{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.1765536723163842,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00035310734463276836,
      "grad_norm": 0.45211324095726013,
      "learning_rate": 5e-05,
      "loss": 3.273,
      "step": 1
    },
    {
      "epoch": 0.0007062146892655367,
      "grad_norm": 0.30982279777526855,
      "learning_rate": 4.99e-05,
      "loss": 2.5125,
      "step": 2
    },
    {
      "epoch": 0.001059322033898305,
      "grad_norm": 0.39342790842056274,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 2.8347,
      "step": 3
    },
    {
      "epoch": 0.0014124293785310734,
      "grad_norm": 0.3912785053253174,
      "learning_rate": 4.97e-05,
      "loss": 2.6189,
      "step": 4
    },
    {
      "epoch": 0.0017655367231638418,
      "grad_norm": 0.43560200929641724,
      "learning_rate": 4.96e-05,
      "loss": 2.8784,
      "step": 5
    },
    {
      "epoch": 0.00211864406779661,
      "grad_norm": 0.5163903832435608,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 3.0496,
      "step": 6
    },
    {
      "epoch": 0.0024717514124293787,
      "grad_norm": 0.3364083468914032,
      "learning_rate": 4.94e-05,
      "loss": 2.0403,
      "step": 7
    },
    {
      "epoch": 0.002824858757062147,
      "grad_norm": 0.4723369777202606,
      "learning_rate": 4.93e-05,
      "loss": 2.5065,
      "step": 8
    },
    {
      "epoch": 0.0031779661016949155,
      "grad_norm": 0.5686460137367249,
      "learning_rate": 4.92e-05,
      "loss": 2.8529,
      "step": 9
    },
    {
      "epoch": 0.0035310734463276836,
      "grad_norm": 0.7666249871253967,
      "learning_rate": 4.91e-05,
      "loss": 3.2163,
      "step": 10
    },
    {
      "epoch": 0.003884180790960452,
      "grad_norm": 0.5175825953483582,
      "learning_rate": 4.9e-05,
      "loss": 2.3817,
      "step": 11
    },
    {
      "epoch": 0.00423728813559322,
      "grad_norm": 0.7693667411804199,
      "learning_rate": 4.89e-05,
      "loss": 3.1173,
      "step": 12
    },
    {
      "epoch": 0.004590395480225989,
      "grad_norm": 0.6010498404502869,
      "learning_rate": 4.88e-05,
      "loss": 2.5264,
      "step": 13
    },
    {
      "epoch": 0.0049435028248587575,
      "grad_norm": 0.8724003434181213,
      "learning_rate": 4.87e-05,
      "loss": 3.1545,
      "step": 14
    },
    {
      "epoch": 0.005296610169491525,
      "grad_norm": 0.6222332119941711,
      "learning_rate": 4.86e-05,
      "loss": 2.4859,
      "step": 15
    },
    {
      "epoch": 0.005649717514124294,
      "grad_norm": 0.5855898261070251,
      "learning_rate": 4.85e-05,
      "loss": 2.2748,
      "step": 16
    },
    {
      "epoch": 0.006002824858757062,
      "grad_norm": 0.8227508068084717,
      "learning_rate": 4.8400000000000004e-05,
      "loss": 2.777,
      "step": 17
    },
    {
      "epoch": 0.006355932203389831,
      "grad_norm": 0.747832179069519,
      "learning_rate": 4.83e-05,
      "loss": 2.6429,
      "step": 18
    },
    {
      "epoch": 0.006709039548022599,
      "grad_norm": 0.8245562314987183,
      "learning_rate": 4.82e-05,
      "loss": 2.8865,
      "step": 19
    },
    {
      "epoch": 0.007062146892655367,
      "grad_norm": 0.8959988951683044,
      "learning_rate": 4.8100000000000004e-05,
      "loss": 2.8727,
      "step": 20
    },
    {
      "epoch": 0.007415254237288136,
      "grad_norm": 0.7936757802963257,
      "learning_rate": 4.8e-05,
      "loss": 2.5376,
      "step": 21
    },
    {
      "epoch": 0.007768361581920904,
      "grad_norm": 0.8037199378013611,
      "learning_rate": 4.79e-05,
      "loss": 2.6522,
      "step": 22
    },
    {
      "epoch": 0.008121468926553672,
      "grad_norm": 0.7104085087776184,
      "learning_rate": 4.78e-05,
      "loss": 2.3427,
      "step": 23
    },
    {
      "epoch": 0.00847457627118644,
      "grad_norm": 0.8215932846069336,
      "learning_rate": 4.77e-05,
      "loss": 2.7918,
      "step": 24
    },
    {
      "epoch": 0.00882768361581921,
      "grad_norm": 1.0282623767852783,
      "learning_rate": 4.76e-05,
      "loss": 2.9389,
      "step": 25
    },
    {
      "epoch": 0.009180790960451978,
      "grad_norm": 0.7932501435279846,
      "learning_rate": 4.75e-05,
      "loss": 2.7691,
      "step": 26
    },
    {
      "epoch": 0.009533898305084746,
      "grad_norm": 0.6620106101036072,
      "learning_rate": 4.74e-05,
      "loss": 2.1335,
      "step": 27
    },
    {
      "epoch": 0.009887005649717515,
      "grad_norm": 0.8058229684829712,
      "learning_rate": 4.73e-05,
      "loss": 2.4778,
      "step": 28
    },
    {
      "epoch": 0.010240112994350282,
      "grad_norm": 0.7049869894981384,
      "learning_rate": 4.72e-05,
      "loss": 2.3379,
      "step": 29
    },
    {
      "epoch": 0.01059322033898305,
      "grad_norm": 0.6863004565238953,
      "learning_rate": 4.71e-05,
      "loss": 2.0131,
      "step": 30
    },
    {
      "epoch": 0.010946327683615819,
      "grad_norm": 0.9368977546691895,
      "learning_rate": 4.7e-05,
      "loss": 2.4664,
      "step": 31
    },
    {
      "epoch": 0.011299435028248588,
      "grad_norm": 0.7291913628578186,
      "learning_rate": 4.69e-05,
      "loss": 2.2541,
      "step": 32
    },
    {
      "epoch": 0.011652542372881356,
      "grad_norm": 0.6416012048721313,
      "learning_rate": 4.6800000000000006e-05,
      "loss": 1.8727,
      "step": 33
    },
    {
      "epoch": 0.012005649717514125,
      "grad_norm": 0.6020755767822266,
      "learning_rate": 4.6700000000000003e-05,
      "loss": 1.8596,
      "step": 34
    },
    {
      "epoch": 0.012358757062146893,
      "grad_norm": 0.635847270488739,
      "learning_rate": 4.660000000000001e-05,
      "loss": 1.9688,
      "step": 35
    },
    {
      "epoch": 0.012711864406779662,
      "grad_norm": 0.6946876645088196,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 2.6249,
      "step": 36
    },
    {
      "epoch": 0.013064971751412429,
      "grad_norm": 0.7043133974075317,
      "learning_rate": 4.64e-05,
      "loss": 2.5754,
      "step": 37
    },
    {
      "epoch": 0.013418079096045197,
      "grad_norm": 0.5472213625907898,
      "learning_rate": 4.630000000000001e-05,
      "loss": 1.91,
      "step": 38
    },
    {
      "epoch": 0.013771186440677966,
      "grad_norm": 0.5201235413551331,
      "learning_rate": 4.6200000000000005e-05,
      "loss": 2.0296,
      "step": 39
    },
    {
      "epoch": 0.014124293785310734,
      "grad_norm": 0.5103974342346191,
      "learning_rate": 4.61e-05,
      "loss": 1.7567,
      "step": 40
    },
    {
      "epoch": 0.014477401129943503,
      "grad_norm": 0.42137110233306885,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.7719,
      "step": 41
    },
    {
      "epoch": 0.014830508474576272,
      "grad_norm": 0.6730819940567017,
      "learning_rate": 4.5900000000000004e-05,
      "loss": 2.4245,
      "step": 42
    },
    {
      "epoch": 0.01518361581920904,
      "grad_norm": 0.6370269060134888,
      "learning_rate": 4.58e-05,
      "loss": 2.501,
      "step": 43
    },
    {
      "epoch": 0.015536723163841809,
      "grad_norm": 0.5084403157234192,
      "learning_rate": 4.5700000000000006e-05,
      "loss": 2.0893,
      "step": 44
    },
    {
      "epoch": 0.015889830508474576,
      "grad_norm": 0.6537813544273376,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 2.3681,
      "step": 45
    },
    {
      "epoch": 0.016242937853107344,
      "grad_norm": 0.5805037617683411,
      "learning_rate": 4.55e-05,
      "loss": 2.3028,
      "step": 46
    },
    {
      "epoch": 0.016596045197740113,
      "grad_norm": 0.5214577913284302,
      "learning_rate": 4.5400000000000006e-05,
      "loss": 2.1601,
      "step": 47
    },
    {
      "epoch": 0.01694915254237288,
      "grad_norm": 0.7660169005393982,
      "learning_rate": 4.53e-05,
      "loss": 2.4195,
      "step": 48
    },
    {
      "epoch": 0.01730225988700565,
      "grad_norm": 0.537522554397583,
      "learning_rate": 4.52e-05,
      "loss": 2.3828,
      "step": 49
    },
    {
      "epoch": 0.01765536723163842,
      "grad_norm": 0.5875993371009827,
      "learning_rate": 4.5100000000000005e-05,
      "loss": 2.4529,
      "step": 50
    },
    {
      "epoch": 0.018008474576271187,
      "grad_norm": 0.480882465839386,
      "learning_rate": 4.5e-05,
      "loss": 1.7377,
      "step": 51
    },
    {
      "epoch": 0.018361581920903956,
      "grad_norm": 0.4598454236984253,
      "learning_rate": 4.49e-05,
      "loss": 1.8277,
      "step": 52
    },
    {
      "epoch": 0.018714689265536724,
      "grad_norm": 0.4913174510002136,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 2.0882,
      "step": 53
    },
    {
      "epoch": 0.019067796610169493,
      "grad_norm": 0.4395110011100769,
      "learning_rate": 4.47e-05,
      "loss": 1.8825,
      "step": 54
    },
    {
      "epoch": 0.01942090395480226,
      "grad_norm": 0.5200910568237305,
      "learning_rate": 4.46e-05,
      "loss": 2.1883,
      "step": 55
    },
    {
      "epoch": 0.01977401129943503,
      "grad_norm": 0.7119576930999756,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 2.3147,
      "step": 56
    },
    {
      "epoch": 0.020127118644067795,
      "grad_norm": 0.5865232348442078,
      "learning_rate": 4.44e-05,
      "loss": 2.4823,
      "step": 57
    },
    {
      "epoch": 0.020480225988700564,
      "grad_norm": 0.5342106223106384,
      "learning_rate": 4.43e-05,
      "loss": 2.3112,
      "step": 58
    },
    {
      "epoch": 0.020833333333333332,
      "grad_norm": 0.63221275806427,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 2.4642,
      "step": 59
    },
    {
      "epoch": 0.0211864406779661,
      "grad_norm": 0.5381503701210022,
      "learning_rate": 4.41e-05,
      "loss": 2.1702,
      "step": 60
    },
    {
      "epoch": 0.02153954802259887,
      "grad_norm": 0.34933093190193176,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.6217,
      "step": 61
    },
    {
      "epoch": 0.021892655367231638,
      "grad_norm": 0.4486427307128906,
      "learning_rate": 4.39e-05,
      "loss": 1.9545,
      "step": 62
    },
    {
      "epoch": 0.022245762711864406,
      "grad_norm": 0.39275842905044556,
      "learning_rate": 4.38e-05,
      "loss": 1.692,
      "step": 63
    },
    {
      "epoch": 0.022598870056497175,
      "grad_norm": 0.5706101655960083,
      "learning_rate": 4.3700000000000005e-05,
      "loss": 2.3182,
      "step": 64
    },
    {
      "epoch": 0.022951977401129944,
      "grad_norm": 0.43103381991386414,
      "learning_rate": 4.36e-05,
      "loss": 1.8674,
      "step": 65
    },
    {
      "epoch": 0.023305084745762712,
      "grad_norm": 0.6326637268066406,
      "learning_rate": 4.35e-05,
      "loss": 2.3281,
      "step": 66
    },
    {
      "epoch": 0.02365819209039548,
      "grad_norm": 0.5602039694786072,
      "learning_rate": 4.3400000000000005e-05,
      "loss": 2.24,
      "step": 67
    },
    {
      "epoch": 0.02401129943502825,
      "grad_norm": 0.42276912927627563,
      "learning_rate": 4.33e-05,
      "loss": 1.9933,
      "step": 68
    },
    {
      "epoch": 0.024364406779661018,
      "grad_norm": 0.5240011811256409,
      "learning_rate": 4.32e-05,
      "loss": 2.1384,
      "step": 69
    },
    {
      "epoch": 0.024717514124293787,
      "grad_norm": 0.579456627368927,
      "learning_rate": 4.3100000000000004e-05,
      "loss": 2.2305,
      "step": 70
    },
    {
      "epoch": 0.025070621468926555,
      "grad_norm": 0.5782936215400696,
      "learning_rate": 4.3e-05,
      "loss": 2.2021,
      "step": 71
    },
    {
      "epoch": 0.025423728813559324,
      "grad_norm": 0.6299965977668762,
      "learning_rate": 4.29e-05,
      "loss": 2.3214,
      "step": 72
    },
    {
      "epoch": 0.02577683615819209,
      "grad_norm": 0.634605884552002,
      "learning_rate": 4.2800000000000004e-05,
      "loss": 2.3198,
      "step": 73
    },
    {
      "epoch": 0.026129943502824857,
      "grad_norm": 0.5833197832107544,
      "learning_rate": 4.27e-05,
      "loss": 2.1705,
      "step": 74
    },
    {
      "epoch": 0.026483050847457626,
      "grad_norm": 0.5400241017341614,
      "learning_rate": 4.26e-05,
      "loss": 1.8534,
      "step": 75
    },
    {
      "epoch": 0.026836158192090395,
      "grad_norm": 0.6428772807121277,
      "learning_rate": 4.25e-05,
      "loss": 2.3044,
      "step": 76
    },
    {
      "epoch": 0.027189265536723163,
      "grad_norm": 0.611422061920166,
      "learning_rate": 4.24e-05,
      "loss": 2.1591,
      "step": 77
    },
    {
      "epoch": 0.02754237288135593,
      "grad_norm": 0.602570652961731,
      "learning_rate": 4.23e-05,
      "loss": 2.1249,
      "step": 78
    },
    {
      "epoch": 0.0278954802259887,
      "grad_norm": 0.48029282689094543,
      "learning_rate": 4.22e-05,
      "loss": 1.905,
      "step": 79
    },
    {
      "epoch": 0.02824858757062147,
      "grad_norm": 0.6572628617286682,
      "learning_rate": 4.21e-05,
      "loss": 2.2358,
      "step": 80
    },
    {
      "epoch": 0.028601694915254237,
      "grad_norm": 0.5220134258270264,
      "learning_rate": 4.2e-05,
      "loss": 1.9954,
      "step": 81
    },
    {
      "epoch": 0.028954802259887006,
      "grad_norm": 0.4906415641307831,
      "learning_rate": 4.19e-05,
      "loss": 1.8642,
      "step": 82
    },
    {
      "epoch": 0.029307909604519775,
      "grad_norm": 0.6776078343391418,
      "learning_rate": 4.18e-05,
      "loss": 2.2061,
      "step": 83
    },
    {
      "epoch": 0.029661016949152543,
      "grad_norm": 0.496427059173584,
      "learning_rate": 4.17e-05,
      "loss": 1.9494,
      "step": 84
    },
    {
      "epoch": 0.03001412429378531,
      "grad_norm": 0.4432898759841919,
      "learning_rate": 4.16e-05,
      "loss": 1.6809,
      "step": 85
    },
    {
      "epoch": 0.03036723163841808,
      "grad_norm": 0.6649871468544006,
      "learning_rate": 4.15e-05,
      "loss": 2.0615,
      "step": 86
    },
    {
      "epoch": 0.03072033898305085,
      "grad_norm": 0.6379185318946838,
      "learning_rate": 4.14e-05,
      "loss": 1.9796,
      "step": 87
    },
    {
      "epoch": 0.031073446327683617,
      "grad_norm": 0.5551559925079346,
      "learning_rate": 4.13e-05,
      "loss": 1.874,
      "step": 88
    },
    {
      "epoch": 0.031426553672316386,
      "grad_norm": 0.7269681096076965,
      "learning_rate": 4.12e-05,
      "loss": 2.1407,
      "step": 89
    },
    {
      "epoch": 0.03177966101694915,
      "grad_norm": 0.6741663217544556,
      "learning_rate": 4.11e-05,
      "loss": 1.9975,
      "step": 90
    },
    {
      "epoch": 0.03213276836158192,
      "grad_norm": 0.5726102590560913,
      "learning_rate": 4.1e-05,
      "loss": 1.8897,
      "step": 91
    },
    {
      "epoch": 0.03248587570621469,
      "grad_norm": 0.7033076286315918,
      "learning_rate": 4.09e-05,
      "loss": 2.0044,
      "step": 92
    },
    {
      "epoch": 0.03283898305084746,
      "grad_norm": 0.6314510107040405,
      "learning_rate": 4.08e-05,
      "loss": 2.1114,
      "step": 93
    },
    {
      "epoch": 0.033192090395480225,
      "grad_norm": 0.49348896741867065,
      "learning_rate": 4.07e-05,
      "loss": 1.5142,
      "step": 94
    },
    {
      "epoch": 0.033545197740113,
      "grad_norm": 0.5636611580848694,
      "learning_rate": 4.0600000000000004e-05,
      "loss": 1.7442,
      "step": 95
    },
    {
      "epoch": 0.03389830508474576,
      "grad_norm": 0.5502078533172607,
      "learning_rate": 4.05e-05,
      "loss": 1.744,
      "step": 96
    },
    {
      "epoch": 0.03425141242937853,
      "grad_norm": 0.7853062152862549,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 2.0471,
      "step": 97
    },
    {
      "epoch": 0.0346045197740113,
      "grad_norm": 0.7374521493911743,
      "learning_rate": 4.0300000000000004e-05,
      "loss": 1.9658,
      "step": 98
    },
    {
      "epoch": 0.034957627118644065,
      "grad_norm": 0.5843427181243896,
      "learning_rate": 4.02e-05,
      "loss": 1.698,
      "step": 99
    },
    {
      "epoch": 0.03531073446327684,
      "grad_norm": 0.7103281617164612,
      "learning_rate": 4.0100000000000006e-05,
      "loss": 1.8531,
      "step": 100
    },
    {
      "epoch": 0.0356638418079096,
      "grad_norm": 0.547626256942749,
      "learning_rate": 4e-05,
      "loss": 1.6113,
      "step": 101
    },
    {
      "epoch": 0.036016949152542374,
      "grad_norm": 0.7051151990890503,
      "learning_rate": 3.99e-05,
      "loss": 1.8774,
      "step": 102
    },
    {
      "epoch": 0.03637005649717514,
      "grad_norm": 0.5148686766624451,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 1.6082,
      "step": 103
    },
    {
      "epoch": 0.03672316384180791,
      "grad_norm": 0.5013210773468018,
      "learning_rate": 3.97e-05,
      "loss": 1.6964,
      "step": 104
    },
    {
      "epoch": 0.037076271186440676,
      "grad_norm": 0.7677639126777649,
      "learning_rate": 3.960000000000001e-05,
      "loss": 1.8114,
      "step": 105
    },
    {
      "epoch": 0.03742937853107345,
      "grad_norm": 0.6067560911178589,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 1.6947,
      "step": 106
    },
    {
      "epoch": 0.037782485875706213,
      "grad_norm": 0.5241538286209106,
      "learning_rate": 3.94e-05,
      "loss": 1.4953,
      "step": 107
    },
    {
      "epoch": 0.038135593220338986,
      "grad_norm": 0.5037181377410889,
      "learning_rate": 3.9300000000000007e-05,
      "loss": 1.4166,
      "step": 108
    },
    {
      "epoch": 0.03848870056497175,
      "grad_norm": 0.7840654253959656,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 1.8501,
      "step": 109
    },
    {
      "epoch": 0.03884180790960452,
      "grad_norm": 0.6080065369606018,
      "learning_rate": 3.91e-05,
      "loss": 1.6206,
      "step": 110
    },
    {
      "epoch": 0.03919491525423729,
      "grad_norm": 0.616134762763977,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 1.7034,
      "step": 111
    },
    {
      "epoch": 0.03954802259887006,
      "grad_norm": 0.7034599781036377,
      "learning_rate": 3.8900000000000004e-05,
      "loss": 1.9318,
      "step": 112
    },
    {
      "epoch": 0.039901129943502825,
      "grad_norm": 0.8047641515731812,
      "learning_rate": 3.88e-05,
      "loss": 1.7351,
      "step": 113
    },
    {
      "epoch": 0.04025423728813559,
      "grad_norm": 0.8917104005813599,
      "learning_rate": 3.8700000000000006e-05,
      "loss": 1.8398,
      "step": 114
    },
    {
      "epoch": 0.04060734463276836,
      "grad_norm": 0.8026954531669617,
      "learning_rate": 3.86e-05,
      "loss": 1.7842,
      "step": 115
    },
    {
      "epoch": 0.04096045197740113,
      "grad_norm": 0.5217316150665283,
      "learning_rate": 3.85e-05,
      "loss": 1.5885,
      "step": 116
    },
    {
      "epoch": 0.0413135593220339,
      "grad_norm": 0.6288496851921082,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 1.5425,
      "step": 117
    },
    {
      "epoch": 0.041666666666666664,
      "grad_norm": 0.77554851770401,
      "learning_rate": 3.83e-05,
      "loss": 1.6508,
      "step": 118
    },
    {
      "epoch": 0.042019774011299436,
      "grad_norm": 0.5694034695625305,
      "learning_rate": 3.82e-05,
      "loss": 1.2894,
      "step": 119
    },
    {
      "epoch": 0.0423728813559322,
      "grad_norm": 0.6931347846984863,
      "learning_rate": 3.8100000000000005e-05,
      "loss": 1.7874,
      "step": 120
    },
    {
      "epoch": 0.042725988700564974,
      "grad_norm": 0.7180691957473755,
      "learning_rate": 3.8e-05,
      "loss": 1.5679,
      "step": 121
    },
    {
      "epoch": 0.04307909604519774,
      "grad_norm": 0.5494984984397888,
      "learning_rate": 3.79e-05,
      "loss": 1.2885,
      "step": 122
    },
    {
      "epoch": 0.04343220338983051,
      "grad_norm": 0.8779198527336121,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 1.6815,
      "step": 123
    },
    {
      "epoch": 0.043785310734463276,
      "grad_norm": 0.7112749218940735,
      "learning_rate": 3.77e-05,
      "loss": 1.4889,
      "step": 124
    },
    {
      "epoch": 0.04413841807909605,
      "grad_norm": 0.6239631772041321,
      "learning_rate": 3.76e-05,
      "loss": 1.3156,
      "step": 125
    },
    {
      "epoch": 0.04449152542372881,
      "grad_norm": 0.8934807181358337,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.7629,
      "step": 126
    },
    {
      "epoch": 0.044844632768361585,
      "grad_norm": 0.9005821943283081,
      "learning_rate": 3.74e-05,
      "loss": 1.5775,
      "step": 127
    },
    {
      "epoch": 0.04519774011299435,
      "grad_norm": 0.7520875334739685,
      "learning_rate": 3.73e-05,
      "loss": 1.6556,
      "step": 128
    },
    {
      "epoch": 0.045550847457627115,
      "grad_norm": 0.8343580961227417,
      "learning_rate": 3.72e-05,
      "loss": 1.5429,
      "step": 129
    },
    {
      "epoch": 0.04590395480225989,
      "grad_norm": 0.7819499373435974,
      "learning_rate": 3.71e-05,
      "loss": 1.5313,
      "step": 130
    },
    {
      "epoch": 0.04625706214689265,
      "grad_norm": 0.6105003952980042,
      "learning_rate": 3.7e-05,
      "loss": 1.3124,
      "step": 131
    },
    {
      "epoch": 0.046610169491525424,
      "grad_norm": 0.6769589185714722,
      "learning_rate": 3.69e-05,
      "loss": 1.4305,
      "step": 132
    },
    {
      "epoch": 0.04696327683615819,
      "grad_norm": 0.7654910683631897,
      "learning_rate": 3.68e-05,
      "loss": 1.4972,
      "step": 133
    },
    {
      "epoch": 0.04731638418079096,
      "grad_norm": 0.7610757946968079,
      "learning_rate": 3.6700000000000004e-05,
      "loss": 1.4652,
      "step": 134
    },
    {
      "epoch": 0.04766949152542373,
      "grad_norm": 0.6594064831733704,
      "learning_rate": 3.66e-05,
      "loss": 1.2301,
      "step": 135
    },
    {
      "epoch": 0.0480225988700565,
      "grad_norm": 0.719152569770813,
      "learning_rate": 3.65e-05,
      "loss": 1.3399,
      "step": 136
    },
    {
      "epoch": 0.048375706214689264,
      "grad_norm": 0.9311054348945618,
      "learning_rate": 3.6400000000000004e-05,
      "loss": 1.5786,
      "step": 137
    },
    {
      "epoch": 0.048728813559322036,
      "grad_norm": 0.7845869660377502,
      "learning_rate": 3.63e-05,
      "loss": 1.3293,
      "step": 138
    },
    {
      "epoch": 0.0490819209039548,
      "grad_norm": 0.6555350422859192,
      "learning_rate": 3.62e-05,
      "loss": 1.1677,
      "step": 139
    },
    {
      "epoch": 0.04943502824858757,
      "grad_norm": 0.863808810710907,
      "learning_rate": 3.61e-05,
      "loss": 1.3453,
      "step": 140
    },
    {
      "epoch": 0.04978813559322034,
      "grad_norm": 1.1489522457122803,
      "learning_rate": 3.6e-05,
      "loss": 1.5277,
      "step": 141
    },
    {
      "epoch": 0.05014124293785311,
      "grad_norm": 0.8987497091293335,
      "learning_rate": 3.59e-05,
      "loss": 1.3435,
      "step": 142
    },
    {
      "epoch": 0.050494350282485875,
      "grad_norm": 1.1033629179000854,
      "learning_rate": 3.58e-05,
      "loss": 1.5565,
      "step": 143
    },
    {
      "epoch": 0.05084745762711865,
      "grad_norm": 0.9892997145652771,
      "learning_rate": 3.57e-05,
      "loss": 1.5017,
      "step": 144
    },
    {
      "epoch": 0.05120056497175141,
      "grad_norm": 0.8673040866851807,
      "learning_rate": 3.56e-05,
      "loss": 1.3484,
      "step": 145
    },
    {
      "epoch": 0.05155367231638418,
      "grad_norm": 0.8514487147331238,
      "learning_rate": 3.55e-05,
      "loss": 1.4329,
      "step": 146
    },
    {
      "epoch": 0.05190677966101695,
      "grad_norm": 0.7865382432937622,
      "learning_rate": 3.54e-05,
      "loss": 1.3205,
      "step": 147
    },
    {
      "epoch": 0.052259887005649715,
      "grad_norm": 0.9893730878829956,
      "learning_rate": 3.53e-05,
      "loss": 1.2557,
      "step": 148
    },
    {
      "epoch": 0.05261299435028249,
      "grad_norm": 0.7459471225738525,
      "learning_rate": 3.52e-05,
      "loss": 1.398,
      "step": 149
    },
    {
      "epoch": 0.05296610169491525,
      "grad_norm": 1.2132583856582642,
      "learning_rate": 3.51e-05,
      "loss": 1.396,
      "step": 150
    },
    {
      "epoch": 0.053319209039548024,
      "grad_norm": 1.1160212755203247,
      "learning_rate": 3.5e-05,
      "loss": 1.3013,
      "step": 151
    },
    {
      "epoch": 0.05367231638418079,
      "grad_norm": 1.058889389038086,
      "learning_rate": 3.49e-05,
      "loss": 1.4149,
      "step": 152
    },
    {
      "epoch": 0.05402542372881356,
      "grad_norm": 1.065194845199585,
      "learning_rate": 3.48e-05,
      "loss": 1.3913,
      "step": 153
    },
    {
      "epoch": 0.054378531073446326,
      "grad_norm": 0.8159513473510742,
      "learning_rate": 3.4699999999999996e-05,
      "loss": 1.1171,
      "step": 154
    },
    {
      "epoch": 0.0547316384180791,
      "grad_norm": 1.2298649549484253,
      "learning_rate": 3.46e-05,
      "loss": 1.2559,
      "step": 155
    },
    {
      "epoch": 0.05508474576271186,
      "grad_norm": 0.845851480960846,
      "learning_rate": 3.45e-05,
      "loss": 1.1559,
      "step": 156
    },
    {
      "epoch": 0.055437853107344635,
      "grad_norm": 0.8105614185333252,
      "learning_rate": 3.4399999999999996e-05,
      "loss": 1.092,
      "step": 157
    },
    {
      "epoch": 0.0557909604519774,
      "grad_norm": 0.9624556303024292,
      "learning_rate": 3.430000000000001e-05,
      "loss": 1.1862,
      "step": 158
    },
    {
      "epoch": 0.05614406779661017,
      "grad_norm": 1.2695869207382202,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 1.1886,
      "step": 159
    },
    {
      "epoch": 0.05649717514124294,
      "grad_norm": 0.9481382369995117,
      "learning_rate": 3.41e-05,
      "loss": 1.2366,
      "step": 160
    },
    {
      "epoch": 0.0568502824858757,
      "grad_norm": 1.1612368822097778,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 1.3102,
      "step": 161
    },
    {
      "epoch": 0.057203389830508475,
      "grad_norm": 1.296866536140442,
      "learning_rate": 3.3900000000000004e-05,
      "loss": 1.2689,
      "step": 162
    },
    {
      "epoch": 0.05755649717514124,
      "grad_norm": 1.0605868101119995,
      "learning_rate": 3.38e-05,
      "loss": 1.1056,
      "step": 163
    },
    {
      "epoch": 0.05790960451977401,
      "grad_norm": 1.06467604637146,
      "learning_rate": 3.3700000000000006e-05,
      "loss": 1.2329,
      "step": 164
    },
    {
      "epoch": 0.05826271186440678,
      "grad_norm": 0.9635260105133057,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 1.082,
      "step": 165
    },
    {
      "epoch": 0.05861581920903955,
      "grad_norm": 1.2294679880142212,
      "learning_rate": 3.35e-05,
      "loss": 1.3892,
      "step": 166
    },
    {
      "epoch": 0.058968926553672314,
      "grad_norm": 1.4516520500183105,
      "learning_rate": 3.3400000000000005e-05,
      "loss": 1.0772,
      "step": 167
    },
    {
      "epoch": 0.059322033898305086,
      "grad_norm": 1.08195960521698,
      "learning_rate": 3.33e-05,
      "loss": 1.0811,
      "step": 168
    },
    {
      "epoch": 0.05967514124293785,
      "grad_norm": 1.1554380655288696,
      "learning_rate": 3.32e-05,
      "loss": 1.1417,
      "step": 169
    },
    {
      "epoch": 0.06002824858757062,
      "grad_norm": 1.117101788520813,
      "learning_rate": 3.3100000000000005e-05,
      "loss": 1.2163,
      "step": 170
    },
    {
      "epoch": 0.06038135593220339,
      "grad_norm": 1.2141193151474,
      "learning_rate": 3.3e-05,
      "loss": 1.2469,
      "step": 171
    },
    {
      "epoch": 0.06073446327683616,
      "grad_norm": 1.29922616481781,
      "learning_rate": 3.29e-05,
      "loss": 1.2541,
      "step": 172
    },
    {
      "epoch": 0.061087570621468926,
      "grad_norm": 1.3265364170074463,
      "learning_rate": 3.2800000000000004e-05,
      "loss": 1.1884,
      "step": 173
    },
    {
      "epoch": 0.0614406779661017,
      "grad_norm": 1.4586164951324463,
      "learning_rate": 3.27e-05,
      "loss": 1.1861,
      "step": 174
    },
    {
      "epoch": 0.06179378531073446,
      "grad_norm": 1.3976298570632935,
      "learning_rate": 3.26e-05,
      "loss": 1.0568,
      "step": 175
    },
    {
      "epoch": 0.062146892655367235,
      "grad_norm": 1.0947189331054688,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 1.0583,
      "step": 176
    },
    {
      "epoch": 0.0625,
      "grad_norm": 1.0739431381225586,
      "learning_rate": 3.24e-05,
      "loss": 0.973,
      "step": 177
    },
    {
      "epoch": 0.06285310734463277,
      "grad_norm": 1.0500564575195312,
      "learning_rate": 3.2300000000000006e-05,
      "loss": 0.961,
      "step": 178
    },
    {
      "epoch": 0.06320621468926553,
      "grad_norm": 1.3906699419021606,
      "learning_rate": 3.2200000000000003e-05,
      "loss": 1.1366,
      "step": 179
    },
    {
      "epoch": 0.0635593220338983,
      "grad_norm": 1.2436065673828125,
      "learning_rate": 3.21e-05,
      "loss": 1.0769,
      "step": 180
    },
    {
      "epoch": 0.06391242937853107,
      "grad_norm": 1.137102484703064,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.9609,
      "step": 181
    },
    {
      "epoch": 0.06426553672316385,
      "grad_norm": 1.7101986408233643,
      "learning_rate": 3.19e-05,
      "loss": 1.138,
      "step": 182
    },
    {
      "epoch": 0.0646186440677966,
      "grad_norm": 1.1683790683746338,
      "learning_rate": 3.18e-05,
      "loss": 1.1872,
      "step": 183
    },
    {
      "epoch": 0.06497175141242938,
      "grad_norm": 1.6510800123214722,
      "learning_rate": 3.1700000000000005e-05,
      "loss": 1.0618,
      "step": 184
    },
    {
      "epoch": 0.06532485875706215,
      "grad_norm": 1.5926707983016968,
      "learning_rate": 3.16e-05,
      "loss": 1.0558,
      "step": 185
    },
    {
      "epoch": 0.06567796610169492,
      "grad_norm": 1.502793312072754,
      "learning_rate": 3.15e-05,
      "loss": 0.9155,
      "step": 186
    },
    {
      "epoch": 0.06603107344632768,
      "grad_norm": 1.5513490438461304,
      "learning_rate": 3.1400000000000004e-05,
      "loss": 1.1119,
      "step": 187
    },
    {
      "epoch": 0.06638418079096045,
      "grad_norm": 1.7766902446746826,
      "learning_rate": 3.13e-05,
      "loss": 1.0724,
      "step": 188
    },
    {
      "epoch": 0.06673728813559322,
      "grad_norm": 1.6195155382156372,
      "learning_rate": 3.12e-05,
      "loss": 1.0377,
      "step": 189
    },
    {
      "epoch": 0.067090395480226,
      "grad_norm": 1.5613833665847778,
      "learning_rate": 3.1100000000000004e-05,
      "loss": 1.0057,
      "step": 190
    },
    {
      "epoch": 0.06744350282485875,
      "grad_norm": 1.5648159980773926,
      "learning_rate": 3.1e-05,
      "loss": 0.9758,
      "step": 191
    },
    {
      "epoch": 0.06779661016949153,
      "grad_norm": 2.3179798126220703,
      "learning_rate": 3.09e-05,
      "loss": 1.076,
      "step": 192
    },
    {
      "epoch": 0.0681497175141243,
      "grad_norm": 1.6775708198547363,
      "learning_rate": 3.08e-05,
      "loss": 0.8133,
      "step": 193
    },
    {
      "epoch": 0.06850282485875706,
      "grad_norm": 1.6693114042282104,
      "learning_rate": 3.07e-05,
      "loss": 0.8958,
      "step": 194
    },
    {
      "epoch": 0.06885593220338983,
      "grad_norm": 1.623583436012268,
      "learning_rate": 3.06e-05,
      "loss": 0.9226,
      "step": 195
    },
    {
      "epoch": 0.0692090395480226,
      "grad_norm": 1.6135281324386597,
      "learning_rate": 3.05e-05,
      "loss": 0.8707,
      "step": 196
    },
    {
      "epoch": 0.06956214689265537,
      "grad_norm": 1.6629116535186768,
      "learning_rate": 3.04e-05,
      "loss": 0.8883,
      "step": 197
    },
    {
      "epoch": 0.06991525423728813,
      "grad_norm": 1.9516435861587524,
      "learning_rate": 3.03e-05,
      "loss": 0.8909,
      "step": 198
    },
    {
      "epoch": 0.0702683615819209,
      "grad_norm": 1.9378143548965454,
      "learning_rate": 3.02e-05,
      "loss": 0.7697,
      "step": 199
    },
    {
      "epoch": 0.07062146892655367,
      "grad_norm": 2.0332353115081787,
      "learning_rate": 3.01e-05,
      "loss": 0.9999,
      "step": 200
    },
    {
      "epoch": 0.07097457627118645,
      "grad_norm": 1.4929075241088867,
      "learning_rate": 3e-05,
      "loss": 0.8301,
      "step": 201
    },
    {
      "epoch": 0.0713276836158192,
      "grad_norm": 1.9575985670089722,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 1.0613,
      "step": 202
    },
    {
      "epoch": 0.07168079096045198,
      "grad_norm": 1.3801037073135376,
      "learning_rate": 2.98e-05,
      "loss": 1.0363,
      "step": 203
    },
    {
      "epoch": 0.07203389830508475,
      "grad_norm": 1.8629417419433594,
      "learning_rate": 2.97e-05,
      "loss": 0.8268,
      "step": 204
    },
    {
      "epoch": 0.07238700564971752,
      "grad_norm": 1.5077078342437744,
      "learning_rate": 2.96e-05,
      "loss": 0.9769,
      "step": 205
    },
    {
      "epoch": 0.07274011299435028,
      "grad_norm": 1.8674854040145874,
      "learning_rate": 2.95e-05,
      "loss": 1.011,
      "step": 206
    },
    {
      "epoch": 0.07309322033898305,
      "grad_norm": 2.605397939682007,
      "learning_rate": 2.94e-05,
      "loss": 1.2138,
      "step": 207
    },
    {
      "epoch": 0.07344632768361582,
      "grad_norm": 1.3890961408615112,
      "learning_rate": 2.93e-05,
      "loss": 0.9088,
      "step": 208
    },
    {
      "epoch": 0.0737994350282486,
      "grad_norm": 2.1312317848205566,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 1.2366,
      "step": 209
    },
    {
      "epoch": 0.07415254237288135,
      "grad_norm": 1.817219853401184,
      "learning_rate": 2.91e-05,
      "loss": 0.7755,
      "step": 210
    },
    {
      "epoch": 0.07450564971751412,
      "grad_norm": 1.8591907024383545,
      "learning_rate": 2.9e-05,
      "loss": 0.9953,
      "step": 211
    },
    {
      "epoch": 0.0748587570621469,
      "grad_norm": 1.7456353902816772,
      "learning_rate": 2.8899999999999998e-05,
      "loss": 0.7605,
      "step": 212
    },
    {
      "epoch": 0.07521186440677965,
      "grad_norm": 1.6678094863891602,
      "learning_rate": 2.88e-05,
      "loss": 0.779,
      "step": 213
    },
    {
      "epoch": 0.07556497175141243,
      "grad_norm": 1.7547659873962402,
      "learning_rate": 2.87e-05,
      "loss": 0.7277,
      "step": 214
    },
    {
      "epoch": 0.0759180790960452,
      "grad_norm": 2.3935739994049072,
      "learning_rate": 2.86e-05,
      "loss": 1.0243,
      "step": 215
    },
    {
      "epoch": 0.07627118644067797,
      "grad_norm": 2.1737802028656006,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 0.8159,
      "step": 216
    },
    {
      "epoch": 0.07662429378531073,
      "grad_norm": 2.551464557647705,
      "learning_rate": 2.84e-05,
      "loss": 0.7808,
      "step": 217
    },
    {
      "epoch": 0.0769774011299435,
      "grad_norm": 2.0867929458618164,
      "learning_rate": 2.83e-05,
      "loss": 0.7764,
      "step": 218
    },
    {
      "epoch": 0.07733050847457627,
      "grad_norm": 2.001920700073242,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 0.7499,
      "step": 219
    },
    {
      "epoch": 0.07768361581920905,
      "grad_norm": 2.4141225814819336,
      "learning_rate": 2.8100000000000005e-05,
      "loss": 0.7964,
      "step": 220
    },
    {
      "epoch": 0.0780367231638418,
      "grad_norm": 2.371626615524292,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.8997,
      "step": 221
    },
    {
      "epoch": 0.07838983050847458,
      "grad_norm": 1.6987398862838745,
      "learning_rate": 2.7900000000000004e-05,
      "loss": 0.9769,
      "step": 222
    },
    {
      "epoch": 0.07874293785310735,
      "grad_norm": 1.768283486366272,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 0.6651,
      "step": 223
    },
    {
      "epoch": 0.07909604519774012,
      "grad_norm": 1.827958106994629,
      "learning_rate": 2.7700000000000002e-05,
      "loss": 0.6805,
      "step": 224
    },
    {
      "epoch": 0.07944915254237288,
      "grad_norm": 2.7550013065338135,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 0.8034,
      "step": 225
    },
    {
      "epoch": 0.07980225988700565,
      "grad_norm": 1.7423927783966064,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.7345,
      "step": 226
    },
    {
      "epoch": 0.08015536723163842,
      "grad_norm": 1.9607951641082764,
      "learning_rate": 2.7400000000000002e-05,
      "loss": 0.6771,
      "step": 227
    },
    {
      "epoch": 0.08050847457627118,
      "grad_norm": 2.737659215927124,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 1.1056,
      "step": 228
    },
    {
      "epoch": 0.08086158192090395,
      "grad_norm": 3.3492250442504883,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 1.2373,
      "step": 229
    },
    {
      "epoch": 0.08121468926553672,
      "grad_norm": 2.10353946685791,
      "learning_rate": 2.7100000000000005e-05,
      "loss": 0.8509,
      "step": 230
    },
    {
      "epoch": 0.0815677966101695,
      "grad_norm": 2.1462435722351074,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.8893,
      "step": 231
    },
    {
      "epoch": 0.08192090395480225,
      "grad_norm": 2.2392938137054443,
      "learning_rate": 2.6900000000000003e-05,
      "loss": 0.8682,
      "step": 232
    },
    {
      "epoch": 0.08227401129943503,
      "grad_norm": 2.242868661880493,
      "learning_rate": 2.6800000000000004e-05,
      "loss": 0.812,
      "step": 233
    },
    {
      "epoch": 0.0826271186440678,
      "grad_norm": 1.9220484495162964,
      "learning_rate": 2.6700000000000002e-05,
      "loss": 0.7693,
      "step": 234
    },
    {
      "epoch": 0.08298022598870057,
      "grad_norm": 1.5621178150177002,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 0.875,
      "step": 235
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 1.8653405904769897,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.6442,
      "step": 236
    },
    {
      "epoch": 0.0836864406779661,
      "grad_norm": 2.035147190093994,
      "learning_rate": 2.64e-05,
      "loss": 0.8103,
      "step": 237
    },
    {
      "epoch": 0.08403954802259887,
      "grad_norm": 2.0776431560516357,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 0.8276,
      "step": 238
    },
    {
      "epoch": 0.08439265536723164,
      "grad_norm": 1.745665192604065,
      "learning_rate": 2.6200000000000003e-05,
      "loss": 0.6145,
      "step": 239
    },
    {
      "epoch": 0.0847457627118644,
      "grad_norm": 1.9675729274749756,
      "learning_rate": 2.61e-05,
      "loss": 0.7005,
      "step": 240
    },
    {
      "epoch": 0.08509887005649718,
      "grad_norm": 2.2235677242279053,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.5884,
      "step": 241
    },
    {
      "epoch": 0.08545197740112995,
      "grad_norm": 1.8204203844070435,
      "learning_rate": 2.5900000000000003e-05,
      "loss": 0.6029,
      "step": 242
    },
    {
      "epoch": 0.0858050847457627,
      "grad_norm": 1.6106022596359253,
      "learning_rate": 2.58e-05,
      "loss": 0.5933,
      "step": 243
    },
    {
      "epoch": 0.08615819209039548,
      "grad_norm": 1.8598597049713135,
      "learning_rate": 2.57e-05,
      "loss": 0.6792,
      "step": 244
    },
    {
      "epoch": 0.08651129943502825,
      "grad_norm": 2.186007499694824,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 0.8079,
      "step": 245
    },
    {
      "epoch": 0.08686440677966102,
      "grad_norm": 2.206789255142212,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 0.936,
      "step": 246
    },
    {
      "epoch": 0.08721751412429378,
      "grad_norm": 1.865030288696289,
      "learning_rate": 2.54e-05,
      "loss": 0.8269,
      "step": 247
    },
    {
      "epoch": 0.08757062146892655,
      "grad_norm": 3.2279410362243652,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 0.9072,
      "step": 248
    },
    {
      "epoch": 0.08792372881355932,
      "grad_norm": 1.5364686250686646,
      "learning_rate": 2.5200000000000003e-05,
      "loss": 0.5486,
      "step": 249
    },
    {
      "epoch": 0.0882768361581921,
      "grad_norm": 1.7777973413467407,
      "learning_rate": 2.51e-05,
      "loss": 0.5444,
      "step": 250
    },
    {
      "epoch": 0.08862994350282485,
      "grad_norm": 2.1404268741607666,
      "learning_rate": 2.5e-05,
      "loss": 0.6368,
      "step": 251
    },
    {
      "epoch": 0.08898305084745763,
      "grad_norm": 1.80747389793396,
      "learning_rate": 2.4900000000000002e-05,
      "loss": 0.6483,
      "step": 252
    },
    {
      "epoch": 0.0893361581920904,
      "grad_norm": 1.780600666999817,
      "learning_rate": 2.48e-05,
      "loss": 0.7834,
      "step": 253
    },
    {
      "epoch": 0.08968926553672317,
      "grad_norm": 2.232827663421631,
      "learning_rate": 2.47e-05,
      "loss": 0.7617,
      "step": 254
    },
    {
      "epoch": 0.09004237288135593,
      "grad_norm": 1.9321049451828003,
      "learning_rate": 2.46e-05,
      "loss": 0.5936,
      "step": 255
    },
    {
      "epoch": 0.0903954802259887,
      "grad_norm": 1.7767024040222168,
      "learning_rate": 2.45e-05,
      "loss": 0.5475,
      "step": 256
    },
    {
      "epoch": 0.09074858757062147,
      "grad_norm": 1.7460103034973145,
      "learning_rate": 2.44e-05,
      "loss": 0.7268,
      "step": 257
    },
    {
      "epoch": 0.09110169491525423,
      "grad_norm": 2.010267734527588,
      "learning_rate": 2.43e-05,
      "loss": 0.5198,
      "step": 258
    },
    {
      "epoch": 0.091454802259887,
      "grad_norm": 2.2814278602600098,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.8076,
      "step": 259
    },
    {
      "epoch": 0.09180790960451977,
      "grad_norm": 2.043018341064453,
      "learning_rate": 2.41e-05,
      "loss": 0.7212,
      "step": 260
    },
    {
      "epoch": 0.09216101694915255,
      "grad_norm": 1.5442239046096802,
      "learning_rate": 2.4e-05,
      "loss": 0.6238,
      "step": 261
    },
    {
      "epoch": 0.0925141242937853,
      "grad_norm": 2.0493950843811035,
      "learning_rate": 2.39e-05,
      "loss": 0.6279,
      "step": 262
    },
    {
      "epoch": 0.09286723163841808,
      "grad_norm": 1.9058549404144287,
      "learning_rate": 2.38e-05,
      "loss": 0.6636,
      "step": 263
    },
    {
      "epoch": 0.09322033898305085,
      "grad_norm": 1.766516089439392,
      "learning_rate": 2.37e-05,
      "loss": 0.7014,
      "step": 264
    },
    {
      "epoch": 0.09357344632768362,
      "grad_norm": 1.7538968324661255,
      "learning_rate": 2.36e-05,
      "loss": 0.7425,
      "step": 265
    },
    {
      "epoch": 0.09392655367231638,
      "grad_norm": 1.9766415357589722,
      "learning_rate": 2.35e-05,
      "loss": 0.6857,
      "step": 266
    },
    {
      "epoch": 0.09427966101694915,
      "grad_norm": 1.8969347476959229,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 0.509,
      "step": 267
    },
    {
      "epoch": 0.09463276836158192,
      "grad_norm": 2.556140422821045,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 0.9246,
      "step": 268
    },
    {
      "epoch": 0.0949858757062147,
      "grad_norm": 2.192314624786377,
      "learning_rate": 2.32e-05,
      "loss": 0.8475,
      "step": 269
    },
    {
      "epoch": 0.09533898305084745,
      "grad_norm": 1.6340404748916626,
      "learning_rate": 2.3100000000000002e-05,
      "loss": 0.8045,
      "step": 270
    },
    {
      "epoch": 0.09569209039548023,
      "grad_norm": 1.849310040473938,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.6461,
      "step": 271
    },
    {
      "epoch": 0.096045197740113,
      "grad_norm": 1.7643516063690186,
      "learning_rate": 2.29e-05,
      "loss": 0.8039,
      "step": 272
    },
    {
      "epoch": 0.09639830508474577,
      "grad_norm": 2.249030351638794,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 0.6257,
      "step": 273
    },
    {
      "epoch": 0.09675141242937853,
      "grad_norm": 1.8955878019332886,
      "learning_rate": 2.2700000000000003e-05,
      "loss": 0.4904,
      "step": 274
    },
    {
      "epoch": 0.0971045197740113,
      "grad_norm": 2.037266254425049,
      "learning_rate": 2.26e-05,
      "loss": 0.9272,
      "step": 275
    },
    {
      "epoch": 0.09745762711864407,
      "grad_norm": 2.061370611190796,
      "learning_rate": 2.25e-05,
      "loss": 0.5984,
      "step": 276
    },
    {
      "epoch": 0.09781073446327683,
      "grad_norm": 3.4211597442626953,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 0.9706,
      "step": 277
    },
    {
      "epoch": 0.0981638418079096,
      "grad_norm": 2.7527401447296143,
      "learning_rate": 2.23e-05,
      "loss": 0.7066,
      "step": 278
    },
    {
      "epoch": 0.09851694915254237,
      "grad_norm": 1.8547669649124146,
      "learning_rate": 2.22e-05,
      "loss": 0.458,
      "step": 279
    },
    {
      "epoch": 0.09887005649717515,
      "grad_norm": 1.9559056758880615,
      "learning_rate": 2.2100000000000002e-05,
      "loss": 0.453,
      "step": 280
    },
    {
      "epoch": 0.0992231638418079,
      "grad_norm": 1.7714954614639282,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.5063,
      "step": 281
    },
    {
      "epoch": 0.09957627118644068,
      "grad_norm": 3.084165334701538,
      "learning_rate": 2.19e-05,
      "loss": 0.8611,
      "step": 282
    },
    {
      "epoch": 0.09992937853107345,
      "grad_norm": 1.8296337127685547,
      "learning_rate": 2.18e-05,
      "loss": 0.6884,
      "step": 283
    },
    {
      "epoch": 0.10028248587570622,
      "grad_norm": 2.5133588314056396,
      "learning_rate": 2.1700000000000002e-05,
      "loss": 0.6772,
      "step": 284
    },
    {
      "epoch": 0.10063559322033898,
      "grad_norm": 1.612042784690857,
      "learning_rate": 2.16e-05,
      "loss": 0.4726,
      "step": 285
    },
    {
      "epoch": 0.10098870056497175,
      "grad_norm": 1.9312748908996582,
      "learning_rate": 2.15e-05,
      "loss": 0.6412,
      "step": 286
    },
    {
      "epoch": 0.10134180790960452,
      "grad_norm": 2.4446702003479004,
      "learning_rate": 2.1400000000000002e-05,
      "loss": 0.8463,
      "step": 287
    },
    {
      "epoch": 0.1016949152542373,
      "grad_norm": 4.253321647644043,
      "learning_rate": 2.13e-05,
      "loss": 0.6177,
      "step": 288
    },
    {
      "epoch": 0.10204802259887005,
      "grad_norm": 2.584301471710205,
      "learning_rate": 2.12e-05,
      "loss": 0.7182,
      "step": 289
    },
    {
      "epoch": 0.10240112994350282,
      "grad_norm": 1.7066371440887451,
      "learning_rate": 2.11e-05,
      "loss": 0.5563,
      "step": 290
    },
    {
      "epoch": 0.1027542372881356,
      "grad_norm": 3.1515045166015625,
      "learning_rate": 2.1e-05,
      "loss": 0.8663,
      "step": 291
    },
    {
      "epoch": 0.10310734463276836,
      "grad_norm": 2.3074467182159424,
      "learning_rate": 2.09e-05,
      "loss": 0.8209,
      "step": 292
    },
    {
      "epoch": 0.10346045197740113,
      "grad_norm": 2.0748472213745117,
      "learning_rate": 2.08e-05,
      "loss": 0.6029,
      "step": 293
    },
    {
      "epoch": 0.1038135593220339,
      "grad_norm": 1.8115911483764648,
      "learning_rate": 2.07e-05,
      "loss": 0.4101,
      "step": 294
    },
    {
      "epoch": 0.10416666666666667,
      "grad_norm": 4.644301891326904,
      "learning_rate": 2.06e-05,
      "loss": 0.8823,
      "step": 295
    },
    {
      "epoch": 0.10451977401129943,
      "grad_norm": 1.8356761932373047,
      "learning_rate": 2.05e-05,
      "loss": 0.6229,
      "step": 296
    },
    {
      "epoch": 0.1048728813559322,
      "grad_norm": 2.584825038909912,
      "learning_rate": 2.04e-05,
      "loss": 0.7403,
      "step": 297
    },
    {
      "epoch": 0.10522598870056497,
      "grad_norm": 2.1190598011016846,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 0.5843,
      "step": 298
    },
    {
      "epoch": 0.10557909604519775,
      "grad_norm": 2.0052149295806885,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 0.5476,
      "step": 299
    },
    {
      "epoch": 0.1059322033898305,
      "grad_norm": 2.5021302700042725,
      "learning_rate": 2.01e-05,
      "loss": 0.6322,
      "step": 300
    },
    {
      "epoch": 0.10628531073446328,
      "grad_norm": 2.790252208709717,
      "learning_rate": 2e-05,
      "loss": 0.7356,
      "step": 301
    },
    {
      "epoch": 0.10663841807909605,
      "grad_norm": 1.896161437034607,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 0.6414,
      "step": 302
    },
    {
      "epoch": 0.10699152542372882,
      "grad_norm": 2.121238946914673,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 0.4369,
      "step": 303
    },
    {
      "epoch": 0.10734463276836158,
      "grad_norm": 2.465498685836792,
      "learning_rate": 1.97e-05,
      "loss": 0.7176,
      "step": 304
    },
    {
      "epoch": 0.10769774011299435,
      "grad_norm": 2.068796396255493,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.4301,
      "step": 305
    },
    {
      "epoch": 0.10805084745762712,
      "grad_norm": 1.9807084798812866,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 0.4117,
      "step": 306
    },
    {
      "epoch": 0.10840395480225988,
      "grad_norm": 2.1290664672851562,
      "learning_rate": 1.94e-05,
      "loss": 0.636,
      "step": 307
    },
    {
      "epoch": 0.10875706214689265,
      "grad_norm": 1.9659547805786133,
      "learning_rate": 1.93e-05,
      "loss": 0.51,
      "step": 308
    },
    {
      "epoch": 0.10911016949152542,
      "grad_norm": 2.1764960289001465,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.6201,
      "step": 309
    },
    {
      "epoch": 0.1094632768361582,
      "grad_norm": 2.654648542404175,
      "learning_rate": 1.91e-05,
      "loss": 0.7892,
      "step": 310
    },
    {
      "epoch": 0.10981638418079095,
      "grad_norm": 1.8828105926513672,
      "learning_rate": 1.9e-05,
      "loss": 0.6453,
      "step": 311
    },
    {
      "epoch": 0.11016949152542373,
      "grad_norm": 2.0999748706817627,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 0.7547,
      "step": 312
    },
    {
      "epoch": 0.1105225988700565,
      "grad_norm": 1.9529651403427124,
      "learning_rate": 1.88e-05,
      "loss": 0.3852,
      "step": 313
    },
    {
      "epoch": 0.11087570621468927,
      "grad_norm": 2.1099841594696045,
      "learning_rate": 1.87e-05,
      "loss": 0.8429,
      "step": 314
    },
    {
      "epoch": 0.11122881355932203,
      "grad_norm": 2.6168792247772217,
      "learning_rate": 1.86e-05,
      "loss": 0.8384,
      "step": 315
    },
    {
      "epoch": 0.1115819209039548,
      "grad_norm": 1.8378658294677734,
      "learning_rate": 1.85e-05,
      "loss": 0.5949,
      "step": 316
    },
    {
      "epoch": 0.11193502824858757,
      "grad_norm": 2.333801746368408,
      "learning_rate": 1.84e-05,
      "loss": 0.4418,
      "step": 317
    },
    {
      "epoch": 0.11228813559322035,
      "grad_norm": 2.6359400749206543,
      "learning_rate": 1.83e-05,
      "loss": 0.7196,
      "step": 318
    },
    {
      "epoch": 0.1126412429378531,
      "grad_norm": 1.830614447593689,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 0.5038,
      "step": 319
    },
    {
      "epoch": 0.11299435028248588,
      "grad_norm": 1.8708122968673706,
      "learning_rate": 1.81e-05,
      "loss": 0.356,
      "step": 320
    },
    {
      "epoch": 0.11334745762711865,
      "grad_norm": 1.8595423698425293,
      "learning_rate": 1.8e-05,
      "loss": 0.5855,
      "step": 321
    },
    {
      "epoch": 0.1137005649717514,
      "grad_norm": 2.0622501373291016,
      "learning_rate": 1.79e-05,
      "loss": 0.4762,
      "step": 322
    },
    {
      "epoch": 0.11405367231638418,
      "grad_norm": 1.8576161861419678,
      "learning_rate": 1.78e-05,
      "loss": 0.6245,
      "step": 323
    },
    {
      "epoch": 0.11440677966101695,
      "grad_norm": 2.014206886291504,
      "learning_rate": 1.77e-05,
      "loss": 0.4627,
      "step": 324
    },
    {
      "epoch": 0.11475988700564972,
      "grad_norm": 1.8058968782424927,
      "learning_rate": 1.76e-05,
      "loss": 0.3285,
      "step": 325
    },
    {
      "epoch": 0.11511299435028248,
      "grad_norm": 1.7399612665176392,
      "learning_rate": 1.75e-05,
      "loss": 0.4845,
      "step": 326
    },
    {
      "epoch": 0.11546610169491525,
      "grad_norm": 2.8591692447662354,
      "learning_rate": 1.74e-05,
      "loss": 0.7819,
      "step": 327
    },
    {
      "epoch": 0.11581920903954802,
      "grad_norm": 2.9054338932037354,
      "learning_rate": 1.73e-05,
      "loss": 0.6961,
      "step": 328
    },
    {
      "epoch": 0.1161723163841808,
      "grad_norm": 2.231865167617798,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 0.532,
      "step": 329
    },
    {
      "epoch": 0.11652542372881355,
      "grad_norm": 3.112537145614624,
      "learning_rate": 1.7100000000000002e-05,
      "loss": 0.7317,
      "step": 330
    },
    {
      "epoch": 0.11687853107344633,
      "grad_norm": 1.886364221572876,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.4278,
      "step": 331
    },
    {
      "epoch": 0.1172316384180791,
      "grad_norm": 2.2548704147338867,
      "learning_rate": 1.69e-05,
      "loss": 0.4959,
      "step": 332
    },
    {
      "epoch": 0.11758474576271187,
      "grad_norm": 1.778956651687622,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.4093,
      "step": 333
    },
    {
      "epoch": 0.11793785310734463,
      "grad_norm": 3.506840229034424,
      "learning_rate": 1.6700000000000003e-05,
      "loss": 0.889,
      "step": 334
    },
    {
      "epoch": 0.1182909604519774,
      "grad_norm": 1.7313032150268555,
      "learning_rate": 1.66e-05,
      "loss": 0.3059,
      "step": 335
    },
    {
      "epoch": 0.11864406779661017,
      "grad_norm": 2.914469003677368,
      "learning_rate": 1.65e-05,
      "loss": 0.6975,
      "step": 336
    },
    {
      "epoch": 0.11899717514124294,
      "grad_norm": 2.3646645545959473,
      "learning_rate": 1.6400000000000002e-05,
      "loss": 0.3799,
      "step": 337
    },
    {
      "epoch": 0.1193502824858757,
      "grad_norm": 2.708645820617676,
      "learning_rate": 1.63e-05,
      "loss": 0.6651,
      "step": 338
    },
    {
      "epoch": 0.11970338983050847,
      "grad_norm": 2.4956207275390625,
      "learning_rate": 1.62e-05,
      "loss": 0.5777,
      "step": 339
    },
    {
      "epoch": 0.12005649717514125,
      "grad_norm": 2.371713161468506,
      "learning_rate": 1.6100000000000002e-05,
      "loss": 0.477,
      "step": 340
    },
    {
      "epoch": 0.120409604519774,
      "grad_norm": 1.81998872756958,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.3214,
      "step": 341
    },
    {
      "epoch": 0.12076271186440678,
      "grad_norm": 2.2463088035583496,
      "learning_rate": 1.59e-05,
      "loss": 0.602,
      "step": 342
    },
    {
      "epoch": 0.12111581920903955,
      "grad_norm": 2.2001595497131348,
      "learning_rate": 1.58e-05,
      "loss": 0.6024,
      "step": 343
    },
    {
      "epoch": 0.12146892655367232,
      "grad_norm": 2.6866791248321533,
      "learning_rate": 1.5700000000000002e-05,
      "loss": 0.693,
      "step": 344
    },
    {
      "epoch": 0.12182203389830508,
      "grad_norm": 2.6667373180389404,
      "learning_rate": 1.56e-05,
      "loss": 0.6857,
      "step": 345
    },
    {
      "epoch": 0.12217514124293785,
      "grad_norm": 2.1822969913482666,
      "learning_rate": 1.55e-05,
      "loss": 0.51,
      "step": 346
    },
    {
      "epoch": 0.12252824858757062,
      "grad_norm": 2.881263256072998,
      "learning_rate": 1.54e-05,
      "loss": 0.6404,
      "step": 347
    },
    {
      "epoch": 0.1228813559322034,
      "grad_norm": 2.1057329177856445,
      "learning_rate": 1.53e-05,
      "loss": 0.495,
      "step": 348
    },
    {
      "epoch": 0.12323446327683615,
      "grad_norm": 1.6811401844024658,
      "learning_rate": 1.52e-05,
      "loss": 0.2848,
      "step": 349
    },
    {
      "epoch": 0.12358757062146893,
      "grad_norm": 2.3314602375030518,
      "learning_rate": 1.51e-05,
      "loss": 0.5686,
      "step": 350
    },
    {
      "epoch": 0.1239406779661017,
      "grad_norm": 2.8521921634674072,
      "learning_rate": 1.5e-05,
      "loss": 0.7763,
      "step": 351
    },
    {
      "epoch": 0.12429378531073447,
      "grad_norm": 2.379612922668457,
      "learning_rate": 1.49e-05,
      "loss": 0.6625,
      "step": 352
    },
    {
      "epoch": 0.12464689265536723,
      "grad_norm": 1.7267616987228394,
      "learning_rate": 1.48e-05,
      "loss": 0.3089,
      "step": 353
    },
    {
      "epoch": 0.125,
      "grad_norm": 2.078075647354126,
      "learning_rate": 1.47e-05,
      "loss": 0.5671,
      "step": 354
    },
    {
      "epoch": 0.12535310734463276,
      "grad_norm": 2.491132974624634,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 0.5108,
      "step": 355
    },
    {
      "epoch": 0.12570621468926554,
      "grad_norm": 2.2118701934814453,
      "learning_rate": 1.45e-05,
      "loss": 0.4687,
      "step": 356
    },
    {
      "epoch": 0.1260593220338983,
      "grad_norm": 1.556470274925232,
      "learning_rate": 1.44e-05,
      "loss": 0.2878,
      "step": 357
    },
    {
      "epoch": 0.12641242937853106,
      "grad_norm": 3.0352628231048584,
      "learning_rate": 1.43e-05,
      "loss": 0.7796,
      "step": 358
    },
    {
      "epoch": 0.12676553672316385,
      "grad_norm": 2.498322010040283,
      "learning_rate": 1.42e-05,
      "loss": 0.6919,
      "step": 359
    },
    {
      "epoch": 0.1271186440677966,
      "grad_norm": 2.7435500621795654,
      "learning_rate": 1.4099999999999999e-05,
      "loss": 0.6379,
      "step": 360
    },
    {
      "epoch": 0.1274717514124294,
      "grad_norm": 2.1981303691864014,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.4841,
      "step": 361
    },
    {
      "epoch": 0.12782485875706215,
      "grad_norm": 4.253507137298584,
      "learning_rate": 1.3900000000000002e-05,
      "loss": 0.8417,
      "step": 362
    },
    {
      "epoch": 0.1281779661016949,
      "grad_norm": 2.2498624324798584,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.3849,
      "step": 363
    },
    {
      "epoch": 0.1285310734463277,
      "grad_norm": 2.5313210487365723,
      "learning_rate": 1.3700000000000001e-05,
      "loss": 0.7152,
      "step": 364
    },
    {
      "epoch": 0.12888418079096045,
      "grad_norm": 2.427729368209839,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.613,
      "step": 365
    },
    {
      "epoch": 0.1292372881355932,
      "grad_norm": 2.1720805168151855,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.527,
      "step": 366
    },
    {
      "epoch": 0.129590395480226,
      "grad_norm": 2.054292678833008,
      "learning_rate": 1.3400000000000002e-05,
      "loss": 0.4659,
      "step": 367
    },
    {
      "epoch": 0.12994350282485875,
      "grad_norm": 2.235578775405884,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 0.4747,
      "step": 368
    },
    {
      "epoch": 0.13029661016949154,
      "grad_norm": 1.5781618356704712,
      "learning_rate": 1.32e-05,
      "loss": 0.2759,
      "step": 369
    },
    {
      "epoch": 0.1306497175141243,
      "grad_norm": 1.8499678373336792,
      "learning_rate": 1.3100000000000002e-05,
      "loss": 0.4529,
      "step": 370
    },
    {
      "epoch": 0.13100282485875706,
      "grad_norm": 1.901394248008728,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.5179,
      "step": 371
    },
    {
      "epoch": 0.13135593220338984,
      "grad_norm": 2.1701114177703857,
      "learning_rate": 1.29e-05,
      "loss": 0.5104,
      "step": 372
    },
    {
      "epoch": 0.1317090395480226,
      "grad_norm": 2.1720824241638184,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.3809,
      "step": 373
    },
    {
      "epoch": 0.13206214689265536,
      "grad_norm": 2.1540091037750244,
      "learning_rate": 1.27e-05,
      "loss": 0.5311,
      "step": 374
    },
    {
      "epoch": 0.13241525423728814,
      "grad_norm": 2.0049026012420654,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 0.363,
      "step": 375
    },
    {
      "epoch": 0.1327683615819209,
      "grad_norm": 2.1072335243225098,
      "learning_rate": 1.25e-05,
      "loss": 0.4218,
      "step": 376
    },
    {
      "epoch": 0.13312146892655366,
      "grad_norm": 1.9690706729888916,
      "learning_rate": 1.24e-05,
      "loss": 0.4304,
      "step": 377
    },
    {
      "epoch": 0.13347457627118645,
      "grad_norm": 1.7669113874435425,
      "learning_rate": 1.23e-05,
      "loss": 0.3686,
      "step": 378
    },
    {
      "epoch": 0.1338276836158192,
      "grad_norm": 1.9775711297988892,
      "learning_rate": 1.22e-05,
      "loss": 0.503,
      "step": 379
    },
    {
      "epoch": 0.134180790960452,
      "grad_norm": 2.458343744277954,
      "learning_rate": 1.2100000000000001e-05,
      "loss": 0.4038,
      "step": 380
    },
    {
      "epoch": 0.13453389830508475,
      "grad_norm": 1.9653310775756836,
      "learning_rate": 1.2e-05,
      "loss": 0.5042,
      "step": 381
    },
    {
      "epoch": 0.1348870056497175,
      "grad_norm": 1.832046627998352,
      "learning_rate": 1.19e-05,
      "loss": 0.3696,
      "step": 382
    },
    {
      "epoch": 0.1352401129943503,
      "grad_norm": 4.0539374351501465,
      "learning_rate": 1.18e-05,
      "loss": 0.8044,
      "step": 383
    },
    {
      "epoch": 0.13559322033898305,
      "grad_norm": 4.322567462921143,
      "learning_rate": 1.1700000000000001e-05,
      "loss": 0.9251,
      "step": 384
    },
    {
      "epoch": 0.1359463276836158,
      "grad_norm": 1.9108883142471313,
      "learning_rate": 1.16e-05,
      "loss": 0.4827,
      "step": 385
    },
    {
      "epoch": 0.1362994350282486,
      "grad_norm": 1.4768140316009521,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.2541,
      "step": 386
    },
    {
      "epoch": 0.13665254237288135,
      "grad_norm": 2.005392551422119,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 0.5762,
      "step": 387
    },
    {
      "epoch": 0.1370056497175141,
      "grad_norm": 3.0048270225524902,
      "learning_rate": 1.13e-05,
      "loss": 0.6806,
      "step": 388
    },
    {
      "epoch": 0.1373587570621469,
      "grad_norm": 2.6098363399505615,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.5409,
      "step": 389
    },
    {
      "epoch": 0.13771186440677965,
      "grad_norm": 2.0291426181793213,
      "learning_rate": 1.11e-05,
      "loss": 0.4466,
      "step": 390
    },
    {
      "epoch": 0.13806497175141244,
      "grad_norm": 1.4195945262908936,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.2425,
      "step": 391
    },
    {
      "epoch": 0.1384180790960452,
      "grad_norm": 3.1550965309143066,
      "learning_rate": 1.09e-05,
      "loss": 0.6254,
      "step": 392
    },
    {
      "epoch": 0.13877118644067796,
      "grad_norm": 1.7930920124053955,
      "learning_rate": 1.08e-05,
      "loss": 0.3466,
      "step": 393
    },
    {
      "epoch": 0.13912429378531074,
      "grad_norm": 1.9969501495361328,
      "learning_rate": 1.0700000000000001e-05,
      "loss": 0.3996,
      "step": 394
    },
    {
      "epoch": 0.1394774011299435,
      "grad_norm": 2.3298864364624023,
      "learning_rate": 1.06e-05,
      "loss": 0.4196,
      "step": 395
    },
    {
      "epoch": 0.13983050847457626,
      "grad_norm": 4.122328281402588,
      "learning_rate": 1.05e-05,
      "loss": 0.3735,
      "step": 396
    },
    {
      "epoch": 0.14018361581920905,
      "grad_norm": 2.8163208961486816,
      "learning_rate": 1.04e-05,
      "loss": 0.4261,
      "step": 397
    },
    {
      "epoch": 0.1405367231638418,
      "grad_norm": 1.7612152099609375,
      "learning_rate": 1.03e-05,
      "loss": 0.4001,
      "step": 398
    },
    {
      "epoch": 0.1408898305084746,
      "grad_norm": 2.542123317718506,
      "learning_rate": 1.02e-05,
      "loss": 0.5853,
      "step": 399
    },
    {
      "epoch": 0.14124293785310735,
      "grad_norm": 2.013018846511841,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 0.344,
      "step": 400
    },
    {
      "epoch": 0.1415960451977401,
      "grad_norm": 2.6476850509643555,
      "learning_rate": 1e-05,
      "loss": 0.5595,
      "step": 401
    },
    {
      "epoch": 0.1419491525423729,
      "grad_norm": 2.0753297805786133,
      "learning_rate": 9.900000000000002e-06,
      "loss": 0.3958,
      "step": 402
    },
    {
      "epoch": 0.14230225988700565,
      "grad_norm": 1.726521372795105,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.3896,
      "step": 403
    },
    {
      "epoch": 0.1426553672316384,
      "grad_norm": 2.0191128253936768,
      "learning_rate": 9.7e-06,
      "loss": 0.4541,
      "step": 404
    },
    {
      "epoch": 0.1430084745762712,
      "grad_norm": 2.52801775932312,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.4609,
      "step": 405
    },
    {
      "epoch": 0.14336158192090395,
      "grad_norm": 2.738588809967041,
      "learning_rate": 9.5e-06,
      "loss": 0.4313,
      "step": 406
    },
    {
      "epoch": 0.1437146892655367,
      "grad_norm": 2.0913681983947754,
      "learning_rate": 9.4e-06,
      "loss": 0.4176,
      "step": 407
    },
    {
      "epoch": 0.1440677966101695,
      "grad_norm": 1.8325021266937256,
      "learning_rate": 9.3e-06,
      "loss": 0.3419,
      "step": 408
    },
    {
      "epoch": 0.14442090395480225,
      "grad_norm": 2.245701551437378,
      "learning_rate": 9.2e-06,
      "loss": 0.4799,
      "step": 409
    },
    {
      "epoch": 0.14477401129943504,
      "grad_norm": 3.9585647583007812,
      "learning_rate": 9.100000000000001e-06,
      "loss": 0.6476,
      "step": 410
    },
    {
      "epoch": 0.1451271186440678,
      "grad_norm": 3.2401275634765625,
      "learning_rate": 9e-06,
      "loss": 0.6274,
      "step": 411
    },
    {
      "epoch": 0.14548022598870056,
      "grad_norm": 2.1975040435791016,
      "learning_rate": 8.9e-06,
      "loss": 0.4277,
      "step": 412
    },
    {
      "epoch": 0.14583333333333334,
      "grad_norm": 1.8795210123062134,
      "learning_rate": 8.8e-06,
      "loss": 0.2486,
      "step": 413
    },
    {
      "epoch": 0.1461864406779661,
      "grad_norm": 2.162196159362793,
      "learning_rate": 8.7e-06,
      "loss": 0.3372,
      "step": 414
    },
    {
      "epoch": 0.14653954802259886,
      "grad_norm": 2.1754472255706787,
      "learning_rate": 8.599999999999999e-06,
      "loss": 0.4053,
      "step": 415
    },
    {
      "epoch": 0.14689265536723164,
      "grad_norm": 2.0052478313446045,
      "learning_rate": 8.500000000000002e-06,
      "loss": 0.3775,
      "step": 416
    },
    {
      "epoch": 0.1472457627118644,
      "grad_norm": 2.1078007221221924,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.5518,
      "step": 417
    },
    {
      "epoch": 0.1475988700564972,
      "grad_norm": 2.85091495513916,
      "learning_rate": 8.3e-06,
      "loss": 0.8025,
      "step": 418
    },
    {
      "epoch": 0.14795197740112995,
      "grad_norm": 1.9279358386993408,
      "learning_rate": 8.200000000000001e-06,
      "loss": 0.4753,
      "step": 419
    },
    {
      "epoch": 0.1483050847457627,
      "grad_norm": 2.3065125942230225,
      "learning_rate": 8.1e-06,
      "loss": 0.3965,
      "step": 420
    },
    {
      "epoch": 0.1486581920903955,
      "grad_norm": 3.294403076171875,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.6205,
      "step": 421
    },
    {
      "epoch": 0.14901129943502825,
      "grad_norm": 1.7436306476593018,
      "learning_rate": 7.9e-06,
      "loss": 0.239,
      "step": 422
    },
    {
      "epoch": 0.149364406779661,
      "grad_norm": 1.420858383178711,
      "learning_rate": 7.8e-06,
      "loss": 0.2265,
      "step": 423
    },
    {
      "epoch": 0.1497175141242938,
      "grad_norm": 2.940077066421509,
      "learning_rate": 7.7e-06,
      "loss": 0.6006,
      "step": 424
    },
    {
      "epoch": 0.15007062146892655,
      "grad_norm": 2.1452486515045166,
      "learning_rate": 7.6e-06,
      "loss": 0.3333,
      "step": 425
    },
    {
      "epoch": 0.1504237288135593,
      "grad_norm": 4.092085838317871,
      "learning_rate": 7.5e-06,
      "loss": 0.7523,
      "step": 426
    },
    {
      "epoch": 0.1507768361581921,
      "grad_norm": 1.8198660612106323,
      "learning_rate": 7.4e-06,
      "loss": 0.3283,
      "step": 427
    },
    {
      "epoch": 0.15112994350282485,
      "grad_norm": 1.3757944107055664,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 0.2199,
      "step": 428
    },
    {
      "epoch": 0.15148305084745764,
      "grad_norm": 2.0032095909118652,
      "learning_rate": 7.2e-06,
      "loss": 0.3994,
      "step": 429
    },
    {
      "epoch": 0.1518361581920904,
      "grad_norm": 2.9988086223602295,
      "learning_rate": 7.1e-06,
      "loss": 0.3558,
      "step": 430
    },
    {
      "epoch": 0.15218926553672316,
      "grad_norm": 2.7509193420410156,
      "learning_rate": 7.000000000000001e-06,
      "loss": 0.5825,
      "step": 431
    },
    {
      "epoch": 0.15254237288135594,
      "grad_norm": 2.523036003112793,
      "learning_rate": 6.900000000000001e-06,
      "loss": 0.484,
      "step": 432
    },
    {
      "epoch": 0.1528954802259887,
      "grad_norm": 1.8027594089508057,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.382,
      "step": 433
    },
    {
      "epoch": 0.15324858757062146,
      "grad_norm": 2.55008602142334,
      "learning_rate": 6.700000000000001e-06,
      "loss": 0.4458,
      "step": 434
    },
    {
      "epoch": 0.15360169491525424,
      "grad_norm": 2.159984827041626,
      "learning_rate": 6.6e-06,
      "loss": 0.3597,
      "step": 435
    },
    {
      "epoch": 0.153954802259887,
      "grad_norm": 2.2070958614349365,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.483,
      "step": 436
    },
    {
      "epoch": 0.15430790960451976,
      "grad_norm": 2.283949851989746,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.3999,
      "step": 437
    },
    {
      "epoch": 0.15466101694915255,
      "grad_norm": 2.086303472518921,
      "learning_rate": 6.300000000000001e-06,
      "loss": 0.4877,
      "step": 438
    },
    {
      "epoch": 0.1550141242937853,
      "grad_norm": 2.2829809188842773,
      "learning_rate": 6.2e-06,
      "loss": 0.4839,
      "step": 439
    },
    {
      "epoch": 0.1553672316384181,
      "grad_norm": 2.9696552753448486,
      "learning_rate": 6.1e-06,
      "loss": 0.5502,
      "step": 440
    },
    {
      "epoch": 0.15572033898305085,
      "grad_norm": 2.7994322776794434,
      "learning_rate": 6e-06,
      "loss": 0.5695,
      "step": 441
    },
    {
      "epoch": 0.1560734463276836,
      "grad_norm": 1.9340434074401855,
      "learning_rate": 5.9e-06,
      "loss": 0.3296,
      "step": 442
    },
    {
      "epoch": 0.1564265536723164,
      "grad_norm": 2.5606064796447754,
      "learning_rate": 5.8e-06,
      "loss": 0.411,
      "step": 443
    },
    {
      "epoch": 0.15677966101694915,
      "grad_norm": 2.6650333404541016,
      "learning_rate": 5.7000000000000005e-06,
      "loss": 0.3179,
      "step": 444
    },
    {
      "epoch": 0.1571327683615819,
      "grad_norm": 4.912551403045654,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.8535,
      "step": 445
    },
    {
      "epoch": 0.1574858757062147,
      "grad_norm": 2.2671737670898438,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.3959,
      "step": 446
    },
    {
      "epoch": 0.15783898305084745,
      "grad_norm": 2.4728825092315674,
      "learning_rate": 5.4e-06,
      "loss": 0.3118,
      "step": 447
    },
    {
      "epoch": 0.15819209039548024,
      "grad_norm": 1.7785649299621582,
      "learning_rate": 5.3e-06,
      "loss": 0.228,
      "step": 448
    },
    {
      "epoch": 0.158545197740113,
      "grad_norm": 2.288245916366577,
      "learning_rate": 5.2e-06,
      "loss": 0.3586,
      "step": 449
    },
    {
      "epoch": 0.15889830508474576,
      "grad_norm": 1.821871042251587,
      "learning_rate": 5.1e-06,
      "loss": 0.362,
      "step": 450
    },
    {
      "epoch": 0.15925141242937854,
      "grad_norm": 1.9763644933700562,
      "learning_rate": 5e-06,
      "loss": 0.3057,
      "step": 451
    },
    {
      "epoch": 0.1596045197740113,
      "grad_norm": 1.8915050029754639,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.3815,
      "step": 452
    },
    {
      "epoch": 0.15995762711864406,
      "grad_norm": 1.787156343460083,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.2223,
      "step": 453
    },
    {
      "epoch": 0.16031073446327684,
      "grad_norm": 2.0852158069610596,
      "learning_rate": 4.7e-06,
      "loss": 0.4158,
      "step": 454
    },
    {
      "epoch": 0.1606638418079096,
      "grad_norm": 1.9223723411560059,
      "learning_rate": 4.6e-06,
      "loss": 0.3043,
      "step": 455
    },
    {
      "epoch": 0.16101694915254236,
      "grad_norm": 1.8598310947418213,
      "learning_rate": 4.5e-06,
      "loss": 0.3886,
      "step": 456
    },
    {
      "epoch": 0.16137005649717515,
      "grad_norm": 2.242039918899536,
      "learning_rate": 4.4e-06,
      "loss": 0.4703,
      "step": 457
    },
    {
      "epoch": 0.1617231638418079,
      "grad_norm": 1.7315804958343506,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 0.2178,
      "step": 458
    },
    {
      "epoch": 0.1620762711864407,
      "grad_norm": 2.204817295074463,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.3898,
      "step": 459
    },
    {
      "epoch": 0.16242937853107345,
      "grad_norm": 2.1929430961608887,
      "learning_rate": 4.1000000000000006e-06,
      "loss": 0.3549,
      "step": 460
    },
    {
      "epoch": 0.1627824858757062,
      "grad_norm": 1.7955076694488525,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.3815,
      "step": 461
    },
    {
      "epoch": 0.163135593220339,
      "grad_norm": 4.15421724319458,
      "learning_rate": 3.9e-06,
      "loss": 0.6433,
      "step": 462
    },
    {
      "epoch": 0.16348870056497175,
      "grad_norm": 1.9674700498580933,
      "learning_rate": 3.8e-06,
      "loss": 0.3425,
      "step": 463
    },
    {
      "epoch": 0.1638418079096045,
      "grad_norm": 1.909616231918335,
      "learning_rate": 3.7e-06,
      "loss": 0.343,
      "step": 464
    },
    {
      "epoch": 0.1641949152542373,
      "grad_norm": 2.229336977005005,
      "learning_rate": 3.6e-06,
      "loss": 0.4534,
      "step": 465
    },
    {
      "epoch": 0.16454802259887005,
      "grad_norm": 1.3924556970596313,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 0.2077,
      "step": 466
    },
    {
      "epoch": 0.1649011299435028,
      "grad_norm": 1.3659465312957764,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.2068,
      "step": 467
    },
    {
      "epoch": 0.1652542372881356,
      "grad_norm": 1.867235779762268,
      "learning_rate": 3.3e-06,
      "loss": 0.2963,
      "step": 468
    },
    {
      "epoch": 0.16560734463276836,
      "grad_norm": 1.332623839378357,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.2054,
      "step": 469
    },
    {
      "epoch": 0.16596045197740114,
      "grad_norm": 1.6542198657989502,
      "learning_rate": 3.1e-06,
      "loss": 0.2103,
      "step": 470
    },
    {
      "epoch": 0.1663135593220339,
      "grad_norm": 2.120319128036499,
      "learning_rate": 3e-06,
      "loss": 0.4215,
      "step": 471
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 1.8779698610305786,
      "learning_rate": 2.9e-06,
      "loss": 0.298,
      "step": 472
    },
    {
      "epoch": 0.16701977401129944,
      "grad_norm": 2.830299139022827,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.542,
      "step": 473
    },
    {
      "epoch": 0.1673728813559322,
      "grad_norm": 2.909973621368408,
      "learning_rate": 2.7e-06,
      "loss": 0.4126,
      "step": 474
    },
    {
      "epoch": 0.16772598870056496,
      "grad_norm": 4.498685359954834,
      "learning_rate": 2.6e-06,
      "loss": 0.5541,
      "step": 475
    },
    {
      "epoch": 0.16807909604519775,
      "grad_norm": 2.452082395553589,
      "learning_rate": 2.5e-06,
      "loss": 0.3653,
      "step": 476
    },
    {
      "epoch": 0.1684322033898305,
      "grad_norm": 1.9502593278884888,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.3149,
      "step": 477
    },
    {
      "epoch": 0.1687853107344633,
      "grad_norm": 2.2951347827911377,
      "learning_rate": 2.3e-06,
      "loss": 0.3794,
      "step": 478
    },
    {
      "epoch": 0.16913841807909605,
      "grad_norm": 3.0120177268981934,
      "learning_rate": 2.2e-06,
      "loss": 0.5245,
      "step": 479
    },
    {
      "epoch": 0.1694915254237288,
      "grad_norm": 4.143613815307617,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.6444,
      "step": 480
    },
    {
      "epoch": 0.1698446327683616,
      "grad_norm": 1.838753342628479,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.3379,
      "step": 481
    },
    {
      "epoch": 0.17019774011299435,
      "grad_norm": 1.9792391061782837,
      "learning_rate": 1.9e-06,
      "loss": 0.3329,
      "step": 482
    },
    {
      "epoch": 0.1705508474576271,
      "grad_norm": 1.8502469062805176,
      "learning_rate": 1.8e-06,
      "loss": 0.3703,
      "step": 483
    },
    {
      "epoch": 0.1709039548022599,
      "grad_norm": 1.618812918663025,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.2051,
      "step": 484
    },
    {
      "epoch": 0.17125706214689265,
      "grad_norm": 2.346790075302124,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.49,
      "step": 485
    },
    {
      "epoch": 0.1716101694915254,
      "grad_norm": 1.8763465881347656,
      "learning_rate": 1.5e-06,
      "loss": 0.2948,
      "step": 486
    },
    {
      "epoch": 0.1719632768361582,
      "grad_norm": 1.7063709497451782,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.3513,
      "step": 487
    },
    {
      "epoch": 0.17231638418079095,
      "grad_norm": 2.963041067123413,
      "learning_rate": 1.3e-06,
      "loss": 0.6783,
      "step": 488
    },
    {
      "epoch": 0.17266949152542374,
      "grad_norm": 2.9806196689605713,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.5123,
      "step": 489
    },
    {
      "epoch": 0.1730225988700565,
      "grad_norm": 1.6302564144134521,
      "learning_rate": 1.1e-06,
      "loss": 0.204,
      "step": 490
    },
    {
      "epoch": 0.17337570621468926,
      "grad_norm": 1.9233272075653076,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.3037,
      "step": 491
    },
    {
      "epoch": 0.17372881355932204,
      "grad_norm": 1.8432778120040894,
      "learning_rate": 9e-07,
      "loss": 0.3368,
      "step": 492
    },
    {
      "epoch": 0.1740819209039548,
      "grad_norm": 2.5116734504699707,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.5263,
      "step": 493
    },
    {
      "epoch": 0.17443502824858756,
      "grad_norm": 3.0080406665802,
      "learning_rate": 7.000000000000001e-07,
      "loss": 0.3415,
      "step": 494
    },
    {
      "epoch": 0.17478813559322035,
      "grad_norm": 1.982960820198059,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.459,
      "step": 495
    },
    {
      "epoch": 0.1751412429378531,
      "grad_norm": 2.259504795074463,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.4088,
      "step": 496
    },
    {
      "epoch": 0.1754943502824859,
      "grad_norm": 1.937008261680603,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.2903,
      "step": 497
    },
    {
      "epoch": 0.17584745762711865,
      "grad_norm": 1.8799011707305908,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.2905,
      "step": 498
    },
    {
      "epoch": 0.1762005649717514,
      "grad_norm": 2.9519193172454834,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 0.525,
      "step": 499
    },
    {
      "epoch": 0.1765536723163842,
      "grad_norm": 1.2929198741912842,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 0.1987,
      "step": 500
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "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": 4.099784700491894e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}