838180542,
      "learning_rate": 1e-06,
      "loss": 0.2373,
      "step": 264
    },
    {
      "epoch": 0.09188626907073509,
      "grad_norm": 2.0979104042053223,
      "learning_rate": 1e-06,
      "loss": 0.2183,
      "step": 265
    },
    {
      "epoch": 0.09223300970873786,
      "grad_norm": 2.019399642944336,
      "learning_rate": 1e-06,
      "loss": 0.2195,
      "step": 266
    },
    {
      "epoch": 0.09257975034674064,
      "grad_norm": 2.5858969688415527,
      "learning_rate": 1e-06,
      "loss": 0.234,
      "step": 267
    },
    {
      "epoch": 0.09292649098474341,
      "grad_norm": 2.279453754425049,
      "learning_rate": 1e-06,
      "loss": 0.2312,
      "step": 268
    },
    {
      "epoch": 0.09327323162274619,
      "grad_norm": 3.3690133094787598,
      "learning_rate": 1e-06,
      "loss": 0.2499,
      "step": 269
    },
    {
      "epoch": 0.09361997226074896,
      "grad_norm": 1.591291904449463,
      "learning_rate": 1e-06,
      "loss": 0.1986,
      "step": 270
    },
    {
      "epoch": 0.09396671289875173,
      "grad_norm": 2.8157074451446533,
      "learning_rate": 1e-06,
      "loss": 0.2185,
      "step": 271
    },
    {
      "epoch": 0.09431345353675451,
      "grad_norm": 2.9287891387939453,
      "learning_rate": 1e-06,
      "loss": 0.2151,
      "step": 272
    },
    {
      "epoch": 0.09466019417475728,
      "grad_norm": 3.26141357421875,
      "learning_rate": 1e-06,
      "loss": 0.2295,
      "step": 273
    },
    {
      "epoch": 0.09500693481276005,
      "grad_norm": 1.793872594833374,
      "learning_rate": 1e-06,
      "loss": 0.2053,
      "step": 274
    },
    {
      "epoch": 0.09535367545076283,
      "grad_norm": 6.906111717224121,
      "learning_rate": 1e-06,
      "loss": 0.1894,
      "step": 275
    },
    {
      "epoch": 0.0957004160887656,
      "grad_norm": 5.939563751220703,
      "learning_rate": 1e-06,
      "loss": 0.2454,
      "step": 276
    },
    {
      "epoch": 0.09604715672676838,
      "grad_norm": 4.869255065917969,
      "learning_rate": 1e-06,
      "loss": 0.2029,
      "step": 277
    },
    {
      "epoch": 0.09639389736477115,
      "grad_norm": 2.0359835624694824,
      "learning_rate": 1e-06,
      "loss": 0.2193,
      "step": 278
    },
    {
      "epoch": 0.09674063800277392,
      "grad_norm": 1.732591152191162,
      "learning_rate": 1e-06,
      "loss": 0.2223,
      "step": 279
    },
    {
      "epoch": 0.0970873786407767,
      "grad_norm": 2.112185478210449,
      "learning_rate": 1e-06,
      "loss": 0.2261,
      "step": 280
    },
    {
      "epoch": 0.09743411927877947,
      "grad_norm": 1.9228016138076782,
      "learning_rate": 1e-06,
      "loss": 0.2168,
      "step": 281
    },
    {
      "epoch": 0.09778085991678225,
      "grad_norm": 2.5188448429107666,
      "learning_rate": 1e-06,
      "loss": 0.2526,
      "step": 282
    },
    {
      "epoch": 0.09812760055478502,
      "grad_norm": 6.440067291259766,
      "learning_rate": 1e-06,
      "loss": 0.1962,
      "step": 283
    },
    {
      "epoch": 0.09847434119278779,
      "grad_norm": 2.1563827991485596,
      "learning_rate": 1e-06,
      "loss": 0.2155,
      "step": 284
    },
    {
      "epoch": 0.09882108183079057,
      "grad_norm": 2.890427589416504,
      "learning_rate": 1e-06,
      "loss": 0.204,
      "step": 285
    },
    {
      "epoch": 0.09916782246879334,
      "grad_norm": 3.064075469970703,
      "learning_rate": 1e-06,
      "loss": 0.2124,
      "step": 286
    },
    {
      "epoch": 0.09951456310679611,
      "grad_norm": 2.013902187347412,
      "learning_rate": 1e-06,
      "loss": 0.2011,
      "step": 287
    },
    {
      "epoch": 0.09986130374479889,
      "grad_norm": 5.711750507354736,
      "learning_rate": 1e-06,
      "loss": 0.198,
      "step": 288
    },
    {
      "epoch": 0.10020804438280166,
      "grad_norm": 4.605215072631836,
      "learning_rate": 1e-06,
      "loss": 0.2074,
      "step": 289
    },
    {
      "epoch": 0.10055478502080444,
      "grad_norm": 3.5413124561309814,
      "learning_rate": 1e-06,
      "loss": 0.2516,
      "step": 290
    },
    {
      "epoch": 0.10090152565880721,
      "grad_norm": 2.2203333377838135,
      "learning_rate": 1e-06,
      "loss": 0.2353,
      "step": 291
    },
    {
      "epoch": 0.10124826629680998,
      "grad_norm": 3.4780688285827637,
      "learning_rate": 1e-06,
      "loss": 0.2278,
      "step": 292
    },
    {
      "epoch": 0.10159500693481276,
      "grad_norm": 2.2636001110076904,
      "learning_rate": 1e-06,
      "loss": 0.1992,
      "step": 293
    },
    {
      "epoch": 0.10194174757281553,
      "grad_norm": 4.070486545562744,
      "learning_rate": 1e-06,
      "loss": 0.2037,
      "step": 294
    },
    {
      "epoch": 0.1022884882108183,
      "grad_norm": 2.2457258701324463,
      "learning_rate": 1e-06,
      "loss": 0.2409,
      "step": 295
    },
    {
      "epoch": 0.10263522884882108,
      "grad_norm": 2.1032586097717285,
      "learning_rate": 1e-06,
      "loss": 0.2105,
      "step": 296
    },
    {
      "epoch": 0.10298196948682385,
      "grad_norm": 2.8881735801696777,
      "learning_rate": 1e-06,
      "loss": 0.2115,
      "step": 297
    },
    {
      "epoch": 0.10332871012482663,
      "grad_norm": 2.166180372238159,
      "learning_rate": 1e-06,
      "loss": 0.2231,
      "step": 298
    },
    {
      "epoch": 0.1036754507628294,
      "grad_norm": 2.0594778060913086,
      "learning_rate": 1e-06,
      "loss": 0.2409,
      "step": 299
    },
    {
      "epoch": 0.10402219140083217,
      "grad_norm": 1.8544009923934937,
      "learning_rate": 1e-06,
      "loss": 0.1873,
      "step": 300
    },
    {
      "epoch": 0.10436893203883495,
      "grad_norm": 1.979411005973816,
      "learning_rate": 1e-06,
      "loss": 0.2177,
      "step": 301
    },
    {
      "epoch": 0.10471567267683772,
      "grad_norm": 2.5190136432647705,
      "learning_rate": 1e-06,
      "loss": 0.2206,
      "step": 302
    },
    {
      "epoch": 0.1050624133148405,
      "grad_norm": 2.110924482345581,
      "learning_rate": 1e-06,
      "loss": 0.2087,
      "step": 303
    },
    {
      "epoch": 0.10540915395284327,
      "grad_norm": 5.314784526824951,
      "learning_rate": 1e-06,
      "loss": 0.1707,
      "step": 304
    },
    {
      "epoch": 0.10575589459084604,
      "grad_norm": 3.8186211585998535,
      "learning_rate": 1e-06,
      "loss": 0.2255,
      "step": 305
    },
    {
      "epoch": 0.10610263522884882,
      "grad_norm": 2.197397232055664,
      "learning_rate": 1e-06,
      "loss": 0.2026,
      "step": 306
    },
    {
      "epoch": 0.10644937586685159,
      "grad_norm": 3.364658832550049,
      "learning_rate": 1e-06,
      "loss": 0.237,
      "step": 307
    },
    {
      "epoch": 0.10679611650485436,
      "grad_norm": 3.6538567543029785,
      "learning_rate": 1e-06,
      "loss": 0.2135,
      "step": 308
    },
    {
      "epoch": 0.10714285714285714,
      "grad_norm": 2.080467462539673,
      "learning_rate": 1e-06,
      "loss": 0.2011,
      "step": 309
    },
    {
      "epoch": 0.10748959778085991,
      "grad_norm": 2.394052267074585,
      "learning_rate": 1e-06,
      "loss": 0.2081,
      "step": 310
    },
    {
      "epoch": 0.10783633841886268,
      "grad_norm": 3.270670175552368,
      "learning_rate": 1e-06,
      "loss": 0.2139,
      "step": 311
    },
    {
      "epoch": 0.10818307905686546,
      "grad_norm": 4.5176682472229,
      "learning_rate": 1e-06,
      "loss": 0.2214,
      "step": 312
    },
    {
      "epoch": 0.10852981969486823,
      "grad_norm": 2.9834976196289062,
      "learning_rate": 1e-06,
      "loss": 0.2235,
      "step": 313
    },
    {
      "epoch": 0.108876560332871,
      "grad_norm": 3.658803701400757,
      "learning_rate": 1e-06,
      "loss": 0.1903,
      "step": 314
    },
    {
      "epoch": 0.10922330097087378,
      "grad_norm": 2.972036600112915,
      "learning_rate": 1e-06,
      "loss": 0.2044,
      "step": 315
    },
    {
      "epoch": 0.10957004160887657,
      "grad_norm": 2.014233350753784,
      "learning_rate": 1e-06,
      "loss": 0.2177,
      "step": 316
    },
    {
      "epoch": 0.10991678224687934,
      "grad_norm": 3.3329989910125732,
      "learning_rate": 1e-06,
      "loss": 0.212,
      "step": 317
    },
    {
      "epoch": 0.11026352288488211,
      "grad_norm": 4.286015510559082,
      "learning_rate": 1e-06,
      "loss": 0.2164,
      "step": 318
    },
    {
      "epoch": 0.11061026352288489,
      "grad_norm": 2.517414093017578,
      "learning_rate": 1e-06,
      "loss": 0.2443,
      "step": 319
    },
    {
      "epoch": 0.11095700416088766,
      "grad_norm": 2.3605127334594727,
      "learning_rate": 1e-06,
      "loss": 0.2136,
      "step": 320
    },
    {
      "epoch": 0.11130374479889044,
      "grad_norm": 2.0342857837677,
      "learning_rate": 1e-06,
      "loss": 0.2117,
      "step": 321
    },
    {
      "epoch": 0.11165048543689321,
      "grad_norm": 3.313946485519409,
      "learning_rate": 1e-06,
      "loss": 0.2159,
      "step": 322
    },
    {
      "epoch": 0.11199722607489598,
      "grad_norm": 2.8215742111206055,
      "learning_rate": 1e-06,
      "loss": 0.1666,
      "step": 323
    },
    {
      "epoch": 0.11234396671289876,
      "grad_norm": 3.7066569328308105,
      "learning_rate": 1e-06,
      "loss": 0.2115,
      "step": 324
    },
    {
      "epoch": 0.11269070735090153,
      "grad_norm": 2.824295997619629,
      "learning_rate": 1e-06,
      "loss": 0.2068,
      "step": 325
    },
    {
      "epoch": 0.1130374479889043,
      "grad_norm": 3.302109956741333,
      "learning_rate": 1e-06,
      "loss": 0.2007,
      "step": 326
    },
    {
      "epoch": 0.11338418862690708,
      "grad_norm": 1.7297828197479248,
      "learning_rate": 1e-06,
      "loss": 0.1796,
      "step": 327
    },
    {
      "epoch": 0.11373092926490985,
      "grad_norm": 2.3774609565734863,
      "learning_rate": 1e-06,
      "loss": 0.197,
      "step": 328
    },
    {
      "epoch": 0.11407766990291263,
      "grad_norm": 3.876222610473633,
      "learning_rate": 1e-06,
      "loss": 0.2106,
      "step": 329
    },
    {
      "epoch": 0.1144244105409154,
      "grad_norm": 2.891040086746216,
      "learning_rate": 1e-06,
      "loss": 0.1809,
      "step": 330
    },
    {
      "epoch": 0.11477115117891817,
      "grad_norm": 3.9112064838409424,
      "learning_rate": 1e-06,
      "loss": 0.2031,
      "step": 331
    },
    {
      "epoch": 0.11511789181692095,
      "grad_norm": 1.772921085357666,
      "learning_rate": 1e-06,
      "loss": 0.2011,
      "step": 332
    },
    {
      "epoch": 0.11546463245492372,
      "grad_norm": 2.6547513008117676,
      "learning_rate": 1e-06,
      "loss": 0.2123,
      "step": 333
    },
    {
      "epoch": 0.1158113730929265,
      "grad_norm": 5.290209770202637,
      "learning_rate": 1e-06,
      "loss": 0.2275,
      "step": 334
    },
    {
      "epoch": 0.11615811373092927,
      "grad_norm": 5.404000759124756,
      "learning_rate": 1e-06,
      "loss": 0.2422,
      "step": 335
    },
    {
      "epoch": 0.11650485436893204,
      "grad_norm": 2.187218427658081,
      "learning_rate": 1e-06,
      "loss": 0.1854,
      "step": 336
    },
    {
      "epoch": 0.11685159500693482,
      "grad_norm": 3.028299570083618,
      "learning_rate": 1e-06,
      "loss": 0.2331,
      "step": 337
    },
    {
      "epoch": 0.11719833564493759,
      "grad_norm": 2.9846150875091553,
      "learning_rate": 1e-06,
      "loss": 0.2288,
      "step": 338
    },
    {
      "epoch": 0.11754507628294036,
      "grad_norm": 3.630133867263794,
      "learning_rate": 1e-06,
      "loss": 0.1559,
      "step": 339
    },
    {
      "epoch": 0.11789181692094314,
      "grad_norm": 6.968530654907227,
      "learning_rate": 1e-06,
      "loss": 0.2637,
      "step": 340
    },
    {
      "epoch": 0.11823855755894591,
      "grad_norm": 2.9122650623321533,
      "learning_rate": 1e-06,
      "loss": 0.1821,
      "step": 341
    },
    {
      "epoch": 0.11858529819694869,
      "grad_norm": 2.7790746688842773,
      "learning_rate": 1e-06,
      "loss": 0.1917,
      "step": 342
    },
    {
      "epoch": 0.11893203883495146,
      "grad_norm": 4.776673793792725,
      "learning_rate": 1e-06,
      "loss": 0.1928,
      "step": 343
    },
    {
      "epoch": 0.11927877947295423,
      "grad_norm": 2.7039010524749756,
      "learning_rate": 1e-06,
      "loss": 0.2258,
      "step": 344
    },
    {
      "epoch": 0.119625520110957,
      "grad_norm": 2.733076333999634,
      "learning_rate": 1e-06,
      "loss": 0.1988,
      "step": 345
    },
    {
      "epoch": 0.11997226074895978,
      "grad_norm": 3.1806342601776123,
      "learning_rate": 1e-06,
      "loss": 0.2147,
      "step": 346
    },
    {
      "epoch": 0.12031900138696255,
      "grad_norm": 2.9872305393218994,
      "learning_rate": 1e-06,
      "loss": 0.1832,
      "step": 347
    },
    {
      "epoch": 0.12066574202496533,
      "grad_norm": 6.822136402130127,
      "learning_rate": 1e-06,
      "loss": 0.2202,
      "step": 348
    },
    {
      "epoch": 0.1210124826629681,
      "grad_norm": 2.3304481506347656,
      "learning_rate": 1e-06,
      "loss": 0.2352,
      "step": 349
    },
    {
      "epoch": 0.12135922330097088,
      "grad_norm": 2.4243297576904297,
      "learning_rate": 1e-06,
      "loss": 0.2004,
      "step": 350
    },
    {
      "epoch": 0.12170596393897365,
      "grad_norm": 3.050774097442627,
      "learning_rate": 1e-06,
      "loss": 0.2103,
      "step": 351
    },
    {
      "epoch": 0.12205270457697642,
      "grad_norm": 2.177351951599121,
      "learning_rate": 1e-06,
      "loss": 0.2416,
      "step": 352
    },
    {
      "epoch": 0.1223994452149792,
      "grad_norm": 4.790650844573975,
      "learning_rate": 1e-06,
      "loss": 0.2174,
      "step": 353
    },
    {
      "epoch": 0.12274618585298197,
      "grad_norm": 1.6365033388137817,
      "learning_rate": 1e-06,
      "loss": 0.1874,
      "step": 354
    },
    {
      "epoch": 0.12309292649098474,
      "grad_norm": 3.9262518882751465,
      "learning_rate": 1e-06,
      "loss": 0.2175,
      "step": 355
    },
    {
      "epoch": 0.12343966712898752,
      "grad_norm": 1.7626748085021973,
      "learning_rate": 1e-06,
      "loss": 0.2129,
      "step": 356
    },
    {
      "epoch": 0.12378640776699029,
      "grad_norm": 1.7612718343734741,
      "learning_rate": 1e-06,
      "loss": 0.1816,
      "step": 357
    },
    {
      "epoch": 0.12413314840499307,
      "grad_norm": 1.9998083114624023,
      "learning_rate": 1e-06,
      "loss": 0.1955,
      "step": 358
    },
    {
      "epoch": 0.12447988904299584,
      "grad_norm": 1.9432895183563232,
      "learning_rate": 1e-06,
      "loss": 0.1793,
      "step": 359
    },
    {
      "epoch": 0.12482662968099861,
      "grad_norm": 2.2494137287139893,
      "learning_rate": 1e-06,
      "loss": 0.2182,
      "step": 360
    },
    {
      "epoch": 0.1251733703190014,
      "grad_norm": 2.6575872898101807,
      "learning_rate": 1e-06,
      "loss": 0.2126,
      "step": 361
    },
    {
      "epoch": 0.12552011095700416,
      "grad_norm": 2.1404902935028076,
      "learning_rate": 1e-06,
      "loss": 0.1727,
      "step": 362
    },
    {
      "epoch": 0.12586685159500693,
      "grad_norm": 3.4852030277252197,
      "learning_rate": 1e-06,
      "loss": 0.2132,
      "step": 363
    },
    {
      "epoch": 0.1262135922330097,
      "grad_norm": 1.9237537384033203,
      "learning_rate": 1e-06,
      "loss": 0.1882,
      "step": 364
    },
    {
      "epoch": 0.12656033287101248,
      "grad_norm": 2.6047005653381348,
      "learning_rate": 1e-06,
      "loss": 0.2276,
      "step": 365
    },
    {
      "epoch": 0.12690707350901526,
      "grad_norm": 4.0389838218688965,
      "learning_rate": 1e-06,
      "loss": 0.2294,
      "step": 366
    },
    {
      "epoch": 0.12725381414701803,
      "grad_norm": 2.9058427810668945,
      "learning_rate": 1e-06,
      "loss": 0.1704,
      "step": 367
    },
    {
      "epoch": 0.1276005547850208,
      "grad_norm": 2.214973211288452,
      "learning_rate": 1e-06,
      "loss": 0.1933,
      "step": 368
    },
    {
      "epoch": 0.12794729542302358,
      "grad_norm": 2.8029608726501465,
      "learning_rate": 1e-06,
      "loss": 0.238,
      "step": 369
    },
    {
      "epoch": 0.12829403606102635,
      "grad_norm": 3.4244496822357178,
      "learning_rate": 1e-06,
      "loss": 0.2404,
      "step": 370
    },
    {
      "epoch": 0.12864077669902912,
      "grad_norm": 3.9695568084716797,
      "learning_rate": 1e-06,
      "loss": 0.2299,
      "step": 371
    },
    {
      "epoch": 0.1289875173370319,
      "grad_norm": 2.100534439086914,
      "learning_rate": 1e-06,
      "loss": 0.1878,
      "step": 372
    },
    {
      "epoch": 0.12933425797503467,
      "grad_norm": 1.942069172859192,
      "learning_rate": 1e-06,
      "loss": 0.227,
      "step": 373
    },
    {
      "epoch": 0.12968099861303745,
      "grad_norm": 2.7720251083374023,
      "learning_rate": 1e-06,
      "loss": 0.2054,
      "step": 374
    },
    {
      "epoch": 0.13002773925104022,
      "grad_norm": 1.916839599609375,
      "learning_rate": 1e-06,
      "loss": 0.2135,
      "step": 375
    },
    {
      "epoch": 0.130374479889043,
      "grad_norm": 2.759535312652588,
      "learning_rate": 1e-06,
      "loss": 0.2327,
      "step": 376
    },
    {
      "epoch": 0.13072122052704577,
      "grad_norm": 2.059061288833618,
      "learning_rate": 1e-06,
      "loss": 0.199,
      "step": 377
    },
    {
      "epoch": 0.13106796116504854,
      "grad_norm": 1.892815113067627,
      "learning_rate": 1e-06,
      "loss": 0.1972,
      "step": 378
    },
    {
      "epoch": 0.13141470180305131,
      "grad_norm": 2.945600748062134,
      "learning_rate": 1e-06,
      "loss": 0.2046,
      "step": 379
    },
    {
      "epoch": 0.1317614424410541,
      "grad_norm": 2.75258207321167,
      "learning_rate": 1e-06,
      "loss": 0.2171,
      "step": 380
    },
    {
      "epoch": 0.13210818307905686,
      "grad_norm": 3.2966949939727783,
      "learning_rate": 1e-06,
      "loss": 0.2423,
      "step": 381
    },
    {
      "epoch": 0.13245492371705964,
      "grad_norm": 2.6719970703125,
      "learning_rate": 1e-06,
      "loss": 0.1777,
      "step": 382
    },
    {
      "epoch": 0.1328016643550624,
      "grad_norm": 3.278883695602417,
      "learning_rate": 1e-06,
      "loss": 0.16,
      "step": 383
    },
    {
      "epoch": 0.13314840499306518,
      "grad_norm": 3.322058916091919,
      "learning_rate": 1e-06,
      "loss": 0.1984,
      "step": 384
    },
    {
      "epoch": 0.13349514563106796,
      "grad_norm": 2.3730714321136475,
      "learning_rate": 1e-06,
      "loss": 0.1836,
      "step": 385
    },
    {
      "epoch": 0.13384188626907073,
      "grad_norm": 1.6652333736419678,
      "learning_rate": 1e-06,
      "loss": 0.1896,
      "step": 386
    },
    {
      "epoch": 0.1341886269070735,
      "grad_norm": 3.528460741043091,
      "learning_rate": 1e-06,
      "loss": 0.2363,
      "step": 387
    },
    {
      "epoch": 0.13453536754507628,
      "grad_norm": 2.009749174118042,
      "learning_rate": 1e-06,
      "loss": 0.1814,
      "step": 388
    },
    {
      "epoch": 0.13488210818307905,
      "grad_norm": 2.015735149383545,
      "learning_rate": 1e-06,
      "loss": 0.2067,
      "step": 389
    },
    {
      "epoch": 0.13522884882108183,
      "grad_norm": 1.9921092987060547,
      "learning_rate": 1e-06,
      "loss": 0.1973,
      "step": 390
    },
    {
      "epoch": 0.1355755894590846,
      "grad_norm": 2.4047930240631104,
      "learning_rate": 1e-06,
      "loss": 0.1924,
      "step": 391
    },
    {
      "epoch": 0.13592233009708737,
      "grad_norm": 2.010021924972534,
      "learning_rate": 1e-06,
      "loss": 0.1744,
      "step": 392
    },
    {
      "epoch": 0.13626907073509015,
      "grad_norm": 2.5969536304473877,
      "learning_rate": 1e-06,
      "loss": 0.2092,
      "step": 393
    },
    {
      "epoch": 0.13661581137309292,
      "grad_norm": 3.3786892890930176,
      "learning_rate": 1e-06,
      "loss": 0.2044,
      "step": 394
    },
    {
      "epoch": 0.1369625520110957,
      "grad_norm": 1.805240511894226,
      "learning_rate": 1e-06,
      "loss": 0.2015,
      "step": 395
    },
    {
      "epoch": 0.13730929264909847,
      "grad_norm": 1.8784292936325073,
      "learning_rate": 1e-06,
      "loss": 0.2059,
      "step": 396
    },
    {
      "epoch": 0.13765603328710124,
      "grad_norm": 2.3622231483459473,
      "learning_rate": 1e-06,
      "loss": 0.1786,
      "step": 397
    },
    {
      "epoch": 0.13800277392510402,
      "grad_norm": 1.7961145639419556,
      "learning_rate": 1e-06,
      "loss": 0.2168,
      "step": 398
    },
    {
      "epoch": 0.1383495145631068,
      "grad_norm": 1.9970123767852783,
      "learning_rate": 1e-06,
      "loss": 0.1922,
      "step": 399
    },
    {
      "epoch": 0.13869625520110956,
      "grad_norm": 5.615191459655762,
      "learning_rate": 1e-06,
      "loss": 0.2319,
      "step": 400
    },
    {
      "epoch": 0.13904299583911234,
      "grad_norm": 2.6220815181732178,
      "learning_rate": 1e-06,
      "loss": 0.1899,
      "step": 401
    },
    {
      "epoch": 0.1393897364771151,
      "grad_norm": 3.2528562545776367,
      "learning_rate": 1e-06,
      "loss": 0.1744,
      "step": 402
    },
    {
      "epoch": 0.13973647711511789,
      "grad_norm": 2.6345772743225098,
      "learning_rate": 1e-06,
      "loss": 0.2226,
      "step": 403
    },
    {
      "epoch": 0.14008321775312066,
      "grad_norm": 5.005814075469971,
      "learning_rate": 1e-06,
      "loss": 0.2174,
      "step": 404
    },
    {
      "epoch": 0.14042995839112343,
      "grad_norm": 2.4105288982391357,
      "learning_rate": 1e-06,
      "loss": 0.2154,
      "step": 405
    },
    {
      "epoch": 0.1407766990291262,
      "grad_norm": 3.282689332962036,
      "learning_rate": 1e-06,
      "loss": 0.246,
      "step": 406
    },
    {
      "epoch": 0.14112343966712898,
      "grad_norm": 1.84548020362854,
      "learning_rate": 1e-06,
      "loss": 0.204,
      "step": 407
    },
    {
      "epoch": 0.14147018030513175,
      "grad_norm": 3.254951000213623,
      "learning_rate": 1e-06,
      "loss": 0.2355,
      "step": 408
    },
    {
      "epoch": 0.14181692094313453,
      "grad_norm": 2.499875068664551,
      "learning_rate": 1e-06,
      "loss": 0.2076,
      "step": 409
    },
    {
      "epoch": 0.1421636615811373,
      "grad_norm": 1.6304558515548706,
      "learning_rate": 1e-06,
      "loss": 0.1925,
      "step": 410
    },
    {
      "epoch": 0.14251040221914008,
      "grad_norm": 2.8476548194885254,
      "learning_rate": 1e-06,
      "loss": 0.2329,
      "step": 411
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 2.403111696243286,
      "learning_rate": 1e-06,
      "loss": 0.2016,
      "step": 412
    },
    {
      "epoch": 0.14320388349514562,
      "grad_norm": 2.050206422805786,
      "learning_rate": 1e-06,
      "loss": 0.227,
      "step": 413
    },
    {
      "epoch": 0.1435506241331484,
      "grad_norm": 3.8210604190826416,
      "learning_rate": 1e-06,
      "loss": 0.1825,
      "step": 414
    },
    {
      "epoch": 0.14389736477115117,
      "grad_norm": 2.2794923782348633,
      "learning_rate": 1e-06,
      "loss": 0.1822,
      "step": 415
    },
    {
      "epoch": 0.14424410540915394,
      "grad_norm": 2.152036666870117,
      "learning_rate": 1e-06,
      "loss": 0.2087,
      "step": 416
    },
    {
      "epoch": 0.14459084604715672,
      "grad_norm": 2.4530344009399414,
      "learning_rate": 1e-06,
      "loss": 0.2381,
      "step": 417
    },
    {
      "epoch": 0.1449375866851595,
      "grad_norm": 2.0212299823760986,
      "learning_rate": 1e-06,
      "loss": 0.2166,
      "step": 418
    },
    {
      "epoch": 0.14528432732316227,
      "grad_norm": 5.47108268737793,
      "learning_rate": 1e-06,
      "loss": 0.1952,
      "step": 419
    },
    {
      "epoch": 0.14563106796116504,
      "grad_norm": 2.0776853561401367,
      "learning_rate": 1e-06,
      "loss": 0.2232,
      "step": 420
    },
    {
      "epoch": 0.1459778085991678,
      "grad_norm": 5.302248477935791,
      "learning_rate": 1e-06,
      "loss": 0.1972,
      "step": 421
    },
    {
      "epoch": 0.1463245492371706,
      "grad_norm": 6.040057182312012,
      "learning_rate": 1e-06,
      "loss": 0.236,
      "step": 422
    },
    {
      "epoch": 0.14667128987517336,
      "grad_norm": 2.0244548320770264,
      "learning_rate": 1e-06,
      "loss": 0.189,
      "step": 423
    },
    {
      "epoch": 0.14701803051317613,
      "grad_norm": 1.9456743001937866,
      "learning_rate": 1e-06,
      "loss": 0.2195,
      "step": 424
    },
    {
      "epoch": 0.1473647711511789,
      "grad_norm": 2.1498379707336426,
      "learning_rate": 1e-06,
      "loss": 0.1891,
      "step": 425
    },
    {
      "epoch": 0.14771151178918168,
      "grad_norm": 2.6949310302734375,
      "learning_rate": 1e-06,
      "loss": 0.1933,
      "step": 426
    },
    {
      "epoch": 0.14805825242718446,
      "grad_norm": 2.2090682983398438,
      "learning_rate": 1e-06,
      "loss": 0.2001,
      "step": 427
    },
    {
      "epoch": 0.14840499306518723,
      "grad_norm": 2.047943592071533,
      "learning_rate": 1e-06,
      "loss": 0.2008,
      "step": 428
    },
    {
      "epoch": 0.14875173370319,
      "grad_norm": 3.49019718170166,
      "learning_rate": 1e-06,
      "loss": 0.25,
      "step": 429
    },
    {
      "epoch": 0.14909847434119278,
      "grad_norm": 2.07293963432312,
      "learning_rate": 1e-06,
      "loss": 0.176,
      "step": 430
    },
    {
      "epoch": 0.14944521497919555,
      "grad_norm": 1.9637081623077393,
      "learning_rate": 1e-06,
      "loss": 0.1918,
      "step": 431
    },
    {
      "epoch": 0.14979195561719832,
      "grad_norm": 4.7957658767700195,
      "learning_rate": 1e-06,
      "loss": 0.2399,
      "step": 432
    },
    {
      "epoch": 0.1501386962552011,
      "grad_norm": 2.5030736923217773,
      "learning_rate": 1e-06,
      "loss": 0.1782,
      "step": 433
    },
    {
      "epoch": 0.15048543689320387,
      "grad_norm": 1.744457483291626,
      "learning_rate": 1e-06,
      "loss": 0.1907,
      "step": 434
    },
    {
      "epoch": 0.15083217753120665,
      "grad_norm": 2.6001105308532715,
      "learning_rate": 1e-06,
      "loss": 0.1946,
      "step": 435
    },
    {
      "epoch": 0.15117891816920942,
      "grad_norm": 4.785390377044678,
      "learning_rate": 1e-06,
      "loss": 0.2152,
      "step": 436
    },
    {
      "epoch": 0.1515256588072122,
      "grad_norm": 1.5918031930923462,
      "learning_rate": 1e-06,
      "loss": 0.1817,
      "step": 437
    },
    {
      "epoch": 0.15187239944521497,
      "grad_norm": 2.5173332691192627,
      "learning_rate": 1e-06,
      "loss": 0.2512,
      "step": 438
    },
    {
      "epoch": 0.15221914008321774,
      "grad_norm": 2.2436468601226807,
      "learning_rate": 1e-06,
      "loss": 0.2083,
      "step": 439
    },
    {
      "epoch": 0.15256588072122051,
      "grad_norm": 5.61021089553833,
      "learning_rate": 1e-06,
      "loss": 0.2143,
      "step": 440
    },
    {
      "epoch": 0.1529126213592233,
      "grad_norm": 3.092174768447876,
      "learning_rate": 1e-06,
      "loss": 0.2081,
      "step": 441
    },
    {
      "epoch": 0.15325936199722606,
      "grad_norm": 1.9928513765335083,
      "learning_rate": 1e-06,
      "loss": 0.1816,
      "step": 442
    },
    {
      "epoch": 0.15360610263522884,
      "grad_norm": 2.1547915935516357,
      "learning_rate": 1e-06,
      "loss": 0.1919,
      "step": 443
    },
    {
      "epoch": 0.1539528432732316,
      "grad_norm": 4.531169414520264,
      "learning_rate": 1e-06,
      "loss": 0.2085,
      "step": 444
    },
    {
      "epoch": 0.15429958391123438,
      "grad_norm": 1.7542948722839355,
      "learning_rate": 1e-06,
      "loss": 0.1803,
      "step": 445
    },
    {
      "epoch": 0.15464632454923716,
      "grad_norm": 2.1614997386932373,
      "learning_rate": 1e-06,
      "loss": 0.1946,
      "step": 446
    },
    {
      "epoch": 0.15499306518723993,
      "grad_norm": 5.018011093139648,
      "learning_rate": 1e-06,
      "loss": 0.1974,
      "step": 447
    },
    {
      "epoch": 0.1553398058252427,
      "grad_norm": 2.907313346862793,
      "learning_rate": 1e-06,
      "loss": 0.1916,
      "step": 448
    },
    {
      "epoch": 0.15568654646324548,
      "grad_norm": 2.7045724391937256,
      "learning_rate": 1e-06,
      "loss": 0.1847,
      "step": 449
    },
    {
      "epoch": 0.15603328710124825,
      "grad_norm": 1.8427867889404297,
      "learning_rate": 1e-06,
      "loss": 0.1918,
      "step": 450
    },
    {
      "epoch": 0.15638002773925105,
      "grad_norm": 2.3798131942749023,
      "learning_rate": 1e-06,
      "loss": 0.2125,
      "step": 451
    },
    {
      "epoch": 0.15672676837725383,
      "grad_norm": 3.0421640872955322,
      "learning_rate": 1e-06,
      "loss": 0.1729,
      "step": 452
    },
    {
      "epoch": 0.1570735090152566,
      "grad_norm": 2.487687110900879,
      "learning_rate": 1e-06,
      "loss": 0.1791,
      "step": 453
    },
    {
      "epoch": 0.15742024965325938,
      "grad_norm": 3.9708778858184814,
      "learning_rate": 1e-06,
      "loss": 0.1902,
      "step": 454
    },
    {
      "epoch": 0.15776699029126215,
      "grad_norm": 2.342562437057495,
      "learning_rate": 1e-06,
      "loss": 0.1757,
      "step": 455
    },
    {
      "epoch": 0.15811373092926492,
      "grad_norm": 3.1214346885681152,
      "learning_rate": 1e-06,
      "loss": 0.1758,
      "step": 456
    },
    {
      "epoch": 0.1584604715672677,
      "grad_norm": 3.499249219894409,
      "learning_rate": 1e-06,
      "loss": 0.1947,
      "step": 457
    },
    {
      "epoch": 0.15880721220527047,
      "grad_norm": 1.9343475103378296,
      "learning_rate": 1e-06,
      "loss": 0.1835,
      "step": 458
    },
    {
      "epoch": 0.15915395284327324,
      "grad_norm": 3.174394130706787,
      "learning_rate": 1e-06,
      "loss": 0.2083,
      "step": 459
    },
    {
      "epoch": 0.15950069348127602,
      "grad_norm": 1.8774772882461548,
      "learning_rate": 1e-06,
      "loss": 0.1937,
      "step": 460
    },
    {
      "epoch": 0.1598474341192788,
      "grad_norm": 1.7514275312423706,
      "learning_rate": 1e-06,
      "loss": 0.1865,
      "step": 461
    },
    {
      "epoch": 0.16019417475728157,
      "grad_norm": 2.4523355960845947,
      "learning_rate": 1e-06,
      "loss": 0.1719,
      "step": 462
    },
    {
      "epoch": 0.16054091539528434,
      "grad_norm": 5.636368274688721,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 463
    },
    {
      "epoch": 0.1608876560332871,
      "grad_norm": 2.248699426651001,
      "learning_rate": 1e-06,
      "loss": 0.2034,
      "step": 464
    },
    {
      "epoch": 0.1612343966712899,
      "grad_norm": 1.9587477445602417,
      "learning_rate": 1e-06,
      "loss": 0.1991,
      "step": 465
    },
    {
      "epoch": 0.16158113730929266,
      "grad_norm": 1.5505951642990112,
      "learning_rate": 1e-06,
      "loss": 0.1662,
      "step": 466
    },
    {
      "epoch": 0.16192787794729543,
      "grad_norm": 1.778409719467163,
      "learning_rate": 1e-06,
      "loss": 0.206,
      "step": 467
    },
    {
      "epoch": 0.1622746185852982,
      "grad_norm": 3.1874828338623047,
      "learning_rate": 1e-06,
      "loss": 0.1773,
      "step": 468
    },
    {
      "epoch": 0.16262135922330098,
      "grad_norm": 3.800231695175171,
      "learning_rate": 1e-06,
      "loss": 0.209,
      "step": 469
    },
    {
      "epoch": 0.16296809986130376,
      "grad_norm": 2.2624082565307617,
      "learning_rate": 1e-06,
      "loss": 0.1968,
      "step": 470
    },
    {
      "epoch": 0.16331484049930653,
      "grad_norm": 2.0975418090820312,
      "learning_rate": 1e-06,
      "loss": 0.235,
      "step": 471
    },
    {
      "epoch": 0.1636615811373093,
      "grad_norm": 5.558737277984619,
      "learning_rate": 1e-06,
      "loss": 0.1581,
      "step": 472
    },
    {
      "epoch": 0.16400832177531208,
      "grad_norm": 3.688441514968872,
      "learning_rate": 1e-06,
      "loss": 0.2038,
      "step": 473
    },
    {
      "epoch": 0.16435506241331485,
      "grad_norm": 1.7553553581237793,
      "learning_rate": 1e-06,
      "loss": 0.178,
      "step": 474
    },
    {
      "epoch": 0.16470180305131762,
      "grad_norm": 4.0227952003479,
      "learning_rate": 1e-06,
      "loss": 0.1732,
      "step": 475
    },
    {
      "epoch": 0.1650485436893204,
      "grad_norm": 1.9997745752334595,
      "learning_rate": 1e-06,
      "loss": 0.2068,
      "step": 476
    },
    {
      "epoch": 0.16539528432732317,
      "grad_norm": 3.5058529376983643,
      "learning_rate": 1e-06,
      "loss": 0.2066,
      "step": 477
    },
    {
      "epoch": 0.16574202496532595,
      "grad_norm": 3.745401382446289,
      "learning_rate": 1e-06,
      "loss": 0.2022,
      "step": 478
    },
    {
      "epoch": 0.16608876560332872,
      "grad_norm": 1.898235559463501,
      "learning_rate": 1e-06,
      "loss": 0.1787,
      "step": 479
    },
    {
      "epoch": 0.1664355062413315,
      "grad_norm": 2.2373104095458984,
      "learning_rate": 1e-06,
      "loss": 0.202,
      "step": 480
    },
    {
      "epoch": 0.16678224687933427,
      "grad_norm": 3.4441022872924805,
      "learning_rate": 1e-06,
      "loss": 0.2223,
      "step": 481
    },
    {
      "epoch": 0.16712898751733704,
      "grad_norm": 2.0082736015319824,
      "learning_rate": 1e-06,
      "loss": 0.1436,
      "step": 482
    },
    {
      "epoch": 0.16747572815533981,
      "grad_norm": 7.100375175476074,
      "learning_rate": 1e-06,
      "loss": 0.2622,
      "step": 483
    },
    {
      "epoch": 0.1678224687933426,
      "grad_norm": 3.85292649269104,
      "learning_rate": 1e-06,
      "loss": 0.2141,
      "step": 484
    },
    {
      "epoch": 0.16816920943134536,
      "grad_norm": 2.212933301925659,
      "learning_rate": 1e-06,
      "loss": 0.2052,
      "step": 485
    },
    {
      "epoch": 0.16851595006934814,
      "grad_norm": 2.0661942958831787,
      "learning_rate": 1e-06,
      "loss": 0.1676,
      "step": 486
    },
    {
      "epoch": 0.1688626907073509,
      "grad_norm": 4.262602806091309,
      "learning_rate": 1e-06,
      "loss": 0.2382,
      "step": 487
    },
    {
      "epoch": 0.16920943134535368,
      "grad_norm": 2.734023094177246,
      "learning_rate": 1e-06,
      "loss": 0.1833,
      "step": 488
    },
    {
      "epoch": 0.16955617198335646,
      "grad_norm": 1.9461660385131836,
      "learning_rate": 1e-06,
      "loss": 0.1833,
      "step": 489
    },
    {
      "epoch": 0.16990291262135923,
      "grad_norm": 2.6612391471862793,
      "learning_rate": 1e-06,
      "loss": 0.1996,
      "step": 490
    },
    {
      "epoch": 0.170249653259362,
      "grad_norm": 2.2470898628234863,
      "learning_rate": 1e-06,
      "loss": 0.1882,
      "step": 491
    },
    {
      "epoch": 0.17059639389736478,
      "grad_norm": 4.1081743240356445,
      "learning_rate": 1e-06,
      "loss": 0.1783,
      "step": 492
    },
    {
      "epoch": 0.17094313453536755,
      "grad_norm": 2.1462888717651367,
      "learning_rate": 1e-06,
      "loss": 0.2158,
      "step": 493
    },
    {
      "epoch": 0.17128987517337033,
      "grad_norm": 1.8906127214431763,
      "learning_rate": 1e-06,
      "loss": 0.196,
      "step": 494
    },
    {
      "epoch": 0.1716366158113731,
      "grad_norm": 3.4147090911865234,
      "learning_rate": 1e-06,
      "loss": 0.2561,
      "step": 495
    },
    {
      "epoch": 0.17198335644937587,
      "grad_norm": 3.48602032661438,
      "learning_rate": 1e-06,
      "loss": 0.17,
      "step": 496
    },
    {
      "epoch": 0.17233009708737865,
      "grad_norm": 1.873856544494629,
      "learning_rate": 1e-06,
      "loss": 0.1979,
      "step": 497
    },
    {
      "epoch": 0.17267683772538142,
      "grad_norm": 3.3756508827209473,
      "learning_rate": 1e-06,
      "loss": 0.1932,
      "step": 498
    },
    {
      "epoch": 0.1730235783633842,
      "grad_norm": 1.7137960195541382,
      "learning_rate": 1e-06,
      "loss": 0.1957,
      "step": 499
    },
    {
      "epoch": 0.17337031900138697,
      "grad_norm": 2.221604347229004,
      "learning_rate": 1e-06,
      "loss": 0.2075,
      "step": 500
    },
    {
      "epoch": 0.17371705963938974,
      "grad_norm": 1.7929978370666504,
      "learning_rate": 1e-06,
      "loss": 0.179,
      "step": 501
    },
    {
      "epoch": 0.17406380027739252,
      "grad_norm": 2.439054250717163,
      "learning_rate": 1e-06,
      "loss": 0.1962,
      "step": 502
    },
    {
      "epoch": 0.1744105409153953,
      "grad_norm": 2.1946470737457275,
      "learning_rate": 1e-06,
      "loss": 0.1734,
      "step": 503
    },
    {
      "epoch": 0.17475728155339806,
      "grad_norm": 2.1295015811920166,
      "learning_rate": 1e-06,
      "loss": 0.1943,
      "step": 504
    },
    {
      "epoch": 0.17510402219140084,
      "grad_norm": 3.222607135772705,
      "learning_rate": 1e-06,
      "loss": 0.192,
      "step": 505
    },
    {
      "epoch": 0.1754507628294036,
      "grad_norm": 3.0692672729492188,
      "learning_rate": 1e-06,
      "loss": 0.1785,
      "step": 506
    },
    {
      "epoch": 0.17579750346740639,
      "grad_norm": 3.299917459487915,
      "learning_rate": 1e-06,
      "loss": 0.1768,
      "step": 507
    },
    {
      "epoch": 0.17614424410540916,
      "grad_norm": 2.0762133598327637,
      "learning_rate": 1e-06,
      "loss": 0.1948,
      "step": 508
    },
    {
      "epoch": 0.17649098474341193,
      "grad_norm": 2.0623905658721924,
      "learning_rate": 1e-06,
      "loss": 0.2167,
      "step": 509
    },
    {
      "epoch": 0.1768377253814147,
      "grad_norm": 2.0946035385131836,
      "learning_rate": 1e-06,
      "loss": 0.2028,
      "step": 510
    },
    {
      "epoch": 0.17718446601941748,
      "grad_norm": 1.861733078956604,
      "learning_rate": 1e-06,
      "loss": 0.2162,
      "step": 511
    },
    {
      "epoch": 0.17753120665742025,
      "grad_norm": 2.4840567111968994,
      "learning_rate": 1e-06,
      "loss": 0.1863,
      "step": 512
    },
    {
      "epoch": 0.17787794729542303,
      "grad_norm": 1.8076410293579102,
      "learning_rate": 1e-06,
      "loss": 0.2008,
      "step": 513
    },
    {
      "epoch": 0.1782246879334258,
      "grad_norm": 2.526137351989746,
      "learning_rate": 1e-06,
      "loss": 0.1904,
      "step": 514
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 1.856165885925293,
      "learning_rate": 1e-06,
      "loss": 0.2123,
      "step": 515
    },
    {
      "epoch": 0.17891816920943135,
      "grad_norm": 3.9912686347961426,
      "learning_rate": 1e-06,
      "loss": 0.1335,
      "step": 516
    },
    {
      "epoch": 0.17926490984743412,
      "grad_norm": 2.328289270401001,
      "learning_rate": 1e-06,
      "loss": 0.2084,
      "step": 517
    },
    {
      "epoch": 0.1796116504854369,
      "grad_norm": 1.7976534366607666,
      "learning_rate": 1e-06,
      "loss": 0.1879,
      "step": 518
    },
    {
      "epoch": 0.17995839112343967,
      "grad_norm": 2.629750967025757,
      "learning_rate": 1e-06,
      "loss": 0.2154,
      "step": 519
    },
    {
      "epoch": 0.18030513176144244,
      "grad_norm": 2.200639009475708,
      "learning_rate": 1e-06,
      "loss": 0.1987,
      "step": 520
    },
    {
      "epoch": 0.18065187239944522,
      "grad_norm": 2.636253833770752,
      "learning_rate": 1e-06,
      "loss": 0.2092,
      "step": 521
    },
    {
      "epoch": 0.180998613037448,
      "grad_norm": 2.329237222671509,
      "learning_rate": 1e-06,
      "loss": 0.1865,
      "step": 522
    },
    {
      "epoch": 0.18134535367545077,
      "grad_norm": 3.2682480812072754,
      "learning_rate": 1e-06,
      "loss": 0.2204,
      "step": 523
    },
    {
      "epoch": 0.18169209431345354,
      "grad_norm": 2.117788791656494,
      "learning_rate": 1e-06,
      "loss": 0.1871,
      "step": 524
    },
    {
      "epoch": 0.1820388349514563,
      "grad_norm": 1.9039613008499146,
      "learning_rate": 1e-06,
      "loss": 0.1856,
      "step": 525
    },
    {
      "epoch": 0.1823855755894591,
      "grad_norm": 2.3658318519592285,
      "learning_rate": 1e-06,
      "loss": 0.1799,
      "step": 526
    },
    {
      "epoch": 0.18273231622746186,
      "grad_norm": 3.818437337875366,
      "learning_rate": 1e-06,
      "loss": 0.1944,
      "step": 527
    },
    {
      "epoch": 0.18307905686546463,
      "grad_norm": 2.8354179859161377,
      "learning_rate": 1e-06,
      "loss": 0.1592,
      "step": 528
    },
    {
      "epoch": 0.1834257975034674,
      "grad_norm": 1.872366189956665,
      "learning_rate": 1e-06,
      "loss": 0.1782,
      "step": 529
    },
    {
      "epoch": 0.18377253814147018,
      "grad_norm": 2.1778335571289062,
      "learning_rate": 1e-06,
      "loss": 0.1787,
      "step": 530
    },
    {
      "epoch": 0.18411927877947296,
      "grad_norm": 2.6627323627471924,
      "learning_rate": 1e-06,
      "loss": 0.1775,
      "step": 531
    },
    {
      "epoch": 0.18446601941747573,
      "grad_norm": 2.1031336784362793,
      "learning_rate": 1e-06,
      "loss": 0.2051,
      "step": 532
    },
    {
      "epoch": 0.1848127600554785,
      "grad_norm": 4.2585272789001465,
      "learning_rate": 1e-06,
      "loss": 0.2088,
      "step": 533
    },
    {
      "epoch": 0.18515950069348128,
      "grad_norm": 3.1983723640441895,
      "learning_rate": 1e-06,
      "loss": 0.177,
      "step": 534
    },
    {
      "epoch": 0.18550624133148405,
      "grad_norm": 2.121945381164551,
      "learning_rate": 1e-06,
      "loss": 0.212,
      "step": 535
    },
    {
      "epoch": 0.18585298196948682,
      "grad_norm": 4.327321529388428,
      "learning_rate": 1e-06,
      "loss": 0.2301,
      "step": 536
    },
    {
      "epoch": 0.1861997226074896,
      "grad_norm": 3.546032190322876,
      "learning_rate": 1e-06,
      "loss": 0.2165,
      "step": 537
    },
    {
      "epoch": 0.18654646324549237,
      "grad_norm": 2.3898680210113525,
      "learning_rate": 1e-06,
      "loss": 0.1892,
      "step": 538
    },
    {
      "epoch": 0.18689320388349515,
      "grad_norm": 2.9088211059570312,
      "learning_rate": 1e-06,
      "loss": 0.2018,
      "step": 539
    },
    {
      "epoch": 0.18723994452149792,
      "grad_norm": 2.1512014865875244,
      "learning_rate": 1e-06,
      "loss": 0.2037,
      "step": 540
    },
    {
      "epoch": 0.1875866851595007,
      "grad_norm": 2.608860731124878,
      "learning_rate": 1e-06,
      "loss": 0.1905,
      "step": 541
    },
    {
      "epoch": 0.18793342579750347,
      "grad_norm": 2.0679538249969482,
      "learning_rate": 1e-06,
      "loss": 0.1969,
      "step": 542
    },
    {
      "epoch": 0.18828016643550624,
      "grad_norm": 2.928675889968872,
      "learning_rate": 1e-06,
      "loss": 0.2332,
      "step": 543
    },
    {
      "epoch": 0.18862690707350901,
      "grad_norm": 6.560424327850342,
      "learning_rate": 1e-06,
      "loss": 0.2278,
      "step": 544
    },
    {
      "epoch": 0.1889736477115118,
      "grad_norm": 1.9088329076766968,
      "learning_rate": 1e-06,
      "loss": 0.2027,
      "step": 545
    },
    {
      "epoch": 0.18932038834951456,
      "grad_norm": 2.4593074321746826,
      "learning_rate": 1e-06,
      "loss": 0.1856,
      "step": 546
    },
    {
      "epoch": 0.18966712898751734,
      "grad_norm": 4.2560858726501465,
      "learning_rate": 1e-06,
      "loss": 0.1764,
      "step": 547
    },
    {
      "epoch": 0.1900138696255201,
      "grad_norm": 1.7965539693832397,
      "learning_rate": 1e-06,
      "loss": 0.1932,
      "step": 548
    },
    {
      "epoch": 0.19036061026352288,
      "grad_norm": 1.9651305675506592,
      "learning_rate": 1e-06,
      "loss": 0.2102,
      "step": 549
    },
    {
      "epoch": 0.19070735090152566,
      "grad_norm": 1.5734628438949585,
      "learning_rate": 1e-06,
      "loss": 0.1769,
      "step": 550
    },
    {
      "epoch": 0.19105409153952843,
      "grad_norm": 3.6712100505828857,
      "learning_rate": 1e-06,
      "loss": 0.2034,
      "step": 551
    },
    {
      "epoch": 0.1914008321775312,
      "grad_norm": 4.593487739562988,
      "learning_rate": 1e-06,
      "loss": 0.1643,
      "step": 552
    },
    {
      "epoch": 0.19174757281553398,
      "grad_norm": 2.358043909072876,
      "learning_rate": 1e-06,
      "loss": 0.1652,
      "step": 553
    },
    {
      "epoch": 0.19209431345353675,
      "grad_norm": 2.64416241645813,
      "learning_rate": 1e-06,
      "loss": 0.239,
      "step": 554
    },
    {
      "epoch": 0.19244105409153953,
      "grad_norm": 4.924618721008301,
      "learning_rate": 1e-06,
      "loss": 0.1842,
      "step": 555
    },
    {
      "epoch": 0.1927877947295423,
      "grad_norm": 2.018336534500122,
      "learning_rate": 1e-06,
      "loss": 0.2024,
      "step": 556
    },
    {
      "epoch": 0.19313453536754507,
      "grad_norm": 2.0920143127441406,
      "learning_rate": 1e-06,
      "loss": 0.2031,
      "step": 557
    },
    {
      "epoch": 0.19348127600554785,
      "grad_norm": 1.8479957580566406,
      "learning_rate": 1e-06,
      "loss": 0.1916,
      "step": 558
    },
    {
      "epoch": 0.19382801664355062,
      "grad_norm": 5.079397201538086,
      "learning_rate": 1e-06,
      "loss": 0.2203,
      "step": 559
    },
    {
      "epoch": 0.1941747572815534,
      "grad_norm": 4.152866363525391,
      "learning_rate": 1e-06,
      "loss": 0.1936,
      "step": 560
    },
    {
      "epoch": 0.19452149791955617,
      "grad_norm": 1.6684824228286743,
      "learning_rate": 1e-06,
      "loss": 0.163,
      "step": 561
    },
    {
      "epoch": 0.19486823855755894,
      "grad_norm": 2.2925472259521484,
      "learning_rate": 1e-06,
      "loss": 0.1826,
      "step": 562
    },
    {
      "epoch": 0.19521497919556172,
      "grad_norm": 1.73776113986969,
      "learning_rate": 1e-06,
      "loss": 0.175,
      "step": 563
    },
    {
      "epoch": 0.1955617198335645,
      "grad_norm": 6.04637336730957,
      "learning_rate": 1e-06,
      "loss": 0.2135,
      "step": 564
    },
    {
      "epoch": 0.19590846047156726,
      "grad_norm": 2.765383005142212,
      "learning_rate": 1e-06,
      "loss": 0.1885,
      "step": 565
    },
    {
      "epoch": 0.19625520110957004,
      "grad_norm": 2.205034017562866,
      "learning_rate": 1e-06,
      "loss": 0.1965,
      "step": 566
    },
    {
      "epoch": 0.1966019417475728,
      "grad_norm": 2.0300064086914062,
      "learning_rate": 1e-06,
      "loss": 0.206,
      "step": 567
    },
    {
      "epoch": 0.19694868238557559,
      "grad_norm": 1.9674218893051147,
      "learning_rate": 1e-06,
      "loss": 0.162,
      "step": 568
    },
    {
      "epoch": 0.19729542302357836,
      "grad_norm": 2.875797986984253,
      "learning_rate": 1e-06,
      "loss": 0.2251,
      "step": 569
    },
    {
      "epoch": 0.19764216366158113,
      "grad_norm": 1.855223536491394,
      "learning_rate": 1e-06,
      "loss": 0.2211,
      "step": 570
    },
    {
      "epoch": 0.1979889042995839,
      "grad_norm": 2.8994317054748535,
      "learning_rate": 1e-06,
      "loss": 0.206,
      "step": 571
    },
    {
      "epoch": 0.19833564493758668,
      "grad_norm": 2.2230451107025146,
      "learning_rate": 1e-06,
      "loss": 0.1947,
      "step": 572
    },
    {
      "epoch": 0.19868238557558945,
      "grad_norm": 5.828412055969238,
      "learning_rate": 1e-06,
      "loss": 0.21,
      "step": 573
    },
    {
      "epoch": 0.19902912621359223,
      "grad_norm": 2.926950454711914,
      "learning_rate": 1e-06,
      "loss": 0.2018,
      "step": 574
    },
    {
      "epoch": 0.199375866851595,
      "grad_norm": 3.7960805892944336,
      "learning_rate": 1e-06,
      "loss": 0.188,
      "step": 575
    },
    {
      "epoch": 0.19972260748959778,
      "grad_norm": 1.8869881629943848,
      "learning_rate": 1e-06,
      "loss": 0.2025,
      "step": 576
    },
    {
      "epoch": 0.20006934812760055,
      "grad_norm": 3.9236044883728027,
      "learning_rate": 1e-06,
      "loss": 0.1571,
      "step": 577
    },
    {
      "epoch": 0.20041608876560332,
      "grad_norm": 1.937950849533081,
      "learning_rate": 1e-06,
      "loss": 0.1944,
      "step": 578
    },
    {
      "epoch": 0.2007628294036061,
      "grad_norm": 2.139031410217285,
      "learning_rate": 1e-06,
      "loss": 0.1991,
      "step": 579
    },
    {
      "epoch": 0.20110957004160887,
      "grad_norm": 4.336677551269531,
      "learning_rate": 1e-06,
      "loss": 0.2159,
      "step": 580
    },
    {
      "epoch": 0.20145631067961164,
      "grad_norm": 3.819993495941162,
      "learning_rate": 1e-06,
      "loss": 0.1833,
      "step": 581
    },
    {
      "epoch": 0.20180305131761442,
      "grad_norm": 1.7961671352386475,
      "learning_rate": 1e-06,
      "loss": 0.1807,
      "step": 582
    },
    {
      "epoch": 0.2021497919556172,
      "grad_norm": 2.0944950580596924,
      "learning_rate": 1e-06,
      "loss": 0.2336,
      "step": 583
    },
    {
      "epoch": 0.20249653259361997,
      "grad_norm": 2.7267916202545166,
      "learning_rate": 1e-06,
      "loss": 0.2282,
      "step": 584
    },
    {
      "epoch": 0.20284327323162274,
      "grad_norm": 1.8287086486816406,
      "learning_rate": 1e-06,
      "loss": 0.1684,
      "step": 585
    },
    {
      "epoch": 0.2031900138696255,
      "grad_norm": 3.0344362258911133,
      "learning_rate": 1e-06,
      "loss": 0.1965,
      "step": 586
    },
    {
      "epoch": 0.2035367545076283,
      "grad_norm": 4.7607340812683105,
      "learning_rate": 1e-06,
      "loss": 0.193,
      "step": 587
    },
    {
      "epoch": 0.20388349514563106,
      "grad_norm": 1.6755648851394653,
      "learning_rate": 1e-06,
      "loss": 0.1764,
      "step": 588
    },
    {
      "epoch": 0.20423023578363383,
      "grad_norm": 2.2113454341888428,
      "learning_rate": 1e-06,
      "loss": 0.1614,
      "step": 589
    },
    {
      "epoch": 0.2045769764216366,
      "grad_norm": 2.667480945587158,
      "learning_rate": 1e-06,
      "loss": 0.173,
      "step": 590
    },
    {
      "epoch": 0.20492371705963938,
      "grad_norm": 2.3278677463531494,
      "learning_rate": 1e-06,
      "loss": 0.1876,
      "step": 591
    },
    {
      "epoch": 0.20527045769764216,
      "grad_norm": 3.0954275131225586,
      "learning_rate": 1e-06,
      "loss": 0.219,
      "step": 592
    },
    {
      "epoch": 0.20561719833564493,
      "grad_norm": 1.8866382837295532,
      "learning_rate": 1e-06,
      "loss": 0.1709,
      "step": 593
    },
    {
      "epoch": 0.2059639389736477,
      "grad_norm": 2.9739086627960205,
      "learning_rate": 1e-06,
      "loss": 0.1719,
      "step": 594
    },
    {
      "epoch": 0.20631067961165048,
      "grad_norm": 1.8454285860061646,
      "learning_rate": 1e-06,
      "loss": 0.1959,
      "step": 595
    },
    {
      "epoch": 0.20665742024965325,
      "grad_norm": 1.798344373703003,
      "learning_rate": 1e-06,
      "loss": 0.1655,
      "step": 596
    },
    {
      "epoch": 0.20700416088765602,
      "grad_norm": 3.577150821685791,
      "learning_rate": 1e-06,
      "loss": 0.2016,
      "step": 597
    },
    {
      "epoch": 0.2073509015256588,
      "grad_norm": 1.9027544260025024,
      "learning_rate": 1e-06,
      "loss": 0.1939,
      "step": 598
    },
    {
      "epoch": 0.20769764216366157,
      "grad_norm": 4.212296962738037,
      "learning_rate": 1e-06,
      "loss": 0.1663,
      "step": 599
    },
    {
      "epoch": 0.20804438280166435,
      "grad_norm": 2.9982285499572754,
      "learning_rate": 1e-06,
      "loss": 0.2316,
      "step": 600
    },
    {
      "epoch": 0.20839112343966712,
      "grad_norm": 2.4148337841033936,
      "learning_rate": 1e-06,
      "loss": 0.194,
      "step": 601
    },
    {
      "epoch": 0.2087378640776699,
      "grad_norm": 5.278850078582764,
      "learning_rate": 1e-06,
      "loss": 0.1835,
      "step": 602
    },
    {
      "epoch": 0.20908460471567267,
      "grad_norm": 2.83439564704895,
      "learning_rate": 1e-06,
      "loss": 0.1839,
      "step": 603
    },
    {
      "epoch": 0.20943134535367544,
      "grad_norm": 2.2179036140441895,
      "learning_rate": 1e-06,
      "loss": 0.2021,
      "step": 604
    },
    {
      "epoch": 0.20977808599167821,
      "grad_norm": 5.919695854187012,
      "learning_rate": 1e-06,
      "loss": 0.2135,
      "step": 605
    },
    {
      "epoch": 0.210124826629681,
      "grad_norm": 1.5985592603683472,
      "learning_rate": 1e-06,
      "loss": 0.1672,
      "step": 606
    },
    {
      "epoch": 0.21047156726768376,
      "grad_norm": 3.6065750122070312,
      "learning_rate": 1e-06,
      "loss": 0.2133,
      "step": 607
    },
    {
      "epoch": 0.21081830790568654,
      "grad_norm": 3.56968355178833,
      "learning_rate": 1e-06,
      "loss": 0.1867,
      "step": 608
    },
    {
      "epoch": 0.2111650485436893,
      "grad_norm": 2.662672758102417,
      "learning_rate": 1e-06,
      "loss": 0.1773,
      "step": 609
    },
    {
      "epoch": 0.21151178918169208,
      "grad_norm": 1.9275373220443726,
      "learning_rate": 1e-06,
      "loss": 0.1926,
      "step": 610
    },
    {
      "epoch": 0.21185852981969486,
      "grad_norm": 4.133257865905762,
      "learning_rate": 1e-06,
      "loss": 0.1931,
      "step": 611
    },
    {
      "epoch": 0.21220527045769763,
      "grad_norm": 1.791725516319275,
      "learning_rate": 1e-06,
      "loss": 0.1874,
      "step": 612
    },
    {
      "epoch": 0.2125520110957004,
      "grad_norm": 2.296025514602661,
      "learning_rate": 1e-06,
      "loss": 0.1968,
      "step": 613
    },
    {
      "epoch": 0.21289875173370318,
      "grad_norm": 1.7383009195327759,
      "learning_rate": 1e-06,
      "loss": 0.1597,
      "step": 614
    },
    {
      "epoch": 0.21324549237170595,
      "grad_norm": 2.899918556213379,
      "learning_rate": 1e-06,
      "loss": 0.1843,
      "step": 615
    },
    {
      "epoch": 0.21359223300970873,
      "grad_norm": 7.615829944610596,
      "learning_rate": 1e-06,
      "loss": 0.2117,
      "step": 616
    },
    {
      "epoch": 0.2139389736477115,
      "grad_norm": 1.9387335777282715,
      "learning_rate": 1e-06,
      "loss": 0.1798,
      "step": 617
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 4.627980709075928,
      "learning_rate": 1e-06,
      "loss": 0.1829,
      "step": 618
    },
    {
      "epoch": 0.21463245492371705,
      "grad_norm": 7.2196526527404785,
      "learning_rate": 1e-06,
      "loss": 0.2323,
      "step": 619
    },
    {
      "epoch": 0.21497919556171982,
      "grad_norm": 1.4321414232254028,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 620
    },
    {
      "epoch": 0.2153259361997226,
      "grad_norm": 1.6646205186843872,
      "learning_rate": 1e-06,
      "loss": 0.1922,
      "step": 621
    },
    {
      "epoch": 0.21567267683772537,
      "grad_norm": 1.867525577545166,
      "learning_rate": 1e-06,
      "loss": 0.1995,
      "step": 622
    },
    {
      "epoch": 0.21601941747572814,
      "grad_norm": 2.072211503982544,
      "learning_rate": 1e-06,
      "loss": 0.1848,
      "step": 623
    },
    {
      "epoch": 0.21636615811373092,
      "grad_norm": 3.1472299098968506,
      "learning_rate": 1e-06,
      "loss": 0.17,
      "step": 624
    },
    {
      "epoch": 0.2167128987517337,
      "grad_norm": 5.170080184936523,
      "learning_rate": 1e-06,
      "loss": 0.1975,
      "step": 625
    },
    {
      "epoch": 0.21705963938973646,
      "grad_norm": 1.9573615789413452,
      "learning_rate": 1e-06,
      "loss": 0.2022,
      "step": 626
    },
    {
      "epoch": 0.21740638002773924,
      "grad_norm": 4.052754878997803,
      "learning_rate": 1e-06,
      "loss": 0.1595,
      "step": 627
    },
    {
      "epoch": 0.217753120665742,
      "grad_norm": 2.469994306564331,
      "learning_rate": 1e-06,
      "loss": 0.2114,
      "step": 628
    },
    {
      "epoch": 0.21809986130374479,
      "grad_norm": 1.9860141277313232,
      "learning_rate": 1e-06,
      "loss": 0.1842,
      "step": 629
    },
    {
      "epoch": 0.21844660194174756,
      "grad_norm": 3.6322503089904785,
      "learning_rate": 1e-06,
      "loss": 0.183,
      "step": 630
    },
    {
      "epoch": 0.21879334257975036,
      "grad_norm": 3.5005853176116943,
      "learning_rate": 1e-06,
      "loss": 0.1921,
      "step": 631
    },
    {
      "epoch": 0.21914008321775313,
      "grad_norm": 1.7438961267471313,
      "learning_rate": 1e-06,
      "loss": 0.1883,
      "step": 632
    },
    {
      "epoch": 0.2194868238557559,
      "grad_norm": 2.330274820327759,
      "learning_rate": 1e-06,
      "loss": 0.1777,
      "step": 633
    },
    {
      "epoch": 0.21983356449375868,
      "grad_norm": 2.0113935470581055,
      "learning_rate": 1e-06,
      "loss": 0.198,
      "step": 634
    },
    {
      "epoch": 0.22018030513176146,
      "grad_norm": 2.120816946029663,
      "learning_rate": 1e-06,
      "loss": 0.195,
      "step": 635
    },
    {
      "epoch": 0.22052704576976423,
      "grad_norm": 3.88839054107666,
      "learning_rate": 1e-06,
      "loss": 0.2156,
      "step": 636
    },
    {
      "epoch": 0.220873786407767,
      "grad_norm": 2.330038070678711,
      "learning_rate": 1e-06,
      "loss": 0.1525,
      "step": 637
    },
    {
      "epoch": 0.22122052704576978,
      "grad_norm": 2.0380382537841797,
      "learning_rate": 1e-06,
      "loss": 0.1892,
      "step": 638
    },
    {
      "epoch": 0.22156726768377255,
      "grad_norm": 2.988846778869629,
      "learning_rate": 1e-06,
      "loss": 0.1784,
      "step": 639
    },
    {
      "epoch": 0.22191400832177532,
      "grad_norm": 3.193902015686035,
      "learning_rate": 1e-06,
      "loss": 0.2036,
      "step": 640
    },
    {
      "epoch": 0.2222607489597781,
      "grad_norm": 3.138087511062622,
      "learning_rate": 1e-06,
      "loss": 0.1995,
      "step": 641
    },
    {
      "epoch": 0.22260748959778087,
      "grad_norm": 3.0959274768829346,
      "learning_rate": 1e-06,
      "loss": 0.212,
      "step": 642
    },
    {
      "epoch": 0.22295423023578365,
      "grad_norm": 3.1513280868530273,
      "learning_rate": 1e-06,
      "loss": 0.1928,
      "step": 643
    },
    {
      "epoch": 0.22330097087378642,
      "grad_norm": 3.7054238319396973,
      "learning_rate": 1e-06,
      "loss": 0.1547,
      "step": 644
    },
    {
      "epoch": 0.2236477115117892,
      "grad_norm": 1.7885769605636597,
      "learning_rate": 1e-06,
      "loss": 0.1764,
      "step": 645
    },
    {
      "epoch": 0.22399445214979197,
      "grad_norm": 4.171994686126709,
      "learning_rate": 1e-06,
      "loss": 0.1813,
      "step": 646
    },
    {
      "epoch": 0.22434119278779474,
      "grad_norm": 2.313520669937134,
      "learning_rate": 1e-06,
      "loss": 0.2098,
      "step": 647
    },
    {
      "epoch": 0.22468793342579751,
      "grad_norm": 3.154613971710205,
      "learning_rate": 1e-06,
      "loss": 0.1776,
      "step": 648
    },
    {
      "epoch": 0.2250346740638003,
      "grad_norm": 5.905630111694336,
      "learning_rate": 1e-06,
      "loss": 0.1632,
      "step": 649
    },
    {
      "epoch": 0.22538141470180306,
      "grad_norm": 3.555774211883545,
      "learning_rate": 1e-06,
      "loss": 0.1747,
      "step": 650
    },
    {
      "epoch": 0.22572815533980584,
      "grad_norm": 2.0930142402648926,
      "learning_rate": 1e-06,
      "loss": 0.1962,
      "step": 651
    },
    {
      "epoch": 0.2260748959778086,
      "grad_norm": 3.322031021118164,
      "learning_rate": 1e-06,
      "loss": 0.1887,
      "step": 652
    },
    {
      "epoch": 0.22642163661581138,
      "grad_norm": 3.3001041412353516,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 653
    },
    {
      "epoch": 0.22676837725381416,
      "grad_norm": 2.0050160884857178,
      "learning_rate": 1e-06,
      "loss": 0.1696,
      "step": 654
    },
    {
      "epoch": 0.22711511789181693,
      "grad_norm": 2.4040985107421875,
      "learning_rate": 1e-06,
      "loss": 0.1937,
      "step": 655
    },
    {
      "epoch": 0.2274618585298197,
      "grad_norm": 2.627812147140503,
      "learning_rate": 1e-06,
      "loss": 0.1667,
      "step": 656
    },
    {
      "epoch": 0.22780859916782248,
      "grad_norm": 4.34411096572876,
      "learning_rate": 1e-06,
      "loss": 0.1815,
      "step": 657
    },
    {
      "epoch": 0.22815533980582525,
      "grad_norm": 4.466109275817871,
      "learning_rate": 1e-06,
      "loss": 0.1798,
      "step": 658
    },
    {
      "epoch": 0.22850208044382803,
      "grad_norm": 1.537264347076416,
      "learning_rate": 1e-06,
      "loss": 0.1758,
      "step": 659
    },
    {
      "epoch": 0.2288488210818308,
      "grad_norm": 3.8718326091766357,
      "learning_rate": 1e-06,
      "loss": 0.1796,
      "step": 660
    },
    {
      "epoch": 0.22919556171983357,
      "grad_norm": 3.2782208919525146,
      "learning_rate": 1e-06,
      "loss": 0.158,
      "step": 661
    },
    {
      "epoch": 0.22954230235783635,
      "grad_norm": 1.7151904106140137,
      "learning_rate": 1e-06,
      "loss": 0.1769,
      "step": 662
    },
    {
      "epoch": 0.22988904299583912,
      "grad_norm": 1.8331800699234009,
      "learning_rate": 1e-06,
      "loss": 0.1671,
      "step": 663
    },
    {
      "epoch": 0.2302357836338419,
      "grad_norm": 2.3678629398345947,
      "learning_rate": 1e-06,
      "loss": 0.164,
      "step": 664
    },
    {
      "epoch": 0.23058252427184467,
      "grad_norm": 1.5758439302444458,
      "learning_rate": 1e-06,
      "loss": 0.1689,
      "step": 665
    },
    {
      "epoch": 0.23092926490984744,
      "grad_norm": 2.456174850463867,
      "learning_rate": 1e-06,
      "loss": 0.1681,
      "step": 666
    },
    {
      "epoch": 0.23127600554785022,
      "grad_norm": 1.6386404037475586,
      "learning_rate": 1e-06,
      "loss": 0.1722,
      "step": 667
    },
    {
      "epoch": 0.231622746185853,
      "grad_norm": 1.6972070932388306,
      "learning_rate": 1e-06,
      "loss": 0.1699,
      "step": 668
    },
    {
      "epoch": 0.23196948682385576,
      "grad_norm": 2.991753101348877,
      "learning_rate": 1e-06,
      "loss": 0.194,
      "step": 669
    },
    {
      "epoch": 0.23231622746185854,
      "grad_norm": 1.5393069982528687,
      "learning_rate": 1e-06,
      "loss": 0.1717,
      "step": 670
    },
    {
      "epoch": 0.2326629680998613,
      "grad_norm": 2.0264406204223633,
      "learning_rate": 1e-06,
      "loss": 0.1961,
      "step": 671
    },
    {
      "epoch": 0.23300970873786409,
      "grad_norm": 4.824713706970215,
      "learning_rate": 1e-06,
      "loss": 0.1516,
      "step": 672
    },
    {
      "epoch": 0.23335644937586686,
      "grad_norm": 1.823652982711792,
      "learning_rate": 1e-06,
      "loss": 0.1913,
      "step": 673
    },
    {
      "epoch": 0.23370319001386963,
      "grad_norm": 2.0635838508605957,
      "learning_rate": 1e-06,
      "loss": 0.1702,
      "step": 674
    },
    {
      "epoch": 0.2340499306518724,
      "grad_norm": 1.7453879117965698,
      "learning_rate": 1e-06,
      "loss": 0.171,
      "step": 675
    },
    {
      "epoch": 0.23439667128987518,
      "grad_norm": 3.4135522842407227,
      "learning_rate": 1e-06,
      "loss": 0.2,
      "step": 676
    },
    {
      "epoch": 0.23474341192787795,
      "grad_norm": 4.439614295959473,
      "learning_rate": 1e-06,
      "loss": 0.1779,
      "step": 677
    },
    {
      "epoch": 0.23509015256588073,
      "grad_norm": 1.9684357643127441,
      "learning_rate": 1e-06,
      "loss": 0.1828,
      "step": 678
    },
    {
      "epoch": 0.2354368932038835,
      "grad_norm": 1.9556607007980347,
      "learning_rate": 1e-06,
      "loss": 0.1967,
      "step": 679
    },
    {
      "epoch": 0.23578363384188628,
      "grad_norm": 3.652463436126709,
      "learning_rate": 1e-06,
      "loss": 0.2005,
      "step": 680
    },
    {
      "epoch": 0.23613037447988905,
      "grad_norm": 1.9398771524429321,
      "learning_rate": 1e-06,
      "loss": 0.158,
      "step": 681
    },
    {
      "epoch": 0.23647711511789182,
      "grad_norm": 3.950073480606079,
      "learning_rate": 1e-06,
      "loss": 0.192,
      "step": 682
    },
    {
      "epoch": 0.2368238557558946,
      "grad_norm": 3.032841205596924,
      "learning_rate": 1e-06,
      "loss": 0.2069,
      "step": 683
    },
    {
      "epoch": 0.23717059639389737,
      "grad_norm": 1.8892643451690674,
      "learning_rate": 1e-06,
      "loss": 0.1994,
      "step": 684
    },
    {
      "epoch": 0.23751733703190014,
      "grad_norm": 2.4548637866973877,
      "learning_rate": 1e-06,
      "loss": 0.1818,
      "step": 685
    },
    {
      "epoch": 0.23786407766990292,
      "grad_norm": 1.8674260377883911,
      "learning_rate": 1e-06,
      "loss": 0.1712,
      "step": 686
    },
    {
      "epoch": 0.2382108183079057,
      "grad_norm": 5.607412338256836,
      "learning_rate": 1e-06,
      "loss": 0.205,
      "step": 687
    },
    {
      "epoch": 0.23855755894590847,
      "grad_norm": 1.955690860748291,
      "learning_rate": 1e-06,
      "loss": 0.192,
      "step": 688
    },
    {
      "epoch": 0.23890429958391124,
      "grad_norm": 3.3477094173431396,
      "learning_rate": 1e-06,
      "loss": 0.1538,
      "step": 689
    },
    {
      "epoch": 0.239251040221914,
      "grad_norm": 4.284628868103027,
      "learning_rate": 1e-06,
      "loss": 0.1592,
      "step": 690
    },
    {
      "epoch": 0.2395977808599168,
      "grad_norm": 2.2527334690093994,
      "learning_rate": 1e-06,
      "loss": 0.1675,
      "step": 691
    },
    {
      "epoch": 0.23994452149791956,
      "grad_norm": 2.17332124710083,
      "learning_rate": 1e-06,
      "loss": 0.1988,
      "step": 692
    },
    {
      "epoch": 0.24029126213592233,
      "grad_norm": 3.59987211227417,
      "learning_rate": 1e-06,
      "loss": 0.2038,
      "step": 693
    },
    {
      "epoch": 0.2406380027739251,
      "grad_norm": 4.105143070220947,
      "learning_rate": 1e-06,
      "loss": 0.2028,
      "step": 694
    },
    {
      "epoch": 0.24098474341192788,
      "grad_norm": 1.9248536825180054,
      "learning_rate": 1e-06,
      "loss": 0.2031,
      "step": 695
    },
    {
      "epoch": 0.24133148404993066,
      "grad_norm": 2.1153578758239746,
      "learning_rate": 1e-06,
      "loss": 0.2164,
      "step": 696
    },
    {
      "epoch": 0.24167822468793343,
      "grad_norm": 3.2150540351867676,
      "learning_rate": 1e-06,
      "loss": 0.1985,
      "step": 697
    },
    {
      "epoch": 0.2420249653259362,
      "grad_norm": 2.063730478286743,
      "learning_rate": 1e-06,
      "loss": 0.1914,
      "step": 698
    },
    {
      "epoch": 0.24237170596393898,
      "grad_norm": 1.997815728187561,
      "learning_rate": 1e-06,
      "loss": 0.1956,
      "step": 699
    },
    {
      "epoch": 0.24271844660194175,
      "grad_norm": 2.1964597702026367,
      "learning_rate": 1e-06,
      "loss": 0.1665,
      "step": 700
    },
    {
      "epoch": 0.24306518723994452,
      "grad_norm": 1.6123956441879272,
      "learning_rate": 1e-06,
      "loss": 0.1647,
      "step": 701
    },
    {
      "epoch": 0.2434119278779473,
      "grad_norm": 4.534806251525879,
      "learning_rate": 1e-06,
      "loss": 0.1932,
      "step": 702
    },
    {
      "epoch": 0.24375866851595007,
      "grad_norm": 3.9443535804748535,
      "learning_rate": 1e-06,
      "loss": 0.1928,
      "step": 703
    },
    {
      "epoch": 0.24410540915395285,
      "grad_norm": 6.974303245544434,
      "learning_rate": 1e-06,
      "loss": 0.1648,
      "step": 704
    },
    {
      "epoch": 0.24445214979195562,
      "grad_norm": 2.646906614303589,
      "learning_rate": 1e-06,
      "loss": 0.2277,
      "step": 705
    },
    {
      "epoch": 0.2447988904299584,
      "grad_norm": 2.9131650924682617,
      "learning_rate": 1e-06,
      "loss": 0.1663,
      "step": 706
    },
    {
      "epoch": 0.24514563106796117,
      "grad_norm": 1.7801005840301514,
      "learning_rate": 1e-06,
      "loss": 0.1656,
      "step": 707
    },
    {
      "epoch": 0.24549237170596394,
      "grad_norm": 2.639249563217163,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 708
    },
    {
      "epoch": 0.24583911234396671,
      "grad_norm": 1.4407970905303955,
      "learning_rate": 1e-06,
      "loss": 0.1702,
      "step": 709
    },
    {
      "epoch": 0.2461858529819695,
      "grad_norm": 1.4568276405334473,
      "learning_rate": 1e-06,
      "loss": 0.1659,
      "step": 710
    },
    {
      "epoch": 0.24653259361997226,
      "grad_norm": 4.625824928283691,
      "learning_rate": 1e-06,
      "loss": 0.2295,
      "step": 711
    },
    {
      "epoch": 0.24687933425797504,
      "grad_norm": 4.630171775817871,
      "learning_rate": 1e-06,
      "loss": 0.1668,
      "step": 712
    },
    {
      "epoch": 0.2472260748959778,
      "grad_norm": 1.7043591737747192,
      "learning_rate": 1e-06,
      "loss": 0.179,
      "step": 713
    },
    {
      "epoch": 0.24757281553398058,
      "grad_norm": 2.682840585708618,
      "learning_rate": 1e-06,
      "loss": 0.1708,
      "step": 714
    },
    {
      "epoch": 0.24791955617198336,
      "grad_norm": 1.8791319131851196,
      "learning_rate": 1e-06,
      "loss": 0.195,
      "step": 715
    },
    {
      "epoch": 0.24826629680998613,
      "grad_norm": 1.7238434553146362,
      "learning_rate": 1e-06,
      "loss": 0.1711,
      "step": 716
    },
    {
      "epoch": 0.2486130374479889,
      "grad_norm": 3.407539129257202,
      "learning_rate": 1e-06,
      "loss": 0.1666,
      "step": 717
    },
    {
      "epoch": 0.24895977808599168,
      "grad_norm": 5.695379257202148,
      "learning_rate": 1e-06,
      "loss": 0.1979,
      "step": 718
    },
    {
      "epoch": 0.24930651872399445,
      "grad_norm": 4.801501274108887,
      "learning_rate": 1e-06,
      "loss": 0.1888,
      "step": 719
    },
    {
      "epoch": 0.24965325936199723,
      "grad_norm": 4.1193389892578125,
      "learning_rate": 1e-06,
      "loss": 0.202,
      "step": 720
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.602652072906494,
      "learning_rate": 1e-06,
      "loss": 0.1926,
      "step": 721
    },
    {
      "epoch": 0.2503467406380028,
      "grad_norm": 2.5733113288879395,
      "learning_rate": 1e-06,
      "loss": 0.1964,
      "step": 722
    },
    {
      "epoch": 0.25069348127600555,
      "grad_norm": 1.5140520334243774,
      "learning_rate": 1e-06,
      "loss": 0.1719,
      "step": 723
    },
    {
      "epoch": 0.2510402219140083,
      "grad_norm": 1.8638113737106323,
      "learning_rate": 1e-06,
      "loss": 0.1745,
      "step": 724
    },
    {
      "epoch": 0.2513869625520111,
      "grad_norm": 1.9770817756652832,
      "learning_rate": 1e-06,
      "loss": 0.2094,
      "step": 725
    },
    {
      "epoch": 0.25173370319001387,
      "grad_norm": 3.411306619644165,
      "learning_rate": 1e-06,
      "loss": 0.1674,
      "step": 726
    },
    {
      "epoch": 0.25208044382801664,
      "grad_norm": 5.473649024963379,
      "learning_rate": 1e-06,
      "loss": 0.1843,
      "step": 727
    },
    {
      "epoch": 0.2524271844660194,
      "grad_norm": 3.618953227996826,
      "learning_rate": 1e-06,
      "loss": 0.1691,
      "step": 728
    },
    {
      "epoch": 0.2527739251040222,
      "grad_norm": 2.7553508281707764,
      "learning_rate": 1e-06,
      "loss": 0.1745,
      "step": 729
    },
    {
      "epoch": 0.25312066574202496,
      "grad_norm": 3.168771266937256,
      "learning_rate": 1e-06,
      "loss": 0.1952,
      "step": 730
    },
    {
      "epoch": 0.25346740638002774,
      "grad_norm": 2.5667014122009277,
      "learning_rate": 1e-06,
      "loss": 0.1685,
      "step": 731
    },
    {
      "epoch": 0.2538141470180305,
      "grad_norm": 2.024369716644287,
      "learning_rate": 1e-06,
      "loss": 0.1782,
      "step": 732
    },
    {
      "epoch": 0.2541608876560333,
      "grad_norm": 3.3991565704345703,
      "learning_rate": 1e-06,
      "loss": 0.1628,
      "step": 733
    },
    {
      "epoch": 0.25450762829403606,
      "grad_norm": 3.4920365810394287,
      "learning_rate": 1e-06,
      "loss": 0.1869,
      "step": 734
    },
    {
      "epoch": 0.25485436893203883,
      "grad_norm": 2.4127650260925293,
      "learning_rate": 1e-06,
      "loss": 0.2054,
      "step": 735
    },
    {
      "epoch": 0.2552011095700416,
      "grad_norm": 2.5283799171447754,
      "learning_rate": 1e-06,
      "loss": 0.2128,
      "step": 736
    },
    {
      "epoch": 0.2555478502080444,
      "grad_norm": 3.1239235401153564,
      "learning_rate": 1e-06,
      "loss": 0.189,
      "step": 737
    },
    {
      "epoch": 0.25589459084604715,
      "grad_norm": 4.53823184967041,
      "learning_rate": 1e-06,
      "loss": 0.2216,
      "step": 738
    },
    {
      "epoch": 0.25624133148404993,
      "grad_norm": 1.9192531108856201,
      "learning_rate": 1e-06,
      "loss": 0.1937,
      "step": 739
    },
    {
      "epoch": 0.2565880721220527,
      "grad_norm": 3.241487979888916,
      "learning_rate": 1e-06,
      "loss": 0.1744,
      "step": 740
    },
    {
      "epoch": 0.2569348127600555,
      "grad_norm": 4.408164024353027,
      "learning_rate": 1e-06,
      "loss": 0.1792,
      "step": 741
    },
    {
      "epoch": 0.25728155339805825,
      "grad_norm": 2.0512523651123047,
      "learning_rate": 1e-06,
      "loss": 0.158,
      "step": 742
    },
    {
      "epoch": 0.257628294036061,
      "grad_norm": 1.6755858659744263,
      "learning_rate": 1e-06,
      "loss": 0.1819,
      "step": 743
    },
    {
      "epoch": 0.2579750346740638,
      "grad_norm": 1.6671578884124756,
      "learning_rate": 1e-06,
      "loss": 0.1634,
      "step": 744
    },
    {
      "epoch": 0.25832177531206657,
      "grad_norm": 2.0156643390655518,
      "learning_rate": 1e-06,
      "loss": 0.169,
      "step": 745
    },
    {
      "epoch": 0.25866851595006934,
      "grad_norm": 3.1058249473571777,
      "learning_rate": 1e-06,
      "loss": 0.1633,
      "step": 746
    },
    {
      "epoch": 0.2590152565880721,
      "grad_norm": 2.4028232097625732,
      "learning_rate": 1e-06,
      "loss": 0.1907,
      "step": 747
    },
    {
      "epoch": 0.2593619972260749,
      "grad_norm": 5.4095354080200195,
      "learning_rate": 1e-06,
      "loss": 0.1564,
      "step": 748
    },
    {
      "epoch": 0.25970873786407767,
      "grad_norm": 2.967435359954834,
      "learning_rate": 1e-06,
      "loss": 0.1811,
      "step": 749
    },
    {
      "epoch": 0.26005547850208044,
      "grad_norm": 2.7029571533203125,
      "learning_rate": 1e-06,
      "loss": 0.2192,
      "step": 750
    },
    {
      "epoch": 0.2604022191400832,
      "grad_norm": 2.9570839405059814,
      "learning_rate": 1e-06,
      "loss": 0.1605,
      "step": 751
    },
    {
      "epoch": 0.260748959778086,
      "grad_norm": 2.031963348388672,
      "learning_rate": 1e-06,
      "loss": 0.2144,
      "step": 752
    },
    {
      "epoch": 0.26109570041608876,
      "grad_norm": 1.6963039636611938,
      "learning_rate": 1e-06,
      "loss": 0.1506,
      "step": 753
    },
    {
      "epoch": 0.26144244105409153,
      "grad_norm": 1.8354779481887817,
      "learning_rate": 1e-06,
      "loss": 0.1736,
      "step": 754
    },
    {
      "epoch": 0.2617891816920943,
      "grad_norm": 2.998814582824707,
      "learning_rate": 1e-06,
      "loss": 0.1888,
      "step": 755
    },
    {
      "epoch": 0.2621359223300971,
      "grad_norm": 1.5606882572174072,
      "learning_rate": 1e-06,
      "loss": 0.1434,
      "step": 756
    },
    {
      "epoch": 0.26248266296809986,
      "grad_norm": 6.14403772354126,
      "learning_rate": 1e-06,
      "loss": 0.1722,
      "step": 757
    },
    {
      "epoch": 0.26282940360610263,
      "grad_norm": 4.8302321434021,
      "learning_rate": 1e-06,
      "loss": 0.1674,
      "step": 758
    },
    {
      "epoch": 0.2631761442441054,
      "grad_norm": 6.868436813354492,
      "learning_rate": 1e-06,
      "loss": 0.2065,
      "step": 759
    },
    {
      "epoch": 0.2635228848821082,
      "grad_norm": 5.897651672363281,
      "learning_rate": 1e-06,
      "loss": 0.1934,
      "step": 760
    },
    {
      "epoch": 0.26386962552011095,
      "grad_norm": 3.1365790367126465,
      "learning_rate": 1e-06,
      "loss": 0.2086,
      "step": 761
    },
    {
      "epoch": 0.2642163661581137,
      "grad_norm": 4.416586875915527,
      "learning_rate": 1e-06,
      "loss": 0.1834,
      "step": 762
    },
    {
      "epoch": 0.2645631067961165,
      "grad_norm": 3.5971829891204834,
      "learning_rate": 1e-06,
      "loss": 0.1734,
      "step": 763
    },
    {
      "epoch": 0.26490984743411927,
      "grad_norm": 2.017256259918213,
      "learning_rate": 1e-06,
      "loss": 0.2148,
      "step": 764
    },
    {
      "epoch": 0.26525658807212205,
      "grad_norm": 1.7896546125411987,
      "learning_rate": 1e-06,
      "loss": 0.1934,
      "step": 765
    },
    {
      "epoch": 0.2656033287101248,
      "grad_norm": 2.142129898071289,
      "learning_rate": 1e-06,
      "loss": 0.1691,
      "step": 766
    },
    {
      "epoch": 0.2659500693481276,
      "grad_norm": 2.0466034412384033,
      "learning_rate": 1e-06,
      "loss": 0.1889,
      "step": 767
    },
    {
      "epoch": 0.26629680998613037,
      "grad_norm": 5.9090189933776855,
      "learning_rate": 1e-06,
      "loss": 0.181,
      "step": 768
    },
    {
      "epoch": 0.26664355062413314,
      "grad_norm": 3.580826997756958,
      "learning_rate": 1e-06,
      "loss": 0.1827,
      "step": 769
    },
    {
      "epoch": 0.2669902912621359,
      "grad_norm": 4.003768444061279,
      "learning_rate": 1e-06,
      "loss": 0.1917,
      "step": 770
    },
    {
      "epoch": 0.2673370319001387,
      "grad_norm": 1.884700059890747,
      "learning_rate": 1e-06,
      "loss": 0.182,
      "step": 771
    },
    {
      "epoch": 0.26768377253814146,
      "grad_norm": 2.163311243057251,
      "learning_rate": 1e-06,
      "loss": 0.2078,
      "step": 772
    },
    {
      "epoch": 0.26803051317614424,
      "grad_norm": 4.079468250274658,
      "learning_rate": 1e-06,
      "loss": 0.1485,
      "step": 773
    },
    {
      "epoch": 0.268377253814147,
      "grad_norm": 3.486384391784668,
      "learning_rate": 1e-06,
      "loss": 0.1655,
      "step": 774
    },
    {
      "epoch": 0.2687239944521498,
      "grad_norm": 3.0869040489196777,
      "learning_rate": 1e-06,
      "loss": 0.1543,
      "step": 775
    },
    {
      "epoch": 0.26907073509015256,
      "grad_norm": 1.9200345277786255,
      "learning_rate": 1e-06,
      "loss": 0.1797,
      "step": 776
    },
    {
      "epoch": 0.26941747572815533,
      "grad_norm": 2.677326202392578,
      "learning_rate": 1e-06,
      "loss": 0.1624,
      "step": 777
    },
    {
      "epoch": 0.2697642163661581,
      "grad_norm": 4.965575695037842,
      "learning_rate": 1e-06,
      "loss": 0.1843,
      "step": 778
    },
    {
      "epoch": 0.2701109570041609,
      "grad_norm": 3.100609064102173,
      "learning_rate": 1e-06,
      "loss": 0.2043,
      "step": 779
    },
    {
      "epoch": 0.27045769764216365,
      "grad_norm": 3.7525222301483154,
      "learning_rate": 1e-06,
      "loss": 0.1853,
      "step": 780
    },
    {
      "epoch": 0.2708044382801664,
      "grad_norm": 3.449674606323242,
      "learning_rate": 1e-06,
      "loss": 0.1619,
      "step": 781
    },
    {
      "epoch": 0.2711511789181692,
      "grad_norm": 2.731919050216675,
      "learning_rate": 1e-06,
      "loss": 0.2145,
      "step": 782
    },
    {
      "epoch": 0.271497919556172,
      "grad_norm": 5.897451877593994,
      "learning_rate": 1e-06,
      "loss": 0.196,
      "step": 783
    },
    {
      "epoch": 0.27184466019417475,
      "grad_norm": 3.070488452911377,
      "learning_rate": 1e-06,
      "loss": 0.1985,
      "step": 784
    },
    {
      "epoch": 0.2721914008321775,
      "grad_norm": 2.275068521499634,
      "learning_rate": 1e-06,
      "loss": 0.1865,
      "step": 785
    },
    {
      "epoch": 0.2725381414701803,
      "grad_norm": 2.4451684951782227,
      "learning_rate": 1e-06,
      "loss": 0.1983,
      "step": 786
    },
    {
      "epoch": 0.27288488210818307,
      "grad_norm": 2.198415517807007,
      "learning_rate": 1e-06,
      "loss": 0.1975,
      "step": 787
    },
    {
      "epoch": 0.27323162274618584,
      "grad_norm": 4.776117324829102,
      "learning_rate": 1e-06,
      "loss": 0.1747,
      "step": 788
    },
    {
      "epoch": 0.2735783633841886,
      "grad_norm": 4.163407325744629,
      "learning_rate": 1e-06,
      "loss": 0.1743,
      "step": 789
    },
    {
      "epoch": 0.2739251040221914,
      "grad_norm": 1.9131619930267334,
      "learning_rate": 1e-06,
      "loss": 0.1942,
      "step": 790
    },
    {
      "epoch": 0.27427184466019416,
      "grad_norm": 2.953847885131836,
      "learning_rate": 1e-06,
      "loss": 0.1653,
      "step": 791
    },
    {
      "epoch": 0.27461858529819694,
      "grad_norm": 1.98551344871521,
      "learning_rate": 1e-06,
      "loss": 0.1992,
      "step": 792
    },
    {
      "epoch": 0.2749653259361997,
      "grad_norm": 5.063632965087891,
      "learning_rate": 1e-06,
      "loss": 0.1385,
      "step": 793
    },
    {
      "epoch": 0.2753120665742025,
      "grad_norm": 3.6445019245147705,
      "learning_rate": 1e-06,
      "loss": 0.1653,
      "step": 794
    },
    {
      "epoch": 0.27565880721220526,
      "grad_norm": 1.9771358966827393,
      "learning_rate": 1e-06,
      "loss": 0.2067,
      "step": 795
    },
    {
      "epoch": 0.27600554785020803,
      "grad_norm": 1.9333562850952148,
      "learning_rate": 1e-06,
      "loss": 0.1988,
      "step": 796
    },
    {
      "epoch": 0.2763522884882108,
      "grad_norm": 1.9832173585891724,
      "learning_rate": 1e-06,
      "loss": 0.1646,
      "step": 797
    },
    {
      "epoch": 0.2766990291262136,
      "grad_norm": 2.332988977432251,
      "learning_rate": 1e-06,
      "loss": 0.164,
      "step": 798
    },
    {
      "epoch": 0.27704576976421635,
      "grad_norm": 2.837818145751953,
      "learning_rate": 1e-06,
      "loss": 0.1404,
      "step": 799
    },
    {
      "epoch": 0.27739251040221913,
      "grad_norm": 3.3492045402526855,
      "learning_rate": 1e-06,
      "loss": 0.1635,
      "step": 800
    },
    {
      "epoch": 0.2777392510402219,
      "grad_norm": 1.856004238128662,
      "learning_rate": 1e-06,
      "loss": 0.152,
      "step": 801
    },
    {
      "epoch": 0.2780859916782247,
      "grad_norm": 5.804831504821777,
      "learning_rate": 1e-06,
      "loss": 0.1833,
      "step": 802
    },
    {
      "epoch": 0.27843273231622745,
      "grad_norm": 1.7844274044036865,
      "learning_rate": 1e-06,
      "loss": 0.1768,
      "step": 803
    },
    {
      "epoch": 0.2787794729542302,
      "grad_norm": 2.5231499671936035,
      "learning_rate": 1e-06,
      "loss": 0.193,
      "step": 804
    },
    {
      "epoch": 0.279126213592233,
      "grad_norm": 3.219759464263916,
      "learning_rate": 1e-06,
      "loss": 0.1958,
      "step": 805
    },
    {
      "epoch": 0.27947295423023577,
      "grad_norm": 5.376789093017578,
      "learning_rate": 1e-06,
      "loss": 0.197,
      "step": 806
    },
    {
      "epoch": 0.27981969486823854,
      "grad_norm": 2.796621084213257,
      "learning_rate": 1e-06,
      "loss": 0.2056,
      "step": 807
    },
    {
      "epoch": 0.2801664355062413,
      "grad_norm": 1.9890904426574707,
      "learning_rate": 1e-06,
      "loss": 0.1505,
      "step": 808
    },
    {
      "epoch": 0.2805131761442441,
      "grad_norm": 3.3186163902282715,
      "learning_rate": 1e-06,
      "loss": 0.1487,
      "step": 809
    },
    {
      "epoch": 0.28085991678224687,
      "grad_norm": 3.1521759033203125,
      "learning_rate": 1e-06,
      "loss": 0.1981,
      "step": 810
    },
    {
      "epoch": 0.28120665742024964,
      "grad_norm": 1.7996686697006226,
      "learning_rate": 1e-06,
      "loss": 0.2118,
      "step": 811
    },
    {
      "epoch": 0.2815533980582524,
      "grad_norm": 6.712550640106201,
      "learning_rate": 1e-06,
      "loss": 0.1817,
      "step": 812
    },
    {
      "epoch": 0.2819001386962552,
      "grad_norm": 4.741528511047363,
      "learning_rate": 1e-06,
      "loss": 0.1332,
      "step": 813
    },
    {
      "epoch": 0.28224687933425796,
      "grad_norm": 1.8851327896118164,
      "learning_rate": 1e-06,
      "loss": 0.2028,
      "step": 814
    },
    {
      "epoch": 0.28259361997226073,
      "grad_norm": 2.5891757011413574,
      "learning_rate": 1e-06,
      "loss": 0.2061,
      "step": 815
    },
    {
      "epoch": 0.2829403606102635,
      "grad_norm": 2.130458116531372,
      "learning_rate": 1e-06,
      "loss": 0.1825,
      "step": 816
    },
    {
      "epoch": 0.2832871012482663,
      "grad_norm": 2.5108346939086914,
      "learning_rate": 1e-06,
      "loss": 0.1774,
      "step": 817
    },
    {
      "epoch": 0.28363384188626906,
      "grad_norm": 1.7813079357147217,
      "learning_rate": 1e-06,
      "loss": 0.1809,
      "step": 818
    },
    {
      "epoch": 0.28398058252427183,
      "grad_norm": 1.8457926511764526,
      "learning_rate": 1e-06,
      "loss": 0.1468,
      "step": 819
    },
    {
      "epoch": 0.2843273231622746,
      "grad_norm": 2.4858312606811523,
      "learning_rate": 1e-06,
      "loss": 0.1721,
      "step": 820
    },
    {
      "epoch": 0.2846740638002774,
      "grad_norm": 2.8823843002319336,
      "learning_rate": 1e-06,
      "loss": 0.1908,
      "step": 821
    },
    {
      "epoch": 0.28502080443828015,
      "grad_norm": 2.092439889907837,
      "learning_rate": 1e-06,
      "loss": 0.2033,
      "step": 822
    },
    {
      "epoch": 0.2853675450762829,
      "grad_norm": 2.2995755672454834,
      "learning_rate": 1e-06,
      "loss": 0.1974,
      "step": 823
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 1.6263433694839478,
      "learning_rate": 1e-06,
      "loss": 0.1792,
      "step": 824
    },
    {
      "epoch": 0.28606102635228847,
      "grad_norm": 3.589700222015381,
      "learning_rate": 1e-06,
      "loss": 0.1548,
      "step": 825
    },
    {
      "epoch": 0.28640776699029125,
      "grad_norm": 2.6296792030334473,
      "learning_rate": 1e-06,
      "loss": 0.177,
      "step": 826
    },
    {
      "epoch": 0.286754507628294,
      "grad_norm": 3.712181568145752,
      "learning_rate": 1e-06,
      "loss": 0.1846,
      "step": 827
    },
    {
      "epoch": 0.2871012482662968,
      "grad_norm": 8.185032844543457,
      "learning_rate": 1e-06,
      "loss": 0.2371,
      "step": 828
    },
    {
      "epoch": 0.28744798890429957,
      "grad_norm": 3.00469708442688,
      "learning_rate": 1e-06,
      "loss": 0.1415,
      "step": 829
    },
    {
      "epoch": 0.28779472954230234,
      "grad_norm": 3.0410897731781006,
      "learning_rate": 1e-06,
      "loss": 0.1644,
      "step": 830
    },
    {
      "epoch": 0.2881414701803051,
      "grad_norm": 2.678393840789795,
      "learning_rate": 1e-06,
      "loss": 0.2029,
      "step": 831
    },
    {
      "epoch": 0.2884882108183079,
      "grad_norm": 2.8907225131988525,
      "learning_rate": 1e-06,
      "loss": 0.1449,
      "step": 832
    },
    {
      "epoch": 0.28883495145631066,
      "grad_norm": 1.4863390922546387,
      "learning_rate": 1e-06,
      "loss": 0.1468,
      "step": 833
    },
    {
      "epoch": 0.28918169209431344,
      "grad_norm": 1.769559621810913,
      "learning_rate": 1e-06,
      "loss": 0.1932,
      "step": 834
    },
    {
      "epoch": 0.2895284327323162,
      "grad_norm": 1.6298843622207642,
      "learning_rate": 1e-06,
      "loss": 0.1409,
      "step": 835
    },
    {
      "epoch": 0.289875173370319,
      "grad_norm": 4.213280200958252,
      "learning_rate": 1e-06,
      "loss": 0.1718,
      "step": 836
    },
    {
      "epoch": 0.29022191400832176,
      "grad_norm": 5.342798233032227,
      "learning_rate": 1e-06,
      "loss": 0.2073,
      "step": 837
    },
    {
      "epoch": 0.29056865464632453,
      "grad_norm": 2.607419967651367,
      "learning_rate": 1e-06,
      "loss": 0.175,
      "step": 838
    },
    {
      "epoch": 0.2909153952843273,
      "grad_norm": 3.2880568504333496,
      "learning_rate": 1e-06,
      "loss": 0.2236,
      "step": 839
    },
    {
      "epoch": 0.2912621359223301,
      "grad_norm": 2.370192527770996,
      "learning_rate": 1e-06,
      "loss": 0.1806,
      "step": 840
    },
    {
      "epoch": 0.29160887656033285,
      "grad_norm": 1.8948553800582886,
      "learning_rate": 1e-06,
      "loss": 0.1832,
      "step": 841
    },
    {
      "epoch": 0.2919556171983356,
      "grad_norm": 2.092898368835449,
      "learning_rate": 1e-06,
      "loss": 0.1868,
      "step": 842
    },
    {
      "epoch": 0.2923023578363384,
      "grad_norm": 1.5037286281585693,
      "learning_rate": 1e-06,
      "loss": 0.1581,
      "step": 843
    },
    {
      "epoch": 0.2926490984743412,
      "grad_norm": 3.512408494949341,
      "learning_rate": 1e-06,
      "loss": 0.1883,
      "step": 844
    },
    {
      "epoch": 0.29299583911234395,
      "grad_norm": 7.901118278503418,
      "learning_rate": 1e-06,
      "loss": 0.1555,
      "step": 845
    },
    {
      "epoch": 0.2933425797503467,
      "grad_norm": 5.720544815063477,
      "learning_rate": 1e-06,
      "loss": 0.1626,
      "step": 846
    },
    {
      "epoch": 0.2936893203883495,
      "grad_norm": 2.572035312652588,
      "learning_rate": 1e-06,
      "loss": 0.2006,
      "step": 847
    },
    {
      "epoch": 0.29403606102635227,
      "grad_norm": 1.5427334308624268,
      "learning_rate": 1e-06,
      "loss": 0.1666,
      "step": 848
    },
    {
      "epoch": 0.29438280166435504,
      "grad_norm": 2.345066547393799,
      "learning_rate": 1e-06,
      "loss": 0.1867,
      "step": 849
    },
    {
      "epoch": 0.2947295423023578,
      "grad_norm": 2.536377429962158,
      "learning_rate": 1e-06,
      "loss": 0.1918,
      "step": 850
    },
    {
      "epoch": 0.2950762829403606,
      "grad_norm": 1.5710004568099976,
      "learning_rate": 1e-06,
      "loss": 0.1747,
      "step": 851
    },
    {
      "epoch": 0.29542302357836336,
      "grad_norm": 1.5936557054519653,
      "learning_rate": 1e-06,
      "loss": 0.1569,
      "step": 852
    },
    {
      "epoch": 0.29576976421636614,
      "grad_norm": 4.063443660736084,
      "learning_rate": 1e-06,
      "loss": 0.1922,
      "step": 853
    },
    {
      "epoch": 0.2961165048543689,
      "grad_norm": 1.6471939086914062,
      "learning_rate": 1e-06,
      "loss": 0.1688,
      "step": 854
    },
    {
      "epoch": 0.2964632454923717,
      "grad_norm": 3.7626683712005615,
      "learning_rate": 1e-06,
      "loss": 0.1441,
      "step": 855
    },
    {
      "epoch": 0.29680998613037446,
      "grad_norm": 3.8700716495513916,
      "learning_rate": 1e-06,
      "loss": 0.1727,
      "step": 856
    },
    {
      "epoch": 0.29715672676837723,
      "grad_norm": 3.989415407180786,
      "learning_rate": 1e-06,
      "loss": 0.1821,
      "step": 857
    },
    {
      "epoch": 0.29750346740638,
      "grad_norm": 5.3771538734436035,
      "learning_rate": 1e-06,
      "loss": 0.2133,
      "step": 858
    },
    {
      "epoch": 0.2978502080443828,
      "grad_norm": 1.7356910705566406,
      "learning_rate": 1e-06,
      "loss": 0.1702,
      "step": 859
    },
    {
      "epoch": 0.29819694868238555,
      "grad_norm": 1.795180320739746,
      "learning_rate": 1e-06,
      "loss": 0.1836,
      "step": 860
    },
    {
      "epoch": 0.29854368932038833,
      "grad_norm": 4.196115970611572,
      "learning_rate": 1e-06,
      "loss": 0.173,
      "step": 861
    },
    {
      "epoch": 0.2988904299583911,
      "grad_norm": 2.782108783721924,
      "learning_rate": 1e-06,
      "loss": 0.1905,
      "step": 862
    },
    {
      "epoch": 0.2992371705963939,
      "grad_norm": 3.027663230895996,
      "learning_rate": 1e-06,
      "loss": 0.1974,
      "step": 863
    },
    {
      "epoch": 0.29958391123439665,
      "grad_norm": 1.9452866315841675,
      "learning_rate": 1e-06,
      "loss": 0.1655,
      "step": 864
    },
    {
      "epoch": 0.2999306518723994,
      "grad_norm": 2.2103707790374756,
      "learning_rate": 1e-06,
      "loss": 0.2024,
      "step": 865
    },
    {
      "epoch": 0.3002773925104022,
      "grad_norm": 2.538545846939087,
      "learning_rate": 1e-06,
      "loss": 0.1733,
      "step": 866
    },
    {
      "epoch": 0.30062413314840497,
      "grad_norm": 2.4375507831573486,
      "learning_rate": 1e-06,
      "loss": 0.1841,
      "step": 867
    },
    {
      "epoch": 0.30097087378640774,
      "grad_norm": 2.466561794281006,
      "learning_rate": 1e-06,
      "loss": 0.1835,
      "step": 868
    },
    {
      "epoch": 0.3013176144244105,
      "grad_norm": 1.5926258563995361,
      "learning_rate": 1e-06,
      "loss": 0.1591,
      "step": 869
    },
    {
      "epoch": 0.3016643550624133,
      "grad_norm": 1.525660514831543,
      "learning_rate": 1e-06,
      "loss": 0.1355,
      "step": 870
    },
    {
      "epoch": 0.30201109570041607,
      "grad_norm": 1.8583039045333862,
      "learning_rate": 1e-06,
      "loss": 0.164,
      "step": 871
    },
    {
      "epoch": 0.30235783633841884,
      "grad_norm": 7.7814531326293945,
      "learning_rate": 1e-06,
      "loss": 0.2342,
      "step": 872
    },
    {
      "epoch": 0.3027045769764216,
      "grad_norm": 1.736899495124817,
      "learning_rate": 1e-06,
      "loss": 0.1928,
      "step": 873
    },
    {
      "epoch": 0.3030513176144244,
      "grad_norm": 2.617612361907959,
      "learning_rate": 1e-06,
      "loss": 0.2155,
      "step": 874
    },
    {
      "epoch": 0.30339805825242716,
      "grad_norm": 4.537561416625977,
      "learning_rate": 1e-06,
      "loss": 0.1747,
      "step": 875
    },
    {
      "epoch": 0.30374479889042993,
      "grad_norm": 3.117304563522339,
      "learning_rate": 1e-06,
      "loss": 0.1786,
      "step": 876
    },
    {
      "epoch": 0.3040915395284327,
      "grad_norm": 1.9936450719833374,
      "learning_rate": 1e-06,
      "loss": 0.1963,
      "step": 877
    },
    {
      "epoch": 0.3044382801664355,
      "grad_norm": 2.502594470977783,
      "learning_rate": 1e-06,
      "loss": 0.1719,
      "step": 878
    },
    {
      "epoch": 0.30478502080443826,
      "grad_norm": 2.9981226921081543,
      "learning_rate": 1e-06,
      "loss": 0.1975,
      "step": 879
    },
    {
      "epoch": 0.30513176144244103,
      "grad_norm": 1.8054769039154053,
      "learning_rate": 1e-06,
      "loss": 0.1663,
      "step": 880
    },
    {
      "epoch": 0.3054785020804438,
      "grad_norm": 1.8026496171951294,
      "learning_rate": 1e-06,
      "loss": 0.183,
      "step": 881
    },
    {
      "epoch": 0.3058252427184466,
      "grad_norm": 2.0367887020111084,
      "learning_rate": 1e-06,
      "loss": 0.1754,
      "step": 882
    },
    {
      "epoch": 0.30617198335644935,
      "grad_norm": 1.685020089149475,
      "learning_rate": 1e-06,
      "loss": 0.1749,
      "step": 883
    },
    {
      "epoch": 0.3065187239944521,
      "grad_norm": 2.1757829189300537,
      "learning_rate": 1e-06,
      "loss": 0.2091,
      "step": 884
    },
    {
      "epoch": 0.3068654646324549,
      "grad_norm": 5.054877281188965,
      "learning_rate": 1e-06,
      "loss": 0.1787,
      "step": 885
    },
    {
      "epoch": 0.30721220527045767,
      "grad_norm": 1.8964219093322754,
      "learning_rate": 1e-06,
      "loss": 0.1853,
      "step": 886
    },
    {
      "epoch": 0.30755894590846045,
      "grad_norm": 6.154984951019287,
      "learning_rate": 1e-06,
      "loss": 0.1642,
      "step": 887
    },
    {
      "epoch": 0.3079056865464632,
      "grad_norm": 3.810920238494873,
      "learning_rate": 1e-06,
      "loss": 0.1482,
      "step": 888
    },
    {
      "epoch": 0.308252427184466,
      "grad_norm": 1.7902008295059204,
      "learning_rate": 1e-06,
      "loss": 0.2041,
      "step": 889
    },
    {
      "epoch": 0.30859916782246877,
      "grad_norm": 2.2116265296936035,
      "learning_rate": 1e-06,
      "loss": 0.1945,
      "step": 890
    },
    {
      "epoch": 0.30894590846047154,
      "grad_norm": 4.08473014831543,
      "learning_rate": 1e-06,
      "loss": 0.1814,
      "step": 891
    },
    {
      "epoch": 0.3092926490984743,
      "grad_norm": 4.12503719329834,
      "learning_rate": 1e-06,
      "loss": 0.1697,
      "step": 892
    },
    {
      "epoch": 0.3096393897364771,
      "grad_norm": 2.198702096939087,
      "learning_rate": 1e-06,
      "loss": 0.1646,
      "step": 893
    },
    {
      "epoch": 0.30998613037447986,
      "grad_norm": 2.104871988296509,
      "learning_rate": 1e-06,
      "loss": 0.1837,
      "step": 894
    },
    {
      "epoch": 0.31033287101248264,
      "grad_norm": 1.7803840637207031,
      "learning_rate": 1e-06,
      "loss": 0.1687,
      "step": 895
    },
    {
      "epoch": 0.3106796116504854,
      "grad_norm": 2.5232043266296387,
      "learning_rate": 1e-06,
      "loss": 0.1852,
      "step": 896
    },
    {
      "epoch": 0.3110263522884882,
      "grad_norm": 1.5908442735671997,
      "learning_rate": 1e-06,
      "loss": 0.161,
      "step": 897
    },
    {
      "epoch": 0.31137309292649096,
      "grad_norm": 3.704270839691162,
      "learning_rate": 1e-06,
      "loss": 0.1547,
      "step": 898
    },
    {
      "epoch": 0.31171983356449373,
      "grad_norm": 1.6646925210952759,
      "learning_rate": 1e-06,
      "loss": 0.1621,
      "step": 899
    },
    {
      "epoch": 0.3120665742024965,
      "grad_norm": 1.6272737979888916,
      "learning_rate": 1e-06,
      "loss": 0.1738,
      "step": 900
    },
    {
      "epoch": 0.3124133148404993,
      "grad_norm": 2.2414157390594482,
      "learning_rate": 1e-06,
      "loss": 0.1857,
      "step": 901
    },
    {
      "epoch": 0.3127600554785021,
      "grad_norm": 4.613629341125488,
      "learning_rate": 1e-06,
      "loss": 0.2038,
      "step": 902
    },
    {
      "epoch": 0.3131067961165049,
      "grad_norm": 2.6360721588134766,
      "learning_rate": 1e-06,
      "loss": 0.1689,
      "step": 903
    },
    {
      "epoch": 0.31345353675450766,
      "grad_norm": 2.4351446628570557,
      "learning_rate": 1e-06,
      "loss": 0.1699,
      "step": 904
    },
    {
      "epoch": 0.31380027739251043,
      "grad_norm": 1.645674228668213,
      "learning_rate": 1e-06,
      "loss": 0.1747,
      "step": 905
    },
    {
      "epoch": 0.3141470180305132,
      "grad_norm": 2.1112935543060303,
      "learning_rate": 1e-06,
      "loss": 0.1725,
      "step": 906
    },
    {
      "epoch": 0.314493758668516,
      "grad_norm": 1.8450239896774292,
      "learning_rate": 1e-06,
      "loss": 0.1783,
      "step": 907
    },
    {
      "epoch": 0.31484049930651875,
      "grad_norm": 3.959017276763916,
      "learning_rate": 1e-06,
      "loss": 0.165,
      "step": 908
    },
    {
      "epoch": 0.3151872399445215,
      "grad_norm": 2.1411192417144775,
      "learning_rate": 1e-06,
      "loss": 0.1566,
      "step": 909
    },
    {
      "epoch": 0.3155339805825243,
      "grad_norm": 2.4546244144439697,
      "learning_rate": 1e-06,
      "loss": 0.193,
      "step": 910
    },
    {
      "epoch": 0.31588072122052707,
      "grad_norm": 4.3430280685424805,
      "learning_rate": 1e-06,
      "loss": 0.1721,
      "step": 911
    },
    {
      "epoch": 0.31622746185852985,
      "grad_norm": 3.988929271697998,
      "learning_rate": 1e-06,
      "loss": 0.159,
      "step": 912
    },
    {
      "epoch": 0.3165742024965326,
      "grad_norm": 3.2247719764709473,
      "learning_rate": 1e-06,
      "loss": 0.1624,
      "step": 913
    },
    {
      "epoch": 0.3169209431345354,
      "grad_norm": 1.647119402885437,
      "learning_rate": 1e-06,
      "loss": 0.1659,
      "step": 914
    },
    {
      "epoch": 0.31726768377253817,
      "grad_norm": 3.0910604000091553,
      "learning_rate": 1e-06,
      "loss": 0.18,
      "step": 915
    },
    {
      "epoch": 0.31761442441054094,
      "grad_norm": 1.7037394046783447,
      "learning_rate": 1e-06,
      "loss": 0.2035,
      "step": 916
    },
    {
      "epoch": 0.3179611650485437,
      "grad_norm": 3.9424848556518555,
      "learning_rate": 1e-06,
      "loss": 0.1755,
      "step": 917
    },
    {
      "epoch": 0.3183079056865465,
      "grad_norm": 1.8996752500534058,
      "learning_rate": 1e-06,
      "loss": 0.1674,
      "step": 918
    },
    {
      "epoch": 0.31865464632454926,
      "grad_norm": 1.9568631649017334,
      "learning_rate": 1e-06,
      "loss": 0.1741,
      "step": 919
    },
    {
      "epoch": 0.31900138696255204,
      "grad_norm": 2.990325927734375,
      "learning_rate": 1e-06,
      "loss": 0.1907,
      "step": 920
    },
    {
      "epoch": 0.3193481276005548,
      "grad_norm": 1.8824876546859741,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 921
    },
    {
      "epoch": 0.3196948682385576,
      "grad_norm": 1.7028950452804565,
      "learning_rate": 1e-06,
      "loss": 0.1362,
      "step": 922
    },
    {
      "epoch": 0.32004160887656036,
      "grad_norm": 2.010843515396118,
      "learning_rate": 1e-06,
      "loss": 0.1758,
      "step": 923
    },
    {
      "epoch": 0.32038834951456313,
      "grad_norm": 1.7268744707107544,
      "learning_rate": 1e-06,
      "loss": 0.1853,
      "step": 924
    },
    {
      "epoch": 0.3207350901525659,
      "grad_norm": 2.266728162765503,
      "learning_rate": 1e-06,
      "loss": 0.1644,
      "step": 925
    },
    {
      "epoch": 0.3210818307905687,
      "grad_norm": 5.072295188903809,
      "learning_rate": 1e-06,
      "loss": 0.1793,
      "step": 926
    },
    {
      "epoch": 0.32142857142857145,
      "grad_norm": 2.210855484008789,
      "learning_rate": 1e-06,
      "loss": 0.1425,
      "step": 927
    },
    {
      "epoch": 0.3217753120665742,
      "grad_norm": 2.0803961753845215,
      "learning_rate": 1e-06,
      "loss": 0.1651,
      "step": 928
    },
    {
      "epoch": 0.322122052704577,
      "grad_norm": 2.160520076751709,
      "learning_rate": 1e-06,
      "loss": 0.1596,
      "step": 929
    },
    {
      "epoch": 0.3224687933425798,
      "grad_norm": 1.9977649450302124,
      "learning_rate": 1e-06,
      "loss": 0.1876,
      "step": 930
    },
    {
      "epoch": 0.32281553398058255,
      "grad_norm": 1.5417920351028442,
      "learning_rate": 1e-06,
      "loss": 0.1503,
      "step": 931
    },
    {
      "epoch": 0.3231622746185853,
      "grad_norm": 2.099778175354004,
      "learning_rate": 1e-06,
      "loss": 0.1836,
      "step": 932
    },
    {
      "epoch": 0.3235090152565881,
      "grad_norm": 3.9819977283477783,
      "learning_rate": 1e-06,
      "loss": 0.1587,
      "step": 933
    },
    {
      "epoch": 0.32385575589459087,
      "grad_norm": 1.6170867681503296,
      "learning_rate": 1e-06,
      "loss": 0.147,
      "step": 934
    },
    {
      "epoch": 0.32420249653259364,
      "grad_norm": 3.3660576343536377,
      "learning_rate": 1e-06,
      "loss": 0.1445,
      "step": 935
    },
    {
      "epoch": 0.3245492371705964,
      "grad_norm": 2.2826545238494873,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 936
    },
    {
      "epoch": 0.3248959778085992,
      "grad_norm": 5.107556343078613,
      "learning_rate": 1e-06,
      "loss": 0.1964,
      "step": 937
    },
    {
      "epoch": 0.32524271844660196,
      "grad_norm": 3.9381065368652344,
      "learning_rate": 1e-06,
      "loss": 0.2003,
      "step": 938
    },
    {
      "epoch": 0.32558945908460474,
      "grad_norm": 2.5233023166656494,
      "learning_rate": 1e-06,
      "loss": 0.1747,
      "step": 939
    },
    {
      "epoch": 0.3259361997226075,
      "grad_norm": 1.8627601861953735,
      "learning_rate": 1e-06,
      "loss": 0.177,
      "step": 940
    },
    {
      "epoch": 0.3262829403606103,
      "grad_norm": 1.7760263681411743,
      "learning_rate": 1e-06,
      "loss": 0.1548,
      "step": 941
    },
    {
      "epoch": 0.32662968099861306,
      "grad_norm": 2.0792856216430664,
      "learning_rate": 1e-06,
      "loss": 0.1762,
      "step": 942
    },
    {
      "epoch": 0.32697642163661583,
      "grad_norm": 4.968944549560547,
      "learning_rate": 1e-06,
      "loss": 0.163,
      "step": 943
    },
    {
      "epoch": 0.3273231622746186,
      "grad_norm": 2.9295568466186523,
      "learning_rate": 1e-06,
      "loss": 0.1928,
      "step": 944
    },
    {
      "epoch": 0.3276699029126214,
      "grad_norm": 3.2143867015838623,
      "learning_rate": 1e-06,
      "loss": 0.1401,
      "step": 945
    },
    {
      "epoch": 0.32801664355062415,
      "grad_norm": 4.367483139038086,
      "learning_rate": 1e-06,
      "loss": 0.2002,
      "step": 946
    },
    {
      "epoch": 0.32836338418862693,
      "grad_norm": 2.355031967163086,
      "learning_rate": 1e-06,
      "loss": 0.1719,
      "step": 947
    },
    {
      "epoch": 0.3287101248266297,
      "grad_norm": 2.9974913597106934,
      "learning_rate": 1e-06,
      "loss": 0.1559,
      "step": 948
    },
    {
      "epoch": 0.3290568654646325,
      "grad_norm": 2.6766953468322754,
      "learning_rate": 1e-06,
      "loss": 0.1834,
      "step": 949
    },
    {
      "epoch": 0.32940360610263525,
      "grad_norm": 3.276123523712158,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 950
    },
    {
      "epoch": 0.329750346740638,
      "grad_norm": 1.6151905059814453,
      "learning_rate": 1e-06,
      "loss": 0.1809,
      "step": 951
    },
    {
      "epoch": 0.3300970873786408,
      "grad_norm": 1.8168013095855713,
      "learning_rate": 1e-06,
      "loss": 0.1436,
      "step": 952
    },
    {
      "epoch": 0.33044382801664357,
      "grad_norm": 4.220736503601074,
      "learning_rate": 1e-06,
      "loss": 0.1696,
      "step": 953
    },
    {
      "epoch": 0.33079056865464634,
      "grad_norm": 2.646458625793457,
      "learning_rate": 1e-06,
      "loss": 0.14,
      "step": 954
    },
    {
      "epoch": 0.3311373092926491,
      "grad_norm": 5.339474201202393,
      "learning_rate": 1e-06,
      "loss": 0.2126,
      "step": 955
    },
    {
      "epoch": 0.3314840499306519,
      "grad_norm": 3.148782730102539,
      "learning_rate": 1e-06,
      "loss": 0.1692,
      "step": 956
    },
    {
      "epoch": 0.33183079056865467,
      "grad_norm": 3.1843745708465576,
      "learning_rate": 1e-06,
      "loss": 0.1804,
      "step": 957
    },
    {
      "epoch": 0.33217753120665744,
      "grad_norm": 2.0136914253234863,
      "learning_rate": 1e-06,
      "loss": 0.1817,
      "step": 958
    },
    {
      "epoch": 0.3325242718446602,
      "grad_norm": 2.250756025314331,
      "learning_rate": 1e-06,
      "loss": 0.1885,
      "step": 959
    },
    {
      "epoch": 0.332871012482663,
      "grad_norm": 3.832793951034546,
      "learning_rate": 1e-06,
      "loss": 0.1617,
      "step": 960
    },
    {
      "epoch": 0.33321775312066576,
      "grad_norm": 3.097174644470215,
      "learning_rate": 1e-06,
      "loss": 0.163,
      "step": 961
    },
    {
      "epoch": 0.33356449375866853,
      "grad_norm": 3.634685516357422,
      "learning_rate": 1e-06,
      "loss": 0.1465,
      "step": 962
    },
    {
      "epoch": 0.3339112343966713,
      "grad_norm": 1.968932032585144,
      "learning_rate": 1e-06,
      "loss": 0.1819,
      "step": 963
    },
    {
      "epoch": 0.3342579750346741,
      "grad_norm": 3.506141424179077,
      "learning_rate": 1e-06,
      "loss": 0.1623,
      "step": 964
    },
    {
      "epoch": 0.33460471567267686,
      "grad_norm": 2.201019048690796,
      "learning_rate": 1e-06,
      "loss": 0.1987,
      "step": 965
    },
    {
      "epoch": 0.33495145631067963,
      "grad_norm": 1.871758222579956,
      "learning_rate": 1e-06,
      "loss": 0.1651,
      "step": 966
    },
    {
      "epoch": 0.3352981969486824,
      "grad_norm": 2.852182626724243,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 967
    },
    {
      "epoch": 0.3356449375866852,
      "grad_norm": 2.830751657485962,
      "learning_rate": 1e-06,
      "loss": 0.1934,
      "step": 968
    },
    {
      "epoch": 0.33599167822468795,
      "grad_norm": 2.9230191707611084,
      "learning_rate": 1e-06,
      "loss": 0.1689,
      "step": 969
    },
    {
      "epoch": 0.3363384188626907,
      "grad_norm": 3.4389359951019287,
      "learning_rate": 1e-06,
      "loss": 0.1606,
      "step": 970
    },
    {
      "epoch": 0.3366851595006935,
      "grad_norm": 3.4921634197235107,
      "learning_rate": 1e-06,
      "loss": 0.1891,
      "step": 971
    },
    {
      "epoch": 0.33703190013869627,
      "grad_norm": 2.9009199142456055,
      "learning_rate": 1e-06,
      "loss": 0.1508,
      "step": 972
    },
    {
      "epoch": 0.33737864077669905,
      "grad_norm": 1.8419466018676758,
      "learning_rate": 1e-06,
      "loss": 0.173,
      "step": 973
    },
    {
      "epoch": 0.3377253814147018,
      "grad_norm": 4.729164123535156,
      "learning_rate": 1e-06,
      "loss": 0.1863,
      "step": 974
    },
    {
      "epoch": 0.3380721220527046,
      "grad_norm": 4.153334617614746,
      "learning_rate": 1e-06,
      "loss": 0.189,
      "step": 975
    },
    {
      "epoch": 0.33841886269070737,
      "grad_norm": 2.406278371810913,
      "learning_rate": 1e-06,
      "loss": 0.1726,
      "step": 976
    },
    {
      "epoch": 0.33876560332871014,
      "grad_norm": 1.854493498802185,
      "learning_rate": 1e-06,
      "loss": 0.1859,
      "step": 977
    },
    {
      "epoch": 0.3391123439667129,
      "grad_norm": 1.7517307996749878,
      "learning_rate": 1e-06,
      "loss": 0.1849,
      "step": 978
    },
    {
      "epoch": 0.3394590846047157,
      "grad_norm": 3.5556588172912598,
      "learning_rate": 1e-06,
      "loss": 0.1691,
      "step": 979
    },
    {
      "epoch": 0.33980582524271846,
      "grad_norm": 3.863245725631714,
      "learning_rate": 1e-06,
      "loss": 0.1758,
      "step": 980
    },
    {
      "epoch": 0.34015256588072124,
      "grad_norm": 4.660958766937256,
      "learning_rate": 1e-06,
      "loss": 0.1492,
      "step": 981
    },
    {
      "epoch": 0.340499306518724,
      "grad_norm": 2.8175435066223145,
      "learning_rate": 1e-06,
      "loss": 0.1559,
      "step": 982
    },
    {
      "epoch": 0.3408460471567268,
      "grad_norm": 3.012770891189575,
      "learning_rate": 1e-06,
      "loss": 0.1876,
      "step": 983
    },
    {
      "epoch": 0.34119278779472956,
      "grad_norm": 3.963909387588501,
      "learning_rate": 1e-06,
      "loss": 0.1512,
      "step": 984
    },
    {
      "epoch": 0.34153952843273233,
      "grad_norm": 1.6655025482177734,
      "learning_rate": 1e-06,
      "loss": 0.1429,
      "step": 985
    },
    {
      "epoch": 0.3418862690707351,
      "grad_norm": 2.783829689025879,
      "learning_rate": 1e-06,
      "loss": 0.2169,
      "step": 986
    },
    {
      "epoch": 0.3422330097087379,
      "grad_norm": 1.5330981016159058,
      "learning_rate": 1e-06,
      "loss": 0.1479,
      "step": 987
    },
    {
      "epoch": 0.34257975034674065,
      "grad_norm": 1.6944447755813599,
      "learning_rate": 1e-06,
      "loss": 0.1762,
      "step": 988
    },
    {
      "epoch": 0.3429264909847434,
      "grad_norm": 2.4283335208892822,
      "learning_rate": 1e-06,
      "loss": 0.1577,
      "step": 989
    },
    {
      "epoch": 0.3432732316227462,
      "grad_norm": 2.7645480632781982,
      "learning_rate": 1e-06,
      "loss": 0.1697,
      "step": 990
    },
    {
      "epoch": 0.343619972260749,
      "grad_norm": 4.196599960327148,
      "learning_rate": 1e-06,
      "loss": 0.1777,
      "step": 991
    },
    {
      "epoch": 0.34396671289875175,
      "grad_norm": 1.866174340248108,
      "learning_rate": 1e-06,
      "loss": 0.1461,
      "step": 992
    },
    {
      "epoch": 0.3443134535367545,
      "grad_norm": 2.0059845447540283,
      "learning_rate": 1e-06,
      "loss": 0.182,
      "step": 993
    },
    {
      "epoch": 0.3446601941747573,
      "grad_norm": 1.9645408391952515,
      "learning_rate": 1e-06,
      "loss": 0.1908,
      "step": 994
    },
    {
      "epoch": 0.34500693481276007,
      "grad_norm": 2.1238350868225098,
      "learning_rate": 1e-06,
      "loss": 0.2052,
      "step": 995
    },
    {
      "epoch": 0.34535367545076284,
      "grad_norm": 1.8180639743804932,
      "learning_rate": 1e-06,
      "loss": 0.1684,
      "step": 996
    },
    {
      "epoch": 0.3457004160887656,
      "grad_norm": 2.0264620780944824,
      "learning_rate": 1e-06,
      "loss": 0.1827,
      "step": 997
    },
    {
      "epoch": 0.3460471567267684,
      "grad_norm": 2.302675724029541,
      "learning_rate": 1e-06,
      "loss": 0.1999,
      "step": 998
    },
    {
      "epoch": 0.34639389736477116,
      "grad_norm": 2.18147873878479,
      "learning_rate": 1e-06,
      "loss": 0.1694,
      "step": 999
    },
    {
      "epoch": 0.34674063800277394,
      "grad_norm": 7.13607931137085,
      "learning_rate": 1e-06,
      "loss": 0.1692,
      "step": 1000
    },
    {
      "epoch": 0.3470873786407767,
      "grad_norm": 3.1698849201202393,
      "learning_rate": 1e-06,
      "loss": 0.1861,
      "step": 1001
    },
    {
      "epoch": 0.3474341192787795,
      "grad_norm": 2.142284393310547,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 1002
    },
    {
      "epoch": 0.34778085991678226,
      "grad_norm": 2.648524284362793,
      "learning_rate": 1e-06,
      "loss": 0.1809,
      "step": 1003
    },
    {
      "epoch": 0.34812760055478503,
      "grad_norm": 1.6139206886291504,
      "learning_rate": 1e-06,
      "loss": 0.1659,
      "step": 1004
    },
    {
      "epoch": 0.3484743411927878,
      "grad_norm": 1.9868088960647583,
      "learning_rate": 1e-06,
      "loss": 0.142,
      "step": 1005
    },
    {
      "epoch": 0.3488210818307906,
      "grad_norm": 1.7184799909591675,
      "learning_rate": 1e-06,
      "loss": 0.1547,
      "step": 1006
    },
    {
      "epoch": 0.34916782246879335,
      "grad_norm": 1.670831561088562,
      "learning_rate": 1e-06,
      "loss": 0.1624,
      "step": 1007
    },
    {
      "epoch": 0.34951456310679613,
      "grad_norm": 1.932572364807129,
      "learning_rate": 1e-06,
      "loss": 0.1428,
      "step": 1008
    },
    {
      "epoch": 0.3498613037447989,
      "grad_norm": 3.854353189468384,
      "learning_rate": 1e-06,
      "loss": 0.1436,
      "step": 1009
    },
    {
      "epoch": 0.3502080443828017,
      "grad_norm": 1.7538644075393677,
      "learning_rate": 1e-06,
      "loss": 0.1491,
      "step": 1010
    },
    {
      "epoch": 0.35055478502080445,
      "grad_norm": 1.8496301174163818,
      "learning_rate": 1e-06,
      "loss": 0.1756,
      "step": 1011
    },
    {
      "epoch": 0.3509015256588072,
      "grad_norm": 4.00302267074585,
      "learning_rate": 1e-06,
      "loss": 0.1809,
      "step": 1012
    },
    {
      "epoch": 0.35124826629681,
      "grad_norm": 1.931908369064331,
      "learning_rate": 1e-06,
      "loss": 0.1675,
      "step": 1013
    },
    {
      "epoch": 0.35159500693481277,
      "grad_norm": 1.933568000793457,
      "learning_rate": 1e-06,
      "loss": 0.1556,
      "step": 1014
    },
    {
      "epoch": 0.35194174757281554,
      "grad_norm": 1.5976754426956177,
      "learning_rate": 1e-06,
      "loss": 0.1566,
      "step": 1015
    },
    {
      "epoch": 0.3522884882108183,
      "grad_norm": 1.9067904949188232,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 1016
    },
    {
      "epoch": 0.3526352288488211,
      "grad_norm": 2.5264508724212646,
      "learning_rate": 1e-06,
      "loss": 0.1512,
      "step": 1017
    },
    {
      "epoch": 0.35298196948682387,
      "grad_norm": 3.613884925842285,
      "learning_rate": 1e-06,
      "loss": 0.1775,
      "step": 1018
    },
    {
      "epoch": 0.35332871012482664,
      "grad_norm": 2.958832263946533,
      "learning_rate": 1e-06,
      "loss": 0.1838,
      "step": 1019
    },
    {
      "epoch": 0.3536754507628294,
      "grad_norm": 1.8982412815093994,
      "learning_rate": 1e-06,
      "loss": 0.1416,
      "step": 1020
    },
    {
      "epoch": 0.3540221914008322,
      "grad_norm": 2.6603124141693115,
      "learning_rate": 1e-06,
      "loss": 0.1839,
      "step": 1021
    },
    {
      "epoch": 0.35436893203883496,
      "grad_norm": 3.2685227394104004,
      "learning_rate": 1e-06,
      "loss": 0.1893,
      "step": 1022
    },
    {
      "epoch": 0.35471567267683773,
      "grad_norm": 2.6243014335632324,
      "learning_rate": 1e-06,
      "loss": 0.1805,
      "step": 1023
    },
    {
      "epoch": 0.3550624133148405,
      "grad_norm": 3.69974946975708,
      "learning_rate": 1e-06,
      "loss": 0.199,
      "step": 1024
    },
    {
      "epoch": 0.3554091539528433,
      "grad_norm": 1.906920313835144,
      "learning_rate": 1e-06,
      "loss": 0.1526,
      "step": 1025
    },
    {
      "epoch": 0.35575589459084606,
      "grad_norm": 4.538094520568848,
      "learning_rate": 1e-06,
      "loss": 0.2041,
      "step": 1026
    },
    {
      "epoch": 0.35610263522884883,
      "grad_norm": 1.9397320747375488,
      "learning_rate": 1e-06,
      "loss": 0.1619,
      "step": 1027
    },
    {
      "epoch": 0.3564493758668516,
      "grad_norm": 3.0926973819732666,
      "learning_rate": 1e-06,
      "loss": 0.193,
      "step": 1028
    },
    {
      "epoch": 0.3567961165048544,
      "grad_norm": 6.100024223327637,
      "learning_rate": 1e-06,
      "loss": 0.1759,
      "step": 1029
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 1.7386844158172607,
      "learning_rate": 1e-06,
      "loss": 0.1573,
      "step": 1030
    },
    {
      "epoch": 0.3574895977808599,
      "grad_norm": 2.905871629714966,
      "learning_rate": 1e-06,
      "loss": 0.1522,
      "step": 1031
    },
    {
      "epoch": 0.3578363384188627,
      "grad_norm": 2.167205333709717,
      "learning_rate": 1e-06,
      "loss": 0.1628,
      "step": 1032
    },
    {
      "epoch": 0.35818307905686547,
      "grad_norm": 2.418584108352661,
      "learning_rate": 1e-06,
      "loss": 0.1755,
      "step": 1033
    },
    {
      "epoch": 0.35852981969486825,
      "grad_norm": 2.9872934818267822,
      "learning_rate": 1e-06,
      "loss": 0.174,
      "step": 1034
    },
    {
      "epoch": 0.358876560332871,
      "grad_norm": 3.5263776779174805,
      "learning_rate": 1e-06,
      "loss": 0.1802,
      "step": 1035
    },
    {
      "epoch": 0.3592233009708738,
      "grad_norm": 3.891584634780884,
      "learning_rate": 1e-06,
      "loss": 0.1996,
      "step": 1036
    },
    {
      "epoch": 0.35957004160887657,
      "grad_norm": 2.081064224243164,
      "learning_rate": 1e-06,
      "loss": 0.1578,
      "step": 1037
    },
    {
      "epoch": 0.35991678224687934,
      "grad_norm": 2.100369930267334,
      "learning_rate": 1e-06,
      "loss": 0.1572,
      "step": 1038
    },
    {
      "epoch": 0.3602635228848821,
      "grad_norm": 2.148822069168091,
      "learning_rate": 1e-06,
      "loss": 0.1788,
      "step": 1039
    },
    {
      "epoch": 0.3606102635228849,
      "grad_norm": 2.7405498027801514,
      "learning_rate": 1e-06,
      "loss": 0.1863,
      "step": 1040
    },
    {
      "epoch": 0.36095700416088766,
      "grad_norm": 4.158656597137451,
      "learning_rate": 1e-06,
      "loss": 0.1699,
      "step": 1041
    },
    {
      "epoch": 0.36130374479889044,
      "grad_norm": 2.858583688735962,
      "learning_rate": 1e-06,
      "loss": 0.1855,
      "step": 1042
    },
    {
      "epoch": 0.3616504854368932,
      "grad_norm": 5.225622177124023,
      "learning_rate": 1e-06,
      "loss": 0.1613,
      "step": 1043
    },
    {
      "epoch": 0.361997226074896,
      "grad_norm": 2.721017599105835,
      "learning_rate": 1e-06,
      "loss": 0.1532,
      "step": 1044
    },
    {
      "epoch": 0.36234396671289876,
      "grad_norm": 3.189161539077759,
      "learning_rate": 1e-06,
      "loss": 0.1966,
      "step": 1045
    },
    {
      "epoch": 0.36269070735090153,
      "grad_norm": 2.8146345615386963,
      "learning_rate": 1e-06,
      "loss": 0.1786,
      "step": 1046
    },
    {
      "epoch": 0.3630374479889043,
      "grad_norm": 3.777596950531006,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 1047
    },
    {
      "epoch": 0.3633841886269071,
      "grad_norm": 2.2806339263916016,
      "learning_rate": 1e-06,
      "loss": 0.1447,
      "step": 1048
    },
    {
      "epoch": 0.36373092926490985,
      "grad_norm": 2.7215936183929443,
      "learning_rate": 1e-06,
      "loss": 0.187,
      "step": 1049
    },
    {
      "epoch": 0.3640776699029126,
      "grad_norm": 3.6170296669006348,
      "learning_rate": 1e-06,
      "loss": 0.192,
      "step": 1050
    },
    {
      "epoch": 0.3644244105409154,
      "grad_norm": 2.5755667686462402,
      "learning_rate": 1e-06,
      "loss": 0.1592,
      "step": 1051
    },
    {
      "epoch": 0.3647711511789182,
      "grad_norm": 2.1443443298339844,
      "learning_rate": 1e-06,
      "loss": 0.1704,
      "step": 1052
    },
    {
      "epoch": 0.36511789181692095,
      "grad_norm": 2.3216934204101562,
      "learning_rate": 1e-06,
      "loss": 0.1787,
      "step": 1053
    },
    {
      "epoch": 0.3654646324549237,
      "grad_norm": 2.0402567386627197,
      "learning_rate": 1e-06,
      "loss": 0.1474,
      "step": 1054
    },
    {
      "epoch": 0.3658113730929265,
      "grad_norm": 2.2533066272735596,
      "learning_rate": 1e-06,
      "loss": 0.2121,
      "step": 1055
    },
    {
      "epoch": 0.36615811373092927,
      "grad_norm": 2.638406753540039,
      "learning_rate": 1e-06,
      "loss": 0.165,
      "step": 1056
    },
    {
      "epoch": 0.36650485436893204,
      "grad_norm": 1.7857331037521362,
      "learning_rate": 1e-06,
      "loss": 0.1653,
      "step": 1057
    },
    {
      "epoch": 0.3668515950069348,
      "grad_norm": 4.700150966644287,
      "learning_rate": 1e-06,
      "loss": 0.1723,
      "step": 1058
    },
    {
      "epoch": 0.3671983356449376,
      "grad_norm": 1.8558968305587769,
      "learning_rate": 1e-06,
      "loss": 0.1674,
      "step": 1059
    },
    {
      "epoch": 0.36754507628294036,
      "grad_norm": 2.047917604446411,
      "learning_rate": 1e-06,
      "loss": 0.1757,
      "step": 1060
    },
    {
      "epoch": 0.36789181692094314,
      "grad_norm": 2.157421588897705,
      "learning_rate": 1e-06,
      "loss": 0.1655,
      "step": 1061
    },
    {
      "epoch": 0.3682385575589459,
      "grad_norm": 4.857705593109131,
      "learning_rate": 1e-06,
      "loss": 0.2087,
      "step": 1062
    },
    {
      "epoch": 0.3685852981969487,
      "grad_norm": 2.143380641937256,
      "learning_rate": 1e-06,
      "loss": 0.1595,
      "step": 1063
    },
    {
      "epoch": 0.36893203883495146,
      "grad_norm": 3.1953229904174805,
      "learning_rate": 1e-06,
      "loss": 0.1592,
      "step": 1064
    },
    {
      "epoch": 0.36927877947295423,
      "grad_norm": 1.5903429985046387,
      "learning_rate": 1e-06,
      "loss": 0.1705,
      "step": 1065
    },
    {
      "epoch": 0.369625520110957,
      "grad_norm": 2.1189424991607666,
      "learning_rate": 1e-06,
      "loss": 0.1611,
      "step": 1066
    },
    {
      "epoch": 0.3699722607489598,
      "grad_norm": 2.089324712753296,
      "learning_rate": 1e-06,
      "loss": 0.1625,
      "step": 1067
    },
    {
      "epoch": 0.37031900138696255,
      "grad_norm": 2.1971182823181152,
      "learning_rate": 1e-06,
      "loss": 0.2072,
      "step": 1068
    },
    {
      "epoch": 0.37066574202496533,
      "grad_norm": 3.347195625305176,
      "learning_rate": 1e-06,
      "loss": 0.1455,
      "step": 1069
    },
    {
      "epoch": 0.3710124826629681,
      "grad_norm": 3.526571035385132,
      "learning_rate": 1e-06,
      "loss": 0.154,
      "step": 1070
    },
    {
      "epoch": 0.3713592233009709,
      "grad_norm": 3.1888411045074463,
      "learning_rate": 1e-06,
      "loss": 0.1559,
      "step": 1071
    },
    {
      "epoch": 0.37170596393897365,
      "grad_norm": 2.727715253829956,
      "learning_rate": 1e-06,
      "loss": 0.1538,
      "step": 1072
    },
    {
      "epoch": 0.3720527045769764,
      "grad_norm": 2.437049150466919,
      "learning_rate": 1e-06,
      "loss": 0.1637,
      "step": 1073
    },
    {
      "epoch": 0.3723994452149792,
      "grad_norm": 4.141301155090332,
      "learning_rate": 1e-06,
      "loss": 0.192,
      "step": 1074
    },
    {
      "epoch": 0.37274618585298197,
      "grad_norm": 3.7499258518218994,
      "learning_rate": 1e-06,
      "loss": 0.1791,
      "step": 1075
    },
    {
      "epoch": 0.37309292649098474,
      "grad_norm": 1.5035871267318726,
      "learning_rate": 1e-06,
      "loss": 0.1553,
      "step": 1076
    },
    {
      "epoch": 0.3734396671289875,
      "grad_norm": 1.849098563194275,
      "learning_rate": 1e-06,
      "loss": 0.1851,
      "step": 1077
    },
    {
      "epoch": 0.3737864077669903,
      "grad_norm": 3.0276803970336914,
      "learning_rate": 1e-06,
      "loss": 0.1682,
      "step": 1078
    },
    {
      "epoch": 0.37413314840499307,
      "grad_norm": 3.6269848346710205,
      "learning_rate": 1e-06,
      "loss": 0.1584,
      "step": 1079
    },
    {
      "epoch": 0.37447988904299584,
      "grad_norm": 1.632621169090271,
      "learning_rate": 1e-06,
      "loss": 0.1598,
      "step": 1080
    },
    {
      "epoch": 0.3748266296809986,
      "grad_norm": 4.8789472579956055,
      "learning_rate": 1e-06,
      "loss": 0.1871,
      "step": 1081
    },
    {
      "epoch": 0.3751733703190014,
      "grad_norm": 2.575338363647461,
      "learning_rate": 1e-06,
      "loss": 0.1781,
      "step": 1082
    },
    {
      "epoch": 0.37552011095700416,
      "grad_norm": 5.345282077789307,
      "learning_rate": 1e-06,
      "loss": 0.1923,
      "step": 1083
    },
    {
      "epoch": 0.37586685159500693,
      "grad_norm": 1.7190324068069458,
      "learning_rate": 1e-06,
      "loss": 0.155,
      "step": 1084
    },
    {
      "epoch": 0.3762135922330097,
      "grad_norm": 2.7497055530548096,
      "learning_rate": 1e-06,
      "loss": 0.1667,
      "step": 1085
    },
    {
      "epoch": 0.3765603328710125,
      "grad_norm": 2.812009334564209,
      "learning_rate": 1e-06,
      "loss": 0.2117,
      "step": 1086
    },
    {
      "epoch": 0.37690707350901526,
      "grad_norm": 1.6642462015151978,
      "learning_rate": 1e-06,
      "loss": 0.1749,
      "step": 1087
    },
    {
      "epoch": 0.37725381414701803,
      "grad_norm": 1.9300440549850464,
      "learning_rate": 1e-06,
      "loss": 0.1683,
      "step": 1088
    },
    {
      "epoch": 0.3776005547850208,
      "grad_norm": 3.1472578048706055,
      "learning_rate": 1e-06,
      "loss": 0.2015,
      "step": 1089
    },
    {
      "epoch": 0.3779472954230236,
      "grad_norm": 2.2062807083129883,
      "learning_rate": 1e-06,
      "loss": 0.1713,
      "step": 1090
    },
    {
      "epoch": 0.37829403606102635,
      "grad_norm": 2.400383472442627,
      "learning_rate": 1e-06,
      "loss": 0.2022,
      "step": 1091
    },
    {
      "epoch": 0.3786407766990291,
      "grad_norm": 1.869706153869629,
      "learning_rate": 1e-06,
      "loss": 0.1863,
      "step": 1092
    },
    {
      "epoch": 0.3789875173370319,
      "grad_norm": 4.637445449829102,
      "learning_rate": 1e-06,
      "loss": 0.1642,
      "step": 1093
    },
    {
      "epoch": 0.37933425797503467,
      "grad_norm": 3.4108481407165527,
      "learning_rate": 1e-06,
      "loss": 0.1535,
      "step": 1094
    },
    {
      "epoch": 0.37968099861303745,
      "grad_norm": 2.6168668270111084,
      "learning_rate": 1e-06,
      "loss": 0.1479,
      "step": 1095
    },
    {
      "epoch": 0.3800277392510402,
      "grad_norm": 2.7562978267669678,
      "learning_rate": 1e-06,
      "loss": 0.2129,
      "step": 1096
    },
    {
      "epoch": 0.380374479889043,
      "grad_norm": 2.557950019836426,
      "learning_rate": 1e-06,
      "loss": 0.1644,
      "step": 1097
    },
    {
      "epoch": 0.38072122052704577,
      "grad_norm": 4.20210075378418,
      "learning_rate": 1e-06,
      "loss": 0.177,
      "step": 1098
    },
    {
      "epoch": 0.38106796116504854,
      "grad_norm": 2.3209750652313232,
      "learning_rate": 1e-06,
      "loss": 0.1768,
      "step": 1099
    },
    {
      "epoch": 0.3814147018030513,
      "grad_norm": 2.318815231323242,
      "learning_rate": 1e-06,
      "loss": 0.2034,
      "step": 1100
    },
    {
      "epoch": 0.3817614424410541,
      "grad_norm": 4.568731307983398,
      "learning_rate": 1e-06,
      "loss": 0.171,
      "step": 1101
    },
    {
      "epoch": 0.38210818307905686,
      "grad_norm": 1.882871150970459,
      "learning_rate": 1e-06,
      "loss": 0.1263,
      "step": 1102
    },
    {
      "epoch": 0.38245492371705964,
      "grad_norm": 2.3090946674346924,
      "learning_rate": 1e-06,
      "loss": 0.2043,
      "step": 1103
    },
    {
      "epoch": 0.3828016643550624,
      "grad_norm": 2.3740408420562744,
      "learning_rate": 1e-06,
      "loss": 0.1483,
      "step": 1104
    },
    {
      "epoch": 0.3831484049930652,
      "grad_norm": 3.6114842891693115,
      "learning_rate": 1e-06,
      "loss": 0.1848,
      "step": 1105
    },
    {
      "epoch": 0.38349514563106796,
      "grad_norm": 1.9591243267059326,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 1106
    },
    {
      "epoch": 0.38384188626907073,
      "grad_norm": 1.7068411111831665,
      "learning_rate": 1e-06,
      "loss": 0.1449,
      "step": 1107
    },
    {
      "epoch": 0.3841886269070735,
      "grad_norm": 1.9355899095535278,
      "learning_rate": 1e-06,
      "loss": 0.1902,
      "step": 1108
    },
    {
      "epoch": 0.3845353675450763,
      "grad_norm": 6.148166179656982,
      "learning_rate": 1e-06,
      "loss": 0.1864,
      "step": 1109
    },
    {
      "epoch": 0.38488210818307905,
      "grad_norm": 3.4833147525787354,
      "learning_rate": 1e-06,
      "loss": 0.1521,
      "step": 1110
    },
    {
      "epoch": 0.3852288488210818,
      "grad_norm": 5.457380294799805,
      "learning_rate": 1e-06,
      "loss": 0.17,
      "step": 1111
    },
    {
      "epoch": 0.3855755894590846,
      "grad_norm": 2.54488468170166,
      "learning_rate": 1e-06,
      "loss": 0.187,
      "step": 1112
    },
    {
      "epoch": 0.3859223300970874,
      "grad_norm": 1.795543909072876,
      "learning_rate": 1e-06,
      "loss": 0.2005,
      "step": 1113
    },
    {
      "epoch": 0.38626907073509015,
      "grad_norm": 2.521865129470825,
      "learning_rate": 1e-06,
      "loss": 0.2035,
      "step": 1114
    },
    {
      "epoch": 0.3866158113730929,
      "grad_norm": 2.5379583835601807,
      "learning_rate": 1e-06,
      "loss": 0.184,
      "step": 1115
    },
    {
      "epoch": 0.3869625520110957,
      "grad_norm": 1.7145198583602905,
      "learning_rate": 1e-06,
      "loss": 0.1709,
      "step": 1116
    },
    {
      "epoch": 0.38730929264909847,
      "grad_norm": 3.1036429405212402,
      "learning_rate": 1e-06,
      "loss": 0.2025,
      "step": 1117
    },
    {
      "epoch": 0.38765603328710124,
      "grad_norm": 2.6453635692596436,
      "learning_rate": 1e-06,
      "loss": 0.1301,
      "step": 1118
    },
    {
      "epoch": 0.388002773925104,
      "grad_norm": 2.3852477073669434,
      "learning_rate": 1e-06,
      "loss": 0.1355,
      "step": 1119
    },
    {
      "epoch": 0.3883495145631068,
      "grad_norm": 1.9018865823745728,
      "learning_rate": 1e-06,
      "loss": 0.1618,
      "step": 1120
    },
    {
      "epoch": 0.38869625520110956,
      "grad_norm": 2.8111538887023926,
      "learning_rate": 1e-06,
      "loss": 0.1624,
      "step": 1121
    },
    {
      "epoch": 0.38904299583911234,
      "grad_norm": 2.226997137069702,
      "learning_rate": 1e-06,
      "loss": 0.1671,
      "step": 1122
    },
    {
      "epoch": 0.3893897364771151,
      "grad_norm": 2.0597381591796875,
      "learning_rate": 1e-06,
      "loss": 0.1604,
      "step": 1123
    },
    {
      "epoch": 0.3897364771151179,
      "grad_norm": 5.358310699462891,
      "learning_rate": 1e-06,
      "loss": 0.2211,
      "step": 1124
    },
    {
      "epoch": 0.39008321775312066,
      "grad_norm": 2.202601909637451,
      "learning_rate": 1e-06,
      "loss": 0.1544,
      "step": 1125
    },
    {
      "epoch": 0.39042995839112343,
      "grad_norm": 1.6089842319488525,
      "learning_rate": 1e-06,
      "loss": 0.1582,
      "step": 1126
    },
    {
      "epoch": 0.3907766990291262,
      "grad_norm": 2.50658917427063,
      "learning_rate": 1e-06,
      "loss": 0.1792,
      "step": 1127
    },
    {
      "epoch": 0.391123439667129,
      "grad_norm": 1.6953543424606323,
      "learning_rate": 1e-06,
      "loss": 0.1624,
      "step": 1128
    },
    {
      "epoch": 0.39147018030513175,
      "grad_norm": 1.8754124641418457,
      "learning_rate": 1e-06,
      "loss": 0.1586,
      "step": 1129
    },
    {
      "epoch": 0.39181692094313453,
      "grad_norm": 4.671680450439453,
      "learning_rate": 1e-06,
      "loss": 0.1969,
      "step": 1130
    },
    {
      "epoch": 0.3921636615811373,
      "grad_norm": 2.1152446269989014,
      "learning_rate": 1e-06,
      "loss": 0.1634,
      "step": 1131
    },
    {
      "epoch": 0.3925104022191401,
      "grad_norm": 4.22470235824585,
      "learning_rate": 1e-06,
      "loss": 0.1453,
      "step": 1132
    },
    {
      "epoch": 0.39285714285714285,
      "grad_norm": 1.6116793155670166,
      "learning_rate": 1e-06,
      "loss": 0.1672,
      "step": 1133
    },
    {
      "epoch": 0.3932038834951456,
      "grad_norm": 1.9633543491363525,
      "learning_rate": 1e-06,
      "loss": 0.1529,
      "step": 1134
    },
    {
      "epoch": 0.3935506241331484,
      "grad_norm": 2.0340752601623535,
      "learning_rate": 1e-06,
      "loss": 0.1708,
      "step": 1135
    },
    {
      "epoch": 0.39389736477115117,
      "grad_norm": 2.2635862827301025,
      "learning_rate": 1e-06,
      "loss": 0.1787,
      "step": 1136
    },
    {
      "epoch": 0.39424410540915394,
      "grad_norm": 2.268153667449951,
      "learning_rate": 1e-06,
      "loss": 0.1992,
      "step": 1137
    },
    {
      "epoch": 0.3945908460471567,
      "grad_norm": 1.7230418920516968,
      "learning_rate": 1e-06,
      "loss": 0.1372,
      "step": 1138
    },
    {
      "epoch": 0.3949375866851595,
      "grad_norm": 1.6801517009735107,
      "learning_rate": 1e-06,
      "loss": 0.1428,
      "step": 1139
    },
    {
      "epoch": 0.39528432732316227,
      "grad_norm": 2.6721580028533936,
      "learning_rate": 1e-06,
      "loss": 0.1597,
      "step": 1140
    },
    {
      "epoch": 0.39563106796116504,
      "grad_norm": 2.748892307281494,
      "learning_rate": 1e-06,
      "loss": 0.1554,
      "step": 1141
    },
    {
      "epoch": 0.3959778085991678,
      "grad_norm": 2.1693079471588135,
      "learning_rate": 1e-06,
      "loss": 0.1856,
      "step": 1142
    },
    {
      "epoch": 0.3963245492371706,
      "grad_norm": 3.472423553466797,
      "learning_rate": 1e-06,
      "loss": 0.1748,
      "step": 1143
    },
    {
      "epoch": 0.39667128987517336,
      "grad_norm": 3.2652645111083984,
      "learning_rate": 1e-06,
      "loss": 0.1756,
      "step": 1144
    },
    {
      "epoch": 0.39701803051317613,
      "grad_norm": 1.8527421951293945,
      "learning_rate": 1e-06,
      "loss": 0.1743,
      "step": 1145
    },
    {
      "epoch": 0.3973647711511789,
      "grad_norm": 1.7605152130126953,
      "learning_rate": 1e-06,
      "loss": 0.1524,
      "step": 1146
    },
    {
      "epoch": 0.3977115117891817,
      "grad_norm": 1.6430308818817139,
      "learning_rate": 1e-06,
      "loss": 0.1571,
      "step": 1147
    },
    {
      "epoch": 0.39805825242718446,
      "grad_norm": 1.800592064857483,
      "learning_rate": 1e-06,
      "loss": 0.1962,
      "step": 1148
    },
    {
      "epoch": 0.39840499306518723,
      "grad_norm": 1.925662875175476,
      "learning_rate": 1e-06,
      "loss": 0.1923,
      "step": 1149
    },
    {
      "epoch": 0.39875173370319,
      "grad_norm": 2.714669942855835,
      "learning_rate": 1e-06,
      "loss": 0.1562,
      "step": 1150
    },
    {
      "epoch": 0.3990984743411928,
      "grad_norm": 1.951225996017456,
      "learning_rate": 1e-06,
      "loss": 0.16,
      "step": 1151
    },
    {
      "epoch": 0.39944521497919555,
      "grad_norm": 3.028454542160034,
      "learning_rate": 1e-06,
      "loss": 0.1745,
      "step": 1152
    },
    {
      "epoch": 0.3997919556171983,
      "grad_norm": 5.070191860198975,
      "learning_rate": 1e-06,
      "loss": 0.1875,
      "step": 1153
    },
    {
      "epoch": 0.4001386962552011,
      "grad_norm": 1.9278351068496704,
      "learning_rate": 1e-06,
      "loss": 0.1608,
      "step": 1154
    },
    {
      "epoch": 0.40048543689320387,
      "grad_norm": 2.457918643951416,
      "learning_rate": 1e-06,
      "loss": 0.2081,
      "step": 1155
    },
    {
      "epoch": 0.40083217753120665,
      "grad_norm": 2.267364501953125,
      "learning_rate": 1e-06,
      "loss": 0.147,
      "step": 1156
    },
    {
      "epoch": 0.4011789181692094,
      "grad_norm": 2.4611542224884033,
      "learning_rate": 1e-06,
      "loss": 0.1763,
      "step": 1157
    },
    {
      "epoch": 0.4015256588072122,
      "grad_norm": 1.9886502027511597,
      "learning_rate": 1e-06,
      "loss": 0.183,
      "step": 1158
    },
    {
      "epoch": 0.40187239944521497,
      "grad_norm": 2.998033285140991,
      "learning_rate": 1e-06,
      "loss": 0.1725,
      "step": 1159
    },
    {
      "epoch": 0.40221914008321774,
      "grad_norm": 5.4129319190979,
      "learning_rate": 1e-06,
      "loss": 0.2027,
      "step": 1160
    },
    {
      "epoch": 0.4025658807212205,
      "grad_norm": 3.6759653091430664,
      "learning_rate": 1e-06,
      "loss": 0.1682,
      "step": 1161
    },
    {
      "epoch": 0.4029126213592233,
      "grad_norm": 2.2087016105651855,
      "learning_rate": 1e-06,
      "loss": 0.1633,
      "step": 1162
    },
    {
      "epoch": 0.40325936199722606,
      "grad_norm": 3.113842725753784,
      "learning_rate": 1e-06,
      "loss": 0.1823,
      "step": 1163
    },
    {
      "epoch": 0.40360610263522884,
      "grad_norm": 1.9368499517440796,
      "learning_rate": 1e-06,
      "loss": 0.1677,
      "step": 1164
    },
    {
      "epoch": 0.4039528432732316,
      "grad_norm": 3.0826315879821777,
      "learning_rate": 1e-06,
      "loss": 0.1492,
      "step": 1165
    },
    {
      "epoch": 0.4042995839112344,
      "grad_norm": 4.97300386428833,
      "learning_rate": 1e-06,
      "loss": 0.1753,
      "step": 1166
    },
    {
      "epoch": 0.40464632454923716,
      "grad_norm": 4.992250442504883,
      "learning_rate": 1e-06,
      "loss": 0.1532,
      "step": 1167
    },
    {
      "epoch": 0.40499306518723993,
      "grad_norm": 7.608388423919678,
      "learning_rate": 1e-06,
      "loss": 0.1693,
      "step": 1168
    },
    {
      "epoch": 0.4053398058252427,
      "grad_norm": 2.3100359439849854,
      "learning_rate": 1e-06,
      "loss": 0.1581,
      "step": 1169
    },
    {
      "epoch": 0.4056865464632455,
      "grad_norm": 2.648092746734619,
      "learning_rate": 1e-06,
      "loss": 0.1489,
      "step": 1170
    },
    {
      "epoch": 0.40603328710124825,
      "grad_norm": 1.7646437883377075,
      "learning_rate": 1e-06,
      "loss": 0.1578,
      "step": 1171
    },
    {
      "epoch": 0.406380027739251,
      "grad_norm": 2.959559679031372,
      "learning_rate": 1e-06,
      "loss": 0.1574,
      "step": 1172
    },
    {
      "epoch": 0.4067267683772538,
      "grad_norm": 2.033961057662964,
      "learning_rate": 1e-06,
      "loss": 0.1493,
      "step": 1173
    },
    {
      "epoch": 0.4070735090152566,
      "grad_norm": 1.6733283996582031,
      "learning_rate": 1e-06,
      "loss": 0.1473,
      "step": 1174
    },
    {
      "epoch": 0.40742024965325935,
      "grad_norm": 2.8953778743743896,
      "learning_rate": 1e-06,
      "loss": 0.1538,
      "step": 1175
    },
    {
      "epoch": 0.4077669902912621,
      "grad_norm": 2.2271323204040527,
      "learning_rate": 1e-06,
      "loss": 0.1749,
      "step": 1176
    },
    {
      "epoch": 0.4081137309292649,
      "grad_norm": 1.6847202777862549,
      "learning_rate": 1e-06,
      "loss": 0.1325,
      "step": 1177
    },
    {
      "epoch": 0.40846047156726767,
      "grad_norm": 5.325228214263916,
      "learning_rate": 1e-06,
      "loss": 0.2158,
      "step": 1178
    },
    {
      "epoch": 0.40880721220527044,
      "grad_norm": 2.9201323986053467,
      "learning_rate": 1e-06,
      "loss": 0.2057,
      "step": 1179
    },
    {
      "epoch": 0.4091539528432732,
      "grad_norm": 3.783559560775757,
      "learning_rate": 1e-06,
      "loss": 0.1741,
      "step": 1180
    },
    {
      "epoch": 0.409500693481276,
      "grad_norm": 2.83846378326416,
      "learning_rate": 1e-06,
      "loss": 0.1893,
      "step": 1181
    },
    {
      "epoch": 0.40984743411927876,
      "grad_norm": 3.131998300552368,
      "learning_rate": 1e-06,
      "loss": 0.1753,
      "step": 1182
    },
    {
      "epoch": 0.41019417475728154,
      "grad_norm": 3.4768800735473633,
      "learning_rate": 1e-06,
      "loss": 0.1638,
      "step": 1183
    },
    {
      "epoch": 0.4105409153952843,
      "grad_norm": 3.8055405616760254,
      "learning_rate": 1e-06,
      "loss": 0.1453,
      "step": 1184
    },
    {
      "epoch": 0.4108876560332871,
      "grad_norm": 3.457226514816284,
      "learning_rate": 1e-06,
      "loss": 0.1789,
      "step": 1185
    },
    {
      "epoch": 0.41123439667128986,
      "grad_norm": 2.19962477684021,
      "learning_rate": 1e-06,
      "loss": 0.1832,
      "step": 1186
    },
    {
      "epoch": 0.41158113730929263,
      "grad_norm": 4.11622428894043,
      "learning_rate": 1e-06,
      "loss": 0.1351,
      "step": 1187
    },
    {
      "epoch": 0.4119278779472954,
      "grad_norm": 2.369816541671753,
      "learning_rate": 1e-06,
      "loss": 0.1943,
      "step": 1188
    },
    {
      "epoch": 0.4122746185852982,
      "grad_norm": 6.653234958648682,
      "learning_rate": 1e-06,
      "loss": 0.1635,
      "step": 1189
    },
    {
      "epoch": 0.41262135922330095,
      "grad_norm": 7.1791887283325195,
      "learning_rate": 1e-06,
      "loss": 0.1677,
      "step": 1190
    },
    {
      "epoch": 0.41296809986130373,
      "grad_norm": 3.2327146530151367,
      "learning_rate": 1e-06,
      "loss": 0.1546,
      "step": 1191
    },
    {
      "epoch": 0.4133148404993065,
      "grad_norm": 3.696610927581787,
      "learning_rate": 1e-06,
      "loss": 0.1586,
      "step": 1192
    },
    {
      "epoch": 0.4136615811373093,
      "grad_norm": 2.113956928253174,
      "learning_rate": 1e-06,
      "loss": 0.1764,
      "step": 1193
    },
    {
      "epoch": 0.41400832177531205,
      "grad_norm": 3.988048553466797,
      "learning_rate": 1e-06,
      "loss": 0.1883,
      "step": 1194
    },
    {
      "epoch": 0.4143550624133148,
      "grad_norm": 2.866166353225708,
      "learning_rate": 1e-06,
      "loss": 0.1633,
      "step": 1195
    },
    {
      "epoch": 0.4147018030513176,
      "grad_norm": 2.072659492492676,
      "learning_rate": 1e-06,
      "loss": 0.1714,
      "step": 1196
    },
    {
      "epoch": 0.41504854368932037,
      "grad_norm": 1.681881308555603,
      "learning_rate": 1e-06,
      "loss": 0.1443,
      "step": 1197
    },
    {
      "epoch": 0.41539528432732314,
      "grad_norm": 1.793905258178711,
      "learning_rate": 1e-06,
      "loss": 0.1258,
      "step": 1198
    },
    {
      "epoch": 0.4157420249653259,
      "grad_norm": 5.42755126953125,
      "learning_rate": 1e-06,
      "loss": 0.2034,
      "step": 1199
    },
    {
      "epoch": 0.4160887656033287,
      "grad_norm": 2.7761476039886475,
      "learning_rate": 1e-06,
      "loss": 0.1226,
      "step": 1200
    },
    {
      "epoch": 0.41643550624133147,
      "grad_norm": 3.929356575012207,
      "learning_rate": 1e-06,
      "loss": 0.1742,
      "step": 1201
    },
    {
      "epoch": 0.41678224687933424,
      "grad_norm": 1.8514819145202637,
      "learning_rate": 1e-06,
      "loss": 0.1527,
      "step": 1202
    },
    {
      "epoch": 0.417128987517337,
      "grad_norm": 1.546459436416626,
      "learning_rate": 1e-06,
      "loss": 0.1307,
      "step": 1203
    },
    {
      "epoch": 0.4174757281553398,
      "grad_norm": 2.12691330909729,
      "learning_rate": 1e-06,
      "loss": 0.2007,
      "step": 1204
    },
    {
      "epoch": 0.41782246879334256,
      "grad_norm": 5.694792747497559,
      "learning_rate": 1e-06,
      "loss": 0.1964,
      "step": 1205
    },
    {
      "epoch": 0.41816920943134533,
      "grad_norm": 5.02055549621582,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 1206
    },
    {
      "epoch": 0.4185159500693481,
      "grad_norm": 2.231020927429199,
      "learning_rate": 1e-06,
      "loss": 0.1956,
      "step": 1207
    },
    {
      "epoch": 0.4188626907073509,
      "grad_norm": 1.8750131130218506,
      "learning_rate": 1e-06,
      "loss": 0.1278,
      "step": 1208
    },
    {
      "epoch": 0.41920943134535366,
      "grad_norm": 2.0554685592651367,
      "learning_rate": 1e-06,
      "loss": 0.1586,
      "step": 1209
    },
    {
      "epoch": 0.41955617198335643,
      "grad_norm": 1.9559741020202637,
      "learning_rate": 1e-06,
      "loss": 0.1896,
      "step": 1210
    },
    {
      "epoch": 0.4199029126213592,
      "grad_norm": 4.545790672302246,
      "learning_rate": 1e-06,
      "loss": 0.1508,
      "step": 1211
    },
    {
      "epoch": 0.420249653259362,
      "grad_norm": 4.00286340713501,
      "learning_rate": 1e-06,
      "loss": 0.142,
      "step": 1212
    },
    {
      "epoch": 0.42059639389736475,
      "grad_norm": 5.119708061218262,
      "learning_rate": 1e-06,
      "loss": 0.1608,
      "step": 1213
    },
    {
      "epoch": 0.4209431345353675,
      "grad_norm": 1.9498828649520874,
      "learning_rate": 1e-06,
      "loss": 0.1715,
      "step": 1214
    },
    {
      "epoch": 0.4212898751733703,
      "grad_norm": 3.187913417816162,
      "learning_rate": 1e-06,
      "loss": 0.1756,
      "step": 1215
    },
    {
      "epoch": 0.42163661581137307,
      "grad_norm": 2.6275475025177,
      "learning_rate": 1e-06,
      "loss": 0.1494,
      "step": 1216
    },
    {
      "epoch": 0.42198335644937585,
      "grad_norm": 4.876014232635498,
      "learning_rate": 1e-06,
      "loss": 0.2037,
      "step": 1217
    },
    {
      "epoch": 0.4223300970873786,
      "grad_norm": 1.6075856685638428,
      "learning_rate": 1e-06,
      "loss": 0.1337,
      "step": 1218
    },
    {
      "epoch": 0.4226768377253814,
      "grad_norm": 2.4928295612335205,
      "learning_rate": 1e-06,
      "loss": 0.1601,
      "step": 1219
    },
    {
      "epoch": 0.42302357836338417,
      "grad_norm": 2.0485427379608154,
      "learning_rate": 1e-06,
      "loss": 0.1533,
      "step": 1220
    },
    {
      "epoch": 0.42337031900138694,
      "grad_norm": 2.0547242164611816,
      "learning_rate": 1e-06,
      "loss": 0.1723,
      "step": 1221
    },
    {
      "epoch": 0.4237170596393897,
      "grad_norm": 1.4388387203216553,
      "learning_rate": 1e-06,
      "loss": 0.1404,
      "step": 1222
    },
    {
      "epoch": 0.4240638002773925,
      "grad_norm": 3.0887644290924072,
      "learning_rate": 1e-06,
      "loss": 0.185,
      "step": 1223
    },
    {
      "epoch": 0.42441054091539526,
      "grad_norm": 2.907358169555664,
      "learning_rate": 1e-06,
      "loss": 0.1759,
      "step": 1224
    },
    {
      "epoch": 0.42475728155339804,
      "grad_norm": 1.9200416803359985,
      "learning_rate": 1e-06,
      "loss": 0.1861,
      "step": 1225
    },
    {
      "epoch": 0.4251040221914008,
      "grad_norm": 3.701133966445923,
      "learning_rate": 1e-06,
      "loss": 0.1473,
      "step": 1226
    },
    {
      "epoch": 0.4254507628294036,
      "grad_norm": 1.728384256362915,
      "learning_rate": 1e-06,
      "loss": 0.1704,
      "step": 1227
    },
    {
      "epoch": 0.42579750346740636,
      "grad_norm": 2.6459665298461914,
      "learning_rate": 1e-06,
      "loss": 0.1521,
      "step": 1228
    },
    {
      "epoch": 0.42614424410540913,
      "grad_norm": 3.3006246089935303,
      "learning_rate": 1e-06,
      "loss": 0.1552,
      "step": 1229
    },
    {
      "epoch": 0.4264909847434119,
      "grad_norm": 1.900184154510498,
      "learning_rate": 1e-06,
      "loss": 0.1576,
      "step": 1230
    },
    {
      "epoch": 0.4268377253814147,
      "grad_norm": 1.7001935243606567,
      "learning_rate": 1e-06,
      "loss": 0.1413,
      "step": 1231
    },
    {
      "epoch": 0.42718446601941745,
      "grad_norm": 2.1855411529541016,
      "learning_rate": 1e-06,
      "loss": 0.1584,
      "step": 1232
    },
    {
      "epoch": 0.4275312066574202,
      "grad_norm": 1.9435559511184692,
      "learning_rate": 1e-06,
      "loss": 0.1849,
      "step": 1233
    },
    {
      "epoch": 0.427877947295423,
      "grad_norm": 1.7659322023391724,
      "learning_rate": 1e-06,
      "loss": 0.145,
      "step": 1234
    },
    {
      "epoch": 0.4282246879334258,
      "grad_norm": 2.6840810775756836,
      "learning_rate": 1e-06,
      "loss": 0.1511,
      "step": 1235
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 1.843663215637207,
      "learning_rate": 1e-06,
      "loss": 0.1653,
      "step": 1236
    },
    {
      "epoch": 0.4289181692094313,
      "grad_norm": 1.7395163774490356,
      "learning_rate": 1e-06,
      "loss": 0.167,
      "step": 1237
    },
    {
      "epoch": 0.4292649098474341,
      "grad_norm": 5.5851664543151855,
      "learning_rate": 1e-06,
      "loss": 0.1967,
      "step": 1238
    },
    {
      "epoch": 0.42961165048543687,
      "grad_norm": 3.6831417083740234,
      "learning_rate": 1e-06,
      "loss": 0.167,
      "step": 1239
    },
    {
      "epoch": 0.42995839112343964,
      "grad_norm": 2.4813308715820312,
      "learning_rate": 1e-06,
      "loss": 0.1743,
      "step": 1240
    },
    {
      "epoch": 0.4303051317614424,
      "grad_norm": 1.4829434156417847,
      "learning_rate": 1e-06,
      "loss": 0.1298,
      "step": 1241
    },
    {
      "epoch": 0.4306518723994452,
      "grad_norm": 6.014243125915527,
      "learning_rate": 1e-06,
      "loss": 0.1911,
      "step": 1242
    },
    {
      "epoch": 0.43099861303744796,
      "grad_norm": 3.8425610065460205,
      "learning_rate": 1e-06,
      "loss": 0.1447,
      "step": 1243
    },
    {
      "epoch": 0.43134535367545074,
      "grad_norm": 2.8617472648620605,
      "learning_rate": 1e-06,
      "loss": 0.1752,
      "step": 1244
    },
    {
      "epoch": 0.4316920943134535,
      "grad_norm": 1.48086678981781,
      "learning_rate": 1e-06,
      "loss": 0.1398,
      "step": 1245
    },
    {
      "epoch": 0.4320388349514563,
      "grad_norm": 1.9533336162567139,
      "learning_rate": 1e-06,
      "loss": 0.1919,
      "step": 1246
    },
    {
      "epoch": 0.43238557558945906,
      "grad_norm": 2.6993250846862793,
      "learning_rate": 1e-06,
      "loss": 0.1761,
      "step": 1247
    },
    {
      "epoch": 0.43273231622746183,
      "grad_norm": 2.825752019882202,
      "learning_rate": 1e-06,
      "loss": 0.1489,
      "step": 1248
    },
    {
      "epoch": 0.4330790568654646,
      "grad_norm": 2.8645107746124268,
      "learning_rate": 1e-06,
      "loss": 0.1537,
      "step": 1249
    },
    {
      "epoch": 0.4334257975034674,
      "grad_norm": 1.9935320615768433,
      "learning_rate": 1e-06,
      "loss": 0.1725,
      "step": 1250
    },
    {
      "epoch": 0.43377253814147015,
      "grad_norm": 3.51171612739563,
      "learning_rate": 1e-06,
      "loss": 0.1773,
      "step": 1251
    },
    {
      "epoch": 0.43411927877947293,
      "grad_norm": 1.9769134521484375,
      "learning_rate": 1e-06,
      "loss": 0.1814,
      "step": 1252
    },
    {
      "epoch": 0.4344660194174757,
      "grad_norm": 2.077732801437378,
      "learning_rate": 1e-06,
      "loss": 0.1867,
      "step": 1253
    },
    {
      "epoch": 0.4348127600554785,
      "grad_norm": 1.734952688217163,
      "learning_rate": 1e-06,
      "loss": 0.1546,
      "step": 1254
    },
    {
      "epoch": 0.43515950069348125,
      "grad_norm": 1.789803385734558,
      "learning_rate": 1e-06,
      "loss": 0.1589,
      "step": 1255
    },
    {
      "epoch": 0.435506241331484,
      "grad_norm": 2.2564432621002197,
      "learning_rate": 1e-06,
      "loss": 0.2033,
      "step": 1256
    },
    {
      "epoch": 0.4358529819694868,
      "grad_norm": 2.3573691844940186,
      "learning_rate": 1e-06,
      "loss": 0.2015,
      "step": 1257
    },
    {
      "epoch": 0.43619972260748957,
      "grad_norm": 1.9252046346664429,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 1258
    },
    {
      "epoch": 0.43654646324549234,
      "grad_norm": 1.6210649013519287,
      "learning_rate": 1e-06,
      "loss": 0.1694,
      "step": 1259
    },
    {
      "epoch": 0.4368932038834951,
      "grad_norm": 2.1824262142181396,
      "learning_rate": 1e-06,
      "loss": 0.1634,
      "step": 1260
    },
    {
      "epoch": 0.4372399445214979,
      "grad_norm": 1.7464162111282349,
      "learning_rate": 1e-06,
      "loss": 0.1703,
      "step": 1261
    },
    {
      "epoch": 0.4375866851595007,
      "grad_norm": 2.9391233921051025,
      "learning_rate": 1e-06,
      "loss": 0.212,
      "step": 1262
    },
    {
      "epoch": 0.4379334257975035,
      "grad_norm": 5.7660722732543945,
      "learning_rate": 1e-06,
      "loss": 0.1808,
      "step": 1263
    },
    {
      "epoch": 0.43828016643550627,
      "grad_norm": 1.7276206016540527,
      "learning_rate": 1e-06,
      "loss": 0.1597,
      "step": 1264
    },
    {
      "epoch": 0.43862690707350904,
      "grad_norm": 1.5537587404251099,
      "learning_rate": 1e-06,
      "loss": 0.1627,
      "step": 1265
    },
    {
      "epoch": 0.4389736477115118,
      "grad_norm": 2.254763126373291,
      "learning_rate": 1e-06,
      "loss": 0.163,
      "step": 1266
    },
    {
      "epoch": 0.4393203883495146,
      "grad_norm": 1.903600811958313,
      "learning_rate": 1e-06,
      "loss": 0.1533,
      "step": 1267
    },
    {
      "epoch": 0.43966712898751736,
      "grad_norm": 2.922935962677002,
      "learning_rate": 1e-06,
      "loss": 0.1979,
      "step": 1268
    },
    {
      "epoch": 0.44001386962552014,
      "grad_norm": 4.926792621612549,
      "learning_rate": 1e-06,
      "loss": 0.1434,
      "step": 1269
    },
    {
      "epoch": 0.4403606102635229,
      "grad_norm": 2.8258774280548096,
      "learning_rate": 1e-06,
      "loss": 0.1578,
      "step": 1270
    },
    {
      "epoch": 0.4407073509015257,
      "grad_norm": 3.5319342613220215,
      "learning_rate": 1e-06,
      "loss": 0.1085,
      "step": 1271
    },
    {
      "epoch": 0.44105409153952846,
      "grad_norm": 1.9517085552215576,
      "learning_rate": 1e-06,
      "loss": 0.1877,
      "step": 1272
    },
    {
      "epoch": 0.44140083217753123,
      "grad_norm": 1.7757554054260254,
      "learning_rate": 1e-06,
      "loss": 0.1774,
      "step": 1273
    },
    {
      "epoch": 0.441747572815534,
      "grad_norm": 2.927002191543579,
      "learning_rate": 1e-06,
      "loss": 0.1516,
      "step": 1274
    },
    {
      "epoch": 0.4420943134535368,
      "grad_norm": 2.922168493270874,
      "learning_rate": 1e-06,
      "loss": 0.1405,
      "step": 1275
    },
    {
      "epoch": 0.44244105409153955,
      "grad_norm": 1.9187227487564087,
      "learning_rate": 1e-06,
      "loss": 0.1424,
      "step": 1276
    },
    {
      "epoch": 0.44278779472954233,
      "grad_norm": 3.046382427215576,
      "learning_rate": 1e-06,
      "loss": 0.1587,
      "step": 1277
    },
    {
      "epoch": 0.4431345353675451,
      "grad_norm": 2.2099106311798096,
      "learning_rate": 1e-06,
      "loss": 0.1518,
      "step": 1278
    },
    {
      "epoch": 0.4434812760055479,
      "grad_norm": 3.655850887298584,
      "learning_rate": 1e-06,
      "loss": 0.1733,
      "step": 1279
    },
    {
      "epoch": 0.44382801664355065,
      "grad_norm": 3.842125415802002,
      "learning_rate": 1e-06,
      "loss": 0.1927,
      "step": 1280
    },
    {
      "epoch": 0.4441747572815534,
      "grad_norm": 1.9096598625183105,
      "learning_rate": 1e-06,
      "loss": 0.128,
      "step": 1281
    },
    {
      "epoch": 0.4445214979195562,
      "grad_norm": 3.1731748580932617,
      "learning_rate": 1e-06,
      "loss": 0.1563,
      "step": 1282
    },
    {
      "epoch": 0.44486823855755897,
      "grad_norm": 2.3779473304748535,
      "learning_rate": 1e-06,
      "loss": 0.171,
      "step": 1283
    },
    {
      "epoch": 0.44521497919556174,
      "grad_norm": 4.150517463684082,
      "learning_rate": 1e-06,
      "loss": 0.1452,
      "step": 1284
    },
    {
      "epoch": 0.4455617198335645,
      "grad_norm": 3.774120807647705,
      "learning_rate": 1e-06,
      "loss": 0.1926,
      "step": 1285
    },
    {
      "epoch": 0.4459084604715673,
      "grad_norm": 3.833528995513916,
      "learning_rate": 1e-06,
      "loss": 0.1771,
      "step": 1286
    },
    {
      "epoch": 0.44625520110957007,
      "grad_norm": 3.3106303215026855,
      "learning_rate": 1e-06,
      "loss": 0.1918,
      "step": 1287
    },
    {
      "epoch": 0.44660194174757284,
      "grad_norm": 3.2806928157806396,
      "learning_rate": 1e-06,
      "loss": 0.147,
      "step": 1288
    },
    {
      "epoch": 0.4469486823855756,
      "grad_norm": 2.7862508296966553,
      "learning_rate": 1e-06,
      "loss": 0.1768,
      "step": 1289
    },
    {
      "epoch": 0.4472954230235784,
      "grad_norm": 2.7099242210388184,
      "learning_rate": 1e-06,
      "loss": 0.2095,
      "step": 1290
    },
    {
      "epoch": 0.44764216366158116,
      "grad_norm": 3.24961256980896,
      "learning_rate": 1e-06,
      "loss": 0.171,
      "step": 1291
    },
    {
      "epoch": 0.44798890429958393,
      "grad_norm": 3.220797061920166,
      "learning_rate": 1e-06,
      "loss": 0.1641,
      "step": 1292
    },
    {
      "epoch": 0.4483356449375867,
      "grad_norm": 4.3876237869262695,
      "learning_rate": 1e-06,
      "loss": 0.1526,
      "step": 1293
    },
    {
      "epoch": 0.4486823855755895,
      "grad_norm": 2.9639668464660645,
      "learning_rate": 1e-06,
      "loss": 0.1649,
      "step": 1294
    },
    {
      "epoch": 0.44902912621359226,
      "grad_norm": 2.0333526134490967,
      "learning_rate": 1e-06,
      "loss": 0.1648,
      "step": 1295
    },
    {
      "epoch": 0.44937586685159503,
      "grad_norm": 3.781501531600952,
      "learning_rate": 1e-06,
      "loss": 0.1571,
      "step": 1296
    },
    {
      "epoch": 0.4497226074895978,
      "grad_norm": 1.5544551610946655,
      "learning_rate": 1e-06,
      "loss": 0.1496,
      "step": 1297
    },
    {
      "epoch": 0.4500693481276006,
      "grad_norm": 5.4716596603393555,
      "learning_rate": 1e-06,
      "loss": 0.199,
      "step": 1298
    },
    {
      "epoch": 0.45041608876560335,
      "grad_norm": 1.998415470123291,
      "learning_rate": 1e-06,
      "loss": 0.1236,
      "step": 1299
    },
    {
      "epoch": 0.4507628294036061,
      "grad_norm": 5.753319263458252,
      "learning_rate": 1e-06,
      "loss": 0.2119,
      "step": 1300
    },
    {
      "epoch": 0.4511095700416089,
      "grad_norm": 3.596370220184326,
      "learning_rate": 1e-06,
      "loss": 0.1901,
      "step": 1301
    },
    {
      "epoch": 0.45145631067961167,
      "grad_norm": 3.2227611541748047,
      "learning_rate": 1e-06,
      "loss": 0.1738,
      "step": 1302
    },
    {
      "epoch": 0.45180305131761445,
      "grad_norm": 3.6115000247955322,
      "learning_rate": 1e-06,
      "loss": 0.2224,
      "step": 1303
    },
    {
      "epoch": 0.4521497919556172,
      "grad_norm": 3.304997682571411,
      "learning_rate": 1e-06,
      "loss": 0.1662,
      "step": 1304
    },
    {
      "epoch": 0.45249653259362,
      "grad_norm": 2.065157175064087,
      "learning_rate": 1e-06,
      "loss": 0.1958,
      "step": 1305
    },
    {
      "epoch": 0.45284327323162277,
      "grad_norm": 3.3705663681030273,
      "learning_rate": 1e-06,
      "loss": 0.133,
      "step": 1306
    },
    {
      "epoch": 0.45319001386962554,
      "grad_norm": 3.053819417953491,
      "learning_rate": 1e-06,
      "loss": 0.1409,
      "step": 1307
    },
    {
      "epoch": 0.4535367545076283,
      "grad_norm": 2.586345911026001,
      "learning_rate": 1e-06,
      "loss": 0.1698,
      "step": 1308
    },
    {
      "epoch": 0.4538834951456311,
      "grad_norm": 1.9915062189102173,
      "learning_rate": 1e-06,
      "loss": 0.1884,
      "step": 1309
    },
    {
      "epoch": 0.45423023578363386,
      "grad_norm": 2.1440911293029785,
      "learning_rate": 1e-06,
      "loss": 0.1587,
      "step": 1310
    },
    {
      "epoch": 0.45457697642163664,
      "grad_norm": 2.2233524322509766,
      "learning_rate": 1e-06,
      "loss": 0.1383,
      "step": 1311
    },
    {
      "epoch": 0.4549237170596394,
      "grad_norm": 1.7290481328964233,
      "learning_rate": 1e-06,
      "loss": 0.1783,
      "step": 1312
    },
    {
      "epoch": 0.4552704576976422,
      "grad_norm": 1.7241450548171997,
      "learning_rate": 1e-06,
      "loss": 0.1505,
      "step": 1313
    },
    {
      "epoch": 0.45561719833564496,
      "grad_norm": 2.3768575191497803,
      "learning_rate": 1e-06,
      "loss": 0.1672,
      "step": 1314
    },
    {
      "epoch": 0.45596393897364773,
      "grad_norm": 2.1226019859313965,
      "learning_rate": 1e-06,
      "loss": 0.1497,
      "step": 1315
    },
    {
      "epoch": 0.4563106796116505,
      "grad_norm": 1.9038270711898804,
      "learning_rate": 1e-06,
      "loss": 0.1664,
      "step": 1316
    },
    {
      "epoch": 0.4566574202496533,
      "grad_norm": 1.9552680253982544,
      "learning_rate": 1e-06,
      "loss": 0.1473,
      "step": 1317
    },
    {
      "epoch": 0.45700416088765605,
      "grad_norm": 1.8354042768478394,
      "learning_rate": 1e-06,
      "loss": 0.1628,
      "step": 1318
    },
    {
      "epoch": 0.4573509015256588,
      "grad_norm": 2.84702467918396,
      "learning_rate": 1e-06,
      "loss": 0.2017,
      "step": 1319
    },
    {
      "epoch": 0.4576976421636616,
      "grad_norm": 2.1125497817993164,
      "learning_rate": 1e-06,
      "loss": 0.1753,
      "step": 1320
    },
    {
      "epoch": 0.4580443828016644,
      "grad_norm": 1.6211894750595093,
      "learning_rate": 1e-06,
      "loss": 0.1473,
      "step": 1321
    },
    {
      "epoch": 0.45839112343966715,
      "grad_norm": 3.268822431564331,
      "learning_rate": 1e-06,
      "loss": 0.1463,
      "step": 1322
    },
    {
      "epoch": 0.4587378640776699,
      "grad_norm": 1.8956842422485352,
      "learning_rate": 1e-06,
      "loss": 0.1445,
      "step": 1323
    },
    {
      "epoch": 0.4590846047156727,
      "grad_norm": 3.552382230758667,
      "learning_rate": 1e-06,
      "loss": 0.1357,
      "step": 1324
    },
    {
      "epoch": 0.45943134535367547,
      "grad_norm": 1.8421423435211182,
      "learning_rate": 1e-06,
      "loss": 0.1479,
      "step": 1325
    },
    {
      "epoch": 0.45977808599167824,
      "grad_norm": 4.14149808883667,
      "learning_rate": 1e-06,
      "loss": 0.1504,
      "step": 1326
    },
    {
      "epoch": 0.460124826629681,
      "grad_norm": 1.7335238456726074,
      "learning_rate": 1e-06,
      "loss": 0.1548,
      "step": 1327
    },
    {
      "epoch": 0.4604715672676838,
      "grad_norm": 4.671210289001465,
      "learning_rate": 1e-06,
      "loss": 0.2263,
      "step": 1328
    },
    {
      "epoch": 0.46081830790568656,
      "grad_norm": 2.813931941986084,
      "learning_rate": 1e-06,
      "loss": 0.1668,
      "step": 1329
    },
    {
      "epoch": 0.46116504854368934,
      "grad_norm": 2.9509613513946533,
      "learning_rate": 1e-06,
      "loss": 0.1484,
      "step": 1330
    },
    {
      "epoch": 0.4615117891816921,
      "grad_norm": 2.949826240539551,
      "learning_rate": 1e-06,
      "loss": 0.1827,
      "step": 1331
    },
    {
      "epoch": 0.4618585298196949,
      "grad_norm": 1.8698536157608032,
      "learning_rate": 1e-06,
      "loss": 0.168,
      "step": 1332
    },
    {
      "epoch": 0.46220527045769766,
      "grad_norm": 3.174593210220337,
      "learning_rate": 1e-06,
      "loss": 0.2109,
      "step": 1333
    },
    {
      "epoch": 0.46255201109570043,
      "grad_norm": 1.977563738822937,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 1334
    },
    {
      "epoch": 0.4628987517337032,
      "grad_norm": 1.8068509101867676,
      "learning_rate": 1e-06,
      "loss": 0.1172,
      "step": 1335
    },
    {
      "epoch": 0.463245492371706,
      "grad_norm": 2.076296329498291,
      "learning_rate": 1e-06,
      "loss": 0.1608,
      "step": 1336
    },
    {
      "epoch": 0.46359223300970875,
      "grad_norm": 2.3091671466827393,
      "learning_rate": 1e-06,
      "loss": 0.1665,
      "step": 1337
    },
    {
      "epoch": 0.46393897364771153,
      "grad_norm": 1.894162654876709,
      "learning_rate": 1e-06,
      "loss": 0.1522,
      "step": 1338
    },
    {
      "epoch": 0.4642857142857143,
      "grad_norm": 1.876305341720581,
      "learning_rate": 1e-06,
      "loss": 0.1444,
      "step": 1339
    },
    {
      "epoch": 0.4646324549237171,
      "grad_norm": 2.8168094158172607,
      "learning_rate": 1e-06,
      "loss": 0.1686,
      "step": 1340
    },
    {
      "epoch": 0.46497919556171985,
      "grad_norm": 2.988266706466675,
      "learning_rate": 1e-06,
      "loss": 0.1545,
      "step": 1341
    },
    {
      "epoch": 0.4653259361997226,
      "grad_norm": 1.7541871070861816,
      "learning_rate": 1e-06,
      "loss": 0.1423,
      "step": 1342
    },
    {
      "epoch": 0.4656726768377254,
      "grad_norm": 4.07562255859375,
      "learning_rate": 1e-06,
      "loss": 0.1447,
      "step": 1343
    },
    {
      "epoch": 0.46601941747572817,
      "grad_norm": 4.45770263671875,
      "learning_rate": 1e-06,
      "loss": 0.1639,
      "step": 1344
    },
    {
      "epoch": 0.46636615811373094,
      "grad_norm": 1.9045706987380981,
      "learning_rate": 1e-06,
      "loss": 0.1765,
      "step": 1345
    },
    {
      "epoch": 0.4667128987517337,
      "grad_norm": 1.927842378616333,
      "learning_rate": 1e-06,
      "loss": 0.1485,
      "step": 1346
    },
    {
      "epoch": 0.4670596393897365,
      "grad_norm": 7.443863868713379,
      "learning_rate": 1e-06,
      "loss": 0.1642,
      "step": 1347
    },
    {
      "epoch": 0.46740638002773927,
      "grad_norm": 2.3793625831604004,
      "learning_rate": 1e-06,
      "loss": 0.1906,
      "step": 1348
    },
    {
      "epoch": 0.46775312066574204,
      "grad_norm": 2.352898120880127,
      "learning_rate": 1e-06,
      "loss": 0.1796,
      "step": 1349
    },
    {
      "epoch": 0.4680998613037448,
      "grad_norm": 3.9866578578948975,
      "learning_rate": 1e-06,
      "loss": 0.1445,
      "step": 1350
    },
    {
      "epoch": 0.4684466019417476,
      "grad_norm": 1.6399085521697998,
      "learning_rate": 1e-06,
      "loss": 0.1484,
      "step": 1351
    },
    {
      "epoch": 0.46879334257975036,
      "grad_norm": 1.9889776706695557,
      "learning_rate": 1e-06,
      "loss": 0.174,
      "step": 1352
    },
    {
      "epoch": 0.46914008321775313,
      "grad_norm": 2.0575404167175293,
      "learning_rate": 1e-06,
      "loss": 0.1832,
      "step": 1353
    },
    {
      "epoch": 0.4694868238557559,
      "grad_norm": 2.2382092475891113,
      "learning_rate": 1e-06,
      "loss": 0.1551,
      "step": 1354
    },
    {
      "epoch": 0.4698335644937587,
      "grad_norm": 2.102724552154541,
      "learning_rate": 1e-06,
      "loss": 0.1566,
      "step": 1355
    },
    {
      "epoch": 0.47018030513176146,
      "grad_norm": 2.692265033721924,
      "learning_rate": 1e-06,
      "loss": 0.1921,
      "step": 1356
    },
    {
      "epoch": 0.47052704576976423,
      "grad_norm": 1.9280555248260498,
      "learning_rate": 1e-06,
      "loss": 0.1678,
      "step": 1357
    },
    {
      "epoch": 0.470873786407767,
      "grad_norm": 1.9627320766448975,
      "learning_rate": 1e-06,
      "loss": 0.1591,
      "step": 1358
    },
    {
      "epoch": 0.4712205270457698,
      "grad_norm": 1.7975282669067383,
      "learning_rate": 1e-06,
      "loss": 0.1312,
      "step": 1359
    },
    {
      "epoch": 0.47156726768377255,
      "grad_norm": 3.5527241230010986,
      "learning_rate": 1e-06,
      "loss": 0.1574,
      "step": 1360
    },
    {
      "epoch": 0.4719140083217753,
      "grad_norm": 3.669248104095459,
      "learning_rate": 1e-06,
      "loss": 0.2004,
      "step": 1361
    },
    {
      "epoch": 0.4722607489597781,
      "grad_norm": 1.9283638000488281,
      "learning_rate": 1e-06,
      "loss": 0.1776,
      "step": 1362
    },
    {
      "epoch": 0.47260748959778087,
      "grad_norm": 1.8212238550186157,
      "learning_rate": 1e-06,
      "loss": 0.1401,
      "step": 1363
    },
    {
      "epoch": 0.47295423023578365,
      "grad_norm": 3.3076553344726562,
      "learning_rate": 1e-06,
      "loss": 0.1734,
      "step": 1364
    },
    {
      "epoch": 0.4733009708737864,
      "grad_norm": 2.9192631244659424,
      "learning_rate": 1e-06,
      "loss": 0.1458,
      "step": 1365
    },
    {
      "epoch": 0.4736477115117892,
      "grad_norm": 5.290735244750977,
      "learning_rate": 1e-06,
      "loss": 0.18,
      "step": 1366
    },
    {
      "epoch": 0.47399445214979197,
      "grad_norm": 3.645962953567505,
      "learning_rate": 1e-06,
      "loss": 0.1608,
      "step": 1367
    },
    {
      "epoch": 0.47434119278779474,
      "grad_norm": 4.024303913116455,
      "learning_rate": 1e-06,
      "loss": 0.1925,
      "step": 1368
    },
    {
      "epoch": 0.4746879334257975,
      "grad_norm": 2.006380081176758,
      "learning_rate": 1e-06,
      "loss": 0.1215,
      "step": 1369
    },
    {
      "epoch": 0.4750346740638003,
      "grad_norm": 3.3509087562561035,
      "learning_rate": 1e-06,
      "loss": 0.1481,
      "step": 1370
    },
    {
      "epoch": 0.47538141470180306,
      "grad_norm": 1.8464128971099854,
      "learning_rate": 1e-06,
      "loss": 0.1433,
      "step": 1371
    },
    {
      "epoch": 0.47572815533980584,
      "grad_norm": 2.2263832092285156,
      "learning_rate": 1e-06,
      "loss": 0.1968,
      "step": 1372
    },
    {
      "epoch": 0.4760748959778086,
      "grad_norm": 2.3892805576324463,
      "learning_rate": 1e-06,
      "loss": 0.1736,
      "step": 1373
    },
    {
      "epoch": 0.4764216366158114,
      "grad_norm": 2.452542781829834,
      "learning_rate": 1e-06,
      "loss": 0.1808,
      "step": 1374
    },
    {
      "epoch": 0.47676837725381416,
      "grad_norm": 2.9188780784606934,
      "learning_rate": 1e-06,
      "loss": 0.1254,
      "step": 1375
    },
    {
      "epoch": 0.47711511789181693,
      "grad_norm": 1.651590347290039,
      "learning_rate": 1e-06,
      "loss": 0.1679,
      "step": 1376
    },
    {
      "epoch": 0.4774618585298197,
      "grad_norm": 1.9459658861160278,
      "learning_rate": 1e-06,
      "loss": 0.1622,
      "step": 1377
    },
    {
      "epoch": 0.4778085991678225,
      "grad_norm": 2.0248286724090576,
      "learning_rate": 1e-06,
      "loss": 0.1365,
      "step": 1378
    },
    {
      "epoch": 0.47815533980582525,
      "grad_norm": 4.144098281860352,
      "learning_rate": 1e-06,
      "loss": 0.1347,
      "step": 1379
    },
    {
      "epoch": 0.478502080443828,
      "grad_norm": 1.46219801902771,
      "learning_rate": 1e-06,
      "loss": 0.1468,
      "step": 1380
    },
    {
      "epoch": 0.4788488210818308,
      "grad_norm": 2.1017038822174072,
      "learning_rate": 1e-06,
      "loss": 0.1694,
      "step": 1381
    },
    {
      "epoch": 0.4791955617198336,
      "grad_norm": 2.1770832538604736,
      "learning_rate": 1e-06,
      "loss": 0.1829,
      "step": 1382
    },
    {
      "epoch": 0.47954230235783635,
      "grad_norm": 3.3750691413879395,
      "learning_rate": 1e-06,
      "loss": 0.1604,
      "step": 1383
    },
    {
      "epoch": 0.4798890429958391,
      "grad_norm": 2.4006078243255615,
      "learning_rate": 1e-06,
      "loss": 0.1437,
      "step": 1384
    },
    {
      "epoch": 0.4802357836338419,
      "grad_norm": 2.0251758098602295,
      "learning_rate": 1e-06,
      "loss": 0.163,
      "step": 1385
    },
    {
      "epoch": 0.48058252427184467,
      "grad_norm": 1.4575427770614624,
      "learning_rate": 1e-06,
      "loss": 0.144,
      "step": 1386
    },
    {
      "epoch": 0.48092926490984744,
      "grad_norm": 2.5154523849487305,
      "learning_rate": 1e-06,
      "loss": 0.1608,
      "step": 1387
    },
    {
      "epoch": 0.4812760055478502,
      "grad_norm": 2.151294231414795,
      "learning_rate": 1e-06,
      "loss": 0.1779,
      "step": 1388
    },
    {
      "epoch": 0.481622746185853,
      "grad_norm": 2.023602247238159,
      "learning_rate": 1e-06,
      "loss": 0.1693,
      "step": 1389
    },
    {
      "epoch": 0.48196948682385576,
      "grad_norm": 2.615605115890503,
      "learning_rate": 1e-06,
      "loss": 0.1406,
      "step": 1390
    },
    {
      "epoch": 0.48231622746185854,
      "grad_norm": 1.9372549057006836,
      "learning_rate": 1e-06,
      "loss": 0.1482,
      "step": 1391
    },
    {
      "epoch": 0.4826629680998613,
      "grad_norm": 2.0867562294006348,
      "learning_rate": 1e-06,
      "loss": 0.1745,
      "step": 1392
    },
    {
      "epoch": 0.4830097087378641,
      "grad_norm": 2.231527090072632,
      "learning_rate": 1e-06,
      "loss": 0.1585,
      "step": 1393
    },
    {
      "epoch": 0.48335644937586686,
      "grad_norm": 1.7819068431854248,
      "learning_rate": 1e-06,
      "loss": 0.1544,
      "step": 1394
    },
    {
      "epoch": 0.48370319001386963,
      "grad_norm": 2.0249552726745605,
      "learning_rate": 1e-06,
      "loss": 0.1983,
      "step": 1395
    },
    {
      "epoch": 0.4840499306518724,
      "grad_norm": 2.4923202991485596,
      "learning_rate": 1e-06,
      "loss": 0.1355,
      "step": 1396
    },
    {
      "epoch": 0.4843966712898752,
      "grad_norm": 1.6963531970977783,
      "learning_rate": 1e-06,
      "loss": 0.1464,
      "step": 1397
    },
    {
      "epoch": 0.48474341192787795,
      "grad_norm": 2.0951385498046875,
      "learning_rate": 1e-06,
      "loss": 0.1533,
      "step": 1398
    },
    {
      "epoch": 0.48509015256588073,
      "grad_norm": 3.5178959369659424,
      "learning_rate": 1e-06,
      "loss": 0.1273,
      "step": 1399
    },
    {
      "epoch": 0.4854368932038835,
      "grad_norm": 2.0446174144744873,
      "learning_rate": 1e-06,
      "loss": 0.1815,
      "step": 1400
    },
    {
      "epoch": 0.4857836338418863,
      "grad_norm": 4.695765972137451,
      "learning_rate": 1e-06,
      "loss": 0.1558,
      "step": 1401
    },
    {
      "epoch": 0.48613037447988905,
      "grad_norm": 2.182650089263916,
      "learning_rate": 1e-06,
      "loss": 0.1541,
      "step": 1402
    },
    {
      "epoch": 0.4864771151178918,
      "grad_norm": 2.1337432861328125,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 1403
    },
    {
      "epoch": 0.4868238557558946,
      "grad_norm": 2.5917625427246094,
      "learning_rate": 1e-06,
      "loss": 0.1789,
      "step": 1404
    },
    {
      "epoch": 0.48717059639389737,
      "grad_norm": 1.9805132150650024,
      "learning_rate": 1e-06,
      "loss": 0.1449,
      "step": 1405
    },
    {
      "epoch": 0.48751733703190014,
      "grad_norm": 2.243947982788086,
      "learning_rate": 1e-06,
      "loss": 0.1448,
      "step": 1406
    },
    {
      "epoch": 0.4878640776699029,
      "grad_norm": 1.755910873413086,
      "learning_rate": 1e-06,
      "loss": 0.1519,
      "step": 1407
    },
    {
      "epoch": 0.4882108183079057,
      "grad_norm": 1.5781508684158325,
      "learning_rate": 1e-06,
      "loss": 0.1566,
      "step": 1408
    },
    {
      "epoch": 0.48855755894590847,
      "grad_norm": 1.7904160022735596,
      "learning_rate": 1e-06,
      "loss": 0.1246,
      "step": 1409
    },
    {
      "epoch": 0.48890429958391124,
      "grad_norm": 3.9564223289489746,
      "learning_rate": 1e-06,
      "loss": 0.1654,
      "step": 1410
    },
    {
      "epoch": 0.489251040221914,
      "grad_norm": 1.6851087808609009,
      "learning_rate": 1e-06,
      "loss": 0.1368,
      "step": 1411
    },
    {
      "epoch": 0.4895977808599168,
      "grad_norm": 3.7889657020568848,
      "learning_rate": 1e-06,
      "loss": 0.1632,
      "step": 1412
    },
    {
      "epoch": 0.48994452149791956,
      "grad_norm": 4.374680995941162,
      "learning_rate": 1e-06,
      "loss": 0.1564,
      "step": 1413
    },
    {
      "epoch": 0.49029126213592233,
      "grad_norm": 3.0737457275390625,
      "learning_rate": 1e-06,
      "loss": 0.1501,
      "step": 1414
    },
    {
      "epoch": 0.4906380027739251,
      "grad_norm": 2.6151061058044434,
      "learning_rate": 1e-06,
      "loss": 0.1595,
      "step": 1415
    },
    {
      "epoch": 0.4909847434119279,
      "grad_norm": 2.651824951171875,
      "learning_rate": 1e-06,
      "loss": 0.1469,
      "step": 1416
    },
    {
      "epoch": 0.49133148404993066,
      "grad_norm": 2.8488190174102783,
      "learning_rate": 1e-06,
      "loss": 0.1519,
      "step": 1417
    },
    {
      "epoch": 0.49167822468793343,
      "grad_norm": 2.576077699661255,
      "learning_rate": 1e-06,
      "loss": 0.153,
      "step": 1418
    },
    {
      "epoch": 0.4920249653259362,
      "grad_norm": 1.594548225402832,
      "learning_rate": 1e-06,
      "loss": 0.1542,
      "step": 1419
    },
    {
      "epoch": 0.492371705963939,
      "grad_norm": 2.900624990463257,
      "learning_rate": 1e-06,
      "loss": 0.1542,
      "step": 1420
    },
    {
      "epoch": 0.49271844660194175,
      "grad_norm": 3.2703654766082764,
      "learning_rate": 1e-06,
      "loss": 0.1285,
      "step": 1421
    },
    {
      "epoch": 0.4930651872399445,
      "grad_norm": 1.6843549013137817,
      "learning_rate": 1e-06,
      "loss": 0.1513,
      "step": 1422
    },
    {
      "epoch": 0.4934119278779473,
      "grad_norm": 1.9763399362564087,
      "learning_rate": 1e-06,
      "loss": 0.1504,
      "step": 1423
    },
    {
      "epoch": 0.49375866851595007,
      "grad_norm": 1.569488763809204,
      "learning_rate": 1e-06,
      "loss": 0.1452,
      "step": 1424
    },
    {
      "epoch": 0.49410540915395285,
      "grad_norm": 2.2819178104400635,
      "learning_rate": 1e-06,
      "loss": 0.1346,
      "step": 1425
    },
    {
      "epoch": 0.4944521497919556,
      "grad_norm": 1.7116472721099854,
      "learning_rate": 1e-06,
      "loss": 0.1288,
      "step": 1426
    },
    {
      "epoch": 0.4947988904299584,
      "grad_norm": 2.4184670448303223,
      "learning_rate": 1e-06,
      "loss": 0.1365,
      "step": 1427
    },
    {
      "epoch": 0.49514563106796117,
      "grad_norm": 2.8264670372009277,
      "learning_rate": 1e-06,
      "loss": 0.1521,
      "step": 1428
    },
    {
      "epoch": 0.49549237170596394,
      "grad_norm": 1.987285852432251,
      "learning_rate": 1e-06,
      "loss": 0.1617,
      "step": 1429
    },
    {
      "epoch": 0.4958391123439667,
      "grad_norm": 2.5394463539123535,
      "learning_rate": 1e-06,
      "loss": 0.1418,
      "step": 1430
    },
    {
      "epoch": 0.4961858529819695,
      "grad_norm": 2.5165560245513916,
      "learning_rate": 1e-06,
      "loss": 0.1605,
      "step": 1431
    },
    {
      "epoch": 0.49653259361997226,
      "grad_norm": 2.378039836883545,
      "learning_rate": 1e-06,
      "loss": 0.1866,
      "step": 1432
    },
    {
      "epoch": 0.49687933425797504,
      "grad_norm": 2.0523364543914795,
      "learning_rate": 1e-06,
      "loss": 0.1298,
      "step": 1433
    },
    {
      "epoch": 0.4972260748959778,
      "grad_norm": 3.141749382019043,
      "learning_rate": 1e-06,
      "loss": 0.1296,
      "step": 1434
    },
    {
      "epoch": 0.4975728155339806,
      "grad_norm": 2.1089577674865723,
      "learning_rate": 1e-06,
      "loss": 0.135,
      "step": 1435
    },
    {
      "epoch": 0.49791955617198336,
      "grad_norm": 1.6610194444656372,
      "learning_rate": 1e-06,
      "loss": 0.1194,
      "step": 1436
    },
    {
      "epoch": 0.49826629680998613,
      "grad_norm": 3.3861851692199707,
      "learning_rate": 1e-06,
      "loss": 0.1649,
      "step": 1437
    },
    {
      "epoch": 0.4986130374479889,
      "grad_norm": 2.972479820251465,
      "learning_rate": 1e-06,
      "loss": 0.1565,
      "step": 1438
    },
    {
      "epoch": 0.4989597780859917,
      "grad_norm": 4.443920612335205,
      "learning_rate": 1e-06,
      "loss": 0.1518,
      "step": 1439
    },
    {
      "epoch": 0.49930651872399445,
      "grad_norm": 3.384650707244873,
      "learning_rate": 1e-06,
      "loss": 0.1399,
      "step": 1440
    },
    {
      "epoch": 0.4996532593619972,
      "grad_norm": 2.9029901027679443,
      "learning_rate": 1e-06,
      "loss": 0.1523,
      "step": 1441
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.911129117012024,
      "learning_rate": 1e-06,
      "loss": 0.1748,
      "step": 1442
    },
    {
      "epoch": 0.5003467406380028,
      "grad_norm": 3.056776285171509,
      "learning_rate": 1e-06,
      "loss": 0.1394,
      "step": 1443
    },
    {
      "epoch": 0.5006934812760055,
      "grad_norm": 2.821357011795044,
      "learning_rate": 1e-06,
      "loss": 0.1946,
      "step": 1444
    },
    {
      "epoch": 0.5010402219140083,
      "grad_norm": 3.346324920654297,
      "learning_rate": 1e-06,
      "loss": 0.1793,
      "step": 1445
    },
    {
      "epoch": 0.5013869625520111,
      "grad_norm": 3.7908129692077637,
      "learning_rate": 1e-06,
      "loss": 0.1635,
      "step": 1446
    },
    {
      "epoch": 0.5017337031900139,
      "grad_norm": 1.877949595451355,
      "learning_rate": 1e-06,
      "loss": 0.1678,
      "step": 1447
    },
    {
      "epoch": 0.5020804438280166,
      "grad_norm": 1.9401321411132812,
      "learning_rate": 1e-06,
      "loss": 0.1575,
      "step": 1448
    },
    {
      "epoch": 0.5024271844660194,
      "grad_norm": 1.9344282150268555,
      "learning_rate": 1e-06,
      "loss": 0.1646,
      "step": 1449
    },
    {
      "epoch": 0.5027739251040222,
      "grad_norm": 2.282867193222046,
      "learning_rate": 1e-06,
      "loss": 0.1888,
      "step": 1450
    },
    {
      "epoch": 0.503120665742025,
      "grad_norm": 3.797800064086914,
      "learning_rate": 1e-06,
      "loss": 0.1256,
      "step": 1451
    },
    {
      "epoch": 0.5034674063800277,
      "grad_norm": 3.1414899826049805,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 1452
    },
    {
      "epoch": 0.5038141470180305,
      "grad_norm": 1.9722864627838135,
      "learning_rate": 1e-06,
      "loss": 0.1408,
      "step": 1453
    },
    {
      "epoch": 0.5041608876560333,
      "grad_norm": 1.5774776935577393,
      "learning_rate": 1e-06,
      "loss": 0.1137,
      "step": 1454
    },
    {
      "epoch": 0.5045076282940361,
      "grad_norm": 3.0094971656799316,
      "learning_rate": 1e-06,
      "loss": 0.1554,
      "step": 1455
    },
    {
      "epoch": 0.5048543689320388,
      "grad_norm": 3.39162015914917,
      "learning_rate": 1e-06,
      "loss": 0.1611,
      "step": 1456
    },
    {
      "epoch": 0.5052011095700416,
      "grad_norm": 2.489313840866089,
      "learning_rate": 1e-06,
      "loss": 0.1535,
      "step": 1457
    },
    {
      "epoch": 0.5055478502080444,
      "grad_norm": 4.133990287780762,
      "learning_rate": 1e-06,
      "loss": 0.1775,
      "step": 1458
    },
    {
      "epoch": 0.5058945908460472,
      "grad_norm": 2.7449488639831543,
      "learning_rate": 1e-06,
      "loss": 0.1495,
      "step": 1459
    },
    {
      "epoch": 0.5062413314840499,
      "grad_norm": 4.382802486419678,
      "learning_rate": 1e-06,
      "loss": 0.133,
      "step": 1460
    },
    {
      "epoch": 0.5065880721220527,
      "grad_norm": 2.110130786895752,
      "learning_rate": 1e-06,
      "loss": 0.1564,
      "step": 1461
    },
    {
      "epoch": 0.5069348127600555,
      "grad_norm": 2.527834415435791,
      "learning_rate": 1e-06,
      "loss": 0.1757,
      "step": 1462
    },
    {
      "epoch": 0.5072815533980582,
      "grad_norm": 2.9167022705078125,
      "learning_rate": 1e-06,
      "loss": 0.1686,
      "step": 1463
    },
    {
      "epoch": 0.507628294036061,
      "grad_norm": 2.2591302394866943,
      "learning_rate": 1e-06,
      "loss": 0.1355,
      "step": 1464
    },
    {
      "epoch": 0.5079750346740638,
      "grad_norm": 4.148419380187988,
      "learning_rate": 1e-06,
      "loss": 0.1391,
      "step": 1465
    },
    {
      "epoch": 0.5083217753120666,
      "grad_norm": 4.044282913208008,
      "learning_rate": 1e-06,
      "loss": 0.1642,
      "step": 1466
    },
    {
      "epoch": 0.5086685159500693,
      "grad_norm": 1.8170924186706543,
      "learning_rate": 1e-06,
      "loss": 0.1218,
      "step": 1467
    },
    {
      "epoch": 0.5090152565880721,
      "grad_norm": 2.241746664047241,
      "learning_rate": 1e-06,
      "loss": 0.1643,
      "step": 1468
    },
    {
      "epoch": 0.5093619972260749,
      "grad_norm": 3.467628240585327,
      "learning_rate": 1e-06,
      "loss": 0.1355,
      "step": 1469
    },
    {
      "epoch": 0.5097087378640777,
      "grad_norm": 3.2897260189056396,
      "learning_rate": 1e-06,
      "loss": 0.145,
      "step": 1470
    },
    {
      "epoch": 0.5100554785020804,
      "grad_norm": 1.7362409830093384,
      "learning_rate": 1e-06,
      "loss": 0.1376,
      "step": 1471
    },
    {
      "epoch": 0.5104022191400832,
      "grad_norm": 2.2274043560028076,
      "learning_rate": 1e-06,
      "loss": 0.1581,
      "step": 1472
    },
    {
      "epoch": 0.510748959778086,
      "grad_norm": 1.8008100986480713,
      "learning_rate": 1e-06,
      "loss": 0.1359,
      "step": 1473
    },
    {
      "epoch": 0.5110957004160888,
      "grad_norm": 1.5842705965042114,
      "learning_rate": 1e-06,
      "loss": 0.1264,
      "step": 1474
    },
    {
      "epoch": 0.5114424410540915,
      "grad_norm": 3.425924301147461,
      "learning_rate": 1e-06,
      "loss": 0.1281,
      "step": 1475
    },
    {
      "epoch": 0.5117891816920943,
      "grad_norm": 2.418201208114624,
      "learning_rate": 1e-06,
      "loss": 0.1695,
      "step": 1476
    },
    {
      "epoch": 0.5121359223300971,
      "grad_norm": 2.0753753185272217,
      "learning_rate": 1e-06,
      "loss": 0.1415,
      "step": 1477
    },
    {
      "epoch": 0.5124826629680999,
      "grad_norm": 2.378450632095337,
      "learning_rate": 1e-06,
      "loss": 0.2064,
      "step": 1478
    },
    {
      "epoch": 0.5128294036061026,
      "grad_norm": 2.1089658737182617,
      "learning_rate": 1e-06,
      "loss": 0.1471,
      "step": 1479
    },
    {
      "epoch": 0.5131761442441054,
      "grad_norm": 2.442574977874756,
      "learning_rate": 1e-06,
      "loss": 0.175,
      "step": 1480
    },
    {
      "epoch": 0.5135228848821082,
      "grad_norm": 2.503943681716919,
      "learning_rate": 1e-06,
      "loss": 0.1389,
      "step": 1481
    },
    {
      "epoch": 0.513869625520111,
      "grad_norm": 1.4603307247161865,
      "learning_rate": 1e-06,
      "loss": 0.1318,
      "step": 1482
    },
    {
      "epoch": 0.5142163661581137,
      "grad_norm": 2.270416498184204,
      "learning_rate": 1e-06,
      "loss": 0.1572,
      "step": 1483
    },
    {
      "epoch": 0.5145631067961165,
      "grad_norm": 2.438401699066162,
      "learning_rate": 1e-06,
      "loss": 0.1337,
      "step": 1484
    },
    {
      "epoch": 0.5149098474341193,
      "grad_norm": 2.581489086151123,
      "learning_rate": 1e-06,
      "loss": 0.1554,
      "step": 1485
    },
    {
      "epoch": 0.515256588072122,
      "grad_norm": 2.019855260848999,
      "learning_rate": 1e-06,
      "loss": 0.1386,
      "step": 1486
    },
    {
      "epoch": 0.5156033287101248,
      "grad_norm": 4.10310173034668,
      "learning_rate": 1e-06,
      "loss": 0.1458,
      "step": 1487
    },
    {
      "epoch": 0.5159500693481276,
      "grad_norm": 1.7705767154693604,
      "learning_rate": 1e-06,
      "loss": 0.1464,
      "step": 1488
    },
    {
      "epoch": 0.5162968099861304,
      "grad_norm": 2.2230114936828613,
      "learning_rate": 1e-06,
      "loss": 0.1376,
      "step": 1489
    },
    {
      "epoch": 0.5166435506241331,
      "grad_norm": 6.302447319030762,
      "learning_rate": 1e-06,
      "loss": 0.1822,
      "step": 1490
    },
    {
      "epoch": 0.5169902912621359,
      "grad_norm": 3.1601343154907227,
      "learning_rate": 1e-06,
      "loss": 0.1539,
      "step": 1491
    },
    {
      "epoch": 0.5173370319001387,
      "grad_norm": 2.8884189128875732,
      "learning_rate": 1e-06,
      "loss": 0.1727,
      "step": 1492
    },
    {
      "epoch": 0.5176837725381415,
      "grad_norm": 2.00549578666687,
      "learning_rate": 1e-06,
      "loss": 0.135,
      "step": 1493
    },
    {
      "epoch": 0.5180305131761442,
      "grad_norm": 3.5855584144592285,
      "learning_rate": 1e-06,
      "loss": 0.2021,
      "step": 1494
    },
    {
      "epoch": 0.518377253814147,
      "grad_norm": 4.1129069328308105,
      "learning_rate": 1e-06,
      "loss": 0.1407,
      "step": 1495
    },
    {
      "epoch": 0.5187239944521498,
      "grad_norm": 2.877685546875,
      "learning_rate": 1e-06,
      "loss": 0.195,
      "step": 1496
    },
    {
      "epoch": 0.5190707350901526,
      "grad_norm": 2.007178544998169,
      "learning_rate": 1e-06,
      "loss": 0.1874,
      "step": 1497
    },
    {
      "epoch": 0.5194174757281553,
      "grad_norm": 1.9459450244903564,
      "learning_rate": 1e-06,
      "loss": 0.1465,
      "step": 1498
    },
    {
      "epoch": 0.5197642163661581,
      "grad_norm": 3.181014060974121,
      "learning_rate": 1e-06,
      "loss": 0.1346,
      "step": 1499
    },
    {
      "epoch": 0.5201109570041609,
      "grad_norm": 2.48248553276062,
      "learning_rate": 1e-06,
      "loss": 0.1531,
      "step": 1500
    },
    {
      "epoch": 0.5204576976421637,
      "grad_norm": 2.8141188621520996,
      "learning_rate": 1e-06,
      "loss": 0.1957,
      "step": 1501
    },
    {
      "epoch": 0.5208044382801664,
      "grad_norm": 1.9651800394058228,
      "learning_rate": 1e-06,
      "loss": 0.1665,
      "step": 1502
    },
    {
      "epoch": 0.5211511789181692,
      "grad_norm": 1.5798146724700928,
      "learning_rate": 1e-06,
      "loss": 0.1452,
      "step": 1503
    },
    {
      "epoch": 0.521497919556172,
      "grad_norm": 1.6036591529846191,
      "learning_rate": 1e-06,
      "loss": 0.1268,
      "step": 1504
    },
    {
      "epoch": 0.5218446601941747,
      "grad_norm": 2.003415107727051,
      "learning_rate": 1e-06,
      "loss": 0.1337,
      "step": 1505
    },
    {
      "epoch": 0.5221914008321775,
      "grad_norm": 3.0233986377716064,
      "learning_rate": 1e-06,
      "loss": 0.2012,
      "step": 1506
    },
    {
      "epoch": 0.5225381414701803,
      "grad_norm": 1.6863069534301758,
      "learning_rate": 1e-06,
      "loss": 0.1401,
      "step": 1507
    },
    {
      "epoch": 0.5228848821081831,
      "grad_norm": 4.177762985229492,
      "learning_rate": 1e-06,
      "loss": 0.1805,
      "step": 1508
    },
    {
      "epoch": 0.5232316227461858,
      "grad_norm": 1.8863307237625122,
      "learning_rate": 1e-06,
      "loss": 0.1557,
      "step": 1509
    },
    {
      "epoch": 0.5235783633841886,
      "grad_norm": 4.922123432159424,
      "learning_rate": 1e-06,
      "loss": 0.211,
      "step": 1510
    },
    {
      "epoch": 0.5239251040221914,
      "grad_norm": 2.77559232711792,
      "learning_rate": 1e-06,
      "loss": 0.1621,
      "step": 1511
    },
    {
      "epoch": 0.5242718446601942,
      "grad_norm": 1.7231181859970093,
      "learning_rate": 1e-06,
      "loss": 0.1443,
      "step": 1512
    },
    {
      "epoch": 0.5246185852981969,
      "grad_norm": 2.0931448936462402,
      "learning_rate": 1e-06,
      "loss": 0.1776,
      "step": 1513
    },
    {
      "epoch": 0.5249653259361997,
      "grad_norm": 2.012320041656494,
      "learning_rate": 1e-06,
      "loss": 0.1765,
      "step": 1514
    },
    {
      "epoch": 0.5253120665742025,
      "grad_norm": 1.9464123249053955,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 1515
    },
    {
      "epoch": 0.5256588072122053,
      "grad_norm": 4.585052013397217,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 1516
    },
    {
      "epoch": 0.526005547850208,
      "grad_norm": 2.2083330154418945,
      "learning_rate": 1e-06,
      "loss": 0.1476,
      "step": 1517
    },
    {
      "epoch": 0.5263522884882108,
      "grad_norm": 2.2266805171966553,
      "learning_rate": 1e-06,
      "loss": 0.1477,
      "step": 1518
    },
    {
      "epoch": 0.5266990291262136,
      "grad_norm": 2.1281206607818604,
      "learning_rate": 1e-06,
      "loss": 0.137,
      "step": 1519
    },
    {
      "epoch": 0.5270457697642164,
      "grad_norm": 2.8681931495666504,
      "learning_rate": 1e-06,
      "loss": 0.1794,
      "step": 1520
    },
    {
      "epoch": 0.5273925104022191,
      "grad_norm": 2.800225257873535,
      "learning_rate": 1e-06,
      "loss": 0.1278,
      "step": 1521
    },
    {
      "epoch": 0.5277392510402219,
      "grad_norm": 4.79030704498291,
      "learning_rate": 1e-06,
      "loss": 0.2031,
      "step": 1522
    },
    {
      "epoch": 0.5280859916782247,
      "grad_norm": 1.678977370262146,
      "learning_rate": 1e-06,
      "loss": 0.1378,
      "step": 1523
    },
    {
      "epoch": 0.5284327323162274,
      "grad_norm": 2.589163064956665,
      "learning_rate": 1e-06,
      "loss": 0.1535,
      "step": 1524
    },
    {
      "epoch": 0.5287794729542302,
      "grad_norm": 2.825577974319458,
      "learning_rate": 1e-06,
      "loss": 0.1411,
      "step": 1525
    },
    {
      "epoch": 0.529126213592233,
      "grad_norm": 2.0116469860076904,
      "learning_rate": 1e-06,
      "loss": 0.1568,
      "step": 1526
    },
    {
      "epoch": 0.5294729542302358,
      "grad_norm": 1.699302077293396,
      "learning_rate": 1e-06,
      "loss": 0.1558,
      "step": 1527
    },
    {
      "epoch": 0.5298196948682385,
      "grad_norm": 2.1007838249206543,
      "learning_rate": 1e-06,
      "loss": 0.1772,
      "step": 1528
    },
    {
      "epoch": 0.5301664355062413,
      "grad_norm": 2.103917360305786,
      "learning_rate": 1e-06,
      "loss": 0.1601,
      "step": 1529
    },
    {
      "epoch": 0.5305131761442441,
      "grad_norm": 3.3394370079040527,
      "learning_rate": 1e-06,
      "loss": 0.1823,
      "step": 1530
    },
    {
      "epoch": 0.5308599167822469,
      "grad_norm": 2.7562904357910156,
      "learning_rate": 1e-06,
      "loss": 0.1509,
      "step": 1531
    },
    {
      "epoch": 0.5312066574202496,
      "grad_norm": 2.3599746227264404,
      "learning_rate": 1e-06,
      "loss": 0.1563,
      "step": 1532
    },
    {
      "epoch": 0.5315533980582524,
      "grad_norm": 2.014448404312134,
      "learning_rate": 1e-06,
      "loss": 0.1273,
      "step": 1533
    },
    {
      "epoch": 0.5319001386962552,
      "grad_norm": 1.5994056463241577,
      "learning_rate": 1e-06,
      "loss": 0.1494,
      "step": 1534
    },
    {
      "epoch": 0.532246879334258,
      "grad_norm": 1.998543620109558,
      "learning_rate": 1e-06,
      "loss": 0.1479,
      "step": 1535
    },
    {
      "epoch": 0.5325936199722607,
      "grad_norm": 1.7799046039581299,
      "learning_rate": 1e-06,
      "loss": 0.1466,
      "step": 1536
    },
    {
      "epoch": 0.5329403606102635,
      "grad_norm": 2.455554723739624,
      "learning_rate": 1e-06,
      "loss": 0.1201,
      "step": 1537
    },
    {
      "epoch": 0.5332871012482663,
      "grad_norm": 1.9349944591522217,
      "learning_rate": 1e-06,
      "loss": 0.1483,
      "step": 1538
    },
    {
      "epoch": 0.5336338418862691,
      "grad_norm": 2.089660406112671,
      "learning_rate": 1e-06,
      "loss": 0.1631,
      "step": 1539
    },
    {
      "epoch": 0.5339805825242718,
      "grad_norm": 2.356295108795166,
      "learning_rate": 1e-06,
      "loss": 0.1547,
      "step": 1540
    },
    {
      "epoch": 0.5343273231622746,
      "grad_norm": 2.6225013732910156,
      "learning_rate": 1e-06,
      "loss": 0.1634,
      "step": 1541
    },
    {
      "epoch": 0.5346740638002774,
      "grad_norm": 4.054690361022949,
      "learning_rate": 1e-06,
      "loss": 0.1597,
      "step": 1542
    },
    {
      "epoch": 0.5350208044382802,
      "grad_norm": 2.222895860671997,
      "learning_rate": 1e-06,
      "loss": 0.1719,
      "step": 1543
    },
    {
      "epoch": 0.5353675450762829,
      "grad_norm": 1.8733757734298706,
      "learning_rate": 1e-06,
      "loss": 0.1387,
      "step": 1544
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 3.893399715423584,
      "learning_rate": 1e-06,
      "loss": 0.1212,
      "step": 1545
    },
    {
      "epoch": 0.5360610263522885,
      "grad_norm": 2.2785661220550537,
      "learning_rate": 1e-06,
      "loss": 0.1573,
      "step": 1546
    },
    {
      "epoch": 0.5364077669902912,
      "grad_norm": 2.716165065765381,
      "learning_rate": 1e-06,
      "loss": 0.1809,
      "step": 1547
    },
    {
      "epoch": 0.536754507628294,
      "grad_norm": 2.239154100418091,
      "learning_rate": 1e-06,
      "loss": 0.1484,
      "step": 1548
    },
    {
      "epoch": 0.5371012482662968,
      "grad_norm": 2.2777199745178223,
      "learning_rate": 1e-06,
      "loss": 0.1776,
      "step": 1549
    },
    {
      "epoch": 0.5374479889042996,
      "grad_norm": 3.0242958068847656,
      "learning_rate": 1e-06,
      "loss": 0.1308,
      "step": 1550
    },
    {
      "epoch": 0.5377947295423023,
      "grad_norm": 2.226205587387085,
      "learning_rate": 1e-06,
      "loss": 0.1737,
      "step": 1551
    },
    {
      "epoch": 0.5381414701803051,
      "grad_norm": 2.850829601287842,
      "learning_rate": 1e-06,
      "loss": 0.1871,
      "step": 1552
    },
    {
      "epoch": 0.5384882108183079,
      "grad_norm": 3.371608257293701,
      "learning_rate": 1e-06,
      "loss": 0.1564,
      "step": 1553
    },
    {
      "epoch": 0.5388349514563107,
      "grad_norm": 2.1789321899414062,
      "learning_rate": 1e-06,
      "loss": 0.1589,
      "step": 1554
    },
    {
      "epoch": 0.5391816920943134,
      "grad_norm": 1.6266393661499023,
      "learning_rate": 1e-06,
      "loss": 0.1507,
      "step": 1555
    },
    {
      "epoch": 0.5395284327323162,
      "grad_norm": 1.8946685791015625,
      "learning_rate": 1e-06,
      "loss": 0.1556,
      "step": 1556
    },
    {
      "epoch": 0.539875173370319,
      "grad_norm": 3.5388271808624268,
      "learning_rate": 1e-06,
      "loss": 0.1408,
      "step": 1557
    },
    {
      "epoch": 0.5402219140083218,
      "grad_norm": 3.0828049182891846,
      "learning_rate": 1e-06,
      "loss": 0.1673,
      "step": 1558
    },
    {
      "epoch": 0.5405686546463245,
      "grad_norm": 2.063697576522827,
      "learning_rate": 1e-06,
      "loss": 0.1394,
      "step": 1559
    },
    {
      "epoch": 0.5409153952843273,
      "grad_norm": 2.2390551567077637,
      "learning_rate": 1e-06,
      "loss": 0.1453,
      "step": 1560
    },
    {
      "epoch": 0.5412621359223301,
      "grad_norm": 4.824660778045654,
      "learning_rate": 1e-06,
      "loss": 0.1882,
      "step": 1561
    },
    {
      "epoch": 0.5416088765603329,
      "grad_norm": 1.9755252599716187,
      "learning_rate": 1e-06,
      "loss": 0.1711,
      "step": 1562
    },
    {
      "epoch": 0.5419556171983356,
      "grad_norm": 1.8800153732299805,
      "learning_rate": 1e-06,
      "loss": 0.1458,
      "step": 1563
    },
    {
      "epoch": 0.5423023578363384,
      "grad_norm": 3.3188271522521973,
      "learning_rate": 1e-06,
      "loss": 0.1564,
      "step": 1564
    },
    {
      "epoch": 0.5426490984743412,
      "grad_norm": 1.953864574432373,
      "learning_rate": 1e-06,
      "loss": 0.1517,
      "step": 1565
    },
    {
      "epoch": 0.542995839112344,
      "grad_norm": 2.689492702484131,
      "learning_rate": 1e-06,
      "loss": 0.1671,
      "step": 1566
    },
    {
      "epoch": 0.5433425797503467,
      "grad_norm": 2.0836610794067383,
      "learning_rate": 1e-06,
      "loss": 0.1683,
      "step": 1567
    },
    {
      "epoch": 0.5436893203883495,
      "grad_norm": 2.045745849609375,
      "learning_rate": 1e-06,
      "loss": 0.1545,
      "step": 1568
    },
    {
      "epoch": 0.5440360610263523,
      "grad_norm": 4.5482401847839355,
      "learning_rate": 1e-06,
      "loss": 0.1518,
      "step": 1569
    },
    {
      "epoch": 0.544382801664355,
      "grad_norm": 2.2771973609924316,
      "learning_rate": 1e-06,
      "loss": 0.1972,
      "step": 1570
    },
    {
      "epoch": 0.5447295423023578,
      "grad_norm": 2.4855430126190186,
      "learning_rate": 1e-06,
      "loss": 0.1717,
      "step": 1571
    },
    {
      "epoch": 0.5450762829403606,
      "grad_norm": 1.9103244543075562,
      "learning_rate": 1e-06,
      "loss": 0.1662,
      "step": 1572
    },
    {
      "epoch": 0.5454230235783634,
      "grad_norm": 2.2929623126983643,
      "learning_rate": 1e-06,
      "loss": 0.1974,
      "step": 1573
    },
    {
      "epoch": 0.5457697642163661,
      "grad_norm": 2.0059826374053955,
      "learning_rate": 1e-06,
      "loss": 0.1622,
      "step": 1574
    },
    {
      "epoch": 0.5461165048543689,
      "grad_norm": 2.1234233379364014,
      "learning_rate": 1e-06,
      "loss": 0.178,
      "step": 1575
    },
    {
      "epoch": 0.5464632454923717,
      "grad_norm": 4.070279598236084,
      "learning_rate": 1e-06,
      "loss": 0.215,
      "step": 1576
    },
    {
      "epoch": 0.5468099861303745,
      "grad_norm": 2.5643842220306396,
      "learning_rate": 1e-06,
      "loss": 0.159,
      "step": 1577
    },
    {
      "epoch": 0.5471567267683772,
      "grad_norm": 1.6535149812698364,
      "learning_rate": 1e-06,
      "loss": 0.146,
      "step": 1578
    },
    {
      "epoch": 0.54750346740638,
      "grad_norm": 2.068358898162842,
      "learning_rate": 1e-06,
      "loss": 0.1691,
      "step": 1579
    },
    {
      "epoch": 0.5478502080443828,
      "grad_norm": 2.275562286376953,
      "learning_rate": 1e-06,
      "loss": 0.1564,
      "step": 1580
    },
    {
      "epoch": 0.5481969486823856,
      "grad_norm": 2.2625136375427246,
      "learning_rate": 1e-06,
      "loss": 0.1399,
      "step": 1581
    },
    {
      "epoch": 0.5485436893203883,
      "grad_norm": 2.553527593612671,
      "learning_rate": 1e-06,
      "loss": 0.1339,
      "step": 1582
    },
    {
      "epoch": 0.5488904299583911,
      "grad_norm": 2.405510425567627,
      "learning_rate": 1e-06,
      "loss": 0.1518,
      "step": 1583
    },
    {
      "epoch": 0.5492371705963939,
      "grad_norm": 2.205770969390869,
      "learning_rate": 1e-06,
      "loss": 0.1656,
      "step": 1584
    },
    {
      "epoch": 0.5495839112343966,
      "grad_norm": 2.961263418197632,
      "learning_rate": 1e-06,
      "loss": 0.1549,
      "step": 1585
    },
    {
      "epoch": 0.5499306518723994,
      "grad_norm": 2.239316701889038,
      "learning_rate": 1e-06,
      "loss": 0.1637,
      "step": 1586
    },
    {
      "epoch": 0.5502773925104022,
      "grad_norm": 2.7622323036193848,
      "learning_rate": 1e-06,
      "loss": 0.1726,
      "step": 1587
    },
    {
      "epoch": 0.550624133148405,
      "grad_norm": 2.3954594135284424,
      "learning_rate": 1e-06,
      "loss": 0.152,
      "step": 1588
    },
    {
      "epoch": 0.5509708737864077,
      "grad_norm": 2.528770923614502,
      "learning_rate": 1e-06,
      "loss": 0.1328,
      "step": 1589
    },
    {
      "epoch": 0.5513176144244105,
      "grad_norm": 4.014219760894775,
      "learning_rate": 1e-06,
      "loss": 0.1167,
      "step": 1590
    },
    {
      "epoch": 0.5516643550624133,
      "grad_norm": 1.444198727607727,
      "learning_rate": 1e-06,
      "loss": 0.1387,
      "step": 1591
    },
    {
      "epoch": 0.5520110957004161,
      "grad_norm": 1.6705647706985474,
      "learning_rate": 1e-06,
      "loss": 0.1104,
      "step": 1592
    },
    {
      "epoch": 0.5523578363384188,
      "grad_norm": 2.243098735809326,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 1593
    },
    {
      "epoch": 0.5527045769764216,
      "grad_norm": 2.3548195362091064,
      "learning_rate": 1e-06,
      "loss": 0.1432,
      "step": 1594
    },
    {
      "epoch": 0.5530513176144244,
      "grad_norm": 1.7457523345947266,
      "learning_rate": 1e-06,
      "loss": 0.131,
      "step": 1595
    },
    {
      "epoch": 0.5533980582524272,
      "grad_norm": 2.6490869522094727,
      "learning_rate": 1e-06,
      "loss": 0.1571,
      "step": 1596
    },
    {
      "epoch": 0.5537447988904299,
      "grad_norm": 3.868406057357788,
      "learning_rate": 1e-06,
      "loss": 0.1713,
      "step": 1597
    },
    {
      "epoch": 0.5540915395284327,
      "grad_norm": 3.2075355052948,
      "learning_rate": 1e-06,
      "loss": 0.1467,
      "step": 1598
    },
    {
      "epoch": 0.5544382801664355,
      "grad_norm": 2.0262351036071777,
      "learning_rate": 1e-06,
      "loss": 0.1598,
      "step": 1599
    },
    {
      "epoch": 0.5547850208044383,
      "grad_norm": 2.029439926147461,
      "learning_rate": 1e-06,
      "loss": 0.1578,
      "step": 1600
    },
    {
      "epoch": 0.555131761442441,
      "grad_norm": 1.949021339416504,
      "learning_rate": 1e-06,
      "loss": 0.1524,
      "step": 1601
    },
    {
      "epoch": 0.5554785020804438,
      "grad_norm": 4.777602195739746,
      "learning_rate": 1e-06,
      "loss": 0.1553,
      "step": 1602
    },
    {
      "epoch": 0.5558252427184466,
      "grad_norm": 2.1241836547851562,
      "learning_rate": 1e-06,
      "loss": 0.1868,
      "step": 1603
    },
    {
      "epoch": 0.5561719833564494,
      "grad_norm": 2.2839200496673584,
      "learning_rate": 1e-06,
      "loss": 0.151,
      "step": 1604
    },
    {
      "epoch": 0.5565187239944521,
      "grad_norm": 2.7008562088012695,
      "learning_rate": 1e-06,
      "loss": 0.132,
      "step": 1605
    },
    {
      "epoch": 0.5568654646324549,
      "grad_norm": 2.1138343811035156,
      "learning_rate": 1e-06,
      "loss": 0.1349,
      "step": 1606
    },
    {
      "epoch": 0.5572122052704577,
      "grad_norm": 2.0788066387176514,
      "learning_rate": 1e-06,
      "loss": 0.1803,
      "step": 1607
    },
    {
      "epoch": 0.5575589459084604,
      "grad_norm": 2.323345899581909,
      "learning_rate": 1e-06,
      "loss": 0.1749,
      "step": 1608
    },
    {
      "epoch": 0.5579056865464632,
      "grad_norm": 3.4080355167388916,
      "learning_rate": 1e-06,
      "loss": 0.1453,
      "step": 1609
    },
    {
      "epoch": 0.558252427184466,
      "grad_norm": 2.0123281478881836,
      "learning_rate": 1e-06,
      "loss": 0.1345,
      "step": 1610
    },
    {
      "epoch": 0.5585991678224688,
      "grad_norm": 2.1220619678497314,
      "learning_rate": 1e-06,
      "loss": 0.1823,
      "step": 1611
    },
    {
      "epoch": 0.5589459084604715,
      "grad_norm": 2.4971022605895996,
      "learning_rate": 1e-06,
      "loss": 0.1591,
      "step": 1612
    },
    {
      "epoch": 0.5592926490984743,
      "grad_norm": 1.527266025543213,
      "learning_rate": 1e-06,
      "loss": 0.1459,
      "step": 1613
    },
    {
      "epoch": 0.5596393897364771,
      "grad_norm": 2.1191792488098145,
      "learning_rate": 1e-06,
      "loss": 0.1411,
      "step": 1614
    },
    {
      "epoch": 0.5599861303744799,
      "grad_norm": 1.978819489479065,
      "learning_rate": 1e-06,
      "loss": 0.1459,
      "step": 1615
    },
    {
      "epoch": 0.5603328710124826,
      "grad_norm": 2.3482275009155273,
      "learning_rate": 1e-06,
      "loss": 0.1784,
      "step": 1616
    },
    {
      "epoch": 0.5606796116504854,
      "grad_norm": 1.5257140398025513,
      "learning_rate": 1e-06,
      "loss": 0.142,
      "step": 1617
    },
    {
      "epoch": 0.5610263522884882,
      "grad_norm": 3.646221399307251,
      "learning_rate": 1e-06,
      "loss": 0.1756,
      "step": 1618
    },
    {
      "epoch": 0.561373092926491,
      "grad_norm": 1.7928833961486816,
      "learning_rate": 1e-06,
      "loss": 0.1423,
      "step": 1619
    },
    {
      "epoch": 0.5617198335644937,
      "grad_norm": 2.0367002487182617,
      "learning_rate": 1e-06,
      "loss": 0.1382,
      "step": 1620
    },
    {
      "epoch": 0.5620665742024965,
      "grad_norm": 1.9482024908065796,
      "learning_rate": 1e-06,
      "loss": 0.1573,
      "step": 1621
    },
    {
      "epoch": 0.5624133148404993,
      "grad_norm": 2.336115598678589,
      "learning_rate": 1e-06,
      "loss": 0.147,
      "step": 1622
    },
    {
      "epoch": 0.562760055478502,
      "grad_norm": 4.602109909057617,
      "learning_rate": 1e-06,
      "loss": 0.1847,
      "step": 1623
    },
    {
      "epoch": 0.5631067961165048,
      "grad_norm": 2.374413251876831,
      "learning_rate": 1e-06,
      "loss": 0.1668,
      "step": 1624
    },
    {
      "epoch": 0.5634535367545076,
      "grad_norm": 3.371469497680664,
      "learning_rate": 1e-06,
      "loss": 0.1654,
      "step": 1625
    },
    {
      "epoch": 0.5638002773925104,
      "grad_norm": 1.608858585357666,
      "learning_rate": 1e-06,
      "loss": 0.1259,
      "step": 1626
    },
    {
      "epoch": 0.5641470180305131,
      "grad_norm": 1.8223319053649902,
      "learning_rate": 1e-06,
      "loss": 0.1599,
      "step": 1627
    },
    {
      "epoch": 0.5644937586685159,
      "grad_norm": 3.2663540840148926,
      "learning_rate": 1e-06,
      "loss": 0.1525,
      "step": 1628
    },
    {
      "epoch": 0.5648404993065187,
      "grad_norm": 2.5820138454437256,
      "learning_rate": 1e-06,
      "loss": 0.1748,
      "step": 1629
    },
    {
      "epoch": 0.5651872399445215,
      "grad_norm": 2.0734665393829346,
      "learning_rate": 1e-06,
      "loss": 0.1456,
      "step": 1630
    },
    {
      "epoch": 0.5655339805825242,
      "grad_norm": 2.0002732276916504,
      "learning_rate": 1e-06,
      "loss": 0.1547,
      "step": 1631
    },
    {
      "epoch": 0.565880721220527,
      "grad_norm": 2.328211545944214,
      "learning_rate": 1e-06,
      "loss": 0.1603,
      "step": 1632
    },
    {
      "epoch": 0.5662274618585298,
      "grad_norm": 3.090592861175537,
      "learning_rate": 1e-06,
      "loss": 0.1467,
      "step": 1633
    },
    {
      "epoch": 0.5665742024965326,
      "grad_norm": 1.7330372333526611,
      "learning_rate": 1e-06,
      "loss": 0.1289,
      "step": 1634
    },
    {
      "epoch": 0.5669209431345353,
      "grad_norm": 1.6095702648162842,
      "learning_rate": 1e-06,
      "loss": 0.1386,
      "step": 1635
    },
    {
      "epoch": 0.5672676837725381,
      "grad_norm": 2.5205414295196533,
      "learning_rate": 1e-06,
      "loss": 0.1817,
      "step": 1636
    },
    {
      "epoch": 0.5676144244105409,
      "grad_norm": 2.5302205085754395,
      "learning_rate": 1e-06,
      "loss": 0.1515,
      "step": 1637
    },
    {
      "epoch": 0.5679611650485437,
      "grad_norm": 2.0151143074035645,
      "learning_rate": 1e-06,
      "loss": 0.1376,
      "step": 1638
    },
    {
      "epoch": 0.5683079056865464,
      "grad_norm": 4.463467121124268,
      "learning_rate": 1e-06,
      "loss": 0.1865,
      "step": 1639
    },
    {
      "epoch": 0.5686546463245492,
      "grad_norm": 2.7108094692230225,
      "learning_rate": 1e-06,
      "loss": 0.1527,
      "step": 1640
    },
    {
      "epoch": 0.569001386962552,
      "grad_norm": 3.0464956760406494,
      "learning_rate": 1e-06,
      "loss": 0.1671,
      "step": 1641
    },
    {
      "epoch": 0.5693481276005548,
      "grad_norm": 2.9522790908813477,
      "learning_rate": 1e-06,
      "loss": 0.1863,
      "step": 1642
    },
    {
      "epoch": 0.5696948682385575,
      "grad_norm": 2.414750099182129,
      "learning_rate": 1e-06,
      "loss": 0.1537,
      "step": 1643
    },
    {
      "epoch": 0.5700416088765603,
      "grad_norm": 1.9789800643920898,
      "learning_rate": 1e-06,
      "loss": 0.1405,
      "step": 1644
    },
    {
      "epoch": 0.5703883495145631,
      "grad_norm": 2.12092661857605,
      "learning_rate": 1e-06,
      "loss": 0.1452,
      "step": 1645
    },
    {
      "epoch": 0.5707350901525658,
      "grad_norm": 2.106600046157837,
      "learning_rate": 1e-06,
      "loss": 0.1769,
      "step": 1646
    },
    {
      "epoch": 0.5710818307905686,
      "grad_norm": 1.6850626468658447,
      "learning_rate": 1e-06,
      "loss": 0.1415,
      "step": 1647
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 2.36781907081604,
      "learning_rate": 1e-06,
      "loss": 0.1574,
      "step": 1648
    },
    {
      "epoch": 0.5717753120665742,
      "grad_norm": 3.573758840560913,
      "learning_rate": 1e-06,
      "loss": 0.1754,
      "step": 1649
    },
    {
      "epoch": 0.5721220527045769,
      "grad_norm": 2.212064743041992,
      "learning_rate": 1e-06,
      "loss": 0.1398,
      "step": 1650
    },
    {
      "epoch": 0.5724687933425797,
      "grad_norm": 3.7138683795928955,
      "learning_rate": 1e-06,
      "loss": 0.1486,
      "step": 1651
    },
    {
      "epoch": 0.5728155339805825,
      "grad_norm": 2.328507900238037,
      "learning_rate": 1e-06,
      "loss": 0.181,
      "step": 1652
    },
    {
      "epoch": 0.5731622746185853,
      "grad_norm": 2.0306859016418457,
      "learning_rate": 1e-06,
      "loss": 0.1678,
      "step": 1653
    },
    {
      "epoch": 0.573509015256588,
      "grad_norm": 2.6543784141540527,
      "learning_rate": 1e-06,
      "loss": 0.1798,
      "step": 1654
    },
    {
      "epoch": 0.5738557558945908,
      "grad_norm": 2.0634257793426514,
      "learning_rate": 1e-06,
      "loss": 0.1597,
      "step": 1655
    },
    {
      "epoch": 0.5742024965325936,
      "grad_norm": 3.299668788909912,
      "learning_rate": 1e-06,
      "loss": 0.1498,
      "step": 1656
    },
    {
      "epoch": 0.5745492371705964,
      "grad_norm": 1.8128036260604858,
      "learning_rate": 1e-06,
      "loss": 0.1252,
      "step": 1657
    },
    {
      "epoch": 0.5748959778085991,
      "grad_norm": 2.365511894226074,
      "learning_rate": 1e-06,
      "loss": 0.1581,
      "step": 1658
    },
    {
      "epoch": 0.5752427184466019,
      "grad_norm": 2.4852473735809326,
      "learning_rate": 1e-06,
      "loss": 0.1702,
      "step": 1659
    },
    {
      "epoch": 0.5755894590846047,
      "grad_norm": 2.0349810123443604,
      "learning_rate": 1e-06,
      "loss": 0.1515,
      "step": 1660
    },
    {
      "epoch": 0.5759361997226075,
      "grad_norm": 1.8448779582977295,
      "learning_rate": 1e-06,
      "loss": 0.1553,
      "step": 1661
    },
    {
      "epoch": 0.5762829403606102,
      "grad_norm": 3.345630645751953,
      "learning_rate": 1e-06,
      "loss": 0.1508,
      "step": 1662
    },
    {
      "epoch": 0.576629680998613,
      "grad_norm": 2.9367878437042236,
      "learning_rate": 1e-06,
      "loss": 0.1637,
      "step": 1663
    },
    {
      "epoch": 0.5769764216366158,
      "grad_norm": 1.9184025526046753,
      "learning_rate": 1e-06,
      "loss": 0.1452,
      "step": 1664
    },
    {
      "epoch": 0.5773231622746186,
      "grad_norm": 3.0208446979522705,
      "learning_rate": 1e-06,
      "loss": 0.1319,
      "step": 1665
    },
    {
      "epoch": 0.5776699029126213,
      "grad_norm": 1.9824481010437012,
      "learning_rate": 1e-06,
      "loss": 0.1285,
      "step": 1666
    },
    {
      "epoch": 0.5780166435506241,
      "grad_norm": 1.768882393836975,
      "learning_rate": 1e-06,
      "loss": 0.1589,
      "step": 1667
    },
    {
      "epoch": 0.5783633841886269,
      "grad_norm": 5.792402267456055,
      "learning_rate": 1e-06,
      "loss": 0.1643,
      "step": 1668
    },
    {
      "epoch": 0.5787101248266296,
      "grad_norm": 2.2808361053466797,
      "learning_rate": 1e-06,
      "loss": 0.1925,
      "step": 1669
    },
    {
      "epoch": 0.5790568654646324,
      "grad_norm": 1.661110520362854,
      "learning_rate": 1e-06,
      "loss": 0.1383,
      "step": 1670
    },
    {
      "epoch": 0.5794036061026352,
      "grad_norm": 2.764539957046509,
      "learning_rate": 1e-06,
      "loss": 0.1346,
      "step": 1671
    },
    {
      "epoch": 0.579750346740638,
      "grad_norm": 2.591721296310425,
      "learning_rate": 1e-06,
      "loss": 0.1683,
      "step": 1672
    },
    {
      "epoch": 0.5800970873786407,
      "grad_norm": 1.6234064102172852,
      "learning_rate": 1e-06,
      "loss": 0.1638,
      "step": 1673
    },
    {
      "epoch": 0.5804438280166435,
      "grad_norm": 3.9577250480651855,
      "learning_rate": 1e-06,
      "loss": 0.1731,
      "step": 1674
    },
    {
      "epoch": 0.5807905686546463,
      "grad_norm": 3.799262762069702,
      "learning_rate": 1e-06,
      "loss": 0.1556,
      "step": 1675
    },
    {
      "epoch": 0.5811373092926491,
      "grad_norm": 2.9457106590270996,
      "learning_rate": 1e-06,
      "loss": 0.1611,
      "step": 1676
    },
    {
      "epoch": 0.5814840499306518,
      "grad_norm": 2.1177031993865967,
      "learning_rate": 1e-06,
      "loss": 0.1723,
      "step": 1677
    },
    {
      "epoch": 0.5818307905686546,
      "grad_norm": 1.5257512331008911,
      "learning_rate": 1e-06,
      "loss": 0.132,
      "step": 1678
    },
    {
      "epoch": 0.5821775312066574,
      "grad_norm": 2.6540706157684326,
      "learning_rate": 1e-06,
      "loss": 0.1648,
      "step": 1679
    },
    {
      "epoch": 0.5825242718446602,
      "grad_norm": 2.4282069206237793,
      "learning_rate": 1e-06,
      "loss": 0.1462,
      "step": 1680
    },
    {
      "epoch": 0.5828710124826629,
      "grad_norm": 1.7106441259384155,
      "learning_rate": 1e-06,
      "loss": 0.1528,
      "step": 1681
    },
    {
      "epoch": 0.5832177531206657,
      "grad_norm": 2.3406167030334473,
      "learning_rate": 1e-06,
      "loss": 0.1909,
      "step": 1682
    },
    {
      "epoch": 0.5835644937586685,
      "grad_norm": 2.504379987716675,
      "learning_rate": 1e-06,
      "loss": 0.1444,
      "step": 1683
    },
    {
      "epoch": 0.5839112343966713,
      "grad_norm": 4.548439979553223,
      "learning_rate": 1e-06,
      "loss": 0.1736,
      "step": 1684
    },
    {
      "epoch": 0.584257975034674,
      "grad_norm": 2.725050449371338,
      "learning_rate": 1e-06,
      "loss": 0.1406,
      "step": 1685
    },
    {
      "epoch": 0.5846047156726768,
      "grad_norm": 2.026104211807251,
      "learning_rate": 1e-06,
      "loss": 0.1531,
      "step": 1686
    },
    {
      "epoch": 0.5849514563106796,
      "grad_norm": 2.1058292388916016,
      "learning_rate": 1e-06,
      "loss": 0.1434,
      "step": 1687
    },
    {
      "epoch": 0.5852981969486823,
      "grad_norm": 2.1308064460754395,
      "learning_rate": 1e-06,
      "loss": 0.1529,
      "step": 1688
    },
    {
      "epoch": 0.5856449375866851,
      "grad_norm": 4.247373104095459,
      "learning_rate": 1e-06,
      "loss": 0.1962,
      "step": 1689
    },
    {
      "epoch": 0.5859916782246879,
      "grad_norm": 4.272634029388428,
      "learning_rate": 1e-06,
      "loss": 0.1319,
      "step": 1690
    },
    {
      "epoch": 0.5863384188626907,
      "grad_norm": 1.9161146879196167,
      "learning_rate": 1e-06,
      "loss": 0.1794,
      "step": 1691
    },
    {
      "epoch": 0.5866851595006934,
      "grad_norm": 3.2819652557373047,
      "learning_rate": 1e-06,
      "loss": 0.1388,
      "step": 1692
    },
    {
      "epoch": 0.5870319001386962,
      "grad_norm": 2.6425933837890625,
      "learning_rate": 1e-06,
      "loss": 0.1508,
      "step": 1693
    },
    {
      "epoch": 0.587378640776699,
      "grad_norm": 2.2003285884857178,
      "learning_rate": 1e-06,
      "loss": 0.1315,
      "step": 1694
    },
    {
      "epoch": 0.5877253814147018,
      "grad_norm": 2.436519145965576,
      "learning_rate": 1e-06,
      "loss": 0.1888,
      "step": 1695
    },
    {
      "epoch": 0.5880721220527045,
      "grad_norm": 2.4235095977783203,
      "learning_rate": 1e-06,
      "loss": 0.1574,
      "step": 1696
    },
    {
      "epoch": 0.5884188626907073,
      "grad_norm": 2.7736899852752686,
      "learning_rate": 1e-06,
      "loss": 0.1324,
      "step": 1697
    },
    {
      "epoch": 0.5887656033287101,
      "grad_norm": 6.031193256378174,
      "learning_rate": 1e-06,
      "loss": 0.1635,
      "step": 1698
    },
    {
      "epoch": 0.5891123439667129,
      "grad_norm": 1.7690778970718384,
      "learning_rate": 1e-06,
      "loss": 0.1351,
      "step": 1699
    },
    {
      "epoch": 0.5894590846047156,
      "grad_norm": 1.9819053411483765,
      "learning_rate": 1e-06,
      "loss": 0.1649,
      "step": 1700
    },
    {
      "epoch": 0.5898058252427184,
      "grad_norm": 2.2417149543762207,
      "learning_rate": 1e-06,
      "loss": 0.1551,
      "step": 1701
    },
    {
      "epoch": 0.5901525658807212,
      "grad_norm": 1.7193907499313354,
      "learning_rate": 1e-06,
      "loss": 0.1373,
      "step": 1702
    },
    {
      "epoch": 0.590499306518724,
      "grad_norm": 2.16274356842041,
      "learning_rate": 1e-06,
      "loss": 0.1629,
      "step": 1703
    },
    {
      "epoch": 0.5908460471567267,
      "grad_norm": 2.157911539077759,
      "learning_rate": 1e-06,
      "loss": 0.1289,
      "step": 1704
    },
    {
      "epoch": 0.5911927877947295,
      "grad_norm": 3.0244007110595703,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 1705
    },
    {
      "epoch": 0.5915395284327323,
      "grad_norm": 1.9083997011184692,
      "learning_rate": 1e-06,
      "loss": 0.1736,
      "step": 1706
    },
    {
      "epoch": 0.591886269070735,
      "grad_norm": 2.4870333671569824,
      "learning_rate": 1e-06,
      "loss": 0.1541,
      "step": 1707
    },
    {
      "epoch": 0.5922330097087378,
      "grad_norm": 2.714510202407837,
      "learning_rate": 1e-06,
      "loss": 0.1561,
      "step": 1708
    },
    {
      "epoch": 0.5925797503467406,
      "grad_norm": 1.7150081396102905,
      "learning_rate": 1e-06,
      "loss": 0.1437,
      "step": 1709
    },
    {
      "epoch": 0.5929264909847434,
      "grad_norm": 2.5376031398773193,
      "learning_rate": 1e-06,
      "loss": 0.1588,
      "step": 1710
    },
    {
      "epoch": 0.5932732316227461,
      "grad_norm": 2.173816680908203,
      "learning_rate": 1e-06,
      "loss": 0.1668,
      "step": 1711
    },
    {
      "epoch": 0.5936199722607489,
      "grad_norm": 3.2527916431427,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 1712
    },
    {
      "epoch": 0.5939667128987517,
      "grad_norm": 2.0413970947265625,
      "learning_rate": 1e-06,
      "loss": 0.1535,
      "step": 1713
    },
    {
      "epoch": 0.5943134535367545,
      "grad_norm": 2.2500665187835693,
      "learning_rate": 1e-06,
      "loss": 0.1744,
      "step": 1714
    },
    {
      "epoch": 0.5946601941747572,
      "grad_norm": 2.931854248046875,
      "learning_rate": 1e-06,
      "loss": 0.1729,
      "step": 1715
    },
    {
      "epoch": 0.59500693481276,
      "grad_norm": 1.800407886505127,
      "learning_rate": 1e-06,
      "loss": 0.148,
      "step": 1716
    },
    {
      "epoch": 0.5953536754507628,
      "grad_norm": 3.381312370300293,
      "learning_rate": 1e-06,
      "loss": 0.1358,
      "step": 1717
    },
    {
      "epoch": 0.5957004160887656,
      "grad_norm": 3.778513193130493,
      "learning_rate": 1e-06,
      "loss": 0.156,
      "step": 1718
    },
    {
      "epoch": 0.5960471567267683,
      "grad_norm": 1.6460663080215454,
      "learning_rate": 1e-06,
      "loss": 0.1479,
      "step": 1719
    },
    {
      "epoch": 0.5963938973647711,
      "grad_norm": 2.169105052947998,
      "learning_rate": 1e-06,
      "loss": 0.1712,
      "step": 1720
    },
    {
      "epoch": 0.5967406380027739,
      "grad_norm": 1.761500358581543,
      "learning_rate": 1e-06,
      "loss": 0.1442,
      "step": 1721
    },
    {
      "epoch": 0.5970873786407767,
      "grad_norm": 1.7620950937271118,
      "learning_rate": 1e-06,
      "loss": 0.1423,
      "step": 1722
    },
    {
      "epoch": 0.5974341192787794,
      "grad_norm": 2.3129122257232666,
      "learning_rate": 1e-06,
      "loss": 0.1765,
      "step": 1723
    },
    {
      "epoch": 0.5977808599167822,
      "grad_norm": 2.858874559402466,
      "learning_rate": 1e-06,
      "loss": 0.1467,
      "step": 1724
    },
    {
      "epoch": 0.598127600554785,
      "grad_norm": 3.7320244312286377,
      "learning_rate": 1e-06,
      "loss": 0.1796,
      "step": 1725
    },
    {
      "epoch": 0.5984743411927878,
      "grad_norm": 3.407935380935669,
      "learning_rate": 1e-06,
      "loss": 0.1386,
      "step": 1726
    },
    {
      "epoch": 0.5988210818307905,
      "grad_norm": 2.2188820838928223,
      "learning_rate": 1e-06,
      "loss": 0.1723,
      "step": 1727
    },
    {
      "epoch": 0.5991678224687933,
      "grad_norm": 1.947733759880066,
      "learning_rate": 1e-06,
      "loss": 0.1716,
      "step": 1728
    },
    {
      "epoch": 0.5995145631067961,
      "grad_norm": 2.274256467819214,
      "learning_rate": 1e-06,
      "loss": 0.1519,
      "step": 1729
    },
    {
      "epoch": 0.5998613037447988,
      "grad_norm": 1.881277084350586,
      "learning_rate": 1e-06,
      "loss": 0.127,
      "step": 1730
    },
    {
      "epoch": 0.6002080443828016,
      "grad_norm": 2.522675037384033,
      "learning_rate": 1e-06,
      "loss": 0.1806,
      "step": 1731
    },
    {
      "epoch": 0.6005547850208044,
      "grad_norm": 3.52107834815979,
      "learning_rate": 1e-06,
      "loss": 0.1581,
      "step": 1732
    },
    {
      "epoch": 0.6009015256588072,
      "grad_norm": 2.1973533630371094,
      "learning_rate": 1e-06,
      "loss": 0.126,
      "step": 1733
    },
    {
      "epoch": 0.6012482662968099,
      "grad_norm": 2.215819835662842,
      "learning_rate": 1e-06,
      "loss": 0.1454,
      "step": 1734
    },
    {
      "epoch": 0.6015950069348127,
      "grad_norm": 1.8062198162078857,
      "learning_rate": 1e-06,
      "loss": 0.1583,
      "step": 1735
    },
    {
      "epoch": 0.6019417475728155,
      "grad_norm": 2.090193510055542,
      "learning_rate": 1e-06,
      "loss": 0.1434,
      "step": 1736
    },
    {
      "epoch": 0.6022884882108183,
      "grad_norm": 1.7331854104995728,
      "learning_rate": 1e-06,
      "loss": 0.1323,
      "step": 1737
    },
    {
      "epoch": 0.602635228848821,
      "grad_norm": 2.4180166721343994,
      "learning_rate": 1e-06,
      "loss": 0.1384,
      "step": 1738
    },
    {
      "epoch": 0.6029819694868238,
      "grad_norm": 3.263031482696533,
      "learning_rate": 1e-06,
      "loss": 0.1324,
      "step": 1739
    },
    {
      "epoch": 0.6033287101248266,
      "grad_norm": 2.341269016265869,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 1740
    },
    {
      "epoch": 0.6036754507628294,
      "grad_norm": 3.2346274852752686,
      "learning_rate": 1e-06,
      "loss": 0.1382,
      "step": 1741
    },
    {
      "epoch": 0.6040221914008321,
      "grad_norm": 1.9541387557983398,
      "learning_rate": 1e-06,
      "loss": 0.1673,
      "step": 1742
    },
    {
      "epoch": 0.6043689320388349,
      "grad_norm": 3.1931421756744385,
      "learning_rate": 1e-06,
      "loss": 0.1437,
      "step": 1743
    },
    {
      "epoch": 0.6047156726768377,
      "grad_norm": 1.9747329950332642,
      "learning_rate": 1e-06,
      "loss": 0.1346,
      "step": 1744
    },
    {
      "epoch": 0.6050624133148405,
      "grad_norm": 3.360968589782715,
      "learning_rate": 1e-06,
      "loss": 0.1584,
      "step": 1745
    },
    {
      "epoch": 0.6054091539528432,
      "grad_norm": 2.160388469696045,
      "learning_rate": 1e-06,
      "loss": 0.132,
      "step": 1746
    },
    {
      "epoch": 0.605755894590846,
      "grad_norm": 4.136070728302002,
      "learning_rate": 1e-06,
      "loss": 0.1506,
      "step": 1747
    },
    {
      "epoch": 0.6061026352288488,
      "grad_norm": 2.1393015384674072,
      "learning_rate": 1e-06,
      "loss": 0.1568,
      "step": 1748
    },
    {
      "epoch": 0.6064493758668515,
      "grad_norm": 2.9097952842712402,
      "learning_rate": 1e-06,
      "loss": 0.1343,
      "step": 1749
    },
    {
      "epoch": 0.6067961165048543,
      "grad_norm": 1.9220006465911865,
      "learning_rate": 1e-06,
      "loss": 0.1446,
      "step": 1750
    },
    {
      "epoch": 0.6071428571428571,
      "grad_norm": 2.2562577724456787,
      "learning_rate": 1e-06,
      "loss": 0.1751,
      "step": 1751
    },
    {
      "epoch": 0.6074895977808599,
      "grad_norm": 2.5833182334899902,
      "learning_rate": 1e-06,
      "loss": 0.1606,
      "step": 1752
    },
    {
      "epoch": 0.6078363384188626,
      "grad_norm": 6.893828392028809,
      "learning_rate": 1e-06,
      "loss": 0.1516,
      "step": 1753
    },
    {
      "epoch": 0.6081830790568654,
      "grad_norm": 6.116687774658203,
      "learning_rate": 1e-06,
      "loss": 0.1175,
      "step": 1754
    },
    {
      "epoch": 0.6085298196948682,
      "grad_norm": 1.879164218902588,
      "learning_rate": 1e-06,
      "loss": 0.1381,
      "step": 1755
    },
    {
      "epoch": 0.608876560332871,
      "grad_norm": 3.298027753829956,
      "learning_rate": 1e-06,
      "loss": 0.1726,
      "step": 1756
    },
    {
      "epoch": 0.6092233009708737,
      "grad_norm": 2.1097750663757324,
      "learning_rate": 1e-06,
      "loss": 0.165,
      "step": 1757
    },
    {
      "epoch": 0.6095700416088765,
      "grad_norm": 4.932060241699219,
      "learning_rate": 1e-06,
      "loss": 0.1399,
      "step": 1758
    },
    {
      "epoch": 0.6099167822468793,
      "grad_norm": 1.9408305883407593,
      "learning_rate": 1e-06,
      "loss": 0.1661,
      "step": 1759
    },
    {
      "epoch": 0.6102635228848821,
      "grad_norm": 1.8838698863983154,
      "learning_rate": 1e-06,
      "loss": 0.1617,
      "step": 1760
    },
    {
      "epoch": 0.6106102635228848,
      "grad_norm": 2.995274782180786,
      "learning_rate": 1e-06,
      "loss": 0.1361,
      "step": 1761
    },
    {
      "epoch": 0.6109570041608876,
      "grad_norm": 2.9905359745025635,
      "learning_rate": 1e-06,
      "loss": 0.1485,
      "step": 1762
    },
    {
      "epoch": 0.6113037447988904,
      "grad_norm": 2.9674386978149414,
      "learning_rate": 1e-06,
      "loss": 0.1503,
      "step": 1763
    },
    {
      "epoch": 0.6116504854368932,
      "grad_norm": 2.390683889389038,
      "learning_rate": 1e-06,
      "loss": 0.1585,
      "step": 1764
    },
    {
      "epoch": 0.6119972260748959,
      "grad_norm": 2.8863577842712402,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 1765
    },
    {
      "epoch": 0.6123439667128987,
      "grad_norm": 3.183380365371704,
      "learning_rate": 1e-06,
      "loss": 0.1292,
      "step": 1766
    },
    {
      "epoch": 0.6126907073509015,
      "grad_norm": 2.080923318862915,
      "learning_rate": 1e-06,
      "loss": 0.1166,
      "step": 1767
    },
    {
      "epoch": 0.6130374479889042,
      "grad_norm": 2.8261148929595947,
      "learning_rate": 1e-06,
      "loss": 0.159,
      "step": 1768
    },
    {
      "epoch": 0.613384188626907,
      "grad_norm": 2.0629355907440186,
      "learning_rate": 1e-06,
      "loss": 0.1433,
      "step": 1769
    },
    {
      "epoch": 0.6137309292649098,
      "grad_norm": 1.8249051570892334,
      "learning_rate": 1e-06,
      "loss": 0.1448,
      "step": 1770
    },
    {
      "epoch": 0.6140776699029126,
      "grad_norm": 2.489915132522583,
      "learning_rate": 1e-06,
      "loss": 0.1975,
      "step": 1771
    },
    {
      "epoch": 0.6144244105409153,
      "grad_norm": 2.171525716781616,
      "learning_rate": 1e-06,
      "loss": 0.143,
      "step": 1772
    },
    {
      "epoch": 0.6147711511789181,
      "grad_norm": 3.3242056369781494,
      "learning_rate": 1e-06,
      "loss": 0.1337,
      "step": 1773
    },
    {
      "epoch": 0.6151178918169209,
      "grad_norm": 3.56597900390625,
      "learning_rate": 1e-06,
      "loss": 0.1278,
      "step": 1774
    },
    {
      "epoch": 0.6154646324549237,
      "grad_norm": 1.8567633628845215,
      "learning_rate": 1e-06,
      "loss": 0.1546,
      "step": 1775
    },
    {
      "epoch": 0.6158113730929264,
      "grad_norm": 2.3509087562561035,
      "learning_rate": 1e-06,
      "loss": 0.1677,
      "step": 1776
    },
    {
      "epoch": 0.6161581137309292,
      "grad_norm": 2.3500568866729736,
      "learning_rate": 1e-06,
      "loss": 0.1573,
      "step": 1777
    },
    {
      "epoch": 0.616504854368932,
      "grad_norm": 1.687183141708374,
      "learning_rate": 1e-06,
      "loss": 0.1384,
      "step": 1778
    },
    {
      "epoch": 0.6168515950069348,
      "grad_norm": 2.0229380130767822,
      "learning_rate": 1e-06,
      "loss": 0.1231,
      "step": 1779
    },
    {
      "epoch": 0.6171983356449375,
      "grad_norm": 2.1885972023010254,
      "learning_rate": 1e-06,
      "loss": 0.1656,
      "step": 1780
    },
    {
      "epoch": 0.6175450762829403,
      "grad_norm": 4.203535556793213,
      "learning_rate": 1e-06,
      "loss": 0.1725,
      "step": 1781
    },
    {
      "epoch": 0.6178918169209431,
      "grad_norm": 2.7054553031921387,
      "learning_rate": 1e-06,
      "loss": 0.1316,
      "step": 1782
    },
    {
      "epoch": 0.6182385575589459,
      "grad_norm": 2.386046886444092,
      "learning_rate": 1e-06,
      "loss": 0.1086,
      "step": 1783
    },
    {
      "epoch": 0.6185852981969486,
      "grad_norm": 2.22220516204834,
      "learning_rate": 1e-06,
      "loss": 0.199,
      "step": 1784
    },
    {
      "epoch": 0.6189320388349514,
      "grad_norm": 2.0740716457366943,
      "learning_rate": 1e-06,
      "loss": 0.1167,
      "step": 1785
    },
    {
      "epoch": 0.6192787794729542,
      "grad_norm": 3.928887128829956,
      "learning_rate": 1e-06,
      "loss": 0.1659,
      "step": 1786
    },
    {
      "epoch": 0.619625520110957,
      "grad_norm": 3.2319023609161377,
      "learning_rate": 1e-06,
      "loss": 0.1538,
      "step": 1787
    },
    {
      "epoch": 0.6199722607489597,
      "grad_norm": 2.2219529151916504,
      "learning_rate": 1e-06,
      "loss": 0.1689,
      "step": 1788
    },
    {
      "epoch": 0.6203190013869625,
      "grad_norm": 1.9555144309997559,
      "learning_rate": 1e-06,
      "loss": 0.1572,
      "step": 1789
    },
    {
      "epoch": 0.6206657420249653,
      "grad_norm": 1.8324187994003296,
      "learning_rate": 1e-06,
      "loss": 0.1549,
      "step": 1790
    },
    {
      "epoch": 0.621012482662968,
      "grad_norm": 4.268178462982178,
      "learning_rate": 1e-06,
      "loss": 0.1676,
      "step": 1791
    },
    {
      "epoch": 0.6213592233009708,
      "grad_norm": 2.28480863571167,
      "learning_rate": 1e-06,
      "loss": 0.1555,
      "step": 1792
    },
    {
      "epoch": 0.6217059639389736,
      "grad_norm": 2.970492362976074,
      "learning_rate": 1e-06,
      "loss": 0.1369,
      "step": 1793
    },
    {
      "epoch": 0.6220527045769764,
      "grad_norm": 2.2355594635009766,
      "learning_rate": 1e-06,
      "loss": 0.2016,
      "step": 1794
    },
    {
      "epoch": 0.6223994452149791,
      "grad_norm": 2.1921377182006836,
      "learning_rate": 1e-06,
      "loss": 0.1664,
      "step": 1795
    },
    {
      "epoch": 0.6227461858529819,
      "grad_norm": 2.3036246299743652,
      "learning_rate": 1e-06,
      "loss": 0.1617,
      "step": 1796
    },
    {
      "epoch": 0.6230929264909847,
      "grad_norm": 2.2555198669433594,
      "learning_rate": 1e-06,
      "loss": 0.1693,
      "step": 1797
    },
    {
      "epoch": 0.6234396671289875,
      "grad_norm": 2.416224718093872,
      "learning_rate": 1e-06,
      "loss": 0.1745,
      "step": 1798
    },
    {
      "epoch": 0.6237864077669902,
      "grad_norm": 1.9002783298492432,
      "learning_rate": 1e-06,
      "loss": 0.1618,
      "step": 1799
    },
    {
      "epoch": 0.624133148404993,
      "grad_norm": 1.8858561515808105,
      "learning_rate": 1e-06,
      "loss": 0.1885,
      "step": 1800
    },
    {
      "epoch": 0.6244798890429958,
      "grad_norm": 3.2046573162078857,
      "learning_rate": 1e-06,
      "loss": 0.1588,
      "step": 1801
    },
    {
      "epoch": 0.6248266296809986,
      "grad_norm": 3.2242980003356934,
      "learning_rate": 1e-06,
      "loss": 0.1758,
      "step": 1802
    },
    {
      "epoch": 0.6251733703190014,
      "grad_norm": 4.321735858917236,
      "learning_rate": 1e-06,
      "loss": 0.1745,
      "step": 1803
    },
    {
      "epoch": 0.6255201109570042,
      "grad_norm": 5.258564472198486,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 1804
    },
    {
      "epoch": 0.625866851595007,
      "grad_norm": 2.0180282592773438,
      "learning_rate": 1e-06,
      "loss": 0.1327,
      "step": 1805
    },
    {
      "epoch": 0.6262135922330098,
      "grad_norm": 2.150374412536621,
      "learning_rate": 1e-06,
      "loss": 0.129,
      "step": 1806
    },
    {
      "epoch": 0.6265603328710125,
      "grad_norm": 2.426352024078369,
      "learning_rate": 1e-06,
      "loss": 0.155,
      "step": 1807
    },
    {
      "epoch": 0.6269070735090153,
      "grad_norm": 3.8777828216552734,
      "learning_rate": 1e-06,
      "loss": 0.134,
      "step": 1808
    },
    {
      "epoch": 0.6272538141470181,
      "grad_norm": 2.8608689308166504,
      "learning_rate": 1e-06,
      "loss": 0.1723,
      "step": 1809
    },
    {
      "epoch": 0.6276005547850209,
      "grad_norm": 2.592209577560425,
      "learning_rate": 1e-06,
      "loss": 0.1473,
      "step": 1810
    },
    {
      "epoch": 0.6279472954230236,
      "grad_norm": 1.9494845867156982,
      "learning_rate": 1e-06,
      "loss": 0.1771,
      "step": 1811
    },
    {
      "epoch": 0.6282940360610264,
      "grad_norm": 1.7008180618286133,
      "learning_rate": 1e-06,
      "loss": 0.1429,
      "step": 1812
    },
    {
      "epoch": 0.6286407766990292,
      "grad_norm": 1.7186037302017212,
      "learning_rate": 1e-06,
      "loss": 0.161,
      "step": 1813
    },
    {
      "epoch": 0.628987517337032,
      "grad_norm": 2.3491291999816895,
      "learning_rate": 1e-06,
      "loss": 0.1767,
      "step": 1814
    },
    {
      "epoch": 0.6293342579750347,
      "grad_norm": 2.4897589683532715,
      "learning_rate": 1e-06,
      "loss": 0.1528,
      "step": 1815
    },
    {
      "epoch": 0.6296809986130375,
      "grad_norm": 1.9620110988616943,
      "learning_rate": 1e-06,
      "loss": 0.1481,
      "step": 1816
    },
    {
      "epoch": 0.6300277392510403,
      "grad_norm": 2.08915114402771,
      "learning_rate": 1e-06,
      "loss": 0.1465,
      "step": 1817
    },
    {
      "epoch": 0.630374479889043,
      "grad_norm": 3.368116617202759,
      "learning_rate": 1e-06,
      "loss": 0.1516,
      "step": 1818
    },
    {
      "epoch": 0.6307212205270458,
      "grad_norm": 2.13322377204895,
      "learning_rate": 1e-06,
      "loss": 0.153,
      "step": 1819
    },
    {
      "epoch": 0.6310679611650486,
      "grad_norm": 2.479769229888916,
      "learning_rate": 1e-06,
      "loss": 0.1646,
      "step": 1820
    },
    {
      "epoch": 0.6314147018030514,
      "grad_norm": 3.7209994792938232,
      "learning_rate": 1e-06,
      "loss": 0.1795,
      "step": 1821
    },
    {
      "epoch": 0.6317614424410541,
      "grad_norm": 4.989049434661865,
      "learning_rate": 1e-06,
      "loss": 0.1582,
      "step": 1822
    },
    {
      "epoch": 0.6321081830790569,
      "grad_norm": 2.99127459526062,
      "learning_rate": 1e-06,
      "loss": 0.1149,
      "step": 1823
    },
    {
      "epoch": 0.6324549237170597,
      "grad_norm": 3.3212974071502686,
      "learning_rate": 1e-06,
      "loss": 0.1359,
      "step": 1824
    },
    {
      "epoch": 0.6328016643550625,
      "grad_norm": 2.3676962852478027,
      "learning_rate": 1e-06,
      "loss": 0.1448,
      "step": 1825
    },
    {
      "epoch": 0.6331484049930652,
      "grad_norm": 2.0515060424804688,
      "learning_rate": 1e-06,
      "loss": 0.1557,
      "step": 1826
    },
    {
      "epoch": 0.633495145631068,
      "grad_norm": 3.021026372909546,
      "learning_rate": 1e-06,
      "loss": 0.143,
      "step": 1827
    },
    {
      "epoch": 0.6338418862690708,
      "grad_norm": 2.328517198562622,
      "learning_rate": 1e-06,
      "loss": 0.1265,
      "step": 1828
    },
    {
      "epoch": 0.6341886269070736,
      "grad_norm": 1.7838841676712036,
      "learning_rate": 1e-06,
      "loss": 0.1188,
      "step": 1829
    },
    {
      "epoch": 0.6345353675450763,
      "grad_norm": 2.873117685317993,
      "learning_rate": 1e-06,
      "loss": 0.1251,
      "step": 1830
    },
    {
      "epoch": 0.6348821081830791,
      "grad_norm": 2.515712022781372,
      "learning_rate": 1e-06,
      "loss": 0.1258,
      "step": 1831
    },
    {
      "epoch": 0.6352288488210819,
      "grad_norm": 4.897085189819336,
      "learning_rate": 1e-06,
      "loss": 0.1528,
      "step": 1832
    },
    {
      "epoch": 0.6355755894590847,
      "grad_norm": 2.860135793685913,
      "learning_rate": 1e-06,
      "loss": 0.1385,
      "step": 1833
    },
    {
      "epoch": 0.6359223300970874,
      "grad_norm": 2.492570400238037,
      "learning_rate": 1e-06,
      "loss": 0.1761,
      "step": 1834
    },
    {
      "epoch": 0.6362690707350902,
      "grad_norm": 1.7522510290145874,
      "learning_rate": 1e-06,
      "loss": 0.1309,
      "step": 1835
    },
    {
      "epoch": 0.636615811373093,
      "grad_norm": 2.5122199058532715,
      "learning_rate": 1e-06,
      "loss": 0.1502,
      "step": 1836
    },
    {
      "epoch": 0.6369625520110958,
      "grad_norm": 1.8004939556121826,
      "learning_rate": 1e-06,
      "loss": 0.1482,
      "step": 1837
    },
    {
      "epoch": 0.6373092926490985,
      "grad_norm": 2.451542854309082,
      "learning_rate": 1e-06,
      "loss": 0.1656,
      "step": 1838
    },
    {
      "epoch": 0.6376560332871013,
      "grad_norm": 3.232908248901367,
      "learning_rate": 1e-06,
      "loss": 0.1625,
      "step": 1839
    },
    {
      "epoch": 0.6380027739251041,
      "grad_norm": 2.172776460647583,
      "learning_rate": 1e-06,
      "loss": 0.1713,
      "step": 1840
    },
    {
      "epoch": 0.6383495145631068,
      "grad_norm": 2.3084630966186523,
      "learning_rate": 1e-06,
      "loss": 0.1474,
      "step": 1841
    },
    {
      "epoch": 0.6386962552011096,
      "grad_norm": 1.8688488006591797,
      "learning_rate": 1e-06,
      "loss": 0.1386,
      "step": 1842
    },
    {
      "epoch": 0.6390429958391124,
      "grad_norm": 2.7231826782226562,
      "learning_rate": 1e-06,
      "loss": 0.1523,
      "step": 1843
    },
    {
      "epoch": 0.6393897364771152,
      "grad_norm": 2.4542200565338135,
      "learning_rate": 1e-06,
      "loss": 0.1505,
      "step": 1844
    },
    {
      "epoch": 0.6397364771151179,
      "grad_norm": 2.8579893112182617,
      "learning_rate": 1e-06,
      "loss": 0.1565,
      "step": 1845
    },
    {
      "epoch": 0.6400832177531207,
      "grad_norm": 3.261667251586914,
      "learning_rate": 1e-06,
      "loss": 0.1216,
      "step": 1846
    },
    {
      "epoch": 0.6404299583911235,
      "grad_norm": 2.358067035675049,
      "learning_rate": 1e-06,
      "loss": 0.17,
      "step": 1847
    },
    {
      "epoch": 0.6407766990291263,
      "grad_norm": 3.6049678325653076,
      "learning_rate": 1e-06,
      "loss": 0.1488,
      "step": 1848
    },
    {
      "epoch": 0.641123439667129,
      "grad_norm": 2.246796131134033,
      "learning_rate": 1e-06,
      "loss": 0.1601,
      "step": 1849
    },
    {
      "epoch": 0.6414701803051318,
      "grad_norm": 1.8105849027633667,
      "learning_rate": 1e-06,
      "loss": 0.1093,
      "step": 1850
    },
    {
      "epoch": 0.6418169209431346,
      "grad_norm": 1.771470546722412,
      "learning_rate": 1e-06,
      "loss": 0.1511,
      "step": 1851
    },
    {
      "epoch": 0.6421636615811374,
      "grad_norm": 2.7330245971679688,
      "learning_rate": 1e-06,
      "loss": 0.176,
      "step": 1852
    },
    {
      "epoch": 0.6425104022191401,
      "grad_norm": 4.204245567321777,
      "learning_rate": 1e-06,
      "loss": 0.1684,
      "step": 1853
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 2.937845468521118,
      "learning_rate": 1e-06,
      "loss": 0.1684,
      "step": 1854
    },
    {
      "epoch": 0.6432038834951457,
      "grad_norm": 3.5551681518554688,
      "learning_rate": 1e-06,
      "loss": 0.1421,
      "step": 1855
    },
    {
      "epoch": 0.6435506241331485,
      "grad_norm": 3.067307472229004,
      "learning_rate": 1e-06,
      "loss": 0.1413,
      "step": 1856
    },
    {
      "epoch": 0.6438973647711512,
      "grad_norm": 2.2362422943115234,
      "learning_rate": 1e-06,
      "loss": 0.1781,
      "step": 1857
    },
    {
      "epoch": 0.644244105409154,
      "grad_norm": 2.397939443588257,
      "learning_rate": 1e-06,
      "loss": 0.1694,
      "step": 1858
    },
    {
      "epoch": 0.6445908460471568,
      "grad_norm": 2.447730779647827,
      "learning_rate": 1e-06,
      "loss": 0.1727,
      "step": 1859
    },
    {
      "epoch": 0.6449375866851595,
      "grad_norm": 2.1447441577911377,
      "learning_rate": 1e-06,
      "loss": 0.1688,
      "step": 1860
    },
    {
      "epoch": 0.6452843273231623,
      "grad_norm": 2.5228350162506104,
      "learning_rate": 1e-06,
      "loss": 0.1489,
      "step": 1861
    },
    {
      "epoch": 0.6456310679611651,
      "grad_norm": 3.115823268890381,
      "learning_rate": 1e-06,
      "loss": 0.1276,
      "step": 1862
    },
    {
      "epoch": 0.6459778085991679,
      "grad_norm": 1.9809269905090332,
      "learning_rate": 1e-06,
      "loss": 0.1688,
      "step": 1863
    },
    {
      "epoch": 0.6463245492371706,
      "grad_norm": 1.765738606452942,
      "learning_rate": 1e-06,
      "loss": 0.1619,
      "step": 1864
    },
    {
      "epoch": 0.6466712898751734,
      "grad_norm": 2.3409016132354736,
      "learning_rate": 1e-06,
      "loss": 0.1492,
      "step": 1865
    },
    {
      "epoch": 0.6470180305131762,
      "grad_norm": 2.63494873046875,
      "learning_rate": 1e-06,
      "loss": 0.155,
      "step": 1866
    },
    {
      "epoch": 0.647364771151179,
      "grad_norm": 2.3334195613861084,
      "learning_rate": 1e-06,
      "loss": 0.1476,
      "step": 1867
    },
    {
      "epoch": 0.6477115117891817,
      "grad_norm": 2.024077892303467,
      "learning_rate": 1e-06,
      "loss": 0.1419,
      "step": 1868
    },
    {
      "epoch": 0.6480582524271845,
      "grad_norm": 2.0084269046783447,
      "learning_rate": 1e-06,
      "loss": 0.152,
      "step": 1869
    },
    {
      "epoch": 0.6484049930651873,
      "grad_norm": 2.7949881553649902,
      "learning_rate": 1e-06,
      "loss": 0.2076,
      "step": 1870
    },
    {
      "epoch": 0.6487517337031901,
      "grad_norm": 4.183343887329102,
      "learning_rate": 1e-06,
      "loss": 0.1928,
      "step": 1871
    },
    {
      "epoch": 0.6490984743411928,
      "grad_norm": 3.4971086978912354,
      "learning_rate": 1e-06,
      "loss": 0.1486,
      "step": 1872
    },
    {
      "epoch": 0.6494452149791956,
      "grad_norm": 2.022676944732666,
      "learning_rate": 1e-06,
      "loss": 0.1332,
      "step": 1873
    },
    {
      "epoch": 0.6497919556171984,
      "grad_norm": 3.9578323364257812,
      "learning_rate": 1e-06,
      "loss": 0.1666,
      "step": 1874
    },
    {
      "epoch": 0.6501386962552012,
      "grad_norm": 2.3028218746185303,
      "learning_rate": 1e-06,
      "loss": 0.1662,
      "step": 1875
    },
    {
      "epoch": 0.6504854368932039,
      "grad_norm": 2.505436897277832,
      "learning_rate": 1e-06,
      "loss": 0.1291,
      "step": 1876
    },
    {
      "epoch": 0.6508321775312067,
      "grad_norm": 3.1190056800842285,
      "learning_rate": 1e-06,
      "loss": 0.1731,
      "step": 1877
    },
    {
      "epoch": 0.6511789181692095,
      "grad_norm": 2.1968798637390137,
      "learning_rate": 1e-06,
      "loss": 0.1481,
      "step": 1878
    },
    {
      "epoch": 0.6515256588072122,
      "grad_norm": 2.4867115020751953,
      "learning_rate": 1e-06,
      "loss": 0.1194,
      "step": 1879
    },
    {
      "epoch": 0.651872399445215,
      "grad_norm": 2.8347103595733643,
      "learning_rate": 1e-06,
      "loss": 0.1717,
      "step": 1880
    },
    {
      "epoch": 0.6522191400832178,
      "grad_norm": 3.8689327239990234,
      "learning_rate": 1e-06,
      "loss": 0.1226,
      "step": 1881
    },
    {
      "epoch": 0.6525658807212206,
      "grad_norm": 2.115302801132202,
      "learning_rate": 1e-06,
      "loss": 0.1491,
      "step": 1882
    },
    {
      "epoch": 0.6529126213592233,
      "grad_norm": 3.8725671768188477,
      "learning_rate": 1e-06,
      "loss": 0.1623,
      "step": 1883
    },
    {
      "epoch": 0.6532593619972261,
      "grad_norm": 2.957167863845825,
      "learning_rate": 1e-06,
      "loss": 0.1231,
      "step": 1884
    },
    {
      "epoch": 0.6536061026352289,
      "grad_norm": 2.131582021713257,
      "learning_rate": 1e-06,
      "loss": 0.1722,
      "step": 1885
    },
    {
      "epoch": 0.6539528432732317,
      "grad_norm": 2.6788721084594727,
      "learning_rate": 1e-06,
      "loss": 0.143,
      "step": 1886
    },
    {
      "epoch": 0.6542995839112344,
      "grad_norm": 2.4989116191864014,
      "learning_rate": 1e-06,
      "loss": 0.1345,
      "step": 1887
    },
    {
      "epoch": 0.6546463245492372,
      "grad_norm": 2.2083873748779297,
      "learning_rate": 1e-06,
      "loss": 0.1272,
      "step": 1888
    },
    {
      "epoch": 0.65499306518724,
      "grad_norm": 3.461536407470703,
      "learning_rate": 1e-06,
      "loss": 0.1506,
      "step": 1889
    },
    {
      "epoch": 0.6553398058252428,
      "grad_norm": 1.8483829498291016,
      "learning_rate": 1e-06,
      "loss": 0.1281,
      "step": 1890
    },
    {
      "epoch": 0.6556865464632455,
      "grad_norm": 5.3467302322387695,
      "learning_rate": 1e-06,
      "loss": 0.163,
      "step": 1891
    },
    {
      "epoch": 0.6560332871012483,
      "grad_norm": 3.6457359790802,
      "learning_rate": 1e-06,
      "loss": 0.1626,
      "step": 1892
    },
    {
      "epoch": 0.6563800277392511,
      "grad_norm": 3.4049503803253174,
      "learning_rate": 1e-06,
      "loss": 0.1452,
      "step": 1893
    },
    {
      "epoch": 0.6567267683772539,
      "grad_norm": 5.646260738372803,
      "learning_rate": 1e-06,
      "loss": 0.1739,
      "step": 1894
    },
    {
      "epoch": 0.6570735090152566,
      "grad_norm": 3.8253393173217773,
      "learning_rate": 1e-06,
      "loss": 0.1399,
      "step": 1895
    },
    {
      "epoch": 0.6574202496532594,
      "grad_norm": 3.993398904800415,
      "learning_rate": 1e-06,
      "loss": 0.1651,
      "step": 1896
    },
    {
      "epoch": 0.6577669902912622,
      "grad_norm": 4.906624794006348,
      "learning_rate": 1e-06,
      "loss": 0.109,
      "step": 1897
    },
    {
      "epoch": 0.658113730929265,
      "grad_norm": 2.4523298740386963,
      "learning_rate": 1e-06,
      "loss": 0.1374,
      "step": 1898
    },
    {
      "epoch": 0.6584604715672677,
      "grad_norm": 4.38106632232666,
      "learning_rate": 1e-06,
      "loss": 0.1601,
      "step": 1899
    },
    {
      "epoch": 0.6588072122052705,
      "grad_norm": 2.361201047897339,
      "learning_rate": 1e-06,
      "loss": 0.1647,
      "step": 1900
    },
    {
      "epoch": 0.6591539528432733,
      "grad_norm": 1.7961761951446533,
      "learning_rate": 1e-06,
      "loss": 0.1503,
      "step": 1901
    },
    {
      "epoch": 0.659500693481276,
      "grad_norm": 4.200663089752197,
      "learning_rate": 1e-06,
      "loss": 0.1362,
      "step": 1902
    },
    {
      "epoch": 0.6598474341192788,
      "grad_norm": 2.141098976135254,
      "learning_rate": 1e-06,
      "loss": 0.1287,
      "step": 1903
    },
    {
      "epoch": 0.6601941747572816,
      "grad_norm": 3.659383773803711,
      "learning_rate": 1e-06,
      "loss": 0.1679,
      "step": 1904
    },
    {
      "epoch": 0.6605409153952844,
      "grad_norm": 2.159740447998047,
      "learning_rate": 1e-06,
      "loss": 0.151,
      "step": 1905
    },
    {
      "epoch": 0.6608876560332871,
      "grad_norm": 2.612746238708496,
      "learning_rate": 1e-06,
      "loss": 0.1331,
      "step": 1906
    },
    {
      "epoch": 0.6612343966712899,
      "grad_norm": 2.067161798477173,
      "learning_rate": 1e-06,
      "loss": 0.1375,
      "step": 1907
    },
    {
      "epoch": 0.6615811373092927,
      "grad_norm": 2.2887136936187744,
      "learning_rate": 1e-06,
      "loss": 0.1445,
      "step": 1908
    },
    {
      "epoch": 0.6619278779472955,
      "grad_norm": 3.688062906265259,
      "learning_rate": 1e-06,
      "loss": 0.1943,
      "step": 1909
    },
    {
      "epoch": 0.6622746185852982,
      "grad_norm": 2.4100046157836914,
      "learning_rate": 1e-06,
      "loss": 0.1798,
      "step": 1910
    },
    {
      "epoch": 0.662621359223301,
      "grad_norm": 1.7269797325134277,
      "learning_rate": 1e-06,
      "loss": 0.1404,
      "step": 1911
    },
    {
      "epoch": 0.6629680998613038,
      "grad_norm": 1.8225007057189941,
      "learning_rate": 1e-06,
      "loss": 0.1252,
      "step": 1912
    },
    {
      "epoch": 0.6633148404993066,
      "grad_norm": 2.511077404022217,
      "learning_rate": 1e-06,
      "loss": 0.1458,
      "step": 1913
    },
    {
      "epoch": 0.6636615811373093,
      "grad_norm": 4.242014408111572,
      "learning_rate": 1e-06,
      "loss": 0.1833,
      "step": 1914
    },
    {
      "epoch": 0.6640083217753121,
      "grad_norm": 2.4541306495666504,
      "learning_rate": 1e-06,
      "loss": 0.1639,
      "step": 1915
    },
    {
      "epoch": 0.6643550624133149,
      "grad_norm": 4.093621730804443,
      "learning_rate": 1e-06,
      "loss": 0.1533,
      "step": 1916
    },
    {
      "epoch": 0.6647018030513177,
      "grad_norm": 1.6138478517532349,
      "learning_rate": 1e-06,
      "loss": 0.1365,
      "step": 1917
    },
    {
      "epoch": 0.6650485436893204,
      "grad_norm": 2.1072335243225098,
      "learning_rate": 1e-06,
      "loss": 0.1502,
      "step": 1918
    },
    {
      "epoch": 0.6653952843273232,
      "grad_norm": 4.632135391235352,
      "learning_rate": 1e-06,
      "loss": 0.1454,
      "step": 1919
    },
    {
      "epoch": 0.665742024965326,
      "grad_norm": 2.38916015625,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 1920
    },
    {
      "epoch": 0.6660887656033287,
      "grad_norm": 1.7111142873764038,
      "learning_rate": 1e-06,
      "loss": 0.1491,
      "step": 1921
    },
    {
      "epoch": 0.6664355062413315,
      "grad_norm": 1.6361645460128784,
      "learning_rate": 1e-06,
      "loss": 0.1181,
      "step": 1922
    },
    {
      "epoch": 0.6667822468793343,
      "grad_norm": 2.845128297805786,
      "learning_rate": 1e-06,
      "loss": 0.1508,
      "step": 1923
    },
    {
      "epoch": 0.6671289875173371,
      "grad_norm": 3.6008825302124023,
      "learning_rate": 1e-06,
      "loss": 0.1773,
      "step": 1924
    },
    {
      "epoch": 0.6674757281553398,
      "grad_norm": 2.7898175716400146,
      "learning_rate": 1e-06,
      "loss": 0.1485,
      "step": 1925
    },
    {
      "epoch": 0.6678224687933426,
      "grad_norm": 2.1968812942504883,
      "learning_rate": 1e-06,
      "loss": 0.1538,
      "step": 1926
    },
    {
      "epoch": 0.6681692094313454,
      "grad_norm": 3.2181034088134766,
      "learning_rate": 1e-06,
      "loss": 0.1637,
      "step": 1927
    },
    {
      "epoch": 0.6685159500693482,
      "grad_norm": 3.3751060962677,
      "learning_rate": 1e-06,
      "loss": 0.1967,
      "step": 1928
    },
    {
      "epoch": 0.6688626907073509,
      "grad_norm": 4.066105842590332,
      "learning_rate": 1e-06,
      "loss": 0.1598,
      "step": 1929
    },
    {
      "epoch": 0.6692094313453537,
      "grad_norm": 3.0509390830993652,
      "learning_rate": 1e-06,
      "loss": 0.2023,
      "step": 1930
    },
    {
      "epoch": 0.6695561719833565,
      "grad_norm": 2.0706655979156494,
      "learning_rate": 1e-06,
      "loss": 0.1549,
      "step": 1931
    },
    {
      "epoch": 0.6699029126213593,
      "grad_norm": 2.6645240783691406,
      "learning_rate": 1e-06,
      "loss": 0.1451,
      "step": 1932
    },
    {
      "epoch": 0.670249653259362,
      "grad_norm": 2.8021857738494873,
      "learning_rate": 1e-06,
      "loss": 0.1238,
      "step": 1933
    },
    {
      "epoch": 0.6705963938973648,
      "grad_norm": 2.668379068374634,
      "learning_rate": 1e-06,
      "loss": 0.1606,
      "step": 1934
    },
    {
      "epoch": 0.6709431345353676,
      "grad_norm": 2.5608723163604736,
      "learning_rate": 1e-06,
      "loss": 0.1425,
      "step": 1935
    },
    {
      "epoch": 0.6712898751733704,
      "grad_norm": 2.992034435272217,
      "learning_rate": 1e-06,
      "loss": 0.1378,
      "step": 1936
    },
    {
      "epoch": 0.6716366158113731,
      "grad_norm": 1.993120789527893,
      "learning_rate": 1e-06,
      "loss": 0.1504,
      "step": 1937
    },
    {
      "epoch": 0.6719833564493759,
      "grad_norm": 1.990553379058838,
      "learning_rate": 1e-06,
      "loss": 0.1648,
      "step": 1938
    },
    {
      "epoch": 0.6723300970873787,
      "grad_norm": 3.997328758239746,
      "learning_rate": 1e-06,
      "loss": 0.1319,
      "step": 1939
    },
    {
      "epoch": 0.6726768377253814,
      "grad_norm": 3.1100356578826904,
      "learning_rate": 1e-06,
      "loss": 0.1824,
      "step": 1940
    },
    {
      "epoch": 0.6730235783633842,
      "grad_norm": 1.7388405799865723,
      "learning_rate": 1e-06,
      "loss": 0.1388,
      "step": 1941
    },
    {
      "epoch": 0.673370319001387,
      "grad_norm": 2.0104892253875732,
      "learning_rate": 1e-06,
      "loss": 0.1537,
      "step": 1942
    },
    {
      "epoch": 0.6737170596393898,
      "grad_norm": 4.981960296630859,
      "learning_rate": 1e-06,
      "loss": 0.1823,
      "step": 1943
    },
    {
      "epoch": 0.6740638002773925,
      "grad_norm": 1.5539988279342651,
      "learning_rate": 1e-06,
      "loss": 0.1507,
      "step": 1944
    },
    {
      "epoch": 0.6744105409153953,
      "grad_norm": 3.975370168685913,
      "learning_rate": 1e-06,
      "loss": 0.1869,
      "step": 1945
    },
    {
      "epoch": 0.6747572815533981,
      "grad_norm": 2.3512701988220215,
      "learning_rate": 1e-06,
      "loss": 0.1363,
      "step": 1946
    },
    {
      "epoch": 0.6751040221914009,
      "grad_norm": 1.7181259393692017,
      "learning_rate": 1e-06,
      "loss": 0.1472,
      "step": 1947
    },
    {
      "epoch": 0.6754507628294036,
      "grad_norm": 1.7866528034210205,
      "learning_rate": 1e-06,
      "loss": 0.1358,
      "step": 1948
    },
    {
      "epoch": 0.6757975034674064,
      "grad_norm": 2.635709524154663,
      "learning_rate": 1e-06,
      "loss": 0.1659,
      "step": 1949
    },
    {
      "epoch": 0.6761442441054092,
      "grad_norm": 1.9648034572601318,
      "learning_rate": 1e-06,
      "loss": 0.1443,
      "step": 1950
    },
    {
      "epoch": 0.676490984743412,
      "grad_norm": 2.1203794479370117,
      "learning_rate": 1e-06,
      "loss": 0.1283,
      "step": 1951
    },
    {
      "epoch": 0.6768377253814147,
      "grad_norm": 2.54052996635437,
      "learning_rate": 1e-06,
      "loss": 0.1447,
      "step": 1952
    },
    {
      "epoch": 0.6771844660194175,
      "grad_norm": 2.9203107357025146,
      "learning_rate": 1e-06,
      "loss": 0.1362,
      "step": 1953
    },
    {
      "epoch": 0.6775312066574203,
      "grad_norm": 2.1829326152801514,
      "learning_rate": 1e-06,
      "loss": 0.1267,
      "step": 1954
    },
    {
      "epoch": 0.6778779472954231,
      "grad_norm": 2.608271837234497,
      "learning_rate": 1e-06,
      "loss": 0.1673,
      "step": 1955
    },
    {
      "epoch": 0.6782246879334258,
      "grad_norm": 2.003397226333618,
      "learning_rate": 1e-06,
      "loss": 0.1687,
      "step": 1956
    },
    {
      "epoch": 0.6785714285714286,
      "grad_norm": 2.5364773273468018,
      "learning_rate": 1e-06,
      "loss": 0.1331,
      "step": 1957
    },
    {
      "epoch": 0.6789181692094314,
      "grad_norm": 2.5657753944396973,
      "learning_rate": 1e-06,
      "loss": 0.1485,
      "step": 1958
    },
    {
      "epoch": 0.6792649098474342,
      "grad_norm": 2.2307422161102295,
      "learning_rate": 1e-06,
      "loss": 0.16,
      "step": 1959
    },
    {
      "epoch": 0.6796116504854369,
      "grad_norm": 2.0911974906921387,
      "learning_rate": 1e-06,
      "loss": 0.1338,
      "step": 1960
    },
    {
      "epoch": 0.6799583911234397,
      "grad_norm": 3.5447793006896973,
      "learning_rate": 1e-06,
      "loss": 0.1949,
      "step": 1961
    },
    {
      "epoch": 0.6803051317614425,
      "grad_norm": 2.7882168292999268,
      "learning_rate": 1e-06,
      "loss": 0.1912,
      "step": 1962
    },
    {
      "epoch": 0.6806518723994452,
      "grad_norm": 1.985961675643921,
      "learning_rate": 1e-06,
      "loss": 0.1228,
      "step": 1963
    },
    {
      "epoch": 0.680998613037448,
      "grad_norm": 4.209602355957031,
      "learning_rate": 1e-06,
      "loss": 0.1476,
      "step": 1964
    },
    {
      "epoch": 0.6813453536754508,
      "grad_norm": 2.3590171337127686,
      "learning_rate": 1e-06,
      "loss": 0.1246,
      "step": 1965
    },
    {
      "epoch": 0.6816920943134536,
      "grad_norm": 2.680187225341797,
      "learning_rate": 1e-06,
      "loss": 0.1812,
      "step": 1966
    },
    {
      "epoch": 0.6820388349514563,
      "grad_norm": 2.99859881401062,
      "learning_rate": 1e-06,
      "loss": 0.1483,
      "step": 1967
    },
    {
      "epoch": 0.6823855755894591,
      "grad_norm": 2.140812635421753,
      "learning_rate": 1e-06,
      "loss": 0.1315,
      "step": 1968
    },
    {
      "epoch": 0.6827323162274619,
      "grad_norm": 2.1156227588653564,
      "learning_rate": 1e-06,
      "loss": 0.1576,
      "step": 1969
    },
    {
      "epoch": 0.6830790568654647,
      "grad_norm": 2.218289613723755,
      "learning_rate": 1e-06,
      "loss": 0.1378,
      "step": 1970
    },
    {
      "epoch": 0.6834257975034674,
      "grad_norm": 2.0559825897216797,
      "learning_rate": 1e-06,
      "loss": 0.158,
      "step": 1971
    },
    {
      "epoch": 0.6837725381414702,
      "grad_norm": 2.555392026901245,
      "learning_rate": 1e-06,
      "loss": 0.1377,
      "step": 1972
    },
    {
      "epoch": 0.684119278779473,
      "grad_norm": 2.068892240524292,
      "learning_rate": 1e-06,
      "loss": 0.134,
      "step": 1973
    },
    {
      "epoch": 0.6844660194174758,
      "grad_norm": 1.8548657894134521,
      "learning_rate": 1e-06,
      "loss": 0.1438,
      "step": 1974
    },
    {
      "epoch": 0.6848127600554785,
      "grad_norm": 5.5719499588012695,
      "learning_rate": 1e-06,
      "loss": 0.1765,
      "step": 1975
    },
    {
      "epoch": 0.6851595006934813,
      "grad_norm": 3.041820526123047,
      "learning_rate": 1e-06,
      "loss": 0.129,
      "step": 1976
    },
    {
      "epoch": 0.6855062413314841,
      "grad_norm": 7.963346004486084,
      "learning_rate": 1e-06,
      "loss": 0.2188,
      "step": 1977
    },
    {
      "epoch": 0.6858529819694869,
      "grad_norm": 2.143159866333008,
      "learning_rate": 1e-06,
      "loss": 0.1669,
      "step": 1978
    },
    {
      "epoch": 0.6861997226074896,
      "grad_norm": 2.7842423915863037,
      "learning_rate": 1e-06,
      "loss": 0.1568,
      "step": 1979
    },
    {
      "epoch": 0.6865464632454924,
      "grad_norm": 2.4829189777374268,
      "learning_rate": 1e-06,
      "loss": 0.1609,
      "step": 1980
    },
    {
      "epoch": 0.6868932038834952,
      "grad_norm": 2.9691684246063232,
      "learning_rate": 1e-06,
      "loss": 0.1566,
      "step": 1981
    },
    {
      "epoch": 0.687239944521498,
      "grad_norm": 1.885637640953064,
      "learning_rate": 1e-06,
      "loss": 0.1251,
      "step": 1982
    },
    {
      "epoch": 0.6875866851595007,
      "grad_norm": 3.081054210662842,
      "learning_rate": 1e-06,
      "loss": 0.1216,
      "step": 1983
    },
    {
      "epoch": 0.6879334257975035,
      "grad_norm": 2.235093832015991,
      "learning_rate": 1e-06,
      "loss": 0.1424,
      "step": 1984
    },
    {
      "epoch": 0.6882801664355063,
      "grad_norm": 2.051729679107666,
      "learning_rate": 1e-06,
      "loss": 0.1411,
      "step": 1985
    },
    {
      "epoch": 0.688626907073509,
      "grad_norm": 3.0905375480651855,
      "learning_rate": 1e-06,
      "loss": 0.1706,
      "step": 1986
    },
    {
      "epoch": 0.6889736477115118,
      "grad_norm": 2.348816156387329,
      "learning_rate": 1e-06,
      "loss": 0.1609,
      "step": 1987
    },
    {
      "epoch": 0.6893203883495146,
      "grad_norm": 3.2591960430145264,
      "learning_rate": 1e-06,
      "loss": 0.189,
      "step": 1988
    },
    {
      "epoch": 0.6896671289875174,
      "grad_norm": 5.137045860290527,
      "learning_rate": 1e-06,
      "loss": 0.1329,
      "step": 1989
    },
    {
      "epoch": 0.6900138696255201,
      "grad_norm": 3.611876964569092,
      "learning_rate": 1e-06,
      "loss": 0.13,
      "step": 1990
    },
    {
      "epoch": 0.6903606102635229,
      "grad_norm": 3.3672244548797607,
      "learning_rate": 1e-06,
      "loss": 0.1583,
      "step": 1991
    },
    {
      "epoch": 0.6907073509015257,
      "grad_norm": 2.499140501022339,
      "learning_rate": 1e-06,
      "loss": 0.12,
      "step": 1992
    },
    {
      "epoch": 0.6910540915395285,
      "grad_norm": 2.993394136428833,
      "learning_rate": 1e-06,
      "loss": 0.1329,
      "step": 1993
    },
    {
      "epoch": 0.6914008321775312,
      "grad_norm": 3.3767669200897217,
      "learning_rate": 1e-06,
      "loss": 0.1607,
      "step": 1994
    },
    {
      "epoch": 0.691747572815534,
      "grad_norm": 4.418931484222412,
      "learning_rate": 1e-06,
      "loss": 0.17,
      "step": 1995
    },
    {
      "epoch": 0.6920943134535368,
      "grad_norm": 2.544867753982544,
      "learning_rate": 1e-06,
      "loss": 0.1642,
      "step": 1996
    },
    {
      "epoch": 0.6924410540915396,
      "grad_norm": 4.367650508880615,
      "learning_rate": 1e-06,
      "loss": 0.161,
      "step": 1997
    },
    {
      "epoch": 0.6927877947295423,
      "grad_norm": 2.103868007659912,
      "learning_rate": 1e-06,
      "loss": 0.1351,
      "step": 1998
    },
    {
      "epoch": 0.6931345353675451,
      "grad_norm": 2.2191128730773926,
      "learning_rate": 1e-06,
      "loss": 0.1622,
      "step": 1999
    },
    {
      "epoch": 0.6934812760055479,
      "grad_norm": 2.239701747894287,
      "learning_rate": 1e-06,
      "loss": 0.1547,
      "step": 2000
    },
    {
      "epoch": 0.6938280166435506,
      "grad_norm": 2.166487216949463,
      "learning_rate": 1e-06,
      "loss": 0.1431,
      "step": 2001
    },
    {
      "epoch": 0.6941747572815534,
      "grad_norm": 2.3545658588409424,
      "learning_rate": 1e-06,
      "loss": 0.1271,
      "step": 2002
    },
    {
      "epoch": 0.6945214979195562,
      "grad_norm": 2.040910005569458,
      "learning_rate": 1e-06,
      "loss": 0.1502,
      "step": 2003
    },
    {
      "epoch": 0.694868238557559,
      "grad_norm": 2.179579973220825,
      "learning_rate": 1e-06,
      "loss": 0.1385,
      "step": 2004
    },
    {
      "epoch": 0.6952149791955617,
      "grad_norm": 2.0149500370025635,
      "learning_rate": 1e-06,
      "loss": 0.1396,
      "step": 2005
    },
    {
      "epoch": 0.6955617198335645,
      "grad_norm": 3.165092706680298,
      "learning_rate": 1e-06,
      "loss": 0.1217,
      "step": 2006
    },
    {
      "epoch": 0.6959084604715673,
      "grad_norm": 2.45447039604187,
      "learning_rate": 1e-06,
      "loss": 0.1455,
      "step": 2007
    },
    {
      "epoch": 0.6962552011095701,
      "grad_norm": 4.395773410797119,
      "learning_rate": 1e-06,
      "loss": 0.1568,
      "step": 2008
    },
    {
      "epoch": 0.6966019417475728,
      "grad_norm": 4.389420986175537,
      "learning_rate": 1e-06,
      "loss": 0.1294,
      "step": 2009
    },
    {
      "epoch": 0.6969486823855756,
      "grad_norm": 1.8730595111846924,
      "learning_rate": 1e-06,
      "loss": 0.1371,
      "step": 2010
    },
    {
      "epoch": 0.6972954230235784,
      "grad_norm": 2.394237518310547,
      "learning_rate": 1e-06,
      "loss": 0.1588,
      "step": 2011
    },
    {
      "epoch": 0.6976421636615812,
      "grad_norm": 2.2952566146850586,
      "learning_rate": 1e-06,
      "loss": 0.1362,
      "step": 2012
    },
    {
      "epoch": 0.6979889042995839,
      "grad_norm": 3.177366018295288,
      "learning_rate": 1e-06,
      "loss": 0.1169,
      "step": 2013
    },
    {
      "epoch": 0.6983356449375867,
      "grad_norm": 2.6569299697875977,
      "learning_rate": 1e-06,
      "loss": 0.1503,
      "step": 2014
    },
    {
      "epoch": 0.6986823855755895,
      "grad_norm": 2.000425338745117,
      "learning_rate": 1e-06,
      "loss": 0.1399,
      "step": 2015
    },
    {
      "epoch": 0.6990291262135923,
      "grad_norm": 3.407623529434204,
      "learning_rate": 1e-06,
      "loss": 0.1704,
      "step": 2016
    },
    {
      "epoch": 0.699375866851595,
      "grad_norm": 2.73795747756958,
      "learning_rate": 1e-06,
      "loss": 0.1476,
      "step": 2017
    },
    {
      "epoch": 0.6997226074895978,
      "grad_norm": 3.8470706939697266,
      "learning_rate": 1e-06,
      "loss": 0.1694,
      "step": 2018
    },
    {
      "epoch": 0.7000693481276006,
      "grad_norm": 2.0070230960845947,
      "learning_rate": 1e-06,
      "loss": 0.1629,
      "step": 2019
    },
    {
      "epoch": 0.7004160887656034,
      "grad_norm": 2.150434970855713,
      "learning_rate": 1e-06,
      "loss": 0.1228,
      "step": 2020
    },
    {
      "epoch": 0.7007628294036061,
      "grad_norm": 1.901470422744751,
      "learning_rate": 1e-06,
      "loss": 0.114,
      "step": 2021
    },
    {
      "epoch": 0.7011095700416089,
      "grad_norm": 2.0084474086761475,
      "learning_rate": 1e-06,
      "loss": 0.1431,
      "step": 2022
    },
    {
      "epoch": 0.7014563106796117,
      "grad_norm": 2.262711763381958,
      "learning_rate": 1e-06,
      "loss": 0.1298,
      "step": 2023
    },
    {
      "epoch": 0.7018030513176144,
      "grad_norm": 2.019993782043457,
      "learning_rate": 1e-06,
      "loss": 0.1568,
      "step": 2024
    },
    {
      "epoch": 0.7021497919556172,
      "grad_norm": 3.4103362560272217,
      "learning_rate": 1e-06,
      "loss": 0.1283,
      "step": 2025
    },
    {
      "epoch": 0.70249653259362,
      "grad_norm": 2.4157567024230957,
      "learning_rate": 1e-06,
      "loss": 0.1739,
      "step": 2026
    },
    {
      "epoch": 0.7028432732316228,
      "grad_norm": 2.405488967895508,
      "learning_rate": 1e-06,
      "loss": 0.1439,
      "step": 2027
    },
    {
      "epoch": 0.7031900138696255,
      "grad_norm": 1.8161323070526123,
      "learning_rate": 1e-06,
      "loss": 0.1458,
      "step": 2028
    },
    {
      "epoch": 0.7035367545076283,
      "grad_norm": 4.244541168212891,
      "learning_rate": 1e-06,
      "loss": 0.1481,
      "step": 2029
    },
    {
      "epoch": 0.7038834951456311,
      "grad_norm": 2.1027145385742188,
      "learning_rate": 1e-06,
      "loss": 0.1376,
      "step": 2030
    },
    {
      "epoch": 0.7042302357836339,
      "grad_norm": 1.911086916923523,
      "learning_rate": 1e-06,
      "loss": 0.1393,
      "step": 2031
    },
    {
      "epoch": 0.7045769764216366,
      "grad_norm": 3.033745527267456,
      "learning_rate": 1e-06,
      "loss": 0.1313,
      "step": 2032
    },
    {
      "epoch": 0.7049237170596394,
      "grad_norm": 2.4053955078125,
      "learning_rate": 1e-06,
      "loss": 0.1483,
      "step": 2033
    },
    {
      "epoch": 0.7052704576976422,
      "grad_norm": 1.796697735786438,
      "learning_rate": 1e-06,
      "loss": 0.1207,
      "step": 2034
    },
    {
      "epoch": 0.705617198335645,
      "grad_norm": 2.3117001056671143,
      "learning_rate": 1e-06,
      "loss": 0.1391,
      "step": 2035
    },
    {
      "epoch": 0.7059639389736477,
      "grad_norm": 3.2871193885803223,
      "learning_rate": 1e-06,
      "loss": 0.169,
      "step": 2036
    },
    {
      "epoch": 0.7063106796116505,
      "grad_norm": 2.3646295070648193,
      "learning_rate": 1e-06,
      "loss": 0.1827,
      "step": 2037
    },
    {
      "epoch": 0.7066574202496533,
      "grad_norm": 1.7376772165298462,
      "learning_rate": 1e-06,
      "loss": 0.1375,
      "step": 2038
    },
    {
      "epoch": 0.707004160887656,
      "grad_norm": 2.200528383255005,
      "learning_rate": 1e-06,
      "loss": 0.1417,
      "step": 2039
    },
    {
      "epoch": 0.7073509015256588,
      "grad_norm": 2.742612838745117,
      "learning_rate": 1e-06,
      "loss": 0.156,
      "step": 2040
    },
    {
      "epoch": 0.7076976421636616,
      "grad_norm": 2.012239694595337,
      "learning_rate": 1e-06,
      "loss": 0.1768,
      "step": 2041
    },
    {
      "epoch": 0.7080443828016644,
      "grad_norm": 1.6217942237854004,
      "learning_rate": 1e-06,
      "loss": 0.1093,
      "step": 2042
    },
    {
      "epoch": 0.7083911234396671,
      "grad_norm": 3.5539634227752686,
      "learning_rate": 1e-06,
      "loss": 0.1688,
      "step": 2043
    },
    {
      "epoch": 0.7087378640776699,
      "grad_norm": 2.4913277626037598,
      "learning_rate": 1e-06,
      "loss": 0.1956,
      "step": 2044
    },
    {
      "epoch": 0.7090846047156727,
      "grad_norm": 1.9440995454788208,
      "learning_rate": 1e-06,
      "loss": 0.1648,
      "step": 2045
    },
    {
      "epoch": 0.7094313453536755,
      "grad_norm": 2.7142701148986816,
      "learning_rate": 1e-06,
      "loss": 0.1566,
      "step": 2046
    },
    {
      "epoch": 0.7097780859916782,
      "grad_norm": 2.7249319553375244,
      "learning_rate": 1e-06,
      "loss": 0.1664,
      "step": 2047
    },
    {
      "epoch": 0.710124826629681,
      "grad_norm": 2.3271358013153076,
      "learning_rate": 1e-06,
      "loss": 0.1705,
      "step": 2048
    },
    {
      "epoch": 0.7104715672676838,
      "grad_norm": 1.7891216278076172,
      "learning_rate": 1e-06,
      "loss": 0.1293,
      "step": 2049
    },
    {
      "epoch": 0.7108183079056866,
      "grad_norm": 2.274380922317505,
      "learning_rate": 1e-06,
      "loss": 0.1559,
      "step": 2050
    },
    {
      "epoch": 0.7111650485436893,
      "grad_norm": 2.742744207382202,
      "learning_rate": 1e-06,
      "loss": 0.1763,
      "step": 2051
    },
    {
      "epoch": 0.7115117891816921,
      "grad_norm": 1.602330207824707,
      "learning_rate": 1e-06,
      "loss": 0.1516,
      "step": 2052
    },
    {
      "epoch": 0.7118585298196949,
      "grad_norm": 1.677738904953003,
      "learning_rate": 1e-06,
      "loss": 0.1205,
      "step": 2053
    },
    {
      "epoch": 0.7122052704576977,
      "grad_norm": 2.515608549118042,
      "learning_rate": 1e-06,
      "loss": 0.1495,
      "step": 2054
    },
    {
      "epoch": 0.7125520110957004,
      "grad_norm": 2.719072103500366,
      "learning_rate": 1e-06,
      "loss": 0.142,
      "step": 2055
    },
    {
      "epoch": 0.7128987517337032,
      "grad_norm": 2.3802108764648438,
      "learning_rate": 1e-06,
      "loss": 0.1374,
      "step": 2056
    },
    {
      "epoch": 0.713245492371706,
      "grad_norm": 3.3662099838256836,
      "learning_rate": 1e-06,
      "loss": 0.1593,
      "step": 2057
    },
    {
      "epoch": 0.7135922330097088,
      "grad_norm": 2.0012762546539307,
      "learning_rate": 1e-06,
      "loss": 0.1675,
      "step": 2058
    },
    {
      "epoch": 0.7139389736477115,
      "grad_norm": 1.8083027601242065,
      "learning_rate": 1e-06,
      "loss": 0.1215,
      "step": 2059
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 3.830106735229492,
      "learning_rate": 1e-06,
      "loss": 0.1446,
      "step": 2060
    },
    {
      "epoch": 0.7146324549237171,
      "grad_norm": 2.1997570991516113,
      "learning_rate": 1e-06,
      "loss": 0.1455,
      "step": 2061
    },
    {
      "epoch": 0.7149791955617198,
      "grad_norm": 1.8245985507965088,
      "learning_rate": 1e-06,
      "loss": 0.1576,
      "step": 2062
    },
    {
      "epoch": 0.7153259361997226,
      "grad_norm": 2.8333470821380615,
      "learning_rate": 1e-06,
      "loss": 0.1414,
      "step": 2063
    },
    {
      "epoch": 0.7156726768377254,
      "grad_norm": 1.8098995685577393,
      "learning_rate": 1e-06,
      "loss": 0.1155,
      "step": 2064
    },
    {
      "epoch": 0.7160194174757282,
      "grad_norm": 2.2318859100341797,
      "learning_rate": 1e-06,
      "loss": 0.145,
      "step": 2065
    },
    {
      "epoch": 0.7163661581137309,
      "grad_norm": 3.0203652381896973,
      "learning_rate": 1e-06,
      "loss": 0.1889,
      "step": 2066
    },
    {
      "epoch": 0.7167128987517337,
      "grad_norm": 3.068648099899292,
      "learning_rate": 1e-06,
      "loss": 0.1269,
      "step": 2067
    },
    {
      "epoch": 0.7170596393897365,
      "grad_norm": 2.7916111946105957,
      "learning_rate": 1e-06,
      "loss": 0.156,
      "step": 2068
    },
    {
      "epoch": 0.7174063800277393,
      "grad_norm": 1.9404290914535522,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 2069
    },
    {
      "epoch": 0.717753120665742,
      "grad_norm": 2.287555694580078,
      "learning_rate": 1e-06,
      "loss": 0.121,
      "step": 2070
    },
    {
      "epoch": 0.7180998613037448,
      "grad_norm": 2.1457648277282715,
      "learning_rate": 1e-06,
      "loss": 0.1232,
      "step": 2071
    },
    {
      "epoch": 0.7184466019417476,
      "grad_norm": 4.4433441162109375,
      "learning_rate": 1e-06,
      "loss": 0.1315,
      "step": 2072
    },
    {
      "epoch": 0.7187933425797504,
      "grad_norm": 3.2125916481018066,
      "learning_rate": 1e-06,
      "loss": 0.1623,
      "step": 2073
    },
    {
      "epoch": 0.7191400832177531,
      "grad_norm": 2.4481723308563232,
      "learning_rate": 1e-06,
      "loss": 0.1458,
      "step": 2074
    },
    {
      "epoch": 0.7194868238557559,
      "grad_norm": 2.894010543823242,
      "learning_rate": 1e-06,
      "loss": 0.1623,
      "step": 2075
    },
    {
      "epoch": 0.7198335644937587,
      "grad_norm": 2.1684398651123047,
      "learning_rate": 1e-06,
      "loss": 0.1323,
      "step": 2076
    },
    {
      "epoch": 0.7201803051317615,
      "grad_norm": 2.481157064437866,
      "learning_rate": 1e-06,
      "loss": 0.1576,
      "step": 2077
    },
    {
      "epoch": 0.7205270457697642,
      "grad_norm": 2.519806385040283,
      "learning_rate": 1e-06,
      "loss": 0.1507,
      "step": 2078
    },
    {
      "epoch": 0.720873786407767,
      "grad_norm": 2.813413381576538,
      "learning_rate": 1e-06,
      "loss": 0.143,
      "step": 2079
    },
    {
      "epoch": 0.7212205270457698,
      "grad_norm": 2.9213216304779053,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 2080
    },
    {
      "epoch": 0.7215672676837726,
      "grad_norm": 2.8391432762145996,
      "learning_rate": 1e-06,
      "loss": 0.1652,
      "step": 2081
    },
    {
      "epoch": 0.7219140083217753,
      "grad_norm": 2.9582345485687256,
      "learning_rate": 1e-06,
      "loss": 0.1511,
      "step": 2082
    },
    {
      "epoch": 0.7222607489597781,
      "grad_norm": 2.236300230026245,
      "learning_rate": 1e-06,
      "loss": 0.1272,
      "step": 2083
    },
    {
      "epoch": 0.7226074895977809,
      "grad_norm": 2.000270366668701,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 2084
    },
    {
      "epoch": 0.7229542302357836,
      "grad_norm": 2.317187547683716,
      "learning_rate": 1e-06,
      "loss": 0.16,
      "step": 2085
    },
    {
      "epoch": 0.7233009708737864,
      "grad_norm": 2.866093397140503,
      "learning_rate": 1e-06,
      "loss": 0.1539,
      "step": 2086
    },
    {
      "epoch": 0.7236477115117892,
      "grad_norm": 3.5143120288848877,
      "learning_rate": 1e-06,
      "loss": 0.1434,
      "step": 2087
    },
    {
      "epoch": 0.723994452149792,
      "grad_norm": 3.5509307384490967,
      "learning_rate": 1e-06,
      "loss": 0.1357,
      "step": 2088
    },
    {
      "epoch": 0.7243411927877947,
      "grad_norm": 1.9065073728561401,
      "learning_rate": 1e-06,
      "loss": 0.1227,
      "step": 2089
    },
    {
      "epoch": 0.7246879334257975,
      "grad_norm": 2.9716525077819824,
      "learning_rate": 1e-06,
      "loss": 0.1538,
      "step": 2090
    },
    {
      "epoch": 0.7250346740638003,
      "grad_norm": 2.858944892883301,
      "learning_rate": 1e-06,
      "loss": 0.1333,
      "step": 2091
    },
    {
      "epoch": 0.7253814147018031,
      "grad_norm": 4.433228969573975,
      "learning_rate": 1e-06,
      "loss": 0.1487,
      "step": 2092
    },
    {
      "epoch": 0.7257281553398058,
      "grad_norm": 2.223188877105713,
      "learning_rate": 1e-06,
      "loss": 0.152,
      "step": 2093
    },
    {
      "epoch": 0.7260748959778086,
      "grad_norm": 2.040834665298462,
      "learning_rate": 1e-06,
      "loss": 0.1542,
      "step": 2094
    },
    {
      "epoch": 0.7264216366158114,
      "grad_norm": 2.6200249195098877,
      "learning_rate": 1e-06,
      "loss": 0.1551,
      "step": 2095
    },
    {
      "epoch": 0.7267683772538142,
      "grad_norm": 2.9047865867614746,
      "learning_rate": 1e-06,
      "loss": 0.1368,
      "step": 2096
    },
    {
      "epoch": 0.7271151178918169,
      "grad_norm": 2.617482900619507,
      "learning_rate": 1e-06,
      "loss": 0.1762,
      "step": 2097
    },
    {
      "epoch": 0.7274618585298197,
      "grad_norm": 2.3887550830841064,
      "learning_rate": 1e-06,
      "loss": 0.1652,
      "step": 2098
    },
    {
      "epoch": 0.7278085991678225,
      "grad_norm": 2.3276662826538086,
      "learning_rate": 1e-06,
      "loss": 0.1696,
      "step": 2099
    },
    {
      "epoch": 0.7281553398058253,
      "grad_norm": 2.4688990116119385,
      "learning_rate": 1e-06,
      "loss": 0.1865,
      "step": 2100
    },
    {
      "epoch": 0.728502080443828,
      "grad_norm": 3.6115245819091797,
      "learning_rate": 1e-06,
      "loss": 0.1667,
      "step": 2101
    },
    {
      "epoch": 0.7288488210818308,
      "grad_norm": 4.399360179901123,
      "learning_rate": 1e-06,
      "loss": 0.1383,
      "step": 2102
    },
    {
      "epoch": 0.7291955617198336,
      "grad_norm": 5.1829729080200195,
      "learning_rate": 1e-06,
      "loss": 0.1576,
      "step": 2103
    },
    {
      "epoch": 0.7295423023578363,
      "grad_norm": 4.441248416900635,
      "learning_rate": 1e-06,
      "loss": 0.139,
      "step": 2104
    },
    {
      "epoch": 0.7298890429958391,
      "grad_norm": 3.7266438007354736,
      "learning_rate": 1e-06,
      "loss": 0.1497,
      "step": 2105
    },
    {
      "epoch": 0.7302357836338419,
      "grad_norm": 2.000267505645752,
      "learning_rate": 1e-06,
      "loss": 0.1643,
      "step": 2106
    },
    {
      "epoch": 0.7305825242718447,
      "grad_norm": 2.909482955932617,
      "learning_rate": 1e-06,
      "loss": 0.1409,
      "step": 2107
    },
    {
      "epoch": 0.7309292649098474,
      "grad_norm": 3.0315585136413574,
      "learning_rate": 1e-06,
      "loss": 0.1822,
      "step": 2108
    },
    {
      "epoch": 0.7312760055478502,
      "grad_norm": 2.9591143131256104,
      "learning_rate": 1e-06,
      "loss": 0.1946,
      "step": 2109
    },
    {
      "epoch": 0.731622746185853,
      "grad_norm": 3.1562037467956543,
      "learning_rate": 1e-06,
      "loss": 0.1889,
      "step": 2110
    },
    {
      "epoch": 0.7319694868238558,
      "grad_norm": 3.8917124271392822,
      "learning_rate": 1e-06,
      "loss": 0.1746,
      "step": 2111
    },
    {
      "epoch": 0.7323162274618585,
      "grad_norm": 2.4440743923187256,
      "learning_rate": 1e-06,
      "loss": 0.1837,
      "step": 2112
    },
    {
      "epoch": 0.7326629680998613,
      "grad_norm": 2.5348663330078125,
      "learning_rate": 1e-06,
      "loss": 0.1625,
      "step": 2113
    },
    {
      "epoch": 0.7330097087378641,
      "grad_norm": 1.7352176904678345,
      "learning_rate": 1e-06,
      "loss": 0.1518,
      "step": 2114
    },
    {
      "epoch": 0.7333564493758669,
      "grad_norm": 2.3291268348693848,
      "learning_rate": 1e-06,
      "loss": 0.1341,
      "step": 2115
    },
    {
      "epoch": 0.7337031900138696,
      "grad_norm": 3.349982738494873,
      "learning_rate": 1e-06,
      "loss": 0.166,
      "step": 2116
    },
    {
      "epoch": 0.7340499306518724,
      "grad_norm": 2.1478614807128906,
      "learning_rate": 1e-06,
      "loss": 0.161,
      "step": 2117
    },
    {
      "epoch": 0.7343966712898752,
      "grad_norm": 2.126044750213623,
      "learning_rate": 1e-06,
      "loss": 0.1441,
      "step": 2118
    },
    {
      "epoch": 0.734743411927878,
      "grad_norm": 2.5404157638549805,
      "learning_rate": 1e-06,
      "loss": 0.164,
      "step": 2119
    },
    {
      "epoch": 0.7350901525658807,
      "grad_norm": 2.651552677154541,
      "learning_rate": 1e-06,
      "loss": 0.1599,
      "step": 2120
    },
    {
      "epoch": 0.7354368932038835,
      "grad_norm": 4.563192367553711,
      "learning_rate": 1e-06,
      "loss": 0.124,
      "step": 2121
    },
    {
      "epoch": 0.7357836338418863,
      "grad_norm": 1.917719841003418,
      "learning_rate": 1e-06,
      "loss": 0.165,
      "step": 2122
    },
    {
      "epoch": 0.736130374479889,
      "grad_norm": 3.211930274963379,
      "learning_rate": 1e-06,
      "loss": 0.1625,
      "step": 2123
    },
    {
      "epoch": 0.7364771151178918,
      "grad_norm": 2.4502320289611816,
      "learning_rate": 1e-06,
      "loss": 0.1318,
      "step": 2124
    },
    {
      "epoch": 0.7368238557558946,
      "grad_norm": 2.0151150226593018,
      "learning_rate": 1e-06,
      "loss": 0.1832,
      "step": 2125
    },
    {
      "epoch": 0.7371705963938974,
      "grad_norm": 3.2877564430236816,
      "learning_rate": 1e-06,
      "loss": 0.149,
      "step": 2126
    },
    {
      "epoch": 0.7375173370319001,
      "grad_norm": 2.4384498596191406,
      "learning_rate": 1e-06,
      "loss": 0.1717,
      "step": 2127
    },
    {
      "epoch": 0.7378640776699029,
      "grad_norm": 1.685181736946106,
      "learning_rate": 1e-06,
      "loss": 0.1361,
      "step": 2128
    },
    {
      "epoch": 0.7382108183079057,
      "grad_norm": 1.982335090637207,
      "learning_rate": 1e-06,
      "loss": 0.1598,
      "step": 2129
    },
    {
      "epoch": 0.7385575589459085,
      "grad_norm": 1.7395703792572021,
      "learning_rate": 1e-06,
      "loss": 0.1523,
      "step": 2130
    },
    {
      "epoch": 0.7389042995839112,
      "grad_norm": 4.207993507385254,
      "learning_rate": 1e-06,
      "loss": 0.1749,
      "step": 2131
    },
    {
      "epoch": 0.739251040221914,
      "grad_norm": 2.950206995010376,
      "learning_rate": 1e-06,
      "loss": 0.1748,
      "step": 2132
    },
    {
      "epoch": 0.7395977808599168,
      "grad_norm": 2.3146812915802,
      "learning_rate": 1e-06,
      "loss": 0.1379,
      "step": 2133
    },
    {
      "epoch": 0.7399445214979196,
      "grad_norm": 3.0385754108428955,
      "learning_rate": 1e-06,
      "loss": 0.149,
      "step": 2134
    },
    {
      "epoch": 0.7402912621359223,
      "grad_norm": 2.4939308166503906,
      "learning_rate": 1e-06,
      "loss": 0.1621,
      "step": 2135
    },
    {
      "epoch": 0.7406380027739251,
      "grad_norm": 3.453312397003174,
      "learning_rate": 1e-06,
      "loss": 0.1817,
      "step": 2136
    },
    {
      "epoch": 0.7409847434119279,
      "grad_norm": 1.8451257944107056,
      "learning_rate": 1e-06,
      "loss": 0.1169,
      "step": 2137
    },
    {
      "epoch": 0.7413314840499307,
      "grad_norm": 1.7449294328689575,
      "learning_rate": 1e-06,
      "loss": 0.1307,
      "step": 2138
    },
    {
      "epoch": 0.7416782246879334,
      "grad_norm": 2.023353338241577,
      "learning_rate": 1e-06,
      "loss": 0.1229,
      "step": 2139
    },
    {
      "epoch": 0.7420249653259362,
      "grad_norm": 2.990633726119995,
      "learning_rate": 1e-06,
      "loss": 0.1469,
      "step": 2140
    },
    {
      "epoch": 0.742371705963939,
      "grad_norm": 2.25154709815979,
      "learning_rate": 1e-06,
      "loss": 0.1447,
      "step": 2141
    },
    {
      "epoch": 0.7427184466019418,
      "grad_norm": 2.082965135574341,
      "learning_rate": 1e-06,
      "loss": 0.1315,
      "step": 2142
    },
    {
      "epoch": 0.7430651872399445,
      "grad_norm": 2.342172145843506,
      "learning_rate": 1e-06,
      "loss": 0.1613,
      "step": 2143
    },
    {
      "epoch": 0.7434119278779473,
      "grad_norm": 1.9159072637557983,
      "learning_rate": 1e-06,
      "loss": 0.128,
      "step": 2144
    },
    {
      "epoch": 0.7437586685159501,
      "grad_norm": 2.7071194648742676,
      "learning_rate": 1e-06,
      "loss": 0.1023,
      "step": 2145
    },
    {
      "epoch": 0.7441054091539528,
      "grad_norm": 1.955583095550537,
      "learning_rate": 1e-06,
      "loss": 0.1694,
      "step": 2146
    },
    {
      "epoch": 0.7444521497919556,
      "grad_norm": 1.6116520166397095,
      "learning_rate": 1e-06,
      "loss": 0.1452,
      "step": 2147
    },
    {
      "epoch": 0.7447988904299584,
      "grad_norm": 2.063347101211548,
      "learning_rate": 1e-06,
      "loss": 0.1365,
      "step": 2148
    },
    {
      "epoch": 0.7451456310679612,
      "grad_norm": 2.245347023010254,
      "learning_rate": 1e-06,
      "loss": 0.1437,
      "step": 2149
    },
    {
      "epoch": 0.7454923717059639,
      "grad_norm": 2.005086660385132,
      "learning_rate": 1e-06,
      "loss": 0.1031,
      "step": 2150
    },
    {
      "epoch": 0.7458391123439667,
      "grad_norm": 3.6264290809631348,
      "learning_rate": 1e-06,
      "loss": 0.1509,
      "step": 2151
    },
    {
      "epoch": 0.7461858529819695,
      "grad_norm": 3.764726161956787,
      "learning_rate": 1e-06,
      "loss": 0.1373,
      "step": 2152
    },
    {
      "epoch": 0.7465325936199723,
      "grad_norm": 5.636833667755127,
      "learning_rate": 1e-06,
      "loss": 0.1851,
      "step": 2153
    },
    {
      "epoch": 0.746879334257975,
      "grad_norm": 3.568516969680786,
      "learning_rate": 1e-06,
      "loss": 0.1337,
      "step": 2154
    },
    {
      "epoch": 0.7472260748959778,
      "grad_norm": 1.9982422590255737,
      "learning_rate": 1e-06,
      "loss": 0.1399,
      "step": 2155
    },
    {
      "epoch": 0.7475728155339806,
      "grad_norm": 2.262208938598633,
      "learning_rate": 1e-06,
      "loss": 0.1318,
      "step": 2156
    },
    {
      "epoch": 0.7479195561719834,
      "grad_norm": 2.338360071182251,
      "learning_rate": 1e-06,
      "loss": 0.1532,
      "step": 2157
    },
    {
      "epoch": 0.7482662968099861,
      "grad_norm": 2.570775032043457,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 2158
    },
    {
      "epoch": 0.7486130374479889,
      "grad_norm": 2.5094552040100098,
      "learning_rate": 1e-06,
      "loss": 0.1252,
      "step": 2159
    },
    {
      "epoch": 0.7489597780859917,
      "grad_norm": 3.4843361377716064,
      "learning_rate": 1e-06,
      "loss": 0.1238,
      "step": 2160
    },
    {
      "epoch": 0.7493065187239945,
      "grad_norm": 2.467364549636841,
      "learning_rate": 1e-06,
      "loss": 0.1446,
      "step": 2161
    },
    {
      "epoch": 0.7496532593619972,
      "grad_norm": 2.1688084602355957,
      "learning_rate": 1e-06,
      "loss": 0.146,
      "step": 2162
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.409573793411255,
      "learning_rate": 1e-06,
      "loss": 0.1702,
      "step": 2163
    },
    {
      "epoch": 0.7503467406380028,
      "grad_norm": 3.1284704208374023,
      "learning_rate": 1e-06,
      "loss": 0.1131,
      "step": 2164
    },
    {
      "epoch": 0.7506934812760055,
      "grad_norm": 2.3599703311920166,
      "learning_rate": 1e-06,
      "loss": 0.1569,
      "step": 2165
    },
    {
      "epoch": 0.7510402219140083,
      "grad_norm": 2.9270598888397217,
      "learning_rate": 1e-06,
      "loss": 0.1723,
      "step": 2166
    },
    {
      "epoch": 0.7513869625520111,
      "grad_norm": 3.9771506786346436,
      "learning_rate": 1e-06,
      "loss": 0.1508,
      "step": 2167
    },
    {
      "epoch": 0.7517337031900139,
      "grad_norm": 2.2273499965667725,
      "learning_rate": 1e-06,
      "loss": 0.1313,
      "step": 2168
    },
    {
      "epoch": 0.7520804438280166,
      "grad_norm": 2.1485869884490967,
      "learning_rate": 1e-06,
      "loss": 0.1242,
      "step": 2169
    },
    {
      "epoch": 0.7524271844660194,
      "grad_norm": 3.504756212234497,
      "learning_rate": 1e-06,
      "loss": 0.1515,
      "step": 2170
    },
    {
      "epoch": 0.7527739251040222,
      "grad_norm": 3.7918713092803955,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 2171
    },
    {
      "epoch": 0.753120665742025,
      "grad_norm": 3.0239686965942383,
      "learning_rate": 1e-06,
      "loss": 0.1645,
      "step": 2172
    },
    {
      "epoch": 0.7534674063800277,
      "grad_norm": 2.892434597015381,
      "learning_rate": 1e-06,
      "loss": 0.133,
      "step": 2173
    },
    {
      "epoch": 0.7538141470180305,
      "grad_norm": 3.0266504287719727,
      "learning_rate": 1e-06,
      "loss": 0.1423,
      "step": 2174
    },
    {
      "epoch": 0.7541608876560333,
      "grad_norm": 2.232647180557251,
      "learning_rate": 1e-06,
      "loss": 0.1101,
      "step": 2175
    },
    {
      "epoch": 0.7545076282940361,
      "grad_norm": 2.5927672386169434,
      "learning_rate": 1e-06,
      "loss": 0.1474,
      "step": 2176
    },
    {
      "epoch": 0.7548543689320388,
      "grad_norm": 4.152268409729004,
      "learning_rate": 1e-06,
      "loss": 0.1451,
      "step": 2177
    },
    {
      "epoch": 0.7552011095700416,
      "grad_norm": 2.1622960567474365,
      "learning_rate": 1e-06,
      "loss": 0.1412,
      "step": 2178
    },
    {
      "epoch": 0.7555478502080444,
      "grad_norm": 3.911947250366211,
      "learning_rate": 1e-06,
      "loss": 0.1432,
      "step": 2179
    },
    {
      "epoch": 0.7558945908460472,
      "grad_norm": 3.2365176677703857,
      "learning_rate": 1e-06,
      "loss": 0.1284,
      "step": 2180
    },
    {
      "epoch": 0.7562413314840499,
      "grad_norm": 2.945903778076172,
      "learning_rate": 1e-06,
      "loss": 0.1461,
      "step": 2181
    },
    {
      "epoch": 0.7565880721220527,
      "grad_norm": 3.0714845657348633,
      "learning_rate": 1e-06,
      "loss": 0.1857,
      "step": 2182
    },
    {
      "epoch": 0.7569348127600555,
      "grad_norm": 2.2900753021240234,
      "learning_rate": 1e-06,
      "loss": 0.1419,
      "step": 2183
    },
    {
      "epoch": 0.7572815533980582,
      "grad_norm": 2.031371593475342,
      "learning_rate": 1e-06,
      "loss": 0.1216,
      "step": 2184
    },
    {
      "epoch": 0.757628294036061,
      "grad_norm": 2.7789697647094727,
      "learning_rate": 1e-06,
      "loss": 0.1275,
      "step": 2185
    },
    {
      "epoch": 0.7579750346740638,
      "grad_norm": 2.324058771133423,
      "learning_rate": 1e-06,
      "loss": 0.1307,
      "step": 2186
    },
    {
      "epoch": 0.7583217753120666,
      "grad_norm": 3.2857003211975098,
      "learning_rate": 1e-06,
      "loss": 0.1331,
      "step": 2187
    },
    {
      "epoch": 0.7586685159500693,
      "grad_norm": 3.0428361892700195,
      "learning_rate": 1e-06,
      "loss": 0.1491,
      "step": 2188
    },
    {
      "epoch": 0.7590152565880721,
      "grad_norm": 2.756199598312378,
      "learning_rate": 1e-06,
      "loss": 0.115,
      "step": 2189
    },
    {
      "epoch": 0.7593619972260749,
      "grad_norm": 2.3134708404541016,
      "learning_rate": 1e-06,
      "loss": 0.1233,
      "step": 2190
    },
    {
      "epoch": 0.7597087378640777,
      "grad_norm": 2.598783016204834,
      "learning_rate": 1e-06,
      "loss": 0.1333,
      "step": 2191
    },
    {
      "epoch": 0.7600554785020804,
      "grad_norm": 3.0439116954803467,
      "learning_rate": 1e-06,
      "loss": 0.1508,
      "step": 2192
    },
    {
      "epoch": 0.7604022191400832,
      "grad_norm": 2.078136682510376,
      "learning_rate": 1e-06,
      "loss": 0.1192,
      "step": 2193
    },
    {
      "epoch": 0.760748959778086,
      "grad_norm": 2.642509937286377,
      "learning_rate": 1e-06,
      "loss": 0.1577,
      "step": 2194
    },
    {
      "epoch": 0.7610957004160888,
      "grad_norm": 2.7496683597564697,
      "learning_rate": 1e-06,
      "loss": 0.1658,
      "step": 2195
    },
    {
      "epoch": 0.7614424410540915,
      "grad_norm": 2.2078957557678223,
      "learning_rate": 1e-06,
      "loss": 0.1016,
      "step": 2196
    },
    {
      "epoch": 0.7617891816920943,
      "grad_norm": 1.7725173234939575,
      "learning_rate": 1e-06,
      "loss": 0.1178,
      "step": 2197
    },
    {
      "epoch": 0.7621359223300971,
      "grad_norm": 2.821902275085449,
      "learning_rate": 1e-06,
      "loss": 0.1306,
      "step": 2198
    },
    {
      "epoch": 0.7624826629680999,
      "grad_norm": 1.9677958488464355,
      "learning_rate": 1e-06,
      "loss": 0.1499,
      "step": 2199
    },
    {
      "epoch": 0.7628294036061026,
      "grad_norm": 1.9492521286010742,
      "learning_rate": 1e-06,
      "loss": 0.1408,
      "step": 2200
    },
    {
      "epoch": 0.7631761442441054,
      "grad_norm": 2.5732200145721436,
      "learning_rate": 1e-06,
      "loss": 0.1239,
      "step": 2201
    },
    {
      "epoch": 0.7635228848821082,
      "grad_norm": 5.270609378814697,
      "learning_rate": 1e-06,
      "loss": 0.1715,
      "step": 2202
    },
    {
      "epoch": 0.763869625520111,
      "grad_norm": 2.55741286277771,
      "learning_rate": 1e-06,
      "loss": 0.1183,
      "step": 2203
    },
    {
      "epoch": 0.7642163661581137,
      "grad_norm": 2.0879695415496826,
      "learning_rate": 1e-06,
      "loss": 0.1571,
      "step": 2204
    },
    {
      "epoch": 0.7645631067961165,
      "grad_norm": 3.306610584259033,
      "learning_rate": 1e-06,
      "loss": 0.1819,
      "step": 2205
    },
    {
      "epoch": 0.7649098474341193,
      "grad_norm": 5.571447849273682,
      "learning_rate": 1e-06,
      "loss": 0.1288,
      "step": 2206
    },
    {
      "epoch": 0.765256588072122,
      "grad_norm": 2.29826283454895,
      "learning_rate": 1e-06,
      "loss": 0.1707,
      "step": 2207
    },
    {
      "epoch": 0.7656033287101248,
      "grad_norm": 2.8895623683929443,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 2208
    },
    {
      "epoch": 0.7659500693481276,
      "grad_norm": 2.1935150623321533,
      "learning_rate": 1e-06,
      "loss": 0.148,
      "step": 2209
    },
    {
      "epoch": 0.7662968099861304,
      "grad_norm": 2.136641502380371,
      "learning_rate": 1e-06,
      "loss": 0.1317,
      "step": 2210
    },
    {
      "epoch": 0.7666435506241331,
      "grad_norm": 2.338508367538452,
      "learning_rate": 1e-06,
      "loss": 0.143,
      "step": 2211
    },
    {
      "epoch": 0.7669902912621359,
      "grad_norm": 1.9527918100357056,
      "learning_rate": 1e-06,
      "loss": 0.129,
      "step": 2212
    },
    {
      "epoch": 0.7673370319001387,
      "grad_norm": 3.5570404529571533,
      "learning_rate": 1e-06,
      "loss": 0.1631,
      "step": 2213
    },
    {
      "epoch": 0.7676837725381415,
      "grad_norm": 1.9036303758621216,
      "learning_rate": 1e-06,
      "loss": 0.1453,
      "step": 2214
    },
    {
      "epoch": 0.7680305131761442,
      "grad_norm": 3.279881238937378,
      "learning_rate": 1e-06,
      "loss": 0.1561,
      "step": 2215
    },
    {
      "epoch": 0.768377253814147,
      "grad_norm": 3.6818034648895264,
      "learning_rate": 1e-06,
      "loss": 0.1404,
      "step": 2216
    },
    {
      "epoch": 0.7687239944521498,
      "grad_norm": 2.868618965148926,
      "learning_rate": 1e-06,
      "loss": 0.1492,
      "step": 2217
    },
    {
      "epoch": 0.7690707350901526,
      "grad_norm": 2.422971725463867,
      "learning_rate": 1e-06,
      "loss": 0.1465,
      "step": 2218
    },
    {
      "epoch": 0.7694174757281553,
      "grad_norm": 3.9464166164398193,
      "learning_rate": 1e-06,
      "loss": 0.1541,
      "step": 2219
    },
    {
      "epoch": 0.7697642163661581,
      "grad_norm": 2.302083969116211,
      "learning_rate": 1e-06,
      "loss": 0.1585,
      "step": 2220
    },
    {
      "epoch": 0.7701109570041609,
      "grad_norm": 2.771343946456909,
      "learning_rate": 1e-06,
      "loss": 0.127,
      "step": 2221
    },
    {
      "epoch": 0.7704576976421637,
      "grad_norm": 2.0313382148742676,
      "learning_rate": 1e-06,
      "loss": 0.1635,
      "step": 2222
    },
    {
      "epoch": 0.7708044382801664,
      "grad_norm": 3.974736213684082,
      "learning_rate": 1e-06,
      "loss": 0.1493,
      "step": 2223
    },
    {
      "epoch": 0.7711511789181692,
      "grad_norm": 2.2979185581207275,
      "learning_rate": 1e-06,
      "loss": 0.1205,
      "step": 2224
    },
    {
      "epoch": 0.771497919556172,
      "grad_norm": 2.6536171436309814,
      "learning_rate": 1e-06,
      "loss": 0.1431,
      "step": 2225
    },
    {
      "epoch": 0.7718446601941747,
      "grad_norm": 2.917592763900757,
      "learning_rate": 1e-06,
      "loss": 0.1473,
      "step": 2226
    },
    {
      "epoch": 0.7721914008321775,
      "grad_norm": 2.4181203842163086,
      "learning_rate": 1e-06,
      "loss": 0.1502,
      "step": 2227
    },
    {
      "epoch": 0.7725381414701803,
      "grad_norm": 2.419362783432007,
      "learning_rate": 1e-06,
      "loss": 0.1183,
      "step": 2228
    },
    {
      "epoch": 0.7728848821081831,
      "grad_norm": 2.5661659240722656,
      "learning_rate": 1e-06,
      "loss": 0.1402,
      "step": 2229
    },
    {
      "epoch": 0.7732316227461858,
      "grad_norm": 1.9720306396484375,
      "learning_rate": 1e-06,
      "loss": 0.1253,
      "step": 2230
    },
    {
      "epoch": 0.7735783633841886,
      "grad_norm": 3.7642719745635986,
      "learning_rate": 1e-06,
      "loss": 0.1552,
      "step": 2231
    },
    {
      "epoch": 0.7739251040221914,
      "grad_norm": 8.331310272216797,
      "learning_rate": 1e-06,
      "loss": 0.2239,
      "step": 2232
    },
    {
      "epoch": 0.7742718446601942,
      "grad_norm": 2.8353569507598877,
      "learning_rate": 1e-06,
      "loss": 0.1348,
      "step": 2233
    },
    {
      "epoch": 0.7746185852981969,
      "grad_norm": 3.4421985149383545,
      "learning_rate": 1e-06,
      "loss": 0.1813,
      "step": 2234
    },
    {
      "epoch": 0.7749653259361997,
      "grad_norm": 3.0186569690704346,
      "learning_rate": 1e-06,
      "loss": 0.1741,
      "step": 2235
    },
    {
      "epoch": 0.7753120665742025,
      "grad_norm": 1.918044924736023,
      "learning_rate": 1e-06,
      "loss": 0.1231,
      "step": 2236
    },
    {
      "epoch": 0.7756588072122053,
      "grad_norm": 2.8227319717407227,
      "learning_rate": 1e-06,
      "loss": 0.141,
      "step": 2237
    },
    {
      "epoch": 0.776005547850208,
      "grad_norm": 1.9563891887664795,
      "learning_rate": 1e-06,
      "loss": 0.1403,
      "step": 2238
    },
    {
      "epoch": 0.7763522884882108,
      "grad_norm": 3.1032557487487793,
      "learning_rate": 1e-06,
      "loss": 0.1455,
      "step": 2239
    },
    {
      "epoch": 0.7766990291262136,
      "grad_norm": 2.261861801147461,
      "learning_rate": 1e-06,
      "loss": 0.1397,
      "step": 2240
    },
    {
      "epoch": 0.7770457697642164,
      "grad_norm": 3.733213186264038,
      "learning_rate": 1e-06,
      "loss": 0.167,
      "step": 2241
    },
    {
      "epoch": 0.7773925104022191,
      "grad_norm": 2.189244270324707,
      "learning_rate": 1e-06,
      "loss": 0.1243,
      "step": 2242
    },
    {
      "epoch": 0.7777392510402219,
      "grad_norm": 3.7104716300964355,
      "learning_rate": 1e-06,
      "loss": 0.1295,
      "step": 2243
    },
    {
      "epoch": 0.7780859916782247,
      "grad_norm": 1.9497910737991333,
      "learning_rate": 1e-06,
      "loss": 0.1408,
      "step": 2244
    },
    {
      "epoch": 0.7784327323162274,
      "grad_norm": 3.3845322132110596,
      "learning_rate": 1e-06,
      "loss": 0.1478,
      "step": 2245
    },
    {
      "epoch": 0.7787794729542302,
      "grad_norm": 1.5930759906768799,
      "learning_rate": 1e-06,
      "loss": 0.1126,
      "step": 2246
    },
    {
      "epoch": 0.779126213592233,
      "grad_norm": 2.993636131286621,
      "learning_rate": 1e-06,
      "loss": 0.1232,
      "step": 2247
    },
    {
      "epoch": 0.7794729542302358,
      "grad_norm": 2.2873573303222656,
      "learning_rate": 1e-06,
      "loss": 0.1456,
      "step": 2248
    },
    {
      "epoch": 0.7798196948682385,
      "grad_norm": 3.0485949516296387,
      "learning_rate": 1e-06,
      "loss": 0.1277,
      "step": 2249
    },
    {
      "epoch": 0.7801664355062413,
      "grad_norm": 3.5445034503936768,
      "learning_rate": 1e-06,
      "loss": 0.145,
      "step": 2250
    },
    {
      "epoch": 0.7805131761442441,
      "grad_norm": 2.8255326747894287,
      "learning_rate": 1e-06,
      "loss": 0.1321,
      "step": 2251
    },
    {
      "epoch": 0.7808599167822469,
      "grad_norm": 2.1605257987976074,
      "learning_rate": 1e-06,
      "loss": 0.1152,
      "step": 2252
    },
    {
      "epoch": 0.7812066574202496,
      "grad_norm": 2.541396379470825,
      "learning_rate": 1e-06,
      "loss": 0.1488,
      "step": 2253
    },
    {
      "epoch": 0.7815533980582524,
      "grad_norm": 2.67930006980896,
      "learning_rate": 1e-06,
      "loss": 0.1334,
      "step": 2254
    },
    {
      "epoch": 0.7819001386962552,
      "grad_norm": 2.717165470123291,
      "learning_rate": 1e-06,
      "loss": 0.1487,
      "step": 2255
    },
    {
      "epoch": 0.782246879334258,
      "grad_norm": 2.5122649669647217,
      "learning_rate": 1e-06,
      "loss": 0.128,
      "step": 2256
    },
    {
      "epoch": 0.7825936199722607,
      "grad_norm": 3.2565906047821045,
      "learning_rate": 1e-06,
      "loss": 0.1541,
      "step": 2257
    },
    {
      "epoch": 0.7829403606102635,
      "grad_norm": 2.227794647216797,
      "learning_rate": 1e-06,
      "loss": 0.1562,
      "step": 2258
    },
    {
      "epoch": 0.7832871012482663,
      "grad_norm": 1.8468554019927979,
      "learning_rate": 1e-06,
      "loss": 0.1258,
      "step": 2259
    },
    {
      "epoch": 0.7836338418862691,
      "grad_norm": 2.6450536251068115,
      "learning_rate": 1e-06,
      "loss": 0.1218,
      "step": 2260
    },
    {
      "epoch": 0.7839805825242718,
      "grad_norm": 3.5764429569244385,
      "learning_rate": 1e-06,
      "loss": 0.1399,
      "step": 2261
    },
    {
      "epoch": 0.7843273231622746,
      "grad_norm": 3.0315048694610596,
      "learning_rate": 1e-06,
      "loss": 0.1651,
      "step": 2262
    },
    {
      "epoch": 0.7846740638002774,
      "grad_norm": 3.2928550243377686,
      "learning_rate": 1e-06,
      "loss": 0.1461,
      "step": 2263
    },
    {
      "epoch": 0.7850208044382802,
      "grad_norm": 2.096841812133789,
      "learning_rate": 1e-06,
      "loss": 0.1297,
      "step": 2264
    },
    {
      "epoch": 0.7853675450762829,
      "grad_norm": 2.753180980682373,
      "learning_rate": 1e-06,
      "loss": 0.1545,
      "step": 2265
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 3.1257476806640625,
      "learning_rate": 1e-06,
      "loss": 0.1371,
      "step": 2266
    },
    {
      "epoch": 0.7860610263522885,
      "grad_norm": 3.747185707092285,
      "learning_rate": 1e-06,
      "loss": 0.148,
      "step": 2267
    },
    {
      "epoch": 0.7864077669902912,
      "grad_norm": 2.7665867805480957,
      "learning_rate": 1e-06,
      "loss": 0.1316,
      "step": 2268
    },
    {
      "epoch": 0.786754507628294,
      "grad_norm": 4.6021318435668945,
      "learning_rate": 1e-06,
      "loss": 0.1613,
      "step": 2269
    },
    {
      "epoch": 0.7871012482662968,
      "grad_norm": 3.030388116836548,
      "learning_rate": 1e-06,
      "loss": 0.1333,
      "step": 2270
    },
    {
      "epoch": 0.7874479889042996,
      "grad_norm": 2.6754000186920166,
      "learning_rate": 1e-06,
      "loss": 0.1391,
      "step": 2271
    },
    {
      "epoch": 0.7877947295423023,
      "grad_norm": 3.4339709281921387,
      "learning_rate": 1e-06,
      "loss": 0.1393,
      "step": 2272
    },
    {
      "epoch": 0.7881414701803051,
      "grad_norm": 2.5777359008789062,
      "learning_rate": 1e-06,
      "loss": 0.1269,
      "step": 2273
    },
    {
      "epoch": 0.7884882108183079,
      "grad_norm": 3.1862616539001465,
      "learning_rate": 1e-06,
      "loss": 0.175,
      "step": 2274
    },
    {
      "epoch": 0.7888349514563107,
      "grad_norm": 2.623753786087036,
      "learning_rate": 1e-06,
      "loss": 0.1374,
      "step": 2275
    },
    {
      "epoch": 0.7891816920943134,
      "grad_norm": 2.4776697158813477,
      "learning_rate": 1e-06,
      "loss": 0.1435,
      "step": 2276
    },
    {
      "epoch": 0.7895284327323162,
      "grad_norm": 2.2111032009124756,
      "learning_rate": 1e-06,
      "loss": 0.1139,
      "step": 2277
    },
    {
      "epoch": 0.789875173370319,
      "grad_norm": 2.097119092941284,
      "learning_rate": 1e-06,
      "loss": 0.1246,
      "step": 2278
    },
    {
      "epoch": 0.7902219140083218,
      "grad_norm": 2.417785882949829,
      "learning_rate": 1e-06,
      "loss": 0.1433,
      "step": 2279
    },
    {
      "epoch": 0.7905686546463245,
      "grad_norm": 3.1172475814819336,
      "learning_rate": 1e-06,
      "loss": 0.1155,
      "step": 2280
    },
    {
      "epoch": 0.7909153952843273,
      "grad_norm": 2.320932626724243,
      "learning_rate": 1e-06,
      "loss": 0.1368,
      "step": 2281
    },
    {
      "epoch": 0.7912621359223301,
      "grad_norm": 1.8932015895843506,
      "learning_rate": 1e-06,
      "loss": 0.1325,
      "step": 2282
    },
    {
      "epoch": 0.7916088765603329,
      "grad_norm": 2.369478225708008,
      "learning_rate": 1e-06,
      "loss": 0.1814,
      "step": 2283
    },
    {
      "epoch": 0.7919556171983356,
      "grad_norm": 1.7178212404251099,
      "learning_rate": 1e-06,
      "loss": 0.1213,
      "step": 2284
    },
    {
      "epoch": 0.7923023578363384,
      "grad_norm": 2.493131399154663,
      "learning_rate": 1e-06,
      "loss": 0.1378,
      "step": 2285
    },
    {
      "epoch": 0.7926490984743412,
      "grad_norm": 2.1028831005096436,
      "learning_rate": 1e-06,
      "loss": 0.137,
      "step": 2286
    },
    {
      "epoch": 0.792995839112344,
      "grad_norm": 3.674328088760376,
      "learning_rate": 1e-06,
      "loss": 0.156,
      "step": 2287
    },
    {
      "epoch": 0.7933425797503467,
      "grad_norm": 2.515397310256958,
      "learning_rate": 1e-06,
      "loss": 0.1278,
      "step": 2288
    },
    {
      "epoch": 0.7936893203883495,
      "grad_norm": 2.559921979904175,
      "learning_rate": 1e-06,
      "loss": 0.1363,
      "step": 2289
    },
    {
      "epoch": 0.7940360610263523,
      "grad_norm": 3.420501232147217,
      "learning_rate": 1e-06,
      "loss": 0.1326,
      "step": 2290
    },
    {
      "epoch": 0.794382801664355,
      "grad_norm": 1.9830430746078491,
      "learning_rate": 1e-06,
      "loss": 0.1242,
      "step": 2291
    },
    {
      "epoch": 0.7947295423023578,
      "grad_norm": 1.8026798963546753,
      "learning_rate": 1e-06,
      "loss": 0.116,
      "step": 2292
    },
    {
      "epoch": 0.7950762829403606,
      "grad_norm": 2.667938232421875,
      "learning_rate": 1e-06,
      "loss": 0.14,
      "step": 2293
    },
    {
      "epoch": 0.7954230235783634,
      "grad_norm": 3.140871524810791,
      "learning_rate": 1e-06,
      "loss": 0.137,
      "step": 2294
    },
    {
      "epoch": 0.7957697642163661,
      "grad_norm": 3.2242789268493652,
      "learning_rate": 1e-06,
      "loss": 0.1559,
      "step": 2295
    },
    {
      "epoch": 0.7961165048543689,
      "grad_norm": 2.4139838218688965,
      "learning_rate": 1e-06,
      "loss": 0.1286,
      "step": 2296
    },
    {
      "epoch": 0.7964632454923717,
      "grad_norm": 1.957815408706665,
      "learning_rate": 1e-06,
      "loss": 0.1292,
      "step": 2297
    },
    {
      "epoch": 0.7968099861303745,
      "grad_norm": 4.165436267852783,
      "learning_rate": 1e-06,
      "loss": 0.1342,
      "step": 2298
    },
    {
      "epoch": 0.7971567267683772,
      "grad_norm": 2.260488748550415,
      "learning_rate": 1e-06,
      "loss": 0.126,
      "step": 2299
    },
    {
      "epoch": 0.79750346740638,
      "grad_norm": 2.7062175273895264,
      "learning_rate": 1e-06,
      "loss": 0.1285,
      "step": 2300
    },
    {
      "epoch": 0.7978502080443828,
      "grad_norm": 3.5681710243225098,
      "learning_rate": 1e-06,
      "loss": 0.1115,
      "step": 2301
    },
    {
      "epoch": 0.7981969486823856,
      "grad_norm": 4.171554088592529,
      "learning_rate": 1e-06,
      "loss": 0.12,
      "step": 2302
    },
    {
      "epoch": 0.7985436893203883,
      "grad_norm": 1.9496839046478271,
      "learning_rate": 1e-06,
      "loss": 0.1323,
      "step": 2303
    },
    {
      "epoch": 0.7988904299583911,
      "grad_norm": 2.0774519443511963,
      "learning_rate": 1e-06,
      "loss": 0.1353,
      "step": 2304
    },
    {
      "epoch": 0.7992371705963939,
      "grad_norm": 2.0676655769348145,
      "learning_rate": 1e-06,
      "loss": 0.1294,
      "step": 2305
    },
    {
      "epoch": 0.7995839112343966,
      "grad_norm": 2.6793768405914307,
      "learning_rate": 1e-06,
      "loss": 0.1349,
      "step": 2306
    },
    {
      "epoch": 0.7999306518723994,
      "grad_norm": 1.8901208639144897,
      "learning_rate": 1e-06,
      "loss": 0.1318,
      "step": 2307
    },
    {
      "epoch": 0.8002773925104022,
      "grad_norm": 2.2167181968688965,
      "learning_rate": 1e-06,
      "loss": 0.1393,
      "step": 2308
    },
    {
      "epoch": 0.800624133148405,
      "grad_norm": 3.0722689628601074,
      "learning_rate": 1e-06,
      "loss": 0.1565,
      "step": 2309
    },
    {
      "epoch": 0.8009708737864077,
      "grad_norm": 2.4346120357513428,
      "learning_rate": 1e-06,
      "loss": 0.1498,
      "step": 2310
    },
    {
      "epoch": 0.8013176144244105,
      "grad_norm": 1.9527393579483032,
      "learning_rate": 1e-06,
      "loss": 0.1459,
      "step": 2311
    },
    {
      "epoch": 0.8016643550624133,
      "grad_norm": 2.2069756984710693,
      "learning_rate": 1e-06,
      "loss": 0.166,
      "step": 2312
    },
    {
      "epoch": 0.8020110957004161,
      "grad_norm": 3.2333059310913086,
      "learning_rate": 1e-06,
      "loss": 0.1786,
      "step": 2313
    },
    {
      "epoch": 0.8023578363384188,
      "grad_norm": 3.749083995819092,
      "learning_rate": 1e-06,
      "loss": 0.1594,
      "step": 2314
    },
    {
      "epoch": 0.8027045769764216,
      "grad_norm": 2.32126522064209,
      "learning_rate": 1e-06,
      "loss": 0.1161,
      "step": 2315
    },
    {
      "epoch": 0.8030513176144244,
      "grad_norm": 3.9740304946899414,
      "learning_rate": 1e-06,
      "loss": 0.1492,
      "step": 2316
    },
    {
      "epoch": 0.8033980582524272,
      "grad_norm": 2.4023282527923584,
      "learning_rate": 1e-06,
      "loss": 0.1319,
      "step": 2317
    },
    {
      "epoch": 0.8037447988904299,
      "grad_norm": 2.1061739921569824,
      "learning_rate": 1e-06,
      "loss": 0.1239,
      "step": 2318
    },
    {
      "epoch": 0.8040915395284327,
      "grad_norm": 2.2139999866485596,
      "learning_rate": 1e-06,
      "loss": 0.1479,
      "step": 2319
    },
    {
      "epoch": 0.8044382801664355,
      "grad_norm": 2.9679105281829834,
      "learning_rate": 1e-06,
      "loss": 0.1466,
      "step": 2320
    },
    {
      "epoch": 0.8047850208044383,
      "grad_norm": 2.461606740951538,
      "learning_rate": 1e-06,
      "loss": 0.1505,
      "step": 2321
    },
    {
      "epoch": 0.805131761442441,
      "grad_norm": 2.9136569499969482,
      "learning_rate": 1e-06,
      "loss": 0.1626,
      "step": 2322
    },
    {
      "epoch": 0.8054785020804438,
      "grad_norm": 1.9064357280731201,
      "learning_rate": 1e-06,
      "loss": 0.138,
      "step": 2323
    },
    {
      "epoch": 0.8058252427184466,
      "grad_norm": 2.3706328868865967,
      "learning_rate": 1e-06,
      "loss": 0.1311,
      "step": 2324
    },
    {
      "epoch": 0.8061719833564494,
      "grad_norm": 3.0558860301971436,
      "learning_rate": 1e-06,
      "loss": 0.1171,
      "step": 2325
    },
    {
      "epoch": 0.8065187239944521,
      "grad_norm": 4.301748275756836,
      "learning_rate": 1e-06,
      "loss": 0.1781,
      "step": 2326
    },
    {
      "epoch": 0.8068654646324549,
      "grad_norm": 3.325167655944824,
      "learning_rate": 1e-06,
      "loss": 0.1443,
      "step": 2327
    },
    {
      "epoch": 0.8072122052704577,
      "grad_norm": 2.0337326526641846,
      "learning_rate": 1e-06,
      "loss": 0.1317,
      "step": 2328
    },
    {
      "epoch": 0.8075589459084604,
      "grad_norm": 2.0139780044555664,
      "learning_rate": 1e-06,
      "loss": 0.163,
      "step": 2329
    },
    {
      "epoch": 0.8079056865464632,
      "grad_norm": 2.4566049575805664,
      "learning_rate": 1e-06,
      "loss": 0.1758,
      "step": 2330
    },
    {
      "epoch": 0.808252427184466,
      "grad_norm": 2.1362626552581787,
      "learning_rate": 1e-06,
      "loss": 0.1435,
      "step": 2331
    },
    {
      "epoch": 0.8085991678224688,
      "grad_norm": 1.9229339361190796,
      "learning_rate": 1e-06,
      "loss": 0.1444,
      "step": 2332
    },
    {
      "epoch": 0.8089459084604715,
      "grad_norm": 3.5861167907714844,
      "learning_rate": 1e-06,
      "loss": 0.1486,
      "step": 2333
    },
    {
      "epoch": 0.8092926490984743,
      "grad_norm": 1.9148346185684204,
      "learning_rate": 1e-06,
      "loss": 0.1355,
      "step": 2334
    },
    {
      "epoch": 0.8096393897364771,
      "grad_norm": 2.1705756187438965,
      "learning_rate": 1e-06,
      "loss": 0.1402,
      "step": 2335
    },
    {
      "epoch": 0.8099861303744799,
      "grad_norm": 1.75710928440094,
      "learning_rate": 1e-06,
      "loss": 0.1188,
      "step": 2336
    },
    {
      "epoch": 0.8103328710124826,
      "grad_norm": 1.7081013917922974,
      "learning_rate": 1e-06,
      "loss": 0.1314,
      "step": 2337
    },
    {
      "epoch": 0.8106796116504854,
      "grad_norm": 3.3994858264923096,
      "learning_rate": 1e-06,
      "loss": 0.1017,
      "step": 2338
    },
    {
      "epoch": 0.8110263522884882,
      "grad_norm": 3.6731457710266113,
      "learning_rate": 1e-06,
      "loss": 0.1161,
      "step": 2339
    },
    {
      "epoch": 0.811373092926491,
      "grad_norm": 2.8746140003204346,
      "learning_rate": 1e-06,
      "loss": 0.1453,
      "step": 2340
    },
    {
      "epoch": 0.8117198335644937,
      "grad_norm": 2.074737310409546,
      "learning_rate": 1e-06,
      "loss": 0.1477,
      "step": 2341
    },
    {
      "epoch": 0.8120665742024965,
      "grad_norm": 3.902578592300415,
      "learning_rate": 1e-06,
      "loss": 0.1372,
      "step": 2342
    },
    {
      "epoch": 0.8124133148404993,
      "grad_norm": 2.1763041019439697,
      "learning_rate": 1e-06,
      "loss": 0.1246,
      "step": 2343
    },
    {
      "epoch": 0.812760055478502,
      "grad_norm": 3.487217903137207,
      "learning_rate": 1e-06,
      "loss": 0.1074,
      "step": 2344
    },
    {
      "epoch": 0.8131067961165048,
      "grad_norm": 1.8151427507400513,
      "learning_rate": 1e-06,
      "loss": 0.0811,
      "step": 2345
    },
    {
      "epoch": 0.8134535367545076,
      "grad_norm": 2.7768118381500244,
      "learning_rate": 1e-06,
      "loss": 0.17,
      "step": 2346
    },
    {
      "epoch": 0.8138002773925104,
      "grad_norm": 2.1280786991119385,
      "learning_rate": 1e-06,
      "loss": 0.152,
      "step": 2347
    },
    {
      "epoch": 0.8141470180305131,
      "grad_norm": 3.410444974899292,
      "learning_rate": 1e-06,
      "loss": 0.1596,
      "step": 2348
    },
    {
      "epoch": 0.8144937586685159,
      "grad_norm": 1.9879850149154663,
      "learning_rate": 1e-06,
      "loss": 0.1277,
      "step": 2349
    },
    {
      "epoch": 0.8148404993065187,
      "grad_norm": 2.854005813598633,
      "learning_rate": 1e-06,
      "loss": 0.1358,
      "step": 2350
    },
    {
      "epoch": 0.8151872399445215,
      "grad_norm": 2.091987133026123,
      "learning_rate": 1e-06,
      "loss": 0.1321,
      "step": 2351
    },
    {
      "epoch": 0.8155339805825242,
      "grad_norm": 2.376237392425537,
      "learning_rate": 1e-06,
      "loss": 0.1499,
      "step": 2352
    },
    {
      "epoch": 0.815880721220527,
      "grad_norm": 2.1092143058776855,
      "learning_rate": 1e-06,
      "loss": 0.1543,
      "step": 2353
    },
    {
      "epoch": 0.8162274618585298,
      "grad_norm": 2.8994827270507812,
      "learning_rate": 1e-06,
      "loss": 0.1463,
      "step": 2354
    },
    {
      "epoch": 0.8165742024965326,
      "grad_norm": 2.607811689376831,
      "learning_rate": 1e-06,
      "loss": 0.1171,
      "step": 2355
    },
    {
      "epoch": 0.8169209431345353,
      "grad_norm": 3.0912883281707764,
      "learning_rate": 1e-06,
      "loss": 0.1444,
      "step": 2356
    },
    {
      "epoch": 0.8172676837725381,
      "grad_norm": 2.976580858230591,
      "learning_rate": 1e-06,
      "loss": 0.1761,
      "step": 2357
    },
    {
      "epoch": 0.8176144244105409,
      "grad_norm": 2.745866060256958,
      "learning_rate": 1e-06,
      "loss": 0.1282,
      "step": 2358
    },
    {
      "epoch": 0.8179611650485437,
      "grad_norm": 2.7969820499420166,
      "learning_rate": 1e-06,
      "loss": 0.117,
      "step": 2359
    },
    {
      "epoch": 0.8183079056865464,
      "grad_norm": 2.0547497272491455,
      "learning_rate": 1e-06,
      "loss": 0.1322,
      "step": 2360
    },
    {
      "epoch": 0.8186546463245492,
      "grad_norm": 2.9290833473205566,
      "learning_rate": 1e-06,
      "loss": 0.18,
      "step": 2361
    },
    {
      "epoch": 0.819001386962552,
      "grad_norm": 2.636199474334717,
      "learning_rate": 1e-06,
      "loss": 0.1636,
      "step": 2362
    },
    {
      "epoch": 0.8193481276005548,
      "grad_norm": 2.2920427322387695,
      "learning_rate": 1e-06,
      "loss": 0.1362,
      "step": 2363
    },
    {
      "epoch": 0.8196948682385575,
      "grad_norm": 2.04429030418396,
      "learning_rate": 1e-06,
      "loss": 0.1486,
      "step": 2364
    },
    {
      "epoch": 0.8200416088765603,
      "grad_norm": 2.0373737812042236,
      "learning_rate": 1e-06,
      "loss": 0.1257,
      "step": 2365
    },
    {
      "epoch": 0.8203883495145631,
      "grad_norm": 2.8232688903808594,
      "learning_rate": 1e-06,
      "loss": 0.1549,
      "step": 2366
    },
    {
      "epoch": 0.8207350901525658,
      "grad_norm": 2.8030526638031006,
      "learning_rate": 1e-06,
      "loss": 0.1429,
      "step": 2367
    },
    {
      "epoch": 0.8210818307905686,
      "grad_norm": 3.216235399246216,
      "learning_rate": 1e-06,
      "loss": 0.1781,
      "step": 2368
    },
    {
      "epoch": 0.8214285714285714,
      "grad_norm": 2.1101338863372803,
      "learning_rate": 1e-06,
      "loss": 0.1353,
      "step": 2369
    },
    {
      "epoch": 0.8217753120665742,
      "grad_norm": 3.3085856437683105,
      "learning_rate": 1e-06,
      "loss": 0.1526,
      "step": 2370
    },
    {
      "epoch": 0.8221220527045769,
      "grad_norm": 2.0516085624694824,
      "learning_rate": 1e-06,
      "loss": 0.148,
      "step": 2371
    },
    {
      "epoch": 0.8224687933425797,
      "grad_norm": 2.2691268920898438,
      "learning_rate": 1e-06,
      "loss": 0.1306,
      "step": 2372
    },
    {
      "epoch": 0.8228155339805825,
      "grad_norm": 2.2179465293884277,
      "learning_rate": 1e-06,
      "loss": 0.1412,
      "step": 2373
    },
    {
      "epoch": 0.8231622746185853,
      "grad_norm": 2.9574997425079346,
      "learning_rate": 1e-06,
      "loss": 0.1928,
      "step": 2374
    },
    {
      "epoch": 0.823509015256588,
      "grad_norm": 2.3397562503814697,
      "learning_rate": 1e-06,
      "loss": 0.1478,
      "step": 2375
    },
    {
      "epoch": 0.8238557558945908,
      "grad_norm": 2.1710567474365234,
      "learning_rate": 1e-06,
      "loss": 0.1619,
      "step": 2376
    },
    {
      "epoch": 0.8242024965325936,
      "grad_norm": 2.402174472808838,
      "learning_rate": 1e-06,
      "loss": 0.1495,
      "step": 2377
    },
    {
      "epoch": 0.8245492371705964,
      "grad_norm": 2.147606372833252,
      "learning_rate": 1e-06,
      "loss": 0.1261,
      "step": 2378
    },
    {
      "epoch": 0.8248959778085991,
      "grad_norm": 2.674955129623413,
      "learning_rate": 1e-06,
      "loss": 0.1174,
      "step": 2379
    },
    {
      "epoch": 0.8252427184466019,
      "grad_norm": 3.098621129989624,
      "learning_rate": 1e-06,
      "loss": 0.1686,
      "step": 2380
    },
    {
      "epoch": 0.8255894590846047,
      "grad_norm": 2.411407232284546,
      "learning_rate": 1e-06,
      "loss": 0.1175,
      "step": 2381
    },
    {
      "epoch": 0.8259361997226075,
      "grad_norm": 3.399182081222534,
      "learning_rate": 1e-06,
      "loss": 0.1345,
      "step": 2382
    },
    {
      "epoch": 0.8262829403606102,
      "grad_norm": 3.7523751258850098,
      "learning_rate": 1e-06,
      "loss": 0.139,
      "step": 2383
    },
    {
      "epoch": 0.826629680998613,
      "grad_norm": 2.860729217529297,
      "learning_rate": 1e-06,
      "loss": 0.1782,
      "step": 2384
    },
    {
      "epoch": 0.8269764216366158,
      "grad_norm": 4.691393852233887,
      "learning_rate": 1e-06,
      "loss": 0.1536,
      "step": 2385
    },
    {
      "epoch": 0.8273231622746186,
      "grad_norm": 2.227851152420044,
      "learning_rate": 1e-06,
      "loss": 0.1156,
      "step": 2386
    },
    {
      "epoch": 0.8276699029126213,
      "grad_norm": 2.429328680038452,
      "learning_rate": 1e-06,
      "loss": 0.1305,
      "step": 2387
    },
    {
      "epoch": 0.8280166435506241,
      "grad_norm": 1.7986654043197632,
      "learning_rate": 1e-06,
      "loss": 0.1271,
      "step": 2388
    },
    {
      "epoch": 0.8283633841886269,
      "grad_norm": 2.666243314743042,
      "learning_rate": 1e-06,
      "loss": 0.1173,
      "step": 2389
    },
    {
      "epoch": 0.8287101248266296,
      "grad_norm": 3.4500293731689453,
      "learning_rate": 1e-06,
      "loss": 0.1979,
      "step": 2390
    },
    {
      "epoch": 0.8290568654646324,
      "grad_norm": 2.057880401611328,
      "learning_rate": 1e-06,
      "loss": 0.1349,
      "step": 2391
    },
    {
      "epoch": 0.8294036061026352,
      "grad_norm": 2.378908634185791,
      "learning_rate": 1e-06,
      "loss": 0.129,
      "step": 2392
    },
    {
      "epoch": 0.829750346740638,
      "grad_norm": 1.690531849861145,
      "learning_rate": 1e-06,
      "loss": 0.1249,
      "step": 2393
    },
    {
      "epoch": 0.8300970873786407,
      "grad_norm": 3.056180953979492,
      "learning_rate": 1e-06,
      "loss": 0.1269,
      "step": 2394
    },
    {
      "epoch": 0.8304438280166435,
      "grad_norm": 4.274582862854004,
      "learning_rate": 1e-06,
      "loss": 0.1123,
      "step": 2395
    },
    {
      "epoch": 0.8307905686546463,
      "grad_norm": 2.3732125759124756,
      "learning_rate": 1e-06,
      "loss": 0.1626,
      "step": 2396
    },
    {
      "epoch": 0.8311373092926491,
      "grad_norm": 2.0551693439483643,
      "learning_rate": 1e-06,
      "loss": 0.1381,
      "step": 2397
    },
    {
      "epoch": 0.8314840499306518,
      "grad_norm": 3.1651957035064697,
      "learning_rate": 1e-06,
      "loss": 0.143,
      "step": 2398
    },
    {
      "epoch": 0.8318307905686546,
      "grad_norm": 1.9696643352508545,
      "learning_rate": 1e-06,
      "loss": 0.1238,
      "step": 2399
    },
    {
      "epoch": 0.8321775312066574,
      "grad_norm": 2.9673922061920166,
      "learning_rate": 1e-06,
      "loss": 0.1699,
      "step": 2400
    },
    {
      "epoch": 0.8325242718446602,
      "grad_norm": 2.1757307052612305,
      "learning_rate": 1e-06,
      "loss": 0.1287,
      "step": 2401
    },
    {
      "epoch": 0.8328710124826629,
      "grad_norm": 2.689530611038208,
      "learning_rate": 1e-06,
      "loss": 0.1407,
      "step": 2402
    },
    {
      "epoch": 0.8332177531206657,
      "grad_norm": 3.808122396469116,
      "learning_rate": 1e-06,
      "loss": 0.1323,
      "step": 2403
    },
    {
      "epoch": 0.8335644937586685,
      "grad_norm": 2.106861114501953,
      "learning_rate": 1e-06,
      "loss": 0.1375,
      "step": 2404
    },
    {
      "epoch": 0.8339112343966713,
      "grad_norm": 3.0327420234680176,
      "learning_rate": 1e-06,
      "loss": 0.1323,
      "step": 2405
    },
    {
      "epoch": 0.834257975034674,
      "grad_norm": 2.0406875610351562,
      "learning_rate": 1e-06,
      "loss": 0.1349,
      "step": 2406
    },
    {
      "epoch": 0.8346047156726768,
      "grad_norm": 5.959853172302246,
      "learning_rate": 1e-06,
      "loss": 0.1603,
      "step": 2407
    },
    {
      "epoch": 0.8349514563106796,
      "grad_norm": 2.409048080444336,
      "learning_rate": 1e-06,
      "loss": 0.1401,
      "step": 2408
    },
    {
      "epoch": 0.8352981969486823,
      "grad_norm": 2.135956287384033,
      "learning_rate": 1e-06,
      "loss": 0.1406,
      "step": 2409
    },
    {
      "epoch": 0.8356449375866851,
      "grad_norm": 2.256753921508789,
      "learning_rate": 1e-06,
      "loss": 0.1348,
      "step": 2410
    },
    {
      "epoch": 0.8359916782246879,
      "grad_norm": 2.43808913230896,
      "learning_rate": 1e-06,
      "loss": 0.1495,
      "step": 2411
    },
    {
      "epoch": 0.8363384188626907,
      "grad_norm": 2.6997463703155518,
      "learning_rate": 1e-06,
      "loss": 0.1227,
      "step": 2412
    },
    {
      "epoch": 0.8366851595006934,
      "grad_norm": 3.3730976581573486,
      "learning_rate": 1e-06,
      "loss": 0.1212,
      "step": 2413
    },
    {
      "epoch": 0.8370319001386962,
      "grad_norm": 3.1687965393066406,
      "learning_rate": 1e-06,
      "loss": 0.1268,
      "step": 2414
    },
    {
      "epoch": 0.837378640776699,
      "grad_norm": 2.4090065956115723,
      "learning_rate": 1e-06,
      "loss": 0.1449,
      "step": 2415
    },
    {
      "epoch": 0.8377253814147018,
      "grad_norm": 2.7174880504608154,
      "learning_rate": 1e-06,
      "loss": 0.1413,
      "step": 2416
    },
    {
      "epoch": 0.8380721220527045,
      "grad_norm": 2.5149476528167725,
      "learning_rate": 1e-06,
      "loss": 0.1568,
      "step": 2417
    },
    {
      "epoch": 0.8384188626907073,
      "grad_norm": 4.081730842590332,
      "learning_rate": 1e-06,
      "loss": 0.1159,
      "step": 2418
    },
    {
      "epoch": 0.8387656033287101,
      "grad_norm": 2.8811569213867188,
      "learning_rate": 1e-06,
      "loss": 0.1729,
      "step": 2419
    },
    {
      "epoch": 0.8391123439667129,
      "grad_norm": 2.758857011795044,
      "learning_rate": 1e-06,
      "loss": 0.1391,
      "step": 2420
    },
    {
      "epoch": 0.8394590846047156,
      "grad_norm": 2.5798659324645996,
      "learning_rate": 1e-06,
      "loss": 0.1211,
      "step": 2421
    },
    {
      "epoch": 0.8398058252427184,
      "grad_norm": 5.093081951141357,
      "learning_rate": 1e-06,
      "loss": 0.1562,
      "step": 2422
    },
    {
      "epoch": 0.8401525658807212,
      "grad_norm": 2.7829928398132324,
      "learning_rate": 1e-06,
      "loss": 0.116,
      "step": 2423
    },
    {
      "epoch": 0.840499306518724,
      "grad_norm": 5.886650562286377,
      "learning_rate": 1e-06,
      "loss": 0.1584,
      "step": 2424
    },
    {
      "epoch": 0.8408460471567267,
      "grad_norm": 2.695272207260132,
      "learning_rate": 1e-06,
      "loss": 0.1519,
      "step": 2425
    },
    {
      "epoch": 0.8411927877947295,
      "grad_norm": 1.8741930723190308,
      "learning_rate": 1e-06,
      "loss": 0.1183,
      "step": 2426
    },
    {
      "epoch": 0.8415395284327323,
      "grad_norm": 2.3263349533081055,
      "learning_rate": 1e-06,
      "loss": 0.1567,
      "step": 2427
    },
    {
      "epoch": 0.841886269070735,
      "grad_norm": 2.279158353805542,
      "learning_rate": 1e-06,
      "loss": 0.1386,
      "step": 2428
    },
    {
      "epoch": 0.8422330097087378,
      "grad_norm": 2.275637626647949,
      "learning_rate": 1e-06,
      "loss": 0.1325,
      "step": 2429
    },
    {
      "epoch": 0.8425797503467406,
      "grad_norm": 2.705381393432617,
      "learning_rate": 1e-06,
      "loss": 0.1842,
      "step": 2430
    },
    {
      "epoch": 0.8429264909847434,
      "grad_norm": 2.2983760833740234,
      "learning_rate": 1e-06,
      "loss": 0.1288,
      "step": 2431
    },
    {
      "epoch": 0.8432732316227461,
      "grad_norm": 2.5210258960723877,
      "learning_rate": 1e-06,
      "loss": 0.1506,
      "step": 2432
    },
    {
      "epoch": 0.8436199722607489,
      "grad_norm": 2.4116249084472656,
      "learning_rate": 1e-06,
      "loss": 0.1377,
      "step": 2433
    },
    {
      "epoch": 0.8439667128987517,
      "grad_norm": 3.0181736946105957,
      "learning_rate": 1e-06,
      "loss": 0.1206,
      "step": 2434
    },
    {
      "epoch": 0.8443134535367545,
      "grad_norm": 2.6214163303375244,
      "learning_rate": 1e-06,
      "loss": 0.1807,
      "step": 2435
    },
    {
      "epoch": 0.8446601941747572,
      "grad_norm": 2.952951669692993,
      "learning_rate": 1e-06,
      "loss": 0.1392,
      "step": 2436
    },
    {
      "epoch": 0.84500693481276,
      "grad_norm": 2.5941317081451416,
      "learning_rate": 1e-06,
      "loss": 0.1359,
      "step": 2437
    },
    {
      "epoch": 0.8453536754507628,
      "grad_norm": 3.742173433303833,
      "learning_rate": 1e-06,
      "loss": 0.1257,
      "step": 2438
    },
    {
      "epoch": 0.8457004160887656,
      "grad_norm": 2.2225148677825928,
      "learning_rate": 1e-06,
      "loss": 0.1341,
      "step": 2439
    },
    {
      "epoch": 0.8460471567267683,
      "grad_norm": 2.1798903942108154,
      "learning_rate": 1e-06,
      "loss": 0.1436,
      "step": 2440
    },
    {
      "epoch": 0.8463938973647711,
      "grad_norm": 2.250204563140869,
      "learning_rate": 1e-06,
      "loss": 0.1274,
      "step": 2441
    },
    {
      "epoch": 0.8467406380027739,
      "grad_norm": 2.7729578018188477,
      "learning_rate": 1e-06,
      "loss": 0.1432,
      "step": 2442
    },
    {
      "epoch": 0.8470873786407767,
      "grad_norm": 2.45955491065979,
      "learning_rate": 1e-06,
      "loss": 0.1196,
      "step": 2443
    },
    {
      "epoch": 0.8474341192787794,
      "grad_norm": 2.97157621383667,
      "learning_rate": 1e-06,
      "loss": 0.1218,
      "step": 2444
    },
    {
      "epoch": 0.8477808599167822,
      "grad_norm": 2.3688995838165283,
      "learning_rate": 1e-06,
      "loss": 0.1171,
      "step": 2445
    },
    {
      "epoch": 0.848127600554785,
      "grad_norm": 2.0294275283813477,
      "learning_rate": 1e-06,
      "loss": 0.1428,
      "step": 2446
    },
    {
      "epoch": 0.8484743411927878,
      "grad_norm": 2.387620687484741,
      "learning_rate": 1e-06,
      "loss": 0.1597,
      "step": 2447
    },
    {
      "epoch": 0.8488210818307905,
      "grad_norm": 2.6711483001708984,
      "learning_rate": 1e-06,
      "loss": 0.1353,
      "step": 2448
    },
    {
      "epoch": 0.8491678224687933,
      "grad_norm": 1.8898664712905884,
      "learning_rate": 1e-06,
      "loss": 0.1174,
      "step": 2449
    },
    {
      "epoch": 0.8495145631067961,
      "grad_norm": 3.391063928604126,
      "learning_rate": 1e-06,
      "loss": 0.1485,
      "step": 2450
    },
    {
      "epoch": 0.8498613037447988,
      "grad_norm": 2.1252994537353516,
      "learning_rate": 1e-06,
      "loss": 0.1212,
      "step": 2451
    },
    {
      "epoch": 0.8502080443828016,
      "grad_norm": 2.0090253353118896,
      "learning_rate": 1e-06,
      "loss": 0.1171,
      "step": 2452
    },
    {
      "epoch": 0.8505547850208044,
      "grad_norm": 2.170214891433716,
      "learning_rate": 1e-06,
      "loss": 0.1487,
      "step": 2453
    },
    {
      "epoch": 0.8509015256588072,
      "grad_norm": 2.922431468963623,
      "learning_rate": 1e-06,
      "loss": 0.1148,
      "step": 2454
    },
    {
      "epoch": 0.8512482662968099,
      "grad_norm": 2.2969465255737305,
      "learning_rate": 1e-06,
      "loss": 0.1417,
      "step": 2455
    },
    {
      "epoch": 0.8515950069348127,
      "grad_norm": 2.0030033588409424,
      "learning_rate": 1e-06,
      "loss": 0.1146,
      "step": 2456
    },
    {
      "epoch": 0.8519417475728155,
      "grad_norm": 2.1828982830047607,
      "learning_rate": 1e-06,
      "loss": 0.1204,
      "step": 2457
    },
    {
      "epoch": 0.8522884882108183,
      "grad_norm": 1.8292583227157593,
      "learning_rate": 1e-06,
      "loss": 0.1263,
      "step": 2458
    },
    {
      "epoch": 0.852635228848821,
      "grad_norm": 2.6252031326293945,
      "learning_rate": 1e-06,
      "loss": 0.1443,
      "step": 2459
    },
    {
      "epoch": 0.8529819694868238,
      "grad_norm": 2.30623459815979,
      "learning_rate": 1e-06,
      "loss": 0.1436,
      "step": 2460
    },
    {
      "epoch": 0.8533287101248266,
      "grad_norm": 2.8300130367279053,
      "learning_rate": 1e-06,
      "loss": 0.1644,
      "step": 2461
    },
    {
      "epoch": 0.8536754507628294,
      "grad_norm": 2.417668104171753,
      "learning_rate": 1e-06,
      "loss": 0.1432,
      "step": 2462
    },
    {
      "epoch": 0.8540221914008321,
      "grad_norm": 2.5126116275787354,
      "learning_rate": 1e-06,
      "loss": 0.1228,
      "step": 2463
    },
    {
      "epoch": 0.8543689320388349,
      "grad_norm": 2.0431835651397705,
      "learning_rate": 1e-06,
      "loss": 0.135,
      "step": 2464
    },
    {
      "epoch": 0.8547156726768377,
      "grad_norm": 3.008413076400757,
      "learning_rate": 1e-06,
      "loss": 0.1218,
      "step": 2465
    },
    {
      "epoch": 0.8550624133148405,
      "grad_norm": 2.4672775268554688,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 2466
    },
    {
      "epoch": 0.8554091539528432,
      "grad_norm": 3.271653652191162,
      "learning_rate": 1e-06,
      "loss": 0.1474,
      "step": 2467
    },
    {
      "epoch": 0.855755894590846,
      "grad_norm": 2.1060640811920166,
      "learning_rate": 1e-06,
      "loss": 0.112,
      "step": 2468
    },
    {
      "epoch": 0.8561026352288488,
      "grad_norm": 1.9942084550857544,
      "learning_rate": 1e-06,
      "loss": 0.1288,
      "step": 2469
    },
    {
      "epoch": 0.8564493758668515,
      "grad_norm": 2.6707653999328613,
      "learning_rate": 1e-06,
      "loss": 0.1613,
      "step": 2470
    },
    {
      "epoch": 0.8567961165048543,
      "grad_norm": 3.239352226257324,
      "learning_rate": 1e-06,
      "loss": 0.1444,
      "step": 2471
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 2.7429757118225098,
      "learning_rate": 1e-06,
      "loss": 0.1516,
      "step": 2472
    },
    {
      "epoch": 0.8574895977808599,
      "grad_norm": 2.1264328956604004,
      "learning_rate": 1e-06,
      "loss": 0.1657,
      "step": 2473
    },
    {
      "epoch": 0.8578363384188626,
      "grad_norm": 2.4109182357788086,
      "learning_rate": 1e-06,
      "loss": 0.1331,
      "step": 2474
    },
    {
      "epoch": 0.8581830790568654,
      "grad_norm": 2.9945809841156006,
      "learning_rate": 1e-06,
      "loss": 0.1349,
      "step": 2475
    },
    {
      "epoch": 0.8585298196948682,
      "grad_norm": 2.7945027351379395,
      "learning_rate": 1e-06,
      "loss": 0.151,
      "step": 2476
    },
    {
      "epoch": 0.858876560332871,
      "grad_norm": 1.5665547847747803,
      "learning_rate": 1e-06,
      "loss": 0.1231,
      "step": 2477
    },
    {
      "epoch": 0.8592233009708737,
      "grad_norm": 2.601025342941284,
      "learning_rate": 1e-06,
      "loss": 0.1207,
      "step": 2478
    },
    {
      "epoch": 0.8595700416088765,
      "grad_norm": 3.027022123336792,
      "learning_rate": 1e-06,
      "loss": 0.1592,
      "step": 2479
    },
    {
      "epoch": 0.8599167822468793,
      "grad_norm": 2.1945347785949707,
      "learning_rate": 1e-06,
      "loss": 0.142,
      "step": 2480
    },
    {
      "epoch": 0.8602635228848821,
      "grad_norm": 2.1606740951538086,
      "learning_rate": 1e-06,
      "loss": 0.1387,
      "step": 2481
    },
    {
      "epoch": 0.8606102635228848,
      "grad_norm": 3.8076674938201904,
      "learning_rate": 1e-06,
      "loss": 0.1128,
      "step": 2482
    },
    {
      "epoch": 0.8609570041608876,
      "grad_norm": 3.2100863456726074,
      "learning_rate": 1e-06,
      "loss": 0.1345,
      "step": 2483
    },
    {
      "epoch": 0.8613037447988904,
      "grad_norm": 2.945937156677246,
      "learning_rate": 1e-06,
      "loss": 0.1253,
      "step": 2484
    },
    {
      "epoch": 0.8616504854368932,
      "grad_norm": 2.1155993938446045,
      "learning_rate": 1e-06,
      "loss": 0.1366,
      "step": 2485
    },
    {
      "epoch": 0.8619972260748959,
      "grad_norm": 2.137275457382202,
      "learning_rate": 1e-06,
      "loss": 0.1434,
      "step": 2486
    },
    {
      "epoch": 0.8623439667128987,
      "grad_norm": 3.105685234069824,
      "learning_rate": 1e-06,
      "loss": 0.1242,
      "step": 2487
    },
    {
      "epoch": 0.8626907073509015,
      "grad_norm": 2.3001792430877686,
      "learning_rate": 1e-06,
      "loss": 0.1322,
      "step": 2488
    },
    {
      "epoch": 0.8630374479889042,
      "grad_norm": 4.752837657928467,
      "learning_rate": 1e-06,
      "loss": 0.1471,
      "step": 2489
    },
    {
      "epoch": 0.863384188626907,
      "grad_norm": 4.404865264892578,
      "learning_rate": 1e-06,
      "loss": 0.1753,
      "step": 2490
    },
    {
      "epoch": 0.8637309292649098,
      "grad_norm": 2.0198276042938232,
      "learning_rate": 1e-06,
      "loss": 0.1351,
      "step": 2491
    },
    {
      "epoch": 0.8640776699029126,
      "grad_norm": 4.447784900665283,
      "learning_rate": 1e-06,
      "loss": 0.1736,
      "step": 2492
    },
    {
      "epoch": 0.8644244105409153,
      "grad_norm": 2.295395612716675,
      "learning_rate": 1e-06,
      "loss": 0.1327,
      "step": 2493
    },
    {
      "epoch": 0.8647711511789181,
      "grad_norm": 1.9984354972839355,
      "learning_rate": 1e-06,
      "loss": 0.1164,
      "step": 2494
    },
    {
      "epoch": 0.8651178918169209,
      "grad_norm": 2.847456693649292,
      "learning_rate": 1e-06,
      "loss": 0.1441,
      "step": 2495
    },
    {
      "epoch": 0.8654646324549237,
      "grad_norm": 3.244466543197632,
      "learning_rate": 1e-06,
      "loss": 0.1434,
      "step": 2496
    },
    {
      "epoch": 0.8658113730929264,
      "grad_norm": 2.110191583633423,
      "learning_rate": 1e-06,
      "loss": 0.1269,
      "step": 2497
    },
    {
      "epoch": 0.8661581137309292,
      "grad_norm": 2.554386615753174,
      "learning_rate": 1e-06,
      "loss": 0.1522,
      "step": 2498
    },
    {
      "epoch": 0.866504854368932,
      "grad_norm": 3.446542501449585,
      "learning_rate": 1e-06,
      "loss": 0.1483,
      "step": 2499
    },
    {
      "epoch": 0.8668515950069348,
      "grad_norm": 2.2236502170562744,
      "learning_rate": 1e-06,
      "loss": 0.1384,
      "step": 2500
    },
    {
      "epoch": 0.8671983356449375,
      "grad_norm": 2.7683444023132324,
      "learning_rate": 1e-06,
      "loss": 0.1283,
      "step": 2501
    },
    {
      "epoch": 0.8675450762829403,
      "grad_norm": 3.002000093460083,
      "learning_rate": 1e-06,
      "loss": 0.1742,
      "step": 2502
    },
    {
      "epoch": 0.8678918169209431,
      "grad_norm": 5.195772647857666,
      "learning_rate": 1e-06,
      "loss": 0.1753,
      "step": 2503
    },
    {
      "epoch": 0.8682385575589459,
      "grad_norm": 3.325066566467285,
      "learning_rate": 1e-06,
      "loss": 0.1238,
      "step": 2504
    },
    {
      "epoch": 0.8685852981969486,
      "grad_norm": 2.8764262199401855,
      "learning_rate": 1e-06,
      "loss": 0.1269,
      "step": 2505
    },
    {
      "epoch": 0.8689320388349514,
      "grad_norm": 4.183605670928955,
      "learning_rate": 1e-06,
      "loss": 0.173,
      "step": 2506
    },
    {
      "epoch": 0.8692787794729542,
      "grad_norm": 2.656033515930176,
      "learning_rate": 1e-06,
      "loss": 0.1767,
      "step": 2507
    },
    {
      "epoch": 0.869625520110957,
      "grad_norm": 2.7101523876190186,
      "learning_rate": 1e-06,
      "loss": 0.1274,
      "step": 2508
    },
    {
      "epoch": 0.8699722607489597,
      "grad_norm": 2.912912368774414,
      "learning_rate": 1e-06,
      "loss": 0.1603,
      "step": 2509
    },
    {
      "epoch": 0.8703190013869625,
      "grad_norm": 2.7677292823791504,
      "learning_rate": 1e-06,
      "loss": 0.1403,
      "step": 2510
    },
    {
      "epoch": 0.8706657420249653,
      "grad_norm": 7.316020965576172,
      "learning_rate": 1e-06,
      "loss": 0.1854,
      "step": 2511
    },
    {
      "epoch": 0.871012482662968,
      "grad_norm": 2.9252498149871826,
      "learning_rate": 1e-06,
      "loss": 0.1385,
      "step": 2512
    },
    {
      "epoch": 0.8713592233009708,
      "grad_norm": 2.6173603534698486,
      "learning_rate": 1e-06,
      "loss": 0.1405,
      "step": 2513
    },
    {
      "epoch": 0.8717059639389736,
      "grad_norm": 3.6852455139160156,
      "learning_rate": 1e-06,
      "loss": 0.1621,
      "step": 2514
    },
    {
      "epoch": 0.8720527045769764,
      "grad_norm": 3.2382307052612305,
      "learning_rate": 1e-06,
      "loss": 0.0994,
      "step": 2515
    },
    {
      "epoch": 0.8723994452149791,
      "grad_norm": 2.688394784927368,
      "learning_rate": 1e-06,
      "loss": 0.1423,
      "step": 2516
    },
    {
      "epoch": 0.8727461858529819,
      "grad_norm": 3.035778522491455,
      "learning_rate": 1e-06,
      "loss": 0.1563,
      "step": 2517
    },
    {
      "epoch": 0.8730929264909847,
      "grad_norm": 2.3126778602600098,
      "learning_rate": 1e-06,
      "loss": 0.1394,
      "step": 2518
    },
    {
      "epoch": 0.8734396671289875,
      "grad_norm": 2.3858559131622314,
      "learning_rate": 1e-06,
      "loss": 0.1215,
      "step": 2519
    },
    {
      "epoch": 0.8737864077669902,
      "grad_norm": 3.0119643211364746,
      "learning_rate": 1e-06,
      "loss": 0.1157,
      "step": 2520
    },
    {
      "epoch": 0.874133148404993,
      "grad_norm": 2.6011157035827637,
      "learning_rate": 1e-06,
      "loss": 0.1646,
      "step": 2521
    },
    {
      "epoch": 0.8744798890429958,
      "grad_norm": 2.6988086700439453,
      "learning_rate": 1e-06,
      "loss": 0.1424,
      "step": 2522
    },
    {
      "epoch": 0.8748266296809986,
      "grad_norm": 2.830876588821411,
      "learning_rate": 1e-06,
      "loss": 0.1378,
      "step": 2523
    },
    {
      "epoch": 0.8751733703190014,
      "grad_norm": 1.7022074460983276,
      "learning_rate": 1e-06,
      "loss": 0.1197,
      "step": 2524
    },
    {
      "epoch": 0.8755201109570042,
      "grad_norm": 2.1738150119781494,
      "learning_rate": 1e-06,
      "loss": 0.123,
      "step": 2525
    },
    {
      "epoch": 0.875866851595007,
      "grad_norm": 3.662080764770508,
      "learning_rate": 1e-06,
      "loss": 0.1436,
      "step": 2526
    },
    {
      "epoch": 0.8762135922330098,
      "grad_norm": 1.7804447412490845,
      "learning_rate": 1e-06,
      "loss": 0.0925,
      "step": 2527
    },
    {
      "epoch": 0.8765603328710125,
      "grad_norm": 2.030134916305542,
      "learning_rate": 1e-06,
      "loss": 0.1294,
      "step": 2528
    },
    {
      "epoch": 0.8769070735090153,
      "grad_norm": 1.968540072441101,
      "learning_rate": 1e-06,
      "loss": 0.1247,
      "step": 2529
    },
    {
      "epoch": 0.8772538141470181,
      "grad_norm": 3.905348539352417,
      "learning_rate": 1e-06,
      "loss": 0.1248,
      "step": 2530
    },
    {
      "epoch": 0.8776005547850209,
      "grad_norm": 2.988689661026001,
      "learning_rate": 1e-06,
      "loss": 0.1408,
      "step": 2531
    },
    {
      "epoch": 0.8779472954230236,
      "grad_norm": 2.6646010875701904,
      "learning_rate": 1e-06,
      "loss": 0.137,
      "step": 2532
    },
    {
      "epoch": 0.8782940360610264,
      "grad_norm": 3.523129940032959,
      "learning_rate": 1e-06,
      "loss": 0.1305,
      "step": 2533
    },
    {
      "epoch": 0.8786407766990292,
      "grad_norm": 7.069967269897461,
      "learning_rate": 1e-06,
      "loss": 0.151,
      "step": 2534
    },
    {
      "epoch": 0.878987517337032,
      "grad_norm": 2.2202308177948,
      "learning_rate": 1e-06,
      "loss": 0.1281,
      "step": 2535
    },
    {
      "epoch": 0.8793342579750347,
      "grad_norm": 4.117976188659668,
      "learning_rate": 1e-06,
      "loss": 0.1423,
      "step": 2536
    },
    {
      "epoch": 0.8796809986130375,
      "grad_norm": 2.210732936859131,
      "learning_rate": 1e-06,
      "loss": 0.1138,
      "step": 2537
    },
    {
      "epoch": 0.8800277392510403,
      "grad_norm": 2.42452335357666,
      "learning_rate": 1e-06,
      "loss": 0.1292,
      "step": 2538
    },
    {
      "epoch": 0.880374479889043,
      "grad_norm": 4.303539276123047,
      "learning_rate": 1e-06,
      "loss": 0.1631,
      "step": 2539
    },
    {
      "epoch": 0.8807212205270458,
      "grad_norm": 3.3438968658447266,
      "learning_rate": 1e-06,
      "loss": 0.1535,
      "step": 2540
    },
    {
      "epoch": 0.8810679611650486,
      "grad_norm": 3.1318681240081787,
      "learning_rate": 1e-06,
      "loss": 0.1806,
      "step": 2541
    },
    {
      "epoch": 0.8814147018030514,
      "grad_norm": 2.541802167892456,
      "learning_rate": 1e-06,
      "loss": 0.132,
      "step": 2542
    },
    {
      "epoch": 0.8817614424410541,
      "grad_norm": 3.502878427505493,
      "learning_rate": 1e-06,
      "loss": 0.127,
      "step": 2543
    },
    {
      "epoch": 0.8821081830790569,
      "grad_norm": 5.988008499145508,
      "learning_rate": 1e-06,
      "loss": 0.1334,
      "step": 2544
    },
    {
      "epoch": 0.8824549237170597,
      "grad_norm": 3.4589040279388428,
      "learning_rate": 1e-06,
      "loss": 0.1198,
      "step": 2545
    },
    {
      "epoch": 0.8828016643550625,
      "grad_norm": 4.258084774017334,
      "learning_rate": 1e-06,
      "loss": 0.1556,
      "step": 2546
    },
    {
      "epoch": 0.8831484049930652,
      "grad_norm": 2.9290645122528076,
      "learning_rate": 1e-06,
      "loss": 0.1383,
      "step": 2547
    },
    {
      "epoch": 0.883495145631068,
      "grad_norm": 5.24904727935791,
      "learning_rate": 1e-06,
      "loss": 0.1404,
      "step": 2548
    },
    {
      "epoch": 0.8838418862690708,
      "grad_norm": 2.883603572845459,
      "learning_rate": 1e-06,
      "loss": 0.1282,
      "step": 2549
    },
    {
      "epoch": 0.8841886269070736,
      "grad_norm": 2.8196918964385986,
      "learning_rate": 1e-06,
      "loss": 0.1478,
      "step": 2550
    },
    {
      "epoch": 0.8845353675450763,
      "grad_norm": 7.683722972869873,
      "learning_rate": 1e-06,
      "loss": 0.2069,
      "step": 2551
    },
    {
      "epoch": 0.8848821081830791,
      "grad_norm": 3.080634832382202,
      "learning_rate": 1e-06,
      "loss": 0.1451,
      "step": 2552
    },
    {
      "epoch": 0.8852288488210819,
      "grad_norm": 2.3532392978668213,
      "learning_rate": 1e-06,
      "loss": 0.1343,
      "step": 2553
    },
    {
      "epoch": 0.8855755894590847,
      "grad_norm": 2.798870325088501,
      "learning_rate": 1e-06,
      "loss": 0.1315,
      "step": 2554
    },
    {
      "epoch": 0.8859223300970874,
      "grad_norm": 2.4134509563446045,
      "learning_rate": 1e-06,
      "loss": 0.1507,
      "step": 2555
    },
    {
      "epoch": 0.8862690707350902,
      "grad_norm": 3.6513829231262207,
      "learning_rate": 1e-06,
      "loss": 0.1658,
      "step": 2556
    },
    {
      "epoch": 0.886615811373093,
      "grad_norm": 4.6206817626953125,
      "learning_rate": 1e-06,
      "loss": 0.1489,
      "step": 2557
    },
    {
      "epoch": 0.8869625520110958,
      "grad_norm": 4.091174125671387,
      "learning_rate": 1e-06,
      "loss": 0.152,
      "step": 2558
    },
    {
      "epoch": 0.8873092926490985,
      "grad_norm": 1.8325977325439453,
      "learning_rate": 1e-06,
      "loss": 0.1384,
      "step": 2559
    },
    {
      "epoch": 0.8876560332871013,
      "grad_norm": 2.203538656234741,
      "learning_rate": 1e-06,
      "loss": 0.1476,
      "step": 2560
    },
    {
      "epoch": 0.8880027739251041,
      "grad_norm": 3.418199300765991,
      "learning_rate": 1e-06,
      "loss": 0.11,
      "step": 2561
    },
    {
      "epoch": 0.8883495145631068,
      "grad_norm": 3.521099805831909,
      "learning_rate": 1e-06,
      "loss": 0.1088,
      "step": 2562
    },
    {
      "epoch": 0.8886962552011096,
      "grad_norm": 5.303309917449951,
      "learning_rate": 1e-06,
      "loss": 0.1504,
      "step": 2563
    },
    {
      "epoch": 0.8890429958391124,
      "grad_norm": 2.706254243850708,
      "learning_rate": 1e-06,
      "loss": 0.1522,
      "step": 2564
    },
    {
      "epoch": 0.8893897364771152,
      "grad_norm": 3.3216984272003174,
      "learning_rate": 1e-06,
      "loss": 0.1324,
      "step": 2565
    },
    {
      "epoch": 0.8897364771151179,
      "grad_norm": 3.598803997039795,
      "learning_rate": 1e-06,
      "loss": 0.1523,
      "step": 2566
    },
    {
      "epoch": 0.8900832177531207,
      "grad_norm": 2.6814308166503906,
      "learning_rate": 1e-06,
      "loss": 0.1539,
      "step": 2567
    },
    {
      "epoch": 0.8904299583911235,
      "grad_norm": 2.5854735374450684,
      "learning_rate": 1e-06,
      "loss": 0.1025,
      "step": 2568
    },
    {
      "epoch": 0.8907766990291263,
      "grad_norm": 2.8812289237976074,
      "learning_rate": 1e-06,
      "loss": 0.1313,
      "step": 2569
    },
    {
      "epoch": 0.891123439667129,
      "grad_norm": 3.4628384113311768,
      "learning_rate": 1e-06,
      "loss": 0.1494,
      "step": 2570
    },
    {
      "epoch": 0.8914701803051318,
      "grad_norm": 2.1708106994628906,
      "learning_rate": 1e-06,
      "loss": 0.1214,
      "step": 2571
    },
    {
      "epoch": 0.8918169209431346,
      "grad_norm": 2.2976863384246826,
      "learning_rate": 1e-06,
      "loss": 0.1184,
      "step": 2572
    },
    {
      "epoch": 0.8921636615811374,
      "grad_norm": 2.7640933990478516,
      "learning_rate": 1e-06,
      "loss": 0.1234,
      "step": 2573
    },
    {
      "epoch": 0.8925104022191401,
      "grad_norm": 3.4392740726470947,
      "learning_rate": 1e-06,
      "loss": 0.1495,
      "step": 2574
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 3.151865005493164,
      "learning_rate": 1e-06,
      "loss": 0.1204,
      "step": 2575
    },
    {
      "epoch": 0.8932038834951457,
      "grad_norm": 2.665188789367676,
      "learning_rate": 1e-06,
      "loss": 0.1312,
      "step": 2576
    },
    {
      "epoch": 0.8935506241331485,
      "grad_norm": 2.2830886840820312,
      "learning_rate": 1e-06,
      "loss": 0.1069,
      "step": 2577
    },
    {
      "epoch": 0.8938973647711512,
      "grad_norm": 3.063302755355835,
      "learning_rate": 1e-06,
      "loss": 0.1506,
      "step": 2578
    },
    {
      "epoch": 0.894244105409154,
      "grad_norm": 2.6693129539489746,
      "learning_rate": 1e-06,
      "loss": 0.1242,
      "step": 2579
    },
    {
      "epoch": 0.8945908460471568,
      "grad_norm": 3.1677896976470947,
      "learning_rate": 1e-06,
      "loss": 0.127,
      "step": 2580
    },
    {
      "epoch": 0.8949375866851595,
      "grad_norm": 3.562784194946289,
      "learning_rate": 1e-06,
      "loss": 0.1502,
      "step": 2581
    },
    {
      "epoch": 0.8952843273231623,
      "grad_norm": 2.2113330364227295,
      "learning_rate": 1e-06,
      "loss": 0.1146,
      "step": 2582
    },
    {
      "epoch": 0.8956310679611651,
      "grad_norm": 3.4925997257232666,
      "learning_rate": 1e-06,
      "loss": 0.133,
      "step": 2583
    },
    {
      "epoch": 0.8959778085991679,
      "grad_norm": 2.26218843460083,
      "learning_rate": 1e-06,
      "loss": 0.14,
      "step": 2584
    },
    {
      "epoch": 0.8963245492371706,
      "grad_norm": 2.437492847442627,
      "learning_rate": 1e-06,
      "loss": 0.1292,
      "step": 2585
    },
    {
      "epoch": 0.8966712898751734,
      "grad_norm": 3.4927961826324463,
      "learning_rate": 1e-06,
      "loss": 0.1195,
      "step": 2586
    },
    {
      "epoch": 0.8970180305131762,
      "grad_norm": 4.102070331573486,
      "learning_rate": 1e-06,
      "loss": 0.1356,
      "step": 2587
    },
    {
      "epoch": 0.897364771151179,
      "grad_norm": 5.60887336730957,
      "learning_rate": 1e-06,
      "loss": 0.1479,
      "step": 2588
    },
    {
      "epoch": 0.8977115117891817,
      "grad_norm": 2.31508731842041,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 2589
    },
    {
      "epoch": 0.8980582524271845,
      "grad_norm": 2.7380573749542236,
      "learning_rate": 1e-06,
      "loss": 0.1463,
      "step": 2590
    },
    {
      "epoch": 0.8984049930651873,
      "grad_norm": 1.9896982908248901,
      "learning_rate": 1e-06,
      "loss": 0.1162,
      "step": 2591
    },
    {
      "epoch": 0.8987517337031901,
      "grad_norm": 2.7056241035461426,
      "learning_rate": 1e-06,
      "loss": 0.1591,
      "step": 2592
    },
    {
      "epoch": 0.8990984743411928,
      "grad_norm": 1.9622806310653687,
      "learning_rate": 1e-06,
      "loss": 0.1138,
      "step": 2593
    },
    {
      "epoch": 0.8994452149791956,
      "grad_norm": 3.957826614379883,
      "learning_rate": 1e-06,
      "loss": 0.1561,
      "step": 2594
    },
    {
      "epoch": 0.8997919556171984,
      "grad_norm": 2.8110544681549072,
      "learning_rate": 1e-06,
      "loss": 0.1379,
      "step": 2595
    },
    {
      "epoch": 0.9001386962552012,
      "grad_norm": 3.7810873985290527,
      "learning_rate": 1e-06,
      "loss": 0.1533,
      "step": 2596
    },
    {
      "epoch": 0.9004854368932039,
      "grad_norm": 2.166959762573242,
      "learning_rate": 1e-06,
      "loss": 0.1411,
      "step": 2597
    },
    {
      "epoch": 0.9008321775312067,
      "grad_norm": 2.021667718887329,
      "learning_rate": 1e-06,
      "loss": 0.1501,
      "step": 2598
    },
    {
      "epoch": 0.9011789181692095,
      "grad_norm": 3.390395402908325,
      "learning_rate": 1e-06,
      "loss": 0.1468,
      "step": 2599
    },
    {
      "epoch": 0.9015256588072122,
      "grad_norm": 2.1694798469543457,
      "learning_rate": 1e-06,
      "loss": 0.1283,
      "step": 2600
    },
    {
      "epoch": 0.901872399445215,
      "grad_norm": 3.0354669094085693,
      "learning_rate": 1e-06,
      "loss": 0.1327,
      "step": 2601
    },
    {
      "epoch": 0.9022191400832178,
      "grad_norm": 3.5150482654571533,
      "learning_rate": 1e-06,
      "loss": 0.1492,
      "step": 2602
    },
    {
      "epoch": 0.9025658807212206,
      "grad_norm": 2.1176624298095703,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 2603
    },
    {
      "epoch": 0.9029126213592233,
      "grad_norm": 2.6938793659210205,
      "learning_rate": 1e-06,
      "loss": 0.1662,
      "step": 2604
    },
    {
      "epoch": 0.9032593619972261,
      "grad_norm": 3.2767770290374756,
      "learning_rate": 1e-06,
      "loss": 0.1132,
      "step": 2605
    },
    {
      "epoch": 0.9036061026352289,
      "grad_norm": 3.4244790077209473,
      "learning_rate": 1e-06,
      "loss": 0.152,
      "step": 2606
    },
    {
      "epoch": 0.9039528432732317,
      "grad_norm": 2.3270678520202637,
      "learning_rate": 1e-06,
      "loss": 0.164,
      "step": 2607
    },
    {
      "epoch": 0.9042995839112344,
      "grad_norm": 2.2908077239990234,
      "learning_rate": 1e-06,
      "loss": 0.1561,
      "step": 2608
    },
    {
      "epoch": 0.9046463245492372,
      "grad_norm": 1.976806879043579,
      "learning_rate": 1e-06,
      "loss": 0.1219,
      "step": 2609
    },
    {
      "epoch": 0.90499306518724,
      "grad_norm": 2.578418731689453,
      "learning_rate": 1e-06,
      "loss": 0.1454,
      "step": 2610
    },
    {
      "epoch": 0.9053398058252428,
      "grad_norm": 2.3116960525512695,
      "learning_rate": 1e-06,
      "loss": 0.1499,
      "step": 2611
    },
    {
      "epoch": 0.9056865464632455,
      "grad_norm": 3.2435667514801025,
      "learning_rate": 1e-06,
      "loss": 0.1213,
      "step": 2612
    },
    {
      "epoch": 0.9060332871012483,
      "grad_norm": 1.842995524406433,
      "learning_rate": 1e-06,
      "loss": 0.1074,
      "step": 2613
    },
    {
      "epoch": 0.9063800277392511,
      "grad_norm": 2.7041714191436768,
      "learning_rate": 1e-06,
      "loss": 0.1504,
      "step": 2614
    },
    {
      "epoch": 0.9067267683772539,
      "grad_norm": 3.233328104019165,
      "learning_rate": 1e-06,
      "loss": 0.1284,
      "step": 2615
    },
    {
      "epoch": 0.9070735090152566,
      "grad_norm": 2.0158610343933105,
      "learning_rate": 1e-06,
      "loss": 0.1097,
      "step": 2616
    },
    {
      "epoch": 0.9074202496532594,
      "grad_norm": 2.651226043701172,
      "learning_rate": 1e-06,
      "loss": 0.1276,
      "step": 2617
    },
    {
      "epoch": 0.9077669902912622,
      "grad_norm": 3.511500835418701,
      "learning_rate": 1e-06,
      "loss": 0.1523,
      "step": 2618
    },
    {
      "epoch": 0.908113730929265,
      "grad_norm": 2.868360996246338,
      "learning_rate": 1e-06,
      "loss": 0.1482,
      "step": 2619
    },
    {
      "epoch": 0.9084604715672677,
      "grad_norm": 5.283854961395264,
      "learning_rate": 1e-06,
      "loss": 0.1529,
      "step": 2620
    },
    {
      "epoch": 0.9088072122052705,
      "grad_norm": 2.312080144882202,
      "learning_rate": 1e-06,
      "loss": 0.1197,
      "step": 2621
    },
    {
      "epoch": 0.9091539528432733,
      "grad_norm": 2.4903831481933594,
      "learning_rate": 1e-06,
      "loss": 0.1488,
      "step": 2622
    },
    {
      "epoch": 0.909500693481276,
      "grad_norm": 3.8020405769348145,
      "learning_rate": 1e-06,
      "loss": 0.1167,
      "step": 2623
    },
    {
      "epoch": 0.9098474341192788,
      "grad_norm": 4.8802666664123535,
      "learning_rate": 1e-06,
      "loss": 0.1598,
      "step": 2624
    },
    {
      "epoch": 0.9101941747572816,
      "grad_norm": 2.2983484268188477,
      "learning_rate": 1e-06,
      "loss": 0.1255,
      "step": 2625
    },
    {
      "epoch": 0.9105409153952844,
      "grad_norm": 2.7895936965942383,
      "learning_rate": 1e-06,
      "loss": 0.1464,
      "step": 2626
    },
    {
      "epoch": 0.9108876560332871,
      "grad_norm": 3.0631260871887207,
      "learning_rate": 1e-06,
      "loss": 0.1165,
      "step": 2627
    },
    {
      "epoch": 0.9112343966712899,
      "grad_norm": 5.066497802734375,
      "learning_rate": 1e-06,
      "loss": 0.1336,
      "step": 2628
    },
    {
      "epoch": 0.9115811373092927,
      "grad_norm": 2.471651792526245,
      "learning_rate": 1e-06,
      "loss": 0.1418,
      "step": 2629
    },
    {
      "epoch": 0.9119278779472955,
      "grad_norm": 2.3465819358825684,
      "learning_rate": 1e-06,
      "loss": 0.1309,
      "step": 2630
    },
    {
      "epoch": 0.9122746185852982,
      "grad_norm": 2.3085386753082275,
      "learning_rate": 1e-06,
      "loss": 0.1099,
      "step": 2631
    },
    {
      "epoch": 0.912621359223301,
      "grad_norm": 6.42535400390625,
      "learning_rate": 1e-06,
      "loss": 0.1657,
      "step": 2632
    },
    {
      "epoch": 0.9129680998613038,
      "grad_norm": 2.8673324584960938,
      "learning_rate": 1e-06,
      "loss": 0.1509,
      "step": 2633
    },
    {
      "epoch": 0.9133148404993066,
      "grad_norm": 2.016261339187622,
      "learning_rate": 1e-06,
      "loss": 0.1464,
      "step": 2634
    },
    {
      "epoch": 0.9136615811373093,
      "grad_norm": 2.854090690612793,
      "learning_rate": 1e-06,
      "loss": 0.1341,
      "step": 2635
    },
    {
      "epoch": 0.9140083217753121,
      "grad_norm": 2.522303342819214,
      "learning_rate": 1e-06,
      "loss": 0.1272,
      "step": 2636
    },
    {
      "epoch": 0.9143550624133149,
      "grad_norm": 2.4291818141937256,
      "learning_rate": 1e-06,
      "loss": 0.1476,
      "step": 2637
    },
    {
      "epoch": 0.9147018030513177,
      "grad_norm": 1.8062353134155273,
      "learning_rate": 1e-06,
      "loss": 0.1041,
      "step": 2638
    },
    {
      "epoch": 0.9150485436893204,
      "grad_norm": 2.9814038276672363,
      "learning_rate": 1e-06,
      "loss": 0.1324,
      "step": 2639
    },
    {
      "epoch": 0.9153952843273232,
      "grad_norm": 2.3664748668670654,
      "learning_rate": 1e-06,
      "loss": 0.1533,
      "step": 2640
    },
    {
      "epoch": 0.915742024965326,
      "grad_norm": 2.347658634185791,
      "learning_rate": 1e-06,
      "loss": 0.1217,
      "step": 2641
    },
    {
      "epoch": 0.9160887656033287,
      "grad_norm": 3.068976640701294,
      "learning_rate": 1e-06,
      "loss": 0.1383,
      "step": 2642
    },
    {
      "epoch": 0.9164355062413315,
      "grad_norm": 2.92535400390625,
      "learning_rate": 1e-06,
      "loss": 0.1459,
      "step": 2643
    },
    {
      "epoch": 0.9167822468793343,
      "grad_norm": 2.56888747215271,
      "learning_rate": 1e-06,
      "loss": 0.1308,
      "step": 2644
    },
    {
      "epoch": 0.9171289875173371,
      "grad_norm": 2.389007806777954,
      "learning_rate": 1e-06,
      "loss": 0.1628,
      "step": 2645
    },
    {
      "epoch": 0.9174757281553398,
      "grad_norm": 2.4458770751953125,
      "learning_rate": 1e-06,
      "loss": 0.1135,
      "step": 2646
    },
    {
      "epoch": 0.9178224687933426,
      "grad_norm": 2.0837292671203613,
      "learning_rate": 1e-06,
      "loss": 0.1154,
      "step": 2647
    },
    {
      "epoch": 0.9181692094313454,
      "grad_norm": 5.257343769073486,
      "learning_rate": 1e-06,
      "loss": 0.1318,
      "step": 2648
    },
    {
      "epoch": 0.9185159500693482,
      "grad_norm": 4.321478366851807,
      "learning_rate": 1e-06,
      "loss": 0.1418,
      "step": 2649
    },
    {
      "epoch": 0.9188626907073509,
      "grad_norm": 2.9317991733551025,
      "learning_rate": 1e-06,
      "loss": 0.1491,
      "step": 2650
    },
    {
      "epoch": 0.9192094313453537,
      "grad_norm": 3.1797657012939453,
      "learning_rate": 1e-06,
      "loss": 0.1566,
      "step": 2651
    },
    {
      "epoch": 0.9195561719833565,
      "grad_norm": 2.146343231201172,
      "learning_rate": 1e-06,
      "loss": 0.1431,
      "step": 2652
    },
    {
      "epoch": 0.9199029126213593,
      "grad_norm": 2.9985647201538086,
      "learning_rate": 1e-06,
      "loss": 0.1404,
      "step": 2653
    },
    {
      "epoch": 0.920249653259362,
      "grad_norm": 1.452409267425537,
      "learning_rate": 1e-06,
      "loss": 0.0951,
      "step": 2654
    },
    {
      "epoch": 0.9205963938973648,
      "grad_norm": 3.1620914936065674,
      "learning_rate": 1e-06,
      "loss": 0.1582,
      "step": 2655
    },
    {
      "epoch": 0.9209431345353676,
      "grad_norm": 4.072134971618652,
      "learning_rate": 1e-06,
      "loss": 0.1415,
      "step": 2656
    },
    {
      "epoch": 0.9212898751733704,
      "grad_norm": 2.393535852432251,
      "learning_rate": 1e-06,
      "loss": 0.1214,
      "step": 2657
    },
    {
      "epoch": 0.9216366158113731,
      "grad_norm": 1.983525037765503,
      "learning_rate": 1e-06,
      "loss": 0.1196,
      "step": 2658
    },
    {
      "epoch": 0.9219833564493759,
      "grad_norm": 2.7609565258026123,
      "learning_rate": 1e-06,
      "loss": 0.1005,
      "step": 2659
    },
    {
      "epoch": 0.9223300970873787,
      "grad_norm": 2.900486707687378,
      "learning_rate": 1e-06,
      "loss": 0.1112,
      "step": 2660
    },
    {
      "epoch": 0.9226768377253814,
      "grad_norm": 2.1307523250579834,
      "learning_rate": 1e-06,
      "loss": 0.114,
      "step": 2661
    },
    {
      "epoch": 0.9230235783633842,
      "grad_norm": 3.795403242111206,
      "learning_rate": 1e-06,
      "loss": 0.1341,
      "step": 2662
    },
    {
      "epoch": 0.923370319001387,
      "grad_norm": 2.027860403060913,
      "learning_rate": 1e-06,
      "loss": 0.1359,
      "step": 2663
    },
    {
      "epoch": 0.9237170596393898,
      "grad_norm": 1.9218776226043701,
      "learning_rate": 1e-06,
      "loss": 0.1206,
      "step": 2664
    },
    {
      "epoch": 0.9240638002773925,
      "grad_norm": 2.235786199569702,
      "learning_rate": 1e-06,
      "loss": 0.1399,
      "step": 2665
    },
    {
      "epoch": 0.9244105409153953,
      "grad_norm": 2.2658088207244873,
      "learning_rate": 1e-06,
      "loss": 0.1469,
      "step": 2666
    },
    {
      "epoch": 0.9247572815533981,
      "grad_norm": 2.2411727905273438,
      "learning_rate": 1e-06,
      "loss": 0.1334,
      "step": 2667
    },
    {
      "epoch": 0.9251040221914009,
      "grad_norm": 2.2818808555603027,
      "learning_rate": 1e-06,
      "loss": 0.1323,
      "step": 2668
    },
    {
      "epoch": 0.9254507628294036,
      "grad_norm": 2.430490255355835,
      "learning_rate": 1e-06,
      "loss": 0.129,
      "step": 2669
    },
    {
      "epoch": 0.9257975034674064,
      "grad_norm": 2.8982584476470947,
      "learning_rate": 1e-06,
      "loss": 0.1231,
      "step": 2670
    },
    {
      "epoch": 0.9261442441054092,
      "grad_norm": 3.195540428161621,
      "learning_rate": 1e-06,
      "loss": 0.1309,
      "step": 2671
    },
    {
      "epoch": 0.926490984743412,
      "grad_norm": 2.102155923843384,
      "learning_rate": 1e-06,
      "loss": 0.1392,
      "step": 2672
    },
    {
      "epoch": 0.9268377253814147,
      "grad_norm": 2.8941521644592285,
      "learning_rate": 1e-06,
      "loss": 0.1402,
      "step": 2673
    },
    {
      "epoch": 0.9271844660194175,
      "grad_norm": 1.982313632965088,
      "learning_rate": 1e-06,
      "loss": 0.1215,
      "step": 2674
    },
    {
      "epoch": 0.9275312066574203,
      "grad_norm": 3.4784913063049316,
      "learning_rate": 1e-06,
      "loss": 0.1597,
      "step": 2675
    },
    {
      "epoch": 0.9278779472954231,
      "grad_norm": 2.5161221027374268,
      "learning_rate": 1e-06,
      "loss": 0.1721,
      "step": 2676
    },
    {
      "epoch": 0.9282246879334258,
      "grad_norm": 2.5937342643737793,
      "learning_rate": 1e-06,
      "loss": 0.1466,
      "step": 2677
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 2.7261319160461426,
      "learning_rate": 1e-06,
      "loss": 0.1446,
      "step": 2678
    },
    {
      "epoch": 0.9289181692094314,
      "grad_norm": 2.406301498413086,
      "learning_rate": 1e-06,
      "loss": 0.143,
      "step": 2679
    },
    {
      "epoch": 0.9292649098474342,
      "grad_norm": 3.284876823425293,
      "learning_rate": 1e-06,
      "loss": 0.1859,
      "step": 2680
    },
    {
      "epoch": 0.9296116504854369,
      "grad_norm": 2.8308169841766357,
      "learning_rate": 1e-06,
      "loss": 0.1483,
      "step": 2681
    },
    {
      "epoch": 0.9299583911234397,
      "grad_norm": 2.309736490249634,
      "learning_rate": 1e-06,
      "loss": 0.1341,
      "step": 2682
    },
    {
      "epoch": 0.9303051317614425,
      "grad_norm": 2.3416709899902344,
      "learning_rate": 1e-06,
      "loss": 0.09,
      "step": 2683
    },
    {
      "epoch": 0.9306518723994452,
      "grad_norm": 2.0100162029266357,
      "learning_rate": 1e-06,
      "loss": 0.1185,
      "step": 2684
    },
    {
      "epoch": 0.930998613037448,
      "grad_norm": 2.5902180671691895,
      "learning_rate": 1e-06,
      "loss": 0.1393,
      "step": 2685
    },
    {
      "epoch": 0.9313453536754508,
      "grad_norm": 4.1081767082214355,
      "learning_rate": 1e-06,
      "loss": 0.1468,
      "step": 2686
    },
    {
      "epoch": 0.9316920943134536,
      "grad_norm": 2.6473231315612793,
      "learning_rate": 1e-06,
      "loss": 0.1258,
      "step": 2687
    },
    {
      "epoch": 0.9320388349514563,
      "grad_norm": 2.6274592876434326,
      "learning_rate": 1e-06,
      "loss": 0.1216,
      "step": 2688
    },
    {
      "epoch": 0.9323855755894591,
      "grad_norm": 2.321573495864868,
      "learning_rate": 1e-06,
      "loss": 0.1156,
      "step": 2689
    },
    {
      "epoch": 0.9327323162274619,
      "grad_norm": 2.83542799949646,
      "learning_rate": 1e-06,
      "loss": 0.113,
      "step": 2690
    },
    {
      "epoch": 0.9330790568654647,
      "grad_norm": 3.1709511280059814,
      "learning_rate": 1e-06,
      "loss": 0.1644,
      "step": 2691
    },
    {
      "epoch": 0.9334257975034674,
      "grad_norm": 1.8596209287643433,
      "learning_rate": 1e-06,
      "loss": 0.126,
      "step": 2692
    },
    {
      "epoch": 0.9337725381414702,
      "grad_norm": 2.7414655685424805,
      "learning_rate": 1e-06,
      "loss": 0.1305,
      "step": 2693
    },
    {
      "epoch": 0.934119278779473,
      "grad_norm": 3.0172135829925537,
      "learning_rate": 1e-06,
      "loss": 0.1463,
      "step": 2694
    },
    {
      "epoch": 0.9344660194174758,
      "grad_norm": 2.8152413368225098,
      "learning_rate": 1e-06,
      "loss": 0.1396,
      "step": 2695
    },
    {
      "epoch": 0.9348127600554785,
      "grad_norm": 2.0890872478485107,
      "learning_rate": 1e-06,
      "loss": 0.1034,
      "step": 2696
    },
    {
      "epoch": 0.9351595006934813,
      "grad_norm": 2.5807907581329346,
      "learning_rate": 1e-06,
      "loss": 0.1488,
      "step": 2697
    },
    {
      "epoch": 0.9355062413314841,
      "grad_norm": 2.900175094604492,
      "learning_rate": 1e-06,
      "loss": 0.1427,
      "step": 2698
    },
    {
      "epoch": 0.9358529819694869,
      "grad_norm": 2.0167062282562256,
      "learning_rate": 1e-06,
      "loss": 0.1071,
      "step": 2699
    },
    {
      "epoch": 0.9361997226074896,
      "grad_norm": 2.26310658454895,
      "learning_rate": 1e-06,
      "loss": 0.1393,
      "step": 2700
    },
    {
      "epoch": 0.9365464632454924,
      "grad_norm": 2.191679000854492,
      "learning_rate": 1e-06,
      "loss": 0.1299,
      "step": 2701
    },
    {
      "epoch": 0.9368932038834952,
      "grad_norm": 2.8167102336883545,
      "learning_rate": 1e-06,
      "loss": 0.1317,
      "step": 2702
    },
    {
      "epoch": 0.937239944521498,
      "grad_norm": 2.700887441635132,
      "learning_rate": 1e-06,
      "loss": 0.1726,
      "step": 2703
    },
    {
      "epoch": 0.9375866851595007,
      "grad_norm": 2.511220932006836,
      "learning_rate": 1e-06,
      "loss": 0.1579,
      "step": 2704
    },
    {
      "epoch": 0.9379334257975035,
      "grad_norm": 3.294839859008789,
      "learning_rate": 1e-06,
      "loss": 0.142,
      "step": 2705
    },
    {
      "epoch": 0.9382801664355063,
      "grad_norm": 2.9182393550872803,
      "learning_rate": 1e-06,
      "loss": 0.1242,
      "step": 2706
    },
    {
      "epoch": 0.938626907073509,
      "grad_norm": 1.9845600128173828,
      "learning_rate": 1e-06,
      "loss": 0.1103,
      "step": 2707
    },
    {
      "epoch": 0.9389736477115118,
      "grad_norm": 3.3602540493011475,
      "learning_rate": 1e-06,
      "loss": 0.1374,
      "step": 2708
    },
    {
      "epoch": 0.9393203883495146,
      "grad_norm": 2.03686261177063,
      "learning_rate": 1e-06,
      "loss": 0.1203,
      "step": 2709
    },
    {
      "epoch": 0.9396671289875174,
      "grad_norm": 2.6273975372314453,
      "learning_rate": 1e-06,
      "loss": 0.1296,
      "step": 2710
    },
    {
      "epoch": 0.9400138696255201,
      "grad_norm": 2.5899906158447266,
      "learning_rate": 1e-06,
      "loss": 0.1411,
      "step": 2711
    },
    {
      "epoch": 0.9403606102635229,
      "grad_norm": 2.5806844234466553,
      "learning_rate": 1e-06,
      "loss": 0.1371,
      "step": 2712
    },
    {
      "epoch": 0.9407073509015257,
      "grad_norm": 2.3415279388427734,
      "learning_rate": 1e-06,
      "loss": 0.1231,
      "step": 2713
    },
    {
      "epoch": 0.9410540915395285,
      "grad_norm": 3.5045316219329834,
      "learning_rate": 1e-06,
      "loss": 0.1382,
      "step": 2714
    },
    {
      "epoch": 0.9414008321775312,
      "grad_norm": 2.0843143463134766,
      "learning_rate": 1e-06,
      "loss": 0.1078,
      "step": 2715
    },
    {
      "epoch": 0.941747572815534,
      "grad_norm": 2.7572216987609863,
      "learning_rate": 1e-06,
      "loss": 0.1407,
      "step": 2716
    },
    {
      "epoch": 0.9420943134535368,
      "grad_norm": 2.4735355377197266,
      "learning_rate": 1e-06,
      "loss": 0.1305,
      "step": 2717
    },
    {
      "epoch": 0.9424410540915396,
      "grad_norm": 2.1515815258026123,
      "learning_rate": 1e-06,
      "loss": 0.1231,
      "step": 2718
    },
    {
      "epoch": 0.9427877947295423,
      "grad_norm": 3.494328737258911,
      "learning_rate": 1e-06,
      "loss": 0.1017,
      "step": 2719
    },
    {
      "epoch": 0.9431345353675451,
      "grad_norm": 2.391047239303589,
      "learning_rate": 1e-06,
      "loss": 0.125,
      "step": 2720
    },
    {
      "epoch": 0.9434812760055479,
      "grad_norm": 1.7509092092514038,
      "learning_rate": 1e-06,
      "loss": 0.1099,
      "step": 2721
    },
    {
      "epoch": 0.9438280166435506,
      "grad_norm": 2.7417235374450684,
      "learning_rate": 1e-06,
      "loss": 0.1167,
      "step": 2722
    },
    {
      "epoch": 0.9441747572815534,
      "grad_norm": 2.740318536758423,
      "learning_rate": 1e-06,
      "loss": 0.1291,
      "step": 2723
    },
    {
      "epoch": 0.9445214979195562,
      "grad_norm": 3.362915515899658,
      "learning_rate": 1e-06,
      "loss": 0.1361,
      "step": 2724
    },
    {
      "epoch": 0.944868238557559,
      "grad_norm": 2.27724552154541,
      "learning_rate": 1e-06,
      "loss": 0.1218,
      "step": 2725
    },
    {
      "epoch": 0.9452149791955617,
      "grad_norm": 2.4609215259552,
      "learning_rate": 1e-06,
      "loss": 0.1397,
      "step": 2726
    },
    {
      "epoch": 0.9455617198335645,
      "grad_norm": 3.4139716625213623,
      "learning_rate": 1e-06,
      "loss": 0.1238,
      "step": 2727
    },
    {
      "epoch": 0.9459084604715673,
      "grad_norm": 2.5708959102630615,
      "learning_rate": 1e-06,
      "loss": 0.1173,
      "step": 2728
    },
    {
      "epoch": 0.9462552011095701,
      "grad_norm": 2.1245861053466797,
      "learning_rate": 1e-06,
      "loss": 0.1325,
      "step": 2729
    },
    {
      "epoch": 0.9466019417475728,
      "grad_norm": 3.1224300861358643,
      "learning_rate": 1e-06,
      "loss": 0.1674,
      "step": 2730
    },
    {
      "epoch": 0.9469486823855756,
      "grad_norm": 2.9298243522644043,
      "learning_rate": 1e-06,
      "loss": 0.1628,
      "step": 2731
    },
    {
      "epoch": 0.9472954230235784,
      "grad_norm": 2.5283334255218506,
      "learning_rate": 1e-06,
      "loss": 0.1238,
      "step": 2732
    },
    {
      "epoch": 0.9476421636615812,
      "grad_norm": 1.95565927028656,
      "learning_rate": 1e-06,
      "loss": 0.1154,
      "step": 2733
    },
    {
      "epoch": 0.9479889042995839,
      "grad_norm": 2.800057888031006,
      "learning_rate": 1e-06,
      "loss": 0.1088,
      "step": 2734
    },
    {
      "epoch": 0.9483356449375867,
      "grad_norm": 2.9999020099639893,
      "learning_rate": 1e-06,
      "loss": 0.1263,
      "step": 2735
    },
    {
      "epoch": 0.9486823855755895,
      "grad_norm": 2.755236864089966,
      "learning_rate": 1e-06,
      "loss": 0.1282,
      "step": 2736
    },
    {
      "epoch": 0.9490291262135923,
      "grad_norm": 2.161933422088623,
      "learning_rate": 1e-06,
      "loss": 0.1441,
      "step": 2737
    },
    {
      "epoch": 0.949375866851595,
      "grad_norm": 2.8767547607421875,
      "learning_rate": 1e-06,
      "loss": 0.1047,
      "step": 2738
    },
    {
      "epoch": 0.9497226074895978,
      "grad_norm": 2.172853946685791,
      "learning_rate": 1e-06,
      "loss": 0.1201,
      "step": 2739
    },
    {
      "epoch": 0.9500693481276006,
      "grad_norm": 2.792728900909424,
      "learning_rate": 1e-06,
      "loss": 0.1136,
      "step": 2740
    },
    {
      "epoch": 0.9504160887656034,
      "grad_norm": 2.2738940715789795,
      "learning_rate": 1e-06,
      "loss": 0.0994,
      "step": 2741
    },
    {
      "epoch": 0.9507628294036061,
      "grad_norm": 3.087293863296509,
      "learning_rate": 1e-06,
      "loss": 0.1417,
      "step": 2742
    },
    {
      "epoch": 0.9511095700416089,
      "grad_norm": 2.7275638580322266,
      "learning_rate": 1e-06,
      "loss": 0.1498,
      "step": 2743
    },
    {
      "epoch": 0.9514563106796117,
      "grad_norm": 3.0619494915008545,
      "learning_rate": 1e-06,
      "loss": 0.1445,
      "step": 2744
    },
    {
      "epoch": 0.9518030513176144,
      "grad_norm": 2.5986380577087402,
      "learning_rate": 1e-06,
      "loss": 0.1363,
      "step": 2745
    },
    {
      "epoch": 0.9521497919556172,
      "grad_norm": 3.417404890060425,
      "learning_rate": 1e-06,
      "loss": 0.1448,
      "step": 2746
    },
    {
      "epoch": 0.95249653259362,
      "grad_norm": 4.566203594207764,
      "learning_rate": 1e-06,
      "loss": 0.1653,
      "step": 2747
    },
    {
      "epoch": 0.9528432732316228,
      "grad_norm": 3.027130126953125,
      "learning_rate": 1e-06,
      "loss": 0.1543,
      "step": 2748
    },
    {
      "epoch": 0.9531900138696255,
      "grad_norm": 3.81862211227417,
      "learning_rate": 1e-06,
      "loss": 0.1651,
      "step": 2749
    },
    {
      "epoch": 0.9535367545076283,
      "grad_norm": 2.1442902088165283,
      "learning_rate": 1e-06,
      "loss": 0.1094,
      "step": 2750
    },
    {
      "epoch": 0.9538834951456311,
      "grad_norm": 2.5918867588043213,
      "learning_rate": 1e-06,
      "loss": 0.1348,
      "step": 2751
    },
    {
      "epoch": 0.9542302357836339,
      "grad_norm": 2.379676103591919,
      "learning_rate": 1e-06,
      "loss": 0.1376,
      "step": 2752
    },
    {
      "epoch": 0.9545769764216366,
      "grad_norm": 2.339801549911499,
      "learning_rate": 1e-06,
      "loss": 0.1239,
      "step": 2753
    },
    {
      "epoch": 0.9549237170596394,
      "grad_norm": 3.523301839828491,
      "learning_rate": 1e-06,
      "loss": 0.1241,
      "step": 2754
    },
    {
      "epoch": 0.9552704576976422,
      "grad_norm": 5.052760124206543,
      "learning_rate": 1e-06,
      "loss": 0.132,
      "step": 2755
    },
    {
      "epoch": 0.955617198335645,
      "grad_norm": 2.946012020111084,
      "learning_rate": 1e-06,
      "loss": 0.1742,
      "step": 2756
    },
    {
      "epoch": 0.9559639389736477,
      "grad_norm": 3.70878529548645,
      "learning_rate": 1e-06,
      "loss": 0.1526,
      "step": 2757
    },
    {
      "epoch": 0.9563106796116505,
      "grad_norm": 2.7530357837677,
      "learning_rate": 1e-06,
      "loss": 0.1757,
      "step": 2758
    },
    {
      "epoch": 0.9566574202496533,
      "grad_norm": 5.642550945281982,
      "learning_rate": 1e-06,
      "loss": 0.1614,
      "step": 2759
    },
    {
      "epoch": 0.957004160887656,
      "grad_norm": 3.5328824520111084,
      "learning_rate": 1e-06,
      "loss": 0.1212,
      "step": 2760
    },
    {
      "epoch": 0.9573509015256588,
      "grad_norm": 2.6719205379486084,
      "learning_rate": 1e-06,
      "loss": 0.136,
      "step": 2761
    },
    {
      "epoch": 0.9576976421636616,
      "grad_norm": 3.643240213394165,
      "learning_rate": 1e-06,
      "loss": 0.1279,
      "step": 2762
    },
    {
      "epoch": 0.9580443828016644,
      "grad_norm": 3.318526029586792,
      "learning_rate": 1e-06,
      "loss": 0.1327,
      "step": 2763
    },
    {
      "epoch": 0.9583911234396671,
      "grad_norm": 2.845430850982666,
      "learning_rate": 1e-06,
      "loss": 0.1291,
      "step": 2764
    },
    {
      "epoch": 0.9587378640776699,
      "grad_norm": 2.6309032440185547,
      "learning_rate": 1e-06,
      "loss": 0.1295,
      "step": 2765
    },
    {
      "epoch": 0.9590846047156727,
      "grad_norm": 2.3172414302825928,
      "learning_rate": 1e-06,
      "loss": 0.1494,
      "step": 2766
    },
    {
      "epoch": 0.9594313453536755,
      "grad_norm": 2.4579062461853027,
      "learning_rate": 1e-06,
      "loss": 0.1252,
      "step": 2767
    },
    {
      "epoch": 0.9597780859916782,
      "grad_norm": 2.277562379837036,
      "learning_rate": 1e-06,
      "loss": 0.127,
      "step": 2768
    },
    {
      "epoch": 0.960124826629681,
      "grad_norm": 2.561744451522827,
      "learning_rate": 1e-06,
      "loss": 0.1183,
      "step": 2769
    },
    {
      "epoch": 0.9604715672676838,
      "grad_norm": 2.368710517883301,
      "learning_rate": 1e-06,
      "loss": 0.1209,
      "step": 2770
    },
    {
      "epoch": 0.9608183079056866,
      "grad_norm": 2.4350576400756836,
      "learning_rate": 1e-06,
      "loss": 0.1615,
      "step": 2771
    },
    {
      "epoch": 0.9611650485436893,
      "grad_norm": 2.933718204498291,
      "learning_rate": 1e-06,
      "loss": 0.1183,
      "step": 2772
    },
    {
      "epoch": 0.9615117891816921,
      "grad_norm": 2.6752936840057373,
      "learning_rate": 1e-06,
      "loss": 0.1351,
      "step": 2773
    },
    {
      "epoch": 0.9618585298196949,
      "grad_norm": 2.8626444339752197,
      "learning_rate": 1e-06,
      "loss": 0.1445,
      "step": 2774
    },
    {
      "epoch": 0.9622052704576977,
      "grad_norm": 2.7919156551361084,
      "learning_rate": 1e-06,
      "loss": 0.1015,
      "step": 2775
    },
    {
      "epoch": 0.9625520110957004,
      "grad_norm": 2.687023878097534,
      "learning_rate": 1e-06,
      "loss": 0.1398,
      "step": 2776
    },
    {
      "epoch": 0.9628987517337032,
      "grad_norm": 3.4772815704345703,
      "learning_rate": 1e-06,
      "loss": 0.1276,
      "step": 2777
    },
    {
      "epoch": 0.963245492371706,
      "grad_norm": 3.7592716217041016,
      "learning_rate": 1e-06,
      "loss": 0.1196,
      "step": 2778
    },
    {
      "epoch": 0.9635922330097088,
      "grad_norm": 2.2832236289978027,
      "learning_rate": 1e-06,
      "loss": 0.1149,
      "step": 2779
    },
    {
      "epoch": 0.9639389736477115,
      "grad_norm": 2.2533531188964844,
      "learning_rate": 1e-06,
      "loss": 0.1361,
      "step": 2780
    },
    {
      "epoch": 0.9642857142857143,
      "grad_norm": 2.458859920501709,
      "learning_rate": 1e-06,
      "loss": 0.1248,
      "step": 2781
    },
    {
      "epoch": 0.9646324549237171,
      "grad_norm": 2.3881120681762695,
      "learning_rate": 1e-06,
      "loss": 0.1365,
      "step": 2782
    },
    {
      "epoch": 0.9649791955617198,
      "grad_norm": 2.8706626892089844,
      "learning_rate": 1e-06,
      "loss": 0.1324,
      "step": 2783
    },
    {
      "epoch": 0.9653259361997226,
      "grad_norm": 2.1797640323638916,
      "learning_rate": 1e-06,
      "loss": 0.1178,
      "step": 2784
    },
    {
      "epoch": 0.9656726768377254,
      "grad_norm": 3.3081295490264893,
      "learning_rate": 1e-06,
      "loss": 0.1058,
      "step": 2785
    },
    {
      "epoch": 0.9660194174757282,
      "grad_norm": 4.193579196929932,
      "learning_rate": 1e-06,
      "loss": 0.1522,
      "step": 2786
    },
    {
      "epoch": 0.9663661581137309,
      "grad_norm": 2.6109237670898438,
      "learning_rate": 1e-06,
      "loss": 0.1091,
      "step": 2787
    },
    {
      "epoch": 0.9667128987517337,
      "grad_norm": 2.3282434940338135,
      "learning_rate": 1e-06,
      "loss": 0.1333,
      "step": 2788
    },
    {
      "epoch": 0.9670596393897365,
      "grad_norm": 2.578936815261841,
      "learning_rate": 1e-06,
      "loss": 0.1161,
      "step": 2789
    },
    {
      "epoch": 0.9674063800277393,
      "grad_norm": 3.2605860233306885,
      "learning_rate": 1e-06,
      "loss": 0.1535,
      "step": 2790
    },
    {
      "epoch": 0.967753120665742,
      "grad_norm": 2.45969295501709,
      "learning_rate": 1e-06,
      "loss": 0.1199,
      "step": 2791
    },
    {
      "epoch": 0.9680998613037448,
      "grad_norm": 3.5829577445983887,
      "learning_rate": 1e-06,
      "loss": 0.1001,
      "step": 2792
    },
    {
      "epoch": 0.9684466019417476,
      "grad_norm": 2.413280963897705,
      "learning_rate": 1e-06,
      "loss": 0.1261,
      "step": 2793
    },
    {
      "epoch": 0.9687933425797504,
      "grad_norm": 2.541856050491333,
      "learning_rate": 1e-06,
      "loss": 0.106,
      "step": 2794
    },
    {
      "epoch": 0.9691400832177531,
      "grad_norm": 1.7989451885223389,
      "learning_rate": 1e-06,
      "loss": 0.1139,
      "step": 2795
    },
    {
      "epoch": 0.9694868238557559,
      "grad_norm": 5.392777442932129,
      "learning_rate": 1e-06,
      "loss": 0.1379,
      "step": 2796
    },
    {
      "epoch": 0.9698335644937587,
      "grad_norm": 2.561255931854248,
      "learning_rate": 1e-06,
      "loss": 0.1301,
      "step": 2797
    },
    {
      "epoch": 0.9701803051317615,
      "grad_norm": 2.3186416625976562,
      "learning_rate": 1e-06,
      "loss": 0.1049,
      "step": 2798
    },
    {
      "epoch": 0.9705270457697642,
      "grad_norm": 2.255427122116089,
      "learning_rate": 1e-06,
      "loss": 0.1084,
      "step": 2799
    },
    {
      "epoch": 0.970873786407767,
      "grad_norm": 3.2925469875335693,
      "learning_rate": 1e-06,
      "loss": 0.1509,
      "step": 2800
    },
    {
      "epoch": 0.9712205270457698,
      "grad_norm": 2.2866995334625244,
      "learning_rate": 1e-06,
      "loss": 0.1221,
      "step": 2801
    },
    {
      "epoch": 0.9715672676837726,
      "grad_norm": 2.409914016723633,
      "learning_rate": 1e-06,
      "loss": 0.1279,
      "step": 2802
    },
    {
      "epoch": 0.9719140083217753,
      "grad_norm": 2.746307849884033,
      "learning_rate": 1e-06,
      "loss": 0.1119,
      "step": 2803
    },
    {
      "epoch": 0.9722607489597781,
      "grad_norm": 2.264927387237549,
      "learning_rate": 1e-06,
      "loss": 0.1333,
      "step": 2804
    },
    {
      "epoch": 0.9726074895977809,
      "grad_norm": 1.9397131204605103,
      "learning_rate": 1e-06,
      "loss": 0.0929,
      "step": 2805
    },
    {
      "epoch": 0.9729542302357836,
      "grad_norm": 4.742498397827148,
      "learning_rate": 1e-06,
      "loss": 0.1577,
      "step": 2806
    },
    {
      "epoch": 0.9733009708737864,
      "grad_norm": 3.283451795578003,
      "learning_rate": 1e-06,
      "loss": 0.114,
      "step": 2807
    },
    {
      "epoch": 0.9736477115117892,
      "grad_norm": 2.6818947792053223,
      "learning_rate": 1e-06,
      "loss": 0.1446,
      "step": 2808
    },
    {
      "epoch": 0.973994452149792,
      "grad_norm": 4.0017476081848145,
      "learning_rate": 1e-06,
      "loss": 0.1241,
      "step": 2809
    },
    {
      "epoch": 0.9743411927877947,
      "grad_norm": 2.343931198120117,
      "learning_rate": 1e-06,
      "loss": 0.1157,
      "step": 2810
    },
    {
      "epoch": 0.9746879334257975,
      "grad_norm": 2.369596004486084,
      "learning_rate": 1e-06,
      "loss": 0.1451,
      "step": 2811
    },
    {
      "epoch": 0.9750346740638003,
      "grad_norm": 2.386434316635132,
      "learning_rate": 1e-06,
      "loss": 0.1205,
      "step": 2812
    },
    {
      "epoch": 0.9753814147018031,
      "grad_norm": 1.8188420534133911,
      "learning_rate": 1e-06,
      "loss": 0.0877,
      "step": 2813
    },
    {
      "epoch": 0.9757281553398058,
      "grad_norm": 2.927330493927002,
      "learning_rate": 1e-06,
      "loss": 0.1493,
      "step": 2814
    },
    {
      "epoch": 0.9760748959778086,
      "grad_norm": 2.0784151554107666,
      "learning_rate": 1e-06,
      "loss": 0.1269,
      "step": 2815
    },
    {
      "epoch": 0.9764216366158114,
      "grad_norm": 2.062995433807373,
      "learning_rate": 1e-06,
      "loss": 0.1287,
      "step": 2816
    },
    {
      "epoch": 0.9767683772538142,
      "grad_norm": 2.0508055686950684,
      "learning_rate": 1e-06,
      "loss": 0.0901,
      "step": 2817
    },
    {
      "epoch": 0.9771151178918169,
      "grad_norm": 3.7478532791137695,
      "learning_rate": 1e-06,
      "loss": 0.1304,
      "step": 2818
    },
    {
      "epoch": 0.9774618585298197,
      "grad_norm": 2.1847145557403564,
      "learning_rate": 1e-06,
      "loss": 0.1275,
      "step": 2819
    },
    {
      "epoch": 0.9778085991678225,
      "grad_norm": 2.580650568008423,
      "learning_rate": 1e-06,
      "loss": 0.157,
      "step": 2820
    },
    {
      "epoch": 0.9781553398058253,
      "grad_norm": 2.9201722145080566,
      "learning_rate": 1e-06,
      "loss": 0.1486,
      "step": 2821
    },
    {
      "epoch": 0.978502080443828,
      "grad_norm": 5.508189678192139,
      "learning_rate": 1e-06,
      "loss": 0.1458,
      "step": 2822
    },
    {
      "epoch": 0.9788488210818308,
      "grad_norm": 3.1913814544677734,
      "learning_rate": 1e-06,
      "loss": 0.1162,
      "step": 2823
    },
    {
      "epoch": 0.9791955617198336,
      "grad_norm": 4.063955783843994,
      "learning_rate": 1e-06,
      "loss": 0.159,
      "step": 2824
    },
    {
      "epoch": 0.9795423023578363,
      "grad_norm": 2.553622245788574,
      "learning_rate": 1e-06,
      "loss": 0.1343,
      "step": 2825
    },
    {
      "epoch": 0.9798890429958391,
      "grad_norm": 2.749812364578247,
      "learning_rate": 1e-06,
      "loss": 0.1393,
      "step": 2826
    },
    {
      "epoch": 0.9802357836338419,
      "grad_norm": 3.2445449829101562,
      "learning_rate": 1e-06,
      "loss": 0.1138,
      "step": 2827
    },
    {
      "epoch": 0.9805825242718447,
      "grad_norm": 2.1383461952209473,
      "learning_rate": 1e-06,
      "loss": 0.1053,
      "step": 2828
    },
    {
      "epoch": 0.9809292649098474,
      "grad_norm": 3.216644763946533,
      "learning_rate": 1e-06,
      "loss": 0.0934,
      "step": 2829
    },
    {
      "epoch": 0.9812760055478502,
      "grad_norm": 2.851259231567383,
      "learning_rate": 1e-06,
      "loss": 0.1455,
      "step": 2830
    },
    {
      "epoch": 0.981622746185853,
      "grad_norm": 3.037447452545166,
      "learning_rate": 1e-06,
      "loss": 0.1392,
      "step": 2831
    },
    {
      "epoch": 0.9819694868238558,
      "grad_norm": 3.316592216491699,
      "learning_rate": 1e-06,
      "loss": 0.1772,
      "step": 2832
    },
    {
      "epoch": 0.9823162274618585,
      "grad_norm": 2.238222599029541,
      "learning_rate": 1e-06,
      "loss": 0.1162,
      "step": 2833
    },
    {
      "epoch": 0.9826629680998613,
      "grad_norm": 2.7036538124084473,
      "learning_rate": 1e-06,
      "loss": 0.1116,
      "step": 2834
    },
    {
      "epoch": 0.9830097087378641,
      "grad_norm": 2.6380455493927,
      "learning_rate": 1e-06,
      "loss": 0.0934,
      "step": 2835
    },
    {
      "epoch": 0.9833564493758669,
      "grad_norm": 2.5659232139587402,
      "learning_rate": 1e-06,
      "loss": 0.1418,
      "step": 2836
    },
    {
      "epoch": 0.9837031900138696,
      "grad_norm": 3.315857172012329,
      "learning_rate": 1e-06,
      "loss": 0.16,
      "step": 2837
    },
    {
      "epoch": 0.9840499306518724,
      "grad_norm": 2.2236742973327637,
      "learning_rate": 1e-06,
      "loss": 0.1433,
      "step": 2838
    },
    {
      "epoch": 0.9843966712898752,
      "grad_norm": 2.794739246368408,
      "learning_rate": 1e-06,
      "loss": 0.1276,
      "step": 2839
    },
    {
      "epoch": 0.984743411927878,
      "grad_norm": 1.8826152086257935,
      "learning_rate": 1e-06,
      "loss": 0.1023,
      "step": 2840
    },
    {
      "epoch": 0.9850901525658807,
      "grad_norm": 2.971917152404785,
      "learning_rate": 1e-06,
      "loss": 0.1238,
      "step": 2841
    },
    {
      "epoch": 0.9854368932038835,
      "grad_norm": 3.801140069961548,
      "learning_rate": 1e-06,
      "loss": 0.1215,
      "step": 2842
    },
    {
      "epoch": 0.9857836338418863,
      "grad_norm": 5.8529276847839355,
      "learning_rate": 1e-06,
      "loss": 0.1557,
      "step": 2843
    },
    {
      "epoch": 0.986130374479889,
      "grad_norm": 3.4524965286254883,
      "learning_rate": 1e-06,
      "loss": 0.1246,
      "step": 2844
    },
    {
      "epoch": 0.9864771151178918,
      "grad_norm": 2.8966329097747803,
      "learning_rate": 1e-06,
      "loss": 0.1083,
      "step": 2845
    },
    {
      "epoch": 0.9868238557558946,
      "grad_norm": 3.0084590911865234,
      "learning_rate": 1e-06,
      "loss": 0.1523,
      "step": 2846
    },
    {
      "epoch": 0.9871705963938974,
      "grad_norm": 2.85929536819458,
      "learning_rate": 1e-06,
      "loss": 0.1506,
      "step": 2847
    },
    {
      "epoch": 0.9875173370319001,
      "grad_norm": 3.1562740802764893,
      "learning_rate": 1e-06,
      "loss": 0.1013,
      "step": 2848
    },
    {
      "epoch": 0.9878640776699029,
      "grad_norm": 2.7616796493530273,
      "learning_rate": 1e-06,
      "loss": 0.139,
      "step": 2849
    },
    {
      "epoch": 0.9882108183079057,
      "grad_norm": 2.781879186630249,
      "learning_rate": 1e-06,
      "loss": 0.164,
      "step": 2850
    },
    {
      "epoch": 0.9885575589459085,
      "grad_norm": 2.50597882270813,
      "learning_rate": 1e-06,
      "loss": 0.1022,
      "step": 2851
    },
    {
      "epoch": 0.9889042995839112,
      "grad_norm": 4.0737833976745605,
      "learning_rate": 1e-06,
      "loss": 0.1652,
      "step": 2852
    },
    {
      "epoch": 0.989251040221914,
      "grad_norm": 3.9901700019836426,
      "learning_rate": 1e-06,
      "loss": 0.1411,
      "step": 2853
    },
    {
      "epoch": 0.9895977808599168,
      "grad_norm": 2.159637928009033,
      "learning_rate": 1e-06,
      "loss": 0.1058,
      "step": 2854
    },
    {
      "epoch": 0.9899445214979196,
      "grad_norm": 3.102248430252075,
      "learning_rate": 1e-06,
      "loss": 0.136,
      "step": 2855
    },
    {
      "epoch": 0.9902912621359223,
      "grad_norm": 2.2245311737060547,
      "learning_rate": 1e-06,
      "loss": 0.1498,
      "step": 2856
    },
    {
      "epoch": 0.9906380027739251,
      "grad_norm": 3.013869524002075,
      "learning_rate": 1e-06,
      "loss": 0.1276,
      "step": 2857
    },
    {
      "epoch": 0.9909847434119279,
      "grad_norm": 2.9369828701019287,
      "learning_rate": 1e-06,
      "loss": 0.1353,
      "step": 2858
    },
    {
      "epoch": 0.9913314840499307,
      "grad_norm": 2.4606969356536865,
      "learning_rate": 1e-06,
      "loss": 0.1251,
      "step": 2859
    },
    {
      "epoch": 0.9916782246879334,
      "grad_norm": 2.1041226387023926,
      "learning_rate": 1e-06,
      "loss": 0.1168,
      "step": 2860
    },
    {
      "epoch": 0.9920249653259362,
      "grad_norm": 2.6059367656707764,
      "learning_rate": 1e-06,
      "loss": 0.1419,
      "step": 2861
    },
    {
      "epoch": 0.992371705963939,
      "grad_norm": 2.68571138381958,
      "learning_rate": 1e-06,
      "loss": 0.147,
      "step": 2862
    },
    {
      "epoch": 0.9927184466019418,
      "grad_norm": 2.2065820693969727,
      "learning_rate": 1e-06,
      "loss": 0.1141,
      "step": 2863
    },
    {
      "epoch": 0.9930651872399445,
      "grad_norm": 2.7700083255767822,
      "learning_rate": 1e-06,
      "loss": 0.1321,
      "step": 2864
    },
    {
      "epoch": 0.9934119278779473,
      "grad_norm": 1.815712571144104,
      "learning_rate": 1e-06,
      "loss": 0.104,
      "step": 2865
    },
    {
      "epoch": 0.9937586685159501,
      "grad_norm": 2.593801259994507,
      "learning_rate": 1e-06,
      "loss": 0.1425,
      "step": 2866
    },
    {
      "epoch": 0.9941054091539528,
      "grad_norm": 2.9171457290649414,
      "learning_rate": 1e-06,
      "loss": 0.1342,
      "step": 2867
    },
    {
      "epoch": 0.9944521497919556,
      "grad_norm": 4.014968395233154,
      "learning_rate": 1e-06,
      "loss": 0.1385,
      "step": 2868
    },
    {
      "epoch": 0.9947988904299584,
      "grad_norm": 2.8752143383026123,
      "learning_rate": 1e-06,
      "loss": 0.1386,
      "step": 2869
    },
    {
      "epoch": 0.9951456310679612,
      "grad_norm": 3.8755412101745605,
      "learning_rate": 1e-06,
      "loss": 0.1815,
      "step": 2870
    },
    {
      "epoch": 0.9954923717059639,
      "grad_norm": 2.6354281902313232,
      "learning_rate": 1e-06,
      "loss": 0.1115,
      "step": 2871
    },
    {
      "epoch": 0.9958391123439667,
      "grad_norm": 2.4390392303466797,
      "learning_rate": 1e-06,
      "loss": 0.1543,
      "step": 2872
    },
    {
      "epoch": 0.9961858529819695,
      "grad_norm": 2.396690607070923,
      "learning_rate": 1e-06,
      "loss": 0.1096,
      "step": 2873
    },
    {
      "epoch": 0.9965325936199723,
      "grad_norm": 2.379737615585327,
      "learning_rate": 1e-06,
      "loss": 0.1426,
      "step": 2874
    },
    {
      "epoch": 0.996879334257975,
      "grad_norm": 3.5744974613189697,
      "learning_rate": 1e-06,
      "loss": 0.1485,
      "step": 2875
    },
    {
      "epoch": 0.9972260748959778,
      "grad_norm": 2.513230323791504,
      "learning_rate": 1e-06,
      "loss": 0.1598,
      "step": 2876
    },
    {
      "epoch": 0.9975728155339806,
      "grad_norm": 2.260457992553711,
      "learning_rate": 1e-06,
      "loss": 0.1226,
      "step": 2877
    },
    {
      "epoch": 0.9979195561719834,
      "grad_norm": 2.7690110206604004,
      "learning_rate": 1e-06,
      "loss": 0.1366,
      "step": 2878
    },
    {
      "epoch": 0.9982662968099861,
      "grad_norm": 2.5981392860412598,
      "learning_rate": 1e-06,
      "loss": 0.1333,
      "step": 2879
    },
    {
      "epoch": 0.9986130374479889,
      "grad_norm": 3.166869878768921,
      "learning_rate": 1e-06,
      "loss": 0.1061,
      "step": 2880
    },
    {
      "epoch": 0.9989597780859917,
      "grad_norm": 1.9656025171279907,
      "learning_rate": 1e-06,
      "loss": 0.12,
      "step": 2881
    },
    {
      "epoch": 0.9993065187239945,
      "grad_norm": 2.704719066619873,
      "learning_rate": 1e-06,
      "loss": 0.124,
      "step": 2882
    },
    {
      "epoch": 0.9996532593619972,
      "grad_norm": 2.3944320678710938,
      "learning_rate": 1e-06,
      "loss": 0.1216,
      "step": 2883
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.362536907196045,
      "learning_rate": 1e-06,
      "loss": 0.1326,
      "step": 2884
    },
    {
      "epoch": 1.0,
      "step": 2884,
      "total_flos": 9.774378103106175e+18,
      "train_loss": 0.16740735881443908,
      "train_runtime": 15269.0756,
      "train_samples_per_second": 24.172,
      "train_steps_per_second": 0.189
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 2884,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.774378103106175e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}