model-nemo / trainer_state.json
Kendamarron's picture
Upload folder using huggingface_hub
d38780e verified
Raw History Blame Contribute Delete
85.4 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 492,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0040650406504065045,
"grad_norm": 7.252714940173002,
"learning_rate": 2.0000000000000002e-07,
"loss": 1.7848,
"step": 1
},
{
"epoch": 0.008130081300813009,
"grad_norm": 8.041954031416958,
"learning_rate": 4.0000000000000003e-07,
"loss": 1.9339,
"step": 2
},
{
"epoch": 0.012195121951219513,
"grad_norm": 7.730138743538377,
"learning_rate": 6.000000000000001e-07,
"loss": 1.7886,
"step": 3
},
{
"epoch": 0.016260162601626018,
"grad_norm": 7.583424866406282,
"learning_rate": 8.000000000000001e-07,
"loss": 1.6851,
"step": 4
},
{
"epoch": 0.02032520325203252,
"grad_norm": 8.096824351961159,
"learning_rate": 1.0000000000000002e-06,
"loss": 1.8599,
"step": 5
},
{
"epoch": 0.024390243902439025,
"grad_norm": 7.079805446171174,
"learning_rate": 1.2000000000000002e-06,
"loss": 1.6364,
"step": 6
},
{
"epoch": 0.028455284552845527,
"grad_norm": 8.059662590377085,
"learning_rate": 1.4000000000000001e-06,
"loss": 1.8774,
"step": 7
},
{
"epoch": 0.032520325203252036,
"grad_norm": 7.784056695808858,
"learning_rate": 1.6000000000000001e-06,
"loss": 1.8972,
"step": 8
},
{
"epoch": 0.036585365853658534,
"grad_norm": 6.700684708686177,
"learning_rate": 1.8000000000000001e-06,
"loss": 1.8278,
"step": 9
},
{
"epoch": 0.04065040650406504,
"grad_norm": 6.514102409318334,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.6222,
"step": 10
},
{
"epoch": 0.044715447154471545,
"grad_norm": 5.64452086813206,
"learning_rate": 2.2e-06,
"loss": 1.7499,
"step": 11
},
{
"epoch": 0.04878048780487805,
"grad_norm": 6.304162192868725,
"learning_rate": 2.4000000000000003e-06,
"loss": 1.9921,
"step": 12
},
{
"epoch": 0.052845528455284556,
"grad_norm": 5.660742187171789,
"learning_rate": 2.6e-06,
"loss": 1.5938,
"step": 13
},
{
"epoch": 0.056910569105691054,
"grad_norm": 5.84135884044965,
"learning_rate": 2.8000000000000003e-06,
"loss": 1.7175,
"step": 14
},
{
"epoch": 0.06097560975609756,
"grad_norm": 5.8883921035199664,
"learning_rate": 3e-06,
"loss": 1.7167,
"step": 15
},
{
"epoch": 0.06504065040650407,
"grad_norm": 4.872577028754156,
"learning_rate": 3.2000000000000003e-06,
"loss": 1.6396,
"step": 16
},
{
"epoch": 0.06910569105691057,
"grad_norm": 5.033828898246033,
"learning_rate": 3.4000000000000005e-06,
"loss": 1.7618,
"step": 17
},
{
"epoch": 0.07317073170731707,
"grad_norm": 5.618404672136545,
"learning_rate": 3.6000000000000003e-06,
"loss": 1.8478,
"step": 18
},
{
"epoch": 0.07723577235772358,
"grad_norm": 4.644359087577748,
"learning_rate": 3.8000000000000005e-06,
"loss": 1.7212,
"step": 19
},
{
"epoch": 0.08130081300813008,
"grad_norm": 4.61494509739694,
"learning_rate": 4.000000000000001e-06,
"loss": 1.7749,
"step": 20
},
{
"epoch": 0.08536585365853659,
"grad_norm": 4.915181497872769,
"learning_rate": 4.2000000000000004e-06,
"loss": 1.7796,
"step": 21
},
{
"epoch": 0.08943089430894309,
"grad_norm": 4.971626559916134,
"learning_rate": 4.4e-06,
"loss": 1.7369,
"step": 22
},
{
"epoch": 0.09349593495934959,
"grad_norm": 4.506844268450052,
"learning_rate": 4.600000000000001e-06,
"loss": 1.7173,
"step": 23
},
{
"epoch": 0.0975609756097561,
"grad_norm": 5.566132011081365,
"learning_rate": 4.800000000000001e-06,
"loss": 1.7578,
"step": 24
},
{
"epoch": 0.1016260162601626,
"grad_norm": 4.285061174044111,
"learning_rate": 5e-06,
"loss": 1.6751,
"step": 25
},
{
"epoch": 0.10569105691056911,
"grad_norm": 3.2399830013754887,
"learning_rate": 5.2e-06,
"loss": 1.3605,
"step": 26
},
{
"epoch": 0.10975609756097561,
"grad_norm": 3.539848061855604,
"learning_rate": 5.400000000000001e-06,
"loss": 1.38,
"step": 27
},
{
"epoch": 0.11382113821138211,
"grad_norm": 4.546848096544689,
"learning_rate": 5.600000000000001e-06,
"loss": 1.5411,
"step": 28
},
{
"epoch": 0.11788617886178862,
"grad_norm": 4.677103196959982,
"learning_rate": 5.8e-06,
"loss": 1.6256,
"step": 29
},
{
"epoch": 0.12195121951219512,
"grad_norm": 4.1219944114782345,
"learning_rate": 6e-06,
"loss": 1.6101,
"step": 30
},
{
"epoch": 0.12601626016260162,
"grad_norm": 4.386185353107293,
"learning_rate": 6.200000000000001e-06,
"loss": 1.6272,
"step": 31
},
{
"epoch": 0.13008130081300814,
"grad_norm": 4.117211442239609,
"learning_rate": 6.4000000000000006e-06,
"loss": 1.4924,
"step": 32
},
{
"epoch": 0.13414634146341464,
"grad_norm": 4.679432150474269,
"learning_rate": 6.600000000000001e-06,
"loss": 1.6172,
"step": 33
},
{
"epoch": 0.13821138211382114,
"grad_norm": 4.141719392574872,
"learning_rate": 6.800000000000001e-06,
"loss": 1.6287,
"step": 34
},
{
"epoch": 0.14227642276422764,
"grad_norm": 14.149512334739653,
"learning_rate": 7e-06,
"loss": 1.7817,
"step": 35
},
{
"epoch": 0.14634146341463414,
"grad_norm": 4.256590260435325,
"learning_rate": 7.2000000000000005e-06,
"loss": 1.5054,
"step": 36
},
{
"epoch": 0.15040650406504066,
"grad_norm": 3.9378372902054215,
"learning_rate": 7.4e-06,
"loss": 1.4642,
"step": 37
},
{
"epoch": 0.15447154471544716,
"grad_norm": 5.073065583562386,
"learning_rate": 7.600000000000001e-06,
"loss": 1.6987,
"step": 38
},
{
"epoch": 0.15853658536585366,
"grad_norm": 3.955080579841974,
"learning_rate": 7.800000000000002e-06,
"loss": 1.4779,
"step": 39
},
{
"epoch": 0.16260162601626016,
"grad_norm": 4.679066673569764,
"learning_rate": 8.000000000000001e-06,
"loss": 1.7102,
"step": 40
},
{
"epoch": 0.16666666666666666,
"grad_norm": 4.4740512918032636,
"learning_rate": 8.2e-06,
"loss": 1.5423,
"step": 41
},
{
"epoch": 0.17073170731707318,
"grad_norm": 4.960822682002186,
"learning_rate": 8.400000000000001e-06,
"loss": 1.6839,
"step": 42
},
{
"epoch": 0.17479674796747968,
"grad_norm": 4.0320252512557095,
"learning_rate": 8.6e-06,
"loss": 1.6922,
"step": 43
},
{
"epoch": 0.17886178861788618,
"grad_norm": 4.3372067875156395,
"learning_rate": 8.8e-06,
"loss": 1.4837,
"step": 44
},
{
"epoch": 0.18292682926829268,
"grad_norm": 4.47419442001546,
"learning_rate": 9e-06,
"loss": 1.5454,
"step": 45
},
{
"epoch": 0.18699186991869918,
"grad_norm": 4.406750306634826,
"learning_rate": 9.200000000000002e-06,
"loss": 1.5301,
"step": 46
},
{
"epoch": 0.1910569105691057,
"grad_norm": 3.9021397483272655,
"learning_rate": 9.4e-06,
"loss": 1.6126,
"step": 47
},
{
"epoch": 0.1951219512195122,
"grad_norm": 3.653827972520367,
"learning_rate": 9.600000000000001e-06,
"loss": 1.4816,
"step": 48
},
{
"epoch": 0.1991869918699187,
"grad_norm": 4.505273199692054,
"learning_rate": 9.800000000000001e-06,
"loss": 1.5949,
"step": 49
},
{
"epoch": 0.2032520325203252,
"grad_norm": 3.565918103537592,
"learning_rate": 1e-05,
"loss": 1.5271,
"step": 50
},
{
"epoch": 0.2073170731707317,
"grad_norm": 4.012380931266431,
"learning_rate": 9.999873702897528e-06,
"loss": 1.5027,
"step": 51
},
{
"epoch": 0.21138211382113822,
"grad_norm": 4.888550731359106,
"learning_rate": 9.999494817970498e-06,
"loss": 1.612,
"step": 52
},
{
"epoch": 0.21544715447154472,
"grad_norm": 4.1912784450852225,
"learning_rate": 9.998863364359734e-06,
"loss": 1.4866,
"step": 53
},
{
"epoch": 0.21951219512195122,
"grad_norm": 4.178267006955291,
"learning_rate": 9.997979373965542e-06,
"loss": 1.653,
"step": 54
},
{
"epoch": 0.22357723577235772,
"grad_norm": 4.87798113769922,
"learning_rate": 9.996842891446092e-06,
"loss": 1.6177,
"step": 55
},
{
"epoch": 0.22764227642276422,
"grad_norm": 3.8484589931839857,
"learning_rate": 9.995453974215164e-06,
"loss": 1.5229,
"step": 56
},
{
"epoch": 0.23170731707317074,
"grad_norm": 4.101650487240815,
"learning_rate": 9.993812692439247e-06,
"loss": 1.5214,
"step": 57
},
{
"epoch": 0.23577235772357724,
"grad_norm": 4.362611564068637,
"learning_rate": 9.991919129033994e-06,
"loss": 1.6307,
"step": 58
},
{
"epoch": 0.23983739837398374,
"grad_norm": 4.655614475337303,
"learning_rate": 9.98977337966003e-06,
"loss": 1.5714,
"step": 59
},
{
"epoch": 0.24390243902439024,
"grad_norm": 4.342587357240943,
"learning_rate": 9.987375552718133e-06,
"loss": 1.4715,
"step": 60
},
{
"epoch": 0.24796747967479674,
"grad_norm": 4.238066429606384,
"learning_rate": 9.984725769343737e-06,
"loss": 1.5536,
"step": 61
},
{
"epoch": 0.25203252032520324,
"grad_norm": 4.301601899528779,
"learning_rate": 9.981824163400827e-06,
"loss": 1.4955,
"step": 62
},
{
"epoch": 0.25609756097560976,
"grad_norm": 3.911274471894397,
"learning_rate": 9.978670881475173e-06,
"loss": 1.5308,
"step": 63
},
{
"epoch": 0.2601626016260163,
"grad_norm": 4.118832227167338,
"learning_rate": 9.975266082866923e-06,
"loss": 1.5555,
"step": 64
},
{
"epoch": 0.26422764227642276,
"grad_norm": 4.432642562527081,
"learning_rate": 9.971609939582556e-06,
"loss": 1.658,
"step": 65
},
{
"epoch": 0.2682926829268293,
"grad_norm": 3.5208387359907727,
"learning_rate": 9.967702636326195e-06,
"loss": 1.4747,
"step": 66
},
{
"epoch": 0.27235772357723576,
"grad_norm": 4.2526896847696305,
"learning_rate": 9.96354437049027e-06,
"loss": 1.523,
"step": 67
},
{
"epoch": 0.2764227642276423,
"grad_norm": 4.217466839694022,
"learning_rate": 9.959135352145552e-06,
"loss": 1.5447,
"step": 68
},
{
"epoch": 0.2804878048780488,
"grad_norm": 4.519363889301345,
"learning_rate": 9.954475804030539e-06,
"loss": 1.6282,
"step": 69
},
{
"epoch": 0.2845528455284553,
"grad_norm": 4.405830508618357,
"learning_rate": 9.9495659615402e-06,
"loss": 1.6676,
"step": 70
},
{
"epoch": 0.2886178861788618,
"grad_norm": 4.063392227520675,
"learning_rate": 9.944406072714086e-06,
"loss": 1.5793,
"step": 71
},
{
"epoch": 0.2926829268292683,
"grad_norm": 3.7279811430490013,
"learning_rate": 9.938996398223802e-06,
"loss": 1.4448,
"step": 72
},
{
"epoch": 0.2967479674796748,
"grad_norm": 4.7340318449158,
"learning_rate": 9.933337211359833e-06,
"loss": 1.5116,
"step": 73
},
{
"epoch": 0.3008130081300813,
"grad_norm": 4.125862373262926,
"learning_rate": 9.927428798017738e-06,
"loss": 1.6252,
"step": 74
},
{
"epoch": 0.3048780487804878,
"grad_norm": 4.503060221923278,
"learning_rate": 9.921271456683716e-06,
"loss": 1.5883,
"step": 75
},
{
"epoch": 0.3089430894308943,
"grad_norm": 4.030964240265612,
"learning_rate": 9.91486549841951e-06,
"loss": 1.4513,
"step": 76
},
{
"epoch": 0.3130081300813008,
"grad_norm": 4.279705027198175,
"learning_rate": 9.908211246846708e-06,
"loss": 1.3775,
"step": 77
},
{
"epoch": 0.3170731707317073,
"grad_norm": 4.368927809869299,
"learning_rate": 9.901309038130392e-06,
"loss": 1.5471,
"step": 78
},
{
"epoch": 0.32113821138211385,
"grad_norm": 3.55743157069438,
"learning_rate": 9.894159220962138e-06,
"loss": 1.3894,
"step": 79
},
{
"epoch": 0.3252032520325203,
"grad_norm": 4.004435314321132,
"learning_rate": 9.886762156542428e-06,
"loss": 1.4869,
"step": 80
},
{
"epoch": 0.32926829268292684,
"grad_norm": 4.27950260857538,
"learning_rate": 9.879118218562384e-06,
"loss": 1.3831,
"step": 81
},
{
"epoch": 0.3333333333333333,
"grad_norm": 4.178642916274787,
"learning_rate": 9.871227793184893e-06,
"loss": 1.3701,
"step": 82
},
{
"epoch": 0.33739837398373984,
"grad_norm": 4.654225154626956,
"learning_rate": 9.863091279025095e-06,
"loss": 1.5812,
"step": 83
},
{
"epoch": 0.34146341463414637,
"grad_norm": 4.033763945760902,
"learning_rate": 9.854709087130261e-06,
"loss": 1.5003,
"step": 84
},
{
"epoch": 0.34552845528455284,
"grad_norm": 4.134788720970193,
"learning_rate": 9.846081640959008e-06,
"loss": 1.4811,
"step": 85
},
{
"epoch": 0.34959349593495936,
"grad_norm": 4.004503453499489,
"learning_rate": 9.837209376359918e-06,
"loss": 1.5358,
"step": 86
},
{
"epoch": 0.35365853658536583,
"grad_norm": 3.7677878191637233,
"learning_rate": 9.828092741549513e-06,
"loss": 1.4858,
"step": 87
},
{
"epoch": 0.35772357723577236,
"grad_norm": 4.578911874711249,
"learning_rate": 9.81873219708962e-06,
"loss": 1.5514,
"step": 88
},
{
"epoch": 0.3617886178861789,
"grad_norm": 4.9270432646197735,
"learning_rate": 9.809128215864096e-06,
"loss": 1.6168,
"step": 89
},
{
"epoch": 0.36585365853658536,
"grad_norm": 4.193560229748371,
"learning_rate": 9.79928128305494e-06,
"loss": 1.5689,
"step": 90
},
{
"epoch": 0.3699186991869919,
"grad_norm": 3.858624299461163,
"learning_rate": 9.789191896117786e-06,
"loss": 1.5973,
"step": 91
},
{
"epoch": 0.37398373983739835,
"grad_norm": 3.6862911668206046,
"learning_rate": 9.778860564756769e-06,
"loss": 1.448,
"step": 92
},
{
"epoch": 0.3780487804878049,
"grad_norm": 4.044958145902229,
"learning_rate": 9.768287810898773e-06,
"loss": 1.4225,
"step": 93
},
{
"epoch": 0.3821138211382114,
"grad_norm": 4.635836143476677,
"learning_rate": 9.757474168667072e-06,
"loss": 1.6042,
"step": 94
},
{
"epoch": 0.3861788617886179,
"grad_norm": 4.237166478010606,
"learning_rate": 9.746420184354334e-06,
"loss": 1.4636,
"step": 95
},
{
"epoch": 0.3902439024390244,
"grad_norm": 3.6945776453925885,
"learning_rate": 9.73512641639504e-06,
"loss": 1.5352,
"step": 96
},
{
"epoch": 0.3943089430894309,
"grad_norm": 4.2788684169490105,
"learning_rate": 9.723593435337252e-06,
"loss": 1.4831,
"step": 97
},
{
"epoch": 0.3983739837398374,
"grad_norm": 4.237562757783015,
"learning_rate": 9.711821823813812e-06,
"loss": 1.4187,
"step": 98
},
{
"epoch": 0.4024390243902439,
"grad_norm": 4.0121887080303695,
"learning_rate": 9.699812176512887e-06,
"loss": 1.5175,
"step": 99
},
{
"epoch": 0.4065040650406504,
"grad_norm": 4.089988976445491,
"learning_rate": 9.68756510014794e-06,
"loss": 1.5804,
"step": 100
},
{
"epoch": 0.4105691056910569,
"grad_norm": 4.2230335971432185,
"learning_rate": 9.675081213427076e-06,
"loss": 1.486,
"step": 101
},
{
"epoch": 0.4146341463414634,
"grad_norm": 4.31542184716973,
"learning_rate": 9.66236114702178e-06,
"loss": 1.5065,
"step": 102
},
{
"epoch": 0.4186991869918699,
"grad_norm": 3.7651861366457844,
"learning_rate": 9.649405543535067e-06,
"loss": 1.447,
"step": 103
},
{
"epoch": 0.42276422764227645,
"grad_norm": 3.565995005609616,
"learning_rate": 9.636215057469009e-06,
"loss": 1.5025,
"step": 104
},
{
"epoch": 0.4268292682926829,
"grad_norm": 4.755520093055592,
"learning_rate": 9.622790355191672e-06,
"loss": 1.5984,
"step": 105
},
{
"epoch": 0.43089430894308944,
"grad_norm": 4.560871118656236,
"learning_rate": 9.609132114903458e-06,
"loss": 1.6039,
"step": 106
},
{
"epoch": 0.4349593495934959,
"grad_norm": 4.14097350938773,
"learning_rate": 9.595241026602836e-06,
"loss": 1.5437,
"step": 107
},
{
"epoch": 0.43902439024390244,
"grad_norm": 4.134377632845812,
"learning_rate": 9.581117792051487e-06,
"loss": 1.5875,
"step": 108
},
{
"epoch": 0.44308943089430897,
"grad_norm": 4.181076547736983,
"learning_rate": 9.56676312473885e-06,
"loss": 1.4929,
"step": 109
},
{
"epoch": 0.44715447154471544,
"grad_norm": 4.484563052187801,
"learning_rate": 9.552177749846083e-06,
"loss": 1.652,
"step": 110
},
{
"epoch": 0.45121951219512196,
"grad_norm": 4.097649732485832,
"learning_rate": 9.537362404209419e-06,
"loss": 1.4188,
"step": 111
},
{
"epoch": 0.45528455284552843,
"grad_norm": 3.7058287794583844,
"learning_rate": 9.522317836282949e-06,
"loss": 1.386,
"step": 112
},
{
"epoch": 0.45934959349593496,
"grad_norm": 4.493281379042812,
"learning_rate": 9.507044806100806e-06,
"loss": 1.5926,
"step": 113
},
{
"epoch": 0.4634146341463415,
"grad_norm": 4.316018698694785,
"learning_rate": 9.491544085238778e-06,
"loss": 1.5401,
"step": 114
},
{
"epoch": 0.46747967479674796,
"grad_norm": 4.894081838523462,
"learning_rate": 9.475816456775313e-06,
"loss": 1.4803,
"step": 115
},
{
"epoch": 0.4715447154471545,
"grad_norm": 3.7742812759076876,
"learning_rate": 9.459862715251973e-06,
"loss": 1.2609,
"step": 116
},
{
"epoch": 0.47560975609756095,
"grad_norm": 4.211778484845177,
"learning_rate": 9.44368366663329e-06,
"loss": 1.5086,
"step": 117
},
{
"epoch": 0.4796747967479675,
"grad_norm": 4.178202675025514,
"learning_rate": 9.427280128266049e-06,
"loss": 1.4877,
"step": 118
},
{
"epoch": 0.483739837398374,
"grad_norm": 3.81342990030962,
"learning_rate": 9.410652928837998e-06,
"loss": 1.556,
"step": 119
},
{
"epoch": 0.4878048780487805,
"grad_norm": 3.6819015497865455,
"learning_rate": 9.393802908335978e-06,
"loss": 1.2884,
"step": 120
},
{
"epoch": 0.491869918699187,
"grad_norm": 4.617462995606693,
"learning_rate": 9.376730918003495e-06,
"loss": 1.6417,
"step": 121
},
{
"epoch": 0.4959349593495935,
"grad_norm": 4.226492412729268,
"learning_rate": 9.359437820297716e-06,
"loss": 1.5866,
"step": 122
},
{
"epoch": 0.5,
"grad_norm": 3.5584733289799235,
"learning_rate": 9.341924488845892e-06,
"loss": 1.5407,
"step": 123
},
{
"epoch": 0.5040650406504065,
"grad_norm": 4.135798632090438,
"learning_rate": 9.324191808401235e-06,
"loss": 1.4992,
"step": 124
},
{
"epoch": 0.508130081300813,
"grad_norm": 4.210768157038929,
"learning_rate": 9.306240674798203e-06,
"loss": 1.5432,
"step": 125
},
{
"epoch": 0.5121951219512195,
"grad_norm": 4.38733920704163,
"learning_rate": 9.288071994907262e-06,
"loss": 1.5541,
"step": 126
},
{
"epoch": 0.516260162601626,
"grad_norm": 4.481255250958727,
"learning_rate": 9.269686686589063e-06,
"loss": 1.4207,
"step": 127
},
{
"epoch": 0.5203252032520326,
"grad_norm": 2.6132141931469346,
"learning_rate": 9.251085678648072e-06,
"loss": 1.0761,
"step": 128
},
{
"epoch": 0.524390243902439,
"grad_norm": 4.48309463777525,
"learning_rate": 9.232269910785651e-06,
"loss": 1.4887,
"step": 129
},
{
"epoch": 0.5284552845528455,
"grad_norm": 4.429844303156119,
"learning_rate": 9.213240333552589e-06,
"loss": 1.7002,
"step": 130
},
{
"epoch": 0.532520325203252,
"grad_norm": 3.8214018897562654,
"learning_rate": 9.193997908301069e-06,
"loss": 1.6757,
"step": 131
},
{
"epoch": 0.5365853658536586,
"grad_norm": 4.2570506828256605,
"learning_rate": 9.174543607136111e-06,
"loss": 1.4434,
"step": 132
},
{
"epoch": 0.540650406504065,
"grad_norm": 4.968270587243067,
"learning_rate": 9.154878412866465e-06,
"loss": 1.5012,
"step": 133
},
{
"epoch": 0.5447154471544715,
"grad_norm": 4.927472957532449,
"learning_rate": 9.135003318954954e-06,
"loss": 1.5436,
"step": 134
},
{
"epoch": 0.5487804878048781,
"grad_norm": 3.6784871911807313,
"learning_rate": 9.114919329468283e-06,
"loss": 1.5054,
"step": 135
},
{
"epoch": 0.5528455284552846,
"grad_norm": 3.909457661323846,
"learning_rate": 9.094627459026326e-06,
"loss": 1.6024,
"step": 136
},
{
"epoch": 0.556910569105691,
"grad_norm": 3.6189177817219274,
"learning_rate": 9.074128732750859e-06,
"loss": 1.4737,
"step": 137
},
{
"epoch": 0.5609756097560976,
"grad_norm": 4.396763959996382,
"learning_rate": 9.053424186213776e-06,
"loss": 1.6196,
"step": 138
},
{
"epoch": 0.5650406504065041,
"grad_norm": 4.05959424566033,
"learning_rate": 9.032514865384767e-06,
"loss": 1.501,
"step": 139
},
{
"epoch": 0.5691056910569106,
"grad_norm": 3.8692860407635323,
"learning_rate": 9.011401826578492e-06,
"loss": 1.5761,
"step": 140
},
{
"epoch": 0.573170731707317,
"grad_norm": 4.130928540510955,
"learning_rate": 8.990086136401199e-06,
"loss": 1.4981,
"step": 141
},
{
"epoch": 0.5772357723577236,
"grad_norm": 3.4615012538912073,
"learning_rate": 8.968568871696847e-06,
"loss": 1.5123,
"step": 142
},
{
"epoch": 0.5813008130081301,
"grad_norm": 4.194843955078523,
"learning_rate": 8.946851119492717e-06,
"loss": 1.3952,
"step": 143
},
{
"epoch": 0.5853658536585366,
"grad_norm": 3.8748342201448014,
"learning_rate": 8.924933976944474e-06,
"loss": 1.4191,
"step": 144
},
{
"epoch": 0.5894308943089431,
"grad_norm": 3.9671232026128327,
"learning_rate": 8.902818551280758e-06,
"loss": 1.4265,
"step": 145
},
{
"epoch": 0.5934959349593496,
"grad_norm": 3.7789529942834434,
"learning_rate": 8.880505959747245e-06,
"loss": 1.5675,
"step": 146
},
{
"epoch": 0.5975609756097561,
"grad_norm": 3.9650111560868946,
"learning_rate": 8.857997329550195e-06,
"loss": 1.4827,
"step": 147
},
{
"epoch": 0.6016260162601627,
"grad_norm": 4.206949987820223,
"learning_rate": 8.835293797799517e-06,
"loss": 1.6229,
"step": 148
},
{
"epoch": 0.6056910569105691,
"grad_norm": 5.392821026840922,
"learning_rate": 8.812396511451324e-06,
"loss": 1.5615,
"step": 149
},
{
"epoch": 0.6097560975609756,
"grad_norm": 3.4784769352898297,
"learning_rate": 8.789306627249985e-06,
"loss": 1.3398,
"step": 150
},
{
"epoch": 0.6138211382113821,
"grad_norm": 4.545003974920788,
"learning_rate": 8.766025311669685e-06,
"loss": 1.5171,
"step": 151
},
{
"epoch": 0.6178861788617886,
"grad_norm": 3.492957975610403,
"learning_rate": 8.742553740855507e-06,
"loss": 1.4777,
"step": 152
},
{
"epoch": 0.6219512195121951,
"grad_norm": 3.5683430523130943,
"learning_rate": 8.718893100564002e-06,
"loss": 1.5097,
"step": 153
},
{
"epoch": 0.6260162601626016,
"grad_norm": 3.615648900985645,
"learning_rate": 8.695044586103297e-06,
"loss": 1.3382,
"step": 154
},
{
"epoch": 0.6300813008130082,
"grad_norm": 3.9273873075270926,
"learning_rate": 8.6710094022727e-06,
"loss": 1.4144,
"step": 155
},
{
"epoch": 0.6341463414634146,
"grad_norm": 4.520103275610601,
"learning_rate": 8.646788763301842e-06,
"loss": 1.4159,
"step": 156
},
{
"epoch": 0.6382113821138211,
"grad_norm": 3.7370573534301554,
"learning_rate": 8.62238389278933e-06,
"loss": 1.4078,
"step": 157
},
{
"epoch": 0.6422764227642277,
"grad_norm": 3.7448394640213785,
"learning_rate": 8.59779602364094e-06,
"loss": 1.5356,
"step": 158
},
{
"epoch": 0.6463414634146342,
"grad_norm": 3.531064969394623,
"learning_rate": 8.573026398007323e-06,
"loss": 1.5079,
"step": 159
},
{
"epoch": 0.6504065040650406,
"grad_norm": 4.206132375250735,
"learning_rate": 8.548076267221258e-06,
"loss": 1.5113,
"step": 160
},
{
"epoch": 0.6544715447154471,
"grad_norm": 4.3840587897407834,
"learning_rate": 8.52294689173443e-06,
"loss": 1.566,
"step": 161
},
{
"epoch": 0.6585365853658537,
"grad_norm": 3.7017661184350334,
"learning_rate": 8.497639541053769e-06,
"loss": 1.4671,
"step": 162
},
{
"epoch": 0.6626016260162602,
"grad_norm": 3.7543935791161176,
"learning_rate": 8.472155493677299e-06,
"loss": 1.4916,
"step": 163
},
{
"epoch": 0.6666666666666666,
"grad_norm": 4.6253028926456405,
"learning_rate": 8.446496037029555e-06,
"loss": 1.5772,
"step": 164
},
{
"epoch": 0.6707317073170732,
"grad_norm": 3.9393510425216682,
"learning_rate": 8.420662467396548e-06,
"loss": 1.5256,
"step": 165
},
{
"epoch": 0.6747967479674797,
"grad_norm": 3.6125553513227113,
"learning_rate": 8.394656089860274e-06,
"loss": 1.5097,
"step": 166
},
{
"epoch": 0.6788617886178862,
"grad_norm": 4.16306624806542,
"learning_rate": 8.368478218232787e-06,
"loss": 1.4933,
"step": 167
},
{
"epoch": 0.6829268292682927,
"grad_norm": 3.953302118531983,
"learning_rate": 8.342130174989819e-06,
"loss": 1.4468,
"step": 168
},
{
"epoch": 0.6869918699186992,
"grad_norm": 4.690723855965399,
"learning_rate": 8.315613291203977e-06,
"loss": 1.5799,
"step": 169
},
{
"epoch": 0.6910569105691057,
"grad_norm": 3.9042113692341416,
"learning_rate": 8.288928906477497e-06,
"loss": 1.3188,
"step": 170
},
{
"epoch": 0.6951219512195121,
"grad_norm": 3.819625555583647,
"learning_rate": 8.262078368874566e-06,
"loss": 1.5533,
"step": 171
},
{
"epoch": 0.6991869918699187,
"grad_norm": 2.9715877508305724,
"learning_rate": 8.235063034853228e-06,
"loss": 1.1148,
"step": 172
},
{
"epoch": 0.7032520325203252,
"grad_norm": 3.606400141092476,
"learning_rate": 8.207884269196845e-06,
"loss": 1.4893,
"step": 173
},
{
"epoch": 0.7073170731707317,
"grad_norm": 4.461915743399059,
"learning_rate": 8.180543444945154e-06,
"loss": 1.5427,
"step": 174
},
{
"epoch": 0.7113821138211383,
"grad_norm": 4.779705190716492,
"learning_rate": 8.153041943324912e-06,
"loss": 1.414,
"step": 175
},
{
"epoch": 0.7154471544715447,
"grad_norm": 4.276486153672041,
"learning_rate": 8.125381153680103e-06,
"loss": 1.5302,
"step": 176
},
{
"epoch": 0.7195121951219512,
"grad_norm": 4.238840116007087,
"learning_rate": 8.097562473401764e-06,
"loss": 1.3102,
"step": 177
},
{
"epoch": 0.7235772357723578,
"grad_norm": 4.237740381579647,
"learning_rate": 8.069587307857377e-06,
"loss": 1.4436,
"step": 178
},
{
"epoch": 0.7276422764227642,
"grad_norm": 3.7162216825525927,
"learning_rate": 8.041457070319884e-06,
"loss": 1.2793,
"step": 179
},
{
"epoch": 0.7317073170731707,
"grad_norm": 3.844333529762707,
"learning_rate": 8.013173181896283e-06,
"loss": 1.5374,
"step": 180
},
{
"epoch": 0.7357723577235772,
"grad_norm": 3.8220385386917615,
"learning_rate": 7.984737071455834e-06,
"loss": 1.4163,
"step": 181
},
{
"epoch": 0.7398373983739838,
"grad_norm": 3.6685146114993943,
"learning_rate": 7.95615017555788e-06,
"loss": 1.4275,
"step": 182
},
{
"epoch": 0.7439024390243902,
"grad_norm": 4.052963640332902,
"learning_rate": 7.927413938379268e-06,
"loss": 1.429,
"step": 183
},
{
"epoch": 0.7479674796747967,
"grad_norm": 4.465867327817124,
"learning_rate": 7.898529811641393e-06,
"loss": 1.6021,
"step": 184
},
{
"epoch": 0.7520325203252033,
"grad_norm": 4.262132065127325,
"learning_rate": 7.869499254536865e-06,
"loss": 1.4571,
"step": 185
},
{
"epoch": 0.7560975609756098,
"grad_norm": 4.088434419556255,
"learning_rate": 7.84032373365578e-06,
"loss": 1.5395,
"step": 186
},
{
"epoch": 0.7601626016260162,
"grad_norm": 5.08541650116389,
"learning_rate": 7.811004722911637e-06,
"loss": 1.5279,
"step": 187
},
{
"epoch": 0.7642276422764228,
"grad_norm": 4.434520321233798,
"learning_rate": 7.781543703466881e-06,
"loss": 1.6533,
"step": 188
},
{
"epoch": 0.7682926829268293,
"grad_norm": 4.156561821204787,
"learning_rate": 7.751942163658066e-06,
"loss": 1.4536,
"step": 189
},
{
"epoch": 0.7723577235772358,
"grad_norm": 4.052792569245924,
"learning_rate": 7.722201598920673e-06,
"loss": 1.5543,
"step": 190
},
{
"epoch": 0.7764227642276422,
"grad_norm": 5.1250039974702775,
"learning_rate": 7.692323511713568e-06,
"loss": 1.5276,
"step": 191
},
{
"epoch": 0.7804878048780488,
"grad_norm": 3.8490656548861635,
"learning_rate": 7.662309411443084e-06,
"loss": 1.4725,
"step": 192
},
{
"epoch": 0.7845528455284553,
"grad_norm": 3.5599844603279434,
"learning_rate": 7.63216081438678e-06,
"loss": 1.4958,
"step": 193
},
{
"epoch": 0.7886178861788617,
"grad_norm": 3.89224947478226,
"learning_rate": 7.601879243616838e-06,
"loss": 1.4679,
"step": 194
},
{
"epoch": 0.7926829268292683,
"grad_norm": 4.291686100205089,
"learning_rate": 7.571466228923115e-06,
"loss": 1.5106,
"step": 195
},
{
"epoch": 0.7967479674796748,
"grad_norm": 3.7108822171665863,
"learning_rate": 7.540923306735868e-06,
"loss": 1.487,
"step": 196
},
{
"epoch": 0.8008130081300813,
"grad_norm": 3.0914010089470825,
"learning_rate": 7.510252020048121e-06,
"loss": 1.229,
"step": 197
},
{
"epoch": 0.8048780487804879,
"grad_norm": 4.516517851884792,
"learning_rate": 7.479453918337733e-06,
"loss": 1.4602,
"step": 198
},
{
"epoch": 0.8089430894308943,
"grad_norm": 4.130400559771834,
"learning_rate": 7.448530557489105e-06,
"loss": 1.396,
"step": 199
},
{
"epoch": 0.8130081300813008,
"grad_norm": 4.4322915898599655,
"learning_rate": 7.417483499714589e-06,
"loss": 1.5604,
"step": 200
},
{
"epoch": 0.8170731707317073,
"grad_norm": 3.8094872673493767,
"learning_rate": 7.386314313475557e-06,
"loss": 1.5942,
"step": 201
},
{
"epoch": 0.8211382113821138,
"grad_norm": 3.544830361462593,
"learning_rate": 7.355024573403174e-06,
"loss": 1.5303,
"step": 202
},
{
"epoch": 0.8252032520325203,
"grad_norm": 4.250695002729135,
"learning_rate": 7.323615860218844e-06,
"loss": 1.4203,
"step": 203
},
{
"epoch": 0.8292682926829268,
"grad_norm": 4.425292970743814,
"learning_rate": 7.292089760654352e-06,
"loss": 1.5367,
"step": 204
},
{
"epoch": 0.8333333333333334,
"grad_norm": 3.7498722104475686,
"learning_rate": 7.2604478673717095e-06,
"loss": 1.5462,
"step": 205
},
{
"epoch": 0.8373983739837398,
"grad_norm": 4.21611622145359,
"learning_rate": 7.2286917788826926e-06,
"loss": 1.66,
"step": 206
},
{
"epoch": 0.8414634146341463,
"grad_norm": 3.7234454752744686,
"learning_rate": 7.196823099468084e-06,
"loss": 1.4327,
"step": 207
},
{
"epoch": 0.8455284552845529,
"grad_norm": 4.305427329850879,
"learning_rate": 7.1648434390966356e-06,
"loss": 1.3825,
"step": 208
},
{
"epoch": 0.8495934959349594,
"grad_norm": 4.371006316975152,
"learning_rate": 7.132754413343721e-06,
"loss": 1.619,
"step": 209
},
{
"epoch": 0.8536585365853658,
"grad_norm": 3.944138562661673,
"learning_rate": 7.100557643309732e-06,
"loss": 1.468,
"step": 210
},
{
"epoch": 0.8577235772357723,
"grad_norm": 3.9244009579137975,
"learning_rate": 7.0682547555381734e-06,
"loss": 1.4736,
"step": 211
},
{
"epoch": 0.8617886178861789,
"grad_norm": 3.9797118366214597,
"learning_rate": 7.035847381933494e-06,
"loss": 1.5529,
"step": 212
},
{
"epoch": 0.8658536585365854,
"grad_norm": 4.5651684626691,
"learning_rate": 7.003337159678649e-06,
"loss": 1.5253,
"step": 213
},
{
"epoch": 0.8699186991869918,
"grad_norm": 3.9234590079518483,
"learning_rate": 6.970725731152389e-06,
"loss": 1.5733,
"step": 214
},
{
"epoch": 0.8739837398373984,
"grad_norm": 3.8509361368556236,
"learning_rate": 6.938014743846285e-06,
"loss": 1.5955,
"step": 215
},
{
"epoch": 0.8780487804878049,
"grad_norm": 4.110184642785762,
"learning_rate": 6.905205850281502e-06,
"loss": 1.5388,
"step": 216
},
{
"epoch": 0.8821138211382114,
"grad_norm": 3.6055230082787624,
"learning_rate": 6.872300707925319e-06,
"loss": 1.4322,
"step": 217
},
{
"epoch": 0.8861788617886179,
"grad_norm": 4.063053038982541,
"learning_rate": 6.8393009791073895e-06,
"loss": 1.4822,
"step": 218
},
{
"epoch": 0.8902439024390244,
"grad_norm": 3.5220426836484253,
"learning_rate": 6.806208330935766e-06,
"loss": 1.3724,
"step": 219
},
{
"epoch": 0.8943089430894309,
"grad_norm": 3.7775619539984384,
"learning_rate": 6.773024435212678e-06,
"loss": 1.4603,
"step": 220
},
{
"epoch": 0.8983739837398373,
"grad_norm": 4.598762256770478,
"learning_rate": 6.739750968350081e-06,
"loss": 1.5066,
"step": 221
},
{
"epoch": 0.9024390243902439,
"grad_norm": 4.198093521281738,
"learning_rate": 6.706389611284953e-06,
"loss": 1.5804,
"step": 222
},
{
"epoch": 0.9065040650406504,
"grad_norm": 3.8645227144297998,
"learning_rate": 6.6729420493943875e-06,
"loss": 1.4994,
"step": 223
},
{
"epoch": 0.9105691056910569,
"grad_norm": 3.3258313240727646,
"learning_rate": 6.639409972410446e-06,
"loss": 1.3242,
"step": 224
},
{
"epoch": 0.9146341463414634,
"grad_norm": 4.036090909687347,
"learning_rate": 6.605795074334793e-06,
"loss": 1.5058,
"step": 225
},
{
"epoch": 0.9186991869918699,
"grad_norm": 4.275435429861942,
"learning_rate": 6.57209905335312e-06,
"loss": 1.4772,
"step": 226
},
{
"epoch": 0.9227642276422764,
"grad_norm": 3.6089292407316997,
"learning_rate": 6.538323611749351e-06,
"loss": 1.4726,
"step": 227
},
{
"epoch": 0.926829268292683,
"grad_norm": 3.611573503513016,
"learning_rate": 6.504470455819651e-06,
"loss": 1.4672,
"step": 228
},
{
"epoch": 0.9308943089430894,
"grad_norm": 3.8214510249995777,
"learning_rate": 6.470541295786222e-06,
"loss": 1.3973,
"step": 229
},
{
"epoch": 0.9349593495934959,
"grad_norm": 3.835773184688124,
"learning_rate": 6.436537845710904e-06,
"loss": 1.5484,
"step": 230
},
{
"epoch": 0.9390243902439024,
"grad_norm": 4.04783334864951,
"learning_rate": 6.402461823408584e-06,
"loss": 1.5051,
"step": 231
},
{
"epoch": 0.943089430894309,
"grad_norm": 4.662221169315537,
"learning_rate": 6.368314950360416e-06,
"loss": 1.4658,
"step": 232
},
{
"epoch": 0.9471544715447154,
"grad_norm": 3.6403213977944127,
"learning_rate": 6.334098951626847e-06,
"loss": 1.4847,
"step": 233
},
{
"epoch": 0.9512195121951219,
"grad_norm": 3.790783459144055,
"learning_rate": 6.299815555760478e-06,
"loss": 1.431,
"step": 234
},
{
"epoch": 0.9552845528455285,
"grad_norm": 3.501495900552076,
"learning_rate": 6.265466494718731e-06,
"loss": 1.2882,
"step": 235
},
{
"epoch": 0.959349593495935,
"grad_norm": 3.9957053415971613,
"learning_rate": 6.231053503776363e-06,
"loss": 1.3099,
"step": 236
},
{
"epoch": 0.9634146341463414,
"grad_norm": 3.844161677005182,
"learning_rate": 6.1965783214377895e-06,
"loss": 1.5197,
"step": 237
},
{
"epoch": 0.967479674796748,
"grad_norm": 3.8384616545529986,
"learning_rate": 6.1620426893492645e-06,
"loss": 1.5219,
"step": 238
},
{
"epoch": 0.9715447154471545,
"grad_norm": 3.346881287038091,
"learning_rate": 6.127448352210894e-06,
"loss": 1.2766,
"step": 239
},
{
"epoch": 0.975609756097561,
"grad_norm": 3.841358671953983,
"learning_rate": 6.092797057688496e-06,
"loss": 1.4856,
"step": 240
},
{
"epoch": 0.9796747967479674,
"grad_norm": 3.2273602717235796,
"learning_rate": 6.058090556325305e-06,
"loss": 1.3838,
"step": 241
},
{
"epoch": 0.983739837398374,
"grad_norm": 3.781592629619726,
"learning_rate": 6.0233306014535505e-06,
"loss": 1.4198,
"step": 242
},
{
"epoch": 0.9878048780487805,
"grad_norm": 4.052768925377713,
"learning_rate": 5.988518949105862e-06,
"loss": 1.5785,
"step": 243
},
{
"epoch": 0.991869918699187,
"grad_norm": 3.8803305720836336,
"learning_rate": 5.953657357926569e-06,
"loss": 1.4233,
"step": 244
},
{
"epoch": 0.9959349593495935,
"grad_norm": 5.103389183399254,
"learning_rate": 5.918747589082853e-06,
"loss": 1.5313,
"step": 245
},
{
"epoch": 1.0,
"grad_norm": 3.5704015527732627,
"learning_rate": 5.883791406175775e-06,
"loss": 1.2762,
"step": 246
},
{
"epoch": 1.0040650406504066,
"grad_norm": 4.216632291136199,
"learning_rate": 5.848790575151181e-06,
"loss": 0.9355,
"step": 247
},
{
"epoch": 1.008130081300813,
"grad_norm": 4.106578314905918,
"learning_rate": 5.813746864210489e-06,
"loss": 1.0987,
"step": 248
},
{
"epoch": 1.0121951219512195,
"grad_norm": 4.290456410003726,
"learning_rate": 5.778662043721359e-06,
"loss": 1.2815,
"step": 249
},
{
"epoch": 1.016260162601626,
"grad_norm": 3.4931202641139367,
"learning_rate": 5.743537886128258e-06,
"loss": 1.0524,
"step": 250
},
{
"epoch": 1.0203252032520325,
"grad_norm": 3.83506643883134,
"learning_rate": 5.708376165862921e-06,
"loss": 0.9908,
"step": 251
},
{
"epoch": 1.024390243902439,
"grad_norm": 3.399638782538431,
"learning_rate": 5.673178659254698e-06,
"loss": 1.0546,
"step": 252
},
{
"epoch": 1.0284552845528456,
"grad_norm": 3.8422737797255873,
"learning_rate": 5.637947144440832e-06,
"loss": 1.0086,
"step": 253
},
{
"epoch": 1.032520325203252,
"grad_norm": 4.323870126328155,
"learning_rate": 5.6026834012766155e-06,
"loss": 1.0805,
"step": 254
},
{
"epoch": 1.0365853658536586,
"grad_norm": 3.5450826685712817,
"learning_rate": 5.567389211245486e-06,
"loss": 0.8563,
"step": 255
},
{
"epoch": 1.040650406504065,
"grad_norm": 3.2888573965449437,
"learning_rate": 5.532066357369012e-06,
"loss": 1.0105,
"step": 256
},
{
"epoch": 1.0447154471544715,
"grad_norm": 4.190107194544142,
"learning_rate": 5.496716624116836e-06,
"loss": 0.9597,
"step": 257
},
{
"epoch": 1.048780487804878,
"grad_norm": 3.6862994007034238,
"learning_rate": 5.46134179731651e-06,
"loss": 1.0305,
"step": 258
},
{
"epoch": 1.0528455284552845,
"grad_norm": 3.555468688002694,
"learning_rate": 5.425943664063284e-06,
"loss": 0.8827,
"step": 259
},
{
"epoch": 1.056910569105691,
"grad_norm": 4.094504105572881,
"learning_rate": 5.390524012629824e-06,
"loss": 1.068,
"step": 260
},
{
"epoch": 1.0609756097560976,
"grad_norm": 3.5250058893374794,
"learning_rate": 5.3550846323758666e-06,
"loss": 0.8818,
"step": 261
},
{
"epoch": 1.065040650406504,
"grad_norm": 3.542846431805926,
"learning_rate": 5.319627313657829e-06,
"loss": 0.9169,
"step": 262
},
{
"epoch": 1.0691056910569106,
"grad_norm": 3.8586777254657783,
"learning_rate": 5.284153847738356e-06,
"loss": 1.0591,
"step": 263
},
{
"epoch": 1.0731707317073171,
"grad_norm": 4.235851716612974,
"learning_rate": 5.248666026695835e-06,
"loss": 1.141,
"step": 264
},
{
"epoch": 1.0772357723577235,
"grad_norm": 3.7250854058603085,
"learning_rate": 5.213165643333851e-06,
"loss": 1.0495,
"step": 265
},
{
"epoch": 1.08130081300813,
"grad_norm": 3.8263762456879125,
"learning_rate": 5.177654491090627e-06,
"loss": 1.0217,
"step": 266
},
{
"epoch": 1.0853658536585367,
"grad_norm": 4.412392915802862,
"learning_rate": 5.1421343639484165e-06,
"loss": 1.0547,
"step": 267
},
{
"epoch": 1.089430894308943,
"grad_norm": 4.181058115462962,
"learning_rate": 5.1066070563428736e-06,
"loss": 0.9316,
"step": 268
},
{
"epoch": 1.0934959349593496,
"grad_norm": 3.3937572080987177,
"learning_rate": 5.071074363072403e-06,
"loss": 0.9298,
"step": 269
},
{
"epoch": 1.0975609756097562,
"grad_norm": 4.355853356906208,
"learning_rate": 5.035538079207488e-06,
"loss": 0.8765,
"step": 270
},
{
"epoch": 1.1016260162601625,
"grad_norm": 3.298839176747146,
"learning_rate": 5e-06,
"loss": 0.9391,
"step": 271
},
{
"epoch": 1.1056910569105691,
"grad_norm": 4.504229350855425,
"learning_rate": 4.964461920792512e-06,
"loss": 1.1586,
"step": 272
},
{
"epoch": 1.1097560975609757,
"grad_norm": 3.5656359190998557,
"learning_rate": 4.928925636927597e-06,
"loss": 1.0144,
"step": 273
},
{
"epoch": 1.113821138211382,
"grad_norm": 3.618874611859637,
"learning_rate": 4.893392943657127e-06,
"loss": 0.8568,
"step": 274
},
{
"epoch": 1.1178861788617886,
"grad_norm": 4.169072041098469,
"learning_rate": 4.857865636051586e-06,
"loss": 1.0775,
"step": 275
},
{
"epoch": 1.1219512195121952,
"grad_norm": 3.443485206553656,
"learning_rate": 4.822345508909376e-06,
"loss": 0.9325,
"step": 276
},
{
"epoch": 1.1260162601626016,
"grad_norm": 3.2881344998560027,
"learning_rate": 4.786834356666153e-06,
"loss": 0.9096,
"step": 277
},
{
"epoch": 1.1300813008130082,
"grad_norm": 3.795586925465744,
"learning_rate": 4.751333973304166e-06,
"loss": 0.9672,
"step": 278
},
{
"epoch": 1.1341463414634148,
"grad_norm": 3.6242186524308955,
"learning_rate": 4.715846152261645e-06,
"loss": 0.9583,
"step": 279
},
{
"epoch": 1.1382113821138211,
"grad_norm": 4.631530737465761,
"learning_rate": 4.680372686342173e-06,
"loss": 1.0386,
"step": 280
},
{
"epoch": 1.1422764227642277,
"grad_norm": 3.931455744651018,
"learning_rate": 4.644915367624134e-06,
"loss": 0.9738,
"step": 281
},
{
"epoch": 1.146341463414634,
"grad_norm": 3.886080843660189,
"learning_rate": 4.609475987370177e-06,
"loss": 1.0012,
"step": 282
},
{
"epoch": 1.1504065040650406,
"grad_norm": 4.228319224768889,
"learning_rate": 4.5740563359367164e-06,
"loss": 1.1627,
"step": 283
},
{
"epoch": 1.1544715447154472,
"grad_norm": 3.514944636631718,
"learning_rate": 4.53865820268349e-06,
"loss": 0.956,
"step": 284
},
{
"epoch": 1.1585365853658536,
"grad_norm": 3.538951910470513,
"learning_rate": 4.503283375883165e-06,
"loss": 0.9759,
"step": 285
},
{
"epoch": 1.1626016260162602,
"grad_norm": 4.274123074176085,
"learning_rate": 4.467933642630989e-06,
"loss": 1.0171,
"step": 286
},
{
"epoch": 1.1666666666666667,
"grad_norm": 3.836022987364166,
"learning_rate": 4.432610788754517e-06,
"loss": 1.0258,
"step": 287
},
{
"epoch": 1.170731707317073,
"grad_norm": 3.678535995693643,
"learning_rate": 4.397316598723385e-06,
"loss": 1.0198,
"step": 288
},
{
"epoch": 1.1747967479674797,
"grad_norm": 3.393770150546556,
"learning_rate": 4.362052855559171e-06,
"loss": 0.9732,
"step": 289
},
{
"epoch": 1.1788617886178863,
"grad_norm": 3.858038880135667,
"learning_rate": 4.326821340745304e-06,
"loss": 1.0039,
"step": 290
},
{
"epoch": 1.1829268292682926,
"grad_norm": 4.00764244121564,
"learning_rate": 4.291623834137082e-06,
"loss": 0.9767,
"step": 291
},
{
"epoch": 1.1869918699186992,
"grad_norm": 3.933188571362055,
"learning_rate": 4.256462113871741e-06,
"loss": 1.0594,
"step": 292
},
{
"epoch": 1.1910569105691058,
"grad_norm": 4.501928479732958,
"learning_rate": 4.2213379562786406e-06,
"loss": 0.9411,
"step": 293
},
{
"epoch": 1.1951219512195121,
"grad_norm": 3.7101337632487152,
"learning_rate": 4.186253135789511e-06,
"loss": 0.9728,
"step": 294
},
{
"epoch": 1.1991869918699187,
"grad_norm": 3.8236826876602303,
"learning_rate": 4.15120942484882e-06,
"loss": 0.9036,
"step": 295
},
{
"epoch": 1.203252032520325,
"grad_norm": 4.024473693532734,
"learning_rate": 4.116208593824227e-06,
"loss": 1.0781,
"step": 296
},
{
"epoch": 1.2073170731707317,
"grad_norm": 3.6369388820395283,
"learning_rate": 4.081252410917148e-06,
"loss": 0.9891,
"step": 297
},
{
"epoch": 1.2113821138211383,
"grad_norm": 3.7010541156424708,
"learning_rate": 4.046342642073433e-06,
"loss": 0.9724,
"step": 298
},
{
"epoch": 1.2154471544715446,
"grad_norm": 3.5911225035162815,
"learning_rate": 4.01148105089414e-06,
"loss": 1.0239,
"step": 299
},
{
"epoch": 1.2195121951219512,
"grad_norm": 3.521662940956276,
"learning_rate": 3.976669398546451e-06,
"loss": 1.0109,
"step": 300
},
{
"epoch": 1.2235772357723578,
"grad_norm": 3.855644275693675,
"learning_rate": 3.941909443674696e-06,
"loss": 1.0498,
"step": 301
},
{
"epoch": 1.2276422764227641,
"grad_norm": 3.4599309481583935,
"learning_rate": 3.907202942311506e-06,
"loss": 0.9521,
"step": 302
},
{
"epoch": 1.2317073170731707,
"grad_norm": 3.491476989084066,
"learning_rate": 3.872551647789108e-06,
"loss": 0.9474,
"step": 303
},
{
"epoch": 1.2357723577235773,
"grad_norm": 32.262705414195246,
"learning_rate": 3.837957310650738e-06,
"loss": 0.906,
"step": 304
},
{
"epoch": 1.2398373983739837,
"grad_norm": 3.84250057643824,
"learning_rate": 3.803421678562213e-06,
"loss": 0.9264,
"step": 305
},
{
"epoch": 1.2439024390243902,
"grad_norm": 3.1494724710918827,
"learning_rate": 3.7689464962236367e-06,
"loss": 0.8365,
"step": 306
},
{
"epoch": 1.2479674796747968,
"grad_norm": 3.3982902219496354,
"learning_rate": 3.734533505281269e-06,
"loss": 0.9983,
"step": 307
},
{
"epoch": 1.2520325203252032,
"grad_norm": 3.893213160762182,
"learning_rate": 3.700184444239524e-06,
"loss": 0.9916,
"step": 308
},
{
"epoch": 1.2560975609756098,
"grad_norm": 4.084636899332558,
"learning_rate": 3.6659010483731543e-06,
"loss": 0.9333,
"step": 309
},
{
"epoch": 1.2601626016260163,
"grad_norm": 3.5547285795900025,
"learning_rate": 3.6316850496395863e-06,
"loss": 1.0306,
"step": 310
},
{
"epoch": 1.2642276422764227,
"grad_norm": 4.387856612059098,
"learning_rate": 3.597538176591417e-06,
"loss": 1.0213,
"step": 311
},
{
"epoch": 1.2682926829268293,
"grad_norm": 3.7034317042973757,
"learning_rate": 3.563462154289098e-06,
"loss": 1.0348,
"step": 312
},
{
"epoch": 1.2723577235772359,
"grad_norm": 3.6811952215617447,
"learning_rate": 3.5294587042137796e-06,
"loss": 1.058,
"step": 313
},
{
"epoch": 1.2764227642276422,
"grad_norm": 3.1183018069231636,
"learning_rate": 3.49552954418035e-06,
"loss": 0.8768,
"step": 314
},
{
"epoch": 1.2804878048780488,
"grad_norm": 3.444034542353001,
"learning_rate": 3.461676388250651e-06,
"loss": 0.9825,
"step": 315
},
{
"epoch": 1.2845528455284554,
"grad_norm": 3.7392115265278165,
"learning_rate": 3.4279009466468825e-06,
"loss": 0.8815,
"step": 316
},
{
"epoch": 1.2886178861788617,
"grad_norm": 4.182871042947326,
"learning_rate": 3.3942049256652093e-06,
"loss": 0.9786,
"step": 317
},
{
"epoch": 1.2926829268292683,
"grad_norm": 3.5630464583189823,
"learning_rate": 3.3605900275895565e-06,
"loss": 1.008,
"step": 318
},
{
"epoch": 1.296747967479675,
"grad_norm": 3.8298454034397715,
"learning_rate": 3.3270579506056146e-06,
"loss": 1.0363,
"step": 319
},
{
"epoch": 1.3008130081300813,
"grad_norm": 3.8540799671085284,
"learning_rate": 3.2936103887150484e-06,
"loss": 0.9924,
"step": 320
},
{
"epoch": 1.3048780487804879,
"grad_norm": 3.7964601094131143,
"learning_rate": 3.2602490316499197e-06,
"loss": 1.0857,
"step": 321
},
{
"epoch": 1.3089430894308944,
"grad_norm": 3.6666310225840664,
"learning_rate": 3.226975564787322e-06,
"loss": 1.0222,
"step": 322
},
{
"epoch": 1.3130081300813008,
"grad_norm": 4.010214555221014,
"learning_rate": 3.1937916690642356e-06,
"loss": 1.085,
"step": 323
},
{
"epoch": 1.3170731707317074,
"grad_norm": 3.7791909242435544,
"learning_rate": 3.1606990208926125e-06,
"loss": 0.9724,
"step": 324
},
{
"epoch": 1.321138211382114,
"grad_norm": 3.064503593689533,
"learning_rate": 3.127699292074683e-06,
"loss": 0.9301,
"step": 325
},
{
"epoch": 1.3252032520325203,
"grad_norm": 3.9981146911437273,
"learning_rate": 3.0947941497184985e-06,
"loss": 1.1049,
"step": 326
},
{
"epoch": 1.329268292682927,
"grad_norm": 3.792220384973029,
"learning_rate": 3.0619852561537165e-06,
"loss": 1.0321,
"step": 327
},
{
"epoch": 1.3333333333333333,
"grad_norm": 4.289296789774771,
"learning_rate": 3.0292742688476125e-06,
"loss": 0.9311,
"step": 328
},
{
"epoch": 1.3373983739837398,
"grad_norm": 3.6446929722740666,
"learning_rate": 2.9966628403213528e-06,
"loss": 1.025,
"step": 329
},
{
"epoch": 1.3414634146341464,
"grad_norm": 4.490033319208552,
"learning_rate": 2.964152618066508e-06,
"loss": 1.0568,
"step": 330
},
{
"epoch": 1.3455284552845528,
"grad_norm": 3.450965579472706,
"learning_rate": 2.93174524446183e-06,
"loss": 0.8784,
"step": 331
},
{
"epoch": 1.3495934959349594,
"grad_norm": 4.161559788405042,
"learning_rate": 2.899442356690271e-06,
"loss": 0.9934,
"step": 332
},
{
"epoch": 1.3536585365853657,
"grad_norm": 4.526731975106362,
"learning_rate": 2.8672455866562797e-06,
"loss": 1.0624,
"step": 333
},
{
"epoch": 1.3577235772357723,
"grad_norm": 4.128986988767476,
"learning_rate": 2.835156560903365e-06,
"loss": 0.9091,
"step": 334
},
{
"epoch": 1.3617886178861789,
"grad_norm": 4.116808081237394,
"learning_rate": 2.803176900531915e-06,
"loss": 1.0205,
"step": 335
},
{
"epoch": 1.3658536585365852,
"grad_norm": 3.5446252267366765,
"learning_rate": 2.771308221117309e-06,
"loss": 0.9307,
"step": 336
},
{
"epoch": 1.3699186991869918,
"grad_norm": 3.9548465167028586,
"learning_rate": 2.7395521326282913e-06,
"loss": 1.0483,
"step": 337
},
{
"epoch": 1.3739837398373984,
"grad_norm": 3.5472423921579748,
"learning_rate": 2.7079102393456503e-06,
"loss": 0.9656,
"step": 338
},
{
"epoch": 1.3780487804878048,
"grad_norm": 3.853516813124923,
"learning_rate": 2.6763841397811576e-06,
"loss": 0.9221,
"step": 339
},
{
"epoch": 1.3821138211382114,
"grad_norm": 4.060475171502202,
"learning_rate": 2.6449754265968263e-06,
"loss": 1.0565,
"step": 340
},
{
"epoch": 1.386178861788618,
"grad_norm": 3.0492356938168377,
"learning_rate": 2.6136856865244443e-06,
"loss": 0.8312,
"step": 341
},
{
"epoch": 1.3902439024390243,
"grad_norm": 3.9578106322634974,
"learning_rate": 2.5825165002854124e-06,
"loss": 1.0317,
"step": 342
},
{
"epoch": 1.3943089430894309,
"grad_norm": 4.353093806474984,
"learning_rate": 2.5514694425108968e-06,
"loss": 1.1481,
"step": 343
},
{
"epoch": 1.3983739837398375,
"grad_norm": 3.513880546700832,
"learning_rate": 2.5205460816622684e-06,
"loss": 0.9368,
"step": 344
},
{
"epoch": 1.4024390243902438,
"grad_norm": 3.8414653794228872,
"learning_rate": 2.4897479799518797e-06,
"loss": 1.0788,
"step": 345
},
{
"epoch": 1.4065040650406504,
"grad_norm": 46.63836467878444,
"learning_rate": 2.4590766932641353e-06,
"loss": 0.9008,
"step": 346
},
{
"epoch": 1.410569105691057,
"grad_norm": 4.45225089824771,
"learning_rate": 2.4285337710768843e-06,
"loss": 0.9595,
"step": 347
},
{
"epoch": 1.4146341463414633,
"grad_norm": 3.3936120852821734,
"learning_rate": 2.3981207563831633e-06,
"loss": 0.956,
"step": 348
},
{
"epoch": 1.41869918699187,
"grad_norm": 4.694796697002138,
"learning_rate": 2.3678391856132203e-06,
"loss": 0.9804,
"step": 349
},
{
"epoch": 1.4227642276422765,
"grad_norm": 3.9039345975261712,
"learning_rate": 2.3376905885569185e-06,
"loss": 1.1112,
"step": 350
},
{
"epoch": 1.4268292682926829,
"grad_norm": 4.375882192486315,
"learning_rate": 2.3076764882864333e-06,
"loss": 0.9284,
"step": 351
},
{
"epoch": 1.4308943089430894,
"grad_norm": 4.504132135708646,
"learning_rate": 2.2777984010793264e-06,
"loss": 1.125,
"step": 352
},
{
"epoch": 1.434959349593496,
"grad_norm": 4.182412088395119,
"learning_rate": 2.2480578363419363e-06,
"loss": 1.1081,
"step": 353
},
{
"epoch": 1.4390243902439024,
"grad_norm": 4.585452168929015,
"learning_rate": 2.2184562965331203e-06,
"loss": 1.0475,
"step": 354
},
{
"epoch": 1.443089430894309,
"grad_norm": 4.0885701813647515,
"learning_rate": 2.1889952770883644e-06,
"loss": 1.124,
"step": 355
},
{
"epoch": 1.4471544715447155,
"grad_norm": 3.5056911821730226,
"learning_rate": 2.159676266344222e-06,
"loss": 0.9659,
"step": 356
},
{
"epoch": 1.451219512195122,
"grad_norm": 3.9979580432930684,
"learning_rate": 2.130500745463136e-06,
"loss": 0.994,
"step": 357
},
{
"epoch": 1.4552845528455285,
"grad_norm": 4.039513265865325,
"learning_rate": 2.1014701883586087e-06,
"loss": 1.0819,
"step": 358
},
{
"epoch": 1.459349593495935,
"grad_norm": 3.4944639246235147,
"learning_rate": 2.072586061620735e-06,
"loss": 0.9812,
"step": 359
},
{
"epoch": 1.4634146341463414,
"grad_norm": 4.01097044265515,
"learning_rate": 2.043849824442124e-06,
"loss": 0.9238,
"step": 360
},
{
"epoch": 1.467479674796748,
"grad_norm": 3.8206482337293397,
"learning_rate": 2.0152629285441668e-06,
"loss": 1.035,
"step": 361
},
{
"epoch": 1.4715447154471546,
"grad_norm": 3.2549782865289236,
"learning_rate": 1.9868268181037186e-06,
"loss": 0.8686,
"step": 362
},
{
"epoch": 1.475609756097561,
"grad_norm": 4.213188910356952,
"learning_rate": 1.958542929680117e-06,
"loss": 0.9639,
"step": 363
},
{
"epoch": 1.4796747967479675,
"grad_norm": 3.9202184083066087,
"learning_rate": 1.9304126921426235e-06,
"loss": 1.0089,
"step": 364
},
{
"epoch": 1.4837398373983741,
"grad_norm": 3.900579324779219,
"learning_rate": 1.9024375265982386e-06,
"loss": 0.8801,
"step": 365
},
{
"epoch": 1.4878048780487805,
"grad_norm": 3.7497758212254215,
"learning_rate": 1.8746188463198983e-06,
"loss": 1.0532,
"step": 366
},
{
"epoch": 1.491869918699187,
"grad_norm": 3.316320439440319,
"learning_rate": 1.8469580566750911e-06,
"loss": 0.9393,
"step": 367
},
{
"epoch": 1.4959349593495934,
"grad_norm": 3.6503770546307432,
"learning_rate": 1.8194565550548477e-06,
"loss": 0.9727,
"step": 368
},
{
"epoch": 1.5,
"grad_norm": 4.359109037614491,
"learning_rate": 1.7921157308031567e-06,
"loss": 1.0279,
"step": 369
},
{
"epoch": 1.5040650406504064,
"grad_norm": 2.882183049368505,
"learning_rate": 1.764936965146773e-06,
"loss": 0.8451,
"step": 370
},
{
"epoch": 1.5081300813008132,
"grad_norm": 4.457745491095926,
"learning_rate": 1.7379216311254339e-06,
"loss": 1.0752,
"step": 371
},
{
"epoch": 1.5121951219512195,
"grad_norm": 3.3162867863381913,
"learning_rate": 1.7110710935225055e-06,
"loss": 0.952,
"step": 372
},
{
"epoch": 1.5162601626016259,
"grad_norm": 4.003446669515271,
"learning_rate": 1.6843867087960252e-06,
"loss": 1.0375,
"step": 373
},
{
"epoch": 1.5203252032520327,
"grad_norm": 3.6017712032571256,
"learning_rate": 1.6578698250101828e-06,
"loss": 0.9802,
"step": 374
},
{
"epoch": 1.524390243902439,
"grad_norm": 4.140971003588729,
"learning_rate": 1.6315217817672142e-06,
"loss": 1.067,
"step": 375
},
{
"epoch": 1.5284552845528454,
"grad_norm": 4.0705004013893875,
"learning_rate": 1.6053439101397257e-06,
"loss": 0.915,
"step": 376
},
{
"epoch": 1.532520325203252,
"grad_norm": 4.1303831727413645,
"learning_rate": 1.5793375326034539e-06,
"loss": 0.9664,
"step": 377
},
{
"epoch": 1.5365853658536586,
"grad_norm": 3.7129753807072143,
"learning_rate": 1.5535039629704467e-06,
"loss": 1.064,
"step": 378
},
{
"epoch": 1.540650406504065,
"grad_norm": 3.5034854214257183,
"learning_rate": 1.5278445063227038e-06,
"loss": 1.0013,
"step": 379
},
{
"epoch": 1.5447154471544715,
"grad_norm": 3.1215335264280286,
"learning_rate": 1.502360458946232e-06,
"loss": 0.9222,
"step": 380
},
{
"epoch": 1.548780487804878,
"grad_norm": 3.955236224458765,
"learning_rate": 1.4770531082655704e-06,
"loss": 0.9712,
"step": 381
},
{
"epoch": 1.5528455284552845,
"grad_norm": 4.198292135711804,
"learning_rate": 1.451923732778745e-06,
"loss": 0.8991,
"step": 382
},
{
"epoch": 1.556910569105691,
"grad_norm": 3.102574689879648,
"learning_rate": 1.4269736019926778e-06,
"loss": 1.0298,
"step": 383
},
{
"epoch": 1.5609756097560976,
"grad_norm": 3.372846780482694,
"learning_rate": 1.4022039763590595e-06,
"loss": 0.8567,
"step": 384
},
{
"epoch": 1.565040650406504,
"grad_norm": 3.688513986136554,
"learning_rate": 1.3776161072106703e-06,
"loss": 0.9762,
"step": 385
},
{
"epoch": 1.5691056910569106,
"grad_norm": 3.694767191664073,
"learning_rate": 1.3532112366981598e-06,
"loss": 0.9979,
"step": 386
},
{
"epoch": 1.5731707317073171,
"grad_norm": 4.0857246463701955,
"learning_rate": 1.3289905977273027e-06,
"loss": 0.9511,
"step": 387
},
{
"epoch": 1.5772357723577235,
"grad_norm": 3.53499666737931,
"learning_rate": 1.3049554138967052e-06,
"loss": 0.8967,
"step": 388
},
{
"epoch": 1.58130081300813,
"grad_norm": 3.5347496328031465,
"learning_rate": 1.2811068994359992e-06,
"loss": 0.7356,
"step": 389
},
{
"epoch": 1.5853658536585367,
"grad_norm": 4.381582039757114,
"learning_rate": 1.257446259144494e-06,
"loss": 1.0999,
"step": 390
},
{
"epoch": 1.589430894308943,
"grad_norm": 3.7466328104969278,
"learning_rate": 1.233974688330315e-06,
"loss": 0.9674,
"step": 391
},
{
"epoch": 1.5934959349593496,
"grad_norm": 4.277062923168833,
"learning_rate": 1.210693372750017e-06,
"loss": 0.9929,
"step": 392
},
{
"epoch": 1.5975609756097562,
"grad_norm": 3.8305888215053754,
"learning_rate": 1.1876034885486764e-06,
"loss": 0.9858,
"step": 393
},
{
"epoch": 1.6016260162601625,
"grad_norm": 3.4914255971094876,
"learning_rate": 1.1647062022004845e-06,
"loss": 0.9851,
"step": 394
},
{
"epoch": 1.6056910569105691,
"grad_norm": 3.9890311938988643,
"learning_rate": 1.1420026704498077e-06,
"loss": 0.8826,
"step": 395
},
{
"epoch": 1.6097560975609757,
"grad_norm": 3.6808318567839367,
"learning_rate": 1.1194940402527566e-06,
"loss": 1.0323,
"step": 396
},
{
"epoch": 1.613821138211382,
"grad_norm": 3.7719643553355926,
"learning_rate": 1.0971814487192429e-06,
"loss": 0.9259,
"step": 397
},
{
"epoch": 1.6178861788617886,
"grad_norm": 4.179966730043697,
"learning_rate": 1.075066023055527e-06,
"loss": 1.0841,
"step": 398
},
{
"epoch": 1.6219512195121952,
"grad_norm": 3.4390515759083544,
"learning_rate": 1.0531488805072848e-06,
"loss": 0.9617,
"step": 399
},
{
"epoch": 1.6260162601626016,
"grad_norm": 3.433401797605647,
"learning_rate": 1.0314311283031531e-06,
"loss": 0.9369,
"step": 400
},
{
"epoch": 1.6300813008130082,
"grad_norm": 3.9442267675520313,
"learning_rate": 1.0099138635988026e-06,
"loss": 1.0009,
"step": 401
},
{
"epoch": 1.6341463414634148,
"grad_norm": 4.142825713403116,
"learning_rate": 9.885981734215094e-07,
"loss": 0.9275,
"step": 402
},
{
"epoch": 1.6382113821138211,
"grad_norm": 4.202690423258989,
"learning_rate": 9.67485134615232e-07,
"loss": 1.05,
"step": 403
},
{
"epoch": 1.6422764227642277,
"grad_norm": 3.839411663010975,
"learning_rate": 9.465758137862264e-07,
"loss": 0.9513,
"step": 404
},
{
"epoch": 1.6463414634146343,
"grad_norm": 3.672326700992681,
"learning_rate": 9.258712672491416e-07,
"loss": 1.024,
"step": 405
},
{
"epoch": 1.6504065040650406,
"grad_norm": 4.073445323984085,
"learning_rate": 9.053725409736752e-07,
"loss": 0.9947,
"step": 406
},
{
"epoch": 1.654471544715447,
"grad_norm": 4.010315161100075,
"learning_rate": 8.850806705317183e-07,
"loss": 1.0415,
"step": 407
},
{
"epoch": 1.6585365853658538,
"grad_norm": 3.8293127353729015,
"learning_rate": 8.649966810450472e-07,
"loss": 0.9706,
"step": 408
},
{
"epoch": 1.6626016260162602,
"grad_norm": 4.28837815719892,
"learning_rate": 8.451215871335355e-07,
"loss": 1.0411,
"step": 409
},
{
"epoch": 1.6666666666666665,
"grad_norm": 4.0132337255657555,
"learning_rate": 8.254563928638892e-07,
"loss": 1.1221,
"step": 410
},
{
"epoch": 1.6707317073170733,
"grad_norm": 3.7377426068899564,
"learning_rate": 8.060020916989331e-07,
"loss": 1.0049,
"step": 411
},
{
"epoch": 1.6747967479674797,
"grad_norm": 3.4438718525721255,
"learning_rate": 7.86759666447412e-07,
"loss": 0.9957,
"step": 412
},
{
"epoch": 1.678861788617886,
"grad_norm": 3.6205519453520956,
"learning_rate": 7.677300892143485e-07,
"loss": 0.9857,
"step": 413
},
{
"epoch": 1.6829268292682928,
"grad_norm": 3.8982585731415083,
"learning_rate": 7.489143213519301e-07,
"loss": 0.9346,
"step": 414
},
{
"epoch": 1.6869918699186992,
"grad_norm": 4.142796108192852,
"learning_rate": 7.303133134109391e-07,
"loss": 1.0084,
"step": 415
},
{
"epoch": 1.6910569105691056,
"grad_norm": 4.225439613721668,
"learning_rate": 7.119280050927407e-07,
"loss": 0.9657,
"step": 416
},
{
"epoch": 1.6951219512195121,
"grad_norm": 3.3574575459381895,
"learning_rate": 6.937593252017983e-07,
"loss": 0.9819,
"step": 417
},
{
"epoch": 1.6991869918699187,
"grad_norm": 4.914841000477497,
"learning_rate": 6.758081915987669e-07,
"loss": 0.8859,
"step": 418
},
{
"epoch": 1.703252032520325,
"grad_norm": 3.781768313712973,
"learning_rate": 6.580755111541076e-07,
"loss": 1.0215,
"step": 419
},
{
"epoch": 1.7073170731707317,
"grad_norm": 3.7238954715457537,
"learning_rate": 6.405621797022848e-07,
"loss": 1.0156,
"step": 420
},
{
"epoch": 1.7113821138211383,
"grad_norm": 3.553159127802151,
"learning_rate": 6.232690819965065e-07,
"loss": 1.1136,
"step": 421
},
{
"epoch": 1.7154471544715446,
"grad_norm": 3.612323933609705,
"learning_rate": 6.061970916640236e-07,
"loss": 0.8911,
"step": 422
},
{
"epoch": 1.7195121951219512,
"grad_norm": 3.587845262671792,
"learning_rate": 5.893470711620036e-07,
"loss": 0.9443,
"step": 423
},
{
"epoch": 1.7235772357723578,
"grad_norm": 3.5104402409616715,
"learning_rate": 5.727198717339511e-07,
"loss": 0.9253,
"step": 424
},
{
"epoch": 1.7276422764227641,
"grad_norm": 3.114895367636644,
"learning_rate": 5.563163333667098e-07,
"loss": 0.9985,
"step": 425
},
{
"epoch": 1.7317073170731707,
"grad_norm": 4.204476593228392,
"learning_rate": 5.401372847480285e-07,
"loss": 1.1147,
"step": 426
},
{
"epoch": 1.7357723577235773,
"grad_norm": 5.856252832998231,
"learning_rate": 5.241835432246888e-07,
"loss": 0.7672,
"step": 427
},
{
"epoch": 1.7398373983739837,
"grad_norm": 4.584343185417779,
"learning_rate": 5.084559147612244e-07,
"loss": 1.0433,
"step": 428
},
{
"epoch": 1.7439024390243902,
"grad_norm": 3.5601385211052863,
"learning_rate": 4.929551938991945e-07,
"loss": 0.9034,
"step": 429
},
{
"epoch": 1.7479674796747968,
"grad_norm": 3.7438242901268834,
"learning_rate": 4.776821637170525e-07,
"loss": 0.9871,
"step": 430
},
{
"epoch": 1.7520325203252032,
"grad_norm": 3.726788471257809,
"learning_rate": 4.626375957905821e-07,
"loss": 1.0139,
"step": 431
},
{
"epoch": 1.7560975609756098,
"grad_norm": 3.500194615705022,
"learning_rate": 4.4782225015391754e-07,
"loss": 0.9387,
"step": 432
},
{
"epoch": 1.7601626016260163,
"grad_norm": 4.036120727657588,
"learning_rate": 4.3323687526115045e-07,
"loss": 1.0591,
"step": 433
},
{
"epoch": 1.7642276422764227,
"grad_norm": 3.7179550108214308,
"learning_rate": 4.1888220794851386e-07,
"loss": 0.9231,
"step": 434
},
{
"epoch": 1.7682926829268293,
"grad_norm": 4.116011774790538,
"learning_rate": 4.0475897339716466e-07,
"loss": 1.096,
"step": 435
},
{
"epoch": 1.7723577235772359,
"grad_norm": 3.4904100752060825,
"learning_rate": 3.908678850965425e-07,
"loss": 1.0487,
"step": 436
},
{
"epoch": 1.7764227642276422,
"grad_norm": 3.4213015383298506,
"learning_rate": 3.7720964480832847e-07,
"loss": 1.0293,
"step": 437
},
{
"epoch": 1.7804878048780488,
"grad_norm": 3.6796253516773088,
"learning_rate": 3.6378494253099307e-07,
"loss": 1.019,
"step": 438
},
{
"epoch": 1.7845528455284554,
"grad_norm": 3.4820509030406415,
"learning_rate": 3.505944564649344e-07,
"loss": 1.0071,
"step": 439
},
{
"epoch": 1.7886178861788617,
"grad_norm": 4.178141072634524,
"learning_rate": 3.3763885297822153e-07,
"loss": 1.0619,
"step": 440
},
{
"epoch": 1.7926829268292683,
"grad_norm": 4.236971625953427,
"learning_rate": 3.2491878657292643e-07,
"loss": 1.0626,
"step": 441
},
{
"epoch": 1.796747967479675,
"grad_norm": 3.6519879305302227,
"learning_rate": 3.1243489985206097e-07,
"loss": 0.8886,
"step": 442
},
{
"epoch": 1.8008130081300813,
"grad_norm": 3.586686228856645,
"learning_rate": 3.001878234871147e-07,
"loss": 0.9034,
"step": 443
},
{
"epoch": 1.8048780487804879,
"grad_norm": 4.238214426688419,
"learning_rate": 2.8817817618618846e-07,
"loss": 0.9862,
"step": 444
},
{
"epoch": 1.8089430894308944,
"grad_norm": 3.490951187385756,
"learning_rate": 2.7640656466274785e-07,
"loss": 0.9977,
"step": 445
},
{
"epoch": 1.8130081300813008,
"grad_norm": 3.6485834724251394,
"learning_rate": 2.648735836049615e-07,
"loss": 0.9395,
"step": 446
},
{
"epoch": 1.8170731707317072,
"grad_norm": 3.3220985022269667,
"learning_rate": 2.5357981564566647e-07,
"loss": 0.8872,
"step": 447
},
{
"epoch": 1.821138211382114,
"grad_norm": 4.123832064865542,
"learning_rate": 2.4252583133292927e-07,
"loss": 1.0679,
"step": 448
},
{
"epoch": 1.8252032520325203,
"grad_norm": 3.492052368274594,
"learning_rate": 2.3171218910122695e-07,
"loss": 0.879,
"step": 449
},
{
"epoch": 1.8292682926829267,
"grad_norm": 4.036773726834657,
"learning_rate": 2.2113943524323167e-07,
"loss": 0.9786,
"step": 450
},
{
"epoch": 1.8333333333333335,
"grad_norm": 4.136219271416021,
"learning_rate": 2.1080810388221406e-07,
"loss": 1.1111,
"step": 451
},
{
"epoch": 1.8373983739837398,
"grad_norm": 3.8933772306252967,
"learning_rate": 2.007187169450603e-07,
"loss": 0.9266,
"step": 452
},
{
"epoch": 1.8414634146341462,
"grad_norm": 3.4850123477897865,
"learning_rate": 1.908717841359048e-07,
"loss": 0.8836,
"step": 453
},
{
"epoch": 1.845528455284553,
"grad_norm": 3.824450626647066,
"learning_rate": 1.8126780291038037e-07,
"loss": 0.9532,
"step": 454
},
{
"epoch": 1.8495934959349594,
"grad_norm": 4.340689760572588,
"learning_rate": 1.7190725845048827e-07,
"loss": 0.951,
"step": 455
},
{
"epoch": 1.8536585365853657,
"grad_norm": 3.6380391806167287,
"learning_rate": 1.6279062364008446e-07,
"loss": 1.0206,
"step": 456
},
{
"epoch": 1.8577235772357723,
"grad_norm": 3.483222380757324,
"learning_rate": 1.5391835904099316e-07,
"loss": 0.9323,
"step": 457
},
{
"epoch": 1.8617886178861789,
"grad_norm": 3.83318933725723,
"learning_rate": 1.4529091286973994e-07,
"loss": 0.9291,
"step": 458
},
{
"epoch": 1.8658536585365852,
"grad_norm": 5.25876322386921,
"learning_rate": 1.3690872097490481e-07,
"loss": 1.0843,
"step": 459
},
{
"epoch": 1.8699186991869918,
"grad_norm": 3.8163606429748405,
"learning_rate": 1.2877220681510927e-07,
"loss": 1.1322,
"step": 460
},
{
"epoch": 1.8739837398373984,
"grad_norm": 3.745271876845421,
"learning_rate": 1.208817814376162e-07,
"loss": 0.9812,
"step": 461
},
{
"epoch": 1.8780487804878048,
"grad_norm": 3.938076078503757,
"learning_rate": 1.1323784345757205e-07,
"loss": 1.0328,
"step": 462
},
{
"epoch": 1.8821138211382114,
"grad_norm": 3.8600868786911193,
"learning_rate": 1.0584077903786238e-07,
"loss": 1.0218,
"step": 463
},
{
"epoch": 1.886178861788618,
"grad_norm": 4.079832178269913,
"learning_rate": 9.869096186961025e-08,
"loss": 1.0443,
"step": 464
},
{
"epoch": 1.8902439024390243,
"grad_norm": 3.68360826380147,
"learning_rate": 9.178875315329183e-08,
"loss": 0.9284,
"step": 465
},
{
"epoch": 1.8943089430894309,
"grad_norm": 3.6046022549114705,
"learning_rate": 8.513450158049109e-08,
"loss": 0.9411,
"step": 466
},
{
"epoch": 1.8983739837398375,
"grad_norm": 4.024583530140874,
"learning_rate": 7.872854331628599e-08,
"loss": 1.1105,
"step": 467
},
{
"epoch": 1.9024390243902438,
"grad_norm": 4.15926484895413,
"learning_rate": 7.257120198226219e-08,
"loss": 1.066,
"step": 468
},
{
"epoch": 1.9065040650406504,
"grad_norm": 3.7639785430318566,
"learning_rate": 6.666278864016884e-08,
"loss": 1.059,
"step": 469
},
{
"epoch": 1.910569105691057,
"grad_norm": 3.953800325677805,
"learning_rate": 6.100360177619946e-08,
"loss": 1.0305,
"step": 470
},
{
"epoch": 1.9146341463414633,
"grad_norm": 3.4067262115402164,
"learning_rate": 5.559392728591517e-08,
"loss": 1.0381,
"step": 471
},
{
"epoch": 1.91869918699187,
"grad_norm": 3.140993551988303,
"learning_rate": 5.0434038459801213e-08,
"loss": 0.7932,
"step": 472
},
{
"epoch": 1.9227642276422765,
"grad_norm": 3.5873575446175754,
"learning_rate": 4.5524195969461895e-08,
"loss": 0.9524,
"step": 473
},
{
"epoch": 1.9268292682926829,
"grad_norm": 4.698767984892136,
"learning_rate": 4.086464785444777e-08,
"loss": 0.8371,
"step": 474
},
{
"epoch": 1.9308943089430894,
"grad_norm": 3.7978393128890997,
"learning_rate": 3.645562950973014e-08,
"loss": 0.9666,
"step": 475
},
{
"epoch": 1.934959349593496,
"grad_norm": 3.5823042491668717,
"learning_rate": 3.229736367380498e-08,
"loss": 0.9649,
"step": 476
},
{
"epoch": 1.9390243902439024,
"grad_norm": 3.981781985201543,
"learning_rate": 2.8390060417443632e-08,
"loss": 0.9881,
"step": 477
},
{
"epoch": 1.943089430894309,
"grad_norm": 4.065041283971589,
"learning_rate": 2.4733917133077378e-08,
"loss": 1.0041,
"step": 478
},
{
"epoch": 1.9471544715447155,
"grad_norm": 4.2020322856448225,
"learning_rate": 2.1329118524827662e-08,
"loss": 0.95,
"step": 479
},
{
"epoch": 1.951219512195122,
"grad_norm": 4.245506598983614,
"learning_rate": 1.8175836599173545e-08,
"loss": 1.1151,
"step": 480
},
{
"epoch": 1.9552845528455285,
"grad_norm": 3.4124284954497606,
"learning_rate": 1.5274230656263656e-08,
"loss": 1.0035,
"step": 481
},
{
"epoch": 1.959349593495935,
"grad_norm": 4.074743275699229,
"learning_rate": 1.2624447281867625e-08,
"loss": 1.0419,
"step": 482
},
{
"epoch": 1.9634146341463414,
"grad_norm": 3.783549511402003,
"learning_rate": 1.0226620339969795e-08,
"loss": 0.998,
"step": 483
},
{
"epoch": 1.967479674796748,
"grad_norm": 3.929853822852935,
"learning_rate": 8.080870966008513e-09,
"loss": 0.954,
"step": 484
},
{
"epoch": 1.9715447154471546,
"grad_norm": 3.3589007769177326,
"learning_rate": 6.187307560754363e-09,
"loss": 0.9721,
"step": 485
},
{
"epoch": 1.975609756097561,
"grad_norm": 3.3788764384929824,
"learning_rate": 4.546025784837316e-09,
"loss": 0.9284,
"step": 486
},
{
"epoch": 1.9796747967479673,
"grad_norm": 4.091963781011746,
"learning_rate": 3.1571085539089384e-09,
"loss": 1.0655,
"step": 487
},
{
"epoch": 1.9837398373983741,
"grad_norm": 3.776958414812629,
"learning_rate": 2.0206260344590724e-09,
"loss": 1.005,
"step": 488
},
{
"epoch": 1.9878048780487805,
"grad_norm": 4.375338523817764,
"learning_rate": 1.136635640267003e-09,
"loss": 0.9906,
"step": 489
},
{
"epoch": 1.9918699186991868,
"grad_norm": 3.86863750217483,
"learning_rate": 5.051820295032262e-10,
"loss": 0.9767,
"step": 490
},
{
"epoch": 1.9959349593495936,
"grad_norm": 3.6645935176814772,
"learning_rate": 1.2629710247180626e-10,
"loss": 1.0582,
"step": 491
},
{
"epoch": 2.0,
"grad_norm": 3.894299701332893,
"learning_rate": 0.0,
"loss": 1.0572,
"step": 492
},
{
"epoch": 2.0,
"step": 492,
"total_flos": 8984667488256.0,
"train_loss": 1.2601688853124293,
"train_runtime": 414.9647,
"train_samples_per_second": 37.921,
"train_steps_per_second": 1.186
}
],
"logging_steps": 1,
"max_steps": 492,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 50000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 8984667488256.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}