amburger66's picture
LoRA fine-tune on RobotSmith task02 reaching
933e925 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5714285714285714,
"eval_steps": 50,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0005714285714285715,
"grad_norm": 30.909149169921875,
"learning_rate": 0.0,
"loss": 2.3983,
"step": 1
},
{
"epoch": 0.001142857142857143,
"grad_norm": 20.294902801513672,
"learning_rate": 2.0000000000000002e-07,
"loss": 2.1261,
"step": 2
},
{
"epoch": 0.0017142857142857142,
"grad_norm": 9.14809513092041,
"learning_rate": 4.0000000000000003e-07,
"loss": 0.9738,
"step": 3
},
{
"epoch": 0.002285714285714286,
"grad_norm": 23.925573348999023,
"learning_rate": 6.000000000000001e-07,
"loss": 2.3375,
"step": 4
},
{
"epoch": 0.002857142857142857,
"grad_norm": 12.694210052490234,
"learning_rate": 8.000000000000001e-07,
"loss": 1.0047,
"step": 5
},
{
"epoch": 0.0034285714285714284,
"grad_norm": 11.333460807800293,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.989,
"step": 6
},
{
"epoch": 0.004,
"grad_norm": 13.627532005310059,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.1729,
"step": 7
},
{
"epoch": 0.004571428571428572,
"grad_norm": 11.870243072509766,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.1999,
"step": 8
},
{
"epoch": 0.005142857142857143,
"grad_norm": 13.33997917175293,
"learning_rate": 1.6000000000000001e-06,
"loss": 0.7093,
"step": 9
},
{
"epoch": 0.005714285714285714,
"grad_norm": 14.159869194030762,
"learning_rate": 1.8000000000000001e-06,
"loss": 1.6969,
"step": 10
},
{
"epoch": 0.006285714285714286,
"grad_norm": 15.325421333312988,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.3447,
"step": 11
},
{
"epoch": 0.006857142857142857,
"grad_norm": 11.186184883117676,
"learning_rate": 2.2e-06,
"loss": 1.342,
"step": 12
},
{
"epoch": 0.0074285714285714285,
"grad_norm": 10.01712703704834,
"learning_rate": 2.4000000000000003e-06,
"loss": 0.7386,
"step": 13
},
{
"epoch": 0.008,
"grad_norm": 6.775805950164795,
"learning_rate": 2.6e-06,
"loss": 0.4932,
"step": 14
},
{
"epoch": 0.008571428571428572,
"grad_norm": 10.35350227355957,
"learning_rate": 2.8000000000000003e-06,
"loss": 0.7022,
"step": 15
},
{
"epoch": 0.009142857142857144,
"grad_norm": 12.818032264709473,
"learning_rate": 3e-06,
"loss": 0.9708,
"step": 16
},
{
"epoch": 0.009714285714285713,
"grad_norm": 15.664475440979004,
"learning_rate": 3.2000000000000003e-06,
"loss": 0.8846,
"step": 17
},
{
"epoch": 0.010285714285714285,
"grad_norm": 13.012126922607422,
"learning_rate": 3.4000000000000005e-06,
"loss": 0.5485,
"step": 18
},
{
"epoch": 0.010857142857142857,
"grad_norm": 16.6413631439209,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.6856,
"step": 19
},
{
"epoch": 0.011428571428571429,
"grad_norm": 19.3553409576416,
"learning_rate": 3.8000000000000005e-06,
"loss": 1.5569,
"step": 20
},
{
"epoch": 0.012,
"grad_norm": 25.6441707611084,
"learning_rate": 4.000000000000001e-06,
"loss": 2.0246,
"step": 21
},
{
"epoch": 0.012571428571428572,
"grad_norm": 14.960867881774902,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.5696,
"step": 22
},
{
"epoch": 0.013142857142857144,
"grad_norm": 10.267494201660156,
"learning_rate": 4.4e-06,
"loss": 0.7388,
"step": 23
},
{
"epoch": 0.013714285714285714,
"grad_norm": 13.805790901184082,
"learning_rate": 4.600000000000001e-06,
"loss": 1.2968,
"step": 24
},
{
"epoch": 0.014285714285714285,
"grad_norm": 18.01674461364746,
"learning_rate": 4.800000000000001e-06,
"loss": 1.7857,
"step": 25
},
{
"epoch": 0.014857142857142857,
"grad_norm": 14.632944107055664,
"learning_rate": 5e-06,
"loss": 1.0962,
"step": 26
},
{
"epoch": 0.015428571428571429,
"grad_norm": 13.173576354980469,
"learning_rate": 5.2e-06,
"loss": 0.5657,
"step": 27
},
{
"epoch": 0.016,
"grad_norm": 10.291979789733887,
"learning_rate": 5.400000000000001e-06,
"loss": 0.6613,
"step": 28
},
{
"epoch": 0.01657142857142857,
"grad_norm": 16.297080993652344,
"learning_rate": 5.600000000000001e-06,
"loss": 0.783,
"step": 29
},
{
"epoch": 0.017142857142857144,
"grad_norm": 20.519371032714844,
"learning_rate": 5.8e-06,
"loss": 2.1529,
"step": 30
},
{
"epoch": 0.017714285714285714,
"grad_norm": 16.223588943481445,
"learning_rate": 6e-06,
"loss": 1.6447,
"step": 31
},
{
"epoch": 0.018285714285714287,
"grad_norm": 8.131673812866211,
"learning_rate": 6.200000000000001e-06,
"loss": 0.6595,
"step": 32
},
{
"epoch": 0.018857142857142857,
"grad_norm": 11.96106243133545,
"learning_rate": 6.4000000000000006e-06,
"loss": 1.5358,
"step": 33
},
{
"epoch": 0.019428571428571427,
"grad_norm": 12.273172378540039,
"learning_rate": 6.600000000000001e-06,
"loss": 1.0581,
"step": 34
},
{
"epoch": 0.02,
"grad_norm": 12.379732131958008,
"learning_rate": 6.800000000000001e-06,
"loss": 0.8671,
"step": 35
},
{
"epoch": 0.02057142857142857,
"grad_norm": 13.036276817321777,
"learning_rate": 7e-06,
"loss": 0.9137,
"step": 36
},
{
"epoch": 0.021142857142857144,
"grad_norm": 9.701287269592285,
"learning_rate": 7.2000000000000005e-06,
"loss": 0.8232,
"step": 37
},
{
"epoch": 0.021714285714285714,
"grad_norm": 14.830707550048828,
"learning_rate": 7.4e-06,
"loss": 0.9335,
"step": 38
},
{
"epoch": 0.022285714285714287,
"grad_norm": 7.232598304748535,
"learning_rate": 7.600000000000001e-06,
"loss": 0.362,
"step": 39
},
{
"epoch": 0.022857142857142857,
"grad_norm": 12.945651054382324,
"learning_rate": 7.800000000000002e-06,
"loss": 0.9652,
"step": 40
},
{
"epoch": 0.023428571428571427,
"grad_norm": 20.990615844726562,
"learning_rate": 8.000000000000001e-06,
"loss": 1.3337,
"step": 41
},
{
"epoch": 0.024,
"grad_norm": 15.247276306152344,
"learning_rate": 8.2e-06,
"loss": 0.8544,
"step": 42
},
{
"epoch": 0.02457142857142857,
"grad_norm": 12.356481552124023,
"learning_rate": 8.400000000000001e-06,
"loss": 0.3631,
"step": 43
},
{
"epoch": 0.025142857142857144,
"grad_norm": 12.7052001953125,
"learning_rate": 8.6e-06,
"loss": 1.0408,
"step": 44
},
{
"epoch": 0.025714285714285714,
"grad_norm": 13.289191246032715,
"learning_rate": 8.8e-06,
"loss": 0.6977,
"step": 45
},
{
"epoch": 0.026285714285714287,
"grad_norm": 12.667672157287598,
"learning_rate": 9e-06,
"loss": 0.7253,
"step": 46
},
{
"epoch": 0.026857142857142857,
"grad_norm": 12.897276878356934,
"learning_rate": 9.200000000000002e-06,
"loss": 1.0248,
"step": 47
},
{
"epoch": 0.027428571428571427,
"grad_norm": 6.145381927490234,
"learning_rate": 9.4e-06,
"loss": 0.5721,
"step": 48
},
{
"epoch": 0.028,
"grad_norm": 5.274697303771973,
"learning_rate": 9.600000000000001e-06,
"loss": 0.2348,
"step": 49
},
{
"epoch": 0.02857142857142857,
"grad_norm": 9.598197937011719,
"learning_rate": 9.800000000000001e-06,
"loss": 0.4528,
"step": 50
},
{
"epoch": 0.029142857142857144,
"grad_norm": 7.818594932556152,
"learning_rate": 1e-05,
"loss": 0.5808,
"step": 51
},
{
"epoch": 0.029714285714285714,
"grad_norm": 11.814608573913574,
"learning_rate": 1.02e-05,
"loss": 0.4948,
"step": 52
},
{
"epoch": 0.030285714285714287,
"grad_norm": 6.551736831665039,
"learning_rate": 1.04e-05,
"loss": 0.2274,
"step": 53
},
{
"epoch": 0.030857142857142857,
"grad_norm": 3.589810371398926,
"learning_rate": 1.0600000000000002e-05,
"loss": 0.2852,
"step": 54
},
{
"epoch": 0.03142857142857143,
"grad_norm": 4.471236228942871,
"learning_rate": 1.0800000000000002e-05,
"loss": 0.2213,
"step": 55
},
{
"epoch": 0.032,
"grad_norm": 1.8295689821243286,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.1675,
"step": 56
},
{
"epoch": 0.03257142857142857,
"grad_norm": 5.301313877105713,
"learning_rate": 1.1200000000000001e-05,
"loss": 0.2866,
"step": 57
},
{
"epoch": 0.03314285714285714,
"grad_norm": 4.540458679199219,
"learning_rate": 1.14e-05,
"loss": 0.3812,
"step": 58
},
{
"epoch": 0.03371428571428572,
"grad_norm": 6.007534980773926,
"learning_rate": 1.16e-05,
"loss": 0.6284,
"step": 59
},
{
"epoch": 0.03428571428571429,
"grad_norm": 7.267792224884033,
"learning_rate": 1.18e-05,
"loss": 0.3929,
"step": 60
},
{
"epoch": 0.03485714285714286,
"grad_norm": 3.145653486251831,
"learning_rate": 1.2e-05,
"loss": 0.3219,
"step": 61
},
{
"epoch": 0.03542857142857143,
"grad_norm": 6.697690963745117,
"learning_rate": 1.22e-05,
"loss": 0.4401,
"step": 62
},
{
"epoch": 0.036,
"grad_norm": 3.3847038745880127,
"learning_rate": 1.2400000000000002e-05,
"loss": 0.2292,
"step": 63
},
{
"epoch": 0.036571428571428574,
"grad_norm": 4.518666744232178,
"learning_rate": 1.2600000000000001e-05,
"loss": 0.2249,
"step": 64
},
{
"epoch": 0.037142857142857144,
"grad_norm": 3.9876348972320557,
"learning_rate": 1.2800000000000001e-05,
"loss": 0.2247,
"step": 65
},
{
"epoch": 0.037714285714285714,
"grad_norm": 7.24136209487915,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.2969,
"step": 66
},
{
"epoch": 0.038285714285714284,
"grad_norm": 3.0262629985809326,
"learning_rate": 1.3200000000000002e-05,
"loss": 0.1615,
"step": 67
},
{
"epoch": 0.038857142857142854,
"grad_norm": 6.103122711181641,
"learning_rate": 1.3400000000000002e-05,
"loss": 0.262,
"step": 68
},
{
"epoch": 0.03942857142857143,
"grad_norm": 5.345109939575195,
"learning_rate": 1.3600000000000002e-05,
"loss": 0.2203,
"step": 69
},
{
"epoch": 0.04,
"grad_norm": 5.043422698974609,
"learning_rate": 1.38e-05,
"loss": 0.2371,
"step": 70
},
{
"epoch": 0.04057142857142857,
"grad_norm": 2.627903461456299,
"learning_rate": 1.4e-05,
"loss": 0.081,
"step": 71
},
{
"epoch": 0.04114285714285714,
"grad_norm": 5.247978687286377,
"learning_rate": 1.4200000000000001e-05,
"loss": 0.1598,
"step": 72
},
{
"epoch": 0.04171428571428572,
"grad_norm": 2.8443515300750732,
"learning_rate": 1.4400000000000001e-05,
"loss": 0.153,
"step": 73
},
{
"epoch": 0.04228571428571429,
"grad_norm": 2.9642412662506104,
"learning_rate": 1.46e-05,
"loss": 0.0674,
"step": 74
},
{
"epoch": 0.04285714285714286,
"grad_norm": 2.0256309509277344,
"learning_rate": 1.48e-05,
"loss": 0.1526,
"step": 75
},
{
"epoch": 0.04342857142857143,
"grad_norm": 11.363311767578125,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.5037,
"step": 76
},
{
"epoch": 0.044,
"grad_norm": 1.9041038751602173,
"learning_rate": 1.5200000000000002e-05,
"loss": 0.0555,
"step": 77
},
{
"epoch": 0.044571428571428574,
"grad_norm": 6.341301441192627,
"learning_rate": 1.54e-05,
"loss": 0.3671,
"step": 78
},
{
"epoch": 0.045142857142857144,
"grad_norm": 5.188065528869629,
"learning_rate": 1.5600000000000003e-05,
"loss": 0.1099,
"step": 79
},
{
"epoch": 0.045714285714285714,
"grad_norm": 7.643540382385254,
"learning_rate": 1.58e-05,
"loss": 0.3079,
"step": 80
},
{
"epoch": 0.046285714285714284,
"grad_norm": 2.2631261348724365,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.1477,
"step": 81
},
{
"epoch": 0.046857142857142854,
"grad_norm": 7.709822654724121,
"learning_rate": 1.62e-05,
"loss": 0.2929,
"step": 82
},
{
"epoch": 0.04742857142857143,
"grad_norm": 8.411426544189453,
"learning_rate": 1.64e-05,
"loss": 0.1567,
"step": 83
},
{
"epoch": 0.048,
"grad_norm": 1.3031255006790161,
"learning_rate": 1.66e-05,
"loss": 0.0912,
"step": 84
},
{
"epoch": 0.04857142857142857,
"grad_norm": 2.7839410305023193,
"learning_rate": 1.6800000000000002e-05,
"loss": 0.0698,
"step": 85
},
{
"epoch": 0.04914285714285714,
"grad_norm": 2.203723907470703,
"learning_rate": 1.7e-05,
"loss": 0.0637,
"step": 86
},
{
"epoch": 0.04971428571428571,
"grad_norm": 1.733355164527893,
"learning_rate": 1.72e-05,
"loss": 0.0425,
"step": 87
},
{
"epoch": 0.05028571428571429,
"grad_norm": 2.262359857559204,
"learning_rate": 1.7400000000000003e-05,
"loss": 0.1445,
"step": 88
},
{
"epoch": 0.05085714285714286,
"grad_norm": 3.33724045753479,
"learning_rate": 1.76e-05,
"loss": 0.1892,
"step": 89
},
{
"epoch": 0.05142857142857143,
"grad_norm": 0.8401795625686646,
"learning_rate": 1.7800000000000002e-05,
"loss": 0.0189,
"step": 90
},
{
"epoch": 0.052,
"grad_norm": 8.2905855178833,
"learning_rate": 1.8e-05,
"loss": 0.3271,
"step": 91
},
{
"epoch": 0.052571428571428575,
"grad_norm": 6.528992176055908,
"learning_rate": 1.8200000000000002e-05,
"loss": 0.3583,
"step": 92
},
{
"epoch": 0.053142857142857144,
"grad_norm": 2.188873052597046,
"learning_rate": 1.8400000000000003e-05,
"loss": 0.1387,
"step": 93
},
{
"epoch": 0.053714285714285714,
"grad_norm": 5.40103816986084,
"learning_rate": 1.86e-05,
"loss": 0.1803,
"step": 94
},
{
"epoch": 0.054285714285714284,
"grad_norm": 2.9757156372070312,
"learning_rate": 1.88e-05,
"loss": 0.1179,
"step": 95
},
{
"epoch": 0.054857142857142854,
"grad_norm": 1.7169569730758667,
"learning_rate": 1.9e-05,
"loss": 0.1005,
"step": 96
},
{
"epoch": 0.05542857142857143,
"grad_norm": 2.179975986480713,
"learning_rate": 1.9200000000000003e-05,
"loss": 0.1208,
"step": 97
},
{
"epoch": 0.056,
"grad_norm": 9.388154983520508,
"learning_rate": 1.94e-05,
"loss": 0.62,
"step": 98
},
{
"epoch": 0.05657142857142857,
"grad_norm": 2.3877623081207275,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.1066,
"step": 99
},
{
"epoch": 0.05714285714285714,
"grad_norm": 1.3145651817321777,
"learning_rate": 1.98e-05,
"loss": 0.0801,
"step": 100
},
{
"epoch": 0.05771428571428571,
"grad_norm": 1.1048351526260376,
"learning_rate": 2e-05,
"loss": 0.0724,
"step": 101
},
{
"epoch": 0.05828571428571429,
"grad_norm": 1.5472813844680786,
"learning_rate": 1.9999939076577906e-05,
"loss": 0.0997,
"step": 102
},
{
"epoch": 0.05885714285714286,
"grad_norm": 1.6679071187973022,
"learning_rate": 1.9999756307053947e-05,
"loss": 0.0343,
"step": 103
},
{
"epoch": 0.05942857142857143,
"grad_norm": 4.251124858856201,
"learning_rate": 1.9999451693655125e-05,
"loss": 0.2327,
"step": 104
},
{
"epoch": 0.06,
"grad_norm": 1.3014769554138184,
"learning_rate": 1.9999025240093045e-05,
"loss": 0.0743,
"step": 105
},
{
"epoch": 0.060571428571428575,
"grad_norm": 2.053699016571045,
"learning_rate": 1.9998476951563914e-05,
"loss": 0.1389,
"step": 106
},
{
"epoch": 0.061142857142857145,
"grad_norm": 3.8256070613861084,
"learning_rate": 1.9997806834748455e-05,
"loss": 0.0565,
"step": 107
},
{
"epoch": 0.061714285714285715,
"grad_norm": 6.095253944396973,
"learning_rate": 1.9997014897811834e-05,
"loss": 0.1043,
"step": 108
},
{
"epoch": 0.062285714285714285,
"grad_norm": 0.6538562774658203,
"learning_rate": 1.9996101150403543e-05,
"loss": 0.0131,
"step": 109
},
{
"epoch": 0.06285714285714286,
"grad_norm": 0.6610456109046936,
"learning_rate": 1.9995065603657317e-05,
"loss": 0.0167,
"step": 110
},
{
"epoch": 0.06342857142857143,
"grad_norm": 0.5874118208885193,
"learning_rate": 1.999390827019096e-05,
"loss": 0.0424,
"step": 111
},
{
"epoch": 0.064,
"grad_norm": 3.385213851928711,
"learning_rate": 1.999262916410621e-05,
"loss": 0.1516,
"step": 112
},
{
"epoch": 0.06457142857142857,
"grad_norm": 7.320718288421631,
"learning_rate": 1.9991228300988586e-05,
"loss": 0.2522,
"step": 113
},
{
"epoch": 0.06514285714285714,
"grad_norm": 1.3559902906417847,
"learning_rate": 1.998970569790715e-05,
"loss": 0.061,
"step": 114
},
{
"epoch": 0.06571428571428571,
"grad_norm": 6.078310966491699,
"learning_rate": 1.9988061373414342e-05,
"loss": 0.1653,
"step": 115
},
{
"epoch": 0.06628571428571428,
"grad_norm": 1.9504550695419312,
"learning_rate": 1.9986295347545738e-05,
"loss": 0.1853,
"step": 116
},
{
"epoch": 0.06685714285714285,
"grad_norm": 4.628220558166504,
"learning_rate": 1.9984407641819812e-05,
"loss": 0.1841,
"step": 117
},
{
"epoch": 0.06742857142857143,
"grad_norm": 1.2002516984939575,
"learning_rate": 1.9982398279237657e-05,
"loss": 0.0872,
"step": 118
},
{
"epoch": 0.068,
"grad_norm": 0.15886422991752625,
"learning_rate": 1.9980267284282718e-05,
"loss": 0.0058,
"step": 119
},
{
"epoch": 0.06857142857142857,
"grad_norm": 8.15700626373291,
"learning_rate": 1.9978014682920503e-05,
"loss": 0.1933,
"step": 120
},
{
"epoch": 0.06914285714285714,
"grad_norm": 17.520111083984375,
"learning_rate": 1.9975640502598243e-05,
"loss": 0.4609,
"step": 121
},
{
"epoch": 0.06971428571428571,
"grad_norm": 11.971997261047363,
"learning_rate": 1.997314477224458e-05,
"loss": 0.2453,
"step": 122
},
{
"epoch": 0.07028571428571428,
"grad_norm": 2.911884069442749,
"learning_rate": 1.9970527522269204e-05,
"loss": 0.1814,
"step": 123
},
{
"epoch": 0.07085714285714285,
"grad_norm": 4.586652755737305,
"learning_rate": 1.9967788784562474e-05,
"loss": 0.0555,
"step": 124
},
{
"epoch": 0.07142857142857142,
"grad_norm": 1.1485469341278076,
"learning_rate": 1.9964928592495046e-05,
"loss": 0.0593,
"step": 125
},
{
"epoch": 0.072,
"grad_norm": 0.36329418420791626,
"learning_rate": 1.9961946980917457e-05,
"loss": 0.0083,
"step": 126
},
{
"epoch": 0.07257142857142856,
"grad_norm": 2.137957811355591,
"learning_rate": 1.9958843986159705e-05,
"loss": 0.0721,
"step": 127
},
{
"epoch": 0.07314285714285715,
"grad_norm": 1.1695330142974854,
"learning_rate": 1.99556196460308e-05,
"loss": 0.0807,
"step": 128
},
{
"epoch": 0.07371428571428572,
"grad_norm": 0.8388393521308899,
"learning_rate": 1.9952273999818312e-05,
"loss": 0.0561,
"step": 129
},
{
"epoch": 0.07428571428571429,
"grad_norm": 18.02346420288086,
"learning_rate": 1.9948807088287884e-05,
"loss": 0.1778,
"step": 130
},
{
"epoch": 0.07485714285714286,
"grad_norm": 0.14909222722053528,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.0008,
"step": 131
},
{
"epoch": 0.07542857142857143,
"grad_norm": 11.780855178833008,
"learning_rate": 1.9941509639723155e-05,
"loss": 0.4011,
"step": 132
},
{
"epoch": 0.076,
"grad_norm": 0.3538531959056854,
"learning_rate": 1.9937679191605964e-05,
"loss": 0.0057,
"step": 133
},
{
"epoch": 0.07657142857142857,
"grad_norm": 2.768669843673706,
"learning_rate": 1.9933727656003964e-05,
"loss": 0.1097,
"step": 134
},
{
"epoch": 0.07714285714285714,
"grad_norm": 0.6383763551712036,
"learning_rate": 1.992965508106537e-05,
"loss": 0.0673,
"step": 135
},
{
"epoch": 0.07771428571428571,
"grad_norm": 7.484769821166992,
"learning_rate": 1.9925461516413224e-05,
"loss": 0.1674,
"step": 136
},
{
"epoch": 0.07828571428571429,
"grad_norm": 0.20327331125736237,
"learning_rate": 1.9921147013144782e-05,
"loss": 0.0016,
"step": 137
},
{
"epoch": 0.07885714285714286,
"grad_norm": 11.761346817016602,
"learning_rate": 1.9916711623830904e-05,
"loss": 0.0838,
"step": 138
},
{
"epoch": 0.07942857142857143,
"grad_norm": 1.335802674293518,
"learning_rate": 1.991215540251542e-05,
"loss": 0.0649,
"step": 139
},
{
"epoch": 0.08,
"grad_norm": 3.0511791706085205,
"learning_rate": 1.9907478404714438e-05,
"loss": 0.0992,
"step": 140
},
{
"epoch": 0.08057142857142857,
"grad_norm": 2.190948724746704,
"learning_rate": 1.9902680687415704e-05,
"loss": 0.0256,
"step": 141
},
{
"epoch": 0.08114285714285714,
"grad_norm": 1.2879616022109985,
"learning_rate": 1.989776230907789e-05,
"loss": 0.1104,
"step": 142
},
{
"epoch": 0.08171428571428571,
"grad_norm": 3.2425875663757324,
"learning_rate": 1.9892723329629885e-05,
"loss": 0.1528,
"step": 143
},
{
"epoch": 0.08228571428571428,
"grad_norm": 0.6139562726020813,
"learning_rate": 1.988756381047006e-05,
"loss": 0.0081,
"step": 144
},
{
"epoch": 0.08285714285714285,
"grad_norm": 4.402163982391357,
"learning_rate": 1.988228381446553e-05,
"loss": 0.1535,
"step": 145
},
{
"epoch": 0.08342857142857144,
"grad_norm": 1.4351650476455688,
"learning_rate": 1.9876883405951378e-05,
"loss": 0.024,
"step": 146
},
{
"epoch": 0.084,
"grad_norm": 0.8728482723236084,
"learning_rate": 1.987136265072988e-05,
"loss": 0.0722,
"step": 147
},
{
"epoch": 0.08457142857142858,
"grad_norm": 0.10692993551492691,
"learning_rate": 1.9865721616069695e-05,
"loss": 0.0011,
"step": 148
},
{
"epoch": 0.08514285714285715,
"grad_norm": 0.1746189445257187,
"learning_rate": 1.985996037070505e-05,
"loss": 0.0025,
"step": 149
},
{
"epoch": 0.08571428571428572,
"grad_norm": 12.185018539428711,
"learning_rate": 1.9854078984834904e-05,
"loss": 0.6066,
"step": 150
},
{
"epoch": 0.08628571428571429,
"grad_norm": 15.888092994689941,
"learning_rate": 1.9848077530122083e-05,
"loss": 0.3562,
"step": 151
},
{
"epoch": 0.08685714285714285,
"grad_norm": 2.089656352996826,
"learning_rate": 1.984195607969242e-05,
"loss": 0.126,
"step": 152
},
{
"epoch": 0.08742857142857142,
"grad_norm": 0.7072321772575378,
"learning_rate": 1.983571470813386e-05,
"loss": 0.0096,
"step": 153
},
{
"epoch": 0.088,
"grad_norm": 1.8031573295593262,
"learning_rate": 1.9829353491495545e-05,
"loss": 0.024,
"step": 154
},
{
"epoch": 0.08857142857142856,
"grad_norm": 9.206666946411133,
"learning_rate": 1.982287250728689e-05,
"loss": 0.188,
"step": 155
},
{
"epoch": 0.08914285714285715,
"grad_norm": 1.1634976863861084,
"learning_rate": 1.9816271834476642e-05,
"loss": 0.0584,
"step": 156
},
{
"epoch": 0.08971428571428572,
"grad_norm": 0.49177199602127075,
"learning_rate": 1.9809551553491918e-05,
"loss": 0.053,
"step": 157
},
{
"epoch": 0.09028571428571429,
"grad_norm": 0.20078575611114502,
"learning_rate": 1.9802711746217222e-05,
"loss": 0.0019,
"step": 158
},
{
"epoch": 0.09085714285714286,
"grad_norm": 5.038985252380371,
"learning_rate": 1.979575249599344e-05,
"loss": 0.1496,
"step": 159
},
{
"epoch": 0.09142857142857143,
"grad_norm": 1.8690109252929688,
"learning_rate": 1.9788673887616852e-05,
"loss": 0.0875,
"step": 160
},
{
"epoch": 0.092,
"grad_norm": 0.7377235889434814,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.0389,
"step": 161
},
{
"epoch": 0.09257142857142857,
"grad_norm": 2.0583860874176025,
"learning_rate": 1.9774158942860962e-05,
"loss": 0.0836,
"step": 162
},
{
"epoch": 0.09314285714285714,
"grad_norm": 0.7028142809867859,
"learning_rate": 1.9766722783341682e-05,
"loss": 0.0496,
"step": 163
},
{
"epoch": 0.09371428571428571,
"grad_norm": 0.5549291968345642,
"learning_rate": 1.9759167619387474e-05,
"loss": 0.0474,
"step": 164
},
{
"epoch": 0.09428571428571429,
"grad_norm": 0.6520454287528992,
"learning_rate": 1.9751493543055634e-05,
"loss": 0.0347,
"step": 165
},
{
"epoch": 0.09485714285714286,
"grad_norm": 0.5341629981994629,
"learning_rate": 1.9743700647852356e-05,
"loss": 0.0068,
"step": 166
},
{
"epoch": 0.09542857142857143,
"grad_norm": 1.3352835178375244,
"learning_rate": 1.9735789028731603e-05,
"loss": 0.0885,
"step": 167
},
{
"epoch": 0.096,
"grad_norm": 0.43347397446632385,
"learning_rate": 1.972775878209397e-05,
"loss": 0.0588,
"step": 168
},
{
"epoch": 0.09657142857142857,
"grad_norm": 0.9168206453323364,
"learning_rate": 1.9719610005785466e-05,
"loss": 0.0359,
"step": 169
},
{
"epoch": 0.09714285714285714,
"grad_norm": 1.348345160484314,
"learning_rate": 1.971134279909636e-05,
"loss": 0.087,
"step": 170
},
{
"epoch": 0.09771428571428571,
"grad_norm": 13.968751907348633,
"learning_rate": 1.9702957262759964e-05,
"loss": 0.1889,
"step": 171
},
{
"epoch": 0.09828571428571428,
"grad_norm": 0.5930889844894409,
"learning_rate": 1.9694453498951392e-05,
"loss": 0.0071,
"step": 172
},
{
"epoch": 0.09885714285714285,
"grad_norm": 0.9004945158958435,
"learning_rate": 1.9685831611286312e-05,
"loss": 0.0858,
"step": 173
},
{
"epoch": 0.09942857142857142,
"grad_norm": 0.7755551338195801,
"learning_rate": 1.9677091704819714e-05,
"loss": 0.0076,
"step": 174
},
{
"epoch": 0.1,
"grad_norm": 2.2851498126983643,
"learning_rate": 1.9668233886044597e-05,
"loss": 0.0659,
"step": 175
},
{
"epoch": 0.10057142857142858,
"grad_norm": 0.7138407230377197,
"learning_rate": 1.9659258262890683e-05,
"loss": 0.0467,
"step": 176
},
{
"epoch": 0.10114285714285715,
"grad_norm": 7.585505962371826,
"learning_rate": 1.9650164944723116e-05,
"loss": 0.6741,
"step": 177
},
{
"epoch": 0.10171428571428572,
"grad_norm": 0.6501437425613403,
"learning_rate": 1.96409540423411e-05,
"loss": 0.0762,
"step": 178
},
{
"epoch": 0.10228571428571429,
"grad_norm": 1.519418716430664,
"learning_rate": 1.9631625667976584e-05,
"loss": 0.1078,
"step": 179
},
{
"epoch": 0.10285714285714286,
"grad_norm": 0.1101742833852768,
"learning_rate": 1.9622179935292855e-05,
"loss": 0.0018,
"step": 180
},
{
"epoch": 0.10342857142857143,
"grad_norm": 0.9454023241996765,
"learning_rate": 1.961261695938319e-05,
"loss": 0.0545,
"step": 181
},
{
"epoch": 0.104,
"grad_norm": 0.9549784660339355,
"learning_rate": 1.9602936856769432e-05,
"loss": 0.0818,
"step": 182
},
{
"epoch": 0.10457142857142857,
"grad_norm": 1.0671652555465698,
"learning_rate": 1.9593139745400575e-05,
"loss": 0.0963,
"step": 183
},
{
"epoch": 0.10514285714285715,
"grad_norm": 0.9419152140617371,
"learning_rate": 1.9583225744651334e-05,
"loss": 0.0989,
"step": 184
},
{
"epoch": 0.10571428571428572,
"grad_norm": 14.619680404663086,
"learning_rate": 1.9573194975320672e-05,
"loss": 0.0822,
"step": 185
},
{
"epoch": 0.10628571428571429,
"grad_norm": 0.46461963653564453,
"learning_rate": 1.9563047559630356e-05,
"loss": 0.0559,
"step": 186
},
{
"epoch": 0.10685714285714286,
"grad_norm": 0.9543680548667908,
"learning_rate": 1.9552783621223437e-05,
"loss": 0.0572,
"step": 187
},
{
"epoch": 0.10742857142857143,
"grad_norm": 1.1674288511276245,
"learning_rate": 1.954240328516277e-05,
"loss": 0.0669,
"step": 188
},
{
"epoch": 0.108,
"grad_norm": 8.751070976257324,
"learning_rate": 1.9531906677929472e-05,
"loss": 0.1785,
"step": 189
},
{
"epoch": 0.10857142857142857,
"grad_norm": 1.5539981126785278,
"learning_rate": 1.9521293927421388e-05,
"loss": 0.1183,
"step": 190
},
{
"epoch": 0.10914285714285714,
"grad_norm": 1.202494740486145,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.0115,
"step": 191
},
{
"epoch": 0.10971428571428571,
"grad_norm": 0.8840175867080688,
"learning_rate": 1.9499720515246524e-05,
"loss": 0.0826,
"step": 192
},
{
"epoch": 0.11028571428571429,
"grad_norm": 1.159778356552124,
"learning_rate": 1.9488760116444966e-05,
"loss": 0.0752,
"step": 193
},
{
"epoch": 0.11085714285714286,
"grad_norm": 16.495677947998047,
"learning_rate": 1.947768410009586e-05,
"loss": 0.1508,
"step": 194
},
{
"epoch": 0.11142857142857143,
"grad_norm": 0.9508908987045288,
"learning_rate": 1.9466492601156964e-05,
"loss": 0.0569,
"step": 195
},
{
"epoch": 0.112,
"grad_norm": 7.683191299438477,
"learning_rate": 1.945518575599317e-05,
"loss": 0.2294,
"step": 196
},
{
"epoch": 0.11257142857142857,
"grad_norm": 1.4274559020996094,
"learning_rate": 1.944376370237481e-05,
"loss": 0.0586,
"step": 197
},
{
"epoch": 0.11314285714285714,
"grad_norm": 1.057436466217041,
"learning_rate": 1.943222657947601e-05,
"loss": 0.0554,
"step": 198
},
{
"epoch": 0.11371428571428571,
"grad_norm": 2.561751365661621,
"learning_rate": 1.942057452787297e-05,
"loss": 0.067,
"step": 199
},
{
"epoch": 0.11428571428571428,
"grad_norm": 3.2694220542907715,
"learning_rate": 1.9408807689542257e-05,
"loss": 0.0174,
"step": 200
},
{
"epoch": 0.11485714285714285,
"grad_norm": 0.09531368315219879,
"learning_rate": 1.9396926207859085e-05,
"loss": 0.0011,
"step": 201
},
{
"epoch": 0.11542857142857142,
"grad_norm": 0.32422834634780884,
"learning_rate": 1.938493022759556e-05,
"loss": 0.0024,
"step": 202
},
{
"epoch": 0.116,
"grad_norm": 16.233394622802734,
"learning_rate": 1.937281989491892e-05,
"loss": 0.5513,
"step": 203
},
{
"epoch": 0.11657142857142858,
"grad_norm": 1.5187242031097412,
"learning_rate": 1.9360595357389735e-05,
"loss": 0.0956,
"step": 204
},
{
"epoch": 0.11714285714285715,
"grad_norm": 29.309446334838867,
"learning_rate": 1.9348256763960146e-05,
"loss": 0.3799,
"step": 205
},
{
"epoch": 0.11771428571428572,
"grad_norm": 0.45075494050979614,
"learning_rate": 1.9335804264972018e-05,
"loss": 0.0462,
"step": 206
},
{
"epoch": 0.11828571428571429,
"grad_norm": 1.227132797241211,
"learning_rate": 1.9323238012155125e-05,
"loss": 0.0806,
"step": 207
},
{
"epoch": 0.11885714285714286,
"grad_norm": 0.06569048017263412,
"learning_rate": 1.9310558158625286e-05,
"loss": 0.0013,
"step": 208
},
{
"epoch": 0.11942857142857143,
"grad_norm": 15.181023597717285,
"learning_rate": 1.9297764858882516e-05,
"loss": 0.5529,
"step": 209
},
{
"epoch": 0.12,
"grad_norm": 0.2923243045806885,
"learning_rate": 1.9284858268809135e-05,
"loss": 0.0394,
"step": 210
},
{
"epoch": 0.12057142857142857,
"grad_norm": 43.48744201660156,
"learning_rate": 1.9271838545667876e-05,
"loss": 0.1257,
"step": 211
},
{
"epoch": 0.12114285714285715,
"grad_norm": 0.47158148884773254,
"learning_rate": 1.925870584809995e-05,
"loss": 0.0573,
"step": 212
},
{
"epoch": 0.12171428571428572,
"grad_norm": 8.10323429107666,
"learning_rate": 1.9245460336123136e-05,
"loss": 0.2109,
"step": 213
},
{
"epoch": 0.12228571428571429,
"grad_norm": 0.5204163193702698,
"learning_rate": 1.923210217112981e-05,
"loss": 0.0385,
"step": 214
},
{
"epoch": 0.12285714285714286,
"grad_norm": 0.610913097858429,
"learning_rate": 1.9218631515885007e-05,
"loss": 0.0379,
"step": 215
},
{
"epoch": 0.12342857142857143,
"grad_norm": 0.5050652623176575,
"learning_rate": 1.9205048534524405e-05,
"loss": 0.0404,
"step": 216
},
{
"epoch": 0.124,
"grad_norm": 13.898507118225098,
"learning_rate": 1.9191353392552346e-05,
"loss": 0.2726,
"step": 217
},
{
"epoch": 0.12457142857142857,
"grad_norm": 1.1407400369644165,
"learning_rate": 1.9177546256839814e-05,
"loss": 0.0932,
"step": 218
},
{
"epoch": 0.12514285714285714,
"grad_norm": 1.0922890901565552,
"learning_rate": 1.9163627295622397e-05,
"loss": 0.016,
"step": 219
},
{
"epoch": 0.12571428571428572,
"grad_norm": 14.770203590393066,
"learning_rate": 1.914959667849825e-05,
"loss": 0.4285,
"step": 220
},
{
"epoch": 0.12628571428571428,
"grad_norm": 2.7945454120635986,
"learning_rate": 1.913545457642601e-05,
"loss": 0.1545,
"step": 221
},
{
"epoch": 0.12685714285714286,
"grad_norm": 0.910795271396637,
"learning_rate": 1.9121201161722732e-05,
"loss": 0.0697,
"step": 222
},
{
"epoch": 0.12742857142857142,
"grad_norm": 0.7550761103630066,
"learning_rate": 1.910683660806177e-05,
"loss": 0.0451,
"step": 223
},
{
"epoch": 0.128,
"grad_norm": 0.6814684271812439,
"learning_rate": 1.9092361090470688e-05,
"loss": 0.0407,
"step": 224
},
{
"epoch": 0.12857142857142856,
"grad_norm": 10.898219108581543,
"learning_rate": 1.907777478532909e-05,
"loss": 0.2644,
"step": 225
},
{
"epoch": 0.12914285714285714,
"grad_norm": 3.2427897453308105,
"learning_rate": 1.9063077870366504e-05,
"loss": 0.0531,
"step": 226
},
{
"epoch": 0.12971428571428573,
"grad_norm": 7.148037433624268,
"learning_rate": 1.9048270524660197e-05,
"loss": 0.2163,
"step": 227
},
{
"epoch": 0.13028571428571428,
"grad_norm": 7.682395935058594,
"learning_rate": 1.903335292863301e-05,
"loss": 0.1073,
"step": 228
},
{
"epoch": 0.13085714285714287,
"grad_norm": 1.644207239151001,
"learning_rate": 1.901832526405114e-05,
"loss": 0.1071,
"step": 229
},
{
"epoch": 0.13142857142857142,
"grad_norm": 2.6188018321990967,
"learning_rate": 1.9003187714021936e-05,
"loss": 0.0391,
"step": 230
},
{
"epoch": 0.132,
"grad_norm": 0.8990005254745483,
"learning_rate": 1.8987940462991673e-05,
"loss": 0.0522,
"step": 231
},
{
"epoch": 0.13257142857142856,
"grad_norm": 0.5779718160629272,
"learning_rate": 1.8972583696743284e-05,
"loss": 0.0454,
"step": 232
},
{
"epoch": 0.13314285714285715,
"grad_norm": 0.7823575139045715,
"learning_rate": 1.895711760239413e-05,
"loss": 0.0476,
"step": 233
},
{
"epoch": 0.1337142857142857,
"grad_norm": 3.680964708328247,
"learning_rate": 1.8941542368393683e-05,
"loss": 0.1,
"step": 234
},
{
"epoch": 0.13428571428571429,
"grad_norm": 18.032867431640625,
"learning_rate": 1.892585818452126e-05,
"loss": 0.3108,
"step": 235
},
{
"epoch": 0.13485714285714287,
"grad_norm": 2.4249937534332275,
"learning_rate": 1.891006524188368e-05,
"loss": 0.095,
"step": 236
},
{
"epoch": 0.13542857142857143,
"grad_norm": 1.4989759922027588,
"learning_rate": 1.889416373291298e-05,
"loss": 0.0865,
"step": 237
},
{
"epoch": 0.136,
"grad_norm": 0.2266465574502945,
"learning_rate": 1.8878153851364013e-05,
"loss": 0.0048,
"step": 238
},
{
"epoch": 0.13657142857142857,
"grad_norm": 3.880814790725708,
"learning_rate": 1.8862035792312148e-05,
"loss": 0.1073,
"step": 239
},
{
"epoch": 0.13714285714285715,
"grad_norm": 1.1137272119522095,
"learning_rate": 1.884580975215084e-05,
"loss": 0.0143,
"step": 240
},
{
"epoch": 0.1377142857142857,
"grad_norm": 0.6647632122039795,
"learning_rate": 1.8829475928589272e-05,
"loss": 0.0557,
"step": 241
},
{
"epoch": 0.1382857142857143,
"grad_norm": 1.5873721837997437,
"learning_rate": 1.8813034520649923e-05,
"loss": 0.0466,
"step": 242
},
{
"epoch": 0.13885714285714285,
"grad_norm": 3.342252016067505,
"learning_rate": 1.879648572866617e-05,
"loss": 0.0338,
"step": 243
},
{
"epoch": 0.13942857142857143,
"grad_norm": 8.770962715148926,
"learning_rate": 1.8779829754279806e-05,
"loss": 0.2685,
"step": 244
},
{
"epoch": 0.14,
"grad_norm": 1.3860154151916504,
"learning_rate": 1.8763066800438638e-05,
"loss": 0.0803,
"step": 245
},
{
"epoch": 0.14057142857142857,
"grad_norm": 0.9029803276062012,
"learning_rate": 1.874619707139396e-05,
"loss": 0.0146,
"step": 246
},
{
"epoch": 0.14114285714285715,
"grad_norm": 0.6893714070320129,
"learning_rate": 1.8729220772698096e-05,
"loss": 0.0734,
"step": 247
},
{
"epoch": 0.1417142857142857,
"grad_norm": 0.7808796167373657,
"learning_rate": 1.8712138111201898e-05,
"loss": 0.0705,
"step": 248
},
{
"epoch": 0.1422857142857143,
"grad_norm": 0.7740113139152527,
"learning_rate": 1.869494929505219e-05,
"loss": 0.0628,
"step": 249
},
{
"epoch": 0.14285714285714285,
"grad_norm": 1.2257181406021118,
"learning_rate": 1.8677654533689287e-05,
"loss": 0.0365,
"step": 250
},
{
"epoch": 0.14342857142857143,
"grad_norm": 4.592135906219482,
"learning_rate": 1.866025403784439e-05,
"loss": 0.0893,
"step": 251
},
{
"epoch": 0.144,
"grad_norm": 1.1102668046951294,
"learning_rate": 1.864274801953705e-05,
"loss": 0.08,
"step": 252
},
{
"epoch": 0.14457142857142857,
"grad_norm": 1.7475136518478394,
"learning_rate": 1.8625136692072577e-05,
"loss": 0.1169,
"step": 253
},
{
"epoch": 0.14514285714285713,
"grad_norm": 0.647798478603363,
"learning_rate": 1.860742027003944e-05,
"loss": 0.0385,
"step": 254
},
{
"epoch": 0.1457142857142857,
"grad_norm": 0.5438143610954285,
"learning_rate": 1.8589598969306646e-05,
"loss": 0.0602,
"step": 255
},
{
"epoch": 0.1462857142857143,
"grad_norm": 0.5790269374847412,
"learning_rate": 1.8571673007021124e-05,
"loss": 0.0514,
"step": 256
},
{
"epoch": 0.14685714285714285,
"grad_norm": 0.5948929786682129,
"learning_rate": 1.855364260160507e-05,
"loss": 0.0384,
"step": 257
},
{
"epoch": 0.14742857142857144,
"grad_norm": 0.7339144945144653,
"learning_rate": 1.8535507972753275e-05,
"loss": 0.0402,
"step": 258
},
{
"epoch": 0.148,
"grad_norm": 0.25453007221221924,
"learning_rate": 1.851726934143048e-05,
"loss": 0.0027,
"step": 259
},
{
"epoch": 0.14857142857142858,
"grad_norm": 0.5421114563941956,
"learning_rate": 1.849892692986864e-05,
"loss": 0.0587,
"step": 260
},
{
"epoch": 0.14914285714285713,
"grad_norm": 0.11004189401865005,
"learning_rate": 1.848048096156426e-05,
"loss": 0.0006,
"step": 261
},
{
"epoch": 0.14971428571428572,
"grad_norm": 0.3382515013217926,
"learning_rate": 1.8461931661275642e-05,
"loss": 0.0497,
"step": 262
},
{
"epoch": 0.15028571428571427,
"grad_norm": 1.3797106742858887,
"learning_rate": 1.8443279255020153e-05,
"loss": 0.0952,
"step": 263
},
{
"epoch": 0.15085714285714286,
"grad_norm": 0.6782971620559692,
"learning_rate": 1.842452397007148e-05,
"loss": 0.0403,
"step": 264
},
{
"epoch": 0.15142857142857144,
"grad_norm": 1.8418234586715698,
"learning_rate": 1.8405666034956842e-05,
"loss": 0.0723,
"step": 265
},
{
"epoch": 0.152,
"grad_norm": 0.2565213739871979,
"learning_rate": 1.8386705679454243e-05,
"loss": 0.0049,
"step": 266
},
{
"epoch": 0.15257142857142858,
"grad_norm": 0.14162056148052216,
"learning_rate": 1.836764313458962e-05,
"loss": 0.0009,
"step": 267
},
{
"epoch": 0.15314285714285714,
"grad_norm": 1.6656711101531982,
"learning_rate": 1.8348478632634067e-05,
"loss": 0.0967,
"step": 268
},
{
"epoch": 0.15371428571428572,
"grad_norm": 0.3504199683666229,
"learning_rate": 1.8329212407100996e-05,
"loss": 0.0292,
"step": 269
},
{
"epoch": 0.15428571428571428,
"grad_norm": 1.6510212421417236,
"learning_rate": 1.8309844692743283e-05,
"loss": 0.0124,
"step": 270
},
{
"epoch": 0.15485714285714286,
"grad_norm": 0.5507133603096008,
"learning_rate": 1.8290375725550417e-05,
"loss": 0.0673,
"step": 271
},
{
"epoch": 0.15542857142857142,
"grad_norm": 0.36375293135643005,
"learning_rate": 1.827080574274562e-05,
"loss": 0.03,
"step": 272
},
{
"epoch": 0.156,
"grad_norm": 1.03709077835083,
"learning_rate": 1.8251134982782952e-05,
"loss": 0.0574,
"step": 273
},
{
"epoch": 0.15657142857142858,
"grad_norm": 0.39899778366088867,
"learning_rate": 1.8231363685344422e-05,
"loss": 0.0463,
"step": 274
},
{
"epoch": 0.15714285714285714,
"grad_norm": 1.0305384397506714,
"learning_rate": 1.821149209133704e-05,
"loss": 0.0777,
"step": 275
},
{
"epoch": 0.15771428571428572,
"grad_norm": 0.3350299894809723,
"learning_rate": 1.819152044288992e-05,
"loss": 0.0266,
"step": 276
},
{
"epoch": 0.15828571428571428,
"grad_norm": 0.14408257603645325,
"learning_rate": 1.8171448983351284e-05,
"loss": 0.0026,
"step": 277
},
{
"epoch": 0.15885714285714286,
"grad_norm": 0.18242938816547394,
"learning_rate": 1.815127795728554e-05,
"loss": 0.002,
"step": 278
},
{
"epoch": 0.15942857142857142,
"grad_norm": 0.24773648381233215,
"learning_rate": 1.8131007610470278e-05,
"loss": 0.0256,
"step": 279
},
{
"epoch": 0.16,
"grad_norm": 0.41257333755493164,
"learning_rate": 1.8110638189893267e-05,
"loss": 0.0489,
"step": 280
},
{
"epoch": 0.16057142857142856,
"grad_norm": 0.09563997387886047,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.0004,
"step": 281
},
{
"epoch": 0.16114285714285714,
"grad_norm": 0.11035250127315521,
"learning_rate": 1.806960312143802e-05,
"loss": 0.0012,
"step": 282
},
{
"epoch": 0.16171428571428573,
"grad_norm": 0.2648049294948578,
"learning_rate": 1.804893797355914e-05,
"loss": 0.0259,
"step": 283
},
{
"epoch": 0.16228571428571428,
"grad_norm": 0.7870336771011353,
"learning_rate": 1.8028174751911147e-05,
"loss": 0.0104,
"step": 284
},
{
"epoch": 0.16285714285714287,
"grad_norm": 6.117203712463379,
"learning_rate": 1.8007313709487334e-05,
"loss": 0.1224,
"step": 285
},
{
"epoch": 0.16342857142857142,
"grad_norm": 0.08172111213207245,
"learning_rate": 1.798635510047293e-05,
"loss": 0.0008,
"step": 286
},
{
"epoch": 0.164,
"grad_norm": 0.9074399471282959,
"learning_rate": 1.7965299180241963e-05,
"loss": 0.0306,
"step": 287
},
{
"epoch": 0.16457142857142856,
"grad_norm": 0.6775059103965759,
"learning_rate": 1.7944146205354182e-05,
"loss": 0.0664,
"step": 288
},
{
"epoch": 0.16514285714285715,
"grad_norm": 1.1402727365493774,
"learning_rate": 1.792289643355191e-05,
"loss": 0.078,
"step": 289
},
{
"epoch": 0.1657142857142857,
"grad_norm": 0.7922158241271973,
"learning_rate": 1.7901550123756906e-05,
"loss": 0.0758,
"step": 290
},
{
"epoch": 0.1662857142857143,
"grad_norm": 0.7050178647041321,
"learning_rate": 1.788010753606722e-05,
"loss": 0.0634,
"step": 291
},
{
"epoch": 0.16685714285714287,
"grad_norm": 0.7817528247833252,
"learning_rate": 1.785856893175402e-05,
"loss": 0.0697,
"step": 292
},
{
"epoch": 0.16742857142857143,
"grad_norm": 0.6317795515060425,
"learning_rate": 1.78369345732584e-05,
"loss": 0.0351,
"step": 293
},
{
"epoch": 0.168,
"grad_norm": 0.07126916199922562,
"learning_rate": 1.781520472418819e-05,
"loss": 0.0008,
"step": 294
},
{
"epoch": 0.16857142857142857,
"grad_norm": 0.10297396779060364,
"learning_rate": 1.7793379649314743e-05,
"loss": 0.002,
"step": 295
},
{
"epoch": 0.16914285714285715,
"grad_norm": 0.1650630384683609,
"learning_rate": 1.777145961456971e-05,
"loss": 0.0004,
"step": 296
},
{
"epoch": 0.1697142857142857,
"grad_norm": 0.039217106997966766,
"learning_rate": 1.7749444887041797e-05,
"loss": 0.0005,
"step": 297
},
{
"epoch": 0.1702857142857143,
"grad_norm": 0.7831671833992004,
"learning_rate": 1.7727335734973512e-05,
"loss": 0.0705,
"step": 298
},
{
"epoch": 0.17085714285714285,
"grad_norm": 0.11735610663890839,
"learning_rate": 1.7705132427757895e-05,
"loss": 0.0025,
"step": 299
},
{
"epoch": 0.17142857142857143,
"grad_norm": 0.6778116822242737,
"learning_rate": 1.7682835235935236e-05,
"loss": 0.0689,
"step": 300
},
{
"epoch": 0.172,
"grad_norm": 0.5785430669784546,
"learning_rate": 1.766044443118978e-05,
"loss": 0.0648,
"step": 301
},
{
"epoch": 0.17257142857142857,
"grad_norm": 0.24186182022094727,
"learning_rate": 1.7637960286346423e-05,
"loss": 0.0271,
"step": 302
},
{
"epoch": 0.17314285714285715,
"grad_norm": 9.603544235229492,
"learning_rate": 1.761538307536737e-05,
"loss": 0.1241,
"step": 303
},
{
"epoch": 0.1737142857142857,
"grad_norm": 0.4855884909629822,
"learning_rate": 1.759271307334881e-05,
"loss": 0.0372,
"step": 304
},
{
"epoch": 0.1742857142857143,
"grad_norm": 0.26609915494918823,
"learning_rate": 1.7569950556517566e-05,
"loss": 0.0288,
"step": 305
},
{
"epoch": 0.17485714285714285,
"grad_norm": 3.3161323070526123,
"learning_rate": 1.7547095802227723e-05,
"loss": 0.0227,
"step": 306
},
{
"epoch": 0.17542857142857143,
"grad_norm": 0.3556174039840698,
"learning_rate": 1.7524149088957244e-05,
"loss": 0.0435,
"step": 307
},
{
"epoch": 0.176,
"grad_norm": 0.10047593712806702,
"learning_rate": 1.7501110696304598e-05,
"loss": 0.0005,
"step": 308
},
{
"epoch": 0.17657142857142857,
"grad_norm": 0.17550890147686005,
"learning_rate": 1.747798090498532e-05,
"loss": 0.0005,
"step": 309
},
{
"epoch": 0.17714285714285713,
"grad_norm": 0.28629225492477417,
"learning_rate": 1.7454759996828622e-05,
"loss": 0.0454,
"step": 310
},
{
"epoch": 0.1777142857142857,
"grad_norm": 0.6481816172599792,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.0592,
"step": 311
},
{
"epoch": 0.1782857142857143,
"grad_norm": 5.234211444854736,
"learning_rate": 1.74080459628675e-05,
"loss": 0.0352,
"step": 312
},
{
"epoch": 0.17885714285714285,
"grad_norm": 0.7525281310081482,
"learning_rate": 1.7384553406258842e-05,
"loss": 0.0771,
"step": 313
},
{
"epoch": 0.17942857142857144,
"grad_norm": 0.2318568080663681,
"learning_rate": 1.7360970871197347e-05,
"loss": 0.0383,
"step": 314
},
{
"epoch": 0.18,
"grad_norm": 12.53724193572998,
"learning_rate": 1.7337298645028764e-05,
"loss": 0.1838,
"step": 315
},
{
"epoch": 0.18057142857142858,
"grad_norm": 0.4352404773235321,
"learning_rate": 1.7313537016191706e-05,
"loss": 0.0366,
"step": 316
},
{
"epoch": 0.18114285714285713,
"grad_norm": 0.5024876594543457,
"learning_rate": 1.7289686274214116e-05,
"loss": 0.0532,
"step": 317
},
{
"epoch": 0.18171428571428572,
"grad_norm": 0.2753225564956665,
"learning_rate": 1.7265746709709762e-05,
"loss": 0.0057,
"step": 318
},
{
"epoch": 0.18228571428571427,
"grad_norm": 0.4302017390727997,
"learning_rate": 1.7241718614374678e-05,
"loss": 0.0534,
"step": 319
},
{
"epoch": 0.18285714285714286,
"grad_norm": 8.177531242370605,
"learning_rate": 1.7217602280983622e-05,
"loss": 0.3677,
"step": 320
},
{
"epoch": 0.18342857142857144,
"grad_norm": 0.9254489541053772,
"learning_rate": 1.7193398003386514e-05,
"loss": 0.0341,
"step": 321
},
{
"epoch": 0.184,
"grad_norm": 0.258419394493103,
"learning_rate": 1.716910607650483e-05,
"loss": 0.0463,
"step": 322
},
{
"epoch": 0.18457142857142858,
"grad_norm": 0.10824820399284363,
"learning_rate": 1.7144726796328034e-05,
"loss": 0.0006,
"step": 323
},
{
"epoch": 0.18514285714285714,
"grad_norm": 0.3367154002189636,
"learning_rate": 1.712026045990997e-05,
"loss": 0.0464,
"step": 324
},
{
"epoch": 0.18571428571428572,
"grad_norm": 4.240468978881836,
"learning_rate": 1.709570736536521e-05,
"loss": 0.0705,
"step": 325
},
{
"epoch": 0.18628571428571428,
"grad_norm": 0.36025333404541016,
"learning_rate": 1.7071067811865477e-05,
"loss": 0.0311,
"step": 326
},
{
"epoch": 0.18685714285714286,
"grad_norm": 1.5349019765853882,
"learning_rate": 1.7046342099635948e-05,
"loss": 0.0542,
"step": 327
},
{
"epoch": 0.18742857142857142,
"grad_norm": 0.39610278606414795,
"learning_rate": 1.7021530529951627e-05,
"loss": 0.0457,
"step": 328
},
{
"epoch": 0.188,
"grad_norm": 1.067283034324646,
"learning_rate": 1.6996633405133656e-05,
"loss": 0.0474,
"step": 329
},
{
"epoch": 0.18857142857142858,
"grad_norm": 0.03407033905386925,
"learning_rate": 1.697165102854565e-05,
"loss": 0.0011,
"step": 330
},
{
"epoch": 0.18914285714285714,
"grad_norm": 0.0733943060040474,
"learning_rate": 1.6946583704589973e-05,
"loss": 0.0017,
"step": 331
},
{
"epoch": 0.18971428571428572,
"grad_norm": 0.5164986848831177,
"learning_rate": 1.692143173870407e-05,
"loss": 0.0708,
"step": 332
},
{
"epoch": 0.19028571428571428,
"grad_norm": 0.4167448878288269,
"learning_rate": 1.68961954373567e-05,
"loss": 0.0328,
"step": 333
},
{
"epoch": 0.19085714285714286,
"grad_norm": 0.3320442736148834,
"learning_rate": 1.6870875108044233e-05,
"loss": 0.0298,
"step": 334
},
{
"epoch": 0.19142857142857142,
"grad_norm": 0.04326656833291054,
"learning_rate": 1.684547105928689e-05,
"loss": 0.0007,
"step": 335
},
{
"epoch": 0.192,
"grad_norm": 0.31868818402290344,
"learning_rate": 1.6819983600624986e-05,
"loss": 0.0281,
"step": 336
},
{
"epoch": 0.19257142857142856,
"grad_norm": 0.3172406256198883,
"learning_rate": 1.6794413042615168e-05,
"loss": 0.0383,
"step": 337
},
{
"epoch": 0.19314285714285714,
"grad_norm": 11.330007553100586,
"learning_rate": 1.6768759696826608e-05,
"loss": 0.1246,
"step": 338
},
{
"epoch": 0.19371428571428573,
"grad_norm": 0.17406456172466278,
"learning_rate": 1.6743023875837233e-05,
"loss": 0.0007,
"step": 339
},
{
"epoch": 0.19428571428571428,
"grad_norm": 0.22507016360759735,
"learning_rate": 1.6717205893229904e-05,
"loss": 0.0331,
"step": 340
},
{
"epoch": 0.19485714285714287,
"grad_norm": 0.22983767092227936,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.0278,
"step": 341
},
{
"epoch": 0.19542857142857142,
"grad_norm": 2.142090320587158,
"learning_rate": 1.6665324702494524e-05,
"loss": 0.0749,
"step": 342
},
{
"epoch": 0.196,
"grad_norm": 0.23252758383750916,
"learning_rate": 1.6639262126522417e-05,
"loss": 0.0252,
"step": 343
},
{
"epoch": 0.19657142857142856,
"grad_norm": 16.77885627746582,
"learning_rate": 1.661311865323652e-05,
"loss": 0.1841,
"step": 344
},
{
"epoch": 0.19714285714285715,
"grad_norm": 0.2985742390155792,
"learning_rate": 1.6586894601186804e-05,
"loss": 0.0262,
"step": 345
},
{
"epoch": 0.1977142857142857,
"grad_norm": 1.7478572130203247,
"learning_rate": 1.6560590289905074e-05,
"loss": 0.1249,
"step": 346
},
{
"epoch": 0.1982857142857143,
"grad_norm": 1.703766107559204,
"learning_rate": 1.6534206039901057e-05,
"loss": 0.1193,
"step": 347
},
{
"epoch": 0.19885714285714284,
"grad_norm": 11.683262825012207,
"learning_rate": 1.650774217265851e-05,
"loss": 0.5635,
"step": 348
},
{
"epoch": 0.19942857142857143,
"grad_norm": 0.5566651821136475,
"learning_rate": 1.6481199010631312e-05,
"loss": 0.0201,
"step": 349
},
{
"epoch": 0.2,
"grad_norm": 2.675867795944214,
"learning_rate": 1.645457687723951e-05,
"loss": 0.1104,
"step": 350
},
{
"epoch": 0.20057142857142857,
"grad_norm": 2.5131735801696777,
"learning_rate": 1.6427876096865394e-05,
"loss": 0.0141,
"step": 351
},
{
"epoch": 0.20114285714285715,
"grad_norm": 1.8795198202133179,
"learning_rate": 1.6401096994849558e-05,
"loss": 0.0709,
"step": 352
},
{
"epoch": 0.2017142857142857,
"grad_norm": 0.1060282364487648,
"learning_rate": 1.63742398974869e-05,
"loss": 0.0022,
"step": 353
},
{
"epoch": 0.2022857142857143,
"grad_norm": 0.7330822944641113,
"learning_rate": 1.6347305132022677e-05,
"loss": 0.0692,
"step": 354
},
{
"epoch": 0.20285714285714285,
"grad_norm": 0.07668916881084442,
"learning_rate": 1.632029302664851e-05,
"loss": 0.0002,
"step": 355
},
{
"epoch": 0.20342857142857143,
"grad_norm": 0.28914695978164673,
"learning_rate": 1.6293203910498375e-05,
"loss": 0.0266,
"step": 356
},
{
"epoch": 0.204,
"grad_norm": 36.421905517578125,
"learning_rate": 1.6266038113644605e-05,
"loss": 0.2924,
"step": 357
},
{
"epoch": 0.20457142857142857,
"grad_norm": 0.8194921016693115,
"learning_rate": 1.6238795967093865e-05,
"loss": 0.0753,
"step": 358
},
{
"epoch": 0.20514285714285715,
"grad_norm": 0.4537048041820526,
"learning_rate": 1.6211477802783105e-05,
"loss": 0.0341,
"step": 359
},
{
"epoch": 0.2057142857142857,
"grad_norm": 1.3236613273620605,
"learning_rate": 1.6184083953575543e-05,
"loss": 0.0813,
"step": 360
},
{
"epoch": 0.2062857142857143,
"grad_norm": 1.0030516386032104,
"learning_rate": 1.6156614753256583e-05,
"loss": 0.0793,
"step": 361
},
{
"epoch": 0.20685714285714285,
"grad_norm": 0.15925391018390656,
"learning_rate": 1.6129070536529767e-05,
"loss": 0.0022,
"step": 362
},
{
"epoch": 0.20742857142857143,
"grad_norm": 0.17983612418174744,
"learning_rate": 1.610145163901268e-05,
"loss": 0.0046,
"step": 363
},
{
"epoch": 0.208,
"grad_norm": 1.4236613512039185,
"learning_rate": 1.607375839723287e-05,
"loss": 0.0476,
"step": 364
},
{
"epoch": 0.20857142857142857,
"grad_norm": 1.415261149406433,
"learning_rate": 1.6045991148623752e-05,
"loss": 0.0811,
"step": 365
},
{
"epoch": 0.20914285714285713,
"grad_norm": 0.7466470003128052,
"learning_rate": 1.6018150231520486e-05,
"loss": 0.0677,
"step": 366
},
{
"epoch": 0.20971428571428571,
"grad_norm": 0.4385108947753906,
"learning_rate": 1.599023598515586e-05,
"loss": 0.0371,
"step": 367
},
{
"epoch": 0.2102857142857143,
"grad_norm": 9.935927391052246,
"learning_rate": 1.5962248749656158e-05,
"loss": 0.209,
"step": 368
},
{
"epoch": 0.21085714285714285,
"grad_norm": 1.705437421798706,
"learning_rate": 1.5934188866037017e-05,
"loss": 0.0129,
"step": 369
},
{
"epoch": 0.21142857142857144,
"grad_norm": 0.8051179051399231,
"learning_rate": 1.5906056676199256e-05,
"loss": 0.0352,
"step": 370
},
{
"epoch": 0.212,
"grad_norm": 2.643200635910034,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.072,
"step": 371
},
{
"epoch": 0.21257142857142858,
"grad_norm": 0.3560256063938141,
"learning_rate": 1.584957674987216e-05,
"loss": 0.0437,
"step": 372
},
{
"epoch": 0.21314285714285713,
"grad_norm": 0.9192269444465637,
"learning_rate": 1.5821229701572897e-05,
"loss": 0.0386,
"step": 373
},
{
"epoch": 0.21371428571428572,
"grad_norm": 0.026125172153115273,
"learning_rate": 1.5792811723426787e-05,
"loss": 0.0013,
"step": 374
},
{
"epoch": 0.21428571428571427,
"grad_norm": 0.8916493058204651,
"learning_rate": 1.5764323161697933e-05,
"loss": 0.0374,
"step": 375
},
{
"epoch": 0.21485714285714286,
"grad_norm": 0.05699808523058891,
"learning_rate": 1.573576436351046e-05,
"loss": 0.0004,
"step": 376
},
{
"epoch": 0.21542857142857144,
"grad_norm": 1.3239928483963013,
"learning_rate": 1.570713567684432e-05,
"loss": 0.0684,
"step": 377
},
{
"epoch": 0.216,
"grad_norm": 0.6361359357833862,
"learning_rate": 1.5678437450531014e-05,
"loss": 0.0718,
"step": 378
},
{
"epoch": 0.21657142857142858,
"grad_norm": 0.12468484789133072,
"learning_rate": 1.564967003424938e-05,
"loss": 0.0016,
"step": 379
},
{
"epoch": 0.21714285714285714,
"grad_norm": 0.7894086241722107,
"learning_rate": 1.5620833778521306e-05,
"loss": 0.0566,
"step": 380
},
{
"epoch": 0.21771428571428572,
"grad_norm": 1.0137920379638672,
"learning_rate": 1.5591929034707468e-05,
"loss": 0.0812,
"step": 381
},
{
"epoch": 0.21828571428571428,
"grad_norm": 0.5464143753051758,
"learning_rate": 1.556295615500305e-05,
"loss": 0.0521,
"step": 382
},
{
"epoch": 0.21885714285714286,
"grad_norm": 0.2746932804584503,
"learning_rate": 1.553391549243344e-05,
"loss": 0.0326,
"step": 383
},
{
"epoch": 0.21942857142857142,
"grad_norm": 0.5185433626174927,
"learning_rate": 1.5504807400849957e-05,
"loss": 0.0071,
"step": 384
},
{
"epoch": 0.22,
"grad_norm": 1.3760286569595337,
"learning_rate": 1.5475632234925505e-05,
"loss": 0.0515,
"step": 385
},
{
"epoch": 0.22057142857142858,
"grad_norm": 8.069782257080078,
"learning_rate": 1.5446390350150272e-05,
"loss": 0.0751,
"step": 386
},
{
"epoch": 0.22114285714285714,
"grad_norm": 0.8513094186782837,
"learning_rate": 1.54170821028274e-05,
"loss": 0.0488,
"step": 387
},
{
"epoch": 0.22171428571428572,
"grad_norm": 1.081343650817871,
"learning_rate": 1.5387707850068633e-05,
"loss": 0.0358,
"step": 388
},
{
"epoch": 0.22228571428571428,
"grad_norm": 0.05808674544095993,
"learning_rate": 1.5358267949789968e-05,
"loss": 0.0004,
"step": 389
},
{
"epoch": 0.22285714285714286,
"grad_norm": 0.9796769618988037,
"learning_rate": 1.53287627607073e-05,
"loss": 0.0415,
"step": 390
},
{
"epoch": 0.22342857142857142,
"grad_norm": 0.02601262368261814,
"learning_rate": 1.529919264233205e-05,
"loss": 0.001,
"step": 391
},
{
"epoch": 0.224,
"grad_norm": 0.8839792013168335,
"learning_rate": 1.5269557954966777e-05,
"loss": 0.0688,
"step": 392
},
{
"epoch": 0.22457142857142856,
"grad_norm": 0.29274043440818787,
"learning_rate": 1.5239859059700794e-05,
"loss": 0.0366,
"step": 393
},
{
"epoch": 0.22514285714285714,
"grad_norm": 0.3418540954589844,
"learning_rate": 1.5210096318405768e-05,
"loss": 0.0226,
"step": 394
},
{
"epoch": 0.2257142857142857,
"grad_norm": 0.21711991727352142,
"learning_rate": 1.5180270093731305e-05,
"loss": 0.0266,
"step": 395
},
{
"epoch": 0.22628571428571428,
"grad_norm": 0.07126382738351822,
"learning_rate": 1.5150380749100545e-05,
"loss": 0.0011,
"step": 396
},
{
"epoch": 0.22685714285714287,
"grad_norm": 0.06386347115039825,
"learning_rate": 1.5120428648705716e-05,
"loss": 0.0007,
"step": 397
},
{
"epoch": 0.22742857142857142,
"grad_norm": 0.8950487375259399,
"learning_rate": 1.5090414157503715e-05,
"loss": 0.0356,
"step": 398
},
{
"epoch": 0.228,
"grad_norm": 0.5003771781921387,
"learning_rate": 1.5060337641211637e-05,
"loss": 0.0493,
"step": 399
},
{
"epoch": 0.22857142857142856,
"grad_norm": 0.11169221997261047,
"learning_rate": 1.5030199466302354e-05,
"loss": 0.0011,
"step": 400
},
{
"epoch": 0.22914285714285715,
"grad_norm": 0.43462786078453064,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.0322,
"step": 401
},
{
"epoch": 0.2297142857142857,
"grad_norm": 0.1936434805393219,
"learning_rate": 1.4969739610275556e-05,
"loss": 0.0404,
"step": 402
},
{
"epoch": 0.2302857142857143,
"grad_norm": 2.385303020477295,
"learning_rate": 1.493941866584231e-05,
"loss": 0.1299,
"step": 403
},
{
"epoch": 0.23085714285714284,
"grad_norm": 0.1085561066865921,
"learning_rate": 1.490903753615141e-05,
"loss": 0.0008,
"step": 404
},
{
"epoch": 0.23142857142857143,
"grad_norm": 0.3743264675140381,
"learning_rate": 1.4878596591387329e-05,
"loss": 0.0531,
"step": 405
},
{
"epoch": 0.232,
"grad_norm": 0.04981555789709091,
"learning_rate": 1.4848096202463373e-05,
"loss": 0.001,
"step": 406
},
{
"epoch": 0.23257142857142857,
"grad_norm": 0.037978705018758774,
"learning_rate": 1.4817536741017153e-05,
"loss": 0.0007,
"step": 407
},
{
"epoch": 0.23314285714285715,
"grad_norm": 0.2091723382472992,
"learning_rate": 1.478691857940607e-05,
"loss": 0.0256,
"step": 408
},
{
"epoch": 0.2337142857142857,
"grad_norm": 7.492735385894775,
"learning_rate": 1.4756242090702756e-05,
"loss": 0.0712,
"step": 409
},
{
"epoch": 0.2342857142857143,
"grad_norm": 0.5465371608734131,
"learning_rate": 1.4725507648690542e-05,
"loss": 0.0681,
"step": 410
},
{
"epoch": 0.23485714285714285,
"grad_norm": 0.35538962483406067,
"learning_rate": 1.469471562785891e-05,
"loss": 0.0206,
"step": 411
},
{
"epoch": 0.23542857142857143,
"grad_norm": 0.3554655611515045,
"learning_rate": 1.4663866403398915e-05,
"loss": 0.0521,
"step": 412
},
{
"epoch": 0.236,
"grad_norm": 0.33988815546035767,
"learning_rate": 1.463296035119862e-05,
"loss": 0.0194,
"step": 413
},
{
"epoch": 0.23657142857142857,
"grad_norm": 1.1167902946472168,
"learning_rate": 1.4601997847838518e-05,
"loss": 0.0729,
"step": 414
},
{
"epoch": 0.23714285714285716,
"grad_norm": 13.823290824890137,
"learning_rate": 1.4570979270586944e-05,
"loss": 0.0582,
"step": 415
},
{
"epoch": 0.2377142857142857,
"grad_norm": 0.3877246081829071,
"learning_rate": 1.4539904997395468e-05,
"loss": 0.0543,
"step": 416
},
{
"epoch": 0.2382857142857143,
"grad_norm": 0.3316015303134918,
"learning_rate": 1.4508775406894308e-05,
"loss": 0.0506,
"step": 417
},
{
"epoch": 0.23885714285714285,
"grad_norm": 4.192994594573975,
"learning_rate": 1.4477590878387697e-05,
"loss": 0.0778,
"step": 418
},
{
"epoch": 0.23942857142857144,
"grad_norm": 0.1991989016532898,
"learning_rate": 1.4446351791849276e-05,
"loss": 0.0445,
"step": 419
},
{
"epoch": 0.24,
"grad_norm": 2.449563980102539,
"learning_rate": 1.4415058527917454e-05,
"loss": 0.0614,
"step": 420
},
{
"epoch": 0.24057142857142857,
"grad_norm": 0.5236610770225525,
"learning_rate": 1.4383711467890776e-05,
"loss": 0.0353,
"step": 421
},
{
"epoch": 0.24114285714285713,
"grad_norm": 0.19651485979557037,
"learning_rate": 1.4352310993723277e-05,
"loss": 0.0245,
"step": 422
},
{
"epoch": 0.24171428571428571,
"grad_norm": 0.06734167039394379,
"learning_rate": 1.4320857488019826e-05,
"loss": 0.002,
"step": 423
},
{
"epoch": 0.2422857142857143,
"grad_norm": 0.07878410071134567,
"learning_rate": 1.4289351334031461e-05,
"loss": 0.0011,
"step": 424
},
{
"epoch": 0.24285714285714285,
"grad_norm": 0.2995428144931793,
"learning_rate": 1.4257792915650728e-05,
"loss": 0.0475,
"step": 425
},
{
"epoch": 0.24342857142857144,
"grad_norm": 0.635193407535553,
"learning_rate": 1.4226182617406996e-05,
"loss": 0.0617,
"step": 426
},
{
"epoch": 0.244,
"grad_norm": 0.025279924273490906,
"learning_rate": 1.4194520824461773e-05,
"loss": 0.0005,
"step": 427
},
{
"epoch": 0.24457142857142858,
"grad_norm": 9.663908004760742,
"learning_rate": 1.4162807922604014e-05,
"loss": 0.1698,
"step": 428
},
{
"epoch": 0.24514285714285713,
"grad_norm": 4.746012210845947,
"learning_rate": 1.413104429824542e-05,
"loss": 0.0649,
"step": 429
},
{
"epoch": 0.24571428571428572,
"grad_norm": 0.6950699090957642,
"learning_rate": 1.4099230338415728e-05,
"loss": 0.0661,
"step": 430
},
{
"epoch": 0.24628571428571427,
"grad_norm": 0.10010308027267456,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.0006,
"step": 431
},
{
"epoch": 0.24685714285714286,
"grad_norm": 10.790380477905273,
"learning_rate": 1.4035452963523903e-05,
"loss": 0.1404,
"step": 432
},
{
"epoch": 0.24742857142857144,
"grad_norm": 0.1662398725748062,
"learning_rate": 1.4003490325568953e-05,
"loss": 0.0027,
"step": 433
},
{
"epoch": 0.248,
"grad_norm": 0.3029087781906128,
"learning_rate": 1.3971478906347806e-05,
"loss": 0.0459,
"step": 434
},
{
"epoch": 0.24857142857142858,
"grad_norm": 0.34970352053642273,
"learning_rate": 1.3939419095909513e-05,
"loss": 0.0316,
"step": 435
},
{
"epoch": 0.24914285714285714,
"grad_norm": 0.23229295015335083,
"learning_rate": 1.3907311284892737e-05,
"loss": 0.0329,
"step": 436
},
{
"epoch": 0.24971428571428572,
"grad_norm": 0.043262191116809845,
"learning_rate": 1.3875155864521031e-05,
"loss": 0.0004,
"step": 437
},
{
"epoch": 0.2502857142857143,
"grad_norm": 0.27559104561805725,
"learning_rate": 1.3842953226598036e-05,
"loss": 0.0308,
"step": 438
},
{
"epoch": 0.25085714285714283,
"grad_norm": 0.45299819111824036,
"learning_rate": 1.3810703763502744e-05,
"loss": 0.0368,
"step": 439
},
{
"epoch": 0.25142857142857145,
"grad_norm": 0.22716845571994781,
"learning_rate": 1.3778407868184674e-05,
"loss": 0.0265,
"step": 440
},
{
"epoch": 0.252,
"grad_norm": 0.4089091420173645,
"learning_rate": 1.3746065934159123e-05,
"loss": 0.0326,
"step": 441
},
{
"epoch": 0.25257142857142856,
"grad_norm": 0.37597551941871643,
"learning_rate": 1.371367835550235e-05,
"loss": 0.0309,
"step": 442
},
{
"epoch": 0.25314285714285717,
"grad_norm": 0.18417738378047943,
"learning_rate": 1.3681245526846782e-05,
"loss": 0.0391,
"step": 443
},
{
"epoch": 0.2537142857142857,
"grad_norm": 0.057312093675136566,
"learning_rate": 1.3648767843376196e-05,
"loss": 0.0009,
"step": 444
},
{
"epoch": 0.2542857142857143,
"grad_norm": 0.03737432137131691,
"learning_rate": 1.3616245700820922e-05,
"loss": 0.0002,
"step": 445
},
{
"epoch": 0.25485714285714284,
"grad_norm": 0.16385550796985626,
"learning_rate": 1.3583679495453e-05,
"loss": 0.0018,
"step": 446
},
{
"epoch": 0.25542857142857145,
"grad_norm": 3.4600512981414795,
"learning_rate": 1.3551069624081372e-05,
"loss": 0.0496,
"step": 447
},
{
"epoch": 0.256,
"grad_norm": 0.3149811923503876,
"learning_rate": 1.3518416484047018e-05,
"loss": 0.0424,
"step": 448
},
{
"epoch": 0.25657142857142856,
"grad_norm": 0.40836548805236816,
"learning_rate": 1.3485720473218153e-05,
"loss": 0.0512,
"step": 449
},
{
"epoch": 0.2571428571428571,
"grad_norm": 0.03359152004122734,
"learning_rate": 1.3452981989985347e-05,
"loss": 0.0003,
"step": 450
},
{
"epoch": 0.25771428571428573,
"grad_norm": 0.31589624285697937,
"learning_rate": 1.342020143325669e-05,
"loss": 0.0264,
"step": 451
},
{
"epoch": 0.2582857142857143,
"grad_norm": 1.1004728078842163,
"learning_rate": 1.3387379202452917e-05,
"loss": 0.0059,
"step": 452
},
{
"epoch": 0.25885714285714284,
"grad_norm": 0.06722299754619598,
"learning_rate": 1.3354515697502552e-05,
"loss": 0.0004,
"step": 453
},
{
"epoch": 0.25942857142857145,
"grad_norm": 0.0998629555106163,
"learning_rate": 1.3321611318837033e-05,
"loss": 0.0001,
"step": 454
},
{
"epoch": 0.26,
"grad_norm": 1.229622483253479,
"learning_rate": 1.3288666467385834e-05,
"loss": 0.0453,
"step": 455
},
{
"epoch": 0.26057142857142856,
"grad_norm": 0.13116692006587982,
"learning_rate": 1.3255681544571568e-05,
"loss": 0.002,
"step": 456
},
{
"epoch": 0.2611428571428571,
"grad_norm": 1.4979506731033325,
"learning_rate": 1.3222656952305113e-05,
"loss": 0.1051,
"step": 457
},
{
"epoch": 0.26171428571428573,
"grad_norm": 0.167456716299057,
"learning_rate": 1.3189593092980701e-05,
"loss": 0.0355,
"step": 458
},
{
"epoch": 0.2622857142857143,
"grad_norm": 0.08939709514379501,
"learning_rate": 1.3156490369471026e-05,
"loss": 0.0015,
"step": 459
},
{
"epoch": 0.26285714285714284,
"grad_norm": 0.3433065116405487,
"learning_rate": 1.3123349185122328e-05,
"loss": 0.0374,
"step": 460
},
{
"epoch": 0.2634285714285714,
"grad_norm": 2.149732828140259,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.0155,
"step": 461
},
{
"epoch": 0.264,
"grad_norm": 0.06379012018442154,
"learning_rate": 1.3056953049631059e-05,
"loss": 0.001,
"step": 462
},
{
"epoch": 0.26457142857142857,
"grad_norm": 0.30211102962493896,
"learning_rate": 1.3023698907504447e-05,
"loss": 0.0289,
"step": 463
},
{
"epoch": 0.2651428571428571,
"grad_norm": 0.32812952995300293,
"learning_rate": 1.2990407922560869e-05,
"loss": 0.0476,
"step": 464
},
{
"epoch": 0.26571428571428574,
"grad_norm": 0.43371203541755676,
"learning_rate": 1.2957080500440469e-05,
"loss": 0.0192,
"step": 465
},
{
"epoch": 0.2662857142857143,
"grad_norm": 0.3082626163959503,
"learning_rate": 1.2923717047227368e-05,
"loss": 0.0275,
"step": 466
},
{
"epoch": 0.26685714285714285,
"grad_norm": 0.7108474373817444,
"learning_rate": 1.2890317969444716e-05,
"loss": 0.035,
"step": 467
},
{
"epoch": 0.2674285714285714,
"grad_norm": 0.7971264719963074,
"learning_rate": 1.2856883674049736e-05,
"loss": 0.0682,
"step": 468
},
{
"epoch": 0.268,
"grad_norm": 3.764845371246338,
"learning_rate": 1.2823414568428767e-05,
"loss": 0.0344,
"step": 469
},
{
"epoch": 0.26857142857142857,
"grad_norm": 0.2067488729953766,
"learning_rate": 1.2789911060392295e-05,
"loss": 0.0242,
"step": 470
},
{
"epoch": 0.26914285714285713,
"grad_norm": 0.2490624338388443,
"learning_rate": 1.2756373558169992e-05,
"loss": 0.0264,
"step": 471
},
{
"epoch": 0.26971428571428574,
"grad_norm": 0.3789197504520416,
"learning_rate": 1.2722802470405744e-05,
"loss": 0.02,
"step": 472
},
{
"epoch": 0.2702857142857143,
"grad_norm": 0.03925817459821701,
"learning_rate": 1.2689198206152657e-05,
"loss": 0.0008,
"step": 473
},
{
"epoch": 0.27085714285714285,
"grad_norm": 0.06289826333522797,
"learning_rate": 1.265556117486809e-05,
"loss": 0.0006,
"step": 474
},
{
"epoch": 0.2714285714285714,
"grad_norm": 0.08280473202466965,
"learning_rate": 1.2621891786408648e-05,
"loss": 0.0015,
"step": 475
},
{
"epoch": 0.272,
"grad_norm": 0.6400070190429688,
"learning_rate": 1.2588190451025209e-05,
"loss": 0.0636,
"step": 476
},
{
"epoch": 0.2725714285714286,
"grad_norm": 0.035704053938388824,
"learning_rate": 1.2554457579357906e-05,
"loss": 0.0012,
"step": 477
},
{
"epoch": 0.27314285714285713,
"grad_norm": 0.04904402792453766,
"learning_rate": 1.252069358243114e-05,
"loss": 0.0008,
"step": 478
},
{
"epoch": 0.2737142857142857,
"grad_norm": 0.3700973391532898,
"learning_rate": 1.2486898871648552e-05,
"loss": 0.017,
"step": 479
},
{
"epoch": 0.2742857142857143,
"grad_norm": 1.22110915184021,
"learning_rate": 1.2453073858788027e-05,
"loss": 0.0799,
"step": 480
},
{
"epoch": 0.27485714285714286,
"grad_norm": 0.0287148617208004,
"learning_rate": 1.2419218955996677e-05,
"loss": 0.0005,
"step": 481
},
{
"epoch": 0.2754285714285714,
"grad_norm": 0.039590056985616684,
"learning_rate": 1.238533457578581e-05,
"loss": 0.0004,
"step": 482
},
{
"epoch": 0.276,
"grad_norm": 0.033223431557416916,
"learning_rate": 1.23514211310259e-05,
"loss": 0.0004,
"step": 483
},
{
"epoch": 0.2765714285714286,
"grad_norm": 0.17166505753993988,
"learning_rate": 1.2317479034941572e-05,
"loss": 0.0227,
"step": 484
},
{
"epoch": 0.27714285714285714,
"grad_norm": 0.018312368541955948,
"learning_rate": 1.2283508701106559e-05,
"loss": 0.0007,
"step": 485
},
{
"epoch": 0.2777142857142857,
"grad_norm": 0.30562272667884827,
"learning_rate": 1.2249510543438652e-05,
"loss": 0.0259,
"step": 486
},
{
"epoch": 0.2782857142857143,
"grad_norm": 0.6024219989776611,
"learning_rate": 1.2215484976194675e-05,
"loss": 0.029,
"step": 487
},
{
"epoch": 0.27885714285714286,
"grad_norm": 0.6533939242362976,
"learning_rate": 1.2181432413965428e-05,
"loss": 0.0627,
"step": 488
},
{
"epoch": 0.2794285714285714,
"grad_norm": 3.891444444656372,
"learning_rate": 1.2147353271670634e-05,
"loss": 0.0743,
"step": 489
},
{
"epoch": 0.28,
"grad_norm": 8.625120162963867,
"learning_rate": 1.211324796455389e-05,
"loss": 0.7704,
"step": 490
},
{
"epoch": 0.2805714285714286,
"grad_norm": 0.03895373269915581,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0002,
"step": 491
},
{
"epoch": 0.28114285714285714,
"grad_norm": 0.02734975703060627,
"learning_rate": 1.2044960518417902e-05,
"loss": 0.0005,
"step": 492
},
{
"epoch": 0.2817142857142857,
"grad_norm": 1.216913104057312,
"learning_rate": 1.2010779211459649e-05,
"loss": 0.0917,
"step": 493
},
{
"epoch": 0.2822857142857143,
"grad_norm": 0.3201730251312256,
"learning_rate": 1.1976573403791263e-05,
"loss": 0.0158,
"step": 494
},
{
"epoch": 0.28285714285714286,
"grad_norm": 0.28107988834381104,
"learning_rate": 1.194234351219972e-05,
"loss": 0.0259,
"step": 495
},
{
"epoch": 0.2834285714285714,
"grad_norm": 0.030528711155056953,
"learning_rate": 1.190808995376545e-05,
"loss": 0.0008,
"step": 496
},
{
"epoch": 0.284,
"grad_norm": 0.16229555010795593,
"learning_rate": 1.187381314585725e-05,
"loss": 0.0031,
"step": 497
},
{
"epoch": 0.2845714285714286,
"grad_norm": 0.033382948487997055,
"learning_rate": 1.1839513506127202e-05,
"loss": 0.0011,
"step": 498
},
{
"epoch": 0.28514285714285714,
"grad_norm": 0.4741803705692291,
"learning_rate": 1.1805191452505602e-05,
"loss": 0.0472,
"step": 499
},
{
"epoch": 0.2857142857142857,
"grad_norm": 0.3844754993915558,
"learning_rate": 1.1770847403195836e-05,
"loss": 0.0525,
"step": 500
},
{
"epoch": 0.2862857142857143,
"grad_norm": 0.5735390186309814,
"learning_rate": 1.1736481776669307e-05,
"loss": 0.0616,
"step": 501
},
{
"epoch": 0.28685714285714287,
"grad_norm": 0.02491656504571438,
"learning_rate": 1.1702094991660326e-05,
"loss": 0.0004,
"step": 502
},
{
"epoch": 0.2874285714285714,
"grad_norm": 0.16720333695411682,
"learning_rate": 1.1667687467161025e-05,
"loss": 0.0243,
"step": 503
},
{
"epoch": 0.288,
"grad_norm": 0.3294336199760437,
"learning_rate": 1.1633259622416224e-05,
"loss": 0.0276,
"step": 504
},
{
"epoch": 0.2885714285714286,
"grad_norm": 0.06401587277650833,
"learning_rate": 1.159881187691835e-05,
"loss": 0.0014,
"step": 505
},
{
"epoch": 0.28914285714285715,
"grad_norm": 1.1847069263458252,
"learning_rate": 1.156434465040231e-05,
"loss": 0.0335,
"step": 506
},
{
"epoch": 0.2897142857142857,
"grad_norm": 0.23786140978336334,
"learning_rate": 1.1529858362840383e-05,
"loss": 0.0248,
"step": 507
},
{
"epoch": 0.29028571428571426,
"grad_norm": 0.4833123981952667,
"learning_rate": 1.1495353434437098e-05,
"loss": 0.0601,
"step": 508
},
{
"epoch": 0.29085714285714287,
"grad_norm": 0.21625486016273499,
"learning_rate": 1.1460830285624119e-05,
"loss": 0.024,
"step": 509
},
{
"epoch": 0.2914285714285714,
"grad_norm": 0.23418869078159332,
"learning_rate": 1.1426289337055119e-05,
"loss": 0.0335,
"step": 510
},
{
"epoch": 0.292,
"grad_norm": 0.08382096886634827,
"learning_rate": 1.1391731009600655e-05,
"loss": 0.0016,
"step": 511
},
{
"epoch": 0.2925714285714286,
"grad_norm": 0.034197673201560974,
"learning_rate": 1.1357155724343046e-05,
"loss": 0.0005,
"step": 512
},
{
"epoch": 0.29314285714285715,
"grad_norm": 0.3640075623989105,
"learning_rate": 1.1322563902571227e-05,
"loss": 0.0535,
"step": 513
},
{
"epoch": 0.2937142857142857,
"grad_norm": 0.42086243629455566,
"learning_rate": 1.128795596577563e-05,
"loss": 0.0529,
"step": 514
},
{
"epoch": 0.29428571428571426,
"grad_norm": 0.2768264412879944,
"learning_rate": 1.1253332335643043e-05,
"loss": 0.0266,
"step": 515
},
{
"epoch": 0.2948571428571429,
"grad_norm": 0.398020476102829,
"learning_rate": 1.1218693434051475e-05,
"loss": 0.0511,
"step": 516
},
{
"epoch": 0.29542857142857143,
"grad_norm": 0.5826377868652344,
"learning_rate": 1.1184039683065014e-05,
"loss": 0.0291,
"step": 517
},
{
"epoch": 0.296,
"grad_norm": 0.027462149038910866,
"learning_rate": 1.1149371504928667e-05,
"loss": 0.0009,
"step": 518
},
{
"epoch": 0.2965714285714286,
"grad_norm": 0.30697697401046753,
"learning_rate": 1.1114689322063255e-05,
"loss": 0.0265,
"step": 519
},
{
"epoch": 0.29714285714285715,
"grad_norm": 0.0325193777680397,
"learning_rate": 1.1079993557060228e-05,
"loss": 0.001,
"step": 520
},
{
"epoch": 0.2977142857142857,
"grad_norm": 0.2791683077812195,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.0335,
"step": 521
},
{
"epoch": 0.29828571428571427,
"grad_norm": 0.19040565192699432,
"learning_rate": 1.1010562971829464e-05,
"loss": 0.0364,
"step": 522
},
{
"epoch": 0.2988571428571429,
"grad_norm": 2.0151286125183105,
"learning_rate": 1.0975828997591496e-05,
"loss": 0.0142,
"step": 523
},
{
"epoch": 0.29942857142857143,
"grad_norm": 2.2987611293792725,
"learning_rate": 1.0941083133185146e-05,
"loss": 0.0572,
"step": 524
},
{
"epoch": 0.3,
"grad_norm": 0.4411211609840393,
"learning_rate": 1.0906325801977804e-05,
"loss": 0.003,
"step": 525
},
{
"epoch": 0.30057142857142854,
"grad_norm": 0.9109564423561096,
"learning_rate": 1.0871557427476585e-05,
"loss": 0.0422,
"step": 526
},
{
"epoch": 0.30114285714285716,
"grad_norm": 0.37157005071640015,
"learning_rate": 1.083677843332316e-05,
"loss": 0.0554,
"step": 527
},
{
"epoch": 0.3017142857142857,
"grad_norm": 0.2623600661754608,
"learning_rate": 1.0801989243288588e-05,
"loss": 0.026,
"step": 528
},
{
"epoch": 0.30228571428571427,
"grad_norm": 0.5420719981193542,
"learning_rate": 1.0767190281268187e-05,
"loss": 0.0598,
"step": 529
},
{
"epoch": 0.3028571428571429,
"grad_norm": 20.352373123168945,
"learning_rate": 1.0732381971276318e-05,
"loss": 0.1899,
"step": 530
},
{
"epoch": 0.30342857142857144,
"grad_norm": 0.35633453726768494,
"learning_rate": 1.0697564737441254e-05,
"loss": 0.0356,
"step": 531
},
{
"epoch": 0.304,
"grad_norm": 0.26073285937309265,
"learning_rate": 1.0662739004000005e-05,
"loss": 0.025,
"step": 532
},
{
"epoch": 0.30457142857142855,
"grad_norm": 0.039687953889369965,
"learning_rate": 1.0627905195293135e-05,
"loss": 0.0006,
"step": 533
},
{
"epoch": 0.30514285714285716,
"grad_norm": 0.04971844702959061,
"learning_rate": 1.0593063735759619e-05,
"loss": 0.0013,
"step": 534
},
{
"epoch": 0.3057142857142857,
"grad_norm": 0.2272334098815918,
"learning_rate": 1.055821504993164e-05,
"loss": 0.0122,
"step": 535
},
{
"epoch": 0.3062857142857143,
"grad_norm": 1.2308039665222168,
"learning_rate": 1.0523359562429441e-05,
"loss": 0.0107,
"step": 536
},
{
"epoch": 0.3068571428571429,
"grad_norm": 0.32566148042678833,
"learning_rate": 1.0488497697956134e-05,
"loss": 0.0533,
"step": 537
},
{
"epoch": 0.30742857142857144,
"grad_norm": 0.02191033959388733,
"learning_rate": 1.0453629881292537e-05,
"loss": 0.0007,
"step": 538
},
{
"epoch": 0.308,
"grad_norm": 1.6923489570617676,
"learning_rate": 1.0418756537291996e-05,
"loss": 0.0067,
"step": 539
},
{
"epoch": 0.30857142857142855,
"grad_norm": 0.3687346577644348,
"learning_rate": 1.03838780908752e-05,
"loss": 0.0295,
"step": 540
},
{
"epoch": 0.30914285714285716,
"grad_norm": 0.03369167447090149,
"learning_rate": 1.0348994967025012e-05,
"loss": 0.0009,
"step": 541
},
{
"epoch": 0.3097142857142857,
"grad_norm": 0.08542211353778839,
"learning_rate": 1.0314107590781284e-05,
"loss": 0.0011,
"step": 542
},
{
"epoch": 0.3102857142857143,
"grad_norm": 0.20910456776618958,
"learning_rate": 1.0279216387235691e-05,
"loss": 0.0405,
"step": 543
},
{
"epoch": 0.31085714285714283,
"grad_norm": 0.0294099822640419,
"learning_rate": 1.0244321781526533e-05,
"loss": 0.0015,
"step": 544
},
{
"epoch": 0.31142857142857144,
"grad_norm": 0.20268481969833374,
"learning_rate": 1.0209424198833571e-05,
"loss": 0.0035,
"step": 545
},
{
"epoch": 0.312,
"grad_norm": 0.285698801279068,
"learning_rate": 1.0174524064372837e-05,
"loss": 0.0132,
"step": 546
},
{
"epoch": 0.31257142857142856,
"grad_norm": 11.620792388916016,
"learning_rate": 1.0139621803391454e-05,
"loss": 0.1615,
"step": 547
},
{
"epoch": 0.31314285714285717,
"grad_norm": 0.5434874296188354,
"learning_rate": 1.010471784116246e-05,
"loss": 0.0001,
"step": 548
},
{
"epoch": 0.3137142857142857,
"grad_norm": 0.01930747553706169,
"learning_rate": 1.0069812602979617e-05,
"loss": 0.0006,
"step": 549
},
{
"epoch": 0.3142857142857143,
"grad_norm": 0.05357842147350311,
"learning_rate": 1.0034906514152239e-05,
"loss": 0.0016,
"step": 550
},
{
"epoch": 0.31485714285714284,
"grad_norm": 0.30916890501976013,
"learning_rate": 1e-05,
"loss": 0.0488,
"step": 551
},
{
"epoch": 0.31542857142857145,
"grad_norm": 0.28261929750442505,
"learning_rate": 9.965093485847766e-06,
"loss": 0.031,
"step": 552
},
{
"epoch": 0.316,
"grad_norm": 0.12484882771968842,
"learning_rate": 9.930187397020385e-06,
"loss": 0.002,
"step": 553
},
{
"epoch": 0.31657142857142856,
"grad_norm": 0.15603572130203247,
"learning_rate": 9.895282158837545e-06,
"loss": 0.0225,
"step": 554
},
{
"epoch": 0.3171428571428571,
"grad_norm": 0.22738535702228546,
"learning_rate": 9.860378196608549e-06,
"loss": 0.0242,
"step": 555
},
{
"epoch": 0.3177142857142857,
"grad_norm": 0.235463947057724,
"learning_rate": 9.825475935627165e-06,
"loss": 0.0249,
"step": 556
},
{
"epoch": 0.3182857142857143,
"grad_norm": 0.053955066949129105,
"learning_rate": 9.790575801166432e-06,
"loss": 0.0011,
"step": 557
},
{
"epoch": 0.31885714285714284,
"grad_norm": 0.4049069285392761,
"learning_rate": 9.75567821847347e-06,
"loss": 0.0307,
"step": 558
},
{
"epoch": 0.31942857142857145,
"grad_norm": 0.23345054686069489,
"learning_rate": 9.720783612764314e-06,
"loss": 0.0118,
"step": 559
},
{
"epoch": 0.32,
"grad_norm": 0.5557246208190918,
"learning_rate": 9.685892409218718e-06,
"loss": 0.0009,
"step": 560
},
{
"epoch": 0.32057142857142856,
"grad_norm": 21.1752872467041,
"learning_rate": 9.651005032974994e-06,
"loss": 0.1139,
"step": 561
},
{
"epoch": 0.3211428571428571,
"grad_norm": 0.029627056792378426,
"learning_rate": 9.616121909124801e-06,
"loss": 0.0009,
"step": 562
},
{
"epoch": 0.32171428571428573,
"grad_norm": 19.0615177154541,
"learning_rate": 9.581243462708007e-06,
"loss": 0.6573,
"step": 563
},
{
"epoch": 0.3222857142857143,
"grad_norm": 0.015228305011987686,
"learning_rate": 9.546370118707463e-06,
"loss": 0.001,
"step": 564
},
{
"epoch": 0.32285714285714284,
"grad_norm": 51.37855529785156,
"learning_rate": 9.511502302043867e-06,
"loss": 0.1337,
"step": 565
},
{
"epoch": 0.32342857142857145,
"grad_norm": 0.41825437545776367,
"learning_rate": 9.476640437570562e-06,
"loss": 0.0486,
"step": 566
},
{
"epoch": 0.324,
"grad_norm": 0.15489810705184937,
"learning_rate": 9.441784950068362e-06,
"loss": 0.023,
"step": 567
},
{
"epoch": 0.32457142857142857,
"grad_norm": 0.3880479037761688,
"learning_rate": 9.406936264240386e-06,
"loss": 0.0262,
"step": 568
},
{
"epoch": 0.3251428571428571,
"grad_norm": 0.09499350190162659,
"learning_rate": 9.372094804706867e-06,
"loss": 0.0008,
"step": 569
},
{
"epoch": 0.32571428571428573,
"grad_norm": 0.05607116222381592,
"learning_rate": 9.337260996000002e-06,
"loss": 0.0,
"step": 570
},
{
"epoch": 0.3262857142857143,
"grad_norm": 0.2441244125366211,
"learning_rate": 9.302435262558748e-06,
"loss": 0.0343,
"step": 571
},
{
"epoch": 0.32685714285714285,
"grad_norm": 0.4808112382888794,
"learning_rate": 9.267618028723687e-06,
"loss": 0.0041,
"step": 572
},
{
"epoch": 0.3274285714285714,
"grad_norm": 0.0376475527882576,
"learning_rate": 9.232809718731815e-06,
"loss": 0.0006,
"step": 573
},
{
"epoch": 0.328,
"grad_norm": 0.3752923607826233,
"learning_rate": 9.198010756711413e-06,
"loss": 0.0485,
"step": 574
},
{
"epoch": 0.32857142857142857,
"grad_norm": 0.4132545590400696,
"learning_rate": 9.163221566676847e-06,
"loss": 0.0562,
"step": 575
},
{
"epoch": 0.3291428571428571,
"grad_norm": 0.3536934554576874,
"learning_rate": 9.128442572523418e-06,
"loss": 0.0355,
"step": 576
},
{
"epoch": 0.32971428571428574,
"grad_norm": 0.3179832994937897,
"learning_rate": 9.093674198022201e-06,
"loss": 0.0453,
"step": 577
},
{
"epoch": 0.3302857142857143,
"grad_norm": 0.049885306507349014,
"learning_rate": 9.058916866814857e-06,
"loss": 0.0009,
"step": 578
},
{
"epoch": 0.33085714285714285,
"grad_norm": 0.3605054020881653,
"learning_rate": 9.024171002408507e-06,
"loss": 0.0021,
"step": 579
},
{
"epoch": 0.3314285714285714,
"grad_norm": 0.5206215977668762,
"learning_rate": 8.989437028170537e-06,
"loss": 0.0529,
"step": 580
},
{
"epoch": 0.332,
"grad_norm": 0.25148093700408936,
"learning_rate": 8.954715367323468e-06,
"loss": 0.0343,
"step": 581
},
{
"epoch": 0.3325714285714286,
"grad_norm": 0.4498104155063629,
"learning_rate": 8.920006442939772e-06,
"loss": 0.0538,
"step": 582
},
{
"epoch": 0.33314285714285713,
"grad_norm": 0.16591976583003998,
"learning_rate": 8.885310677936746e-06,
"loss": 0.0225,
"step": 583
},
{
"epoch": 0.33371428571428574,
"grad_norm": 1.2063405513763428,
"learning_rate": 8.850628495071336e-06,
"loss": 0.0716,
"step": 584
},
{
"epoch": 0.3342857142857143,
"grad_norm": 0.11695819348096848,
"learning_rate": 8.815960316934991e-06,
"loss": 0.0023,
"step": 585
},
{
"epoch": 0.33485714285714285,
"grad_norm": 0.22231541574001312,
"learning_rate": 8.781306565948528e-06,
"loss": 0.0111,
"step": 586
},
{
"epoch": 0.3354285714285714,
"grad_norm": 1.479414463043213,
"learning_rate": 8.746667664356957e-06,
"loss": 0.0306,
"step": 587
},
{
"epoch": 0.336,
"grad_norm": 0.39188915491104126,
"learning_rate": 8.712044034224374e-06,
"loss": 0.0534,
"step": 588
},
{
"epoch": 0.3365714285714286,
"grad_norm": 0.9432989954948425,
"learning_rate": 8.677436097428775e-06,
"loss": 0.0961,
"step": 589
},
{
"epoch": 0.33714285714285713,
"grad_norm": 0.48859044909477234,
"learning_rate": 8.642844275656957e-06,
"loss": 0.0338,
"step": 590
},
{
"epoch": 0.3377142857142857,
"grad_norm": 0.1998908519744873,
"learning_rate": 8.60826899039935e-06,
"loss": 0.0382,
"step": 591
},
{
"epoch": 0.3382857142857143,
"grad_norm": 0.19373755156993866,
"learning_rate": 8.573710662944884e-06,
"loss": 0.0414,
"step": 592
},
{
"epoch": 0.33885714285714286,
"grad_norm": 0.095662422478199,
"learning_rate": 8.539169714375885e-06,
"loss": 0.0015,
"step": 593
},
{
"epoch": 0.3394285714285714,
"grad_norm": 0.044992219656705856,
"learning_rate": 8.504646565562907e-06,
"loss": 0.0012,
"step": 594
},
{
"epoch": 0.34,
"grad_norm": 0.3015923500061035,
"learning_rate": 8.47014163715962e-06,
"loss": 0.0426,
"step": 595
},
{
"epoch": 0.3405714285714286,
"grad_norm": 0.19642798602581024,
"learning_rate": 8.43565534959769e-06,
"loss": 0.0152,
"step": 596
},
{
"epoch": 0.34114285714285714,
"grad_norm": 0.9269797801971436,
"learning_rate": 8.401188123081653e-06,
"loss": 0.0089,
"step": 597
},
{
"epoch": 0.3417142857142857,
"grad_norm": 10.957679748535156,
"learning_rate": 8.366740377583781e-06,
"loss": 0.0274,
"step": 598
},
{
"epoch": 0.3422857142857143,
"grad_norm": 0.09165984392166138,
"learning_rate": 8.332312532838978e-06,
"loss": 0.0011,
"step": 599
},
{
"epoch": 0.34285714285714286,
"grad_norm": 0.2328844964504242,
"learning_rate": 8.297905008339677e-06,
"loss": 0.043,
"step": 600
},
{
"epoch": 0.3434285714285714,
"grad_norm": 0.3168167173862457,
"learning_rate": 8.263518223330698e-06,
"loss": 0.0291,
"step": 601
},
{
"epoch": 0.344,
"grad_norm": 0.17060230672359467,
"learning_rate": 8.22915259680417e-06,
"loss": 0.0356,
"step": 602
},
{
"epoch": 0.3445714285714286,
"grad_norm": 0.2511342763900757,
"learning_rate": 8.194808547494401e-06,
"loss": 0.0252,
"step": 603
},
{
"epoch": 0.34514285714285714,
"grad_norm": 0.03633415326476097,
"learning_rate": 8.1604864938728e-06,
"loss": 0.0005,
"step": 604
},
{
"epoch": 0.3457142857142857,
"grad_norm": 12.919232368469238,
"learning_rate": 8.126186854142752e-06,
"loss": 0.0878,
"step": 605
},
{
"epoch": 0.3462857142857143,
"grad_norm": 0.04500804841518402,
"learning_rate": 8.091910046234552e-06,
"loss": 0.0012,
"step": 606
},
{
"epoch": 0.34685714285714286,
"grad_norm": 0.33597058057785034,
"learning_rate": 8.057656487800283e-06,
"loss": 0.0497,
"step": 607
},
{
"epoch": 0.3474285714285714,
"grad_norm": 0.17340946197509766,
"learning_rate": 8.023426596208739e-06,
"loss": 0.0409,
"step": 608
},
{
"epoch": 0.348,
"grad_norm": 0.3938964903354645,
"learning_rate": 7.989220788540356e-06,
"loss": 0.0303,
"step": 609
},
{
"epoch": 0.3485714285714286,
"grad_norm": 0.03415045887231827,
"learning_rate": 7.955039481582098e-06,
"loss": 0.0006,
"step": 610
},
{
"epoch": 0.34914285714285714,
"grad_norm": 0.40274906158447266,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0319,
"step": 611
},
{
"epoch": 0.3497142857142857,
"grad_norm": 0.27646389603614807,
"learning_rate": 7.886752035446116e-06,
"loss": 0.0524,
"step": 612
},
{
"epoch": 0.3502857142857143,
"grad_norm": 0.07450652867555618,
"learning_rate": 7.852646728329368e-06,
"loss": 0.0018,
"step": 613
},
{
"epoch": 0.35085714285714287,
"grad_norm": 0.06656794995069504,
"learning_rate": 7.818567586034578e-06,
"loss": 0.001,
"step": 614
},
{
"epoch": 0.3514285714285714,
"grad_norm": 0.03880167007446289,
"learning_rate": 7.784515023805328e-06,
"loss": 0.0,
"step": 615
},
{
"epoch": 0.352,
"grad_norm": 0.05078655108809471,
"learning_rate": 7.750489456561351e-06,
"loss": 0.0009,
"step": 616
},
{
"epoch": 0.3525714285714286,
"grad_norm": 0.11694870889186859,
"learning_rate": 7.716491298893443e-06,
"loss": 0.0022,
"step": 617
},
{
"epoch": 0.35314285714285715,
"grad_norm": 0.9499900341033936,
"learning_rate": 7.68252096505843e-06,
"loss": 0.0685,
"step": 618
},
{
"epoch": 0.3537142857142857,
"grad_norm": 0.18475067615509033,
"learning_rate": 7.6485788689741e-06,
"loss": 0.0409,
"step": 619
},
{
"epoch": 0.35428571428571426,
"grad_norm": 0.3411336839199066,
"learning_rate": 7.6146654242141935e-06,
"loss": 0.0452,
"step": 620
},
{
"epoch": 0.35485714285714287,
"grad_norm": 1.0038598775863647,
"learning_rate": 7.580781044003324e-06,
"loss": 0.0326,
"step": 621
},
{
"epoch": 0.3554285714285714,
"grad_norm": 1.7273715734481812,
"learning_rate": 7.546926141211975e-06,
"loss": 0.1192,
"step": 622
},
{
"epoch": 0.356,
"grad_norm": 0.39912572503089905,
"learning_rate": 7.513101128351454e-06,
"loss": 0.0535,
"step": 623
},
{
"epoch": 0.3565714285714286,
"grad_norm": 0.05744081735610962,
"learning_rate": 7.4793064175688635e-06,
"loss": 0.0016,
"step": 624
},
{
"epoch": 0.35714285714285715,
"grad_norm": 0.3906503915786743,
"learning_rate": 7.445542420642097e-06,
"loss": 0.0463,
"step": 625
},
{
"epoch": 0.3577142857142857,
"grad_norm": 0.07783088088035583,
"learning_rate": 7.411809548974792e-06,
"loss": 0.0027,
"step": 626
},
{
"epoch": 0.35828571428571426,
"grad_norm": 7.119018077850342,
"learning_rate": 7.378108213591355e-06,
"loss": 0.0981,
"step": 627
},
{
"epoch": 0.3588571428571429,
"grad_norm": 0.31130439043045044,
"learning_rate": 7.344438825131912e-06,
"loss": 0.0361,
"step": 628
},
{
"epoch": 0.35942857142857143,
"grad_norm": 0.033409979194402695,
"learning_rate": 7.310801793847344e-06,
"loss": 0.0005,
"step": 629
},
{
"epoch": 0.36,
"grad_norm": 0.12457820773124695,
"learning_rate": 7.277197529594257e-06,
"loss": 0.0338,
"step": 630
},
{
"epoch": 0.3605714285714286,
"grad_norm": 0.3446294665336609,
"learning_rate": 7.243626441830009e-06,
"loss": 0.0487,
"step": 631
},
{
"epoch": 0.36114285714285715,
"grad_norm": 0.03503461182117462,
"learning_rate": 7.210088939607709e-06,
"loss": 0.0007,
"step": 632
},
{
"epoch": 0.3617142857142857,
"grad_norm": 0.303161084651947,
"learning_rate": 7.176585431571235e-06,
"loss": 0.0324,
"step": 633
},
{
"epoch": 0.36228571428571427,
"grad_norm": 0.16304157674312592,
"learning_rate": 7.143116325950266e-06,
"loss": 0.0391,
"step": 634
},
{
"epoch": 0.3628571428571429,
"grad_norm": 0.43130961060523987,
"learning_rate": 7.109682030555283e-06,
"loss": 0.0331,
"step": 635
},
{
"epoch": 0.36342857142857143,
"grad_norm": 0.2779071629047394,
"learning_rate": 7.076282952772634e-06,
"loss": 0.0396,
"step": 636
},
{
"epoch": 0.364,
"grad_norm": 0.3328837752342224,
"learning_rate": 7.042919499559538e-06,
"loss": 0.0514,
"step": 637
},
{
"epoch": 0.36457142857142855,
"grad_norm": 0.34643644094467163,
"learning_rate": 7.009592077439135e-06,
"loss": 0.052,
"step": 638
},
{
"epoch": 0.36514285714285716,
"grad_norm": 0.02169707603752613,
"learning_rate": 6.976301092495556e-06,
"loss": 0.0005,
"step": 639
},
{
"epoch": 0.3657142857142857,
"grad_norm": 0.42019087076187134,
"learning_rate": 6.943046950368944e-06,
"loss": 0.0327,
"step": 640
},
{
"epoch": 0.36628571428571427,
"grad_norm": 0.1543130725622177,
"learning_rate": 6.909830056250527e-06,
"loss": 0.0268,
"step": 641
},
{
"epoch": 0.3668571428571429,
"grad_norm": 0.4150610566139221,
"learning_rate": 6.876650814877675e-06,
"loss": 0.0326,
"step": 642
},
{
"epoch": 0.36742857142857144,
"grad_norm": 1.1808620691299438,
"learning_rate": 6.843509630528977e-06,
"loss": 0.0371,
"step": 643
},
{
"epoch": 0.368,
"grad_norm": 0.30772918462753296,
"learning_rate": 6.8104069070193e-06,
"loss": 0.0271,
"step": 644
},
{
"epoch": 0.36857142857142855,
"grad_norm": 0.2798471450805664,
"learning_rate": 6.777343047694891e-06,
"loss": 0.0264,
"step": 645
},
{
"epoch": 0.36914285714285716,
"grad_norm": 0.11227493733167648,
"learning_rate": 6.744318455428436e-06,
"loss": 0.0006,
"step": 646
},
{
"epoch": 0.3697142857142857,
"grad_norm": 0.013413318432867527,
"learning_rate": 6.711333532614168e-06,
"loss": 0.0012,
"step": 647
},
{
"epoch": 0.3702857142857143,
"grad_norm": 0.017493678256869316,
"learning_rate": 6.67838868116297e-06,
"loss": 0.0005,
"step": 648
},
{
"epoch": 0.37085714285714283,
"grad_norm": 1.4520387649536133,
"learning_rate": 6.645484302497452e-06,
"loss": 0.009,
"step": 649
},
{
"epoch": 0.37142857142857144,
"grad_norm": 0.014982587657868862,
"learning_rate": 6.612620797547087e-06,
"loss": 0.0012,
"step": 650
},
{
"epoch": 0.372,
"grad_norm": 0.21275031566619873,
"learning_rate": 6.579798566743314e-06,
"loss": 0.0257,
"step": 651
},
{
"epoch": 0.37257142857142855,
"grad_norm": 0.1510966271162033,
"learning_rate": 6.547018010014654e-06,
"loss": 0.0221,
"step": 652
},
{
"epoch": 0.37314285714285716,
"grad_norm": 0.7374513745307922,
"learning_rate": 6.5142795267818505e-06,
"loss": 0.0174,
"step": 653
},
{
"epoch": 0.3737142857142857,
"grad_norm": 0.0682489350438118,
"learning_rate": 6.481583515952983e-06,
"loss": 0.0012,
"step": 654
},
{
"epoch": 0.3742857142857143,
"grad_norm": 0.04243054240942001,
"learning_rate": 6.448930375918632e-06,
"loss": 0.0007,
"step": 655
},
{
"epoch": 0.37485714285714283,
"grad_norm": 0.025167247280478477,
"learning_rate": 6.4163205045469975e-06,
"loss": 0.0003,
"step": 656
},
{
"epoch": 0.37542857142857144,
"grad_norm": 0.49935224652290344,
"learning_rate": 6.383754299179079e-06,
"loss": 0.0024,
"step": 657
},
{
"epoch": 0.376,
"grad_norm": 11.183856964111328,
"learning_rate": 6.351232156623803e-06,
"loss": 0.1425,
"step": 658
},
{
"epoch": 0.37657142857142856,
"grad_norm": 0.3608929514884949,
"learning_rate": 6.318754473153221e-06,
"loss": 0.051,
"step": 659
},
{
"epoch": 0.37714285714285717,
"grad_norm": 0.36358442902565,
"learning_rate": 6.286321644497655e-06,
"loss": 0.0289,
"step": 660
},
{
"epoch": 0.3777142857142857,
"grad_norm": 0.026930978521704674,
"learning_rate": 6.25393406584088e-06,
"loss": 0.0004,
"step": 661
},
{
"epoch": 0.3782857142857143,
"grad_norm": 0.31257733702659607,
"learning_rate": 6.22159213181533e-06,
"loss": 0.0481,
"step": 662
},
{
"epoch": 0.37885714285714284,
"grad_norm": 0.018814746290445328,
"learning_rate": 6.18929623649726e-06,
"loss": 0.0007,
"step": 663
},
{
"epoch": 0.37942857142857145,
"grad_norm": 0.1458250880241394,
"learning_rate": 6.157046773401964e-06,
"loss": 0.022,
"step": 664
},
{
"epoch": 0.38,
"grad_norm": 5.4401350021362305,
"learning_rate": 6.124844135478971e-06,
"loss": 0.0151,
"step": 665
},
{
"epoch": 0.38057142857142856,
"grad_norm": 0.2785094082355499,
"learning_rate": 6.092688715107265e-06,
"loss": 0.0492,
"step": 666
},
{
"epoch": 0.3811428571428571,
"grad_norm": 0.027348000556230545,
"learning_rate": 6.06058090409049e-06,
"loss": 0.0013,
"step": 667
},
{
"epoch": 0.38171428571428573,
"grad_norm": 0.1804162710905075,
"learning_rate": 6.028521093652195e-06,
"loss": 0.0229,
"step": 668
},
{
"epoch": 0.3822857142857143,
"grad_norm": 0.35327380895614624,
"learning_rate": 5.996509674431053e-06,
"loss": 0.0529,
"step": 669
},
{
"epoch": 0.38285714285714284,
"grad_norm": 13.218514442443848,
"learning_rate": 5.9645470364761e-06,
"loss": 0.1987,
"step": 670
},
{
"epoch": 0.38342857142857145,
"grad_norm": 0.05170697718858719,
"learning_rate": 5.932633569242e-06,
"loss": 0.0016,
"step": 671
},
{
"epoch": 0.384,
"grad_norm": 0.16022132337093353,
"learning_rate": 5.900769661584273e-06,
"loss": 0.0361,
"step": 672
},
{
"epoch": 0.38457142857142856,
"grad_norm": 0.2107943296432495,
"learning_rate": 5.868955701754584e-06,
"loss": 0.0003,
"step": 673
},
{
"epoch": 0.3851428571428571,
"grad_norm": 0.1718650460243225,
"learning_rate": 5.83719207739599e-06,
"loss": 0.0352,
"step": 674
},
{
"epoch": 0.38571428571428573,
"grad_norm": 0.03544533997774124,
"learning_rate": 5.8054791755382286e-06,
"loss": 0.0009,
"step": 675
},
{
"epoch": 0.3862857142857143,
"grad_norm": 0.360246479511261,
"learning_rate": 5.773817382593008e-06,
"loss": 0.0277,
"step": 676
},
{
"epoch": 0.38685714285714284,
"grad_norm": 0.023578330874443054,
"learning_rate": 5.742207084349274e-06,
"loss": 0.0003,
"step": 677
},
{
"epoch": 0.38742857142857146,
"grad_norm": 0.22791242599487305,
"learning_rate": 5.710648665968543e-06,
"loss": 0.0442,
"step": 678
},
{
"epoch": 0.388,
"grad_norm": 0.2239474654197693,
"learning_rate": 5.679142511980176e-06,
"loss": 0.0091,
"step": 679
},
{
"epoch": 0.38857142857142857,
"grad_norm": 0.22868044674396515,
"learning_rate": 5.647689006276727e-06,
"loss": 0.0253,
"step": 680
},
{
"epoch": 0.3891428571428571,
"grad_norm": 22.384647369384766,
"learning_rate": 5.616288532109225e-06,
"loss": 0.2859,
"step": 681
},
{
"epoch": 0.38971428571428574,
"grad_norm": 0.15848632156848907,
"learning_rate": 5.584941472082549e-06,
"loss": 0.0021,
"step": 682
},
{
"epoch": 0.3902857142857143,
"grad_norm": 0.09392625093460083,
"learning_rate": 5.553648208150728e-06,
"loss": 0.0014,
"step": 683
},
{
"epoch": 0.39085714285714285,
"grad_norm": 0.029840093106031418,
"learning_rate": 5.522409121612304e-06,
"loss": 0.0011,
"step": 684
},
{
"epoch": 0.3914285714285714,
"grad_norm": 0.1226116493344307,
"learning_rate": 5.491224593105695e-06,
"loss": 0.0213,
"step": 685
},
{
"epoch": 0.392,
"grad_norm": 0.19050206243991852,
"learning_rate": 5.460095002604533e-06,
"loss": 0.0316,
"step": 686
},
{
"epoch": 0.39257142857142857,
"grad_norm": 1.0030931234359741,
"learning_rate": 5.429020729413062e-06,
"loss": 0.0662,
"step": 687
},
{
"epoch": 0.3931428571428571,
"grad_norm": 2.1091763973236084,
"learning_rate": 5.398002152161484e-06,
"loss": 0.0065,
"step": 688
},
{
"epoch": 0.39371428571428574,
"grad_norm": 22.139360427856445,
"learning_rate": 5.367039648801386e-06,
"loss": 0.1181,
"step": 689
},
{
"epoch": 0.3942857142857143,
"grad_norm": 0.03205937147140503,
"learning_rate": 5.336133596601089e-06,
"loss": 0.0009,
"step": 690
},
{
"epoch": 0.39485714285714285,
"grad_norm": 0.020037170499563217,
"learning_rate": 5.305284372141095e-06,
"loss": 0.0005,
"step": 691
},
{
"epoch": 0.3954285714285714,
"grad_norm": 0.2879824638366699,
"learning_rate": 5.274492351309462e-06,
"loss": 0.0462,
"step": 692
},
{
"epoch": 0.396,
"grad_norm": 1.364968180656433,
"learning_rate": 5.243757909297247e-06,
"loss": 0.0874,
"step": 693
},
{
"epoch": 0.3965714285714286,
"grad_norm": 0.11832071840763092,
"learning_rate": 5.213081420593933e-06,
"loss": 0.0212,
"step": 694
},
{
"epoch": 0.39714285714285713,
"grad_norm": 0.5274502038955688,
"learning_rate": 5.1824632589828465e-06,
"loss": 0.0247,
"step": 695
},
{
"epoch": 0.3977142857142857,
"grad_norm": 0.42063596844673157,
"learning_rate": 5.151903797536631e-06,
"loss": 0.0534,
"step": 696
},
{
"epoch": 0.3982857142857143,
"grad_norm": 0.041581861674785614,
"learning_rate": 5.121403408612672e-06,
"loss": 0.0008,
"step": 697
},
{
"epoch": 0.39885714285714285,
"grad_norm": 0.459882915019989,
"learning_rate": 5.090962463848592e-06,
"loss": 0.0585,
"step": 698
},
{
"epoch": 0.3994285714285714,
"grad_norm": 0.28338322043418884,
"learning_rate": 5.060581334157693e-06,
"loss": 0.0303,
"step": 699
},
{
"epoch": 0.4,
"grad_norm": 0.357791930437088,
"learning_rate": 5.030260389724447e-06,
"loss": 0.044,
"step": 700
},
{
"epoch": 0.4005714285714286,
"grad_norm": 0.3839028775691986,
"learning_rate": 5.000000000000003e-06,
"loss": 0.0269,
"step": 701
},
{
"epoch": 0.40114285714285713,
"grad_norm": 0.022097395732998848,
"learning_rate": 4.96980053369765e-06,
"loss": 0.0004,
"step": 702
},
{
"epoch": 0.4017142857142857,
"grad_norm": 0.09089873731136322,
"learning_rate": 4.939662358788364e-06,
"loss": 0.0016,
"step": 703
},
{
"epoch": 0.4022857142857143,
"grad_norm": 0.02409243956208229,
"learning_rate": 4.909585842496287e-06,
"loss": 0.0009,
"step": 704
},
{
"epoch": 0.40285714285714286,
"grad_norm": 0.3183746635913849,
"learning_rate": 4.879571351294287e-06,
"loss": 0.0283,
"step": 705
},
{
"epoch": 0.4034285714285714,
"grad_norm": 0.2865620255470276,
"learning_rate": 4.849619250899458e-06,
"loss": 0.0434,
"step": 706
},
{
"epoch": 0.404,
"grad_norm": 0.03536757081747055,
"learning_rate": 4.8197299062687e-06,
"loss": 0.0005,
"step": 707
},
{
"epoch": 0.4045714285714286,
"grad_norm": 0.047931671142578125,
"learning_rate": 4.78990368159424e-06,
"loss": 0.0013,
"step": 708
},
{
"epoch": 0.40514285714285714,
"grad_norm": 0.09240229427814484,
"learning_rate": 4.76014094029921e-06,
"loss": 0.0008,
"step": 709
},
{
"epoch": 0.4057142857142857,
"grad_norm": 0.19956842064857483,
"learning_rate": 4.7304420450332244e-06,
"loss": 0.024,
"step": 710
},
{
"epoch": 0.4062857142857143,
"grad_norm": 0.47484588623046875,
"learning_rate": 4.700807357667953e-06,
"loss": 0.0358,
"step": 711
},
{
"epoch": 0.40685714285714286,
"grad_norm": 3.1876862049102783,
"learning_rate": 4.671237239292699e-06,
"loss": 0.0243,
"step": 712
},
{
"epoch": 0.4074285714285714,
"grad_norm": 0.16582611203193665,
"learning_rate": 4.641732050210032e-06,
"loss": 0.0008,
"step": 713
},
{
"epoch": 0.408,
"grad_norm": 0.25967761874198914,
"learning_rate": 4.612292149931369e-06,
"loss": 0.0267,
"step": 714
},
{
"epoch": 0.4085714285714286,
"grad_norm": 0.06273027509450912,
"learning_rate": 4.582917897172603e-06,
"loss": 0.0011,
"step": 715
},
{
"epoch": 0.40914285714285714,
"grad_norm": 0.031824495643377304,
"learning_rate": 4.5536096498497295e-06,
"loss": 0.0006,
"step": 716
},
{
"epoch": 0.4097142857142857,
"grad_norm": 0.020821528509259224,
"learning_rate": 4.524367765074499e-06,
"loss": 0.0008,
"step": 717
},
{
"epoch": 0.4102857142857143,
"grad_norm": 0.28375720977783203,
"learning_rate": 4.495192599150045e-06,
"loss": 0.0408,
"step": 718
},
{
"epoch": 0.41085714285714287,
"grad_norm": 0.3123238682746887,
"learning_rate": 4.46608450756656e-06,
"loss": 0.0275,
"step": 719
},
{
"epoch": 0.4114285714285714,
"grad_norm": 0.18190975487232208,
"learning_rate": 4.437043844996952e-06,
"loss": 0.0241,
"step": 720
},
{
"epoch": 0.412,
"grad_norm": 0.465442031621933,
"learning_rate": 4.408070965292534e-06,
"loss": 0.0354,
"step": 721
},
{
"epoch": 0.4125714285714286,
"grad_norm": 0.48212334513664246,
"learning_rate": 4.379166221478697e-06,
"loss": 0.0347,
"step": 722
},
{
"epoch": 0.41314285714285715,
"grad_norm": 0.29567527770996094,
"learning_rate": 4.350329965750622e-06,
"loss": 0.0472,
"step": 723
},
{
"epoch": 0.4137142857142857,
"grad_norm": 0.21704529225826263,
"learning_rate": 4.321562549468991e-06,
"loss": 0.0326,
"step": 724
},
{
"epoch": 0.4142857142857143,
"grad_norm": 0.02085958421230316,
"learning_rate": 4.292864323155684e-06,
"loss": 0.0003,
"step": 725
},
{
"epoch": 0.41485714285714287,
"grad_norm": 0.2143467664718628,
"learning_rate": 4.264235636489542e-06,
"loss": 0.0369,
"step": 726
},
{
"epoch": 0.4154285714285714,
"grad_norm": 0.0682457908987999,
"learning_rate": 4.235676838302069e-06,
"loss": 0.0014,
"step": 727
},
{
"epoch": 0.416,
"grad_norm": 0.025718998163938522,
"learning_rate": 4.207188276573214e-06,
"loss": 0.0006,
"step": 728
},
{
"epoch": 0.4165714285714286,
"grad_norm": 0.47329089045524597,
"learning_rate": 4.178770298427107e-06,
"loss": 0.0521,
"step": 729
},
{
"epoch": 0.41714285714285715,
"grad_norm": 0.12551596760749817,
"learning_rate": 4.150423250127846e-06,
"loss": 0.0319,
"step": 730
},
{
"epoch": 0.4177142857142857,
"grad_norm": 0.22539138793945312,
"learning_rate": 4.12214747707527e-06,
"loss": 0.033,
"step": 731
},
{
"epoch": 0.41828571428571426,
"grad_norm": 0.4434771239757538,
"learning_rate": 4.093943323800746e-06,
"loss": 0.0339,
"step": 732
},
{
"epoch": 0.4188571428571429,
"grad_norm": 0.03221685439348221,
"learning_rate": 4.065811133962987e-06,
"loss": 0.0002,
"step": 733
},
{
"epoch": 0.41942857142857143,
"grad_norm": 0.29555070400238037,
"learning_rate": 4.037751250343841e-06,
"loss": 0.0473,
"step": 734
},
{
"epoch": 0.42,
"grad_norm": 0.39999938011169434,
"learning_rate": 4.009764014844143e-06,
"loss": 0.0557,
"step": 735
},
{
"epoch": 0.4205714285714286,
"grad_norm": 0.2046584188938141,
"learning_rate": 3.981849768479516e-06,
"loss": 0.0317,
"step": 736
},
{
"epoch": 0.42114285714285715,
"grad_norm": 0.04865134879946709,
"learning_rate": 3.954008851376252e-06,
"loss": 0.0013,
"step": 737
},
{
"epoch": 0.4217142857142857,
"grad_norm": 0.4475577473640442,
"learning_rate": 3.9262416027671354e-06,
"loss": 0.034,
"step": 738
},
{
"epoch": 0.42228571428571426,
"grad_norm": 0.4417513310909271,
"learning_rate": 3.898548360987325e-06,
"loss": 0.0327,
"step": 739
},
{
"epoch": 0.4228571428571429,
"grad_norm": 0.3132314085960388,
"learning_rate": 3.8709294634702374e-06,
"loss": 0.029,
"step": 740
},
{
"epoch": 0.42342857142857143,
"grad_norm": 0.12235007435083389,
"learning_rate": 3.8433852467434175e-06,
"loss": 0.0208,
"step": 741
},
{
"epoch": 0.424,
"grad_norm": 0.14897814393043518,
"learning_rate": 3.81591604642446e-06,
"loss": 0.001,
"step": 742
},
{
"epoch": 0.42457142857142854,
"grad_norm": 0.12743142247200012,
"learning_rate": 3.7885221972168974e-06,
"loss": 0.0364,
"step": 743
},
{
"epoch": 0.42514285714285716,
"grad_norm": 0.2557103633880615,
"learning_rate": 3.7612040329061405e-06,
"loss": 0.028,
"step": 744
},
{
"epoch": 0.4257142857142857,
"grad_norm": 0.3860604166984558,
"learning_rate": 3.7339618863553983e-06,
"loss": 0.0516,
"step": 745
},
{
"epoch": 0.42628571428571427,
"grad_norm": 0.21680651605129242,
"learning_rate": 3.7067960895016277e-06,
"loss": 0.0306,
"step": 746
},
{
"epoch": 0.4268571428571429,
"grad_norm": 0.046007659286260605,
"learning_rate": 3.679706973351491e-06,
"loss": 0.001,
"step": 747
},
{
"epoch": 0.42742857142857144,
"grad_norm": 0.1162264421582222,
"learning_rate": 3.6526948679773256e-06,
"loss": 0.0366,
"step": 748
},
{
"epoch": 0.428,
"grad_norm": 0.4734565317630768,
"learning_rate": 3.625760102513103e-06,
"loss": 0.029,
"step": 749
},
{
"epoch": 0.42857142857142855,
"grad_norm": 0.02459256909787655,
"learning_rate": 3.598903005150444e-06,
"loss": 0.0001,
"step": 750
},
{
"epoch": 0.42914285714285716,
"grad_norm": 0.08670744299888611,
"learning_rate": 3.5721239031346067e-06,
"loss": 0.0015,
"step": 751
},
{
"epoch": 0.4297142857142857,
"grad_norm": 0.05315675586462021,
"learning_rate": 3.545423122760493e-06,
"loss": 0.0004,
"step": 752
},
{
"epoch": 0.43028571428571427,
"grad_norm": 0.04562332481145859,
"learning_rate": 3.5188009893686916e-06,
"loss": 0.0007,
"step": 753
},
{
"epoch": 0.4308571428571429,
"grad_norm": 0.2997724115848541,
"learning_rate": 3.492257827341492e-06,
"loss": 0.0444,
"step": 754
},
{
"epoch": 0.43142857142857144,
"grad_norm": 0.3253784477710724,
"learning_rate": 3.4657939600989453e-06,
"loss": 0.0527,
"step": 755
},
{
"epoch": 0.432,
"grad_norm": 1.337095022201538,
"learning_rate": 3.4394097100949286e-06,
"loss": 0.0527,
"step": 756
},
{
"epoch": 0.43257142857142855,
"grad_norm": 0.11799000203609467,
"learning_rate": 3.4131053988131947e-06,
"loss": 0.0223,
"step": 757
},
{
"epoch": 0.43314285714285716,
"grad_norm": 0.044688720256090164,
"learning_rate": 3.3868813467634833e-06,
"loss": 0.0002,
"step": 758
},
{
"epoch": 0.4337142857142857,
"grad_norm": 0.33123695850372314,
"learning_rate": 3.360737873477584e-06,
"loss": 0.0482,
"step": 759
},
{
"epoch": 0.4342857142857143,
"grad_norm": 0.24822980165481567,
"learning_rate": 3.3346752975054763e-06,
"loss": 0.0243,
"step": 760
},
{
"epoch": 0.43485714285714283,
"grad_norm": 0.6182214021682739,
"learning_rate": 3.308693936411421e-06,
"loss": 0.0298,
"step": 761
},
{
"epoch": 0.43542857142857144,
"grad_norm": 0.3130514621734619,
"learning_rate": 3.2827941067700996e-06,
"loss": 0.0442,
"step": 762
},
{
"epoch": 0.436,
"grad_norm": 0.254433274269104,
"learning_rate": 3.2569761241627694e-06,
"loss": 0.0247,
"step": 763
},
{
"epoch": 0.43657142857142855,
"grad_norm": 0.02353322133421898,
"learning_rate": 3.2312403031733943e-06,
"loss": 0.0005,
"step": 764
},
{
"epoch": 0.43714285714285717,
"grad_norm": 0.08542972058057785,
"learning_rate": 3.2055869573848374e-06,
"loss": 0.0011,
"step": 765
},
{
"epoch": 0.4377142857142857,
"grad_norm": 0.2962503135204315,
"learning_rate": 3.1800163993750166e-06,
"loss": 0.0472,
"step": 766
},
{
"epoch": 0.4382857142857143,
"grad_norm": 0.3337412178516388,
"learning_rate": 3.1545289407131128e-06,
"loss": 0.0442,
"step": 767
},
{
"epoch": 0.43885714285714283,
"grad_norm": 0.31081369519233704,
"learning_rate": 3.1291248919557717e-06,
"loss": 0.0514,
"step": 768
},
{
"epoch": 0.43942857142857145,
"grad_norm": 0.020432591438293457,
"learning_rate": 3.103804562643302e-06,
"loss": 0.0005,
"step": 769
},
{
"epoch": 0.44,
"grad_norm": 0.31683558225631714,
"learning_rate": 3.0785682612959334e-06,
"loss": 0.0474,
"step": 770
},
{
"epoch": 0.44057142857142856,
"grad_norm": 0.21815727651119232,
"learning_rate": 3.0534162954100264e-06,
"loss": 0.0415,
"step": 771
},
{
"epoch": 0.44114285714285717,
"grad_norm": 0.3833843171596527,
"learning_rate": 3.028348971454356e-06,
"loss": 0.0465,
"step": 772
},
{
"epoch": 0.4417142857142857,
"grad_norm": 0.3005063831806183,
"learning_rate": 3.003366594866345e-06,
"loss": 0.0437,
"step": 773
},
{
"epoch": 0.4422857142857143,
"grad_norm": 0.10851803421974182,
"learning_rate": 2.978469470048376e-06,
"loss": 0.0204,
"step": 774
},
{
"epoch": 0.44285714285714284,
"grad_norm": 0.3799663484096527,
"learning_rate": 2.953657900364053e-06,
"loss": 0.0481,
"step": 775
},
{
"epoch": 0.44342857142857145,
"grad_norm": 0.19519735872745514,
"learning_rate": 2.9289321881345257e-06,
"loss": 0.0083,
"step": 776
},
{
"epoch": 0.444,
"grad_norm": 0.03894161805510521,
"learning_rate": 2.9042926346347932e-06,
"loss": 0.0009,
"step": 777
},
{
"epoch": 0.44457142857142856,
"grad_norm": 0.03295385092496872,
"learning_rate": 2.8797395400900362e-06,
"loss": 0.0005,
"step": 778
},
{
"epoch": 0.4451428571428571,
"grad_norm": 0.43225622177124023,
"learning_rate": 2.855273203671969e-06,
"loss": 0.0322,
"step": 779
},
{
"epoch": 0.44571428571428573,
"grad_norm": 0.1589963734149933,
"learning_rate": 2.830893923495173e-06,
"loss": 0.0376,
"step": 780
},
{
"epoch": 0.4462857142857143,
"grad_norm": 18.78740882873535,
"learning_rate": 2.8066019966134907e-06,
"loss": 0.0492,
"step": 781
},
{
"epoch": 0.44685714285714284,
"grad_norm": 0.2477787584066391,
"learning_rate": 2.7823977190163788e-06,
"loss": 0.0418,
"step": 782
},
{
"epoch": 0.44742857142857145,
"grad_norm": 0.07063715159893036,
"learning_rate": 2.7582813856253276e-06,
"loss": 0.0008,
"step": 783
},
{
"epoch": 0.448,
"grad_norm": 0.022874176502227783,
"learning_rate": 2.7342532902902418e-06,
"loss": 0.0008,
"step": 784
},
{
"epoch": 0.44857142857142857,
"grad_norm": 0.02444029599428177,
"learning_rate": 2.7103137257858867e-06,
"loss": 0.0009,
"step": 785
},
{
"epoch": 0.4491428571428571,
"grad_norm": 0.47617411613464355,
"learning_rate": 2.6864629838082957e-06,
"loss": 0.0534,
"step": 786
},
{
"epoch": 0.44971428571428573,
"grad_norm": 0.3131197690963745,
"learning_rate": 2.6627013549712355e-06,
"loss": 0.047,
"step": 787
},
{
"epoch": 0.4502857142857143,
"grad_norm": 0.3279697597026825,
"learning_rate": 2.639029128802657e-06,
"loss": 0.0476,
"step": 788
},
{
"epoch": 0.45085714285714285,
"grad_norm": 0.25661271810531616,
"learning_rate": 2.615446593741161e-06,
"loss": 0.0364,
"step": 789
},
{
"epoch": 0.4514285714285714,
"grad_norm": 0.24881282448768616,
"learning_rate": 2.5919540371325005e-06,
"loss": 0.0431,
"step": 790
},
{
"epoch": 0.452,
"grad_norm": 0.02317560464143753,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0005,
"step": 791
},
{
"epoch": 0.45257142857142857,
"grad_norm": 1.093137264251709,
"learning_rate": 2.5452400031713786e-06,
"loss": 0.0657,
"step": 792
},
{
"epoch": 0.4531428571428571,
"grad_norm": 0.3608619272708893,
"learning_rate": 2.522019095014683e-06,
"loss": 0.0357,
"step": 793
},
{
"epoch": 0.45371428571428574,
"grad_norm": 0.016164017841219902,
"learning_rate": 2.4988893036954045e-06,
"loss": 0.0007,
"step": 794
},
{
"epoch": 0.4542857142857143,
"grad_norm": 0.02934977225959301,
"learning_rate": 2.4758509110427576e-06,
"loss": 0.0006,
"step": 795
},
{
"epoch": 0.45485714285714285,
"grad_norm": 0.04249361902475357,
"learning_rate": 2.45290419777228e-06,
"loss": 0.0006,
"step": 796
},
{
"epoch": 0.4554285714285714,
"grad_norm": 0.4457858204841614,
"learning_rate": 2.4300494434824373e-06,
"loss": 0.0338,
"step": 797
},
{
"epoch": 0.456,
"grad_norm": 0.28004032373428345,
"learning_rate": 2.407286926651192e-06,
"loss": 0.0391,
"step": 798
},
{
"epoch": 0.4565714285714286,
"grad_norm": 0.960296630859375,
"learning_rate": 2.3846169246326345e-06,
"loss": 0.0327,
"step": 799
},
{
"epoch": 0.45714285714285713,
"grad_norm": 0.3210156559944153,
"learning_rate": 2.362039713653581e-06,
"loss": 0.0279,
"step": 800
},
{
"epoch": 0.45771428571428574,
"grad_norm": 0.1571497917175293,
"learning_rate": 2.339555568810221e-06,
"loss": 0.0338,
"step": 801
},
{
"epoch": 0.4582857142857143,
"grad_norm": 0.048868756741285324,
"learning_rate": 2.317164764064769e-06,
"loss": 0.0011,
"step": 802
},
{
"epoch": 0.45885714285714285,
"grad_norm": 0.04672923684120178,
"learning_rate": 2.2948675722421086e-06,
"loss": 0.0009,
"step": 803
},
{
"epoch": 0.4594285714285714,
"grad_norm": 0.03390154987573624,
"learning_rate": 2.27266426502649e-06,
"loss": 0.001,
"step": 804
},
{
"epoch": 0.46,
"grad_norm": 0.18708500266075134,
"learning_rate": 2.2505551129582047e-06,
"loss": 0.024,
"step": 805
},
{
"epoch": 0.4605714285714286,
"grad_norm": 0.03318360820412636,
"learning_rate": 2.2285403854302912e-06,
"loss": 0.0004,
"step": 806
},
{
"epoch": 0.46114285714285713,
"grad_norm": 0.027982419356703758,
"learning_rate": 2.206620350685257e-06,
"loss": 0.0004,
"step": 807
},
{
"epoch": 0.4617142857142857,
"grad_norm": 0.13334748148918152,
"learning_rate": 2.1847952758118118e-06,
"loss": 0.0438,
"step": 808
},
{
"epoch": 0.4622857142857143,
"grad_norm": 0.20942264795303345,
"learning_rate": 2.163065426741603e-06,
"loss": 0.0352,
"step": 809
},
{
"epoch": 0.46285714285714286,
"grad_norm": 0.12256276607513428,
"learning_rate": 2.1414310682459805e-06,
"loss": 0.0015,
"step": 810
},
{
"epoch": 0.4634285714285714,
"grad_norm": 0.29164934158325195,
"learning_rate": 2.119892463932781e-06,
"loss": 0.0428,
"step": 811
},
{
"epoch": 0.464,
"grad_norm": 0.3350270688533783,
"learning_rate": 2.098449876243096e-06,
"loss": 0.029,
"step": 812
},
{
"epoch": 0.4645714285714286,
"grad_norm": 0.402536541223526,
"learning_rate": 2.0771035664480944e-06,
"loss": 0.0307,
"step": 813
},
{
"epoch": 0.46514285714285714,
"grad_norm": 0.3149171471595764,
"learning_rate": 2.0558537946458177e-06,
"loss": 0.0456,
"step": 814
},
{
"epoch": 0.4657142857142857,
"grad_norm": 0.0804779976606369,
"learning_rate": 2.0347008197580376e-06,
"loss": 0.0011,
"step": 815
},
{
"epoch": 0.4662857142857143,
"grad_norm": 0.14471156895160675,
"learning_rate": 2.013644899527074e-06,
"loss": 0.0266,
"step": 816
},
{
"epoch": 0.46685714285714286,
"grad_norm": 0.4182524085044861,
"learning_rate": 1.9926862905126663e-06,
"loss": 0.0323,
"step": 817
},
{
"epoch": 0.4674285714285714,
"grad_norm": 0.26969775557518005,
"learning_rate": 1.9718252480888567e-06,
"loss": 0.028,
"step": 818
},
{
"epoch": 0.468,
"grad_norm": 0.601039707660675,
"learning_rate": 1.95106202644086e-06,
"loss": 0.0457,
"step": 819
},
{
"epoch": 0.4685714285714286,
"grad_norm": 0.022038260474801064,
"learning_rate": 1.930396878561983e-06,
"loss": 0.0005,
"step": 820
},
{
"epoch": 0.46914285714285714,
"grad_norm": 0.13229797780513763,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.0218,
"step": 821
},
{
"epoch": 0.4697142857142857,
"grad_norm": 0.06522911041975021,
"learning_rate": 1.8893618101067357e-06,
"loss": 0.0011,
"step": 822
},
{
"epoch": 0.4702857142857143,
"grad_norm": 0.3180411458015442,
"learning_rate": 1.8689923895297247e-06,
"loss": 0.0266,
"step": 823
},
{
"epoch": 0.47085714285714286,
"grad_norm": 0.19166551530361176,
"learning_rate": 1.848722042714457e-06,
"loss": 0.0234,
"step": 824
},
{
"epoch": 0.4714285714285714,
"grad_norm": 0.39556875824928284,
"learning_rate": 1.8285510166487154e-06,
"loss": 0.0298,
"step": 825
},
{
"epoch": 0.472,
"grad_norm": 0.1736728996038437,
"learning_rate": 1.808479557110081e-06,
"loss": 0.0403,
"step": 826
},
{
"epoch": 0.4725714285714286,
"grad_norm": 0.4187528192996979,
"learning_rate": 1.7885079086629598e-06,
"loss": 0.0307,
"step": 827
},
{
"epoch": 0.47314285714285714,
"grad_norm": 0.41673463582992554,
"learning_rate": 1.7686363146555807e-06,
"loss": 0.0314,
"step": 828
},
{
"epoch": 0.4737142857142857,
"grad_norm": 0.15574224293231964,
"learning_rate": 1.7488650172170496e-06,
"loss": 0.0237,
"step": 829
},
{
"epoch": 0.4742857142857143,
"grad_norm": 0.4227770268917084,
"learning_rate": 1.7291942572543806e-06,
"loss": 0.0311,
"step": 830
},
{
"epoch": 0.47485714285714287,
"grad_norm": 0.389302521944046,
"learning_rate": 1.709624274449584e-06,
"loss": 0.0521,
"step": 831
},
{
"epoch": 0.4754285714285714,
"grad_norm": 0.017934251576662064,
"learning_rate": 1.6901553072567189e-06,
"loss": 0.0005,
"step": 832
},
{
"epoch": 0.476,
"grad_norm": 0.5057289004325867,
"learning_rate": 1.6707875928990059e-06,
"loss": 0.0224,
"step": 833
},
{
"epoch": 0.4765714285714286,
"grad_norm": 0.02792900614440441,
"learning_rate": 1.651521367365936e-06,
"loss": 0.0006,
"step": 834
},
{
"epoch": 0.47714285714285715,
"grad_norm": 0.3854008913040161,
"learning_rate": 1.6323568654103838e-06,
"loss": 0.0293,
"step": 835
},
{
"epoch": 0.4777142857142857,
"grad_norm": 1.5571532249450684,
"learning_rate": 1.6132943205457607e-06,
"loss": 0.0991,
"step": 836
},
{
"epoch": 0.47828571428571426,
"grad_norm": 0.4142951965332031,
"learning_rate": 1.5943339650431578e-06,
"loss": 0.055,
"step": 837
},
{
"epoch": 0.47885714285714287,
"grad_norm": 0.24086351692676544,
"learning_rate": 1.5754760299285255e-06,
"loss": 0.0468,
"step": 838
},
{
"epoch": 0.4794285714285714,
"grad_norm": 0.017584173008799553,
"learning_rate": 1.5567207449798517e-06,
"loss": 0.0009,
"step": 839
},
{
"epoch": 0.48,
"grad_norm": 0.16903410851955414,
"learning_rate": 1.538068338724361e-06,
"loss": 0.0376,
"step": 840
},
{
"epoch": 0.4805714285714286,
"grad_norm": 0.05065883323550224,
"learning_rate": 1.5195190384357405e-06,
"loss": 0.0011,
"step": 841
},
{
"epoch": 0.48114285714285715,
"grad_norm": 0.28271520137786865,
"learning_rate": 1.5010730701313626e-06,
"loss": 0.0256,
"step": 842
},
{
"epoch": 0.4817142857142857,
"grad_norm": 0.2893432676792145,
"learning_rate": 1.4827306585695234e-06,
"loss": 0.0504,
"step": 843
},
{
"epoch": 0.48228571428571426,
"grad_norm": 0.032072458416223526,
"learning_rate": 1.4644920272467245e-06,
"loss": 0.0007,
"step": 844
},
{
"epoch": 0.4828571428571429,
"grad_norm": 0.3085223436355591,
"learning_rate": 1.446357398394934e-06,
"loss": 0.0467,
"step": 845
},
{
"epoch": 0.48342857142857143,
"grad_norm": 0.8155536651611328,
"learning_rate": 1.4283269929788779e-06,
"loss": 0.0264,
"step": 846
},
{
"epoch": 0.484,
"grad_norm": 0.3773426413536072,
"learning_rate": 1.4104010306933558e-06,
"loss": 0.0292,
"step": 847
},
{
"epoch": 0.4845714285714286,
"grad_norm": 8.531339645385742,
"learning_rate": 1.3925797299605649e-06,
"loss": 0.046,
"step": 848
},
{
"epoch": 0.48514285714285715,
"grad_norm": 0.3405299782752991,
"learning_rate": 1.3748633079274254e-06,
"loss": 0.0452,
"step": 849
},
{
"epoch": 0.4857142857142857,
"grad_norm": 0.31444427371025085,
"learning_rate": 1.3572519804629537e-06,
"loss": 0.0444,
"step": 850
},
{
"epoch": 0.48628571428571427,
"grad_norm": 0.13080042600631714,
"learning_rate": 1.339745962155613e-06,
"loss": 0.021,
"step": 851
},
{
"epoch": 0.4868571428571429,
"grad_norm": 0.3826882243156433,
"learning_rate": 1.322345466310717e-06,
"loss": 0.0291,
"step": 852
},
{
"epoch": 0.48742857142857143,
"grad_norm": 0.12485809624195099,
"learning_rate": 1.30505070494781e-06,
"loss": 0.0211,
"step": 853
},
{
"epoch": 0.488,
"grad_norm": 0.12976916134357452,
"learning_rate": 1.2878618887981064e-06,
"loss": 0.0441,
"step": 854
},
{
"epoch": 0.48857142857142855,
"grad_norm": 0.49974438548088074,
"learning_rate": 1.2707792273019049e-06,
"loss": 0.0044,
"step": 855
},
{
"epoch": 0.48914285714285716,
"grad_norm": 0.03166588023304939,
"learning_rate": 1.2538029286060428e-06,
"loss": 0.0009,
"step": 856
},
{
"epoch": 0.4897142857142857,
"grad_norm": 0.32709768414497375,
"learning_rate": 1.2369331995613664e-06,
"loss": 0.0447,
"step": 857
},
{
"epoch": 0.49028571428571427,
"grad_norm": 0.03613048791885376,
"learning_rate": 1.2201702457201948e-06,
"loss": 0.0005,
"step": 858
},
{
"epoch": 0.4908571428571429,
"grad_norm": 0.03202739730477333,
"learning_rate": 1.2035142713338366e-06,
"loss": 0.0006,
"step": 859
},
{
"epoch": 0.49142857142857144,
"grad_norm": 0.03778680041432381,
"learning_rate": 1.1869654793500784e-06,
"loss": 0.0006,
"step": 860
},
{
"epoch": 0.492,
"grad_norm": 0.4695800542831421,
"learning_rate": 1.1705240714107301e-06,
"loss": 0.0209,
"step": 861
},
{
"epoch": 0.49257142857142855,
"grad_norm": 0.11646213382482529,
"learning_rate": 1.1541902478491607e-06,
"loss": 0.001,
"step": 862
},
{
"epoch": 0.49314285714285716,
"grad_norm": 0.24151349067687988,
"learning_rate": 1.1379642076878528e-06,
"loss": 0.04,
"step": 863
},
{
"epoch": 0.4937142857142857,
"grad_norm": 0.2839173674583435,
"learning_rate": 1.1218461486359878e-06,
"loss": 0.0361,
"step": 864
},
{
"epoch": 0.4942857142857143,
"grad_norm": 0.027763906866312027,
"learning_rate": 1.1058362670870248e-06,
"loss": 0.0008,
"step": 865
},
{
"epoch": 0.4948571428571429,
"grad_norm": 0.2715831696987152,
"learning_rate": 1.0899347581163222e-06,
"loss": 0.0455,
"step": 866
},
{
"epoch": 0.49542857142857144,
"grad_norm": 0.02777138352394104,
"learning_rate": 1.0741418154787443e-06,
"loss": 0.0002,
"step": 867
},
{
"epoch": 0.496,
"grad_norm": 10.18946647644043,
"learning_rate": 1.058457631606319e-06,
"loss": 0.0656,
"step": 868
},
{
"epoch": 0.49657142857142855,
"grad_norm": 0.019488289952278137,
"learning_rate": 1.042882397605871e-06,
"loss": 0.0003,
"step": 869
},
{
"epoch": 0.49714285714285716,
"grad_norm": 0.14182956516742706,
"learning_rate": 1.0274163032567165e-06,
"loss": 0.0377,
"step": 870
},
{
"epoch": 0.4977142857142857,
"grad_norm": 0.1722799837589264,
"learning_rate": 1.012059537008332e-06,
"loss": 0.0023,
"step": 871
},
{
"epoch": 0.4982857142857143,
"grad_norm": 0.2898975908756256,
"learning_rate": 9.968122859780648e-07,
"loss": 0.0261,
"step": 872
},
{
"epoch": 0.49885714285714283,
"grad_norm": 0.017987089231610298,
"learning_rate": 9.816747359488632e-07,
"loss": 0.001,
"step": 873
},
{
"epoch": 0.49942857142857144,
"grad_norm": 0.45560625195503235,
"learning_rate": 9.666470713669918e-07,
"loss": 0.0558,
"step": 874
},
{
"epoch": 0.5,
"grad_norm": 0.02222052402794361,
"learning_rate": 9.517294753398066e-07,
"loss": 0.0002,
"step": 875
},
{
"epoch": 0.5005714285714286,
"grad_norm": 1.05287766456604,
"learning_rate": 9.369221296335007e-07,
"loss": 0.0495,
"step": 876
},
{
"epoch": 0.5011428571428571,
"grad_norm": 0.36959758400917053,
"learning_rate": 9.222252146709143e-07,
"loss": 0.0286,
"step": 877
},
{
"epoch": 0.5017142857142857,
"grad_norm": 0.3018774390220642,
"learning_rate": 9.076389095293148e-07,
"loss": 0.0509,
"step": 878
},
{
"epoch": 0.5022857142857143,
"grad_norm": 0.34820592403411865,
"learning_rate": 8.931633919382299e-07,
"loss": 0.0465,
"step": 879
},
{
"epoch": 0.5028571428571429,
"grad_norm": 0.036583635956048965,
"learning_rate": 8.787988382772705e-07,
"loss": 0.0009,
"step": 880
},
{
"epoch": 0.5034285714285714,
"grad_norm": 0.3857507109642029,
"learning_rate": 8.645454235739903e-07,
"loss": 0.0304,
"step": 881
},
{
"epoch": 0.504,
"grad_norm": 0.3286173641681671,
"learning_rate": 8.504033215017527e-07,
"loss": 0.0442,
"step": 882
},
{
"epoch": 0.5045714285714286,
"grad_norm": 0.166743203997612,
"learning_rate": 8.363727043776037e-07,
"loss": 0.0081,
"step": 883
},
{
"epoch": 0.5051428571428571,
"grad_norm": 0.6375927329063416,
"learning_rate": 8.224537431601886e-07,
"loss": 0.0637,
"step": 884
},
{
"epoch": 0.5057142857142857,
"grad_norm": 0.2817516624927521,
"learning_rate": 8.086466074476562e-07,
"loss": 0.0271,
"step": 885
},
{
"epoch": 0.5062857142857143,
"grad_norm": 0.17895446717739105,
"learning_rate": 7.949514654755963e-07,
"loss": 0.0162,
"step": 886
},
{
"epoch": 0.5068571428571429,
"grad_norm": 0.22048166394233704,
"learning_rate": 7.81368484114996e-07,
"loss": 0.0214,
"step": 887
},
{
"epoch": 0.5074285714285715,
"grad_norm": 0.29659560322761536,
"learning_rate": 7.678978288701911e-07,
"loss": 0.0265,
"step": 888
},
{
"epoch": 0.508,
"grad_norm": 0.045432429760694504,
"learning_rate": 7.545396638768698e-07,
"loss": 0.0008,
"step": 889
},
{
"epoch": 0.5085714285714286,
"grad_norm": 0.17099474370479584,
"learning_rate": 7.412941519000527e-07,
"loss": 0.0082,
"step": 890
},
{
"epoch": 0.5091428571428571,
"grad_norm": 0.0504331961274147,
"learning_rate": 7.281614543321269e-07,
"loss": 0.0011,
"step": 891
},
{
"epoch": 0.5097142857142857,
"grad_norm": 0.7796383500099182,
"learning_rate": 7.151417311908648e-07,
"loss": 0.0558,
"step": 892
},
{
"epoch": 0.5102857142857142,
"grad_norm": 0.10371389240026474,
"learning_rate": 7.022351411174866e-07,
"loss": 0.0394,
"step": 893
},
{
"epoch": 0.5108571428571429,
"grad_norm": 0.3338377773761749,
"learning_rate": 6.894418413747183e-07,
"loss": 0.0456,
"step": 894
},
{
"epoch": 0.5114285714285715,
"grad_norm": 0.027731921523809433,
"learning_rate": 6.767619878448783e-07,
"loss": 0.0011,
"step": 895
},
{
"epoch": 0.512,
"grad_norm": 0.14995941519737244,
"learning_rate": 6.641957350279838e-07,
"loss": 0.0331,
"step": 896
},
{
"epoch": 0.5125714285714286,
"grad_norm": 1.0035312175750732,
"learning_rate": 6.517432360398556e-07,
"loss": 0.0342,
"step": 897
},
{
"epoch": 0.5131428571428571,
"grad_norm": 0.256251722574234,
"learning_rate": 6.394046426102673e-07,
"loss": 0.0253,
"step": 898
},
{
"epoch": 0.5137142857142857,
"grad_norm": 0.025908537209033966,
"learning_rate": 6.271801050810856e-07,
"loss": 0.0005,
"step": 899
},
{
"epoch": 0.5142857142857142,
"grad_norm": 0.022469278424978256,
"learning_rate": 6.150697724044407e-07,
"loss": 0.0003,
"step": 900
},
{
"epoch": 0.5148571428571429,
"grad_norm": 0.3587414622306824,
"learning_rate": 6.030737921409169e-07,
"loss": 0.0363,
"step": 901
},
{
"epoch": 0.5154285714285715,
"grad_norm": 0.3276163637638092,
"learning_rate": 5.911923104577455e-07,
"loss": 0.0527,
"step": 902
},
{
"epoch": 0.516,
"grad_norm": 0.26531481742858887,
"learning_rate": 5.794254721270331e-07,
"loss": 0.0301,
"step": 903
},
{
"epoch": 0.5165714285714286,
"grad_norm": 0.17951834201812744,
"learning_rate": 5.677734205239904e-07,
"loss": 0.0164,
"step": 904
},
{
"epoch": 0.5171428571428571,
"grad_norm": 0.34895995259284973,
"learning_rate": 5.562362976251901e-07,
"loss": 0.0404,
"step": 905
},
{
"epoch": 0.5177142857142857,
"grad_norm": 0.039577994495630264,
"learning_rate": 5.448142440068316e-07,
"loss": 0.0004,
"step": 906
},
{
"epoch": 0.5182857142857142,
"grad_norm": 0.27763235569000244,
"learning_rate": 5.335073988430373e-07,
"loss": 0.0477,
"step": 907
},
{
"epoch": 0.5188571428571429,
"grad_norm": 0.35665449500083923,
"learning_rate": 5.223158999041444e-07,
"loss": 0.0277,
"step": 908
},
{
"epoch": 0.5194285714285715,
"grad_norm": 0.41862088441848755,
"learning_rate": 5.112398835550348e-07,
"loss": 0.0313,
"step": 909
},
{
"epoch": 0.52,
"grad_norm": 0.7225152254104614,
"learning_rate": 5.002794847534765e-07,
"loss": 0.0591,
"step": 910
},
{
"epoch": 0.5205714285714286,
"grad_norm": 0.31689453125,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0443,
"step": 911
},
{
"epoch": 0.5211428571428571,
"grad_norm": 0.04286804422736168,
"learning_rate": 4.787060725786141e-07,
"loss": 0.0009,
"step": 912
},
{
"epoch": 0.5217142857142857,
"grad_norm": 0.07602091878652573,
"learning_rate": 4.6809332207053083e-07,
"loss": 0.0015,
"step": 913
},
{
"epoch": 0.5222857142857142,
"grad_norm": 0.38041192293167114,
"learning_rate": 4.575967148372318e-07,
"loss": 0.0293,
"step": 914
},
{
"epoch": 0.5228571428571429,
"grad_norm": 0.03401736542582512,
"learning_rate": 4.4721637877656377e-07,
"loss": 0.0006,
"step": 915
},
{
"epoch": 0.5234285714285715,
"grad_norm": 0.3198559284210205,
"learning_rate": 4.3695244036964567e-07,
"loss": 0.0444,
"step": 916
},
{
"epoch": 0.524,
"grad_norm": 0.03157348185777664,
"learning_rate": 4.268050246793276e-07,
"loss": 0.001,
"step": 917
},
{
"epoch": 0.5245714285714286,
"grad_norm": 0.034865133464336395,
"learning_rate": 4.167742553486676e-07,
"loss": 0.0007,
"step": 918
},
{
"epoch": 0.5251428571428571,
"grad_norm": 0.2708113491535187,
"learning_rate": 4.068602545994249e-07,
"loss": 0.0257,
"step": 919
},
{
"epoch": 0.5257142857142857,
"grad_norm": 0.0616893470287323,
"learning_rate": 3.9706314323056936e-07,
"loss": 0.0014,
"step": 920
},
{
"epoch": 0.5262857142857142,
"grad_norm": 0.1888759881258011,
"learning_rate": 3.8738304061681107e-07,
"loss": 0.0403,
"step": 921
},
{
"epoch": 0.5268571428571428,
"grad_norm": 0.01952865906059742,
"learning_rate": 3.7782006470714614e-07,
"loss": 0.0008,
"step": 922
},
{
"epoch": 0.5274285714285715,
"grad_norm": 0.4976491332054138,
"learning_rate": 3.68374332023419e-07,
"loss": 0.0531,
"step": 923
},
{
"epoch": 0.528,
"grad_norm": 0.06850526481866837,
"learning_rate": 3.590459576589e-07,
"loss": 0.0017,
"step": 924
},
{
"epoch": 0.5285714285714286,
"grad_norm": 0.14225688576698303,
"learning_rate": 3.498350552768859e-07,
"loss": 0.0346,
"step": 925
},
{
"epoch": 0.5291428571428571,
"grad_norm": 0.10856902599334717,
"learning_rate": 3.4074173710931804e-07,
"loss": 0.0012,
"step": 926
},
{
"epoch": 0.5297142857142857,
"grad_norm": 0.43952202796936035,
"learning_rate": 3.3176611395540625e-07,
"loss": 0.0551,
"step": 927
},
{
"epoch": 0.5302857142857142,
"grad_norm": 0.03717416152358055,
"learning_rate": 3.2290829518028867e-07,
"loss": 0.0002,
"step": 928
},
{
"epoch": 0.5308571428571428,
"grad_norm": 0.27683863043785095,
"learning_rate": 3.1416838871368925e-07,
"loss": 0.0497,
"step": 929
},
{
"epoch": 0.5314285714285715,
"grad_norm": 0.06211227923631668,
"learning_rate": 3.0554650104861137e-07,
"loss": 0.0009,
"step": 930
},
{
"epoch": 0.532,
"grad_norm": 0.24819982051849365,
"learning_rate": 2.970427372400353e-07,
"loss": 0.0484,
"step": 931
},
{
"epoch": 0.5325714285714286,
"grad_norm": 0.015587291680276394,
"learning_rate": 2.8865720090364037e-07,
"loss": 0.0007,
"step": 932
},
{
"epoch": 0.5331428571428571,
"grad_norm": 0.13387271761894226,
"learning_rate": 2.8038999421453827e-07,
"loss": 0.0426,
"step": 933
},
{
"epoch": 0.5337142857142857,
"grad_norm": 0.16433075070381165,
"learning_rate": 2.7224121790603517e-07,
"loss": 0.0294,
"step": 934
},
{
"epoch": 0.5342857142857143,
"grad_norm": 0.16424797475337982,
"learning_rate": 2.6421097126839714e-07,
"loss": 0.0208,
"step": 935
},
{
"epoch": 0.5348571428571428,
"grad_norm": 0.38066744804382324,
"learning_rate": 2.5629935214764866e-07,
"loss": 0.0292,
"step": 936
},
{
"epoch": 0.5354285714285715,
"grad_norm": 0.13763435184955597,
"learning_rate": 2.4850645694436736e-07,
"loss": 0.0212,
"step": 937
},
{
"epoch": 0.536,
"grad_norm": 0.03070133551955223,
"learning_rate": 2.4083238061252565e-07,
"loss": 0.0006,
"step": 938
},
{
"epoch": 0.5365714285714286,
"grad_norm": 1.3827248811721802,
"learning_rate": 2.332772166583208e-07,
"loss": 0.0073,
"step": 939
},
{
"epoch": 0.5371428571428571,
"grad_norm": 1.5520464181900024,
"learning_rate": 2.2584105713904126e-07,
"loss": 0.0849,
"step": 940
},
{
"epoch": 0.5377142857142857,
"grad_norm": 0.1565760374069214,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.0009,
"step": 941
},
{
"epoch": 0.5382857142857143,
"grad_norm": 0.3835518956184387,
"learning_rate": 2.1132611238315004e-07,
"loss": 0.0516,
"step": 942
},
{
"epoch": 0.5388571428571428,
"grad_norm": 0.40111058950424194,
"learning_rate": 2.0424750400655947e-07,
"loss": 0.0306,
"step": 943
},
{
"epoch": 0.5394285714285715,
"grad_norm": 0.05026775971055031,
"learning_rate": 1.9728825378278248e-07,
"loss": 0.0012,
"step": 944
},
{
"epoch": 0.54,
"grad_norm": 0.043875012546777725,
"learning_rate": 1.9044844650808468e-07,
"loss": 0.001,
"step": 945
},
{
"epoch": 0.5405714285714286,
"grad_norm": 0.043922051787376404,
"learning_rate": 1.8372816552336025e-07,
"loss": 0.0013,
"step": 946
},
{
"epoch": 0.5411428571428571,
"grad_norm": 0.25579139590263367,
"learning_rate": 1.7712749271311392e-07,
"loss": 0.0506,
"step": 947
},
{
"epoch": 0.5417142857142857,
"grad_norm": 0.07949605584144592,
"learning_rate": 1.706465085044584e-07,
"loss": 0.0019,
"step": 948
},
{
"epoch": 0.5422857142857143,
"grad_norm": 0.08321674913167953,
"learning_rate": 1.6428529186614195e-07,
"loss": 0.0369,
"step": 949
},
{
"epoch": 0.5428571428571428,
"grad_norm": 0.029275426641106606,
"learning_rate": 1.580439203075812e-07,
"loss": 0.0004,
"step": 950
},
{
"epoch": 0.5434285714285715,
"grad_norm": 0.375702828168869,
"learning_rate": 1.519224698779198e-07,
"loss": 0.0288,
"step": 951
},
{
"epoch": 0.544,
"grad_norm": 0.09123431146144867,
"learning_rate": 1.4592101516509916e-07,
"loss": 0.0005,
"step": 952
},
{
"epoch": 0.5445714285714286,
"grad_norm": 0.06241918355226517,
"learning_rate": 1.400396292949513e-07,
"loss": 0.0008,
"step": 953
},
{
"epoch": 0.5451428571428572,
"grad_norm": 0.06081584841012955,
"learning_rate": 1.3427838393030634e-07,
"loss": 0.0007,
"step": 954
},
{
"epoch": 0.5457142857142857,
"grad_norm": 0.33235132694244385,
"learning_rate": 1.2863734927012094e-07,
"loss": 0.045,
"step": 955
},
{
"epoch": 0.5462857142857143,
"grad_norm": 0.16899196803569794,
"learning_rate": 1.231165940486234e-07,
"loss": 0.0253,
"step": 956
},
{
"epoch": 0.5468571428571428,
"grad_norm": 0.03071417473256588,
"learning_rate": 1.1771618553447217e-07,
"loss": 0.0007,
"step": 957
},
{
"epoch": 0.5474285714285714,
"grad_norm": 0.816353976726532,
"learning_rate": 1.1243618952994195e-07,
"loss": 0.023,
"step": 958
},
{
"epoch": 0.548,
"grad_norm": 0.283340185880661,
"learning_rate": 1.0727667037011668e-07,
"loss": 0.0258,
"step": 959
},
{
"epoch": 0.5485714285714286,
"grad_norm": 0.2835102379322052,
"learning_rate": 1.0223769092211012e-07,
"loss": 0.0262,
"step": 960
},
{
"epoch": 0.5491428571428572,
"grad_norm": 0.028213156387209892,
"learning_rate": 9.731931258429638e-08,
"loss": 0.0007,
"step": 961
},
{
"epoch": 0.5497142857142857,
"grad_norm": 0.03289434686303139,
"learning_rate": 9.252159528556404e-08,
"loss": 0.001,
"step": 962
},
{
"epoch": 0.5502857142857143,
"grad_norm": 0.8161189556121826,
"learning_rate": 8.784459748458318e-08,
"loss": 0.0225,
"step": 963
},
{
"epoch": 0.5508571428571428,
"grad_norm": 0.03120247833430767,
"learning_rate": 8.328837616909612e-08,
"loss": 0.0007,
"step": 964
},
{
"epoch": 0.5514285714285714,
"grad_norm": 0.03551860526204109,
"learning_rate": 7.885298685522235e-08,
"loss": 0.0003,
"step": 965
},
{
"epoch": 0.552,
"grad_norm": 0.0328202098608017,
"learning_rate": 7.453848358678018e-08,
"loss": 0.001,
"step": 966
},
{
"epoch": 0.5525714285714286,
"grad_norm": 0.290255606174469,
"learning_rate": 7.034491893463059e-08,
"loss": 0.0461,
"step": 967
},
{
"epoch": 0.5531428571428572,
"grad_norm": 0.1132972463965416,
"learning_rate": 6.627234399603554e-08,
"loss": 0.0014,
"step": 968
},
{
"epoch": 0.5537142857142857,
"grad_norm": 0.03143693134188652,
"learning_rate": 6.232080839403631e-08,
"loss": 0.0009,
"step": 969
},
{
"epoch": 0.5542857142857143,
"grad_norm": 0.3172324001789093,
"learning_rate": 5.849036027684607e-08,
"loss": 0.0438,
"step": 970
},
{
"epoch": 0.5548571428571428,
"grad_norm": 0.4574429392814636,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0241,
"step": 971
},
{
"epoch": 0.5554285714285714,
"grad_norm": 0.3745904862880707,
"learning_rate": 5.119291171211793e-08,
"loss": 0.0283,
"step": 972
},
{
"epoch": 0.556,
"grad_norm": 0.06869683414697647,
"learning_rate": 4.772600018168816e-08,
"loss": 0.0016,
"step": 973
},
{
"epoch": 0.5565714285714286,
"grad_norm": 0.035209354013204575,
"learning_rate": 4.438035396920004e-08,
"loss": 0.0006,
"step": 974
},
{
"epoch": 0.5571428571428572,
"grad_norm": 2.8792927265167236,
"learning_rate": 4.115601384029666e-08,
"loss": 0.0103,
"step": 975
},
{
"epoch": 0.5577142857142857,
"grad_norm": 0.1605510115623474,
"learning_rate": 3.805301908254455e-08,
"loss": 0.0239,
"step": 976
},
{
"epoch": 0.5582857142857143,
"grad_norm": 0.3072490692138672,
"learning_rate": 3.50714075049563e-08,
"loss": 0.0442,
"step": 977
},
{
"epoch": 0.5588571428571428,
"grad_norm": 0.3675854802131653,
"learning_rate": 3.22112154375287e-08,
"loss": 0.0356,
"step": 978
},
{
"epoch": 0.5594285714285714,
"grad_norm": 0.637628436088562,
"learning_rate": 2.947247773079753e-08,
"loss": 0.0614,
"step": 979
},
{
"epoch": 0.56,
"grad_norm": 0.02508588321506977,
"learning_rate": 2.6855227755419046e-08,
"loss": 0.0006,
"step": 980
},
{
"epoch": 0.5605714285714286,
"grad_norm": 0.8140731453895569,
"learning_rate": 2.4359497401758026e-08,
"loss": 0.0783,
"step": 981
},
{
"epoch": 0.5611428571428572,
"grad_norm": 0.1536528766155243,
"learning_rate": 2.1985317079500358e-08,
"loss": 0.0398,
"step": 982
},
{
"epoch": 0.5617142857142857,
"grad_norm": 0.03017653338611126,
"learning_rate": 1.973271571728441e-08,
"loss": 0.0005,
"step": 983
},
{
"epoch": 0.5622857142857143,
"grad_norm": 0.06902480125427246,
"learning_rate": 1.7601720762346895e-08,
"loss": 0.0008,
"step": 984
},
{
"epoch": 0.5628571428571428,
"grad_norm": 1.1665534973144531,
"learning_rate": 1.5592358180189782e-08,
"loss": 0.0784,
"step": 985
},
{
"epoch": 0.5634285714285714,
"grad_norm": 0.33477380871772766,
"learning_rate": 1.370465245426167e-08,
"loss": 0.052,
"step": 986
},
{
"epoch": 0.564,
"grad_norm": 0.7823247909545898,
"learning_rate": 1.1938626585660252e-08,
"loss": 0.0283,
"step": 987
},
{
"epoch": 0.5645714285714286,
"grad_norm": 0.27447688579559326,
"learning_rate": 1.0294302092853647e-08,
"loss": 0.0268,
"step": 988
},
{
"epoch": 0.5651428571428572,
"grad_norm": 0.2749784588813782,
"learning_rate": 8.771699011416169e-09,
"loss": 0.0496,
"step": 989
},
{
"epoch": 0.5657142857142857,
"grad_norm": 0.2998272180557251,
"learning_rate": 7.370835893788508e-09,
"loss": 0.0428,
"step": 990
},
{
"epoch": 0.5662857142857143,
"grad_norm": 0.10891908407211304,
"learning_rate": 6.091729809042379e-09,
"loss": 0.0136,
"step": 991
},
{
"epoch": 0.5668571428571428,
"grad_norm": 0.23953084647655487,
"learning_rate": 4.9343963426840006e-09,
"loss": 0.0254,
"step": 992
},
{
"epoch": 0.5674285714285714,
"grad_norm": 0.03346346691250801,
"learning_rate": 3.898849596456477e-09,
"loss": 0.0003,
"step": 993
},
{
"epoch": 0.568,
"grad_norm": 0.3791404366493225,
"learning_rate": 2.9851021881688314e-09,
"loss": 0.0319,
"step": 994
},
{
"epoch": 0.5685714285714286,
"grad_norm": 0.29911234974861145,
"learning_rate": 2.193165251545004e-09,
"loss": 0.0272,
"step": 995
},
{
"epoch": 0.5691428571428572,
"grad_norm": 0.29955676198005676,
"learning_rate": 1.5230484360873043e-09,
"loss": 0.0451,
"step": 996
},
{
"epoch": 0.5697142857142857,
"grad_norm": 0.15668544173240662,
"learning_rate": 9.74759906957612e-10,
"loss": 0.0245,
"step": 997
},
{
"epoch": 0.5702857142857143,
"grad_norm": 0.27871453762054443,
"learning_rate": 5.483063448785686e-10,
"loss": 0.0255,
"step": 998
},
{
"epoch": 0.5708571428571428,
"grad_norm": 0.3131696879863739,
"learning_rate": 2.436929460525317e-10,
"loss": 0.0509,
"step": 999
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.029159685596823692,
"learning_rate": 6.092342209607083e-11,
"loss": 0.0008,
"step": 1000
}
],
"logging_steps": 1,
"max_steps": 1000,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}