sedrickkeh's picture
End of training
7880f90 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.995695839311334,
"eval_steps": 500,
"global_step": 696,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00430416068866571,
"grad_norm": 5.716188907623291,
"learning_rate": 1.4285714285714287e-07,
"loss": 0.7934,
"step": 1
},
{
"epoch": 0.00860832137733142,
"grad_norm": 5.855449199676514,
"learning_rate": 2.8571428571428575e-07,
"loss": 0.8156,
"step": 2
},
{
"epoch": 0.01291248206599713,
"grad_norm": 5.917946815490723,
"learning_rate": 4.285714285714286e-07,
"loss": 0.8216,
"step": 3
},
{
"epoch": 0.01721664275466284,
"grad_norm": 6.008902072906494,
"learning_rate": 5.714285714285715e-07,
"loss": 0.8243,
"step": 4
},
{
"epoch": 0.021520803443328552,
"grad_norm": 6.13193941116333,
"learning_rate": 7.142857142857143e-07,
"loss": 0.8402,
"step": 5
},
{
"epoch": 0.02582496413199426,
"grad_norm": 5.711304664611816,
"learning_rate": 8.571428571428572e-07,
"loss": 0.7946,
"step": 6
},
{
"epoch": 0.03012912482065997,
"grad_norm": 5.6730499267578125,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.8204,
"step": 7
},
{
"epoch": 0.03443328550932568,
"grad_norm": 5.452306747436523,
"learning_rate": 1.142857142857143e-06,
"loss": 0.8125,
"step": 8
},
{
"epoch": 0.03873744619799139,
"grad_norm": 4.5766520500183105,
"learning_rate": 1.2857142857142856e-06,
"loss": 0.7809,
"step": 9
},
{
"epoch": 0.043041606886657105,
"grad_norm": 4.3750481605529785,
"learning_rate": 1.4285714285714286e-06,
"loss": 0.7808,
"step": 10
},
{
"epoch": 0.047345767575322814,
"grad_norm": 3.954547166824341,
"learning_rate": 1.5714285714285714e-06,
"loss": 0.7559,
"step": 11
},
{
"epoch": 0.05164992826398852,
"grad_norm": 2.556342363357544,
"learning_rate": 1.7142857142857145e-06,
"loss": 0.734,
"step": 12
},
{
"epoch": 0.05595408895265423,
"grad_norm": 2.303656578063965,
"learning_rate": 1.8571428571428573e-06,
"loss": 0.7309,
"step": 13
},
{
"epoch": 0.06025824964131994,
"grad_norm": 2.232937812805176,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.7365,
"step": 14
},
{
"epoch": 0.06456241032998565,
"grad_norm": 1.9870636463165283,
"learning_rate": 2.1428571428571427e-06,
"loss": 0.7036,
"step": 15
},
{
"epoch": 0.06886657101865136,
"grad_norm": 1.6483427286148071,
"learning_rate": 2.285714285714286e-06,
"loss": 0.7078,
"step": 16
},
{
"epoch": 0.07317073170731707,
"grad_norm": 2.942274808883667,
"learning_rate": 2.428571428571429e-06,
"loss": 0.695,
"step": 17
},
{
"epoch": 0.07747489239598278,
"grad_norm": 3.4096434116363525,
"learning_rate": 2.571428571428571e-06,
"loss": 0.69,
"step": 18
},
{
"epoch": 0.08177905308464849,
"grad_norm": 3.4464306831359863,
"learning_rate": 2.7142857142857144e-06,
"loss": 0.6966,
"step": 19
},
{
"epoch": 0.08608321377331421,
"grad_norm": 3.2228825092315674,
"learning_rate": 2.8571428571428573e-06,
"loss": 0.6711,
"step": 20
},
{
"epoch": 0.09038737446197992,
"grad_norm": 3.1530938148498535,
"learning_rate": 3e-06,
"loss": 0.7096,
"step": 21
},
{
"epoch": 0.09469153515064563,
"grad_norm": 2.6545050144195557,
"learning_rate": 3.142857142857143e-06,
"loss": 0.6677,
"step": 22
},
{
"epoch": 0.09899569583931134,
"grad_norm": 1.9846737384796143,
"learning_rate": 3.285714285714286e-06,
"loss": 0.7003,
"step": 23
},
{
"epoch": 0.10329985652797705,
"grad_norm": 1.4019083976745605,
"learning_rate": 3.428571428571429e-06,
"loss": 0.6554,
"step": 24
},
{
"epoch": 0.10760401721664276,
"grad_norm": 1.2153626680374146,
"learning_rate": 3.5714285714285718e-06,
"loss": 0.6168,
"step": 25
},
{
"epoch": 0.11190817790530846,
"grad_norm": 1.1019742488861084,
"learning_rate": 3.7142857142857146e-06,
"loss": 0.6308,
"step": 26
},
{
"epoch": 0.11621233859397417,
"grad_norm": 1.05548894405365,
"learning_rate": 3.857142857142858e-06,
"loss": 0.6174,
"step": 27
},
{
"epoch": 0.12051649928263988,
"grad_norm": 1.1632239818572998,
"learning_rate": 4.000000000000001e-06,
"loss": 0.6314,
"step": 28
},
{
"epoch": 0.12482065997130559,
"grad_norm": 1.0276652574539185,
"learning_rate": 4.1428571428571435e-06,
"loss": 0.5832,
"step": 29
},
{
"epoch": 0.1291248206599713,
"grad_norm": 0.8600188493728638,
"learning_rate": 4.2857142857142855e-06,
"loss": 0.6304,
"step": 30
},
{
"epoch": 0.133428981348637,
"grad_norm": 0.6312229037284851,
"learning_rate": 4.428571428571429e-06,
"loss": 0.6215,
"step": 31
},
{
"epoch": 0.13773314203730272,
"grad_norm": 0.6575469970703125,
"learning_rate": 4.571428571428572e-06,
"loss": 0.5946,
"step": 32
},
{
"epoch": 0.14203730272596843,
"grad_norm": 0.8034297227859497,
"learning_rate": 4.714285714285715e-06,
"loss": 0.5832,
"step": 33
},
{
"epoch": 0.14634146341463414,
"grad_norm": 0.8478543162345886,
"learning_rate": 4.857142857142858e-06,
"loss": 0.5804,
"step": 34
},
{
"epoch": 0.15064562410329985,
"grad_norm": 0.6925229430198669,
"learning_rate": 5e-06,
"loss": 0.5901,
"step": 35
},
{
"epoch": 0.15494978479196556,
"grad_norm": 0.637363076210022,
"learning_rate": 5.142857142857142e-06,
"loss": 0.6189,
"step": 36
},
{
"epoch": 0.15925394548063126,
"grad_norm": 0.6068949699401855,
"learning_rate": 5.285714285714286e-06,
"loss": 0.5708,
"step": 37
},
{
"epoch": 0.16355810616929697,
"grad_norm": 0.7009698748588562,
"learning_rate": 5.428571428571429e-06,
"loss": 0.5692,
"step": 38
},
{
"epoch": 0.1678622668579627,
"grad_norm": 0.5943602919578552,
"learning_rate": 5.571428571428572e-06,
"loss": 0.5631,
"step": 39
},
{
"epoch": 0.17216642754662842,
"grad_norm": 0.6478841304779053,
"learning_rate": 5.7142857142857145e-06,
"loss": 0.5581,
"step": 40
},
{
"epoch": 0.17647058823529413,
"grad_norm": 0.5670358538627625,
"learning_rate": 5.857142857142858e-06,
"loss": 0.5764,
"step": 41
},
{
"epoch": 0.18077474892395984,
"grad_norm": 0.47529032826423645,
"learning_rate": 6e-06,
"loss": 0.5248,
"step": 42
},
{
"epoch": 0.18507890961262555,
"grad_norm": 0.664316713809967,
"learning_rate": 6.142857142857144e-06,
"loss": 0.5609,
"step": 43
},
{
"epoch": 0.18938307030129126,
"grad_norm": 0.5606933236122131,
"learning_rate": 6.285714285714286e-06,
"loss": 0.5283,
"step": 44
},
{
"epoch": 0.19368723098995697,
"grad_norm": 0.5545839667320251,
"learning_rate": 6.4285714285714295e-06,
"loss": 0.5505,
"step": 45
},
{
"epoch": 0.19799139167862267,
"grad_norm": 0.5529095530509949,
"learning_rate": 6.571428571428572e-06,
"loss": 0.5599,
"step": 46
},
{
"epoch": 0.20229555236728838,
"grad_norm": 0.5661711096763611,
"learning_rate": 6.714285714285714e-06,
"loss": 0.5741,
"step": 47
},
{
"epoch": 0.2065997130559541,
"grad_norm": 0.535641610622406,
"learning_rate": 6.857142857142858e-06,
"loss": 0.5149,
"step": 48
},
{
"epoch": 0.2109038737446198,
"grad_norm": 0.5135163068771362,
"learning_rate": 7e-06,
"loss": 0.5407,
"step": 49
},
{
"epoch": 0.2152080344332855,
"grad_norm": 0.46351608633995056,
"learning_rate": 7.1428571428571436e-06,
"loss": 0.5325,
"step": 50
},
{
"epoch": 0.21951219512195122,
"grad_norm": 0.49139636754989624,
"learning_rate": 7.285714285714286e-06,
"loss": 0.5325,
"step": 51
},
{
"epoch": 0.22381635581061693,
"grad_norm": 0.4905931353569031,
"learning_rate": 7.428571428571429e-06,
"loss": 0.5468,
"step": 52
},
{
"epoch": 0.22812051649928264,
"grad_norm": 0.46065065264701843,
"learning_rate": 7.571428571428572e-06,
"loss": 0.5144,
"step": 53
},
{
"epoch": 0.23242467718794835,
"grad_norm": 0.42439892888069153,
"learning_rate": 7.714285714285716e-06,
"loss": 0.5478,
"step": 54
},
{
"epoch": 0.23672883787661406,
"grad_norm": 0.4570675194263458,
"learning_rate": 7.857142857142858e-06,
"loss": 0.5162,
"step": 55
},
{
"epoch": 0.24103299856527977,
"grad_norm": 0.36280250549316406,
"learning_rate": 8.000000000000001e-06,
"loss": 0.4896,
"step": 56
},
{
"epoch": 0.24533715925394547,
"grad_norm": 0.42444807291030884,
"learning_rate": 8.142857142857143e-06,
"loss": 0.5184,
"step": 57
},
{
"epoch": 0.24964131994261118,
"grad_norm": 0.45756882429122925,
"learning_rate": 8.285714285714287e-06,
"loss": 0.5381,
"step": 58
},
{
"epoch": 0.2539454806312769,
"grad_norm": 0.42676690220832825,
"learning_rate": 8.428571428571429e-06,
"loss": 0.5344,
"step": 59
},
{
"epoch": 0.2582496413199426,
"grad_norm": 0.44801363348960876,
"learning_rate": 8.571428571428571e-06,
"loss": 0.5177,
"step": 60
},
{
"epoch": 0.26255380200860834,
"grad_norm": 0.42160654067993164,
"learning_rate": 8.714285714285715e-06,
"loss": 0.5148,
"step": 61
},
{
"epoch": 0.266857962697274,
"grad_norm": 0.42496350407600403,
"learning_rate": 8.857142857142858e-06,
"loss": 0.5086,
"step": 62
},
{
"epoch": 0.27116212338593976,
"grad_norm": 0.3931051194667816,
"learning_rate": 9e-06,
"loss": 0.5029,
"step": 63
},
{
"epoch": 0.27546628407460544,
"grad_norm": 0.4256575405597687,
"learning_rate": 9.142857142857144e-06,
"loss": 0.5208,
"step": 64
},
{
"epoch": 0.2797704447632712,
"grad_norm": 0.4178571403026581,
"learning_rate": 9.285714285714288e-06,
"loss": 0.5016,
"step": 65
},
{
"epoch": 0.28407460545193686,
"grad_norm": 0.4503072202205658,
"learning_rate": 9.42857142857143e-06,
"loss": 0.4973,
"step": 66
},
{
"epoch": 0.2883787661406026,
"grad_norm": 0.4047144055366516,
"learning_rate": 9.571428571428573e-06,
"loss": 0.5038,
"step": 67
},
{
"epoch": 0.2926829268292683,
"grad_norm": 0.5516330003738403,
"learning_rate": 9.714285714285715e-06,
"loss": 0.5281,
"step": 68
},
{
"epoch": 0.296987087517934,
"grad_norm": 0.43566030263900757,
"learning_rate": 9.857142857142859e-06,
"loss": 0.5086,
"step": 69
},
{
"epoch": 0.3012912482065997,
"grad_norm": 0.41416284441947937,
"learning_rate": 1e-05,
"loss": 0.5203,
"step": 70
},
{
"epoch": 0.30559540889526543,
"grad_norm": 0.47125673294067383,
"learning_rate": 9.999937036309402e-06,
"loss": 0.498,
"step": 71
},
{
"epoch": 0.3098995695839311,
"grad_norm": 0.38754016160964966,
"learning_rate": 9.999748146823376e-06,
"loss": 0.4988,
"step": 72
},
{
"epoch": 0.31420373027259685,
"grad_norm": 0.4142284393310547,
"learning_rate": 9.999433336299195e-06,
"loss": 0.4885,
"step": 73
},
{
"epoch": 0.31850789096126253,
"grad_norm": 0.41517406702041626,
"learning_rate": 9.99899261266551e-06,
"loss": 0.4906,
"step": 74
},
{
"epoch": 0.32281205164992827,
"grad_norm": 0.3982820212841034,
"learning_rate": 9.99842598702216e-06,
"loss": 0.5063,
"step": 75
},
{
"epoch": 0.32711621233859395,
"grad_norm": 0.3950318396091461,
"learning_rate": 9.997733473639876e-06,
"loss": 0.4885,
"step": 76
},
{
"epoch": 0.3314203730272597,
"grad_norm": 0.40950194001197815,
"learning_rate": 9.996915089959942e-06,
"loss": 0.4966,
"step": 77
},
{
"epoch": 0.3357245337159254,
"grad_norm": 0.4177294373512268,
"learning_rate": 9.995970856593739e-06,
"loss": 0.5168,
"step": 78
},
{
"epoch": 0.3400286944045911,
"grad_norm": 0.4307166039943695,
"learning_rate": 9.994900797322233e-06,
"loss": 0.4935,
"step": 79
},
{
"epoch": 0.34433285509325684,
"grad_norm": 0.452928751707077,
"learning_rate": 9.993704939095376e-06,
"loss": 0.5037,
"step": 80
},
{
"epoch": 0.3486370157819225,
"grad_norm": 0.4000836908817291,
"learning_rate": 9.99238331203143e-06,
"loss": 0.4974,
"step": 81
},
{
"epoch": 0.35294117647058826,
"grad_norm": 0.3994639813899994,
"learning_rate": 9.9909359494162e-06,
"loss": 0.4869,
"step": 82
},
{
"epoch": 0.35724533715925394,
"grad_norm": 0.40204912424087524,
"learning_rate": 9.989362887702203e-06,
"loss": 0.4933,
"step": 83
},
{
"epoch": 0.3615494978479197,
"grad_norm": 0.4123329818248749,
"learning_rate": 9.987664166507749e-06,
"loss": 0.4969,
"step": 84
},
{
"epoch": 0.36585365853658536,
"grad_norm": 0.40340450406074524,
"learning_rate": 9.985839828615937e-06,
"loss": 0.5092,
"step": 85
},
{
"epoch": 0.3701578192252511,
"grad_norm": 0.3741258382797241,
"learning_rate": 9.983889919973586e-06,
"loss": 0.4925,
"step": 86
},
{
"epoch": 0.3744619799139168,
"grad_norm": 0.40813329815864563,
"learning_rate": 9.981814489690077e-06,
"loss": 0.4614,
"step": 87
},
{
"epoch": 0.3787661406025825,
"grad_norm": 0.4270995259284973,
"learning_rate": 9.979613590036108e-06,
"loss": 0.5068,
"step": 88
},
{
"epoch": 0.3830703012912482,
"grad_norm": 0.4126576781272888,
"learning_rate": 9.977287276442385e-06,
"loss": 0.5127,
"step": 89
},
{
"epoch": 0.38737446197991393,
"grad_norm": 0.41353291273117065,
"learning_rate": 9.974835607498224e-06,
"loss": 0.4714,
"step": 90
},
{
"epoch": 0.3916786226685796,
"grad_norm": 0.43727749586105347,
"learning_rate": 9.972258644950074e-06,
"loss": 0.5083,
"step": 91
},
{
"epoch": 0.39598278335724535,
"grad_norm": 0.4587208032608032,
"learning_rate": 9.969556453699966e-06,
"loss": 0.4958,
"step": 92
},
{
"epoch": 0.40028694404591103,
"grad_norm": 0.3831060528755188,
"learning_rate": 9.966729101803872e-06,
"loss": 0.4657,
"step": 93
},
{
"epoch": 0.40459110473457677,
"grad_norm": 0.4357382357120514,
"learning_rate": 9.963776660469996e-06,
"loss": 0.4826,
"step": 94
},
{
"epoch": 0.40889526542324245,
"grad_norm": 0.4969067871570587,
"learning_rate": 9.960699204056978e-06,
"loss": 0.4997,
"step": 95
},
{
"epoch": 0.4131994261119082,
"grad_norm": 0.4527207612991333,
"learning_rate": 9.957496810072027e-06,
"loss": 0.5259,
"step": 96
},
{
"epoch": 0.41750358680057387,
"grad_norm": 0.4595460891723633,
"learning_rate": 9.954169559168958e-06,
"loss": 0.5006,
"step": 97
},
{
"epoch": 0.4218077474892396,
"grad_norm": 0.5569490790367126,
"learning_rate": 9.95071753514617e-06,
"loss": 0.5234,
"step": 98
},
{
"epoch": 0.4261119081779053,
"grad_norm": 0.3885694146156311,
"learning_rate": 9.947140824944533e-06,
"loss": 0.4944,
"step": 99
},
{
"epoch": 0.430416068866571,
"grad_norm": 0.4560917019844055,
"learning_rate": 9.943439518645193e-06,
"loss": 0.4932,
"step": 100
},
{
"epoch": 0.4347202295552367,
"grad_norm": 0.4711724519729614,
"learning_rate": 9.939613709467317e-06,
"loss": 0.4882,
"step": 101
},
{
"epoch": 0.43902439024390244,
"grad_norm": 0.39771363139152527,
"learning_rate": 9.935663493765726e-06,
"loss": 0.4464,
"step": 102
},
{
"epoch": 0.4433285509325681,
"grad_norm": 0.4277415871620178,
"learning_rate": 9.93158897102849e-06,
"loss": 0.4761,
"step": 103
},
{
"epoch": 0.44763271162123386,
"grad_norm": 0.5095153450965881,
"learning_rate": 9.9273902438744e-06,
"loss": 0.48,
"step": 104
},
{
"epoch": 0.4519368723098996,
"grad_norm": 0.5148225426673889,
"learning_rate": 9.923067418050399e-06,
"loss": 0.4836,
"step": 105
},
{
"epoch": 0.4562410329985653,
"grad_norm": 0.48084282875061035,
"learning_rate": 9.918620602428916e-06,
"loss": 0.4873,
"step": 106
},
{
"epoch": 0.460545193687231,
"grad_norm": 0.47235897183418274,
"learning_rate": 9.91404990900512e-06,
"loss": 0.4939,
"step": 107
},
{
"epoch": 0.4648493543758967,
"grad_norm": 0.4564887285232544,
"learning_rate": 9.909355452894098e-06,
"loss": 0.4857,
"step": 108
},
{
"epoch": 0.46915351506456243,
"grad_norm": 0.5023767948150635,
"learning_rate": 9.904537352327968e-06,
"loss": 0.504,
"step": 109
},
{
"epoch": 0.4734576757532281,
"grad_norm": 0.44549015164375305,
"learning_rate": 9.899595728652883e-06,
"loss": 0.476,
"step": 110
},
{
"epoch": 0.47776183644189385,
"grad_norm": 0.4371887743473053,
"learning_rate": 9.894530706325994e-06,
"loss": 0.4842,
"step": 111
},
{
"epoch": 0.48206599713055953,
"grad_norm": 0.5438153147697449,
"learning_rate": 9.889342412912296e-06,
"loss": 0.4878,
"step": 112
},
{
"epoch": 0.48637015781922527,
"grad_norm": 0.5181624293327332,
"learning_rate": 9.88403097908143e-06,
"loss": 0.4924,
"step": 113
},
{
"epoch": 0.49067431850789095,
"grad_norm": 0.4505234658718109,
"learning_rate": 9.878596538604388e-06,
"loss": 0.5077,
"step": 114
},
{
"epoch": 0.4949784791965567,
"grad_norm": 0.546039342880249,
"learning_rate": 9.87303922835014e-06,
"loss": 0.5147,
"step": 115
},
{
"epoch": 0.49928263988522237,
"grad_norm": 0.45228955149650574,
"learning_rate": 9.867359188282193e-06,
"loss": 0.4933,
"step": 116
},
{
"epoch": 0.503586800573888,
"grad_norm": 0.4075925052165985,
"learning_rate": 9.861556561455061e-06,
"loss": 0.4635,
"step": 117
},
{
"epoch": 0.5078909612625538,
"grad_norm": 0.5900917649269104,
"learning_rate": 9.855631494010661e-06,
"loss": 0.495,
"step": 118
},
{
"epoch": 0.5121951219512195,
"grad_norm": 0.5058228969573975,
"learning_rate": 9.849584135174642e-06,
"loss": 0.5013,
"step": 119
},
{
"epoch": 0.5164992826398852,
"grad_norm": 0.526889443397522,
"learning_rate": 9.843414637252615e-06,
"loss": 0.4774,
"step": 120
},
{
"epoch": 0.5208034433285509,
"grad_norm": 0.5015537142753601,
"learning_rate": 9.837123155626323e-06,
"loss": 0.4835,
"step": 121
},
{
"epoch": 0.5251076040172167,
"grad_norm": 0.405774861574173,
"learning_rate": 9.830709848749727e-06,
"loss": 0.4766,
"step": 122
},
{
"epoch": 0.5294117647058824,
"grad_norm": 0.5407970547676086,
"learning_rate": 9.824174878145017e-06,
"loss": 0.4937,
"step": 123
},
{
"epoch": 0.533715925394548,
"grad_norm": 0.47976917028427124,
"learning_rate": 9.817518408398536e-06,
"loss": 0.4854,
"step": 124
},
{
"epoch": 0.5380200860832137,
"grad_norm": 0.49154919385910034,
"learning_rate": 9.810740607156647e-06,
"loss": 0.4837,
"step": 125
},
{
"epoch": 0.5423242467718795,
"grad_norm": 0.4680994749069214,
"learning_rate": 9.803841645121505e-06,
"loss": 0.4777,
"step": 126
},
{
"epoch": 0.5466284074605452,
"grad_norm": 0.47835245728492737,
"learning_rate": 9.796821696046748e-06,
"loss": 0.4878,
"step": 127
},
{
"epoch": 0.5509325681492109,
"grad_norm": 0.5058233737945557,
"learning_rate": 9.78968093673314e-06,
"loss": 0.4846,
"step": 128
},
{
"epoch": 0.5552367288378766,
"grad_norm": 0.4391472637653351,
"learning_rate": 9.782419547024108e-06,
"loss": 0.4835,
"step": 129
},
{
"epoch": 0.5595408895265424,
"grad_norm": 0.5265413522720337,
"learning_rate": 9.775037709801206e-06,
"loss": 0.5061,
"step": 130
},
{
"epoch": 0.563845050215208,
"grad_norm": 0.48637405037879944,
"learning_rate": 9.76753561097952e-06,
"loss": 0.4811,
"step": 131
},
{
"epoch": 0.5681492109038737,
"grad_norm": 0.4475197494029999,
"learning_rate": 9.759913439502982e-06,
"loss": 0.4518,
"step": 132
},
{
"epoch": 0.5724533715925395,
"grad_norm": 0.4481680393218994,
"learning_rate": 9.752171387339612e-06,
"loss": 0.4864,
"step": 133
},
{
"epoch": 0.5767575322812052,
"grad_norm": 0.4027429521083832,
"learning_rate": 9.74430964947668e-06,
"loss": 0.4815,
"step": 134
},
{
"epoch": 0.5810616929698709,
"grad_norm": 0.4797030985355377,
"learning_rate": 9.736328423915797e-06,
"loss": 0.4953,
"step": 135
},
{
"epoch": 0.5853658536585366,
"grad_norm": 0.4120473861694336,
"learning_rate": 9.728227911667934e-06,
"loss": 0.4841,
"step": 136
},
{
"epoch": 0.5896700143472023,
"grad_norm": 0.46151697635650635,
"learning_rate": 9.720008316748344e-06,
"loss": 0.4704,
"step": 137
},
{
"epoch": 0.593974175035868,
"grad_norm": 0.4600050151348114,
"learning_rate": 9.711669846171443e-06,
"loss": 0.4652,
"step": 138
},
{
"epoch": 0.5982783357245337,
"grad_norm": 0.4155121147632599,
"learning_rate": 9.703212709945583e-06,
"loss": 0.4754,
"step": 139
},
{
"epoch": 0.6025824964131994,
"grad_norm": 0.4314727783203125,
"learning_rate": 9.694637121067764e-06,
"loss": 0.4822,
"step": 140
},
{
"epoch": 0.6068866571018652,
"grad_norm": 0.5324988961219788,
"learning_rate": 9.685943295518283e-06,
"loss": 0.4795,
"step": 141
},
{
"epoch": 0.6111908177905309,
"grad_norm": 0.45600083470344543,
"learning_rate": 9.677131452255272e-06,
"loss": 0.4697,
"step": 142
},
{
"epoch": 0.6154949784791965,
"grad_norm": 0.5205792188644409,
"learning_rate": 9.668201813209202e-06,
"loss": 0.4803,
"step": 143
},
{
"epoch": 0.6197991391678622,
"grad_norm": 0.4811612069606781,
"learning_rate": 9.659154603277283e-06,
"loss": 0.4715,
"step": 144
},
{
"epoch": 0.624103299856528,
"grad_norm": 0.4459843933582306,
"learning_rate": 9.649990050317806e-06,
"loss": 0.4775,
"step": 145
},
{
"epoch": 0.6284074605451937,
"grad_norm": 0.5670981407165527,
"learning_rate": 9.640708385144403e-06,
"loss": 0.4789,
"step": 146
},
{
"epoch": 0.6327116212338594,
"grad_norm": 0.5041674971580505,
"learning_rate": 9.631309841520233e-06,
"loss": 0.4936,
"step": 147
},
{
"epoch": 0.6370157819225251,
"grad_norm": 0.5009613633155823,
"learning_rate": 9.62179465615209e-06,
"loss": 0.5072,
"step": 148
},
{
"epoch": 0.6413199426111909,
"grad_norm": 0.5727151036262512,
"learning_rate": 9.612163068684453e-06,
"loss": 0.4764,
"step": 149
},
{
"epoch": 0.6456241032998565,
"grad_norm": 0.6043273210525513,
"learning_rate": 9.602415321693434e-06,
"loss": 0.4818,
"step": 150
},
{
"epoch": 0.6499282639885222,
"grad_norm": 0.445709764957428,
"learning_rate": 9.592551660680687e-06,
"loss": 0.465,
"step": 151
},
{
"epoch": 0.6542324246771879,
"grad_norm": 0.6107316613197327,
"learning_rate": 9.582572334067213e-06,
"loss": 0.49,
"step": 152
},
{
"epoch": 0.6585365853658537,
"grad_norm": 0.4649333655834198,
"learning_rate": 9.572477593187101e-06,
"loss": 0.4583,
"step": 153
},
{
"epoch": 0.6628407460545194,
"grad_norm": 0.499577134847641,
"learning_rate": 9.562267692281212e-06,
"loss": 0.4864,
"step": 154
},
{
"epoch": 0.667144906743185,
"grad_norm": 0.5428804755210876,
"learning_rate": 9.551942888490759e-06,
"loss": 0.4746,
"step": 155
},
{
"epoch": 0.6714490674318508,
"grad_norm": 0.48842427134513855,
"learning_rate": 9.541503441850844e-06,
"loss": 0.5099,
"step": 156
},
{
"epoch": 0.6757532281205165,
"grad_norm": 0.4386967420578003,
"learning_rate": 9.530949615283902e-06,
"loss": 0.478,
"step": 157
},
{
"epoch": 0.6800573888091822,
"grad_norm": 0.48663437366485596,
"learning_rate": 9.520281674593084e-06,
"loss": 0.4968,
"step": 158
},
{
"epoch": 0.6843615494978479,
"grad_norm": 0.4161711633205414,
"learning_rate": 9.509499888455554e-06,
"loss": 0.4686,
"step": 159
},
{
"epoch": 0.6886657101865137,
"grad_norm": 0.4658878445625305,
"learning_rate": 9.498604528415731e-06,
"loss": 0.4669,
"step": 160
},
{
"epoch": 0.6929698708751794,
"grad_norm": 0.33149388432502747,
"learning_rate": 9.487595868878447e-06,
"loss": 0.4801,
"step": 161
},
{
"epoch": 0.697274031563845,
"grad_norm": 0.4957387447357178,
"learning_rate": 9.476474187102033e-06,
"loss": 0.4546,
"step": 162
},
{
"epoch": 0.7015781922525107,
"grad_norm": 0.4245833158493042,
"learning_rate": 9.465239763191345e-06,
"loss": 0.4939,
"step": 163
},
{
"epoch": 0.7058823529411765,
"grad_norm": 0.4390011131763458,
"learning_rate": 9.453892880090696e-06,
"loss": 0.4745,
"step": 164
},
{
"epoch": 0.7101865136298422,
"grad_norm": 0.5233902931213379,
"learning_rate": 9.442433823576741e-06,
"loss": 0.488,
"step": 165
},
{
"epoch": 0.7144906743185079,
"grad_norm": 0.4004855155944824,
"learning_rate": 9.430862882251279e-06,
"loss": 0.4666,
"step": 166
},
{
"epoch": 0.7187948350071736,
"grad_norm": 0.48547080159187317,
"learning_rate": 9.419180347533976e-06,
"loss": 0.475,
"step": 167
},
{
"epoch": 0.7230989956958394,
"grad_norm": 0.5068047046661377,
"learning_rate": 9.40738651365503e-06,
"loss": 0.4952,
"step": 168
},
{
"epoch": 0.727403156384505,
"grad_norm": 0.49898117780685425,
"learning_rate": 9.395481677647767e-06,
"loss": 0.4736,
"step": 169
},
{
"epoch": 0.7317073170731707,
"grad_norm": 0.41030001640319824,
"learning_rate": 9.38346613934115e-06,
"loss": 0.4796,
"step": 170
},
{
"epoch": 0.7360114777618364,
"grad_norm": 0.47655925154685974,
"learning_rate": 9.371340201352234e-06,
"loss": 0.4819,
"step": 171
},
{
"epoch": 0.7403156384505022,
"grad_norm": 0.4729374945163727,
"learning_rate": 9.359104169078541e-06,
"loss": 0.4819,
"step": 172
},
{
"epoch": 0.7446197991391679,
"grad_norm": 0.45545288920402527,
"learning_rate": 9.346758350690373e-06,
"loss": 0.4931,
"step": 173
},
{
"epoch": 0.7489239598278336,
"grad_norm": 0.4552542269229889,
"learning_rate": 9.334303057123044e-06,
"loss": 0.4707,
"step": 174
},
{
"epoch": 0.7532281205164992,
"grad_norm": 0.45603325963020325,
"learning_rate": 9.321738602069057e-06,
"loss": 0.4824,
"step": 175
},
{
"epoch": 0.757532281205165,
"grad_norm": 0.5039983987808228,
"learning_rate": 9.309065301970193e-06,
"loss": 0.4862,
"step": 176
},
{
"epoch": 0.7618364418938307,
"grad_norm": 0.4347105622291565,
"learning_rate": 9.296283476009551e-06,
"loss": 0.481,
"step": 177
},
{
"epoch": 0.7661406025824964,
"grad_norm": 0.507379949092865,
"learning_rate": 9.283393446103506e-06,
"loss": 0.4724,
"step": 178
},
{
"epoch": 0.7704447632711621,
"grad_norm": 0.44477155804634094,
"learning_rate": 9.270395536893599e-06,
"loss": 0.4549,
"step": 179
},
{
"epoch": 0.7747489239598279,
"grad_norm": 0.4589724540710449,
"learning_rate": 9.257290075738365e-06,
"loss": 0.4739,
"step": 180
},
{
"epoch": 0.7790530846484935,
"grad_norm": 0.3952377140522003,
"learning_rate": 9.244077392705085e-06,
"loss": 0.4385,
"step": 181
},
{
"epoch": 0.7833572453371592,
"grad_norm": 0.4503597021102905,
"learning_rate": 9.23075782056147e-06,
"loss": 0.4746,
"step": 182
},
{
"epoch": 0.787661406025825,
"grad_norm": 0.44907307624816895,
"learning_rate": 9.217331694767291e-06,
"loss": 0.4657,
"step": 183
},
{
"epoch": 0.7919655667144907,
"grad_norm": 0.4200761318206787,
"learning_rate": 9.20379935346592e-06,
"loss": 0.4742,
"step": 184
},
{
"epoch": 0.7962697274031564,
"grad_norm": 0.4928026795387268,
"learning_rate": 9.190161137475814e-06,
"loss": 0.4548,
"step": 185
},
{
"epoch": 0.8005738880918221,
"grad_norm": 0.41791635751724243,
"learning_rate": 9.176417390281944e-06,
"loss": 0.4717,
"step": 186
},
{
"epoch": 0.8048780487804879,
"grad_norm": 0.4696124792098999,
"learning_rate": 9.162568458027122e-06,
"loss": 0.4748,
"step": 187
},
{
"epoch": 0.8091822094691535,
"grad_norm": 0.41637182235717773,
"learning_rate": 9.148614689503307e-06,
"loss": 0.4905,
"step": 188
},
{
"epoch": 0.8134863701578192,
"grad_norm": 0.42210012674331665,
"learning_rate": 9.134556436142801e-06,
"loss": 0.4865,
"step": 189
},
{
"epoch": 0.8177905308464849,
"grad_norm": 0.3945121169090271,
"learning_rate": 9.120394052009412e-06,
"loss": 0.4805,
"step": 190
},
{
"epoch": 0.8220946915351507,
"grad_norm": 0.4084663987159729,
"learning_rate": 9.10612789378953e-06,
"loss": 0.4546,
"step": 191
},
{
"epoch": 0.8263988522238164,
"grad_norm": 0.39801859855651855,
"learning_rate": 9.091758320783139e-06,
"loss": 0.4813,
"step": 192
},
{
"epoch": 0.830703012912482,
"grad_norm": 0.374301940202713,
"learning_rate": 9.077285694894786e-06,
"loss": 0.4629,
"step": 193
},
{
"epoch": 0.8350071736011477,
"grad_norm": 0.34360548853874207,
"learning_rate": 9.062710380624439e-06,
"loss": 0.437,
"step": 194
},
{
"epoch": 0.8393113342898135,
"grad_norm": 0.40194225311279297,
"learning_rate": 9.048032745058335e-06,
"loss": 0.4503,
"step": 195
},
{
"epoch": 0.8436154949784792,
"grad_norm": 0.38807958364486694,
"learning_rate": 9.033253157859715e-06,
"loss": 0.4817,
"step": 196
},
{
"epoch": 0.8479196556671449,
"grad_norm": 0.3624962270259857,
"learning_rate": 9.018371991259516e-06,
"loss": 0.4466,
"step": 197
},
{
"epoch": 0.8522238163558106,
"grad_norm": 0.40797358751296997,
"learning_rate": 9.003389620047012e-06,
"loss": 0.4774,
"step": 198
},
{
"epoch": 0.8565279770444764,
"grad_norm": 0.40521928668022156,
"learning_rate": 8.988306421560354e-06,
"loss": 0.4807,
"step": 199
},
{
"epoch": 0.860832137733142,
"grad_norm": 0.4247490465641022,
"learning_rate": 8.973122775677078e-06,
"loss": 0.4626,
"step": 200
},
{
"epoch": 0.8651362984218077,
"grad_norm": 0.49826568365097046,
"learning_rate": 8.957839064804542e-06,
"loss": 0.4802,
"step": 201
},
{
"epoch": 0.8694404591104734,
"grad_norm": 0.4205709397792816,
"learning_rate": 8.942455673870278e-06,
"loss": 0.4867,
"step": 202
},
{
"epoch": 0.8737446197991392,
"grad_norm": 0.42230424284935,
"learning_rate": 8.926972990312314e-06,
"loss": 0.4918,
"step": 203
},
{
"epoch": 0.8780487804878049,
"grad_norm": 0.4043838381767273,
"learning_rate": 8.91139140406941e-06,
"loss": 0.4936,
"step": 204
},
{
"epoch": 0.8823529411764706,
"grad_norm": 0.38208943605422974,
"learning_rate": 8.895711307571235e-06,
"loss": 0.4485,
"step": 205
},
{
"epoch": 0.8866571018651362,
"grad_norm": 0.41558074951171875,
"learning_rate": 8.879933095728485e-06,
"loss": 0.4658,
"step": 206
},
{
"epoch": 0.890961262553802,
"grad_norm": 0.3710712492465973,
"learning_rate": 8.864057165922944e-06,
"loss": 0.4531,
"step": 207
},
{
"epoch": 0.8952654232424677,
"grad_norm": 0.43846508860588074,
"learning_rate": 8.848083917997463e-06,
"loss": 0.4754,
"step": 208
},
{
"epoch": 0.8995695839311334,
"grad_norm": 0.44018861651420593,
"learning_rate": 8.832013754245895e-06,
"loss": 0.4828,
"step": 209
},
{
"epoch": 0.9038737446197992,
"grad_norm": 0.47094517946243286,
"learning_rate": 8.815847079402972e-06,
"loss": 0.4455,
"step": 210
},
{
"epoch": 0.9081779053084649,
"grad_norm": 0.40249302983283997,
"learning_rate": 8.799584300634096e-06,
"loss": 0.4841,
"step": 211
},
{
"epoch": 0.9124820659971306,
"grad_norm": 0.4624997675418854,
"learning_rate": 8.783225827525098e-06,
"loss": 0.4533,
"step": 212
},
{
"epoch": 0.9167862266857962,
"grad_norm": 0.4431861639022827,
"learning_rate": 8.766772072071911e-06,
"loss": 0.4676,
"step": 213
},
{
"epoch": 0.921090387374462,
"grad_norm": 0.4086032211780548,
"learning_rate": 8.750223448670204e-06,
"loss": 0.4625,
"step": 214
},
{
"epoch": 0.9253945480631277,
"grad_norm": 0.496051162481308,
"learning_rate": 8.733580374104936e-06,
"loss": 0.4569,
"step": 215
},
{
"epoch": 0.9296987087517934,
"grad_norm": 0.545372724533081,
"learning_rate": 8.716843267539868e-06,
"loss": 0.4738,
"step": 216
},
{
"epoch": 0.9340028694404591,
"grad_norm": 0.45664075016975403,
"learning_rate": 8.700012550507e-06,
"loss": 0.4887,
"step": 217
},
{
"epoch": 0.9383070301291249,
"grad_norm": 0.4230859577655792,
"learning_rate": 8.683088646895955e-06,
"loss": 0.4349,
"step": 218
},
{
"epoch": 0.9426111908177905,
"grad_norm": 0.453646719455719,
"learning_rate": 8.666071982943306e-06,
"loss": 0.453,
"step": 219
},
{
"epoch": 0.9469153515064562,
"grad_norm": 0.38937658071517944,
"learning_rate": 8.648962987221837e-06,
"loss": 0.4767,
"step": 220
},
{
"epoch": 0.9512195121951219,
"grad_norm": 0.4459855258464813,
"learning_rate": 8.631762090629756e-06,
"loss": 0.4814,
"step": 221
},
{
"epoch": 0.9555236728837877,
"grad_norm": 0.3893451392650604,
"learning_rate": 8.614469726379833e-06,
"loss": 0.4761,
"step": 222
},
{
"epoch": 0.9598278335724534,
"grad_norm": 0.37297987937927246,
"learning_rate": 8.597086329988498e-06,
"loss": 0.4673,
"step": 223
},
{
"epoch": 0.9641319942611191,
"grad_norm": 0.46203163266181946,
"learning_rate": 8.579612339264867e-06,
"loss": 0.4946,
"step": 224
},
{
"epoch": 0.9684361549497847,
"grad_norm": 0.42642268538475037,
"learning_rate": 8.562048194299719e-06,
"loss": 0.4776,
"step": 225
},
{
"epoch": 0.9727403156384505,
"grad_norm": 0.3741326928138733,
"learning_rate": 8.544394337454409e-06,
"loss": 0.4632,
"step": 226
},
{
"epoch": 0.9770444763271162,
"grad_norm": 0.39677464962005615,
"learning_rate": 8.52665121334973e-06,
"loss": 0.4821,
"step": 227
},
{
"epoch": 0.9813486370157819,
"grad_norm": 0.42544877529144287,
"learning_rate": 8.508819268854713e-06,
"loss": 0.4893,
"step": 228
},
{
"epoch": 0.9856527977044476,
"grad_norm": 0.4025561213493347,
"learning_rate": 8.49089895307537e-06,
"loss": 0.4682,
"step": 229
},
{
"epoch": 0.9899569583931134,
"grad_norm": 0.4395465552806854,
"learning_rate": 8.472890717343391e-06,
"loss": 0.4558,
"step": 230
},
{
"epoch": 0.994261119081779,
"grad_norm": 0.3767266571521759,
"learning_rate": 8.454795015204767e-06,
"loss": 0.4616,
"step": 231
},
{
"epoch": 0.9985652797704447,
"grad_norm": 0.42802631855010986,
"learning_rate": 8.436612302408376e-06,
"loss": 0.4594,
"step": 232
},
{
"epoch": 1.0028694404591105,
"grad_norm": 0.7791727781295776,
"learning_rate": 8.418343036894497e-06,
"loss": 0.7338,
"step": 233
},
{
"epoch": 1.007173601147776,
"grad_norm": 0.446108341217041,
"learning_rate": 8.399987678783285e-06,
"loss": 0.4693,
"step": 234
},
{
"epoch": 1.011477761836442,
"grad_norm": 0.44180288910865784,
"learning_rate": 8.381546690363174e-06,
"loss": 0.4326,
"step": 235
},
{
"epoch": 1.0157819225251077,
"grad_norm": 0.4678577184677124,
"learning_rate": 8.36302053607924e-06,
"loss": 0.4729,
"step": 236
},
{
"epoch": 1.0200860832137733,
"grad_norm": 0.4438740015029907,
"learning_rate": 8.344409682521499e-06,
"loss": 0.4454,
"step": 237
},
{
"epoch": 1.024390243902439,
"grad_norm": 0.46316948533058167,
"learning_rate": 8.325714598413169e-06,
"loss": 0.4322,
"step": 238
},
{
"epoch": 1.0286944045911048,
"grad_norm": 0.3941633403301239,
"learning_rate": 8.306935754598838e-06,
"loss": 0.4176,
"step": 239
},
{
"epoch": 1.0329985652797704,
"grad_norm": 0.4438393712043762,
"learning_rate": 8.288073624032634e-06,
"loss": 0.4339,
"step": 240
},
{
"epoch": 1.0373027259684362,
"grad_norm": 0.440329372882843,
"learning_rate": 8.269128681766296e-06,
"loss": 0.4535,
"step": 241
},
{
"epoch": 1.0416068866571018,
"grad_norm": 0.4060475826263428,
"learning_rate": 8.250101404937223e-06,
"loss": 0.4235,
"step": 242
},
{
"epoch": 1.0459110473457676,
"grad_norm": 0.41474077105522156,
"learning_rate": 8.230992272756438e-06,
"loss": 0.4531,
"step": 243
},
{
"epoch": 1.0502152080344334,
"grad_norm": 0.44152936339378357,
"learning_rate": 8.211801766496537e-06,
"loss": 0.4601,
"step": 244
},
{
"epoch": 1.054519368723099,
"grad_norm": 0.42042216658592224,
"learning_rate": 8.192530369479562e-06,
"loss": 0.4376,
"step": 245
},
{
"epoch": 1.0588235294117647,
"grad_norm": 0.49132099747657776,
"learning_rate": 8.17317856706482e-06,
"loss": 0.4598,
"step": 246
},
{
"epoch": 1.0631276901004305,
"grad_norm": 0.44164198637008667,
"learning_rate": 8.153746846636675e-06,
"loss": 0.427,
"step": 247
},
{
"epoch": 1.067431850789096,
"grad_norm": 0.4028110206127167,
"learning_rate": 8.13423569759226e-06,
"loss": 0.435,
"step": 248
},
{
"epoch": 1.0717360114777619,
"grad_norm": 0.41196516156196594,
"learning_rate": 8.114645611329152e-06,
"loss": 0.4283,
"step": 249
},
{
"epoch": 1.0760401721664274,
"grad_norm": 0.42356353998184204,
"learning_rate": 8.094977081233006e-06,
"loss": 0.4165,
"step": 250
},
{
"epoch": 1.0803443328550932,
"grad_norm": 0.44623392820358276,
"learning_rate": 8.075230602665118e-06,
"loss": 0.4703,
"step": 251
},
{
"epoch": 1.084648493543759,
"grad_norm": 0.42129871249198914,
"learning_rate": 8.055406672949957e-06,
"loss": 0.449,
"step": 252
},
{
"epoch": 1.0889526542324246,
"grad_norm": 0.44216564297676086,
"learning_rate": 8.03550579136263e-06,
"loss": 0.4292,
"step": 253
},
{
"epoch": 1.0932568149210904,
"grad_norm": 0.4126872718334198,
"learning_rate": 8.015528459116321e-06,
"loss": 0.4383,
"step": 254
},
{
"epoch": 1.0975609756097562,
"grad_norm": 0.44940850138664246,
"learning_rate": 7.995475179349657e-06,
"loss": 0.4499,
"step": 255
},
{
"epoch": 1.1018651362984218,
"grad_norm": 0.4903542995452881,
"learning_rate": 7.975346457114034e-06,
"loss": 0.4388,
"step": 256
},
{
"epoch": 1.1061692969870875,
"grad_norm": 0.5634150505065918,
"learning_rate": 7.955142799360914e-06,
"loss": 0.4649,
"step": 257
},
{
"epoch": 1.1104734576757531,
"grad_norm": 0.3710925877094269,
"learning_rate": 7.934864714929036e-06,
"loss": 0.4136,
"step": 258
},
{
"epoch": 1.114777618364419,
"grad_norm": 0.5903193354606628,
"learning_rate": 7.914512714531612e-06,
"loss": 0.4384,
"step": 259
},
{
"epoch": 1.1190817790530847,
"grad_norm": 0.435581773519516,
"learning_rate": 7.894087310743468e-06,
"loss": 0.4433,
"step": 260
},
{
"epoch": 1.1233859397417503,
"grad_norm": 0.3968298137187958,
"learning_rate": 7.873589017988124e-06,
"loss": 0.4362,
"step": 261
},
{
"epoch": 1.127690100430416,
"grad_norm": 0.38424989581108093,
"learning_rate": 7.853018352524845e-06,
"loss": 0.4054,
"step": 262
},
{
"epoch": 1.1319942611190819,
"grad_norm": 0.41407302021980286,
"learning_rate": 7.832375832435637e-06,
"loss": 0.4461,
"step": 263
},
{
"epoch": 1.1362984218077474,
"grad_norm": 0.40711095929145813,
"learning_rate": 7.811661977612202e-06,
"loss": 0.4898,
"step": 264
},
{
"epoch": 1.1406025824964132,
"grad_norm": 0.41025951504707336,
"learning_rate": 7.790877309742833e-06,
"loss": 0.4188,
"step": 265
},
{
"epoch": 1.144906743185079,
"grad_norm": 0.49598750472068787,
"learning_rate": 7.770022352299294e-06,
"loss": 0.4233,
"step": 266
},
{
"epoch": 1.1492109038737446,
"grad_norm": 0.41980111598968506,
"learning_rate": 7.749097630523618e-06,
"loss": 0.449,
"step": 267
},
{
"epoch": 1.1535150645624104,
"grad_norm": 0.37297967076301575,
"learning_rate": 7.728103671414889e-06,
"loss": 0.444,
"step": 268
},
{
"epoch": 1.157819225251076,
"grad_norm": 0.4400876462459564,
"learning_rate": 7.707041003715962e-06,
"loss": 0.4478,
"step": 269
},
{
"epoch": 1.1621233859397417,
"grad_norm": 0.40423017740249634,
"learning_rate": 7.685910157900158e-06,
"loss": 0.4579,
"step": 270
},
{
"epoch": 1.1664275466284075,
"grad_norm": 0.3636547029018402,
"learning_rate": 7.66471166615789e-06,
"loss": 0.4137,
"step": 271
},
{
"epoch": 1.170731707317073,
"grad_norm": 0.37857159972190857,
"learning_rate": 7.643446062383273e-06,
"loss": 0.4582,
"step": 272
},
{
"epoch": 1.175035868005739,
"grad_norm": 0.37849536538124084,
"learning_rate": 7.622113882160658e-06,
"loss": 0.4222,
"step": 273
},
{
"epoch": 1.1793400286944047,
"grad_norm": 0.36707597970962524,
"learning_rate": 7.600715662751166e-06,
"loss": 0.4181,
"step": 274
},
{
"epoch": 1.1836441893830703,
"grad_norm": 0.4041919708251953,
"learning_rate": 7.579251943079145e-06,
"loss": 0.4115,
"step": 275
},
{
"epoch": 1.187948350071736,
"grad_norm": 0.4002331793308258,
"learning_rate": 7.557723263718596e-06,
"loss": 0.4649,
"step": 276
},
{
"epoch": 1.1922525107604018,
"grad_norm": 0.38895294070243835,
"learning_rate": 7.536130166879561e-06,
"loss": 0.4328,
"step": 277
},
{
"epoch": 1.1965566714490674,
"grad_norm": 0.35401999950408936,
"learning_rate": 7.514473196394467e-06,
"loss": 0.4379,
"step": 278
},
{
"epoch": 1.2008608321377332,
"grad_norm": 0.4220169484615326,
"learning_rate": 7.492752897704432e-06,
"loss": 0.4216,
"step": 279
},
{
"epoch": 1.2051649928263988,
"grad_norm": 0.39525651931762695,
"learning_rate": 7.470969817845518e-06,
"loss": 0.4707,
"step": 280
},
{
"epoch": 1.2094691535150646,
"grad_norm": 0.4009878635406494,
"learning_rate": 7.4491245054349716e-06,
"loss": 0.4229,
"step": 281
},
{
"epoch": 1.2137733142037304,
"grad_norm": 0.4798336923122406,
"learning_rate": 7.427217510657383e-06,
"loss": 0.4593,
"step": 282
},
{
"epoch": 1.218077474892396,
"grad_norm": 0.3721896708011627,
"learning_rate": 7.405249385250854e-06,
"loss": 0.4281,
"step": 283
},
{
"epoch": 1.2223816355810617,
"grad_norm": 0.40978696942329407,
"learning_rate": 7.383220682493081e-06,
"loss": 0.4507,
"step": 284
},
{
"epoch": 1.2266857962697273,
"grad_norm": 0.4719083905220032,
"learning_rate": 7.361131957187435e-06,
"loss": 0.4228,
"step": 285
},
{
"epoch": 1.230989956958393,
"grad_norm": 0.38987070322036743,
"learning_rate": 7.338983765648985e-06,
"loss": 0.4602,
"step": 286
},
{
"epoch": 1.2352941176470589,
"grad_norm": 0.4016108512878418,
"learning_rate": 7.31677666569048e-06,
"loss": 0.4095,
"step": 287
},
{
"epoch": 1.2395982783357244,
"grad_norm": 0.4143194854259491,
"learning_rate": 7.294511216608308e-06,
"loss": 0.4236,
"step": 288
},
{
"epoch": 1.2439024390243902,
"grad_norm": 0.4156009554862976,
"learning_rate": 7.272187979168408e-06,
"loss": 0.4422,
"step": 289
},
{
"epoch": 1.248206599713056,
"grad_norm": 0.4072224199771881,
"learning_rate": 7.249807515592149e-06,
"loss": 0.389,
"step": 290
},
{
"epoch": 1.2525107604017216,
"grad_norm": 0.37442606687545776,
"learning_rate": 7.227370389542161e-06,
"loss": 0.4358,
"step": 291
},
{
"epoch": 1.2568149210903874,
"grad_norm": 0.40852388739585876,
"learning_rate": 7.2048771661081515e-06,
"loss": 0.4325,
"step": 292
},
{
"epoch": 1.2611190817790532,
"grad_norm": 0.39829540252685547,
"learning_rate": 7.182328411792664e-06,
"loss": 0.4262,
"step": 293
},
{
"epoch": 1.2654232424677188,
"grad_norm": 0.3537748456001282,
"learning_rate": 7.159724694496815e-06,
"loss": 0.4488,
"step": 294
},
{
"epoch": 1.2697274031563845,
"grad_norm": 0.3861173689365387,
"learning_rate": 7.137066583505987e-06,
"loss": 0.4491,
"step": 295
},
{
"epoch": 1.2740315638450501,
"grad_norm": 0.38991910219192505,
"learning_rate": 7.114354649475499e-06,
"loss": 0.4399,
"step": 296
},
{
"epoch": 1.278335724533716,
"grad_norm": 0.3736940622329712,
"learning_rate": 7.091589464416225e-06,
"loss": 0.4123,
"step": 297
},
{
"epoch": 1.2826398852223817,
"grad_norm": 0.35150569677352905,
"learning_rate": 7.068771601680191e-06,
"loss": 0.3991,
"step": 298
},
{
"epoch": 1.2869440459110473,
"grad_norm": 0.38645830750465393,
"learning_rate": 7.04590163594614e-06,
"loss": 0.45,
"step": 299
},
{
"epoch": 1.291248206599713,
"grad_norm": 0.45290979743003845,
"learning_rate": 7.022980143205046e-06,
"loss": 0.4321,
"step": 300
},
{
"epoch": 1.2955523672883786,
"grad_norm": 0.3679206669330597,
"learning_rate": 7.000007700745622e-06,
"loss": 0.4263,
"step": 301
},
{
"epoch": 1.2998565279770444,
"grad_norm": 0.3815595507621765,
"learning_rate": 6.976984887139775e-06,
"loss": 0.453,
"step": 302
},
{
"epoch": 1.3041606886657102,
"grad_norm": 0.4131976068019867,
"learning_rate": 6.9539122822280246e-06,
"loss": 0.4471,
"step": 303
},
{
"epoch": 1.308464849354376,
"grad_norm": 0.3435700237751007,
"learning_rate": 6.930790467104916e-06,
"loss": 0.4367,
"step": 304
},
{
"epoch": 1.3127690100430416,
"grad_norm": 0.36125117540359497,
"learning_rate": 6.907620024104377e-06,
"loss": 0.38,
"step": 305
},
{
"epoch": 1.3170731707317074,
"grad_norm": 0.35791122913360596,
"learning_rate": 6.884401536785045e-06,
"loss": 0.4911,
"step": 306
},
{
"epoch": 1.321377331420373,
"grad_norm": 0.33072197437286377,
"learning_rate": 6.861135589915583e-06,
"loss": 0.429,
"step": 307
},
{
"epoch": 1.3256814921090387,
"grad_norm": 0.3363823890686035,
"learning_rate": 6.837822769459942e-06,
"loss": 0.4131,
"step": 308
},
{
"epoch": 1.3299856527977045,
"grad_norm": 0.3425028324127197,
"learning_rate": 6.814463662562609e-06,
"loss": 0.4293,
"step": 309
},
{
"epoch": 1.33428981348637,
"grad_norm": 0.3752707839012146,
"learning_rate": 6.791058857533814e-06,
"loss": 0.4421,
"step": 310
},
{
"epoch": 1.338593974175036,
"grad_norm": 0.3743986189365387,
"learning_rate": 6.767608943834721e-06,
"loss": 0.4438,
"step": 311
},
{
"epoch": 1.3428981348637015,
"grad_norm": 0.36510851979255676,
"learning_rate": 6.744114512062571e-06,
"loss": 0.4087,
"step": 312
},
{
"epoch": 1.3472022955523673,
"grad_norm": 0.4204147160053253,
"learning_rate": 6.720576153935818e-06,
"loss": 0.4873,
"step": 313
},
{
"epoch": 1.351506456241033,
"grad_norm": 0.3257698118686676,
"learning_rate": 6.696994462279223e-06,
"loss": 0.4085,
"step": 314
},
{
"epoch": 1.3558106169296988,
"grad_norm": 0.40403416752815247,
"learning_rate": 6.673370031008919e-06,
"loss": 0.4829,
"step": 315
},
{
"epoch": 1.3601147776183644,
"grad_norm": 0.3832758665084839,
"learning_rate": 6.6497034551174585e-06,
"loss": 0.421,
"step": 316
},
{
"epoch": 1.3644189383070302,
"grad_norm": 0.35449251532554626,
"learning_rate": 6.625995330658828e-06,
"loss": 0.4128,
"step": 317
},
{
"epoch": 1.3687230989956958,
"grad_norm": 0.37400197982788086,
"learning_rate": 6.602246254733431e-06,
"loss": 0.4257,
"step": 318
},
{
"epoch": 1.3730272596843616,
"grad_norm": 0.35631948709487915,
"learning_rate": 6.578456825473055e-06,
"loss": 0.4299,
"step": 319
},
{
"epoch": 1.3773314203730274,
"grad_norm": 0.33373865485191345,
"learning_rate": 6.554627642025807e-06,
"loss": 0.4412,
"step": 320
},
{
"epoch": 1.381635581061693,
"grad_norm": 0.38946038484573364,
"learning_rate": 6.53075930454102e-06,
"loss": 0.3981,
"step": 321
},
{
"epoch": 1.3859397417503587,
"grad_norm": 0.3750453591346741,
"learning_rate": 6.506852414154138e-06,
"loss": 0.461,
"step": 322
},
{
"epoch": 1.3902439024390243,
"grad_norm": 0.3329356014728546,
"learning_rate": 6.482907572971584e-06,
"loss": 0.3842,
"step": 323
},
{
"epoch": 1.39454806312769,
"grad_norm": 0.352209210395813,
"learning_rate": 6.4589253840555856e-06,
"loss": 0.4419,
"step": 324
},
{
"epoch": 1.3988522238163559,
"grad_norm": 0.3593578636646271,
"learning_rate": 6.434906451408991e-06,
"loss": 0.403,
"step": 325
},
{
"epoch": 1.4031563845050214,
"grad_norm": 0.4038272500038147,
"learning_rate": 6.41085137996006e-06,
"loss": 0.472,
"step": 326
},
{
"epoch": 1.4074605451936872,
"grad_norm": 0.30662405490875244,
"learning_rate": 6.386760775547221e-06,
"loss": 0.406,
"step": 327
},
{
"epoch": 1.4117647058823528,
"grad_norm": 0.3446487486362457,
"learning_rate": 6.362635244903818e-06,
"loss": 0.4379,
"step": 328
},
{
"epoch": 1.4160688665710186,
"grad_norm": 0.3683339059352875,
"learning_rate": 6.338475395642834e-06,
"loss": 0.4427,
"step": 329
},
{
"epoch": 1.4203730272596844,
"grad_norm": 0.3524536192417145,
"learning_rate": 6.314281836241573e-06,
"loss": 0.4482,
"step": 330
},
{
"epoch": 1.4246771879483502,
"grad_norm": 0.381590336561203,
"learning_rate": 6.2900551760263564e-06,
"loss": 0.4505,
"step": 331
},
{
"epoch": 1.4289813486370158,
"grad_norm": 0.3800303041934967,
"learning_rate": 6.265796025157154e-06,
"loss": 0.4342,
"step": 332
},
{
"epoch": 1.4332855093256816,
"grad_norm": 0.357076495885849,
"learning_rate": 6.241504994612237e-06,
"loss": 0.4149,
"step": 333
},
{
"epoch": 1.4375896700143471,
"grad_norm": 0.432861328125,
"learning_rate": 6.217182696172776e-06,
"loss": 0.4795,
"step": 334
},
{
"epoch": 1.441893830703013,
"grad_norm": 0.3767949342727661,
"learning_rate": 6.192829742407442e-06,
"loss": 0.3823,
"step": 335
},
{
"epoch": 1.4461979913916787,
"grad_norm": 0.3795372247695923,
"learning_rate": 6.168446746656973e-06,
"loss": 0.4557,
"step": 336
},
{
"epoch": 1.4505021520803443,
"grad_norm": 0.4189460277557373,
"learning_rate": 6.144034323018728e-06,
"loss": 0.4635,
"step": 337
},
{
"epoch": 1.45480631276901,
"grad_norm": 0.3688943684101105,
"learning_rate": 6.119593086331225e-06,
"loss": 0.372,
"step": 338
},
{
"epoch": 1.4591104734576756,
"grad_norm": 0.3465519845485687,
"learning_rate": 6.095123652158648e-06,
"loss": 0.4509,
"step": 339
},
{
"epoch": 1.4634146341463414,
"grad_norm": 0.42170581221580505,
"learning_rate": 6.070626636775349e-06,
"loss": 0.4685,
"step": 340
},
{
"epoch": 1.4677187948350072,
"grad_norm": 0.3757818639278412,
"learning_rate": 6.046102657150328e-06,
"loss": 0.455,
"step": 341
},
{
"epoch": 1.472022955523673,
"grad_norm": 0.3707659840583801,
"learning_rate": 6.021552330931693e-06,
"loss": 0.4176,
"step": 342
},
{
"epoch": 1.4763271162123386,
"grad_norm": 0.3855435848236084,
"learning_rate": 5.996976276431097e-06,
"loss": 0.4216,
"step": 343
},
{
"epoch": 1.4806312769010044,
"grad_norm": 0.37388697266578674,
"learning_rate": 5.972375112608182e-06,
"loss": 0.4122,
"step": 344
},
{
"epoch": 1.48493543758967,
"grad_norm": 0.38593074679374695,
"learning_rate": 5.947749459054972e-06,
"loss": 0.4459,
"step": 345
},
{
"epoch": 1.4892395982783357,
"grad_norm": 0.4058336615562439,
"learning_rate": 5.923099935980278e-06,
"loss": 0.4616,
"step": 346
},
{
"epoch": 1.4935437589670015,
"grad_norm": 0.3796309232711792,
"learning_rate": 5.898427164194084e-06,
"loss": 0.4336,
"step": 347
},
{
"epoch": 1.497847919655667,
"grad_norm": 0.3734838366508484,
"learning_rate": 5.8737317650918905e-06,
"loss": 0.4126,
"step": 348
},
{
"epoch": 1.502152080344333,
"grad_norm": 0.3841530680656433,
"learning_rate": 5.849014360639087e-06,
"loss": 0.4702,
"step": 349
},
{
"epoch": 1.5064562410329985,
"grad_norm": 0.3562658727169037,
"learning_rate": 5.824275573355278e-06,
"loss": 0.4034,
"step": 350
},
{
"epoch": 1.5107604017216643,
"grad_norm": 0.3666515350341797,
"learning_rate": 5.799516026298601e-06,
"loss": 0.4229,
"step": 351
},
{
"epoch": 1.51506456241033,
"grad_norm": 0.4212346374988556,
"learning_rate": 5.7747363430500395e-06,
"loss": 0.4376,
"step": 352
},
{
"epoch": 1.5193687230989958,
"grad_norm": 0.3742626905441284,
"learning_rate": 5.74993714769772e-06,
"loss": 0.4227,
"step": 353
},
{
"epoch": 1.5236728837876614,
"grad_norm": 0.4419170618057251,
"learning_rate": 5.725119064821185e-06,
"loss": 0.4358,
"step": 354
},
{
"epoch": 1.527977044476327,
"grad_norm": 0.35307490825653076,
"learning_rate": 5.700282719475672e-06,
"loss": 0.4292,
"step": 355
},
{
"epoch": 1.5322812051649928,
"grad_norm": 0.38687315583229065,
"learning_rate": 5.675428737176367e-06,
"loss": 0.398,
"step": 356
},
{
"epoch": 1.5365853658536586,
"grad_norm": 0.4023524224758148,
"learning_rate": 5.65055774388265e-06,
"loss": 0.4389,
"step": 357
},
{
"epoch": 1.5408895265423244,
"grad_norm": 0.3677155375480652,
"learning_rate": 5.625670365982332e-06,
"loss": 0.3917,
"step": 358
},
{
"epoch": 1.54519368723099,
"grad_norm": 0.3998677432537079,
"learning_rate": 5.600767230275878e-06,
"loss": 0.4452,
"step": 359
},
{
"epoch": 1.5494978479196555,
"grad_norm": 0.358852356672287,
"learning_rate": 5.575848963960621e-06,
"loss": 0.4241,
"step": 360
},
{
"epoch": 1.5538020086083213,
"grad_norm": 0.32302260398864746,
"learning_rate": 5.5509161946149635e-06,
"loss": 0.4185,
"step": 361
},
{
"epoch": 1.558106169296987,
"grad_norm": 0.4188165068626404,
"learning_rate": 5.525969550182577e-06,
"loss": 0.4654,
"step": 362
},
{
"epoch": 1.5624103299856529,
"grad_norm": 0.39718538522720337,
"learning_rate": 5.501009658956583e-06,
"loss": 0.4181,
"step": 363
},
{
"epoch": 1.5667144906743187,
"grad_norm": 0.3609788715839386,
"learning_rate": 5.4760371495637256e-06,
"loss": 0.4091,
"step": 364
},
{
"epoch": 1.5710186513629842,
"grad_norm": 0.35094860196113586,
"learning_rate": 5.451052650948549e-06,
"loss": 0.4354,
"step": 365
},
{
"epoch": 1.5753228120516498,
"grad_norm": 0.3525417149066925,
"learning_rate": 5.426056792357552e-06,
"loss": 0.4212,
"step": 366
},
{
"epoch": 1.5796269727403156,
"grad_norm": 0.36507681012153625,
"learning_rate": 5.40105020332333e-06,
"loss": 0.412,
"step": 367
},
{
"epoch": 1.5839311334289814,
"grad_norm": 0.3849915564060211,
"learning_rate": 5.376033513648743e-06,
"loss": 0.4572,
"step": 368
},
{
"epoch": 1.5882352941176472,
"grad_norm": 0.3586712181568146,
"learning_rate": 5.3510073533910344e-06,
"loss": 0.4308,
"step": 369
},
{
"epoch": 1.5925394548063128,
"grad_norm": 0.3649207651615143,
"learning_rate": 5.325972352845965e-06,
"loss": 0.4256,
"step": 370
},
{
"epoch": 1.5968436154949783,
"grad_norm": 0.3403324484825134,
"learning_rate": 5.30092914253195e-06,
"loss": 0.4058,
"step": 371
},
{
"epoch": 1.6011477761836441,
"grad_norm": 0.3676261603832245,
"learning_rate": 5.2758783531741655e-06,
"loss": 0.4122,
"step": 372
},
{
"epoch": 1.60545193687231,
"grad_norm": 0.34809696674346924,
"learning_rate": 5.25082061568867e-06,
"loss": 0.4101,
"step": 373
},
{
"epoch": 1.6097560975609757,
"grad_norm": 0.3801212012767792,
"learning_rate": 5.225756561166521e-06,
"loss": 0.3959,
"step": 374
},
{
"epoch": 1.6140602582496413,
"grad_norm": 0.39335742592811584,
"learning_rate": 5.200686820857862e-06,
"loss": 0.4715,
"step": 375
},
{
"epoch": 1.618364418938307,
"grad_norm": 0.33622682094573975,
"learning_rate": 5.175612026156045e-06,
"loss": 0.4125,
"step": 376
},
{
"epoch": 1.6226685796269726,
"grad_norm": 0.39736634492874146,
"learning_rate": 5.150532808581718e-06,
"loss": 0.4685,
"step": 377
},
{
"epoch": 1.6269727403156384,
"grad_norm": 0.32356390357017517,
"learning_rate": 5.125449799766916e-06,
"loss": 0.3995,
"step": 378
},
{
"epoch": 1.6312769010043042,
"grad_norm": 0.3594512939453125,
"learning_rate": 5.100363631439162e-06,
"loss": 0.4497,
"step": 379
},
{
"epoch": 1.63558106169297,
"grad_norm": 0.4013254642486572,
"learning_rate": 5.075274935405554e-06,
"loss": 0.4195,
"step": 380
},
{
"epoch": 1.6398852223816356,
"grad_norm": 0.3409561514854431,
"learning_rate": 5.0501843435368495e-06,
"loss": 0.4051,
"step": 381
},
{
"epoch": 1.6441893830703012,
"grad_norm": 0.3325178325176239,
"learning_rate": 5.025092487751552e-06,
"loss": 0.4421,
"step": 382
},
{
"epoch": 1.648493543758967,
"grad_norm": 0.3224661946296692,
"learning_rate": 5e-06,
"loss": 0.4121,
"step": 383
},
{
"epoch": 1.6527977044476327,
"grad_norm": 0.3134234845638275,
"learning_rate": 4.974907512248451e-06,
"loss": 0.4057,
"step": 384
},
{
"epoch": 1.6571018651362985,
"grad_norm": 0.351633220911026,
"learning_rate": 4.949815656463151e-06,
"loss": 0.4283,
"step": 385
},
{
"epoch": 1.661406025824964,
"grad_norm": 0.3710573613643646,
"learning_rate": 4.924725064594448e-06,
"loss": 0.4705,
"step": 386
},
{
"epoch": 1.6657101865136297,
"grad_norm": 0.33121249079704285,
"learning_rate": 4.89963636856084e-06,
"loss": 0.409,
"step": 387
},
{
"epoch": 1.6700143472022955,
"grad_norm": 0.3725872039794922,
"learning_rate": 4.874550200233085e-06,
"loss": 0.4838,
"step": 388
},
{
"epoch": 1.6743185078909613,
"grad_norm": 0.33540400862693787,
"learning_rate": 4.8494671914182835e-06,
"loss": 0.3972,
"step": 389
},
{
"epoch": 1.678622668579627,
"grad_norm": 0.3285572826862335,
"learning_rate": 4.824387973843957e-06,
"loss": 0.4366,
"step": 390
},
{
"epoch": 1.6829268292682928,
"grad_norm": 0.338501900434494,
"learning_rate": 4.7993131791421385e-06,
"loss": 0.4344,
"step": 391
},
{
"epoch": 1.6872309899569584,
"grad_norm": 0.3199228048324585,
"learning_rate": 4.7742434388334815e-06,
"loss": 0.4171,
"step": 392
},
{
"epoch": 1.691535150645624,
"grad_norm": 0.3764464855194092,
"learning_rate": 4.749179384311331e-06,
"loss": 0.3928,
"step": 393
},
{
"epoch": 1.6958393113342898,
"grad_norm": 0.3166040778160095,
"learning_rate": 4.724121646825838e-06,
"loss": 0.4239,
"step": 394
},
{
"epoch": 1.7001434720229556,
"grad_norm": 0.40935027599334717,
"learning_rate": 4.699070857468052e-06,
"loss": 0.4921,
"step": 395
},
{
"epoch": 1.7044476327116214,
"grad_norm": 0.3424462080001831,
"learning_rate": 4.674027647154037e-06,
"loss": 0.4045,
"step": 396
},
{
"epoch": 1.708751793400287,
"grad_norm": 0.3410986065864563,
"learning_rate": 4.648992646608968e-06,
"loss": 0.4157,
"step": 397
},
{
"epoch": 1.7130559540889525,
"grad_norm": 0.3328428864479065,
"learning_rate": 4.623966486351257e-06,
"loss": 0.4137,
"step": 398
},
{
"epoch": 1.7173601147776183,
"grad_norm": 0.3369297385215759,
"learning_rate": 4.598949796676672e-06,
"loss": 0.4417,
"step": 399
},
{
"epoch": 1.721664275466284,
"grad_norm": 0.3781452476978302,
"learning_rate": 4.573943207642452e-06,
"loss": 0.4371,
"step": 400
},
{
"epoch": 1.7259684361549499,
"grad_norm": 0.31260576844215393,
"learning_rate": 4.548947349051452e-06,
"loss": 0.4133,
"step": 401
},
{
"epoch": 1.7302725968436155,
"grad_norm": 0.3287489712238312,
"learning_rate": 4.523962850436276e-06,
"loss": 0.3952,
"step": 402
},
{
"epoch": 1.7345767575322812,
"grad_norm": 0.357338011264801,
"learning_rate": 4.498990341043419e-06,
"loss": 0.4311,
"step": 403
},
{
"epoch": 1.7388809182209468,
"grad_norm": 0.33581385016441345,
"learning_rate": 4.474030449817423e-06,
"loss": 0.4234,
"step": 404
},
{
"epoch": 1.7431850789096126,
"grad_norm": 0.32600918412208557,
"learning_rate": 4.449083805385037e-06,
"loss": 0.4112,
"step": 405
},
{
"epoch": 1.7474892395982784,
"grad_norm": 0.33547377586364746,
"learning_rate": 4.424151036039381e-06,
"loss": 0.4262,
"step": 406
},
{
"epoch": 1.7517934002869442,
"grad_norm": 0.3635512888431549,
"learning_rate": 4.3992327697241225e-06,
"loss": 0.4538,
"step": 407
},
{
"epoch": 1.7560975609756098,
"grad_norm": 0.34227895736694336,
"learning_rate": 4.3743296340176694e-06,
"loss": 0.4468,
"step": 408
},
{
"epoch": 1.7604017216642753,
"grad_norm": 0.280943363904953,
"learning_rate": 4.3494422561173515e-06,
"loss": 0.3928,
"step": 409
},
{
"epoch": 1.7647058823529411,
"grad_norm": 0.36496126651763916,
"learning_rate": 4.3245712628236356e-06,
"loss": 0.4792,
"step": 410
},
{
"epoch": 1.769010043041607,
"grad_norm": 0.34355199337005615,
"learning_rate": 4.299717280524329e-06,
"loss": 0.3907,
"step": 411
},
{
"epoch": 1.7733142037302727,
"grad_norm": 0.3443046510219574,
"learning_rate": 4.274880935178817e-06,
"loss": 0.433,
"step": 412
},
{
"epoch": 1.7776183644189383,
"grad_norm": 0.3240680694580078,
"learning_rate": 4.250062852302283e-06,
"loss": 0.4214,
"step": 413
},
{
"epoch": 1.781922525107604,
"grad_norm": 0.3351806700229645,
"learning_rate": 4.225263656949961e-06,
"loss": 0.4287,
"step": 414
},
{
"epoch": 1.7862266857962696,
"grad_norm": 0.30206426978111267,
"learning_rate": 4.200483973701401e-06,
"loss": 0.3835,
"step": 415
},
{
"epoch": 1.7905308464849354,
"grad_norm": 0.3471432328224182,
"learning_rate": 4.175724426644724e-06,
"loss": 0.4838,
"step": 416
},
{
"epoch": 1.7948350071736012,
"grad_norm": 0.30823495984077454,
"learning_rate": 4.150985639360914e-06,
"loss": 0.3922,
"step": 417
},
{
"epoch": 1.799139167862267,
"grad_norm": 0.3459091782569885,
"learning_rate": 4.12626823490811e-06,
"loss": 0.444,
"step": 418
},
{
"epoch": 1.8034433285509326,
"grad_norm": 0.35498541593551636,
"learning_rate": 4.1015728358059185e-06,
"loss": 0.4414,
"step": 419
},
{
"epoch": 1.8077474892395982,
"grad_norm": 0.32903069257736206,
"learning_rate": 4.076900064019721e-06,
"loss": 0.4428,
"step": 420
},
{
"epoch": 1.812051649928264,
"grad_norm": 0.3732050955295563,
"learning_rate": 4.052250540945029e-06,
"loss": 0.4471,
"step": 421
},
{
"epoch": 1.8163558106169297,
"grad_norm": 0.30928969383239746,
"learning_rate": 4.02762488739182e-06,
"loss": 0.3788,
"step": 422
},
{
"epoch": 1.8206599713055955,
"grad_norm": 0.3515232503414154,
"learning_rate": 4.003023723568903e-06,
"loss": 0.4587,
"step": 423
},
{
"epoch": 1.824964131994261,
"grad_norm": 0.32477593421936035,
"learning_rate": 3.978447669068309e-06,
"loss": 0.4536,
"step": 424
},
{
"epoch": 1.8292682926829267,
"grad_norm": 0.35645440220832825,
"learning_rate": 3.953897342849673e-06,
"loss": 0.3964,
"step": 425
},
{
"epoch": 1.8335724533715925,
"grad_norm": 0.3410753309726715,
"learning_rate": 3.929373363224654e-06,
"loss": 0.4749,
"step": 426
},
{
"epoch": 1.8378766140602583,
"grad_norm": 0.3486350178718567,
"learning_rate": 3.904876347841354e-06,
"loss": 0.4494,
"step": 427
},
{
"epoch": 1.842180774748924,
"grad_norm": 0.3229897916316986,
"learning_rate": 3.8804069136687775e-06,
"loss": 0.4181,
"step": 428
},
{
"epoch": 1.8464849354375896,
"grad_norm": 0.3268226087093353,
"learning_rate": 3.8559656769812746e-06,
"loss": 0.4477,
"step": 429
},
{
"epoch": 1.8507890961262554,
"grad_norm": 0.31242066621780396,
"learning_rate": 3.8315532533430285e-06,
"loss": 0.4122,
"step": 430
},
{
"epoch": 1.855093256814921,
"grad_norm": 0.3421877920627594,
"learning_rate": 3.8071702575925594e-06,
"loss": 0.4484,
"step": 431
},
{
"epoch": 1.8593974175035868,
"grad_norm": 0.31596559286117554,
"learning_rate": 3.7828173038272266e-06,
"loss": 0.3822,
"step": 432
},
{
"epoch": 1.8637015781922526,
"grad_norm": 0.32587170600891113,
"learning_rate": 3.7584950053877646e-06,
"loss": 0.4536,
"step": 433
},
{
"epoch": 1.8680057388809184,
"grad_norm": 0.3373381197452545,
"learning_rate": 3.7342039748428473e-06,
"loss": 0.3904,
"step": 434
},
{
"epoch": 1.872309899569584,
"grad_norm": 0.33278679847717285,
"learning_rate": 3.709944823973647e-06,
"loss": 0.3923,
"step": 435
},
{
"epoch": 1.8766140602582495,
"grad_norm": 0.3359312415122986,
"learning_rate": 3.685718163758427e-06,
"loss": 0.464,
"step": 436
},
{
"epoch": 1.8809182209469153,
"grad_norm": 0.3137849271297455,
"learning_rate": 3.6615246043571674e-06,
"loss": 0.3741,
"step": 437
},
{
"epoch": 1.885222381635581,
"grad_norm": 0.3578011095523834,
"learning_rate": 3.6373647550961834e-06,
"loss": 0.4495,
"step": 438
},
{
"epoch": 1.8895265423242469,
"grad_norm": 0.39701202511787415,
"learning_rate": 3.61323922445278e-06,
"loss": 0.4643,
"step": 439
},
{
"epoch": 1.8938307030129125,
"grad_norm": 0.3316022455692291,
"learning_rate": 3.5891486200399413e-06,
"loss": 0.3997,
"step": 440
},
{
"epoch": 1.8981348637015782,
"grad_norm": 0.3504200279712677,
"learning_rate": 3.5650935485910103e-06,
"loss": 0.4349,
"step": 441
},
{
"epoch": 1.9024390243902438,
"grad_norm": 0.33614853024482727,
"learning_rate": 3.5410746159444165e-06,
"loss": 0.4419,
"step": 442
},
{
"epoch": 1.9067431850789096,
"grad_norm": 0.3440527319908142,
"learning_rate": 3.5170924270284166e-06,
"loss": 0.4213,
"step": 443
},
{
"epoch": 1.9110473457675754,
"grad_norm": 0.3421337902545929,
"learning_rate": 3.4931475858458634e-06,
"loss": 0.43,
"step": 444
},
{
"epoch": 1.9153515064562412,
"grad_norm": 0.35556983947753906,
"learning_rate": 3.469240695458983e-06,
"loss": 0.3998,
"step": 445
},
{
"epoch": 1.9196556671449068,
"grad_norm": 0.35267797112464905,
"learning_rate": 3.445372357974194e-06,
"loss": 0.4149,
"step": 446
},
{
"epoch": 1.9239598278335723,
"grad_norm": 0.331310898065567,
"learning_rate": 3.4215431745269463e-06,
"loss": 0.4086,
"step": 447
},
{
"epoch": 1.9282639885222381,
"grad_norm": 0.3272462785243988,
"learning_rate": 3.397753745266571e-06,
"loss": 0.4314,
"step": 448
},
{
"epoch": 1.932568149210904,
"grad_norm": 0.31959936022758484,
"learning_rate": 3.374004669341173e-06,
"loss": 0.4381,
"step": 449
},
{
"epoch": 1.9368723098995697,
"grad_norm": 0.340181827545166,
"learning_rate": 3.350296544882543e-06,
"loss": 0.4569,
"step": 450
},
{
"epoch": 1.9411764705882353,
"grad_norm": 0.3325444459915161,
"learning_rate": 3.326629968991083e-06,
"loss": 0.4094,
"step": 451
},
{
"epoch": 1.9454806312769009,
"grad_norm": 0.32819491624832153,
"learning_rate": 3.303005537720778e-06,
"loss": 0.4435,
"step": 452
},
{
"epoch": 1.9497847919655666,
"grad_norm": 0.3121723234653473,
"learning_rate": 3.2794238460641837e-06,
"loss": 0.4096,
"step": 453
},
{
"epoch": 1.9540889526542324,
"grad_norm": 0.3106290400028229,
"learning_rate": 3.255885487937431e-06,
"loss": 0.3645,
"step": 454
},
{
"epoch": 1.9583931133428982,
"grad_norm": 0.3480536937713623,
"learning_rate": 3.2323910561652798e-06,
"loss": 0.4206,
"step": 455
},
{
"epoch": 1.9626972740315638,
"grad_norm": 0.3264220654964447,
"learning_rate": 3.2089411424661864e-06,
"loss": 0.4522,
"step": 456
},
{
"epoch": 1.9670014347202296,
"grad_norm": 0.29471534490585327,
"learning_rate": 3.185536337437393e-06,
"loss": 0.4188,
"step": 457
},
{
"epoch": 1.9713055954088952,
"grad_norm": 0.31906241178512573,
"learning_rate": 3.1621772305400603e-06,
"loss": 0.4532,
"step": 458
},
{
"epoch": 1.975609756097561,
"grad_norm": 0.30077633261680603,
"learning_rate": 3.138864410084419e-06,
"loss": 0.4052,
"step": 459
},
{
"epoch": 1.9799139167862267,
"grad_norm": 0.30726587772369385,
"learning_rate": 3.1155984632149565e-06,
"loss": 0.3927,
"step": 460
},
{
"epoch": 1.9842180774748925,
"grad_norm": 0.3261038661003113,
"learning_rate": 3.0923799758956265e-06,
"loss": 0.4457,
"step": 461
},
{
"epoch": 1.988522238163558,
"grad_norm": 0.2910383939743042,
"learning_rate": 3.0692095328950843e-06,
"loss": 0.404,
"step": 462
},
{
"epoch": 1.9928263988522237,
"grad_norm": 0.2886095345020294,
"learning_rate": 3.0460877177719763e-06,
"loss": 0.4016,
"step": 463
},
{
"epoch": 1.9971305595408895,
"grad_norm": 0.31026142835617065,
"learning_rate": 3.023015112860228e-06,
"loss": 0.4471,
"step": 464
},
{
"epoch": 2.0014347202295553,
"grad_norm": 0.670920193195343,
"learning_rate": 2.9999922992543777e-06,
"loss": 0.6493,
"step": 465
},
{
"epoch": 2.005738880918221,
"grad_norm": 0.3565112054347992,
"learning_rate": 2.977019856794955e-06,
"loss": 0.3958,
"step": 466
},
{
"epoch": 2.010043041606887,
"grad_norm": 0.38834989070892334,
"learning_rate": 2.9540983640538635e-06,
"loss": 0.4022,
"step": 467
},
{
"epoch": 2.014347202295552,
"grad_norm": 0.35440927743911743,
"learning_rate": 2.93122839831981e-06,
"loss": 0.4266,
"step": 468
},
{
"epoch": 2.018651362984218,
"grad_norm": 0.3439231216907501,
"learning_rate": 2.908410535583777e-06,
"loss": 0.4115,
"step": 469
},
{
"epoch": 2.022955523672884,
"grad_norm": 0.30981284379959106,
"learning_rate": 2.8856453505245018e-06,
"loss": 0.3427,
"step": 470
},
{
"epoch": 2.0272596843615496,
"grad_norm": 0.3863826096057892,
"learning_rate": 2.8629334164940127e-06,
"loss": 0.407,
"step": 471
},
{
"epoch": 2.0315638450502154,
"grad_norm": 0.39452433586120605,
"learning_rate": 2.840275305503186e-06,
"loss": 0.3971,
"step": 472
},
{
"epoch": 2.0358680057388807,
"grad_norm": 0.35829833149909973,
"learning_rate": 2.817671588207338e-06,
"loss": 0.4088,
"step": 473
},
{
"epoch": 2.0401721664275465,
"grad_norm": 0.31108009815216064,
"learning_rate": 2.7951228338918506e-06,
"loss": 0.3774,
"step": 474
},
{
"epoch": 2.0444763271162123,
"grad_norm": 0.4098561704158783,
"learning_rate": 2.77262961045784e-06,
"loss": 0.4484,
"step": 475
},
{
"epoch": 2.048780487804878,
"grad_norm": 0.3637750744819641,
"learning_rate": 2.7501924844078538e-06,
"loss": 0.3895,
"step": 476
},
{
"epoch": 2.053084648493544,
"grad_norm": 0.3582223653793335,
"learning_rate": 2.7278120208315927e-06,
"loss": 0.4144,
"step": 477
},
{
"epoch": 2.0573888091822097,
"grad_norm": 0.3238336443901062,
"learning_rate": 2.7054887833916933e-06,
"loss": 0.3799,
"step": 478
},
{
"epoch": 2.061692969870875,
"grad_norm": 0.32167771458625793,
"learning_rate": 2.6832233343095225e-06,
"loss": 0.3839,
"step": 479
},
{
"epoch": 2.065997130559541,
"grad_norm": 0.30850154161453247,
"learning_rate": 2.6610162343510183e-06,
"loss": 0.3808,
"step": 480
},
{
"epoch": 2.0703012912482066,
"grad_norm": 0.38707244396209717,
"learning_rate": 2.6388680428125657e-06,
"loss": 0.4356,
"step": 481
},
{
"epoch": 2.0746054519368724,
"grad_norm": 0.32622581720352173,
"learning_rate": 2.616779317506921e-06,
"loss": 0.3889,
"step": 482
},
{
"epoch": 2.078909612625538,
"grad_norm": 0.35619160532951355,
"learning_rate": 2.594750614749148e-06,
"loss": 0.4568,
"step": 483
},
{
"epoch": 2.0832137733142035,
"grad_norm": 0.3068256378173828,
"learning_rate": 2.572782489342617e-06,
"loss": 0.3612,
"step": 484
},
{
"epoch": 2.0875179340028693,
"grad_norm": 0.3251802623271942,
"learning_rate": 2.5508754945650305e-06,
"loss": 0.4148,
"step": 485
},
{
"epoch": 2.091822094691535,
"grad_norm": 0.28180989623069763,
"learning_rate": 2.5290301821544826e-06,
"loss": 0.3438,
"step": 486
},
{
"epoch": 2.096126255380201,
"grad_norm": 0.3368677794933319,
"learning_rate": 2.5072471022955703e-06,
"loss": 0.4339,
"step": 487
},
{
"epoch": 2.1004304160688667,
"grad_norm": 0.3352259397506714,
"learning_rate": 2.4855268036055346e-06,
"loss": 0.4137,
"step": 488
},
{
"epoch": 2.104734576757532,
"grad_norm": 0.3315313160419464,
"learning_rate": 2.4638698331204404e-06,
"loss": 0.4156,
"step": 489
},
{
"epoch": 2.109038737446198,
"grad_norm": 0.28988370299339294,
"learning_rate": 2.4422767362814045e-06,
"loss": 0.3509,
"step": 490
},
{
"epoch": 2.1133428981348636,
"grad_norm": 0.31385964155197144,
"learning_rate": 2.420748056920856e-06,
"loss": 0.3794,
"step": 491
},
{
"epoch": 2.1176470588235294,
"grad_norm": 0.3286735713481903,
"learning_rate": 2.3992843372488357e-06,
"loss": 0.4653,
"step": 492
},
{
"epoch": 2.1219512195121952,
"grad_norm": 0.29086706042289734,
"learning_rate": 2.3778861178393453e-06,
"loss": 0.3721,
"step": 493
},
{
"epoch": 2.126255380200861,
"grad_norm": 0.3042701482772827,
"learning_rate": 2.3565539376167295e-06,
"loss": 0.3946,
"step": 494
},
{
"epoch": 2.1305595408895264,
"grad_norm": 0.29335102438926697,
"learning_rate": 2.3352883338421085e-06,
"loss": 0.383,
"step": 495
},
{
"epoch": 2.134863701578192,
"grad_norm": 0.3304314911365509,
"learning_rate": 2.3140898420998425e-06,
"loss": 0.4103,
"step": 496
},
{
"epoch": 2.139167862266858,
"grad_norm": 0.31538569927215576,
"learning_rate": 2.2929589962840375e-06,
"loss": 0.4033,
"step": 497
},
{
"epoch": 2.1434720229555237,
"grad_norm": 0.3131985366344452,
"learning_rate": 2.271896328585114e-06,
"loss": 0.387,
"step": 498
},
{
"epoch": 2.1477761836441895,
"grad_norm": 0.3355647027492523,
"learning_rate": 2.2509023694763844e-06,
"loss": 0.3912,
"step": 499
},
{
"epoch": 2.152080344332855,
"grad_norm": 0.2865258753299713,
"learning_rate": 2.2299776477007073e-06,
"loss": 0.3855,
"step": 500
},
{
"epoch": 2.1563845050215207,
"grad_norm": 0.29301708936691284,
"learning_rate": 2.2091226902571673e-06,
"loss": 0.3829,
"step": 501
},
{
"epoch": 2.1606886657101865,
"grad_norm": 0.3320524990558624,
"learning_rate": 2.1883380223878004e-06,
"loss": 0.4205,
"step": 502
},
{
"epoch": 2.1649928263988523,
"grad_norm": 0.32117411494255066,
"learning_rate": 2.1676241675643627e-06,
"loss": 0.4102,
"step": 503
},
{
"epoch": 2.169296987087518,
"grad_norm": 0.3084646165370941,
"learning_rate": 2.1469816474751566e-06,
"loss": 0.408,
"step": 504
},
{
"epoch": 2.173601147776184,
"grad_norm": 0.30409905314445496,
"learning_rate": 2.1264109820118783e-06,
"loss": 0.3981,
"step": 505
},
{
"epoch": 2.177905308464849,
"grad_norm": 0.28728926181793213,
"learning_rate": 2.105912689256533e-06,
"loss": 0.3768,
"step": 506
},
{
"epoch": 2.182209469153515,
"grad_norm": 0.30246901512145996,
"learning_rate": 2.0854872854683877e-06,
"loss": 0.3924,
"step": 507
},
{
"epoch": 2.186513629842181,
"grad_norm": 0.31316882371902466,
"learning_rate": 2.0651352850709656e-06,
"loss": 0.4331,
"step": 508
},
{
"epoch": 2.1908177905308466,
"grad_norm": 0.31881269812583923,
"learning_rate": 2.0448572006390875e-06,
"loss": 0.3617,
"step": 509
},
{
"epoch": 2.1951219512195124,
"grad_norm": 0.30993711948394775,
"learning_rate": 2.0246535428859652e-06,
"loss": 0.4185,
"step": 510
},
{
"epoch": 2.1994261119081777,
"grad_norm": 0.32910722494125366,
"learning_rate": 2.0045248206503454e-06,
"loss": 0.4396,
"step": 511
},
{
"epoch": 2.2037302725968435,
"grad_norm": 0.31692197918891907,
"learning_rate": 1.984471540883679e-06,
"loss": 0.4397,
"step": 512
},
{
"epoch": 2.2080344332855093,
"grad_norm": 0.26367130875587463,
"learning_rate": 1.964494208637369e-06,
"loss": 0.3652,
"step": 513
},
{
"epoch": 2.212338593974175,
"grad_norm": 0.2917594611644745,
"learning_rate": 1.9445933270500444e-06,
"loss": 0.4279,
"step": 514
},
{
"epoch": 2.216642754662841,
"grad_norm": 0.34075671434402466,
"learning_rate": 1.9247693973348834e-06,
"loss": 0.4294,
"step": 515
},
{
"epoch": 2.2209469153515062,
"grad_norm": 0.2909022271633148,
"learning_rate": 1.905022918766995e-06,
"loss": 0.3783,
"step": 516
},
{
"epoch": 2.225251076040172,
"grad_norm": 0.3097451627254486,
"learning_rate": 1.8853543886708498e-06,
"loss": 0.4208,
"step": 517
},
{
"epoch": 2.229555236728838,
"grad_norm": 0.2984945774078369,
"learning_rate": 1.8657643024077431e-06,
"loss": 0.4076,
"step": 518
},
{
"epoch": 2.2338593974175036,
"grad_norm": 0.2853682041168213,
"learning_rate": 1.8462531533633238e-06,
"loss": 0.3759,
"step": 519
},
{
"epoch": 2.2381635581061694,
"grad_norm": 0.2992599606513977,
"learning_rate": 1.8268214329351797e-06,
"loss": 0.3984,
"step": 520
},
{
"epoch": 2.242467718794835,
"grad_norm": 0.28934013843536377,
"learning_rate": 1.8074696305204397e-06,
"loss": 0.399,
"step": 521
},
{
"epoch": 2.2467718794835005,
"grad_norm": 0.30769601464271545,
"learning_rate": 1.7881982335034625e-06,
"loss": 0.4008,
"step": 522
},
{
"epoch": 2.2510760401721663,
"grad_norm": 0.3330642879009247,
"learning_rate": 1.7690077272435636e-06,
"loss": 0.4239,
"step": 523
},
{
"epoch": 2.255380200860832,
"grad_norm": 0.28844162821769714,
"learning_rate": 1.7498985950627794e-06,
"loss": 0.4153,
"step": 524
},
{
"epoch": 2.259684361549498,
"grad_norm": 0.2728285789489746,
"learning_rate": 1.7308713182337044e-06,
"loss": 0.4052,
"step": 525
},
{
"epoch": 2.2639885222381637,
"grad_norm": 0.32398754358291626,
"learning_rate": 1.7119263759673677e-06,
"loss": 0.4153,
"step": 526
},
{
"epoch": 2.2682926829268295,
"grad_norm": 0.30444568395614624,
"learning_rate": 1.6930642454011647e-06,
"loss": 0.3952,
"step": 527
},
{
"epoch": 2.272596843615495,
"grad_norm": 0.2704574167728424,
"learning_rate": 1.6742854015868349e-06,
"loss": 0.3686,
"step": 528
},
{
"epoch": 2.2769010043041606,
"grad_norm": 0.30521517992019653,
"learning_rate": 1.655590317478501e-06,
"loss": 0.4278,
"step": 529
},
{
"epoch": 2.2812051649928264,
"grad_norm": 0.28752925992012024,
"learning_rate": 1.6369794639207626e-06,
"loss": 0.4023,
"step": 530
},
{
"epoch": 2.2855093256814922,
"grad_norm": 0.2987408936023712,
"learning_rate": 1.6184533096368277e-06,
"loss": 0.414,
"step": 531
},
{
"epoch": 2.289813486370158,
"grad_norm": 0.28139418363571167,
"learning_rate": 1.6000123212167158e-06,
"loss": 0.3826,
"step": 532
},
{
"epoch": 2.2941176470588234,
"grad_norm": 0.3225473165512085,
"learning_rate": 1.581656963105504e-06,
"loss": 0.4275,
"step": 533
},
{
"epoch": 2.298421807747489,
"grad_norm": 0.2836931049823761,
"learning_rate": 1.5633876975916261e-06,
"loss": 0.3899,
"step": 534
},
{
"epoch": 2.302725968436155,
"grad_norm": 0.31612229347229004,
"learning_rate": 1.5452049847952338e-06,
"loss": 0.3921,
"step": 535
},
{
"epoch": 2.3070301291248207,
"grad_norm": 0.2987075448036194,
"learning_rate": 1.5271092826566108e-06,
"loss": 0.4058,
"step": 536
},
{
"epoch": 2.3113342898134865,
"grad_norm": 0.27526581287384033,
"learning_rate": 1.5091010469246303e-06,
"loss": 0.3879,
"step": 537
},
{
"epoch": 2.315638450502152,
"grad_norm": 0.28727975487709045,
"learning_rate": 1.4911807311452874e-06,
"loss": 0.389,
"step": 538
},
{
"epoch": 2.3199426111908177,
"grad_norm": 0.294012188911438,
"learning_rate": 1.4733487866502698e-06,
"loss": 0.3775,
"step": 539
},
{
"epoch": 2.3242467718794835,
"grad_norm": 0.29785969853401184,
"learning_rate": 1.4556056625455922e-06,
"loss": 0.404,
"step": 540
},
{
"epoch": 2.3285509325681493,
"grad_norm": 0.322480171918869,
"learning_rate": 1.4379518057002834e-06,
"loss": 0.4039,
"step": 541
},
{
"epoch": 2.332855093256815,
"grad_norm": 0.29027432203292847,
"learning_rate": 1.4203876607351347e-06,
"loss": 0.4156,
"step": 542
},
{
"epoch": 2.3371592539454804,
"grad_norm": 0.30819234251976013,
"learning_rate": 1.4029136700115031e-06,
"loss": 0.4104,
"step": 543
},
{
"epoch": 2.341463414634146,
"grad_norm": 0.29122260212898254,
"learning_rate": 1.3855302736201686e-06,
"loss": 0.3695,
"step": 544
},
{
"epoch": 2.345767575322812,
"grad_norm": 0.2812859117984772,
"learning_rate": 1.3682379093702447e-06,
"loss": 0.3635,
"step": 545
},
{
"epoch": 2.350071736011478,
"grad_norm": 0.3218896687030792,
"learning_rate": 1.3510370127781635e-06,
"loss": 0.4021,
"step": 546
},
{
"epoch": 2.3543758967001436,
"grad_norm": 0.3179861009120941,
"learning_rate": 1.3339280170566959e-06,
"loss": 0.3962,
"step": 547
},
{
"epoch": 2.3586800573888094,
"grad_norm": 0.2889609932899475,
"learning_rate": 1.3169113531040462e-06,
"loss": 0.4131,
"step": 548
},
{
"epoch": 2.3629842180774747,
"grad_norm": 0.2856752574443817,
"learning_rate": 1.2999874494930004e-06,
"loss": 0.3709,
"step": 549
},
{
"epoch": 2.3672883787661405,
"grad_norm": 0.3044317364692688,
"learning_rate": 1.2831567324601325e-06,
"loss": 0.3988,
"step": 550
},
{
"epoch": 2.3715925394548063,
"grad_norm": 0.3010687828063965,
"learning_rate": 1.266419625895064e-06,
"loss": 0.4016,
"step": 551
},
{
"epoch": 2.375896700143472,
"grad_norm": 0.3065761923789978,
"learning_rate": 1.2497765513297976e-06,
"loss": 0.4227,
"step": 552
},
{
"epoch": 2.380200860832138,
"grad_norm": 0.28684499859809875,
"learning_rate": 1.2332279279280907e-06,
"loss": 0.377,
"step": 553
},
{
"epoch": 2.3845050215208037,
"grad_norm": 0.2905068099498749,
"learning_rate": 1.2167741724749026e-06,
"loss": 0.389,
"step": 554
},
{
"epoch": 2.388809182209469,
"grad_norm": 0.3115670680999756,
"learning_rate": 1.2004156993659028e-06,
"loss": 0.4171,
"step": 555
},
{
"epoch": 2.393113342898135,
"grad_norm": 0.29904720187187195,
"learning_rate": 1.1841529205970281e-06,
"loss": 0.3995,
"step": 556
},
{
"epoch": 2.3974175035868006,
"grad_norm": 0.2813907861709595,
"learning_rate": 1.1679862457541052e-06,
"loss": 0.3995,
"step": 557
},
{
"epoch": 2.4017216642754664,
"grad_norm": 0.3307220935821533,
"learning_rate": 1.1519160820025382e-06,
"loss": 0.412,
"step": 558
},
{
"epoch": 2.406025824964132,
"grad_norm": 0.30656975507736206,
"learning_rate": 1.1359428340770567e-06,
"loss": 0.4061,
"step": 559
},
{
"epoch": 2.4103299856527975,
"grad_norm": 0.30962732434272766,
"learning_rate": 1.1200669042715163e-06,
"loss": 0.4016,
"step": 560
},
{
"epoch": 2.4146341463414633,
"grad_norm": 0.29920291900634766,
"learning_rate": 1.104288692428766e-06,
"loss": 0.3848,
"step": 561
},
{
"epoch": 2.418938307030129,
"grad_norm": 0.32098498940467834,
"learning_rate": 1.0886085959305915e-06,
"loss": 0.4224,
"step": 562
},
{
"epoch": 2.423242467718795,
"grad_norm": 0.28412872552871704,
"learning_rate": 1.0730270096876876e-06,
"loss": 0.3759,
"step": 563
},
{
"epoch": 2.4275466284074607,
"grad_norm": 0.336220920085907,
"learning_rate": 1.057544326129723e-06,
"loss": 0.4437,
"step": 564
},
{
"epoch": 2.431850789096126,
"grad_norm": 0.2975854277610779,
"learning_rate": 1.0421609351954599e-06,
"loss": 0.3788,
"step": 565
},
{
"epoch": 2.436154949784792,
"grad_norm": 0.2922147214412689,
"learning_rate": 1.026877224322923e-06,
"loss": 0.4049,
"step": 566
},
{
"epoch": 2.4404591104734576,
"grad_norm": 0.3053329288959503,
"learning_rate": 1.0116935784396482e-06,
"loss": 0.4196,
"step": 567
},
{
"epoch": 2.4447632711621234,
"grad_norm": 0.2925123870372772,
"learning_rate": 9.966103799529891e-07,
"loss": 0.3872,
"step": 568
},
{
"epoch": 2.4490674318507892,
"grad_norm": 0.25717809796333313,
"learning_rate": 9.816280087404851e-07,
"loss": 0.3512,
"step": 569
},
{
"epoch": 2.4533715925394546,
"grad_norm": 0.2897495925426483,
"learning_rate": 9.66746842140287e-07,
"loss": 0.451,
"step": 570
},
{
"epoch": 2.4576757532281204,
"grad_norm": 0.279236376285553,
"learning_rate": 9.519672549416659e-07,
"loss": 0.3622,
"step": 571
},
{
"epoch": 2.461979913916786,
"grad_norm": 0.29578229784965515,
"learning_rate": 9.372896193755621e-07,
"loss": 0.419,
"step": 572
},
{
"epoch": 2.466284074605452,
"grad_norm": 0.28431037068367004,
"learning_rate": 9.227143051052162e-07,
"loss": 0.3895,
"step": 573
},
{
"epoch": 2.4705882352941178,
"grad_norm": 0.30916470289230347,
"learning_rate": 9.082416792168608e-07,
"loss": 0.3914,
"step": 574
},
{
"epoch": 2.4748923959827835,
"grad_norm": 0.2944723963737488,
"learning_rate": 8.938721062104727e-07,
"loss": 0.4186,
"step": 575
},
{
"epoch": 2.479196556671449,
"grad_norm": 0.2983667254447937,
"learning_rate": 8.7960594799059e-07,
"loss": 0.4174,
"step": 576
},
{
"epoch": 2.4835007173601147,
"grad_norm": 0.3032497763633728,
"learning_rate": 8.654435638572e-07,
"loss": 0.3805,
"step": 577
},
{
"epoch": 2.4878048780487805,
"grad_norm": 0.3096209466457367,
"learning_rate": 8.513853104966951e-07,
"loss": 0.4586,
"step": 578
},
{
"epoch": 2.4921090387374463,
"grad_norm": 0.285931259393692,
"learning_rate": 8.374315419728784e-07,
"loss": 0.365,
"step": 579
},
{
"epoch": 2.496413199426112,
"grad_norm": 0.2848215401172638,
"learning_rate": 8.235826097180566e-07,
"loss": 0.4182,
"step": 580
},
{
"epoch": 2.500717360114778,
"grad_norm": 0.299034059047699,
"learning_rate": 8.098388625241854e-07,
"loss": 0.4259,
"step": 581
},
{
"epoch": 2.505021520803443,
"grad_norm": 0.29463037848472595,
"learning_rate": 7.962006465340821e-07,
"loss": 0.3974,
"step": 582
},
{
"epoch": 2.509325681492109,
"grad_norm": 0.2773649990558624,
"learning_rate": 7.8266830523271e-07,
"loss": 0.3681,
"step": 583
},
{
"epoch": 2.513629842180775,
"grad_norm": 0.2890886068344116,
"learning_rate": 7.692421794385313e-07,
"loss": 0.4115,
"step": 584
},
{
"epoch": 2.5179340028694406,
"grad_norm": 0.2763752043247223,
"learning_rate": 7.559226072949166e-07,
"loss": 0.3985,
"step": 585
},
{
"epoch": 2.5222381635581064,
"grad_norm": 0.28498420119285583,
"learning_rate": 7.427099242616348e-07,
"loss": 0.3819,
"step": 586
},
{
"epoch": 2.5265423242467717,
"grad_norm": 0.2716716527938843,
"learning_rate": 7.296044631064014e-07,
"loss": 0.4024,
"step": 587
},
{
"epoch": 2.5308464849354375,
"grad_norm": 0.295413613319397,
"learning_rate": 7.166065538964955e-07,
"loss": 0.4236,
"step": 588
},
{
"epoch": 2.5351506456241033,
"grad_norm": 0.27543219923973083,
"learning_rate": 7.037165239904514e-07,
"loss": 0.41,
"step": 589
},
{
"epoch": 2.539454806312769,
"grad_norm": 0.2738160789012909,
"learning_rate": 6.909346980298093e-07,
"loss": 0.4002,
"step": 590
},
{
"epoch": 2.543758967001435,
"grad_norm": 0.2854576110839844,
"learning_rate": 6.782613979309443e-07,
"loss": 0.3978,
"step": 591
},
{
"epoch": 2.5480631276901002,
"grad_norm": 0.2689387798309326,
"learning_rate": 6.656969428769567e-07,
"loss": 0.3963,
"step": 592
},
{
"epoch": 2.552367288378766,
"grad_norm": 0.2699349522590637,
"learning_rate": 6.532416493096272e-07,
"loss": 0.4085,
"step": 593
},
{
"epoch": 2.556671449067432,
"grad_norm": 0.25454413890838623,
"learning_rate": 6.408958309214597e-07,
"loss": 0.3766,
"step": 594
},
{
"epoch": 2.5609756097560976,
"grad_norm": 0.3066865801811218,
"learning_rate": 6.286597986477683e-07,
"loss": 0.4144,
"step": 595
},
{
"epoch": 2.5652797704447634,
"grad_norm": 0.2773991823196411,
"learning_rate": 6.165338606588517e-07,
"loss": 0.4158,
"step": 596
},
{
"epoch": 2.5695839311334288,
"grad_norm": 0.28849199414253235,
"learning_rate": 6.045183223522339e-07,
"loss": 0.3781,
"step": 597
},
{
"epoch": 2.5738880918220945,
"grad_norm": 0.32376885414123535,
"learning_rate": 5.926134863449712e-07,
"loss": 0.3767,
"step": 598
},
{
"epoch": 2.5781922525107603,
"grad_norm": 0.2920165956020355,
"learning_rate": 5.808196524660253e-07,
"loss": 0.4263,
"step": 599
},
{
"epoch": 2.582496413199426,
"grad_norm": 0.2848154604434967,
"learning_rate": 5.691371177487215e-07,
"loss": 0.3821,
"step": 600
},
{
"epoch": 2.586800573888092,
"grad_norm": 0.28486862778663635,
"learning_rate": 5.575661764232593e-07,
"loss": 0.4167,
"step": 601
},
{
"epoch": 2.5911047345767573,
"grad_norm": 0.28236979246139526,
"learning_rate": 5.461071199093048e-07,
"loss": 0.3902,
"step": 602
},
{
"epoch": 2.5954088952654235,
"grad_norm": 0.2632196545600891,
"learning_rate": 5.347602368086563e-07,
"loss": 0.3849,
"step": 603
},
{
"epoch": 2.599713055954089,
"grad_norm": 0.29081490635871887,
"learning_rate": 5.235258128979676e-07,
"loss": 0.4345,
"step": 604
},
{
"epoch": 2.6040172166427547,
"grad_norm": 0.2874118983745575,
"learning_rate": 5.124041311215544e-07,
"loss": 0.3843,
"step": 605
},
{
"epoch": 2.6083213773314204,
"grad_norm": 0.2967386841773987,
"learning_rate": 5.0139547158427e-07,
"loss": 0.3959,
"step": 606
},
{
"epoch": 2.6126255380200862,
"grad_norm": 0.27277684211730957,
"learning_rate": 4.905001115444475e-07,
"loss": 0.4323,
"step": 607
},
{
"epoch": 2.616929698708752,
"grad_norm": 0.2767053544521332,
"learning_rate": 4.797183254069176e-07,
"loss": 0.3858,
"step": 608
},
{
"epoch": 2.6212338593974174,
"grad_norm": 0.28239718079566956,
"learning_rate": 4.690503847160982e-07,
"loss": 0.4297,
"step": 609
},
{
"epoch": 2.625538020086083,
"grad_norm": 0.2802097201347351,
"learning_rate": 4.5849655814915683e-07,
"loss": 0.3975,
"step": 610
},
{
"epoch": 2.629842180774749,
"grad_norm": 0.2911134958267212,
"learning_rate": 4.4805711150924304e-07,
"loss": 0.3648,
"step": 611
},
{
"epoch": 2.6341463414634148,
"grad_norm": 0.30303090810775757,
"learning_rate": 4.3773230771879004e-07,
"loss": 0.4397,
"step": 612
},
{
"epoch": 2.6384505021520805,
"grad_norm": 0.29142701625823975,
"learning_rate": 4.2752240681290027e-07,
"loss": 0.3775,
"step": 613
},
{
"epoch": 2.642754662840746,
"grad_norm": 0.28723323345184326,
"learning_rate": 4.1742766593278974e-07,
"loss": 0.4052,
"step": 614
},
{
"epoch": 2.6470588235294117,
"grad_norm": 0.2821033000946045,
"learning_rate": 4.074483393193135e-07,
"loss": 0.3822,
"step": 615
},
{
"epoch": 2.6513629842180775,
"grad_norm": 0.28537237644195557,
"learning_rate": 3.9758467830656623e-07,
"loss": 0.3862,
"step": 616
},
{
"epoch": 2.6556671449067433,
"grad_norm": 0.29671528935432434,
"learning_rate": 3.8783693131554836e-07,
"loss": 0.3844,
"step": 617
},
{
"epoch": 2.659971305595409,
"grad_norm": 0.6448749303817749,
"learning_rate": 3.782053438479094e-07,
"loss": 0.4243,
"step": 618
},
{
"epoch": 2.6642754662840744,
"grad_norm": 0.280321329832077,
"learning_rate": 3.686901584797675e-07,
"loss": 0.3809,
"step": 619
},
{
"epoch": 2.66857962697274,
"grad_norm": 0.2735440135002136,
"learning_rate": 3.5929161485559694e-07,
"loss": 0.3873,
"step": 620
},
{
"epoch": 2.672883787661406,
"grad_norm": 0.2726655602455139,
"learning_rate": 3.5000994968219406e-07,
"loss": 0.3545,
"step": 621
},
{
"epoch": 2.677187948350072,
"grad_norm": 0.30105406045913696,
"learning_rate": 3.4084539672271764e-07,
"loss": 0.4052,
"step": 622
},
{
"epoch": 2.6814921090387376,
"grad_norm": 0.2675243318080902,
"learning_rate": 3.3179818679079936e-07,
"loss": 0.3843,
"step": 623
},
{
"epoch": 2.685796269727403,
"grad_norm": 0.281837522983551,
"learning_rate": 3.228685477447291e-07,
"loss": 0.3866,
"step": 624
},
{
"epoch": 2.6901004304160687,
"grad_norm": 0.27979665994644165,
"learning_rate": 3.140567044817172e-07,
"loss": 0.3993,
"step": 625
},
{
"epoch": 2.6944045911047345,
"grad_norm": 0.2708576023578644,
"learning_rate": 3.0536287893223603e-07,
"loss": 0.3876,
"step": 626
},
{
"epoch": 2.6987087517934003,
"grad_norm": 0.2992185950279236,
"learning_rate": 2.967872900544194e-07,
"loss": 0.3943,
"step": 627
},
{
"epoch": 2.703012912482066,
"grad_norm": 0.2894390821456909,
"learning_rate": 2.883301538285582e-07,
"loss": 0.4108,
"step": 628
},
{
"epoch": 2.7073170731707314,
"grad_norm": 0.27265292406082153,
"learning_rate": 2.799916832516575e-07,
"loss": 0.3659,
"step": 629
},
{
"epoch": 2.7116212338593977,
"grad_norm": 0.2490788698196411,
"learning_rate": 2.717720883320685e-07,
"loss": 0.3805,
"step": 630
},
{
"epoch": 2.715925394548063,
"grad_norm": 0.29212287068367004,
"learning_rate": 2.6367157608420347e-07,
"loss": 0.417,
"step": 631
},
{
"epoch": 2.720229555236729,
"grad_norm": 0.2623874247074127,
"learning_rate": 2.556903505233216e-07,
"loss": 0.3635,
"step": 632
},
{
"epoch": 2.7245337159253946,
"grad_norm": 0.28832924365997314,
"learning_rate": 2.4782861266038904e-07,
"loss": 0.4361,
"step": 633
},
{
"epoch": 2.7288378766140604,
"grad_norm": 0.2711504101753235,
"learning_rate": 2.4008656049701875e-07,
"loss": 0.4118,
"step": 634
},
{
"epoch": 2.733142037302726,
"grad_norm": 0.2761891782283783,
"learning_rate": 2.3246438902048196e-07,
"loss": 0.3608,
"step": 635
},
{
"epoch": 2.7374461979913915,
"grad_norm": 0.2847788631916046,
"learning_rate": 2.2496229019879635e-07,
"loss": 0.4244,
"step": 636
},
{
"epoch": 2.7417503586800573,
"grad_norm": 0.27423596382141113,
"learning_rate": 2.175804529758929e-07,
"loss": 0.4173,
"step": 637
},
{
"epoch": 2.746054519368723,
"grad_norm": 0.27934539318084717,
"learning_rate": 2.1031906326685946e-07,
"loss": 0.399,
"step": 638
},
{
"epoch": 2.750358680057389,
"grad_norm": 0.27947354316711426,
"learning_rate": 2.0317830395325255e-07,
"loss": 0.3802,
"step": 639
},
{
"epoch": 2.7546628407460547,
"grad_norm": 0.29531410336494446,
"learning_rate": 1.9615835487849677e-07,
"loss": 0.3867,
"step": 640
},
{
"epoch": 2.75896700143472,
"grad_norm": 0.29944562911987305,
"learning_rate": 1.8925939284335225e-07,
"loss": 0.3963,
"step": 641
},
{
"epoch": 2.763271162123386,
"grad_norm": 0.2796535789966583,
"learning_rate": 1.824815916014644e-07,
"loss": 0.4116,
"step": 642
},
{
"epoch": 2.7675753228120517,
"grad_norm": 0.2771192789077759,
"learning_rate": 1.7582512185498446e-07,
"loss": 0.3818,
"step": 643
},
{
"epoch": 2.7718794835007174,
"grad_norm": 0.2795744240283966,
"learning_rate": 1.6929015125027314e-07,
"loss": 0.4179,
"step": 644
},
{
"epoch": 2.7761836441893832,
"grad_norm": 0.2809208035469055,
"learning_rate": 1.6287684437367724e-07,
"loss": 0.4116,
"step": 645
},
{
"epoch": 2.7804878048780486,
"grad_norm": 0.31142884492874146,
"learning_rate": 1.5658536274738623e-07,
"loss": 0.4216,
"step": 646
},
{
"epoch": 2.7847919655667144,
"grad_norm": 0.3072478175163269,
"learning_rate": 1.504158648253584e-07,
"loss": 0.4378,
"step": 647
},
{
"epoch": 2.78909612625538,
"grad_norm": 0.2876189947128296,
"learning_rate": 1.443685059893396e-07,
"loss": 0.4068,
"step": 648
},
{
"epoch": 2.793400286944046,
"grad_norm": 0.28023388981819153,
"learning_rate": 1.3844343854494123e-07,
"loss": 0.39,
"step": 649
},
{
"epoch": 2.7977044476327118,
"grad_norm": 0.2727784514427185,
"learning_rate": 1.3264081171780797e-07,
"loss": 0.3683,
"step": 650
},
{
"epoch": 2.802008608321377,
"grad_norm": 0.28944262862205505,
"learning_rate": 1.2696077164986e-07,
"loss": 0.3965,
"step": 651
},
{
"epoch": 2.806312769010043,
"grad_norm": 0.27676907181739807,
"learning_rate": 1.2140346139561277e-07,
"loss": 0.4027,
"step": 652
},
{
"epoch": 2.8106169296987087,
"grad_norm": 0.4450412094593048,
"learning_rate": 1.1596902091857043e-07,
"loss": 0.4236,
"step": 653
},
{
"epoch": 2.8149210903873745,
"grad_norm": 0.27743786573410034,
"learning_rate": 1.1065758708770468e-07,
"loss": 0.4033,
"step": 654
},
{
"epoch": 2.8192252510760403,
"grad_norm": 0.26591548323631287,
"learning_rate": 1.0546929367400705e-07,
"loss": 0.3805,
"step": 655
},
{
"epoch": 2.8235294117647056,
"grad_norm": 0.28023067116737366,
"learning_rate": 1.004042713471165e-07,
"loss": 0.3917,
"step": 656
},
{
"epoch": 2.827833572453372,
"grad_norm": 0.2958438992500305,
"learning_rate": 9.546264767203328e-08,
"loss": 0.4197,
"step": 657
},
{
"epoch": 2.832137733142037,
"grad_norm": 0.26877066493034363,
"learning_rate": 9.064454710590253e-08,
"loss": 0.3731,
"step": 658
},
{
"epoch": 2.836441893830703,
"grad_norm": 0.2890089750289917,
"learning_rate": 8.595009099488238e-08,
"loss": 0.3894,
"step": 659
},
{
"epoch": 2.840746054519369,
"grad_norm": 0.27975407242774963,
"learning_rate": 8.137939757108526e-08,
"loss": 0.4,
"step": 660
},
{
"epoch": 2.8450502152080346,
"grad_norm": 0.3021981418132782,
"learning_rate": 7.693258194960252e-08,
"loss": 0.427,
"step": 661
},
{
"epoch": 2.8493543758967004,
"grad_norm": 0.2643302381038666,
"learning_rate": 7.260975612560173e-08,
"loss": 0.3931,
"step": 662
},
{
"epoch": 2.8536585365853657,
"grad_norm": 0.31636062264442444,
"learning_rate": 6.84110289715112e-08,
"loss": 0.3934,
"step": 663
},
{
"epoch": 2.8579626972740315,
"grad_norm": 0.29942587018013,
"learning_rate": 6.433650623427379e-08,
"loss": 0.4066,
"step": 664
},
{
"epoch": 2.8622668579626973,
"grad_norm": 0.3114181160926819,
"learning_rate": 6.038629053268464e-08,
"loss": 0.4122,
"step": 665
},
{
"epoch": 2.866571018651363,
"grad_norm": 0.2796793282032013,
"learning_rate": 5.6560481354807625e-08,
"loss": 0.3884,
"step": 666
},
{
"epoch": 2.870875179340029,
"grad_norm": 0.2825842797756195,
"learning_rate": 5.285917505546967e-08,
"loss": 0.4035,
"step": 667
},
{
"epoch": 2.8751793400286942,
"grad_norm": 0.28558847308158875,
"learning_rate": 4.928246485383148e-08,
"loss": 0.3689,
"step": 668
},
{
"epoch": 2.87948350071736,
"grad_norm": 0.26647746562957764,
"learning_rate": 4.583044083104282e-08,
"loss": 0.389,
"step": 669
},
{
"epoch": 2.883787661406026,
"grad_norm": 0.29167062044143677,
"learning_rate": 4.250318992797375e-08,
"loss": 0.4015,
"step": 670
},
{
"epoch": 2.8880918220946916,
"grad_norm": 0.2896280288696289,
"learning_rate": 3.9300795943021943e-08,
"loss": 0.3971,
"step": 671
},
{
"epoch": 2.8923959827833574,
"grad_norm": 0.26931023597717285,
"learning_rate": 3.622333953000601e-08,
"loss": 0.369,
"step": 672
},
{
"epoch": 2.8967001434720228,
"grad_norm": 0.28395572304725647,
"learning_rate": 3.3270898196129944e-08,
"loss": 0.403,
"step": 673
},
{
"epoch": 2.9010043041606886,
"grad_norm": 0.29338788986206055,
"learning_rate": 3.0443546300035764e-08,
"loss": 0.4154,
"step": 674
},
{
"epoch": 2.9053084648493543,
"grad_norm": 0.29898518323898315,
"learning_rate": 2.77413550499267e-08,
"loss": 0.4208,
"step": 675
},
{
"epoch": 2.90961262553802,
"grad_norm": 0.26712504029273987,
"learning_rate": 2.516439250177749e-08,
"loss": 0.3764,
"step": 676
},
{
"epoch": 2.913916786226686,
"grad_norm": 0.2811000347137451,
"learning_rate": 2.2712723557616335e-08,
"loss": 0.3865,
"step": 677
},
{
"epoch": 2.9182209469153513,
"grad_norm": 0.2783554792404175,
"learning_rate": 2.038640996389285e-08,
"loss": 0.4393,
"step": 678
},
{
"epoch": 2.922525107604017,
"grad_norm": 0.27642354369163513,
"learning_rate": 1.818551030992377e-08,
"loss": 0.377,
"step": 679
},
{
"epoch": 2.926829268292683,
"grad_norm": 0.30214452743530273,
"learning_rate": 1.6110080026414123e-08,
"loss": 0.3919,
"step": 680
},
{
"epoch": 2.9311334289813487,
"grad_norm": 0.2953914999961853,
"learning_rate": 1.4160171384064447e-08,
"loss": 0.4057,
"step": 681
},
{
"epoch": 2.9354375896700144,
"grad_norm": 0.28328144550323486,
"learning_rate": 1.2335833492252425e-08,
"loss": 0.4022,
"step": 682
},
{
"epoch": 2.93974175035868,
"grad_norm": 0.2768747806549072,
"learning_rate": 1.063711229779718e-08,
"loss": 0.3835,
"step": 683
},
{
"epoch": 2.944045911047346,
"grad_norm": 0.2623550295829773,
"learning_rate": 9.06405058380022e-09,
"loss": 0.3517,
"step": 684
},
{
"epoch": 2.9483500717360114,
"grad_norm": 0.29734596610069275,
"learning_rate": 7.61668796857018e-09,
"loss": 0.4198,
"step": 685
},
{
"epoch": 2.952654232424677,
"grad_norm": 0.27858829498291016,
"learning_rate": 6.295060904623618e-09,
"loss": 0.3796,
"step": 686
},
{
"epoch": 2.956958393113343,
"grad_norm": 0.2829955518245697,
"learning_rate": 5.099202677767978e-09,
"loss": 0.4188,
"step": 687
},
{
"epoch": 2.9612625538020088,
"grad_norm": 0.2835357189178467,
"learning_rate": 4.02914340626226e-09,
"loss": 0.42,
"step": 688
},
{
"epoch": 2.9655667144906745,
"grad_norm": 0.28330615162849426,
"learning_rate": 3.0849100400587307e-09,
"loss": 0.3739,
"step": 689
},
{
"epoch": 2.96987087517934,
"grad_norm": 0.30792394280433655,
"learning_rate": 2.2665263601240328e-09,
"loss": 0.4282,
"step": 690
},
{
"epoch": 2.9741750358680057,
"grad_norm": 0.2824040651321411,
"learning_rate": 1.5740129778413215e-09,
"loss": 0.409,
"step": 691
},
{
"epoch": 2.9784791965566715,
"grad_norm": 0.2838318943977356,
"learning_rate": 1.0073873344895735e-09,
"loss": 0.383,
"step": 692
},
{
"epoch": 2.9827833572453373,
"grad_norm": 0.2872404456138611,
"learning_rate": 5.666637008061582e-10,
"loss": 0.3953,
"step": 693
},
{
"epoch": 2.987087517934003,
"grad_norm": 0.2668392062187195,
"learning_rate": 2.5185317662490547e-10,
"loss": 0.383,
"step": 694
},
{
"epoch": 2.9913916786226684,
"grad_norm": 0.2843668460845947,
"learning_rate": 6.296369059854978e-11,
"loss": 0.4103,
"step": 695
},
{
"epoch": 2.995695839311334,
"grad_norm": 0.27907493710517883,
"learning_rate": 0.0,
"loss": 0.3768,
"step": 696
},
{
"epoch": 2.995695839311334,
"step": 696,
"total_flos": 1029628835528704.0,
"train_loss": 0.45175294526692095,
"train_runtime": 14742.7033,
"train_samples_per_second": 4.534,
"train_steps_per_second": 0.047
}
],
"logging_steps": 1.0,
"max_steps": 696,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1029628835528704.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}