doc / trainer_state.json
Blaze7451's picture
Upload folder using huggingface_hub
0f853df verified
Raw History Blame Contribute Delete
330 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 937,
"global_step": 1874,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0005336179295624333,
"grad_norm": 1.625,
"learning_rate": 0.0,
"loss": 0.13112366199493408,
"step": 1
},
{
"epoch": 0.0010672358591248667,
"grad_norm": 1.453125,
"learning_rate": 4e-08,
"loss": 0.14172744750976562,
"step": 2
},
{
"epoch": 0.0016008537886872999,
"grad_norm": 1.8046875,
"learning_rate": 8e-08,
"loss": 0.15269151329994202,
"step": 3
},
{
"epoch": 0.0021344717182497333,
"grad_norm": 1.1875,
"learning_rate": 1.2e-07,
"loss": 0.14777952432632446,
"step": 4
},
{
"epoch": 0.0026680896478121665,
"grad_norm": 1.8046875,
"learning_rate": 1.6e-07,
"loss": 0.14433835446834564,
"step": 5
},
{
"epoch": 0.0032017075773745998,
"grad_norm": 2.5625,
"learning_rate": 2e-07,
"loss": 0.1490514576435089,
"step": 6
},
{
"epoch": 0.003735325506937033,
"grad_norm": 1.328125,
"learning_rate": 2.4e-07,
"loss": 0.16350091993808746,
"step": 7
},
{
"epoch": 0.004268943436499467,
"grad_norm": 2.890625,
"learning_rate": 2.8e-07,
"loss": 0.14256075024604797,
"step": 8
},
{
"epoch": 0.0048025613660619,
"grad_norm": 3.875,
"learning_rate": 3.2e-07,
"loss": 0.14731445908546448,
"step": 9
},
{
"epoch": 0.005336179295624333,
"grad_norm": 3.125,
"learning_rate": 3.6e-07,
"loss": 0.13776251673698425,
"step": 10
},
{
"epoch": 0.005869797225186766,
"grad_norm": 1.7109375,
"learning_rate": 4e-07,
"loss": 0.14122653007507324,
"step": 11
},
{
"epoch": 0.0064034151547491995,
"grad_norm": 1.3359375,
"learning_rate": 4.3999999999999997e-07,
"loss": 0.16007639467716217,
"step": 12
},
{
"epoch": 0.006937033084311633,
"grad_norm": 1.5078125,
"learning_rate": 4.8e-07,
"loss": 0.13928429782390594,
"step": 13
},
{
"epoch": 0.007470651013874066,
"grad_norm": 2.109375,
"learning_rate": 5.2e-07,
"loss": 0.16247457265853882,
"step": 14
},
{
"epoch": 0.0080042689434365,
"grad_norm": 1.953125,
"learning_rate": 5.6e-07,
"loss": 0.14845246076583862,
"step": 15
},
{
"epoch": 0.008537886872998933,
"grad_norm": 1.484375,
"learning_rate": 6e-07,
"loss": 0.15235859155654907,
"step": 16
},
{
"epoch": 0.009071504802561366,
"grad_norm": 2.8125,
"learning_rate": 6.4e-07,
"loss": 0.14680761098861694,
"step": 17
},
{
"epoch": 0.0096051227321238,
"grad_norm": 1.359375,
"learning_rate": 6.800000000000001e-07,
"loss": 0.15282537043094635,
"step": 18
},
{
"epoch": 0.010138740661686232,
"grad_norm": 1.3359375,
"learning_rate": 7.2e-07,
"loss": 0.1538020670413971,
"step": 19
},
{
"epoch": 0.010672358591248666,
"grad_norm": 2.171875,
"learning_rate": 7.599999999999999e-07,
"loss": 0.13729596138000488,
"step": 20
},
{
"epoch": 0.011205976520811099,
"grad_norm": 1.71875,
"learning_rate": 8e-07,
"loss": 0.14770808815956116,
"step": 21
},
{
"epoch": 0.011739594450373533,
"grad_norm": 1.5234375,
"learning_rate": 8.399999999999999e-07,
"loss": 0.14649993181228638,
"step": 22
},
{
"epoch": 0.012273212379935965,
"grad_norm": 1.171875,
"learning_rate": 8.799999999999999e-07,
"loss": 0.1499944031238556,
"step": 23
},
{
"epoch": 0.012806830309498399,
"grad_norm": 8.125,
"learning_rate": 9.2e-07,
"loss": 0.14750531315803528,
"step": 24
},
{
"epoch": 0.013340448239060833,
"grad_norm": 1.4140625,
"learning_rate": 9.6e-07,
"loss": 0.14075997471809387,
"step": 25
},
{
"epoch": 0.013874066168623266,
"grad_norm": 1.171875,
"learning_rate": 1e-06,
"loss": 0.14724700152873993,
"step": 26
},
{
"epoch": 0.0144076840981857,
"grad_norm": 1.75,
"learning_rate": 1.04e-06,
"loss": 0.13195163011550903,
"step": 27
},
{
"epoch": 0.014941302027748132,
"grad_norm": 1.515625,
"learning_rate": 1.08e-06,
"loss": 0.1568724364042282,
"step": 28
},
{
"epoch": 0.015474919957310566,
"grad_norm": 1.8828125,
"learning_rate": 1.12e-06,
"loss": 0.16248375177383423,
"step": 29
},
{
"epoch": 0.016008537886873,
"grad_norm": 9.0625,
"learning_rate": 1.16e-06,
"loss": 0.13451561331748962,
"step": 30
},
{
"epoch": 0.016542155816435433,
"grad_norm": 1.34375,
"learning_rate": 1.2e-06,
"loss": 0.12884676456451416,
"step": 31
},
{
"epoch": 0.017075773745997867,
"grad_norm": 1.7578125,
"learning_rate": 1.24e-06,
"loss": 0.14001944661140442,
"step": 32
},
{
"epoch": 0.017609391675560297,
"grad_norm": 1.5,
"learning_rate": 1.28e-06,
"loss": 0.13409659266471863,
"step": 33
},
{
"epoch": 0.01814300960512273,
"grad_norm": 3.015625,
"learning_rate": 1.32e-06,
"loss": 0.13101552426815033,
"step": 34
},
{
"epoch": 0.018676627534685165,
"grad_norm": 1.9375,
"learning_rate": 1.3600000000000001e-06,
"loss": 0.15233811736106873,
"step": 35
},
{
"epoch": 0.0192102454642476,
"grad_norm": 2.125,
"learning_rate": 1.4e-06,
"loss": 0.1450912058353424,
"step": 36
},
{
"epoch": 0.019743863393810034,
"grad_norm": 1.6484375,
"learning_rate": 1.44e-06,
"loss": 0.14182355999946594,
"step": 37
},
{
"epoch": 0.020277481323372464,
"grad_norm": 1.34375,
"learning_rate": 1.48e-06,
"loss": 0.15987959504127502,
"step": 38
},
{
"epoch": 0.0208110992529349,
"grad_norm": 1.21875,
"learning_rate": 1.5199999999999998e-06,
"loss": 0.12670841813087463,
"step": 39
},
{
"epoch": 0.021344717182497332,
"grad_norm": 5.03125,
"learning_rate": 1.5599999999999999e-06,
"loss": 0.14987631142139435,
"step": 40
},
{
"epoch": 0.021878335112059766,
"grad_norm": 1.1484375,
"learning_rate": 1.6e-06,
"loss": 0.1366727650165558,
"step": 41
},
{
"epoch": 0.022411953041622197,
"grad_norm": 1.3671875,
"learning_rate": 1.6399999999999998e-06,
"loss": 0.13809248805046082,
"step": 42
},
{
"epoch": 0.02294557097118463,
"grad_norm": 1.0390625,
"learning_rate": 1.6799999999999998e-06,
"loss": 0.12975868582725525,
"step": 43
},
{
"epoch": 0.023479188900747065,
"grad_norm": 1.5,
"learning_rate": 1.7199999999999998e-06,
"loss": 0.12949730455875397,
"step": 44
},
{
"epoch": 0.0240128068303095,
"grad_norm": 2.40625,
"learning_rate": 1.7599999999999999e-06,
"loss": 0.13930079340934753,
"step": 45
},
{
"epoch": 0.02454642475987193,
"grad_norm": 1.328125,
"learning_rate": 1.8e-06,
"loss": 0.1614750474691391,
"step": 46
},
{
"epoch": 0.025080042689434364,
"grad_norm": 1.2265625,
"learning_rate": 1.84e-06,
"loss": 0.14609354734420776,
"step": 47
},
{
"epoch": 0.025613660618996798,
"grad_norm": 3.953125,
"learning_rate": 1.8799999999999998e-06,
"loss": 0.14017945528030396,
"step": 48
},
{
"epoch": 0.026147278548559232,
"grad_norm": 2.15625,
"learning_rate": 1.92e-06,
"loss": 0.12664636969566345,
"step": 49
},
{
"epoch": 0.026680896478121666,
"grad_norm": 1.28125,
"learning_rate": 1.96e-06,
"loss": 0.1593368649482727,
"step": 50
},
{
"epoch": 0.027214514407684097,
"grad_norm": 1.6484375,
"learning_rate": 2e-06,
"loss": 0.14744272828102112,
"step": 51
},
{
"epoch": 0.02774813233724653,
"grad_norm": 2.296875,
"learning_rate": 2.04e-06,
"loss": 0.16213791072368622,
"step": 52
},
{
"epoch": 0.028281750266808965,
"grad_norm": 1.9296875,
"learning_rate": 2.08e-06,
"loss": 0.1429382860660553,
"step": 53
},
{
"epoch": 0.0288153681963714,
"grad_norm": 2.25,
"learning_rate": 2.12e-06,
"loss": 0.14756259322166443,
"step": 54
},
{
"epoch": 0.02934898612593383,
"grad_norm": 1.5,
"learning_rate": 2.16e-06,
"loss": 0.14745742082595825,
"step": 55
},
{
"epoch": 0.029882604055496264,
"grad_norm": 1.1640625,
"learning_rate": 2.2e-06,
"loss": 0.1360693871974945,
"step": 56
},
{
"epoch": 0.030416221985058698,
"grad_norm": 1.5625,
"learning_rate": 2.24e-06,
"loss": 0.1259615421295166,
"step": 57
},
{
"epoch": 0.030949839914621132,
"grad_norm": 1.484375,
"learning_rate": 2.2799999999999998e-06,
"loss": 0.12637874484062195,
"step": 58
},
{
"epoch": 0.03148345784418356,
"grad_norm": 1.3125,
"learning_rate": 2.32e-06,
"loss": 0.1504337191581726,
"step": 59
},
{
"epoch": 0.032017075773746,
"grad_norm": 5.0625,
"learning_rate": 2.36e-06,
"loss": 0.13464780151844025,
"step": 60
},
{
"epoch": 0.03255069370330843,
"grad_norm": 1.65625,
"learning_rate": 2.4e-06,
"loss": 0.13137051463127136,
"step": 61
},
{
"epoch": 0.033084311632870865,
"grad_norm": 1.4375,
"learning_rate": 2.44e-06,
"loss": 0.13932852447032928,
"step": 62
},
{
"epoch": 0.0336179295624333,
"grad_norm": 8.5,
"learning_rate": 2.48e-06,
"loss": 0.13769161701202393,
"step": 63
},
{
"epoch": 0.03415154749199573,
"grad_norm": 2.359375,
"learning_rate": 2.52e-06,
"loss": 0.15295428037643433,
"step": 64
},
{
"epoch": 0.03468516542155817,
"grad_norm": 1.6328125,
"learning_rate": 2.56e-06,
"loss": 0.13566850125789642,
"step": 65
},
{
"epoch": 0.035218783351120594,
"grad_norm": 1.1484375,
"learning_rate": 2.6e-06,
"loss": 0.13369709253311157,
"step": 66
},
{
"epoch": 0.03575240128068303,
"grad_norm": 1.5546875,
"learning_rate": 2.64e-06,
"loss": 0.1427057385444641,
"step": 67
},
{
"epoch": 0.03628601921024546,
"grad_norm": 1.4375,
"learning_rate": 2.68e-06,
"loss": 0.13130415976047516,
"step": 68
},
{
"epoch": 0.0368196371398079,
"grad_norm": 5.875,
"learning_rate": 2.7200000000000002e-06,
"loss": 0.14629092812538147,
"step": 69
},
{
"epoch": 0.03735325506937033,
"grad_norm": 1.4453125,
"learning_rate": 2.76e-06,
"loss": 0.14595183730125427,
"step": 70
},
{
"epoch": 0.037886872998932765,
"grad_norm": 1.640625,
"learning_rate": 2.8e-06,
"loss": 0.12249842286109924,
"step": 71
},
{
"epoch": 0.0384204909284952,
"grad_norm": 2.046875,
"learning_rate": 2.84e-06,
"loss": 0.14652207493782043,
"step": 72
},
{
"epoch": 0.03895410885805763,
"grad_norm": 1.5390625,
"learning_rate": 2.88e-06,
"loss": 0.1352737843990326,
"step": 73
},
{
"epoch": 0.03948772678762007,
"grad_norm": 4.09375,
"learning_rate": 2.92e-06,
"loss": 0.13420942425727844,
"step": 74
},
{
"epoch": 0.040021344717182494,
"grad_norm": 2.453125,
"learning_rate": 2.96e-06,
"loss": 0.1462140679359436,
"step": 75
},
{
"epoch": 0.04055496264674493,
"grad_norm": 14.4375,
"learning_rate": 3e-06,
"loss": 0.12372896820306778,
"step": 76
},
{
"epoch": 0.04108858057630736,
"grad_norm": 12.0,
"learning_rate": 3.0399999999999997e-06,
"loss": 0.15547272562980652,
"step": 77
},
{
"epoch": 0.0416221985058698,
"grad_norm": 1.671875,
"learning_rate": 3.0799999999999997e-06,
"loss": 0.13124234974384308,
"step": 78
},
{
"epoch": 0.04215581643543223,
"grad_norm": 3.28125,
"learning_rate": 3.1199999999999998e-06,
"loss": 0.13380131125450134,
"step": 79
},
{
"epoch": 0.042689434364994665,
"grad_norm": 0.8046875,
"learning_rate": 3.16e-06,
"loss": 0.11884702742099762,
"step": 80
},
{
"epoch": 0.0432230522945571,
"grad_norm": 1.8984375,
"learning_rate": 3.2e-06,
"loss": 0.1355973482131958,
"step": 81
},
{
"epoch": 0.04375667022411953,
"grad_norm": 1.015625,
"learning_rate": 3.24e-06,
"loss": 0.11962362378835678,
"step": 82
},
{
"epoch": 0.04429028815368197,
"grad_norm": 1.3984375,
"learning_rate": 3.2799999999999995e-06,
"loss": 0.1329701989889145,
"step": 83
},
{
"epoch": 0.044823906083244394,
"grad_norm": 2.125,
"learning_rate": 3.3199999999999996e-06,
"loss": 0.1501164585351944,
"step": 84
},
{
"epoch": 0.04535752401280683,
"grad_norm": 3.046875,
"learning_rate": 3.3599999999999996e-06,
"loss": 0.14272454380989075,
"step": 85
},
{
"epoch": 0.04589114194236926,
"grad_norm": 2.15625,
"learning_rate": 3.3999999999999996e-06,
"loss": 0.12242889404296875,
"step": 86
},
{
"epoch": 0.046424759871931696,
"grad_norm": 1.7734375,
"learning_rate": 3.4399999999999997e-06,
"loss": 0.12888413667678833,
"step": 87
},
{
"epoch": 0.04695837780149413,
"grad_norm": 1.4609375,
"learning_rate": 3.4799999999999997e-06,
"loss": 0.1354946345090866,
"step": 88
},
{
"epoch": 0.047491995731056565,
"grad_norm": 2.265625,
"learning_rate": 3.5199999999999998e-06,
"loss": 0.14096581935882568,
"step": 89
},
{
"epoch": 0.048025613660619,
"grad_norm": 1.484375,
"learning_rate": 3.56e-06,
"loss": 0.1245022639632225,
"step": 90
},
{
"epoch": 0.04855923159018143,
"grad_norm": 1.8984375,
"learning_rate": 3.6e-06,
"loss": 0.15014950931072235,
"step": 91
},
{
"epoch": 0.04909284951974386,
"grad_norm": 1.46875,
"learning_rate": 3.64e-06,
"loss": 0.1436699777841568,
"step": 92
},
{
"epoch": 0.049626467449306294,
"grad_norm": 1.234375,
"learning_rate": 3.68e-06,
"loss": 0.12723881006240845,
"step": 93
},
{
"epoch": 0.05016008537886873,
"grad_norm": 1.2890625,
"learning_rate": 3.72e-06,
"loss": 0.13498947024345398,
"step": 94
},
{
"epoch": 0.05069370330843116,
"grad_norm": 0.984375,
"learning_rate": 3.7599999999999996e-06,
"loss": 0.13687095046043396,
"step": 95
},
{
"epoch": 0.051227321237993596,
"grad_norm": 1.765625,
"learning_rate": 3.7999999999999996e-06,
"loss": 0.1502760350704193,
"step": 96
},
{
"epoch": 0.05176093916755603,
"grad_norm": 4.125,
"learning_rate": 3.84e-06,
"loss": 0.12798289954662323,
"step": 97
},
{
"epoch": 0.052294557097118465,
"grad_norm": 1.78125,
"learning_rate": 3.88e-06,
"loss": 0.1594056487083435,
"step": 98
},
{
"epoch": 0.0528281750266809,
"grad_norm": 2.609375,
"learning_rate": 3.92e-06,
"loss": 0.13724134862422943,
"step": 99
},
{
"epoch": 0.05336179295624333,
"grad_norm": 8.4375,
"learning_rate": 3.96e-06,
"loss": 0.1421767920255661,
"step": 100
},
{
"epoch": 0.05389541088580576,
"grad_norm": 9.125,
"learning_rate": 4e-06,
"loss": 0.14196684956550598,
"step": 101
},
{
"epoch": 0.054429028815368194,
"grad_norm": 3.203125,
"learning_rate": 4.0399999999999994e-06,
"loss": 0.12153448164463043,
"step": 102
},
{
"epoch": 0.05496264674493063,
"grad_norm": 1.3046875,
"learning_rate": 4.08e-06,
"loss": 0.14348502457141876,
"step": 103
},
{
"epoch": 0.05549626467449306,
"grad_norm": 2.03125,
"learning_rate": 4.1199999999999995e-06,
"loss": 0.13037782907485962,
"step": 104
},
{
"epoch": 0.056029882604055496,
"grad_norm": 2.015625,
"learning_rate": 4.16e-06,
"loss": 0.15000101923942566,
"step": 105
},
{
"epoch": 0.05656350053361793,
"grad_norm": 1.8125,
"learning_rate": 4.2e-06,
"loss": 0.1348305344581604,
"step": 106
},
{
"epoch": 0.057097118463180364,
"grad_norm": 5.21875,
"learning_rate": 4.24e-06,
"loss": 0.15050621330738068,
"step": 107
},
{
"epoch": 0.0576307363927428,
"grad_norm": 1.7265625,
"learning_rate": 4.28e-06,
"loss": 0.14010144770145416,
"step": 108
},
{
"epoch": 0.05816435432230523,
"grad_norm": 1.5078125,
"learning_rate": 4.32e-06,
"loss": 0.13223057985305786,
"step": 109
},
{
"epoch": 0.05869797225186766,
"grad_norm": 1.3203125,
"learning_rate": 4.36e-06,
"loss": 0.13937245309352875,
"step": 110
},
{
"epoch": 0.059231590181430094,
"grad_norm": 1.0859375,
"learning_rate": 4.4e-06,
"loss": 0.15181338787078857,
"step": 111
},
{
"epoch": 0.05976520811099253,
"grad_norm": 2.0,
"learning_rate": 4.44e-06,
"loss": 0.13395366072654724,
"step": 112
},
{
"epoch": 0.06029882604055496,
"grad_norm": 2.296875,
"learning_rate": 4.48e-06,
"loss": 0.13292331993579865,
"step": 113
},
{
"epoch": 0.060832443970117396,
"grad_norm": 1.1640625,
"learning_rate": 4.519999999999999e-06,
"loss": 0.13701671361923218,
"step": 114
},
{
"epoch": 0.06136606189967983,
"grad_norm": 1.1015625,
"learning_rate": 4.5599999999999995e-06,
"loss": 0.13527025282382965,
"step": 115
},
{
"epoch": 0.061899679829242264,
"grad_norm": 1.0625,
"learning_rate": 4.599999999999999e-06,
"loss": 0.12855729460716248,
"step": 116
},
{
"epoch": 0.0624332977588047,
"grad_norm": 1.109375,
"learning_rate": 4.64e-06,
"loss": 0.1277010291814804,
"step": 117
},
{
"epoch": 0.06296691568836713,
"grad_norm": 1.453125,
"learning_rate": 4.679999999999999e-06,
"loss": 0.14392659068107605,
"step": 118
},
{
"epoch": 0.06350053361792957,
"grad_norm": 1.203125,
"learning_rate": 4.72e-06,
"loss": 0.12997901439666748,
"step": 119
},
{
"epoch": 0.064034151547492,
"grad_norm": 1.1953125,
"learning_rate": 4.759999999999999e-06,
"loss": 0.13523350656032562,
"step": 120
},
{
"epoch": 0.06456776947705443,
"grad_norm": 1.1484375,
"learning_rate": 4.8e-06,
"loss": 0.13294854760169983,
"step": 121
},
{
"epoch": 0.06510138740661686,
"grad_norm": 1.4609375,
"learning_rate": 4.839999999999999e-06,
"loss": 0.14436154067516327,
"step": 122
},
{
"epoch": 0.06563500533617929,
"grad_norm": 1.578125,
"learning_rate": 4.88e-06,
"loss": 0.1320681869983673,
"step": 123
},
{
"epoch": 0.06616862326574173,
"grad_norm": 0.98828125,
"learning_rate": 4.9199999999999995e-06,
"loss": 0.12814241647720337,
"step": 124
},
{
"epoch": 0.06670224119530416,
"grad_norm": 5.65625,
"learning_rate": 4.96e-06,
"loss": 0.14062875509262085,
"step": 125
},
{
"epoch": 0.0672358591248666,
"grad_norm": 4.1875,
"learning_rate": 4.9999999999999996e-06,
"loss": 0.15474671125411987,
"step": 126
},
{
"epoch": 0.06776947705442903,
"grad_norm": 2.25,
"learning_rate": 5.04e-06,
"loss": 0.13234499096870422,
"step": 127
},
{
"epoch": 0.06830309498399147,
"grad_norm": 1.4296875,
"learning_rate": 5.08e-06,
"loss": 0.14397011697292328,
"step": 128
},
{
"epoch": 0.0688367129135539,
"grad_norm": 1.2421875,
"learning_rate": 5.12e-06,
"loss": 0.13815239071846008,
"step": 129
},
{
"epoch": 0.06937033084311633,
"grad_norm": 1.71875,
"learning_rate": 5.16e-06,
"loss": 0.12156923860311508,
"step": 130
},
{
"epoch": 0.06990394877267876,
"grad_norm": 2.625,
"learning_rate": 5.2e-06,
"loss": 0.13927510380744934,
"step": 131
},
{
"epoch": 0.07043756670224119,
"grad_norm": 6.6875,
"learning_rate": 5.24e-06,
"loss": 0.14838266372680664,
"step": 132
},
{
"epoch": 0.07097118463180363,
"grad_norm": 1.09375,
"learning_rate": 5.28e-06,
"loss": 0.13549724221229553,
"step": 133
},
{
"epoch": 0.07150480256136606,
"grad_norm": 1.5703125,
"learning_rate": 5.32e-06,
"loss": 0.12208350747823715,
"step": 134
},
{
"epoch": 0.0720384204909285,
"grad_norm": 1.5,
"learning_rate": 5.36e-06,
"loss": 0.13542619347572327,
"step": 135
},
{
"epoch": 0.07257203842049093,
"grad_norm": 1.8828125,
"learning_rate": 5.4e-06,
"loss": 0.14441967010498047,
"step": 136
},
{
"epoch": 0.07310565635005337,
"grad_norm": 1.4375,
"learning_rate": 5.4400000000000004e-06,
"loss": 0.11140245199203491,
"step": 137
},
{
"epoch": 0.0736392742796158,
"grad_norm": 2.28125,
"learning_rate": 5.48e-06,
"loss": 0.12050978094339371,
"step": 138
},
{
"epoch": 0.07417289220917823,
"grad_norm": 1.65625,
"learning_rate": 5.52e-06,
"loss": 0.1412547528743744,
"step": 139
},
{
"epoch": 0.07470651013874066,
"grad_norm": 2.203125,
"learning_rate": 5.559999999999999e-06,
"loss": 0.12972840666770935,
"step": 140
},
{
"epoch": 0.07524012806830309,
"grad_norm": 1.5,
"learning_rate": 5.6e-06,
"loss": 0.14697983860969543,
"step": 141
},
{
"epoch": 0.07577374599786553,
"grad_norm": 2.078125,
"learning_rate": 5.639999999999999e-06,
"loss": 0.13823923468589783,
"step": 142
},
{
"epoch": 0.07630736392742796,
"grad_norm": 1.203125,
"learning_rate": 5.68e-06,
"loss": 0.13876640796661377,
"step": 143
},
{
"epoch": 0.0768409818569904,
"grad_norm": 1.9609375,
"learning_rate": 5.7199999999999994e-06,
"loss": 0.12860342860221863,
"step": 144
},
{
"epoch": 0.07737459978655283,
"grad_norm": 1.2265625,
"learning_rate": 5.76e-06,
"loss": 0.12727797031402588,
"step": 145
},
{
"epoch": 0.07790821771611527,
"grad_norm": 0.94140625,
"learning_rate": 5.7999999999999995e-06,
"loss": 0.13474544882774353,
"step": 146
},
{
"epoch": 0.0784418356456777,
"grad_norm": 1.7109375,
"learning_rate": 5.84e-06,
"loss": 0.1328914612531662,
"step": 147
},
{
"epoch": 0.07897545357524013,
"grad_norm": 1.3515625,
"learning_rate": 5.88e-06,
"loss": 0.13169237971305847,
"step": 148
},
{
"epoch": 0.07950907150480256,
"grad_norm": 6.46875,
"learning_rate": 5.92e-06,
"loss": 0.1330462396144867,
"step": 149
},
{
"epoch": 0.08004268943436499,
"grad_norm": 1.1484375,
"learning_rate": 5.96e-06,
"loss": 0.1325974017381668,
"step": 150
},
{
"epoch": 0.08057630736392743,
"grad_norm": 2.875,
"learning_rate": 6e-06,
"loss": 0.12827789783477783,
"step": 151
},
{
"epoch": 0.08110992529348986,
"grad_norm": 1.6875,
"learning_rate": 6.04e-06,
"loss": 0.14392346143722534,
"step": 152
},
{
"epoch": 0.0816435432230523,
"grad_norm": 1.4375,
"learning_rate": 6.079999999999999e-06,
"loss": 0.14137224853038788,
"step": 153
},
{
"epoch": 0.08217716115261472,
"grad_norm": 1.2890625,
"learning_rate": 6.12e-06,
"loss": 0.136900395154953,
"step": 154
},
{
"epoch": 0.08271077908217717,
"grad_norm": 1.453125,
"learning_rate": 6.1599999999999995e-06,
"loss": 0.14158356189727783,
"step": 155
},
{
"epoch": 0.0832443970117396,
"grad_norm": 5.0,
"learning_rate": 6.2e-06,
"loss": 0.14513534307479858,
"step": 156
},
{
"epoch": 0.08377801494130203,
"grad_norm": 1.484375,
"learning_rate": 6.2399999999999995e-06,
"loss": 0.13704454898834229,
"step": 157
},
{
"epoch": 0.08431163287086446,
"grad_norm": 7.15625,
"learning_rate": 6.28e-06,
"loss": 0.11171528697013855,
"step": 158
},
{
"epoch": 0.08484525080042689,
"grad_norm": 1.7109375,
"learning_rate": 6.32e-06,
"loss": 0.1400269716978073,
"step": 159
},
{
"epoch": 0.08537886872998933,
"grad_norm": 1.4765625,
"learning_rate": 6.36e-06,
"loss": 0.12547072768211365,
"step": 160
},
{
"epoch": 0.08591248665955176,
"grad_norm": 1.640625,
"learning_rate": 6.4e-06,
"loss": 0.1414448320865631,
"step": 161
},
{
"epoch": 0.0864461045891142,
"grad_norm": 1.4140625,
"learning_rate": 6.44e-06,
"loss": 0.13517022132873535,
"step": 162
},
{
"epoch": 0.08697972251867662,
"grad_norm": 2.015625,
"learning_rate": 6.48e-06,
"loss": 0.1294126808643341,
"step": 163
},
{
"epoch": 0.08751334044823907,
"grad_norm": 0.90625,
"learning_rate": 6.519999999999999e-06,
"loss": 0.129896879196167,
"step": 164
},
{
"epoch": 0.08804695837780149,
"grad_norm": 1.46875,
"learning_rate": 6.559999999999999e-06,
"loss": 0.13506123423576355,
"step": 165
},
{
"epoch": 0.08858057630736393,
"grad_norm": 6.1875,
"learning_rate": 6.5999999999999995e-06,
"loss": 0.13368020951747894,
"step": 166
},
{
"epoch": 0.08911419423692636,
"grad_norm": 1.203125,
"learning_rate": 6.639999999999999e-06,
"loss": 0.13397908210754395,
"step": 167
},
{
"epoch": 0.08964781216648879,
"grad_norm": 0.8515625,
"learning_rate": 6.6799999999999996e-06,
"loss": 0.12838855385780334,
"step": 168
},
{
"epoch": 0.09018143009605123,
"grad_norm": 1.1328125,
"learning_rate": 6.719999999999999e-06,
"loss": 0.1354101300239563,
"step": 169
},
{
"epoch": 0.09071504802561366,
"grad_norm": 1.1171875,
"learning_rate": 6.76e-06,
"loss": 0.13738185167312622,
"step": 170
},
{
"epoch": 0.0912486659551761,
"grad_norm": 0.93359375,
"learning_rate": 6.799999999999999e-06,
"loss": 0.14488762617111206,
"step": 171
},
{
"epoch": 0.09178228388473852,
"grad_norm": 1.4296875,
"learning_rate": 6.84e-06,
"loss": 0.12928706407546997,
"step": 172
},
{
"epoch": 0.09231590181430097,
"grad_norm": 0.9453125,
"learning_rate": 6.879999999999999e-06,
"loss": 0.13206326961517334,
"step": 173
},
{
"epoch": 0.09284951974386339,
"grad_norm": 0.8359375,
"learning_rate": 6.92e-06,
"loss": 0.1297769546508789,
"step": 174
},
{
"epoch": 0.09338313767342583,
"grad_norm": 1.265625,
"learning_rate": 6.9599999999999994e-06,
"loss": 0.1337125152349472,
"step": 175
},
{
"epoch": 0.09391675560298826,
"grad_norm": 3.984375,
"learning_rate": 7e-06,
"loss": 0.13906241953372955,
"step": 176
},
{
"epoch": 0.09445037353255069,
"grad_norm": 2.453125,
"learning_rate": 7.0399999999999995e-06,
"loss": 0.12388654053211212,
"step": 177
},
{
"epoch": 0.09498399146211313,
"grad_norm": 1.21875,
"learning_rate": 7.08e-06,
"loss": 0.14148969948291779,
"step": 178
},
{
"epoch": 0.09551760939167556,
"grad_norm": 1.4375,
"learning_rate": 7.12e-06,
"loss": 0.15286089479923248,
"step": 179
},
{
"epoch": 0.096051227321238,
"grad_norm": 1.5078125,
"learning_rate": 7.16e-06,
"loss": 0.14178329706192017,
"step": 180
},
{
"epoch": 0.09658484525080042,
"grad_norm": 6.1875,
"learning_rate": 7.2e-06,
"loss": 0.14246678352355957,
"step": 181
},
{
"epoch": 0.09711846318036287,
"grad_norm": 1.6015625,
"learning_rate": 7.24e-06,
"loss": 0.1226498931646347,
"step": 182
},
{
"epoch": 0.09765208110992529,
"grad_norm": 1.65625,
"learning_rate": 7.28e-06,
"loss": 0.13879838585853577,
"step": 183
},
{
"epoch": 0.09818569903948772,
"grad_norm": 0.9765625,
"learning_rate": 7.32e-06,
"loss": 0.1319880485534668,
"step": 184
},
{
"epoch": 0.09871931696905016,
"grad_norm": 0.8359375,
"learning_rate": 7.36e-06,
"loss": 0.12739792466163635,
"step": 185
},
{
"epoch": 0.09925293489861259,
"grad_norm": 1.3515625,
"learning_rate": 7.4e-06,
"loss": 0.14962303638458252,
"step": 186
},
{
"epoch": 0.09978655282817503,
"grad_norm": 1.4765625,
"learning_rate": 7.44e-06,
"loss": 0.12272988259792328,
"step": 187
},
{
"epoch": 0.10032017075773746,
"grad_norm": 1.1640625,
"learning_rate": 7.48e-06,
"loss": 0.14004406332969666,
"step": 188
},
{
"epoch": 0.1008537886872999,
"grad_norm": 3.46875,
"learning_rate": 7.519999999999999e-06,
"loss": 0.1399938464164734,
"step": 189
},
{
"epoch": 0.10138740661686232,
"grad_norm": 2.328125,
"learning_rate": 7.56e-06,
"loss": 0.1324838399887085,
"step": 190
},
{
"epoch": 0.10192102454642477,
"grad_norm": 1.1328125,
"learning_rate": 7.599999999999999e-06,
"loss": 0.12993931770324707,
"step": 191
},
{
"epoch": 0.10245464247598719,
"grad_norm": 5.875,
"learning_rate": 7.64e-06,
"loss": 0.1436869353055954,
"step": 192
},
{
"epoch": 0.10298826040554962,
"grad_norm": 15.1875,
"learning_rate": 7.68e-06,
"loss": 0.14680171012878418,
"step": 193
},
{
"epoch": 0.10352187833511206,
"grad_norm": 4.6875,
"learning_rate": 7.719999999999999e-06,
"loss": 0.13149963319301605,
"step": 194
},
{
"epoch": 0.10405549626467449,
"grad_norm": 0.99609375,
"learning_rate": 7.76e-06,
"loss": 0.10434500873088837,
"step": 195
},
{
"epoch": 0.10458911419423693,
"grad_norm": 0.9921875,
"learning_rate": 7.8e-06,
"loss": 0.1360233724117279,
"step": 196
},
{
"epoch": 0.10512273212379936,
"grad_norm": 0.84375,
"learning_rate": 7.84e-06,
"loss": 0.1351269632577896,
"step": 197
},
{
"epoch": 0.1056563500533618,
"grad_norm": 3.0625,
"learning_rate": 7.879999999999999e-06,
"loss": 0.13437964022159576,
"step": 198
},
{
"epoch": 0.10618996798292422,
"grad_norm": 1.640625,
"learning_rate": 7.92e-06,
"loss": 0.14219871163368225,
"step": 199
},
{
"epoch": 0.10672358591248667,
"grad_norm": 1.4609375,
"learning_rate": 7.96e-06,
"loss": 0.13191944360733032,
"step": 200
},
{
"epoch": 0.10725720384204909,
"grad_norm": 1.328125,
"learning_rate": 8e-06,
"loss": 0.13794036209583282,
"step": 201
},
{
"epoch": 0.10779082177161152,
"grad_norm": 17.25,
"learning_rate": 7.999993660409357e-06,
"loss": 0.1176079660654068,
"step": 202
},
{
"epoch": 0.10832443970117396,
"grad_norm": 1.09375,
"learning_rate": 7.999974641659757e-06,
"loss": 0.15095004439353943,
"step": 203
},
{
"epoch": 0.10885805763073639,
"grad_norm": 1.484375,
"learning_rate": 7.999942943818184e-06,
"loss": 0.13592541217803955,
"step": 204
},
{
"epoch": 0.10939167556029883,
"grad_norm": 1.0859375,
"learning_rate": 7.999898566996276e-06,
"loss": 0.13609889149665833,
"step": 205
},
{
"epoch": 0.10992529348986126,
"grad_norm": 1.6171875,
"learning_rate": 7.999841511350329e-06,
"loss": 0.1379818320274353,
"step": 206
},
{
"epoch": 0.1104589114194237,
"grad_norm": 0.7890625,
"learning_rate": 7.999771777081294e-06,
"loss": 0.14876048266887665,
"step": 207
},
{
"epoch": 0.11099252934898612,
"grad_norm": 1.171875,
"learning_rate": 7.999689364434774e-06,
"loss": 0.1292504370212555,
"step": 208
},
{
"epoch": 0.11152614727854857,
"grad_norm": 0.90234375,
"learning_rate": 7.999594273701023e-06,
"loss": 0.1324600875377655,
"step": 209
},
{
"epoch": 0.11205976520811099,
"grad_norm": 1.375,
"learning_rate": 7.999486505214955e-06,
"loss": 0.13433937728405,
"step": 210
},
{
"epoch": 0.11259338313767342,
"grad_norm": 1.625,
"learning_rate": 7.999366059356127e-06,
"loss": 0.14571602642536163,
"step": 211
},
{
"epoch": 0.11312700106723586,
"grad_norm": 1.171875,
"learning_rate": 7.999232936548748e-06,
"loss": 0.12603814899921417,
"step": 212
},
{
"epoch": 0.11366061899679829,
"grad_norm": 0.87109375,
"learning_rate": 7.999087137261678e-06,
"loss": 0.13884830474853516,
"step": 213
},
{
"epoch": 0.11419423692636073,
"grad_norm": 1.1875,
"learning_rate": 7.998928662008421e-06,
"loss": 0.13220319151878357,
"step": 214
},
{
"epoch": 0.11472785485592316,
"grad_norm": 1.171875,
"learning_rate": 7.998757511347125e-06,
"loss": 0.1456248015165329,
"step": 215
},
{
"epoch": 0.1152614727854856,
"grad_norm": 0.8828125,
"learning_rate": 7.998573685880583e-06,
"loss": 0.12541213631629944,
"step": 216
},
{
"epoch": 0.11579509071504802,
"grad_norm": 1.890625,
"learning_rate": 7.998377186256224e-06,
"loss": 0.1334465742111206,
"step": 217
},
{
"epoch": 0.11632870864461047,
"grad_norm": 1.03125,
"learning_rate": 7.998168013166122e-06,
"loss": 0.13426785171031952,
"step": 218
},
{
"epoch": 0.11686232657417289,
"grad_norm": 0.90234375,
"learning_rate": 7.997946167346984e-06,
"loss": 0.14470884203910828,
"step": 219
},
{
"epoch": 0.11739594450373532,
"grad_norm": 1.21875,
"learning_rate": 7.99771164958015e-06,
"loss": 0.12855111062526703,
"step": 220
},
{
"epoch": 0.11792956243329776,
"grad_norm": 1.796875,
"learning_rate": 7.997464460691585e-06,
"loss": 0.14419671893119812,
"step": 221
},
{
"epoch": 0.11846318036286019,
"grad_norm": 1.6328125,
"learning_rate": 7.997204601551894e-06,
"loss": 0.14565066993236542,
"step": 222
},
{
"epoch": 0.11899679829242263,
"grad_norm": 4.125,
"learning_rate": 7.996932073076294e-06,
"loss": 0.13421288132667542,
"step": 223
},
{
"epoch": 0.11953041622198506,
"grad_norm": 1.3046875,
"learning_rate": 7.996646876224633e-06,
"loss": 0.14643988013267517,
"step": 224
},
{
"epoch": 0.1200640341515475,
"grad_norm": 1.828125,
"learning_rate": 7.99634901200137e-06,
"loss": 0.1383713036775589,
"step": 225
},
{
"epoch": 0.12059765208110992,
"grad_norm": 1.890625,
"learning_rate": 7.996038481455584e-06,
"loss": 0.13766659796237946,
"step": 226
},
{
"epoch": 0.12113127001067236,
"grad_norm": 1.2734375,
"learning_rate": 7.995715285680958e-06,
"loss": 0.15362277626991272,
"step": 227
},
{
"epoch": 0.12166488794023479,
"grad_norm": 3.359375,
"learning_rate": 7.99537942581579e-06,
"loss": 0.1268962025642395,
"step": 228
},
{
"epoch": 0.12219850586979722,
"grad_norm": 2.203125,
"learning_rate": 7.995030903042975e-06,
"loss": 0.14100855588912964,
"step": 229
},
{
"epoch": 0.12273212379935966,
"grad_norm": 6.34375,
"learning_rate": 7.994669718590006e-06,
"loss": 0.1319357454776764,
"step": 230
},
{
"epoch": 0.12326574172892209,
"grad_norm": 1.0390625,
"learning_rate": 7.994295873728976e-06,
"loss": 0.1273908168077469,
"step": 231
},
{
"epoch": 0.12379935965848453,
"grad_norm": 2.546875,
"learning_rate": 7.993909369776566e-06,
"loss": 0.13907387852668762,
"step": 232
},
{
"epoch": 0.12433297758804696,
"grad_norm": 0.8984375,
"learning_rate": 7.993510208094036e-06,
"loss": 0.12650838494300842,
"step": 233
},
{
"epoch": 0.1248665955176094,
"grad_norm": 2.421875,
"learning_rate": 7.993098390087235e-06,
"loss": 0.12302593141794205,
"step": 234
},
{
"epoch": 0.12540021344717184,
"grad_norm": 2.71875,
"learning_rate": 7.992673917206583e-06,
"loss": 0.14238953590393066,
"step": 235
},
{
"epoch": 0.12593383137673425,
"grad_norm": 2.359375,
"learning_rate": 7.99223679094707e-06,
"loss": 0.12786713242530823,
"step": 236
},
{
"epoch": 0.1264674493062967,
"grad_norm": 2.125,
"learning_rate": 7.991787012848255e-06,
"loss": 0.1318153291940689,
"step": 237
},
{
"epoch": 0.12700106723585913,
"grad_norm": 2.140625,
"learning_rate": 7.99132458449425e-06,
"loss": 0.13319279253482819,
"step": 238
},
{
"epoch": 0.12753468516542155,
"grad_norm": 9.6875,
"learning_rate": 7.990849507513731e-06,
"loss": 0.1402091681957245,
"step": 239
},
{
"epoch": 0.128068303094984,
"grad_norm": 1.0078125,
"learning_rate": 7.990361783579912e-06,
"loss": 0.131394624710083,
"step": 240
},
{
"epoch": 0.12860192102454643,
"grad_norm": 4.84375,
"learning_rate": 7.989861414410557e-06,
"loss": 0.1344088315963745,
"step": 241
},
{
"epoch": 0.12913553895410887,
"grad_norm": 3.03125,
"learning_rate": 7.989348401767962e-06,
"loss": 0.14013636112213135,
"step": 242
},
{
"epoch": 0.12966915688367128,
"grad_norm": 2.84375,
"learning_rate": 7.988822747458953e-06,
"loss": 0.13389602303504944,
"step": 243
},
{
"epoch": 0.13020277481323372,
"grad_norm": 1.265625,
"learning_rate": 7.988284453334885e-06,
"loss": 0.13655516505241394,
"step": 244
},
{
"epoch": 0.13073639274279616,
"grad_norm": 1.5,
"learning_rate": 7.987733521291628e-06,
"loss": 0.13951365649700165,
"step": 245
},
{
"epoch": 0.13127001067235858,
"grad_norm": 3.59375,
"learning_rate": 7.987169953269558e-06,
"loss": 0.15002214908599854,
"step": 246
},
{
"epoch": 0.13180362860192102,
"grad_norm": 0.98828125,
"learning_rate": 7.986593751253562e-06,
"loss": 0.13093721866607666,
"step": 247
},
{
"epoch": 0.13233724653148346,
"grad_norm": 0.77734375,
"learning_rate": 7.986004917273018e-06,
"loss": 0.13277417421340942,
"step": 248
},
{
"epoch": 0.1328708644610459,
"grad_norm": 4.4375,
"learning_rate": 7.985403453401796e-06,
"loss": 0.13300329446792603,
"step": 249
},
{
"epoch": 0.13340448239060831,
"grad_norm": 19.75,
"learning_rate": 7.984789361758251e-06,
"loss": 0.1418147087097168,
"step": 250
},
{
"epoch": 0.13393810032017076,
"grad_norm": 18.375,
"learning_rate": 7.98416264450521e-06,
"loss": 0.1333741992712021,
"step": 251
},
{
"epoch": 0.1344717182497332,
"grad_norm": 0.96875,
"learning_rate": 7.983523303849966e-06,
"loss": 0.13330569863319397,
"step": 252
},
{
"epoch": 0.13500533617929564,
"grad_norm": 2.046875,
"learning_rate": 7.982871342044276e-06,
"loss": 0.14242327213287354,
"step": 253
},
{
"epoch": 0.13553895410885805,
"grad_norm": 15.9375,
"learning_rate": 7.982206761384344e-06,
"loss": 0.12544983625411987,
"step": 254
},
{
"epoch": 0.1360725720384205,
"grad_norm": 0.890625,
"learning_rate": 7.981529564210822e-06,
"loss": 0.1369403898715973,
"step": 255
},
{
"epoch": 0.13660618996798293,
"grad_norm": 1.1875,
"learning_rate": 7.980839752908793e-06,
"loss": 0.13040348887443542,
"step": 256
},
{
"epoch": 0.13713980789754535,
"grad_norm": 6.84375,
"learning_rate": 7.980137329907769e-06,
"loss": 0.12282237410545349,
"step": 257
},
{
"epoch": 0.1376734258271078,
"grad_norm": 1.234375,
"learning_rate": 7.97942229768168e-06,
"loss": 0.14523519575595856,
"step": 258
},
{
"epoch": 0.13820704375667023,
"grad_norm": 0.95703125,
"learning_rate": 7.978694658748865e-06,
"loss": 0.13352800905704498,
"step": 259
},
{
"epoch": 0.13874066168623267,
"grad_norm": 1.0078125,
"learning_rate": 7.977954415672067e-06,
"loss": 0.13583557307720184,
"step": 260
},
{
"epoch": 0.13927427961579508,
"grad_norm": 1.15625,
"learning_rate": 7.977201571058419e-06,
"loss": 0.13112758100032806,
"step": 261
},
{
"epoch": 0.13980789754535752,
"grad_norm": 2.09375,
"learning_rate": 7.97643612755943e-06,
"loss": 0.13287177681922913,
"step": 262
},
{
"epoch": 0.14034151547491996,
"grad_norm": 0.78515625,
"learning_rate": 7.975658087870995e-06,
"loss": 0.13495852053165436,
"step": 263
},
{
"epoch": 0.14087513340448238,
"grad_norm": 1.6171875,
"learning_rate": 7.97486745473336e-06,
"loss": 0.11777077615261078,
"step": 264
},
{
"epoch": 0.14140875133404482,
"grad_norm": 1.25,
"learning_rate": 7.974064230931136e-06,
"loss": 0.1354389190673828,
"step": 265
},
{
"epoch": 0.14194236926360726,
"grad_norm": 3.234375,
"learning_rate": 7.973248419293269e-06,
"loss": 0.12675046920776367,
"step": 266
},
{
"epoch": 0.1424759871931697,
"grad_norm": 0.90625,
"learning_rate": 7.972420022693044e-06,
"loss": 0.13410872220993042,
"step": 267
},
{
"epoch": 0.14300960512273211,
"grad_norm": 0.8046875,
"learning_rate": 7.971579044048072e-06,
"loss": 0.12337101250886917,
"step": 268
},
{
"epoch": 0.14354322305229456,
"grad_norm": 1.65625,
"learning_rate": 7.970725486320273e-06,
"loss": 0.15455490350723267,
"step": 269
},
{
"epoch": 0.144076840981857,
"grad_norm": 1.234375,
"learning_rate": 7.969859352515874e-06,
"loss": 0.1363036036491394,
"step": 270
},
{
"epoch": 0.14461045891141944,
"grad_norm": 2.09375,
"learning_rate": 7.968980645685392e-06,
"loss": 0.13511885702610016,
"step": 271
},
{
"epoch": 0.14514407684098185,
"grad_norm": 1.0546875,
"learning_rate": 7.96808936892363e-06,
"loss": 0.14115452766418457,
"step": 272
},
{
"epoch": 0.1456776947705443,
"grad_norm": 0.9921875,
"learning_rate": 7.967185525369661e-06,
"loss": 0.12967218458652496,
"step": 273
},
{
"epoch": 0.14621131270010673,
"grad_norm": 1.75,
"learning_rate": 7.966269118206813e-06,
"loss": 0.16108211874961853,
"step": 274
},
{
"epoch": 0.14674493062966915,
"grad_norm": 0.9765625,
"learning_rate": 7.965340150662671e-06,
"loss": 0.12674804031848907,
"step": 275
},
{
"epoch": 0.1472785485592316,
"grad_norm": 1.1484375,
"learning_rate": 7.964398626009053e-06,
"loss": 0.1353631466627121,
"step": 276
},
{
"epoch": 0.14781216648879403,
"grad_norm": 1.359375,
"learning_rate": 7.963444547562004e-06,
"loss": 0.13910847902297974,
"step": 277
},
{
"epoch": 0.14834578441835647,
"grad_norm": 12.375,
"learning_rate": 7.962477918681781e-06,
"loss": 0.14266562461853027,
"step": 278
},
{
"epoch": 0.14887940234791888,
"grad_norm": 1.1796875,
"learning_rate": 7.96149874277285e-06,
"loss": 0.1355934739112854,
"step": 279
},
{
"epoch": 0.14941302027748132,
"grad_norm": 1.7109375,
"learning_rate": 7.960507023283862e-06,
"loss": 0.13579387962818146,
"step": 280
},
{
"epoch": 0.14994663820704376,
"grad_norm": 1.984375,
"learning_rate": 7.959502763707646e-06,
"loss": 0.14226222038269043,
"step": 281
},
{
"epoch": 0.15048025613660618,
"grad_norm": 0.9921875,
"learning_rate": 7.9584859675812e-06,
"loss": 0.14672532677650452,
"step": 282
},
{
"epoch": 0.15101387406616862,
"grad_norm": 1.0,
"learning_rate": 7.957456638485676e-06,
"loss": 0.1337587833404541,
"step": 283
},
{
"epoch": 0.15154749199573106,
"grad_norm": 1.078125,
"learning_rate": 7.956414780046365e-06,
"loss": 0.13988350331783295,
"step": 284
},
{
"epoch": 0.1520811099252935,
"grad_norm": 1.6171875,
"learning_rate": 7.955360395932685e-06,
"loss": 0.14275816082954407,
"step": 285
},
{
"epoch": 0.1526147278548559,
"grad_norm": 1.5078125,
"learning_rate": 7.954293489858175e-06,
"loss": 0.14317359030246735,
"step": 286
},
{
"epoch": 0.15314834578441835,
"grad_norm": 2.109375,
"learning_rate": 7.95321406558047e-06,
"loss": 0.12176140397787094,
"step": 287
},
{
"epoch": 0.1536819637139808,
"grad_norm": 1.2734375,
"learning_rate": 7.952122126901297e-06,
"loss": 0.11602231860160828,
"step": 288
},
{
"epoch": 0.1542155816435432,
"grad_norm": 2.265625,
"learning_rate": 7.951017677666458e-06,
"loss": 0.13399404287338257,
"step": 289
},
{
"epoch": 0.15474919957310565,
"grad_norm": 1.5546875,
"learning_rate": 7.949900721765819e-06,
"loss": 0.12931866943836212,
"step": 290
},
{
"epoch": 0.1552828175026681,
"grad_norm": 1.1640625,
"learning_rate": 7.948771263133294e-06,
"loss": 0.13807906210422516,
"step": 291
},
{
"epoch": 0.15581643543223053,
"grad_norm": 0.80078125,
"learning_rate": 7.947629305746826e-06,
"loss": 0.14229467511177063,
"step": 292
},
{
"epoch": 0.15635005336179295,
"grad_norm": 1.140625,
"learning_rate": 7.946474853628386e-06,
"loss": 0.1323075294494629,
"step": 293
},
{
"epoch": 0.1568836712913554,
"grad_norm": 1.03125,
"learning_rate": 7.945307910843948e-06,
"loss": 0.13296887278556824,
"step": 294
},
{
"epoch": 0.15741728922091783,
"grad_norm": 0.87109375,
"learning_rate": 7.944128481503479e-06,
"loss": 0.13029369711875916,
"step": 295
},
{
"epoch": 0.15795090715048027,
"grad_norm": 1.4765625,
"learning_rate": 7.942936569760922e-06,
"loss": 0.12771499156951904,
"step": 296
},
{
"epoch": 0.15848452508004268,
"grad_norm": 0.6875,
"learning_rate": 7.941732179814185e-06,
"loss": 0.11567488312721252,
"step": 297
},
{
"epoch": 0.15901814300960512,
"grad_norm": 0.859375,
"learning_rate": 7.94051531590512e-06,
"loss": 0.13120344281196594,
"step": 298
},
{
"epoch": 0.15955176093916756,
"grad_norm": 1.25,
"learning_rate": 7.93928598231952e-06,
"loss": 0.12277650088071823,
"step": 299
},
{
"epoch": 0.16008537886872998,
"grad_norm": 1.2265625,
"learning_rate": 7.938044183387088e-06,
"loss": 0.12671735882759094,
"step": 300
},
{
"epoch": 0.16061899679829242,
"grad_norm": 1.96875,
"learning_rate": 7.936789923481437e-06,
"loss": 0.13567811250686646,
"step": 301
},
{
"epoch": 0.16115261472785486,
"grad_norm": 1.8046875,
"learning_rate": 7.935523207020058e-06,
"loss": 0.14428842067718506,
"step": 302
},
{
"epoch": 0.1616862326574173,
"grad_norm": 1.734375,
"learning_rate": 7.934244038464325e-06,
"loss": 0.13490864634513855,
"step": 303
},
{
"epoch": 0.1622198505869797,
"grad_norm": 1.21875,
"learning_rate": 7.93295242231946e-06,
"loss": 0.12998397648334503,
"step": 304
},
{
"epoch": 0.16275346851654215,
"grad_norm": 0.75,
"learning_rate": 7.931648363134532e-06,
"loss": 0.11724256724119186,
"step": 305
},
{
"epoch": 0.1632870864461046,
"grad_norm": 1.0234375,
"learning_rate": 7.930331865502427e-06,
"loss": 0.13621410727500916,
"step": 306
},
{
"epoch": 0.163820704375667,
"grad_norm": 0.796875,
"learning_rate": 7.929002934059845e-06,
"loss": 0.1306716799736023,
"step": 307
},
{
"epoch": 0.16435432230522945,
"grad_norm": 3.203125,
"learning_rate": 7.927661573487274e-06,
"loss": 0.12586519122123718,
"step": 308
},
{
"epoch": 0.1648879402347919,
"grad_norm": 1.078125,
"learning_rate": 7.92630778850898e-06,
"loss": 0.1371864378452301,
"step": 309
},
{
"epoch": 0.16542155816435433,
"grad_norm": 1.25,
"learning_rate": 7.924941583892987e-06,
"loss": 0.13785448670387268,
"step": 310
},
{
"epoch": 0.16595517609391675,
"grad_norm": 0.81640625,
"learning_rate": 7.92356296445106e-06,
"loss": 0.1386108696460724,
"step": 311
},
{
"epoch": 0.1664887940234792,
"grad_norm": 1.1953125,
"learning_rate": 7.922171935038687e-06,
"loss": 0.1251438409090042,
"step": 312
},
{
"epoch": 0.16702241195304163,
"grad_norm": 1.7109375,
"learning_rate": 7.92076850055507e-06,
"loss": 0.12475191056728363,
"step": 313
},
{
"epoch": 0.16755602988260407,
"grad_norm": 1.8046875,
"learning_rate": 7.919352665943098e-06,
"loss": 0.14221517741680145,
"step": 314
},
{
"epoch": 0.16808964781216648,
"grad_norm": 1.328125,
"learning_rate": 7.91792443618933e-06,
"loss": 0.14704571664333344,
"step": 315
},
{
"epoch": 0.16862326574172892,
"grad_norm": 0.88671875,
"learning_rate": 7.916483816323988e-06,
"loss": 0.13959166407585144,
"step": 316
},
{
"epoch": 0.16915688367129136,
"grad_norm": 1.203125,
"learning_rate": 7.915030811420924e-06,
"loss": 0.13804636895656586,
"step": 317
},
{
"epoch": 0.16969050160085378,
"grad_norm": 2.4375,
"learning_rate": 7.913565426597616e-06,
"loss": 0.14321064949035645,
"step": 318
},
{
"epoch": 0.17022411953041622,
"grad_norm": 0.9296875,
"learning_rate": 7.912087667015138e-06,
"loss": 0.14272166788578033,
"step": 319
},
{
"epoch": 0.17075773745997866,
"grad_norm": 1.0546875,
"learning_rate": 7.910597537878156e-06,
"loss": 0.12815096974372864,
"step": 320
},
{
"epoch": 0.1712913553895411,
"grad_norm": 1.671875,
"learning_rate": 7.909095044434898e-06,
"loss": 0.14385640621185303,
"step": 321
},
{
"epoch": 0.1718249733191035,
"grad_norm": 1.140625,
"learning_rate": 7.907580191977131e-06,
"loss": 0.137839213013649,
"step": 322
},
{
"epoch": 0.17235859124866595,
"grad_norm": 0.80078125,
"learning_rate": 7.906052985840166e-06,
"loss": 0.13666321337223053,
"step": 323
},
{
"epoch": 0.1728922091782284,
"grad_norm": 0.89453125,
"learning_rate": 7.904513431402809e-06,
"loss": 0.1519109159708023,
"step": 324
},
{
"epoch": 0.1734258271077908,
"grad_norm": 3.609375,
"learning_rate": 7.902961534087366e-06,
"loss": 0.13765056431293488,
"step": 325
},
{
"epoch": 0.17395944503735325,
"grad_norm": 0.9765625,
"learning_rate": 7.901397299359608e-06,
"loss": 0.1318364292383194,
"step": 326
},
{
"epoch": 0.1744930629669157,
"grad_norm": 0.8046875,
"learning_rate": 7.899820732728766e-06,
"loss": 0.13353854417800903,
"step": 327
},
{
"epoch": 0.17502668089647813,
"grad_norm": 1.3359375,
"learning_rate": 7.898231839747495e-06,
"loss": 0.12594074010849,
"step": 328
},
{
"epoch": 0.17556029882604055,
"grad_norm": 1.640625,
"learning_rate": 7.896630626011872e-06,
"loss": 0.12517009675502777,
"step": 329
},
{
"epoch": 0.17609391675560299,
"grad_norm": 1.5,
"learning_rate": 7.895017097161362e-06,
"loss": 0.14450788497924805,
"step": 330
},
{
"epoch": 0.17662753468516543,
"grad_norm": 1.8046875,
"learning_rate": 7.893391258878802e-06,
"loss": 0.14543895423412323,
"step": 331
},
{
"epoch": 0.17716115261472787,
"grad_norm": 0.828125,
"learning_rate": 7.89175311689039e-06,
"loss": 0.13184189796447754,
"step": 332
},
{
"epoch": 0.17769477054429028,
"grad_norm": 1.515625,
"learning_rate": 7.890102676965651e-06,
"loss": 0.13581451773643494,
"step": 333
},
{
"epoch": 0.17822838847385272,
"grad_norm": 0.9609375,
"learning_rate": 7.888439944917429e-06,
"loss": 0.1421377956867218,
"step": 334
},
{
"epoch": 0.17876200640341516,
"grad_norm": 0.9765625,
"learning_rate": 7.886764926601856e-06,
"loss": 0.13427680730819702,
"step": 335
},
{
"epoch": 0.17929562433297758,
"grad_norm": 1.0859375,
"learning_rate": 7.885077627918335e-06,
"loss": 0.14583437144756317,
"step": 336
},
{
"epoch": 0.17982924226254002,
"grad_norm": 1.046875,
"learning_rate": 7.883378054809526e-06,
"loss": 0.14192166924476624,
"step": 337
},
{
"epoch": 0.18036286019210246,
"grad_norm": 2.609375,
"learning_rate": 7.881666213261317e-06,
"loss": 0.13646627962589264,
"step": 338
},
{
"epoch": 0.1808964781216649,
"grad_norm": 1.53125,
"learning_rate": 7.879942109302806e-06,
"loss": 0.14525869488716125,
"step": 339
},
{
"epoch": 0.1814300960512273,
"grad_norm": 0.91796875,
"learning_rate": 7.878205749006275e-06,
"loss": 0.13107061386108398,
"step": 340
},
{
"epoch": 0.18196371398078975,
"grad_norm": 1.1796875,
"learning_rate": 7.876457138487177e-06,
"loss": 0.13402347266674042,
"step": 341
},
{
"epoch": 0.1824973319103522,
"grad_norm": 1.1484375,
"learning_rate": 7.87469628390411e-06,
"loss": 0.14611411094665527,
"step": 342
},
{
"epoch": 0.1830309498399146,
"grad_norm": 0.98828125,
"learning_rate": 7.872923191458796e-06,
"loss": 0.14112360775470734,
"step": 343
},
{
"epoch": 0.18356456776947705,
"grad_norm": 0.71875,
"learning_rate": 7.871137867396054e-06,
"loss": 0.14248032867908478,
"step": 344
},
{
"epoch": 0.1840981856990395,
"grad_norm": 1.5234375,
"learning_rate": 7.869340318003787e-06,
"loss": 0.13801157474517822,
"step": 345
},
{
"epoch": 0.18463180362860193,
"grad_norm": 1.703125,
"learning_rate": 7.867530549612956e-06,
"loss": 0.14769896864891052,
"step": 346
},
{
"epoch": 0.18516542155816434,
"grad_norm": 1.453125,
"learning_rate": 7.865708568597554e-06,
"loss": 0.12091437727212906,
"step": 347
},
{
"epoch": 0.18569903948772679,
"grad_norm": 2.328125,
"learning_rate": 7.863874381374592e-06,
"loss": 0.12830129265785217,
"step": 348
},
{
"epoch": 0.18623265741728923,
"grad_norm": 2.109375,
"learning_rate": 7.862027994404065e-06,
"loss": 0.13019058108329773,
"step": 349
},
{
"epoch": 0.18676627534685167,
"grad_norm": 2.453125,
"learning_rate": 7.86016941418894e-06,
"loss": 0.14326529204845428,
"step": 350
},
{
"epoch": 0.18729989327641408,
"grad_norm": 3.75,
"learning_rate": 7.858298647275124e-06,
"loss": 0.13222616910934448,
"step": 351
},
{
"epoch": 0.18783351120597652,
"grad_norm": 0.9296875,
"learning_rate": 7.856415700251452e-06,
"loss": 0.1346774697303772,
"step": 352
},
{
"epoch": 0.18836712913553896,
"grad_norm": 1.84375,
"learning_rate": 7.854520579749651e-06,
"loss": 0.14236673712730408,
"step": 353
},
{
"epoch": 0.18890074706510138,
"grad_norm": 3.015625,
"learning_rate": 7.852613292444328e-06,
"loss": 0.13268648087978363,
"step": 354
},
{
"epoch": 0.18943436499466382,
"grad_norm": 1.0546875,
"learning_rate": 7.850693845052937e-06,
"loss": 0.13026493787765503,
"step": 355
},
{
"epoch": 0.18996798292422626,
"grad_norm": 1.7421875,
"learning_rate": 7.84876224433576e-06,
"loss": 0.13641825318336487,
"step": 356
},
{
"epoch": 0.1905016008537887,
"grad_norm": 4.59375,
"learning_rate": 7.846818497095888e-06,
"loss": 0.1294337511062622,
"step": 357
},
{
"epoch": 0.1910352187833511,
"grad_norm": 1.2890625,
"learning_rate": 7.844862610179187e-06,
"loss": 0.12737271189689636,
"step": 358
},
{
"epoch": 0.19156883671291355,
"grad_norm": 3.890625,
"learning_rate": 7.842894590474282e-06,
"loss": 0.13914188742637634,
"step": 359
},
{
"epoch": 0.192102454642476,
"grad_norm": 1.046875,
"learning_rate": 7.840914444912527e-06,
"loss": 0.12608064711093903,
"step": 360
},
{
"epoch": 0.1926360725720384,
"grad_norm": 1.046875,
"learning_rate": 7.838922180467985e-06,
"loss": 0.16115817427635193,
"step": 361
},
{
"epoch": 0.19316969050160085,
"grad_norm": 0.87890625,
"learning_rate": 7.8369178041574e-06,
"loss": 0.13629929721355438,
"step": 362
},
{
"epoch": 0.1937033084311633,
"grad_norm": 0.82421875,
"learning_rate": 7.834901323040175e-06,
"loss": 0.1376287341117859,
"step": 363
},
{
"epoch": 0.19423692636072573,
"grad_norm": 1.0625,
"learning_rate": 7.832872744218348e-06,
"loss": 0.15028135478496552,
"step": 364
},
{
"epoch": 0.19477054429028814,
"grad_norm": 2.640625,
"learning_rate": 7.83083207483656e-06,
"loss": 0.134278804063797,
"step": 365
},
{
"epoch": 0.19530416221985059,
"grad_norm": 1.546875,
"learning_rate": 7.828779322082042e-06,
"loss": 0.1528017222881317,
"step": 366
},
{
"epoch": 0.19583778014941303,
"grad_norm": 6.8125,
"learning_rate": 7.826714493184575e-06,
"loss": 0.12309601157903671,
"step": 367
},
{
"epoch": 0.19637139807897544,
"grad_norm": 3.234375,
"learning_rate": 7.824637595416476e-06,
"loss": 0.12802067399024963,
"step": 368
},
{
"epoch": 0.19690501600853788,
"grad_norm": 1.0078125,
"learning_rate": 7.82254863609257e-06,
"loss": 0.13978977501392365,
"step": 369
},
{
"epoch": 0.19743863393810032,
"grad_norm": 1.2734375,
"learning_rate": 7.820447622570159e-06,
"loss": 0.12919074296951294,
"step": 370
},
{
"epoch": 0.19797225186766276,
"grad_norm": 1.8203125,
"learning_rate": 7.818334562249002e-06,
"loss": 0.1330573856830597,
"step": 371
},
{
"epoch": 0.19850586979722518,
"grad_norm": 1.53125,
"learning_rate": 7.816209462571289e-06,
"loss": 0.13373148441314697,
"step": 372
},
{
"epoch": 0.19903948772678762,
"grad_norm": 0.8671875,
"learning_rate": 7.814072331021606e-06,
"loss": 0.13213931024074554,
"step": 373
},
{
"epoch": 0.19957310565635006,
"grad_norm": 0.9296875,
"learning_rate": 7.811923175126922e-06,
"loss": 0.1331096887588501,
"step": 374
},
{
"epoch": 0.2001067235859125,
"grad_norm": 1.515625,
"learning_rate": 7.809762002456552e-06,
"loss": 0.11967992782592773,
"step": 375
},
{
"epoch": 0.2006403415154749,
"grad_norm": 1.109375,
"learning_rate": 7.807588820622135e-06,
"loss": 0.14182153344154358,
"step": 376
},
{
"epoch": 0.20117395944503735,
"grad_norm": 8.125,
"learning_rate": 7.805403637277606e-06,
"loss": 0.15451596677303314,
"step": 377
},
{
"epoch": 0.2017075773745998,
"grad_norm": 3.546875,
"learning_rate": 7.80320646011917e-06,
"loss": 0.1453040987253189,
"step": 378
},
{
"epoch": 0.2022411953041622,
"grad_norm": 1.078125,
"learning_rate": 7.800997296885274e-06,
"loss": 0.13652533292770386,
"step": 379
},
{
"epoch": 0.20277481323372465,
"grad_norm": 4.8125,
"learning_rate": 7.798776155356576e-06,
"loss": 0.13986679911613464,
"step": 380
},
{
"epoch": 0.2033084311632871,
"grad_norm": 1.0234375,
"learning_rate": 7.796543043355929e-06,
"loss": 0.12649425864219666,
"step": 381
},
{
"epoch": 0.20384204909284953,
"grad_norm": 1.625,
"learning_rate": 7.794297968748339e-06,
"loss": 0.1368485689163208,
"step": 382
},
{
"epoch": 0.20437566702241194,
"grad_norm": 0.83984375,
"learning_rate": 7.792040939440949e-06,
"loss": 0.14622320234775543,
"step": 383
},
{
"epoch": 0.20490928495197439,
"grad_norm": 0.94921875,
"learning_rate": 7.789771963383006e-06,
"loss": 0.13110113143920898,
"step": 384
},
{
"epoch": 0.20544290288153683,
"grad_norm": 1.640625,
"learning_rate": 7.787491048565827e-06,
"loss": 0.14402681589126587,
"step": 385
},
{
"epoch": 0.20597652081109924,
"grad_norm": 1.9453125,
"learning_rate": 7.785198203022783e-06,
"loss": 0.1352449357509613,
"step": 386
},
{
"epoch": 0.20651013874066168,
"grad_norm": 0.859375,
"learning_rate": 7.78289343482927e-06,
"loss": 0.14319977164268494,
"step": 387
},
{
"epoch": 0.20704375667022412,
"grad_norm": 0.828125,
"learning_rate": 7.780576752102664e-06,
"loss": 0.14240705966949463,
"step": 388
},
{
"epoch": 0.20757737459978656,
"grad_norm": 1.078125,
"learning_rate": 7.778248163002311e-06,
"loss": 0.13885857164859772,
"step": 389
},
{
"epoch": 0.20811099252934898,
"grad_norm": 4.34375,
"learning_rate": 7.77590767572949e-06,
"loss": 0.11973259598016739,
"step": 390
},
{
"epoch": 0.20864461045891142,
"grad_norm": 1.0234375,
"learning_rate": 7.773555298527384e-06,
"loss": 0.13473030924797058,
"step": 391
},
{
"epoch": 0.20917822838847386,
"grad_norm": 1.703125,
"learning_rate": 7.771191039681054e-06,
"loss": 0.13241630792617798,
"step": 392
},
{
"epoch": 0.2097118463180363,
"grad_norm": 3.078125,
"learning_rate": 7.768814907517408e-06,
"loss": 0.13049758970737457,
"step": 393
},
{
"epoch": 0.2102454642475987,
"grad_norm": 1.0625,
"learning_rate": 7.766426910405171e-06,
"loss": 0.1319107711315155,
"step": 394
},
{
"epoch": 0.21077908217716115,
"grad_norm": 0.8671875,
"learning_rate": 7.764027056754854e-06,
"loss": 0.14088526368141174,
"step": 395
},
{
"epoch": 0.2113127001067236,
"grad_norm": 8.625,
"learning_rate": 7.761615355018733e-06,
"loss": 0.1338246911764145,
"step": 396
},
{
"epoch": 0.211846318036286,
"grad_norm": 2.265625,
"learning_rate": 7.759191813690807e-06,
"loss": 0.13377243280410767,
"step": 397
},
{
"epoch": 0.21237993596584845,
"grad_norm": 0.77734375,
"learning_rate": 7.756756441306774e-06,
"loss": 0.14791876077651978,
"step": 398
},
{
"epoch": 0.2129135538954109,
"grad_norm": 1.03125,
"learning_rate": 7.754309246444005e-06,
"loss": 0.13322509825229645,
"step": 399
},
{
"epoch": 0.21344717182497333,
"grad_norm": 1.828125,
"learning_rate": 7.75185023772151e-06,
"loss": 0.1464807689189911,
"step": 400
},
{
"epoch": 0.21398078975453574,
"grad_norm": 1.328125,
"learning_rate": 7.749379423799901e-06,
"loss": 0.13722854852676392,
"step": 401
},
{
"epoch": 0.21451440768409819,
"grad_norm": 1.0859375,
"learning_rate": 7.746896813381373e-06,
"loss": 0.12997561693191528,
"step": 402
},
{
"epoch": 0.21504802561366063,
"grad_norm": 3.1875,
"learning_rate": 7.744402415209666e-06,
"loss": 0.13105672597885132,
"step": 403
},
{
"epoch": 0.21558164354322304,
"grad_norm": 0.79296875,
"learning_rate": 7.74189623807004e-06,
"loss": 0.14374282956123352,
"step": 404
},
{
"epoch": 0.21611526147278548,
"grad_norm": 1.78125,
"learning_rate": 7.739378290789235e-06,
"loss": 0.13468097150325775,
"step": 405
},
{
"epoch": 0.21664887940234792,
"grad_norm": 1.2109375,
"learning_rate": 7.73684858223545e-06,
"loss": 0.14835064113140106,
"step": 406
},
{
"epoch": 0.21718249733191036,
"grad_norm": 1.1953125,
"learning_rate": 7.734307121318307e-06,
"loss": 0.14471489191055298,
"step": 407
},
{
"epoch": 0.21771611526147278,
"grad_norm": 1.1015625,
"learning_rate": 7.731753916988813e-06,
"loss": 0.13708409667015076,
"step": 408
},
{
"epoch": 0.21824973319103522,
"grad_norm": 0.7734375,
"learning_rate": 7.729188978239346e-06,
"loss": 0.1285400390625,
"step": 409
},
{
"epoch": 0.21878335112059766,
"grad_norm": 0.9375,
"learning_rate": 7.726612314103604e-06,
"loss": 0.13628476858139038,
"step": 410
},
{
"epoch": 0.2193169690501601,
"grad_norm": 2.71875,
"learning_rate": 7.724023933656585e-06,
"loss": 0.13666772842407227,
"step": 411
},
{
"epoch": 0.2198505869797225,
"grad_norm": 1.1015625,
"learning_rate": 7.721423846014551e-06,
"loss": 0.14051322638988495,
"step": 412
},
{
"epoch": 0.22038420490928495,
"grad_norm": 1.2578125,
"learning_rate": 7.718812060334996e-06,
"loss": 0.12136445939540863,
"step": 413
},
{
"epoch": 0.2209178228388474,
"grad_norm": 1.2578125,
"learning_rate": 7.71618858581662e-06,
"loss": 0.1474105566740036,
"step": 414
},
{
"epoch": 0.2214514407684098,
"grad_norm": 1.28125,
"learning_rate": 7.713553431699284e-06,
"loss": 0.13776057958602905,
"step": 415
},
{
"epoch": 0.22198505869797225,
"grad_norm": 1.1953125,
"learning_rate": 7.710906607263987e-06,
"loss": 0.13818708062171936,
"step": 416
},
{
"epoch": 0.2225186766275347,
"grad_norm": 1.234375,
"learning_rate": 7.70824812183283e-06,
"loss": 0.13769197463989258,
"step": 417
},
{
"epoch": 0.22305229455709713,
"grad_norm": 1.4296875,
"learning_rate": 7.705577984768987e-06,
"loss": 0.14799988269805908,
"step": 418
},
{
"epoch": 0.22358591248665954,
"grad_norm": 1.046875,
"learning_rate": 7.702896205476664e-06,
"loss": 0.11990436166524887,
"step": 419
},
{
"epoch": 0.22411953041622198,
"grad_norm": 1.0390625,
"learning_rate": 7.700202793401076e-06,
"loss": 0.14538419246673584,
"step": 420
},
{
"epoch": 0.22465314834578443,
"grad_norm": 7.5,
"learning_rate": 7.697497758028406e-06,
"loss": 0.1363571286201477,
"step": 421
},
{
"epoch": 0.22518676627534684,
"grad_norm": 1.1484375,
"learning_rate": 7.694781108885775e-06,
"loss": 0.13654100894927979,
"step": 422
},
{
"epoch": 0.22572038420490928,
"grad_norm": 0.96875,
"learning_rate": 7.692052855541206e-06,
"loss": 0.13123245537281036,
"step": 423
},
{
"epoch": 0.22625400213447172,
"grad_norm": 0.8046875,
"learning_rate": 7.689313007603593e-06,
"loss": 0.1410900354385376,
"step": 424
},
{
"epoch": 0.22678762006403416,
"grad_norm": 1.1171875,
"learning_rate": 7.686561574722666e-06,
"loss": 0.14997439086437225,
"step": 425
},
{
"epoch": 0.22732123799359658,
"grad_norm": 1.8359375,
"learning_rate": 7.68379856658896e-06,
"loss": 0.1494731307029724,
"step": 426
},
{
"epoch": 0.22785485592315902,
"grad_norm": 0.8671875,
"learning_rate": 7.681023992933772e-06,
"loss": 0.13702590763568878,
"step": 427
},
{
"epoch": 0.22838847385272146,
"grad_norm": 3.34375,
"learning_rate": 7.678237863529138e-06,
"loss": 0.14781013131141663,
"step": 428
},
{
"epoch": 0.2289220917822839,
"grad_norm": 0.91796875,
"learning_rate": 7.675440188187789e-06,
"loss": 0.14366307854652405,
"step": 429
},
{
"epoch": 0.2294557097118463,
"grad_norm": 2.203125,
"learning_rate": 7.672630976763126e-06,
"loss": 0.1160091757774353,
"step": 430
},
{
"epoch": 0.22998932764140875,
"grad_norm": 2.671875,
"learning_rate": 7.669810239149176e-06,
"loss": 0.13077133893966675,
"step": 431
},
{
"epoch": 0.2305229455709712,
"grad_norm": 0.94921875,
"learning_rate": 7.66697798528056e-06,
"loss": 0.13839778304100037,
"step": 432
},
{
"epoch": 0.2310565635005336,
"grad_norm": 1.8515625,
"learning_rate": 7.664134225132467e-06,
"loss": 0.1568668782711029,
"step": 433
},
{
"epoch": 0.23159018143009605,
"grad_norm": 0.8984375,
"learning_rate": 7.661278968720598e-06,
"loss": 0.13454318046569824,
"step": 434
},
{
"epoch": 0.2321237993596585,
"grad_norm": 1.1015625,
"learning_rate": 7.658412226101156e-06,
"loss": 0.11559347808361053,
"step": 435
},
{
"epoch": 0.23265741728922093,
"grad_norm": 4.375,
"learning_rate": 7.655534007370792e-06,
"loss": 0.15681754052639008,
"step": 436
},
{
"epoch": 0.23319103521878334,
"grad_norm": 2.734375,
"learning_rate": 7.652644322666578e-06,
"loss": 0.11845934391021729,
"step": 437
},
{
"epoch": 0.23372465314834578,
"grad_norm": 1.1953125,
"learning_rate": 7.64974318216597e-06,
"loss": 0.14034810662269592,
"step": 438
},
{
"epoch": 0.23425827107790823,
"grad_norm": 2.78125,
"learning_rate": 7.646830596086768e-06,
"loss": 0.1531040370464325,
"step": 439
},
{
"epoch": 0.23479188900747064,
"grad_norm": 1.3046875,
"learning_rate": 7.643906574687084e-06,
"loss": 0.12844467163085938,
"step": 440
},
{
"epoch": 0.23532550693703308,
"grad_norm": 1.171875,
"learning_rate": 7.640971128265307e-06,
"loss": 0.1423211395740509,
"step": 441
},
{
"epoch": 0.23585912486659552,
"grad_norm": 1.921875,
"learning_rate": 7.638024267160066e-06,
"loss": 0.1630212366580963,
"step": 442
},
{
"epoch": 0.23639274279615796,
"grad_norm": 7.03125,
"learning_rate": 7.635066001750188e-06,
"loss": 0.11993074417114258,
"step": 443
},
{
"epoch": 0.23692636072572038,
"grad_norm": 0.86328125,
"learning_rate": 7.63209634245467e-06,
"loss": 0.12430194020271301,
"step": 444
},
{
"epoch": 0.23745997865528282,
"grad_norm": 5.625,
"learning_rate": 7.629115299732635e-06,
"loss": 0.1398942768573761,
"step": 445
},
{
"epoch": 0.23799359658484526,
"grad_norm": 3.15625,
"learning_rate": 7.626122884083301e-06,
"loss": 0.1338876485824585,
"step": 446
},
{
"epoch": 0.23852721451440767,
"grad_norm": 1.140625,
"learning_rate": 7.623119106045942e-06,
"loss": 0.1404801309108734,
"step": 447
},
{
"epoch": 0.2390608324439701,
"grad_norm": 22.25,
"learning_rate": 7.620103976199845e-06,
"loss": 0.12357479333877563,
"step": 448
},
{
"epoch": 0.23959445037353255,
"grad_norm": 1.796875,
"learning_rate": 7.617077505164283e-06,
"loss": 0.1408863216638565,
"step": 449
},
{
"epoch": 0.240128068303095,
"grad_norm": 0.96875,
"learning_rate": 7.614039703598474e-06,
"loss": 0.1352657973766327,
"step": 450
},
{
"epoch": 0.2406616862326574,
"grad_norm": 1.6015625,
"learning_rate": 7.610990582201536e-06,
"loss": 0.13711000978946686,
"step": 451
},
{
"epoch": 0.24119530416221985,
"grad_norm": 6.125,
"learning_rate": 7.60793015171246e-06,
"loss": 0.13460087776184082,
"step": 452
},
{
"epoch": 0.2417289220917823,
"grad_norm": 1.0859375,
"learning_rate": 7.604858422910066e-06,
"loss": 0.12724679708480835,
"step": 453
},
{
"epoch": 0.24226254002134473,
"grad_norm": 0.984375,
"learning_rate": 7.601775406612966e-06,
"loss": 0.12873651087284088,
"step": 454
},
{
"epoch": 0.24279615795090714,
"grad_norm": 0.78125,
"learning_rate": 7.59868111367953e-06,
"loss": 0.1379387080669403,
"step": 455
},
{
"epoch": 0.24332977588046958,
"grad_norm": 1.2109375,
"learning_rate": 7.5955755550078386e-06,
"loss": 0.14433333277702332,
"step": 456
},
{
"epoch": 0.24386339381003203,
"grad_norm": 2.734375,
"learning_rate": 7.592458741535655e-06,
"loss": 0.14094425737857819,
"step": 457
},
{
"epoch": 0.24439701173959444,
"grad_norm": 3.796875,
"learning_rate": 7.58933068424038e-06,
"loss": 0.1331767737865448,
"step": 458
},
{
"epoch": 0.24493062966915688,
"grad_norm": 1.03125,
"learning_rate": 7.5861913941390144e-06,
"loss": 0.13529697060585022,
"step": 459
},
{
"epoch": 0.24546424759871932,
"grad_norm": 1.53125,
"learning_rate": 7.583040882288122e-06,
"loss": 0.13087782263755798,
"step": 460
},
{
"epoch": 0.24599786552828176,
"grad_norm": 4.0625,
"learning_rate": 7.579879159783791e-06,
"loss": 0.1278803050518036,
"step": 461
},
{
"epoch": 0.24653148345784417,
"grad_norm": 1.8828125,
"learning_rate": 7.576706237761589e-06,
"loss": 0.13922858238220215,
"step": 462
},
{
"epoch": 0.24706510138740662,
"grad_norm": 0.83203125,
"learning_rate": 7.573522127396531e-06,
"loss": 0.14567750692367554,
"step": 463
},
{
"epoch": 0.24759871931696906,
"grad_norm": 0.875,
"learning_rate": 7.570326839903038e-06,
"loss": 0.1256188303232193,
"step": 464
},
{
"epoch": 0.24813233724653147,
"grad_norm": 2.890625,
"learning_rate": 7.567120386534896e-06,
"loss": 0.14363662898540497,
"step": 465
},
{
"epoch": 0.2486659551760939,
"grad_norm": 2.046875,
"learning_rate": 7.563902778585217e-06,
"loss": 0.15219347178936005,
"step": 466
},
{
"epoch": 0.24919957310565635,
"grad_norm": 0.734375,
"learning_rate": 7.560674027386399e-06,
"loss": 0.12688925862312317,
"step": 467
},
{
"epoch": 0.2497331910352188,
"grad_norm": 0.8125,
"learning_rate": 7.557434144310088e-06,
"loss": 0.1362803876399994,
"step": 468
},
{
"epoch": 0.2502668089647812,
"grad_norm": 1.984375,
"learning_rate": 7.554183140767135e-06,
"loss": 0.12919726967811584,
"step": 469
},
{
"epoch": 0.2508004268943437,
"grad_norm": 0.77734375,
"learning_rate": 7.550921028207555e-06,
"loss": 0.1215812936425209,
"step": 470
},
{
"epoch": 0.2513340448239061,
"grad_norm": 0.89453125,
"learning_rate": 7.547647818120495e-06,
"loss": 0.14418412744998932,
"step": 471
},
{
"epoch": 0.2518676627534685,
"grad_norm": 1.203125,
"learning_rate": 7.544363522034184e-06,
"loss": 0.12645360827445984,
"step": 472
},
{
"epoch": 0.25240128068303097,
"grad_norm": 2.53125,
"learning_rate": 7.541068151515893e-06,
"loss": 0.15419961512088776,
"step": 473
},
{
"epoch": 0.2529348986125934,
"grad_norm": 1.4375,
"learning_rate": 7.5377617181719e-06,
"loss": 0.11574093997478485,
"step": 474
},
{
"epoch": 0.2534685165421558,
"grad_norm": 5.78125,
"learning_rate": 7.534444233647448e-06,
"loss": 0.13848531246185303,
"step": 475
},
{
"epoch": 0.25400213447171827,
"grad_norm": 1.3671875,
"learning_rate": 7.531115709626701e-06,
"loss": 0.13065014779567719,
"step": 476
},
{
"epoch": 0.2545357524012807,
"grad_norm": 1.46875,
"learning_rate": 7.527776157832701e-06,
"loss": 0.12038014829158783,
"step": 477
},
{
"epoch": 0.2550693703308431,
"grad_norm": 1.5234375,
"learning_rate": 7.524425590027334e-06,
"loss": 0.11855950951576233,
"step": 478
},
{
"epoch": 0.25560298826040556,
"grad_norm": 0.9765625,
"learning_rate": 7.5210640180112795e-06,
"loss": 0.1307649314403534,
"step": 479
},
{
"epoch": 0.256136606189968,
"grad_norm": 2.578125,
"learning_rate": 7.517691453623979e-06,
"loss": 0.1379428207874298,
"step": 480
},
{
"epoch": 0.25667022411953044,
"grad_norm": 1.2890625,
"learning_rate": 7.514307908743585e-06,
"loss": 0.12999042868614197,
"step": 481
},
{
"epoch": 0.25720384204909286,
"grad_norm": 1.984375,
"learning_rate": 7.510913395286929e-06,
"loss": 0.13386279344558716,
"step": 482
},
{
"epoch": 0.25773745997865527,
"grad_norm": 0.83984375,
"learning_rate": 7.507507925209464e-06,
"loss": 0.1319030225276947,
"step": 483
},
{
"epoch": 0.25827107790821774,
"grad_norm": 1.28125,
"learning_rate": 7.504091510505243e-06,
"loss": 0.13774380087852478,
"step": 484
},
{
"epoch": 0.25880469583778015,
"grad_norm": 1.6796875,
"learning_rate": 7.500664163206857e-06,
"loss": 0.15231990814208984,
"step": 485
},
{
"epoch": 0.25933831376734257,
"grad_norm": 0.79296875,
"learning_rate": 7.497225895385407e-06,
"loss": 0.13475236296653748,
"step": 486
},
{
"epoch": 0.25987193169690503,
"grad_norm": 1.109375,
"learning_rate": 7.493776719150455e-06,
"loss": 0.12044622004032135,
"step": 487
},
{
"epoch": 0.26040554962646745,
"grad_norm": 0.8359375,
"learning_rate": 7.490316646649982e-06,
"loss": 0.12528805434703827,
"step": 488
},
{
"epoch": 0.26093916755602986,
"grad_norm": 4.625,
"learning_rate": 7.486845690070342e-06,
"loss": 0.1450508087873459,
"step": 489
},
{
"epoch": 0.26147278548559233,
"grad_norm": 0.84765625,
"learning_rate": 7.483363861636231e-06,
"loss": 0.11647328734397888,
"step": 490
},
{
"epoch": 0.26200640341515474,
"grad_norm": 0.88671875,
"learning_rate": 7.479871173610628e-06,
"loss": 0.13036462664604187,
"step": 491
},
{
"epoch": 0.26254002134471716,
"grad_norm": 0.8671875,
"learning_rate": 7.476367638294761e-06,
"loss": 0.13066354393959045,
"step": 492
},
{
"epoch": 0.2630736392742796,
"grad_norm": 0.890625,
"learning_rate": 7.472853268028066e-06,
"loss": 0.13258865475654602,
"step": 493
},
{
"epoch": 0.26360725720384204,
"grad_norm": 0.9609375,
"learning_rate": 7.469328075188135e-06,
"loss": 0.132981538772583,
"step": 494
},
{
"epoch": 0.2641408751334045,
"grad_norm": 0.796875,
"learning_rate": 7.465792072190677e-06,
"loss": 0.13903771340847015,
"step": 495
},
{
"epoch": 0.2646744930629669,
"grad_norm": 2.90625,
"learning_rate": 7.462245271489481e-06,
"loss": 0.12945133447647095,
"step": 496
},
{
"epoch": 0.26520811099252933,
"grad_norm": 1.140625,
"learning_rate": 7.4586876855763555e-06,
"loss": 0.14685314893722534,
"step": 497
},
{
"epoch": 0.2657417289220918,
"grad_norm": 1.8203125,
"learning_rate": 7.455119326981103e-06,
"loss": 0.1491100937128067,
"step": 498
},
{
"epoch": 0.2662753468516542,
"grad_norm": 1.5703125,
"learning_rate": 7.451540208271463e-06,
"loss": 0.14856889843940735,
"step": 499
},
{
"epoch": 0.26680896478121663,
"grad_norm": 1.234375,
"learning_rate": 7.447950342053072e-06,
"loss": 0.14342541992664337,
"step": 500
},
{
"epoch": 0.2673425827107791,
"grad_norm": 3.0,
"learning_rate": 7.444349740969423e-06,
"loss": 0.12624087929725647,
"step": 501
},
{
"epoch": 0.2678762006403415,
"grad_norm": 0.89453125,
"learning_rate": 7.440738417701812e-06,
"loss": 0.13238610327243805,
"step": 502
},
{
"epoch": 0.2684098185699039,
"grad_norm": 3.34375,
"learning_rate": 7.4371163849693e-06,
"loss": 0.13621647655963898,
"step": 503
},
{
"epoch": 0.2689434364994664,
"grad_norm": 0.8125,
"learning_rate": 7.4334836555286694e-06,
"loss": 0.12574607133865356,
"step": 504
},
{
"epoch": 0.2694770544290288,
"grad_norm": 2.125,
"learning_rate": 7.429840242174372e-06,
"loss": 0.1469460129737854,
"step": 505
},
{
"epoch": 0.2700106723585913,
"grad_norm": 0.87890625,
"learning_rate": 7.426186157738494e-06,
"loss": 0.14134004712104797,
"step": 506
},
{
"epoch": 0.2705442902881537,
"grad_norm": 0.8046875,
"learning_rate": 7.4225214150907e-06,
"loss": 0.13122451305389404,
"step": 507
},
{
"epoch": 0.2710779082177161,
"grad_norm": 0.94921875,
"learning_rate": 7.418846027138194e-06,
"loss": 0.14471906423568726,
"step": 508
},
{
"epoch": 0.27161152614727857,
"grad_norm": 0.79296875,
"learning_rate": 7.415160006825674e-06,
"loss": 0.15422360599040985,
"step": 509
},
{
"epoch": 0.272145144076841,
"grad_norm": 1.8125,
"learning_rate": 7.411463367135284e-06,
"loss": 0.12365610897541046,
"step": 510
},
{
"epoch": 0.2726787620064034,
"grad_norm": 0.96484375,
"learning_rate": 7.407756121086571e-06,
"loss": 0.14675500988960266,
"step": 511
},
{
"epoch": 0.27321237993596587,
"grad_norm": 0.73046875,
"learning_rate": 7.404038281736435e-06,
"loss": 0.13742336630821228,
"step": 512
},
{
"epoch": 0.2737459978655283,
"grad_norm": 0.8203125,
"learning_rate": 7.400309862179087e-06,
"loss": 0.13093134760856628,
"step": 513
},
{
"epoch": 0.2742796157950907,
"grad_norm": 1.59375,
"learning_rate": 7.396570875546002e-06,
"loss": 0.12805426120758057,
"step": 514
},
{
"epoch": 0.27481323372465316,
"grad_norm": 0.76171875,
"learning_rate": 7.3928213350058705e-06,
"loss": 0.1354788839817047,
"step": 515
},
{
"epoch": 0.2753468516542156,
"grad_norm": 2.78125,
"learning_rate": 7.389061253764555e-06,
"loss": 0.13904979825019836,
"step": 516
},
{
"epoch": 0.275880469583778,
"grad_norm": 1.0,
"learning_rate": 7.385290645065042e-06,
"loss": 0.13468466699123383,
"step": 517
},
{
"epoch": 0.27641408751334046,
"grad_norm": 2.1875,
"learning_rate": 7.3815095221873976e-06,
"loss": 0.13918437063694,
"step": 518
},
{
"epoch": 0.27694770544290287,
"grad_norm": 2.546875,
"learning_rate": 7.377717898448715e-06,
"loss": 0.14391759037971497,
"step": 519
},
{
"epoch": 0.27748132337246534,
"grad_norm": 2.25,
"learning_rate": 7.373915787203073e-06,
"loss": 0.1485878825187683,
"step": 520
},
{
"epoch": 0.27801494130202775,
"grad_norm": 1.03125,
"learning_rate": 7.370103201841489e-06,
"loss": 0.13913270831108093,
"step": 521
},
{
"epoch": 0.27854855923159016,
"grad_norm": 0.9296875,
"learning_rate": 7.366280155791868e-06,
"loss": 0.1319103240966797,
"step": 522
},
{
"epoch": 0.27908217716115263,
"grad_norm": 1.15625,
"learning_rate": 7.362446662518958e-06,
"loss": 0.124852254986763,
"step": 523
},
{
"epoch": 0.27961579509071505,
"grad_norm": 0.94921875,
"learning_rate": 7.358602735524304e-06,
"loss": 0.12713545560836792,
"step": 524
},
{
"epoch": 0.28014941302027746,
"grad_norm": 0.80859375,
"learning_rate": 7.354748388346195e-06,
"loss": 0.12272897362709045,
"step": 525
},
{
"epoch": 0.28068303094983993,
"grad_norm": 0.83984375,
"learning_rate": 7.350883634559623e-06,
"loss": 0.12072448432445526,
"step": 526
},
{
"epoch": 0.28121664887940234,
"grad_norm": 0.78125,
"learning_rate": 7.347008487776229e-06,
"loss": 0.12963704764842987,
"step": 527
},
{
"epoch": 0.28175026680896476,
"grad_norm": 1.2265625,
"learning_rate": 7.3431229616442615e-06,
"loss": 0.12835931777954102,
"step": 528
},
{
"epoch": 0.2822838847385272,
"grad_norm": 5.84375,
"learning_rate": 7.339227069848524e-06,
"loss": 0.13740059733390808,
"step": 529
},
{
"epoch": 0.28281750266808964,
"grad_norm": 0.73828125,
"learning_rate": 7.335320826110326e-06,
"loss": 0.13635815680027008,
"step": 530
},
{
"epoch": 0.2833511205976521,
"grad_norm": 2.53125,
"learning_rate": 7.331404244187437e-06,
"loss": 0.13703462481498718,
"step": 531
},
{
"epoch": 0.2838847385272145,
"grad_norm": 1.5703125,
"learning_rate": 7.327477337874039e-06,
"loss": 0.14396589994430542,
"step": 532
},
{
"epoch": 0.28441835645677693,
"grad_norm": 3.625,
"learning_rate": 7.3235401210006755e-06,
"loss": 0.12196320295333862,
"step": 533
},
{
"epoch": 0.2849519743863394,
"grad_norm": 1.0625,
"learning_rate": 7.319592607434204e-06,
"loss": 0.13355395197868347,
"step": 534
},
{
"epoch": 0.2854855923159018,
"grad_norm": 0.8125,
"learning_rate": 7.315634811077747e-06,
"loss": 0.13911400735378265,
"step": 535
},
{
"epoch": 0.28601921024546423,
"grad_norm": 0.890625,
"learning_rate": 7.311666745870641e-06,
"loss": 0.1493888795375824,
"step": 536
},
{
"epoch": 0.2865528281750267,
"grad_norm": 1.1171875,
"learning_rate": 7.307688425788393e-06,
"loss": 0.12587997317314148,
"step": 537
},
{
"epoch": 0.2870864461045891,
"grad_norm": 1.9296875,
"learning_rate": 7.303699864842625e-06,
"loss": 0.13854238390922546,
"step": 538
},
{
"epoch": 0.2876200640341515,
"grad_norm": 3.109375,
"learning_rate": 7.299701077081029e-06,
"loss": 0.14322435855865479,
"step": 539
},
{
"epoch": 0.288153681963714,
"grad_norm": 0.96484375,
"learning_rate": 7.295692076587313e-06,
"loss": 0.11785995960235596,
"step": 540
},
{
"epoch": 0.2886872998932764,
"grad_norm": 0.80859375,
"learning_rate": 7.291672877481156e-06,
"loss": 0.1385408341884613,
"step": 541
},
{
"epoch": 0.2892209178228389,
"grad_norm": 1.3203125,
"learning_rate": 7.287643493918158e-06,
"loss": 0.13760049641132355,
"step": 542
},
{
"epoch": 0.2897545357524013,
"grad_norm": 1.015625,
"learning_rate": 7.283603940089785e-06,
"loss": 0.14714114367961884,
"step": 543
},
{
"epoch": 0.2902881536819637,
"grad_norm": 2.71875,
"learning_rate": 7.279554230223325e-06,
"loss": 0.1440201699733734,
"step": 544
},
{
"epoch": 0.29082177161152617,
"grad_norm": 1.3671875,
"learning_rate": 7.275494378581835e-06,
"loss": 0.12643572688102722,
"step": 545
},
{
"epoch": 0.2913553895410886,
"grad_norm": 1.1953125,
"learning_rate": 7.271424399464094e-06,
"loss": 0.13203531503677368,
"step": 546
},
{
"epoch": 0.291889007470651,
"grad_norm": 0.87109375,
"learning_rate": 7.267344307204544e-06,
"loss": 0.14118391275405884,
"step": 547
},
{
"epoch": 0.29242262540021347,
"grad_norm": 0.59375,
"learning_rate": 7.263254116173249e-06,
"loss": 0.11809700727462769,
"step": 548
},
{
"epoch": 0.2929562433297759,
"grad_norm": 1.0,
"learning_rate": 7.2591538407758425e-06,
"loss": 0.13624891638755798,
"step": 549
},
{
"epoch": 0.2934898612593383,
"grad_norm": 0.734375,
"learning_rate": 7.2550434954534725e-06,
"loss": 0.13073931634426117,
"step": 550
},
{
"epoch": 0.29402347918890076,
"grad_norm": 1.0390625,
"learning_rate": 7.250923094682754e-06,
"loss": 0.13803747296333313,
"step": 551
},
{
"epoch": 0.2945570971184632,
"grad_norm": 0.85546875,
"learning_rate": 7.246792652975714e-06,
"loss": 0.11895903944969177,
"step": 552
},
{
"epoch": 0.2950907150480256,
"grad_norm": 0.96875,
"learning_rate": 7.242652184879751e-06,
"loss": 0.1323937177658081,
"step": 553
},
{
"epoch": 0.29562433297758806,
"grad_norm": 1.53125,
"learning_rate": 7.2385017049775705e-06,
"loss": 0.128555566072464,
"step": 554
},
{
"epoch": 0.29615795090715047,
"grad_norm": 0.921875,
"learning_rate": 7.234341227887142e-06,
"loss": 0.12250931560993195,
"step": 555
},
{
"epoch": 0.29669156883671294,
"grad_norm": 1.3046875,
"learning_rate": 7.2301707682616405e-06,
"loss": 0.13445773720741272,
"step": 556
},
{
"epoch": 0.29722518676627535,
"grad_norm": 3.484375,
"learning_rate": 7.225990340789408e-06,
"loss": 0.1285247802734375,
"step": 557
},
{
"epoch": 0.29775880469583776,
"grad_norm": 2.296875,
"learning_rate": 7.221799960193885e-06,
"loss": 0.14772188663482666,
"step": 558
},
{
"epoch": 0.29829242262540023,
"grad_norm": 1.5703125,
"learning_rate": 7.217599641233572e-06,
"loss": 0.1407204270362854,
"step": 559
},
{
"epoch": 0.29882604055496265,
"grad_norm": 1.3671875,
"learning_rate": 7.213389398701969e-06,
"loss": 0.15340951085090637,
"step": 560
},
{
"epoch": 0.29935965848452506,
"grad_norm": 1.1328125,
"learning_rate": 7.20916924742753e-06,
"loss": 0.12952452898025513,
"step": 561
},
{
"epoch": 0.29989327641408753,
"grad_norm": 1.84375,
"learning_rate": 7.204939202273605e-06,
"loss": 0.12380887567996979,
"step": 562
},
{
"epoch": 0.30042689434364994,
"grad_norm": 1.265625,
"learning_rate": 7.200699278138391e-06,
"loss": 0.13221827149391174,
"step": 563
},
{
"epoch": 0.30096051227321235,
"grad_norm": 1.125,
"learning_rate": 7.1964494899548785e-06,
"loss": 0.1264842301607132,
"step": 564
},
{
"epoch": 0.3014941302027748,
"grad_norm": 0.78515625,
"learning_rate": 7.192189852690801e-06,
"loss": 0.11648708581924438,
"step": 565
},
{
"epoch": 0.30202774813233724,
"grad_norm": 1.7421875,
"learning_rate": 7.187920381348576e-06,
"loss": 0.14156103134155273,
"step": 566
},
{
"epoch": 0.3025613660618997,
"grad_norm": 1.3984375,
"learning_rate": 7.183641090965262e-06,
"loss": 0.13588982820510864,
"step": 567
},
{
"epoch": 0.3030949839914621,
"grad_norm": 1.2578125,
"learning_rate": 7.179351996612495e-06,
"loss": 0.14885035157203674,
"step": 568
},
{
"epoch": 0.30362860192102453,
"grad_norm": 2.90625,
"learning_rate": 7.175053113396445e-06,
"loss": 0.133467435836792,
"step": 569
},
{
"epoch": 0.304162219850587,
"grad_norm": 1.2421875,
"learning_rate": 7.170744456457755e-06,
"loss": 0.12200412154197693,
"step": 570
},
{
"epoch": 0.3046958377801494,
"grad_norm": 11.4375,
"learning_rate": 7.166426040971494e-06,
"loss": 0.14917732775211334,
"step": 571
},
{
"epoch": 0.3052294557097118,
"grad_norm": 2.25,
"learning_rate": 7.162097882147099e-06,
"loss": 0.1339735984802246,
"step": 572
},
{
"epoch": 0.3057630736392743,
"grad_norm": 2.203125,
"learning_rate": 7.15775999522832e-06,
"loss": 0.15396058559417725,
"step": 573
},
{
"epoch": 0.3062966915688367,
"grad_norm": 2.0625,
"learning_rate": 7.153412395493175e-06,
"loss": 0.13321927189826965,
"step": 574
},
{
"epoch": 0.3068303094983991,
"grad_norm": 0.80859375,
"learning_rate": 7.149055098253887e-06,
"loss": 0.1440630853176117,
"step": 575
},
{
"epoch": 0.3073639274279616,
"grad_norm": 1.1796875,
"learning_rate": 7.144688118856833e-06,
"loss": 0.14327430725097656,
"step": 576
},
{
"epoch": 0.307897545357524,
"grad_norm": 1.140625,
"learning_rate": 7.140311472682494e-06,
"loss": 0.14091786742210388,
"step": 577
},
{
"epoch": 0.3084311632870864,
"grad_norm": 4.15625,
"learning_rate": 7.1359251751453925e-06,
"loss": 0.1411212682723999,
"step": 578
},
{
"epoch": 0.3089647812166489,
"grad_norm": 23.5,
"learning_rate": 7.131529241694047e-06,
"loss": 0.12116609513759613,
"step": 579
},
{
"epoch": 0.3094983991462113,
"grad_norm": 1.5625,
"learning_rate": 7.127123687810913e-06,
"loss": 0.1400996595621109,
"step": 580
},
{
"epoch": 0.31003201707577377,
"grad_norm": 0.83203125,
"learning_rate": 7.122708529012326e-06,
"loss": 0.1530320644378662,
"step": 581
},
{
"epoch": 0.3105656350053362,
"grad_norm": 1.328125,
"learning_rate": 7.118283780848454e-06,
"loss": 0.1385822594165802,
"step": 582
},
{
"epoch": 0.3110992529348986,
"grad_norm": 1.8046875,
"learning_rate": 7.113849458903238e-06,
"loss": 0.12508198618888855,
"step": 583
},
{
"epoch": 0.31163287086446106,
"grad_norm": 0.8828125,
"learning_rate": 7.1094055787943346e-06,
"loss": 0.14618520438671112,
"step": 584
},
{
"epoch": 0.3121664887940235,
"grad_norm": 0.8515625,
"learning_rate": 7.104952156173068e-06,
"loss": 0.13759379088878632,
"step": 585
},
{
"epoch": 0.3127001067235859,
"grad_norm": 1.25,
"learning_rate": 7.100489206724369e-06,
"loss": 0.13000528514385223,
"step": 586
},
{
"epoch": 0.31323372465314836,
"grad_norm": 1.546875,
"learning_rate": 7.096016746166723e-06,
"loss": 0.1266419142484665,
"step": 587
},
{
"epoch": 0.3137673425827108,
"grad_norm": 1.03125,
"learning_rate": 7.0915347902521126e-06,
"loss": 0.1421024203300476,
"step": 588
},
{
"epoch": 0.3143009605122732,
"grad_norm": 0.8046875,
"learning_rate": 7.087043354765961e-06,
"loss": 0.14126747846603394,
"step": 589
},
{
"epoch": 0.31483457844183566,
"grad_norm": 3.453125,
"learning_rate": 7.082542455527084e-06,
"loss": 0.1435851752758026,
"step": 590
},
{
"epoch": 0.31536819637139807,
"grad_norm": 1.234375,
"learning_rate": 7.078032108387623e-06,
"loss": 0.14625194668769836,
"step": 591
},
{
"epoch": 0.31590181430096054,
"grad_norm": 1.3203125,
"learning_rate": 7.073512329232999e-06,
"loss": 0.12826813757419586,
"step": 592
},
{
"epoch": 0.31643543223052295,
"grad_norm": 3.109375,
"learning_rate": 7.068983133981849e-06,
"loss": 0.11779461801052094,
"step": 593
},
{
"epoch": 0.31696905016008536,
"grad_norm": 1.0,
"learning_rate": 7.064444538585977e-06,
"loss": 0.1403738558292389,
"step": 594
},
{
"epoch": 0.31750266808964783,
"grad_norm": 1.5625,
"learning_rate": 7.05989655903029e-06,
"loss": 0.12142419815063477,
"step": 595
},
{
"epoch": 0.31803628601921025,
"grad_norm": 1.8828125,
"learning_rate": 7.055339211332749e-06,
"loss": 0.13068827986717224,
"step": 596
},
{
"epoch": 0.31856990394877266,
"grad_norm": 1.8125,
"learning_rate": 7.050772511544309e-06,
"loss": 0.14096495509147644,
"step": 597
},
{
"epoch": 0.31910352187833513,
"grad_norm": 0.8671875,
"learning_rate": 7.046196475748866e-06,
"loss": 0.15037214756011963,
"step": 598
},
{
"epoch": 0.31963713980789754,
"grad_norm": 0.8671875,
"learning_rate": 7.04161112006319e-06,
"loss": 0.12753714621067047,
"step": 599
},
{
"epoch": 0.32017075773745995,
"grad_norm": 1.375,
"learning_rate": 7.037016460636881e-06,
"loss": 0.13992270827293396,
"step": 600
},
{
"epoch": 0.3207043756670224,
"grad_norm": 2.296875,
"learning_rate": 7.0324125136523075e-06,
"loss": 0.14280618727207184,
"step": 601
},
{
"epoch": 0.32123799359658484,
"grad_norm": 3.75,
"learning_rate": 7.0277992953245444e-06,
"loss": 0.14896650612354279,
"step": 602
},
{
"epoch": 0.3217716115261473,
"grad_norm": 1.21875,
"learning_rate": 7.023176821901326e-06,
"loss": 0.1543242186307907,
"step": 603
},
{
"epoch": 0.3223052294557097,
"grad_norm": 0.671875,
"learning_rate": 7.018545109662975e-06,
"loss": 0.13416439294815063,
"step": 604
},
{
"epoch": 0.32283884738527213,
"grad_norm": 0.85546875,
"learning_rate": 7.013904174922362e-06,
"loss": 0.1385742723941803,
"step": 605
},
{
"epoch": 0.3233724653148346,
"grad_norm": 1.09375,
"learning_rate": 7.0092540340248315e-06,
"loss": 0.13524127006530762,
"step": 606
},
{
"epoch": 0.323906083244397,
"grad_norm": 2.78125,
"learning_rate": 7.004594703348158e-06,
"loss": 0.13422159850597382,
"step": 607
},
{
"epoch": 0.3244397011739594,
"grad_norm": 1.6875,
"learning_rate": 6.99992619930248e-06,
"loss": 0.14427542686462402,
"step": 608
},
{
"epoch": 0.3249733191035219,
"grad_norm": 1.8671875,
"learning_rate": 6.995248538330243e-06,
"loss": 0.12678152322769165,
"step": 609
},
{
"epoch": 0.3255069370330843,
"grad_norm": 0.87109375,
"learning_rate": 6.990561736906146e-06,
"loss": 0.12315723299980164,
"step": 610
},
{
"epoch": 0.3260405549626467,
"grad_norm": 0.82421875,
"learning_rate": 6.985865811537078e-06,
"loss": 0.14885815978050232,
"step": 611
},
{
"epoch": 0.3265741728922092,
"grad_norm": 0.86328125,
"learning_rate": 6.981160778762066e-06,
"loss": 0.13341489434242249,
"step": 612
},
{
"epoch": 0.3271077908217716,
"grad_norm": 0.9453125,
"learning_rate": 6.976446655152208e-06,
"loss": 0.1255360245704651,
"step": 613
},
{
"epoch": 0.327641408751334,
"grad_norm": 0.8984375,
"learning_rate": 6.9717234573106235e-06,
"loss": 0.12982356548309326,
"step": 614
},
{
"epoch": 0.3281750266808965,
"grad_norm": 1.15625,
"learning_rate": 6.966991201872392e-06,
"loss": 0.12769514322280884,
"step": 615
},
{
"epoch": 0.3287086446104589,
"grad_norm": 0.76171875,
"learning_rate": 6.962249905504493e-06,
"loss": 0.12232755124568939,
"step": 616
},
{
"epoch": 0.32924226254002137,
"grad_norm": 0.96484375,
"learning_rate": 6.957499584905746e-06,
"loss": 0.1321721076965332,
"step": 617
},
{
"epoch": 0.3297758804695838,
"grad_norm": 0.9296875,
"learning_rate": 6.952740256806757e-06,
"loss": 0.1344086229801178,
"step": 618
},
{
"epoch": 0.3303094983991462,
"grad_norm": 1.1015625,
"learning_rate": 6.947971937969852e-06,
"loss": 0.13376781344413757,
"step": 619
},
{
"epoch": 0.33084311632870866,
"grad_norm": 0.796875,
"learning_rate": 6.943194645189028e-06,
"loss": 0.1444610208272934,
"step": 620
},
{
"epoch": 0.3313767342582711,
"grad_norm": 2.359375,
"learning_rate": 6.938408395289886e-06,
"loss": 0.13440082967281342,
"step": 621
},
{
"epoch": 0.3319103521878335,
"grad_norm": 0.84375,
"learning_rate": 6.933613205129569e-06,
"loss": 0.1302616447210312,
"step": 622
},
{
"epoch": 0.33244397011739596,
"grad_norm": 0.84765625,
"learning_rate": 6.9288090915967174e-06,
"loss": 0.12852656841278076,
"step": 623
},
{
"epoch": 0.3329775880469584,
"grad_norm": 1.046875,
"learning_rate": 6.92399607161139e-06,
"loss": 0.13498349487781525,
"step": 624
},
{
"epoch": 0.3335112059765208,
"grad_norm": 1.1171875,
"learning_rate": 6.919174162125018e-06,
"loss": 0.1340879499912262,
"step": 625
},
{
"epoch": 0.33404482390608325,
"grad_norm": 3.203125,
"learning_rate": 6.914343380120344e-06,
"loss": 0.1381758749485016,
"step": 626
},
{
"epoch": 0.33457844183564567,
"grad_norm": 2.046875,
"learning_rate": 6.909503742611356e-06,
"loss": 0.13247257471084595,
"step": 627
},
{
"epoch": 0.33511205976520814,
"grad_norm": 1.015625,
"learning_rate": 6.904655266643234e-06,
"loss": 0.15059956908226013,
"step": 628
},
{
"epoch": 0.33564567769477055,
"grad_norm": 1.984375,
"learning_rate": 6.899797969292281e-06,
"loss": 0.15011340379714966,
"step": 629
},
{
"epoch": 0.33617929562433296,
"grad_norm": 2.484375,
"learning_rate": 6.894931867665876e-06,
"loss": 0.13521313667297363,
"step": 630
},
{
"epoch": 0.33671291355389543,
"grad_norm": 0.83203125,
"learning_rate": 6.890056978902402e-06,
"loss": 0.13881130516529083,
"step": 631
},
{
"epoch": 0.33724653148345785,
"grad_norm": 0.953125,
"learning_rate": 6.885173320171194e-06,
"loss": 0.14302894473075867,
"step": 632
},
{
"epoch": 0.33778014941302026,
"grad_norm": 1.2421875,
"learning_rate": 6.880280908672471e-06,
"loss": 0.15118177235126495,
"step": 633
},
{
"epoch": 0.3383137673425827,
"grad_norm": 1.2578125,
"learning_rate": 6.875379761637282e-06,
"loss": 0.13843782246112823,
"step": 634
},
{
"epoch": 0.33884738527214514,
"grad_norm": 2.15625,
"learning_rate": 6.8704698963274415e-06,
"loss": 0.13262371718883514,
"step": 635
},
{
"epoch": 0.33938100320170755,
"grad_norm": 1.1015625,
"learning_rate": 6.865551330035467e-06,
"loss": 0.1603950560092926,
"step": 636
},
{
"epoch": 0.33991462113127,
"grad_norm": 1.7109375,
"learning_rate": 6.8606240800845255e-06,
"loss": 0.1283581256866455,
"step": 637
},
{
"epoch": 0.34044823906083244,
"grad_norm": 1.546875,
"learning_rate": 6.855688163828367e-06,
"loss": 0.13318578898906708,
"step": 638
},
{
"epoch": 0.3409818569903949,
"grad_norm": 2.328125,
"learning_rate": 6.85074359865126e-06,
"loss": 0.1422712504863739,
"step": 639
},
{
"epoch": 0.3415154749199573,
"grad_norm": 1.1328125,
"learning_rate": 6.845790401967939e-06,
"loss": 0.13412456214427948,
"step": 640
},
{
"epoch": 0.34204909284951973,
"grad_norm": 1.21875,
"learning_rate": 6.840828591223538e-06,
"loss": 0.1477305293083191,
"step": 641
},
{
"epoch": 0.3425827107790822,
"grad_norm": 0.89453125,
"learning_rate": 6.835858183893527e-06,
"loss": 0.13645461201667786,
"step": 642
},
{
"epoch": 0.3431163287086446,
"grad_norm": 0.7421875,
"learning_rate": 6.830879197483653e-06,
"loss": 0.12939335405826569,
"step": 643
},
{
"epoch": 0.343649946638207,
"grad_norm": 0.75,
"learning_rate": 6.825891649529885e-06,
"loss": 0.12402007728815079,
"step": 644
},
{
"epoch": 0.3441835645677695,
"grad_norm": 2.078125,
"learning_rate": 6.820895557598337e-06,
"loss": 0.1287146806716919,
"step": 645
},
{
"epoch": 0.3447171824973319,
"grad_norm": 0.9453125,
"learning_rate": 6.815890939285221e-06,
"loss": 0.1469326913356781,
"step": 646
},
{
"epoch": 0.3452508004268943,
"grad_norm": 3.09375,
"learning_rate": 6.810877812216775e-06,
"loss": 0.1412142515182495,
"step": 647
},
{
"epoch": 0.3457844183564568,
"grad_norm": 0.83984375,
"learning_rate": 6.805856194049208e-06,
"loss": 0.14642395079135895,
"step": 648
},
{
"epoch": 0.3463180362860192,
"grad_norm": 1.296875,
"learning_rate": 6.800826102468631e-06,
"loss": 0.13384662568569183,
"step": 649
},
{
"epoch": 0.3468516542155816,
"grad_norm": 3.21875,
"learning_rate": 6.795787555191003e-06,
"loss": 0.15146517753601074,
"step": 650
},
{
"epoch": 0.3473852721451441,
"grad_norm": 0.7578125,
"learning_rate": 6.790740569962061e-06,
"loss": 0.12009504437446594,
"step": 651
},
{
"epoch": 0.3479188900747065,
"grad_norm": 3.609375,
"learning_rate": 6.785685164557258e-06,
"loss": 0.1521005630493164,
"step": 652
},
{
"epoch": 0.34845250800426897,
"grad_norm": 1.0859375,
"learning_rate": 6.780621356781708e-06,
"loss": 0.13955560326576233,
"step": 653
},
{
"epoch": 0.3489861259338314,
"grad_norm": 0.76171875,
"learning_rate": 6.775549164470116e-06,
"loss": 0.12796378135681152,
"step": 654
},
{
"epoch": 0.3495197438633938,
"grad_norm": 1.28125,
"learning_rate": 6.770468605486714e-06,
"loss": 0.14650356769561768,
"step": 655
},
{
"epoch": 0.35005336179295626,
"grad_norm": 32.0,
"learning_rate": 6.7653796977252084e-06,
"loss": 0.13852408528327942,
"step": 656
},
{
"epoch": 0.3505869797225187,
"grad_norm": 1.140625,
"learning_rate": 6.7602824591087034e-06,
"loss": 0.1345336139202118,
"step": 657
},
{
"epoch": 0.3511205976520811,
"grad_norm": 1.2734375,
"learning_rate": 6.755176907589647e-06,
"loss": 0.14335167407989502,
"step": 658
},
{
"epoch": 0.35165421558164356,
"grad_norm": 0.9765625,
"learning_rate": 6.750063061149766e-06,
"loss": 0.11590323597192764,
"step": 659
},
{
"epoch": 0.35218783351120597,
"grad_norm": 0.75390625,
"learning_rate": 6.744940937800001e-06,
"loss": 0.14610867202281952,
"step": 660
},
{
"epoch": 0.3527214514407684,
"grad_norm": 1.0078125,
"learning_rate": 6.739810555580441e-06,
"loss": 0.13389088213443756,
"step": 661
},
{
"epoch": 0.35325506937033085,
"grad_norm": 1.96875,
"learning_rate": 6.7346719325602694e-06,
"loss": 0.12581652402877808,
"step": 662
},
{
"epoch": 0.35378868729989327,
"grad_norm": 1.1796875,
"learning_rate": 6.729525086837688e-06,
"loss": 0.14650002121925354,
"step": 663
},
{
"epoch": 0.35432230522945574,
"grad_norm": 1.015625,
"learning_rate": 6.7243700365398595e-06,
"loss": 0.13003048300743103,
"step": 664
},
{
"epoch": 0.35485592315901815,
"grad_norm": 0.8515625,
"learning_rate": 6.7192067998228465e-06,
"loss": 0.1287475973367691,
"step": 665
},
{
"epoch": 0.35538954108858056,
"grad_norm": 0.89453125,
"learning_rate": 6.714035394871542e-06,
"loss": 0.129266619682312,
"step": 666
},
{
"epoch": 0.35592315901814303,
"grad_norm": 1.390625,
"learning_rate": 6.708855839899605e-06,
"loss": 0.13571710884571075,
"step": 667
},
{
"epoch": 0.35645677694770544,
"grad_norm": 0.765625,
"learning_rate": 6.703668153149403e-06,
"loss": 0.1408233940601349,
"step": 668
},
{
"epoch": 0.35699039487726786,
"grad_norm": 0.96875,
"learning_rate": 6.69847235289194e-06,
"loss": 0.12228646874427795,
"step": 669
},
{
"epoch": 0.3575240128068303,
"grad_norm": 1.234375,
"learning_rate": 6.6932684574268e-06,
"loss": 0.1440693736076355,
"step": 670
},
{
"epoch": 0.35805763073639274,
"grad_norm": 1.5546875,
"learning_rate": 6.688056485082074e-06,
"loss": 0.13073007762432098,
"step": 671
},
{
"epoch": 0.35859124866595515,
"grad_norm": 1.1640625,
"learning_rate": 6.682836454214302e-06,
"loss": 0.13988971710205078,
"step": 672
},
{
"epoch": 0.3591248665955176,
"grad_norm": 1.421875,
"learning_rate": 6.6776083832084055e-06,
"loss": 0.13271759450435638,
"step": 673
},
{
"epoch": 0.35965848452508004,
"grad_norm": 0.8125,
"learning_rate": 6.672372290477625e-06,
"loss": 0.12594851851463318,
"step": 674
},
{
"epoch": 0.36019210245464245,
"grad_norm": 9.625,
"learning_rate": 6.667128194463449e-06,
"loss": 0.1308656483888626,
"step": 675
},
{
"epoch": 0.3607257203842049,
"grad_norm": 4.875,
"learning_rate": 6.661876113635558e-06,
"loss": 0.14575400948524475,
"step": 676
},
{
"epoch": 0.36125933831376733,
"grad_norm": 2.578125,
"learning_rate": 6.656616066491753e-06,
"loss": 0.146834135055542,
"step": 677
},
{
"epoch": 0.3617929562433298,
"grad_norm": 1.0234375,
"learning_rate": 6.651348071557892e-06,
"loss": 0.13680881261825562,
"step": 678
},
{
"epoch": 0.3623265741728922,
"grad_norm": 0.890625,
"learning_rate": 6.646072147387828e-06,
"loss": 0.13356544077396393,
"step": 679
},
{
"epoch": 0.3628601921024546,
"grad_norm": 0.7890625,
"learning_rate": 6.640788312563337e-06,
"loss": 0.14331340789794922,
"step": 680
},
{
"epoch": 0.3633938100320171,
"grad_norm": 0.98828125,
"learning_rate": 6.635496585694057e-06,
"loss": 0.13060417771339417,
"step": 681
},
{
"epoch": 0.3639274279615795,
"grad_norm": 0.77734375,
"learning_rate": 6.630196985417424e-06,
"loss": 0.13111916184425354,
"step": 682
},
{
"epoch": 0.3644610458911419,
"grad_norm": 0.7109375,
"learning_rate": 6.624889530398602e-06,
"loss": 0.1279551088809967,
"step": 683
},
{
"epoch": 0.3649946638207044,
"grad_norm": 0.80078125,
"learning_rate": 6.61957423933042e-06,
"loss": 0.14413878321647644,
"step": 684
},
{
"epoch": 0.3655282817502668,
"grad_norm": 0.703125,
"learning_rate": 6.614251130933305e-06,
"loss": 0.1491723358631134,
"step": 685
},
{
"epoch": 0.3660618996798292,
"grad_norm": 0.75390625,
"learning_rate": 6.608920223955218e-06,
"loss": 0.14138883352279663,
"step": 686
},
{
"epoch": 0.3665955176093917,
"grad_norm": 0.65625,
"learning_rate": 6.603581537171586e-06,
"loss": 0.12726926803588867,
"step": 687
},
{
"epoch": 0.3671291355389541,
"grad_norm": 1.609375,
"learning_rate": 6.598235089385235e-06,
"loss": 0.12939371168613434,
"step": 688
},
{
"epoch": 0.36766275346851657,
"grad_norm": 5.375,
"learning_rate": 6.592880899426327e-06,
"loss": 0.12853409349918365,
"step": 689
},
{
"epoch": 0.368196371398079,
"grad_norm": 0.8984375,
"learning_rate": 6.58751898615229e-06,
"loss": 0.13626736402511597,
"step": 690
},
{
"epoch": 0.3687299893276414,
"grad_norm": 1.1484375,
"learning_rate": 6.5821493684477564e-06,
"loss": 0.13427415490150452,
"step": 691
},
{
"epoch": 0.36926360725720386,
"grad_norm": 1.515625,
"learning_rate": 6.576772065224492e-06,
"loss": 0.13539370894432068,
"step": 692
},
{
"epoch": 0.3697972251867663,
"grad_norm": 1.2578125,
"learning_rate": 6.571387095421327e-06,
"loss": 0.12400360405445099,
"step": 693
},
{
"epoch": 0.3703308431163287,
"grad_norm": 1.4296875,
"learning_rate": 6.5659944780041005e-06,
"loss": 0.14972186088562012,
"step": 694
},
{
"epoch": 0.37086446104589116,
"grad_norm": 0.85546875,
"learning_rate": 6.560594231965584e-06,
"loss": 0.1333673596382141,
"step": 695
},
{
"epoch": 0.37139807897545357,
"grad_norm": 0.81640625,
"learning_rate": 6.555186376325414e-06,
"loss": 0.1278214305639267,
"step": 696
},
{
"epoch": 0.371931696905016,
"grad_norm": 0.87109375,
"learning_rate": 6.54977093013003e-06,
"loss": 0.13962261378765106,
"step": 697
},
{
"epoch": 0.37246531483457845,
"grad_norm": 0.828125,
"learning_rate": 6.544347912452606e-06,
"loss": 0.123472198843956,
"step": 698
},
{
"epoch": 0.37299893276414087,
"grad_norm": 15.6875,
"learning_rate": 6.5389173423929816e-06,
"loss": 0.13458308577537537,
"step": 699
},
{
"epoch": 0.37353255069370334,
"grad_norm": 0.625,
"learning_rate": 6.533479239077596e-06,
"loss": 0.13197621703147888,
"step": 700
},
{
"epoch": 0.37406616862326575,
"grad_norm": 0.87109375,
"learning_rate": 6.528033621659421e-06,
"loss": 0.13895131647586823,
"step": 701
},
{
"epoch": 0.37459978655282816,
"grad_norm": 6.09375,
"learning_rate": 6.522580509317894e-06,
"loss": 0.11831393837928772,
"step": 702
},
{
"epoch": 0.37513340448239063,
"grad_norm": 1.6875,
"learning_rate": 6.517119921258847e-06,
"loss": 0.13115744292736053,
"step": 703
},
{
"epoch": 0.37566702241195304,
"grad_norm": 4.3125,
"learning_rate": 6.511651876714444e-06,
"loss": 0.13348263502120972,
"step": 704
},
{
"epoch": 0.37620064034151546,
"grad_norm": 1.4140625,
"learning_rate": 6.506176394943107e-06,
"loss": 0.13316692411899567,
"step": 705
},
{
"epoch": 0.3767342582710779,
"grad_norm": 0.87109375,
"learning_rate": 6.500693495229457e-06,
"loss": 0.13452118635177612,
"step": 706
},
{
"epoch": 0.37726787620064034,
"grad_norm": 1.0,
"learning_rate": 6.4952031968842366e-06,
"loss": 0.1287904679775238,
"step": 707
},
{
"epoch": 0.37780149413020275,
"grad_norm": 1.9375,
"learning_rate": 6.48970551924425e-06,
"loss": 0.13382476568222046,
"step": 708
},
{
"epoch": 0.3783351120597652,
"grad_norm": 1.15625,
"learning_rate": 6.484200481672287e-06,
"loss": 0.13561445474624634,
"step": 709
},
{
"epoch": 0.37886872998932764,
"grad_norm": 1.6640625,
"learning_rate": 6.478688103557062e-06,
"loss": 0.13765332102775574,
"step": 710
},
{
"epoch": 0.37940234791889005,
"grad_norm": 1.109375,
"learning_rate": 6.473168404313143e-06,
"loss": 0.12656934559345245,
"step": 711
},
{
"epoch": 0.3799359658484525,
"grad_norm": 1.234375,
"learning_rate": 6.467641403380881e-06,
"loss": 0.12931948900222778,
"step": 712
},
{
"epoch": 0.38046958377801493,
"grad_norm": 1.0625,
"learning_rate": 6.462107120226346e-06,
"loss": 0.12264946103096008,
"step": 713
},
{
"epoch": 0.3810032017075774,
"grad_norm": 0.984375,
"learning_rate": 6.456565574341252e-06,
"loss": 0.13883116841316223,
"step": 714
},
{
"epoch": 0.3815368196371398,
"grad_norm": 2.3125,
"learning_rate": 6.451016785242897e-06,
"loss": 0.13053050637245178,
"step": 715
},
{
"epoch": 0.3820704375667022,
"grad_norm": 1.6328125,
"learning_rate": 6.445460772474086e-06,
"loss": 0.13091818988323212,
"step": 716
},
{
"epoch": 0.3826040554962647,
"grad_norm": 1.3984375,
"learning_rate": 6.439897555603068e-06,
"loss": 0.12202519178390503,
"step": 717
},
{
"epoch": 0.3831376734258271,
"grad_norm": 1.015625,
"learning_rate": 6.434327154223464e-06,
"loss": 0.13297173380851746,
"step": 718
},
{
"epoch": 0.3836712913553895,
"grad_norm": 0.890625,
"learning_rate": 6.4287495879541965e-06,
"loss": 0.12986010313034058,
"step": 719
},
{
"epoch": 0.384204909284952,
"grad_norm": 0.80078125,
"learning_rate": 6.423164876439429e-06,
"loss": 0.1417757272720337,
"step": 720
},
{
"epoch": 0.3847385272145144,
"grad_norm": 0.703125,
"learning_rate": 6.417573039348483e-06,
"loss": 0.12481685727834702,
"step": 721
},
{
"epoch": 0.3852721451440768,
"grad_norm": 3.140625,
"learning_rate": 6.411974096375782e-06,
"loss": 0.13195040822029114,
"step": 722
},
{
"epoch": 0.3858057630736393,
"grad_norm": 1.125,
"learning_rate": 6.4063680672407705e-06,
"loss": 0.13351789116859436,
"step": 723
},
{
"epoch": 0.3863393810032017,
"grad_norm": 1.6953125,
"learning_rate": 6.400754971687858e-06,
"loss": 0.12507925927639008,
"step": 724
},
{
"epoch": 0.38687299893276417,
"grad_norm": 1.6875,
"learning_rate": 6.395134829486336e-06,
"loss": 0.14225712418556213,
"step": 725
},
{
"epoch": 0.3874066168623266,
"grad_norm": 1.03125,
"learning_rate": 6.389507660430315e-06,
"loss": 0.13449649512767792,
"step": 726
},
{
"epoch": 0.387940234791889,
"grad_norm": 1.015625,
"learning_rate": 6.383873484338658e-06,
"loss": 0.13416430354118347,
"step": 727
},
{
"epoch": 0.38847385272145146,
"grad_norm": 1.28125,
"learning_rate": 6.378232321054901e-06,
"loss": 0.13658566772937775,
"step": 728
},
{
"epoch": 0.3890074706510139,
"grad_norm": 0.9921875,
"learning_rate": 6.372584190447195e-06,
"loss": 0.13421979546546936,
"step": 729
},
{
"epoch": 0.3895410885805763,
"grad_norm": 0.9609375,
"learning_rate": 6.366929112408223e-06,
"loss": 0.15014611184597015,
"step": 730
},
{
"epoch": 0.39007470651013876,
"grad_norm": 0.9921875,
"learning_rate": 6.361267106855142e-06,
"loss": 0.13813111186027527,
"step": 731
},
{
"epoch": 0.39060832443970117,
"grad_norm": 0.97265625,
"learning_rate": 6.3555981937295074e-06,
"loss": 0.12961573898792267,
"step": 732
},
{
"epoch": 0.3911419423692636,
"grad_norm": 1.1015625,
"learning_rate": 6.349922392997201e-06,
"loss": 0.13667255640029907,
"step": 733
},
{
"epoch": 0.39167556029882605,
"grad_norm": 1.0625,
"learning_rate": 6.344239724648362e-06,
"loss": 0.12993544340133667,
"step": 734
},
{
"epoch": 0.39220917822838847,
"grad_norm": 0.70703125,
"learning_rate": 6.33855020869732e-06,
"loss": 0.12575972080230713,
"step": 735
},
{
"epoch": 0.3927427961579509,
"grad_norm": 0.765625,
"learning_rate": 6.332853865182521e-06,
"loss": 0.12264174222946167,
"step": 736
},
{
"epoch": 0.39327641408751335,
"grad_norm": 0.93359375,
"learning_rate": 6.327150714166456e-06,
"loss": 0.1263483762741089,
"step": 737
},
{
"epoch": 0.39381003201707576,
"grad_norm": 0.828125,
"learning_rate": 6.321440775735594e-06,
"loss": 0.14554113149642944,
"step": 738
},
{
"epoch": 0.39434364994663823,
"grad_norm": 1.078125,
"learning_rate": 6.31572407000031e-06,
"loss": 0.14253820478916168,
"step": 739
},
{
"epoch": 0.39487726787620064,
"grad_norm": 1.234375,
"learning_rate": 6.31000061709481e-06,
"loss": 0.1496836543083191,
"step": 740
},
{
"epoch": 0.39541088580576306,
"grad_norm": 1.125,
"learning_rate": 6.304270437177064e-06,
"loss": 0.14495335519313812,
"step": 741
},
{
"epoch": 0.3959445037353255,
"grad_norm": 0.8984375,
"learning_rate": 6.29853355042874e-06,
"loss": 0.15374426543712616,
"step": 742
},
{
"epoch": 0.39647812166488794,
"grad_norm": 1.1640625,
"learning_rate": 6.29278997705512e-06,
"loss": 0.14872749149799347,
"step": 743
},
{
"epoch": 0.39701173959445035,
"grad_norm": 0.85546875,
"learning_rate": 6.287039737285039e-06,
"loss": 0.14169594645500183,
"step": 744
},
{
"epoch": 0.3975453575240128,
"grad_norm": 2.71875,
"learning_rate": 6.281282851370814e-06,
"loss": 0.1371726244688034,
"step": 745
},
{
"epoch": 0.39807897545357523,
"grad_norm": 1.671875,
"learning_rate": 6.275519339588167e-06,
"loss": 0.14400893449783325,
"step": 746
},
{
"epoch": 0.39861259338313765,
"grad_norm": 1.3046875,
"learning_rate": 6.269749222236155e-06,
"loss": 0.11802718043327332,
"step": 747
},
{
"epoch": 0.3991462113127001,
"grad_norm": 1.140625,
"learning_rate": 6.263972519637103e-06,
"loss": 0.13210949301719666,
"step": 748
},
{
"epoch": 0.39967982924226253,
"grad_norm": 0.72265625,
"learning_rate": 6.2581892521365265e-06,
"loss": 0.1358518749475479,
"step": 749
},
{
"epoch": 0.400213447171825,
"grad_norm": 13.25,
"learning_rate": 6.2523994401030646e-06,
"loss": 0.1365397870540619,
"step": 750
},
{
"epoch": 0.4007470651013874,
"grad_norm": 1.0625,
"learning_rate": 6.246603103928402e-06,
"loss": 0.14109095931053162,
"step": 751
},
{
"epoch": 0.4012806830309498,
"grad_norm": 0.734375,
"learning_rate": 6.240800264027208e-06,
"loss": 0.14907963573932648,
"step": 752
},
{
"epoch": 0.4018143009605123,
"grad_norm": 2.875,
"learning_rate": 6.234990940837054e-06,
"loss": 0.14945384860038757,
"step": 753
},
{
"epoch": 0.4023479188900747,
"grad_norm": 2.09375,
"learning_rate": 6.2291751548183455e-06,
"loss": 0.12680521607398987,
"step": 754
},
{
"epoch": 0.4028815368196371,
"grad_norm": 1.03125,
"learning_rate": 6.223352926454251e-06,
"loss": 0.1281912624835968,
"step": 755
},
{
"epoch": 0.4034151547491996,
"grad_norm": 1.0390625,
"learning_rate": 6.2175242762506275e-06,
"loss": 0.13143879175186157,
"step": 756
},
{
"epoch": 0.403948772678762,
"grad_norm": 0.71875,
"learning_rate": 6.211689224735952e-06,
"loss": 0.13112524151802063,
"step": 757
},
{
"epoch": 0.4044823906083244,
"grad_norm": 5.625,
"learning_rate": 6.205847792461244e-06,
"loss": 0.13007766008377075,
"step": 758
},
{
"epoch": 0.4050160085378869,
"grad_norm": 14.0,
"learning_rate": 6.2e-06,
"loss": 0.11255597323179245,
"step": 759
},
{
"epoch": 0.4055496264674493,
"grad_norm": 1.0703125,
"learning_rate": 6.194145867948113e-06,
"loss": 0.13527894020080566,
"step": 760
},
{
"epoch": 0.40608324439701177,
"grad_norm": 0.61328125,
"learning_rate": 6.188285416923805e-06,
"loss": 0.13225793838500977,
"step": 761
},
{
"epoch": 0.4066168623265742,
"grad_norm": 0.70703125,
"learning_rate": 6.182418667567555e-06,
"loss": 0.13888144493103027,
"step": 762
},
{
"epoch": 0.4071504802561366,
"grad_norm": 1.2890625,
"learning_rate": 6.176545640542025e-06,
"loss": 0.14986646175384521,
"step": 763
},
{
"epoch": 0.40768409818569906,
"grad_norm": 0.94921875,
"learning_rate": 6.170666356531984e-06,
"loss": 0.12005878984928131,
"step": 764
},
{
"epoch": 0.4082177161152615,
"grad_norm": 1.078125,
"learning_rate": 6.1647808362442404e-06,
"loss": 0.14381523430347443,
"step": 765
},
{
"epoch": 0.4087513340448239,
"grad_norm": 0.99609375,
"learning_rate": 6.1588891004075654e-06,
"loss": 0.13813461363315582,
"step": 766
},
{
"epoch": 0.40928495197438636,
"grad_norm": 1.0625,
"learning_rate": 6.152991169772623e-06,
"loss": 0.13620099425315857,
"step": 767
},
{
"epoch": 0.40981856990394877,
"grad_norm": 0.89453125,
"learning_rate": 6.147087065111892e-06,
"loss": 0.130196213722229,
"step": 768
},
{
"epoch": 0.4103521878335112,
"grad_norm": 0.96484375,
"learning_rate": 6.1411768072196e-06,
"loss": 0.13783004879951477,
"step": 769
},
{
"epoch": 0.41088580576307365,
"grad_norm": 1.015625,
"learning_rate": 6.135260416911645e-06,
"loss": 0.13868732750415802,
"step": 770
},
{
"epoch": 0.41141942369263607,
"grad_norm": 0.859375,
"learning_rate": 6.129337915025523e-06,
"loss": 0.13803982734680176,
"step": 771
},
{
"epoch": 0.4119530416221985,
"grad_norm": 1.375,
"learning_rate": 6.123409322420253e-06,
"loss": 0.13283872604370117,
"step": 772
},
{
"epoch": 0.41248665955176095,
"grad_norm": 1.8203125,
"learning_rate": 6.1174746599763105e-06,
"loss": 0.13682179152965546,
"step": 773
},
{
"epoch": 0.41302027748132336,
"grad_norm": 1.3046875,
"learning_rate": 6.111533948595542e-06,
"loss": 0.12449363619089127,
"step": 774
},
{
"epoch": 0.41355389541088583,
"grad_norm": 3.671875,
"learning_rate": 6.105587209201106e-06,
"loss": 0.12873905897140503,
"step": 775
},
{
"epoch": 0.41408751334044824,
"grad_norm": 8.1875,
"learning_rate": 6.0996344627373836e-06,
"loss": 0.12084288895130157,
"step": 776
},
{
"epoch": 0.41462113127001066,
"grad_norm": 0.875,
"learning_rate": 6.09367573016992e-06,
"loss": 0.11813464760780334,
"step": 777
},
{
"epoch": 0.4151547491995731,
"grad_norm": 0.69140625,
"learning_rate": 6.087711032485339e-06,
"loss": 0.13246241211891174,
"step": 778
},
{
"epoch": 0.41568836712913554,
"grad_norm": 0.890625,
"learning_rate": 6.0817403906912745e-06,
"loss": 0.14209003746509552,
"step": 779
},
{
"epoch": 0.41622198505869795,
"grad_norm": 0.6796875,
"learning_rate": 6.075763825816299e-06,
"loss": 0.129084974527359,
"step": 780
},
{
"epoch": 0.4167556029882604,
"grad_norm": 0.80078125,
"learning_rate": 6.069781358909837e-06,
"loss": 0.12760519981384277,
"step": 781
},
{
"epoch": 0.41728922091782283,
"grad_norm": 0.6953125,
"learning_rate": 6.063793011042112e-06,
"loss": 0.1295960247516632,
"step": 782
},
{
"epoch": 0.41782283884738525,
"grad_norm": 1.5078125,
"learning_rate": 6.05779880330405e-06,
"loss": 0.15884271264076233,
"step": 783
},
{
"epoch": 0.4183564567769477,
"grad_norm": 1.625,
"learning_rate": 6.051798756807222e-06,
"loss": 0.14957112073898315,
"step": 784
},
{
"epoch": 0.41889007470651013,
"grad_norm": 1.1640625,
"learning_rate": 6.045792892683758e-06,
"loss": 0.1319868564605713,
"step": 785
},
{
"epoch": 0.4194236926360726,
"grad_norm": 0.95703125,
"learning_rate": 6.039781232086281e-06,
"loss": 0.14036314189434052,
"step": 786
},
{
"epoch": 0.419957310565635,
"grad_norm": 1.109375,
"learning_rate": 6.033763796187831e-06,
"loss": 0.14379188418388367,
"step": 787
},
{
"epoch": 0.4204909284951974,
"grad_norm": 0.66015625,
"learning_rate": 6.0277406061817835e-06,
"loss": 0.12042712420225143,
"step": 788
},
{
"epoch": 0.4210245464247599,
"grad_norm": 0.89453125,
"learning_rate": 6.021711683281783e-06,
"loss": 0.1475035846233368,
"step": 789
},
{
"epoch": 0.4215581643543223,
"grad_norm": 1.1875,
"learning_rate": 6.015677048721664e-06,
"loss": 0.14344163239002228,
"step": 790
},
{
"epoch": 0.4220917822838847,
"grad_norm": 1.0859375,
"learning_rate": 6.009636723755378e-06,
"loss": 0.1469307392835617,
"step": 791
},
{
"epoch": 0.4226254002134472,
"grad_norm": 1.6484375,
"learning_rate": 6.003590729656919e-06,
"loss": 0.14220210909843445,
"step": 792
},
{
"epoch": 0.4231590181430096,
"grad_norm": 1.546875,
"learning_rate": 5.997539087720246e-06,
"loss": 0.12574386596679688,
"step": 793
},
{
"epoch": 0.423692636072572,
"grad_norm": 1.09375,
"learning_rate": 5.991481819259213e-06,
"loss": 0.12123759835958481,
"step": 794
},
{
"epoch": 0.4242262540021345,
"grad_norm": 1.6953125,
"learning_rate": 5.9854189456074835e-06,
"loss": 0.1509871929883957,
"step": 795
},
{
"epoch": 0.4247598719316969,
"grad_norm": 1.2578125,
"learning_rate": 5.97935048811847e-06,
"loss": 0.13558119535446167,
"step": 796
},
{
"epoch": 0.4252934898612593,
"grad_norm": 0.8984375,
"learning_rate": 5.9732764681652475e-06,
"loss": 0.13685007393360138,
"step": 797
},
{
"epoch": 0.4258271077908218,
"grad_norm": 3.5625,
"learning_rate": 5.967196907140483e-06,
"loss": 0.12988555431365967,
"step": 798
},
{
"epoch": 0.4263607257203842,
"grad_norm": 1.6640625,
"learning_rate": 5.961111826456359e-06,
"loss": 0.1256590634584427,
"step": 799
},
{
"epoch": 0.42689434364994666,
"grad_norm": 1.0859375,
"learning_rate": 5.955021247544496e-06,
"loss": 0.13018158078193665,
"step": 800
},
{
"epoch": 0.4274279615795091,
"grad_norm": 0.7421875,
"learning_rate": 5.948925191855885e-06,
"loss": 0.12844260036945343,
"step": 801
},
{
"epoch": 0.4279615795090715,
"grad_norm": 0.85546875,
"learning_rate": 5.9428236808608e-06,
"loss": 0.14438281953334808,
"step": 802
},
{
"epoch": 0.42849519743863396,
"grad_norm": 4.0,
"learning_rate": 5.9367167360487315e-06,
"loss": 0.15138579905033112,
"step": 803
},
{
"epoch": 0.42902881536819637,
"grad_norm": 0.953125,
"learning_rate": 5.930604378928306e-06,
"loss": 0.1499411165714264,
"step": 804
},
{
"epoch": 0.4295624332977588,
"grad_norm": 0.859375,
"learning_rate": 5.924486631027217e-06,
"loss": 0.1303812563419342,
"step": 805
},
{
"epoch": 0.43009605122732125,
"grad_norm": 1.1875,
"learning_rate": 5.918363513892136e-06,
"loss": 0.13025110960006714,
"step": 806
},
{
"epoch": 0.43062966915688367,
"grad_norm": 0.9609375,
"learning_rate": 5.912235049088653e-06,
"loss": 0.12970994412899017,
"step": 807
},
{
"epoch": 0.4311632870864461,
"grad_norm": 0.8984375,
"learning_rate": 5.906101258201188e-06,
"loss": 0.1339489221572876,
"step": 808
},
{
"epoch": 0.43169690501600855,
"grad_norm": 0.82421875,
"learning_rate": 5.899962162832923e-06,
"loss": 0.14164133369922638,
"step": 809
},
{
"epoch": 0.43223052294557096,
"grad_norm": 0.81640625,
"learning_rate": 5.893817784605717e-06,
"loss": 0.130917489528656,
"step": 810
},
{
"epoch": 0.43276414087513343,
"grad_norm": 1.21875,
"learning_rate": 5.887668145160039e-06,
"loss": 0.13007573783397675,
"step": 811
},
{
"epoch": 0.43329775880469584,
"grad_norm": 0.8984375,
"learning_rate": 5.8815132661548884e-06,
"loss": 0.12279300391674042,
"step": 812
},
{
"epoch": 0.43383137673425826,
"grad_norm": 0.9609375,
"learning_rate": 5.8753531692677145e-06,
"loss": 0.130149245262146,
"step": 813
},
{
"epoch": 0.4343649946638207,
"grad_norm": 0.8984375,
"learning_rate": 5.869187876194349e-06,
"loss": 0.14086826145648956,
"step": 814
},
{
"epoch": 0.43489861259338314,
"grad_norm": 0.75,
"learning_rate": 5.863017408648921e-06,
"loss": 0.13157956302165985,
"step": 815
},
{
"epoch": 0.43543223052294555,
"grad_norm": 1.1484375,
"learning_rate": 5.8568417883637845e-06,
"loss": 0.13584381341934204,
"step": 816
},
{
"epoch": 0.435965848452508,
"grad_norm": 2.3125,
"learning_rate": 5.850661037089445e-06,
"loss": 0.1278839111328125,
"step": 817
},
{
"epoch": 0.43649946638207043,
"grad_norm": 2.90625,
"learning_rate": 5.8444751765944726e-06,
"loss": 0.13323771953582764,
"step": 818
},
{
"epoch": 0.43703308431163285,
"grad_norm": 1.78125,
"learning_rate": 5.838284228665437e-06,
"loss": 0.1598339080810547,
"step": 819
},
{
"epoch": 0.4375667022411953,
"grad_norm": 1.234375,
"learning_rate": 5.832088215106824e-06,
"loss": 0.13254910707473755,
"step": 820
},
{
"epoch": 0.43810032017075773,
"grad_norm": 1.0078125,
"learning_rate": 5.825887157740964e-06,
"loss": 0.1257910579442978,
"step": 821
},
{
"epoch": 0.4386339381003202,
"grad_norm": 0.74609375,
"learning_rate": 5.819681078407947e-06,
"loss": 0.1194000318646431,
"step": 822
},
{
"epoch": 0.4391675560298826,
"grad_norm": 1.09375,
"learning_rate": 5.813469998965551e-06,
"loss": 0.13598917424678802,
"step": 823
},
{
"epoch": 0.439701173959445,
"grad_norm": 2.296875,
"learning_rate": 5.807253941289167e-06,
"loss": 0.1345856785774231,
"step": 824
},
{
"epoch": 0.4402347918890075,
"grad_norm": 0.921875,
"learning_rate": 5.801032927271718e-06,
"loss": 0.12210307270288467,
"step": 825
},
{
"epoch": 0.4407684098185699,
"grad_norm": 1.5546875,
"learning_rate": 5.794806978823581e-06,
"loss": 0.1445075273513794,
"step": 826
},
{
"epoch": 0.4413020277481323,
"grad_norm": 1.3125,
"learning_rate": 5.788576117872516e-06,
"loss": 0.12882518768310547,
"step": 827
},
{
"epoch": 0.4418356456776948,
"grad_norm": 1.40625,
"learning_rate": 5.782340366363584e-06,
"loss": 0.1334984451532364,
"step": 828
},
{
"epoch": 0.4423692636072572,
"grad_norm": 3.9375,
"learning_rate": 5.776099746259066e-06,
"loss": 0.15669122338294983,
"step": 829
},
{
"epoch": 0.4429028815368196,
"grad_norm": 2.453125,
"learning_rate": 5.7698542795383965e-06,
"loss": 0.14190684258937836,
"step": 830
},
{
"epoch": 0.4434364994663821,
"grad_norm": 1.515625,
"learning_rate": 5.763603988198073e-06,
"loss": 0.13538184762001038,
"step": 831
},
{
"epoch": 0.4439701173959445,
"grad_norm": 0.98828125,
"learning_rate": 5.757348894251595e-06,
"loss": 0.1276673972606659,
"step": 832
},
{
"epoch": 0.4445037353255069,
"grad_norm": 0.83984375,
"learning_rate": 5.751089019729366e-06,
"loss": 0.12626248598098755,
"step": 833
},
{
"epoch": 0.4450373532550694,
"grad_norm": 2.890625,
"learning_rate": 5.744824386678632e-06,
"loss": 0.1427343636751175,
"step": 834
},
{
"epoch": 0.4455709711846318,
"grad_norm": 1.1640625,
"learning_rate": 5.738555017163401e-06,
"loss": 0.14048369228839874,
"step": 835
},
{
"epoch": 0.44610458911419426,
"grad_norm": 2.375,
"learning_rate": 5.732280933264358e-06,
"loss": 0.12917697429656982,
"step": 836
},
{
"epoch": 0.4466382070437567,
"grad_norm": 3.28125,
"learning_rate": 5.7260021570787935e-06,
"loss": 0.12306460738182068,
"step": 837
},
{
"epoch": 0.4471718249733191,
"grad_norm": 1.0078125,
"learning_rate": 5.719718710720526e-06,
"loss": 0.13587269186973572,
"step": 838
},
{
"epoch": 0.44770544290288156,
"grad_norm": 0.8984375,
"learning_rate": 5.713430616319819e-06,
"loss": 0.13805776834487915,
"step": 839
},
{
"epoch": 0.44823906083244397,
"grad_norm": 1.125,
"learning_rate": 5.707137896023308e-06,
"loss": 0.14297237992286682,
"step": 840
},
{
"epoch": 0.4487726787620064,
"grad_norm": 1.703125,
"learning_rate": 5.700840571993923e-06,
"loss": 0.1339704692363739,
"step": 841
},
{
"epoch": 0.44930629669156885,
"grad_norm": 2.078125,
"learning_rate": 5.694538666410807e-06,
"loss": 0.15773418545722961,
"step": 842
},
{
"epoch": 0.44983991462113126,
"grad_norm": 2.28125,
"learning_rate": 5.688232201469237e-06,
"loss": 0.13986408710479736,
"step": 843
},
{
"epoch": 0.4503735325506937,
"grad_norm": 1.4296875,
"learning_rate": 5.681921199380551e-06,
"loss": 0.12986308336257935,
"step": 844
},
{
"epoch": 0.45090715048025615,
"grad_norm": 4.59375,
"learning_rate": 5.675605682372064e-06,
"loss": 0.1184689998626709,
"step": 845
},
{
"epoch": 0.45144076840981856,
"grad_norm": 0.6875,
"learning_rate": 5.669285672686994e-06,
"loss": 0.12625913321971893,
"step": 846
},
{
"epoch": 0.45197438633938103,
"grad_norm": 1.59375,
"learning_rate": 5.662961192584384e-06,
"loss": 0.13546067476272583,
"step": 847
},
{
"epoch": 0.45250800426894344,
"grad_norm": 1.1953125,
"learning_rate": 5.65663226433902e-06,
"loss": 0.14226895570755005,
"step": 848
},
{
"epoch": 0.45304162219850586,
"grad_norm": 1.46875,
"learning_rate": 5.650298910241352e-06,
"loss": 0.12281224131584167,
"step": 849
},
{
"epoch": 0.4535752401280683,
"grad_norm": 4.25,
"learning_rate": 5.643961152597424e-06,
"loss": 0.1291586458683014,
"step": 850
},
{
"epoch": 0.45410885805763074,
"grad_norm": 1.3671875,
"learning_rate": 5.637619013728782e-06,
"loss": 0.1480562686920166,
"step": 851
},
{
"epoch": 0.45464247598719315,
"grad_norm": 1.21875,
"learning_rate": 5.6312725159724075e-06,
"loss": 0.1352003812789917,
"step": 852
},
{
"epoch": 0.4551760939167556,
"grad_norm": 2.28125,
"learning_rate": 5.624921681680633e-06,
"loss": 0.13184835016727448,
"step": 853
},
{
"epoch": 0.45570971184631803,
"grad_norm": 0.7890625,
"learning_rate": 5.618566533221062e-06,
"loss": 0.13229535520076752,
"step": 854
},
{
"epoch": 0.45624332977588045,
"grad_norm": 4.3125,
"learning_rate": 5.612207092976497e-06,
"loss": 0.14021338522434235,
"step": 855
},
{
"epoch": 0.4567769477054429,
"grad_norm": 1.0625,
"learning_rate": 5.605843383344852e-06,
"loss": 0.14466454088687897,
"step": 856
},
{
"epoch": 0.45731056563500533,
"grad_norm": 0.89453125,
"learning_rate": 5.5994754267390785e-06,
"loss": 0.14731889963150024,
"step": 857
},
{
"epoch": 0.4578441835645678,
"grad_norm": 0.953125,
"learning_rate": 5.593103245587088e-06,
"loss": 0.1401660293340683,
"step": 858
},
{
"epoch": 0.4583778014941302,
"grad_norm": 3.9375,
"learning_rate": 5.58672686233167e-06,
"loss": 0.13546955585479736,
"step": 859
},
{
"epoch": 0.4589114194236926,
"grad_norm": 1.125,
"learning_rate": 5.580346299430411e-06,
"loss": 0.12634322047233582,
"step": 860
},
{
"epoch": 0.4594450373532551,
"grad_norm": 0.99609375,
"learning_rate": 5.573961579355622e-06,
"loss": 0.12239423394203186,
"step": 861
},
{
"epoch": 0.4599786552828175,
"grad_norm": 0.7890625,
"learning_rate": 5.5675727245942525e-06,
"loss": 0.12961003184318542,
"step": 862
},
{
"epoch": 0.4605122732123799,
"grad_norm": 4.3125,
"learning_rate": 5.561179757647818e-06,
"loss": 0.1358482539653778,
"step": 863
},
{
"epoch": 0.4610458911419424,
"grad_norm": 1.1015625,
"learning_rate": 5.554782701032312e-06,
"loss": 0.13748842477798462,
"step": 864
},
{
"epoch": 0.4615795090715048,
"grad_norm": 1.140625,
"learning_rate": 5.548381577278137e-06,
"loss": 0.12954974174499512,
"step": 865
},
{
"epoch": 0.4621131270010672,
"grad_norm": 1.0703125,
"learning_rate": 5.5419764089300155e-06,
"loss": 0.13842041790485382,
"step": 866
},
{
"epoch": 0.4626467449306297,
"grad_norm": 0.97265625,
"learning_rate": 5.53556721854692e-06,
"loss": 0.13455192744731903,
"step": 867
},
{
"epoch": 0.4631803628601921,
"grad_norm": 0.6796875,
"learning_rate": 5.529154028701983e-06,
"loss": 0.11836317181587219,
"step": 868
},
{
"epoch": 0.4637139807897545,
"grad_norm": 0.92578125,
"learning_rate": 5.522736861982427e-06,
"loss": 0.12298063933849335,
"step": 869
},
{
"epoch": 0.464247598719317,
"grad_norm": 0.89453125,
"learning_rate": 5.516315740989479e-06,
"loss": 0.14253345131874084,
"step": 870
},
{
"epoch": 0.4647812166488794,
"grad_norm": 1.9296875,
"learning_rate": 5.509890688338295e-06,
"loss": 0.13991788029670715,
"step": 871
},
{
"epoch": 0.46531483457844186,
"grad_norm": 0.84765625,
"learning_rate": 5.503461726657876e-06,
"loss": 0.1314542442560196,
"step": 872
},
{
"epoch": 0.4658484525080043,
"grad_norm": 0.92578125,
"learning_rate": 5.497028878590993e-06,
"loss": 0.12288765609264374,
"step": 873
},
{
"epoch": 0.4663820704375667,
"grad_norm": 7.9375,
"learning_rate": 5.490592166794103e-06,
"loss": 0.12817099690437317,
"step": 874
},
{
"epoch": 0.46691568836712916,
"grad_norm": 1.5234375,
"learning_rate": 5.48415161393727e-06,
"loss": 0.15515467524528503,
"step": 875
},
{
"epoch": 0.46744930629669157,
"grad_norm": 1.1953125,
"learning_rate": 5.477707242704089e-06,
"loss": 0.12695574760437012,
"step": 876
},
{
"epoch": 0.467982924226254,
"grad_norm": 2.640625,
"learning_rate": 5.471259075791602e-06,
"loss": 0.13264472782611847,
"step": 877
},
{
"epoch": 0.46851654215581645,
"grad_norm": 1.5,
"learning_rate": 5.464807135910217e-06,
"loss": 0.1282857060432434,
"step": 878
},
{
"epoch": 0.46905016008537886,
"grad_norm": 1.4375,
"learning_rate": 5.458351445783637e-06,
"loss": 0.14127972722053528,
"step": 879
},
{
"epoch": 0.4695837780149413,
"grad_norm": 0.96875,
"learning_rate": 5.451892028148766e-06,
"loss": 0.12471657991409302,
"step": 880
},
{
"epoch": 0.47011739594450375,
"grad_norm": 0.80859375,
"learning_rate": 5.4454289057556395e-06,
"loss": 0.1276896893978119,
"step": 881
},
{
"epoch": 0.47065101387406616,
"grad_norm": 1.4453125,
"learning_rate": 5.438962101367342e-06,
"loss": 0.13574856519699097,
"step": 882
},
{
"epoch": 0.47118463180362863,
"grad_norm": 2.21875,
"learning_rate": 5.432491637759925e-06,
"loss": 0.13857027888298035,
"step": 883
},
{
"epoch": 0.47171824973319104,
"grad_norm": 0.8515625,
"learning_rate": 5.426017537722327e-06,
"loss": 0.12701499462127686,
"step": 884
},
{
"epoch": 0.47225186766275346,
"grad_norm": 1.1796875,
"learning_rate": 5.419539824056294e-06,
"loss": 0.14010781049728394,
"step": 885
},
{
"epoch": 0.4727854855923159,
"grad_norm": 1.1171875,
"learning_rate": 5.413058519576302e-06,
"loss": 0.1314486563205719,
"step": 886
},
{
"epoch": 0.47331910352187834,
"grad_norm": 1.2109375,
"learning_rate": 5.40657364710947e-06,
"loss": 0.15350569784641266,
"step": 887
},
{
"epoch": 0.47385272145144075,
"grad_norm": 4.0,
"learning_rate": 5.400085229495486e-06,
"loss": 0.13060224056243896,
"step": 888
},
{
"epoch": 0.4743863393810032,
"grad_norm": 1.203125,
"learning_rate": 5.393593289586522e-06,
"loss": 0.13233044743537903,
"step": 889
},
{
"epoch": 0.47491995731056563,
"grad_norm": 1.03125,
"learning_rate": 5.3870978502471575e-06,
"loss": 0.14206142723560333,
"step": 890
},
{
"epoch": 0.47545357524012805,
"grad_norm": 0.953125,
"learning_rate": 5.380598934354296e-06,
"loss": 0.1302148848772049,
"step": 891
},
{
"epoch": 0.4759871931696905,
"grad_norm": 0.91796875,
"learning_rate": 5.374096564797085e-06,
"loss": 0.13493700325489044,
"step": 892
},
{
"epoch": 0.47652081109925293,
"grad_norm": 1.046875,
"learning_rate": 5.367590764476837e-06,
"loss": 0.13939698040485382,
"step": 893
},
{
"epoch": 0.47705442902881534,
"grad_norm": 1.7421875,
"learning_rate": 5.361081556306946e-06,
"loss": 0.1327444314956665,
"step": 894
},
{
"epoch": 0.4775880469583778,
"grad_norm": 4.5625,
"learning_rate": 5.35456896321281e-06,
"loss": 0.13143512606620789,
"step": 895
},
{
"epoch": 0.4781216648879402,
"grad_norm": 0.9921875,
"learning_rate": 5.348053008131748e-06,
"loss": 0.1298195868730545,
"step": 896
},
{
"epoch": 0.4786552828175027,
"grad_norm": 0.9609375,
"learning_rate": 5.34153371401292e-06,
"loss": 0.13759765028953552,
"step": 897
},
{
"epoch": 0.4791889007470651,
"grad_norm": 2.4375,
"learning_rate": 5.335011103817245e-06,
"loss": 0.14482876658439636,
"step": 898
},
{
"epoch": 0.4797225186766275,
"grad_norm": 0.8828125,
"learning_rate": 5.328485200517325e-06,
"loss": 0.13685324788093567,
"step": 899
},
{
"epoch": 0.48025613660619,
"grad_norm": 0.95703125,
"learning_rate": 5.3219560270973545e-06,
"loss": 0.13253626227378845,
"step": 900
},
{
"epoch": 0.4807897545357524,
"grad_norm": 0.984375,
"learning_rate": 5.315423606553048e-06,
"loss": 0.1298791766166687,
"step": 901
},
{
"epoch": 0.4813233724653148,
"grad_norm": 1.203125,
"learning_rate": 5.308887961891561e-06,
"loss": 0.1377299427986145,
"step": 902
},
{
"epoch": 0.4818569903948773,
"grad_norm": 0.9296875,
"learning_rate": 5.302349116131393e-06,
"loss": 0.14348457753658295,
"step": 903
},
{
"epoch": 0.4823906083244397,
"grad_norm": 2.21875,
"learning_rate": 5.2958070923023315e-06,
"loss": 0.1252109259366989,
"step": 904
},
{
"epoch": 0.4829242262540021,
"grad_norm": 1.109375,
"learning_rate": 5.289261913445348e-06,
"loss": 0.139956995844841,
"step": 905
},
{
"epoch": 0.4834578441835646,
"grad_norm": 1.1015625,
"learning_rate": 5.282713602612528e-06,
"loss": 0.13205938041210175,
"step": 906
},
{
"epoch": 0.483991462113127,
"grad_norm": 1.8046875,
"learning_rate": 5.276162182866986e-06,
"loss": 0.1294470727443695,
"step": 907
},
{
"epoch": 0.48452508004268946,
"grad_norm": 0.60546875,
"learning_rate": 5.269607677282793e-06,
"loss": 0.13724714517593384,
"step": 908
},
{
"epoch": 0.4850586979722519,
"grad_norm": 1.125,
"learning_rate": 5.263050108944882e-06,
"loss": 0.1366155743598938,
"step": 909
},
{
"epoch": 0.4855923159018143,
"grad_norm": 2.671875,
"learning_rate": 5.256489500948972e-06,
"loss": 0.13106200098991394,
"step": 910
},
{
"epoch": 0.48612593383137676,
"grad_norm": 1.0859375,
"learning_rate": 5.2499258764014925e-06,
"loss": 0.156961590051651,
"step": 911
},
{
"epoch": 0.48665955176093917,
"grad_norm": 1.671875,
"learning_rate": 5.2433592584194955e-06,
"loss": 0.14239418506622314,
"step": 912
},
{
"epoch": 0.4871931696905016,
"grad_norm": 1.0546875,
"learning_rate": 5.236789670130575e-06,
"loss": 0.128268301486969,
"step": 913
},
{
"epoch": 0.48772678762006405,
"grad_norm": 0.875,
"learning_rate": 5.230217134672785e-06,
"loss": 0.13807916641235352,
"step": 914
},
{
"epoch": 0.48826040554962646,
"grad_norm": 1.2890625,
"learning_rate": 5.2236416751945626e-06,
"loss": 0.13521814346313477,
"step": 915
},
{
"epoch": 0.4887940234791889,
"grad_norm": 1.8515625,
"learning_rate": 5.217063314854642e-06,
"loss": 0.1347956359386444,
"step": 916
},
{
"epoch": 0.48932764140875135,
"grad_norm": 1.1875,
"learning_rate": 5.210482076821974e-06,
"loss": 0.15106873214244843,
"step": 917
},
{
"epoch": 0.48986125933831376,
"grad_norm": 1.3203125,
"learning_rate": 5.2038979842756454e-06,
"loss": 0.13903489708900452,
"step": 918
},
{
"epoch": 0.49039487726787623,
"grad_norm": 0.98046875,
"learning_rate": 5.197311060404795e-06,
"loss": 0.12791642546653748,
"step": 919
},
{
"epoch": 0.49092849519743864,
"grad_norm": 2.234375,
"learning_rate": 5.1907213284085346e-06,
"loss": 0.13822868466377258,
"step": 920
},
{
"epoch": 0.49146211312700105,
"grad_norm": 1.953125,
"learning_rate": 5.1841288114958654e-06,
"loss": 0.14124801754951477,
"step": 921
},
{
"epoch": 0.4919957310565635,
"grad_norm": 1.515625,
"learning_rate": 5.177533532885599e-06,
"loss": 0.13764949142932892,
"step": 922
},
{
"epoch": 0.49252934898612594,
"grad_norm": 1.4375,
"learning_rate": 5.170935515806272e-06,
"loss": 0.13255469501018524,
"step": 923
},
{
"epoch": 0.49306296691568835,
"grad_norm": 1.71875,
"learning_rate": 5.164334783496065e-06,
"loss": 0.13303151726722717,
"step": 924
},
{
"epoch": 0.4935965848452508,
"grad_norm": 1.65625,
"learning_rate": 5.157731359202725e-06,
"loss": 0.12191550433635712,
"step": 925
},
{
"epoch": 0.49413020277481323,
"grad_norm": 1.625,
"learning_rate": 5.151125266183477e-06,
"loss": 0.12975803017616272,
"step": 926
},
{
"epoch": 0.49466382070437565,
"grad_norm": 1.40625,
"learning_rate": 5.1445165277049465e-06,
"loss": 0.12646839022636414,
"step": 927
},
{
"epoch": 0.4951974386339381,
"grad_norm": 0.98828125,
"learning_rate": 5.137905167043074e-06,
"loss": 0.1241583377122879,
"step": 928
},
{
"epoch": 0.4957310565635005,
"grad_norm": 1.03125,
"learning_rate": 5.131291207483041e-06,
"loss": 0.12851932644844055,
"step": 929
},
{
"epoch": 0.49626467449306294,
"grad_norm": 1.1875,
"learning_rate": 5.124674672319175e-06,
"loss": 0.12480233609676361,
"step": 930
},
{
"epoch": 0.4967982924226254,
"grad_norm": 1.015625,
"learning_rate": 5.118055584854882e-06,
"loss": 0.12151680886745453,
"step": 931
},
{
"epoch": 0.4973319103521878,
"grad_norm": 0.75390625,
"learning_rate": 5.111433968402553e-06,
"loss": 0.13377901911735535,
"step": 932
},
{
"epoch": 0.4978655282817503,
"grad_norm": 1.7265625,
"learning_rate": 5.104809846283483e-06,
"loss": 0.13202190399169922,
"step": 933
},
{
"epoch": 0.4983991462113127,
"grad_norm": 0.8359375,
"learning_rate": 5.0981832418278e-06,
"loss": 0.13428083062171936,
"step": 934
},
{
"epoch": 0.4989327641408751,
"grad_norm": 1.140625,
"learning_rate": 5.091554178374368e-06,
"loss": 0.13188795745372772,
"step": 935
},
{
"epoch": 0.4994663820704376,
"grad_norm": 2.296875,
"learning_rate": 5.084922679270716e-06,
"loss": 0.14230987429618835,
"step": 936
},
{
"epoch": 0.5,
"grad_norm": 0.8046875,
"learning_rate": 5.078288767872947e-06,
"loss": 0.12688730657100677,
"step": 937
},
{
"epoch": 0.5,
"eval_loss": 0.13445857167243958,
"eval_runtime": 11.2191,
"eval_samples_per_second": 5.348,
"eval_steps_per_second": 0.178,
"step": 937
},
{
"epoch": 0.5005336179295624,
"grad_norm": 0.92578125,
"learning_rate": 5.071652467545663e-06,
"loss": 0.1254737377166748,
"step": 938
},
{
"epoch": 0.5010672358591248,
"grad_norm": 0.90625,
"learning_rate": 5.065013801661882e-06,
"loss": 0.13981440663337708,
"step": 939
},
{
"epoch": 0.5016008537886874,
"grad_norm": 1.578125,
"learning_rate": 5.058372793602947e-06,
"loss": 0.12979918718338013,
"step": 940
},
{
"epoch": 0.5021344717182498,
"grad_norm": 1.78125,
"learning_rate": 5.051729466758456e-06,
"loss": 0.14440438151359558,
"step": 941
},
{
"epoch": 0.5026680896478122,
"grad_norm": 0.84765625,
"learning_rate": 5.0450838445261706e-06,
"loss": 0.12817810475826263,
"step": 942
},
{
"epoch": 0.5032017075773746,
"grad_norm": 0.86328125,
"learning_rate": 5.038435950311939e-06,
"loss": 0.13537698984146118,
"step": 943
},
{
"epoch": 0.503735325506937,
"grad_norm": 0.96875,
"learning_rate": 5.031785807529609e-06,
"loss": 0.13686254620552063,
"step": 944
},
{
"epoch": 0.5042689434364994,
"grad_norm": 0.953125,
"learning_rate": 5.02513343960095e-06,
"loss": 0.13954181969165802,
"step": 945
},
{
"epoch": 0.5048025613660619,
"grad_norm": 0.9921875,
"learning_rate": 5.018478869955565e-06,
"loss": 0.12591418623924255,
"step": 946
},
{
"epoch": 0.5053361792956244,
"grad_norm": 2.65625,
"learning_rate": 5.0118221220308155e-06,
"loss": 0.13516512513160706,
"step": 947
},
{
"epoch": 0.5058697972251868,
"grad_norm": 0.859375,
"learning_rate": 5.005163219271734e-06,
"loss": 0.13018985092639923,
"step": 948
},
{
"epoch": 0.5064034151547492,
"grad_norm": 1.2265625,
"learning_rate": 4.998502185130939e-06,
"loss": 0.13040566444396973,
"step": 949
},
{
"epoch": 0.5069370330843116,
"grad_norm": 0.80078125,
"learning_rate": 4.991839043068559e-06,
"loss": 0.1382044106721878,
"step": 950
},
{
"epoch": 0.5074706510138741,
"grad_norm": 0.96484375,
"learning_rate": 4.985173816552147e-06,
"loss": 0.14346781373023987,
"step": 951
},
{
"epoch": 0.5080042689434365,
"grad_norm": 0.953125,
"learning_rate": 4.9785065290565955e-06,
"loss": 0.14343935251235962,
"step": 952
},
{
"epoch": 0.508537886872999,
"grad_norm": 0.97265625,
"learning_rate": 4.971837204064054e-06,
"loss": 0.12788984179496765,
"step": 953
},
{
"epoch": 0.5090715048025614,
"grad_norm": 0.80078125,
"learning_rate": 4.965165865063853e-06,
"loss": 0.12848851084709167,
"step": 954
},
{
"epoch": 0.5096051227321238,
"grad_norm": 1.265625,
"learning_rate": 4.958492535552415e-06,
"loss": 0.12936604022979736,
"step": 955
},
{
"epoch": 0.5101387406616862,
"grad_norm": 2.25,
"learning_rate": 4.951817239033169e-06,
"loss": 0.13092266023159027,
"step": 956
},
{
"epoch": 0.5106723585912487,
"grad_norm": 3.46875,
"learning_rate": 4.945139999016476e-06,
"loss": 0.1326400339603424,
"step": 957
},
{
"epoch": 0.5112059765208111,
"grad_norm": 2.875,
"learning_rate": 4.9384608390195395e-06,
"loss": 0.12369441986083984,
"step": 958
},
{
"epoch": 0.5117395944503735,
"grad_norm": 7.625,
"learning_rate": 4.931779782566328e-06,
"loss": 0.13840287923812866,
"step": 959
},
{
"epoch": 0.512273212379936,
"grad_norm": 2.296875,
"learning_rate": 4.925096853187487e-06,
"loss": 0.1434604823589325,
"step": 960
},
{
"epoch": 0.5128068303094984,
"grad_norm": 1.3203125,
"learning_rate": 4.918412074420259e-06,
"loss": 0.1397111862897873,
"step": 961
},
{
"epoch": 0.5133404482390609,
"grad_norm": 1.296875,
"learning_rate": 4.9117254698084e-06,
"loss": 0.1331661492586136,
"step": 962
},
{
"epoch": 0.5138740661686233,
"grad_norm": 0.984375,
"learning_rate": 4.905037062902095e-06,
"loss": 0.1459713578224182,
"step": 963
},
{
"epoch": 0.5144076840981857,
"grad_norm": 1.1484375,
"learning_rate": 4.898346877257881e-06,
"loss": 0.13905799388885498,
"step": 964
},
{
"epoch": 0.5149413020277481,
"grad_norm": 1.2265625,
"learning_rate": 4.891654936438555e-06,
"loss": 0.12892545759677887,
"step": 965
},
{
"epoch": 0.5154749199573105,
"grad_norm": 2.78125,
"learning_rate": 4.8849612640131e-06,
"loss": 0.1311328113079071,
"step": 966
},
{
"epoch": 0.516008537886873,
"grad_norm": 2.1875,
"learning_rate": 4.8782658835565925e-06,
"loss": 0.13056829571723938,
"step": 967
},
{
"epoch": 0.5165421558164355,
"grad_norm": 1.6328125,
"learning_rate": 4.871568818650131e-06,
"loss": 0.12647056579589844,
"step": 968
},
{
"epoch": 0.5170757737459979,
"grad_norm": 1.0390625,
"learning_rate": 4.864870092880741e-06,
"loss": 0.13085322082042694,
"step": 969
},
{
"epoch": 0.5176093916755603,
"grad_norm": 1.203125,
"learning_rate": 4.8581697298413e-06,
"loss": 0.14101335406303406,
"step": 970
},
{
"epoch": 0.5181430096051227,
"grad_norm": 1.03125,
"learning_rate": 4.8514677531304535e-06,
"loss": 0.13168872892856598,
"step": 971
},
{
"epoch": 0.5186766275346851,
"grad_norm": 0.88671875,
"learning_rate": 4.8447641863525276e-06,
"loss": 0.12382831424474716,
"step": 972
},
{
"epoch": 0.5192102454642475,
"grad_norm": 0.765625,
"learning_rate": 4.8380590531174495e-06,
"loss": 0.1534726619720459,
"step": 973
},
{
"epoch": 0.5197438633938101,
"grad_norm": 1.03125,
"learning_rate": 4.831352377040661e-06,
"loss": 0.1461876481771469,
"step": 974
},
{
"epoch": 0.5202774813233725,
"grad_norm": 1.234375,
"learning_rate": 4.8246441817430465e-06,
"loss": 0.14096978306770325,
"step": 975
},
{
"epoch": 0.5208110992529349,
"grad_norm": 0.8671875,
"learning_rate": 4.817934490850828e-06,
"loss": 0.11423539370298386,
"step": 976
},
{
"epoch": 0.5213447171824973,
"grad_norm": 0.8046875,
"learning_rate": 4.811223327995508e-06,
"loss": 0.12851789593696594,
"step": 977
},
{
"epoch": 0.5218783351120597,
"grad_norm": 3.234375,
"learning_rate": 4.804510716813763e-06,
"loss": 0.1357826292514801,
"step": 978
},
{
"epoch": 0.5224119530416222,
"grad_norm": 0.984375,
"learning_rate": 4.797796680947377e-06,
"loss": 0.15453936159610748,
"step": 979
},
{
"epoch": 0.5229455709711847,
"grad_norm": 0.8125,
"learning_rate": 4.7910812440431495e-06,
"loss": 0.12888383865356445,
"step": 980
},
{
"epoch": 0.5234791889007471,
"grad_norm": 1.5390625,
"learning_rate": 4.7843644297528115e-06,
"loss": 0.13885925710201263,
"step": 981
},
{
"epoch": 0.5240128068303095,
"grad_norm": 2.890625,
"learning_rate": 4.777646261732952e-06,
"loss": 0.13108965754508972,
"step": 982
},
{
"epoch": 0.5245464247598719,
"grad_norm": 0.8125,
"learning_rate": 4.77092676364492e-06,
"loss": 0.12988567352294922,
"step": 983
},
{
"epoch": 0.5250800426894343,
"grad_norm": 13.875,
"learning_rate": 4.764205959154756e-06,
"loss": 0.14293351769447327,
"step": 984
},
{
"epoch": 0.5256136606189968,
"grad_norm": 1.0859375,
"learning_rate": 4.757483871933097e-06,
"loss": 0.12875567376613617,
"step": 985
},
{
"epoch": 0.5261472785485592,
"grad_norm": 2.515625,
"learning_rate": 4.7507605256551e-06,
"loss": 0.14340457320213318,
"step": 986
},
{
"epoch": 0.5266808964781217,
"grad_norm": 0.91015625,
"learning_rate": 4.744035944000354e-06,
"loss": 0.13566985726356506,
"step": 987
},
{
"epoch": 0.5272145144076841,
"grad_norm": 3.375,
"learning_rate": 4.737310150652801e-06,
"loss": 0.1393333077430725,
"step": 988
},
{
"epoch": 0.5277481323372465,
"grad_norm": 1.140625,
"learning_rate": 4.730583169300653e-06,
"loss": 0.12942199409008026,
"step": 989
},
{
"epoch": 0.528281750266809,
"grad_norm": 1.0234375,
"learning_rate": 4.723855023636299e-06,
"loss": 0.1364246904850006,
"step": 990
},
{
"epoch": 0.5288153681963714,
"grad_norm": 2.578125,
"learning_rate": 4.717125737356237e-06,
"loss": 0.13817454874515533,
"step": 991
},
{
"epoch": 0.5293489861259338,
"grad_norm": 1.4453125,
"learning_rate": 4.710395334160976e-06,
"loss": 0.1424347460269928,
"step": 992
},
{
"epoch": 0.5298826040554963,
"grad_norm": 0.96484375,
"learning_rate": 4.703663837754961e-06,
"loss": 0.13900761306285858,
"step": 993
},
{
"epoch": 0.5304162219850587,
"grad_norm": 2.984375,
"learning_rate": 4.6969312718464895e-06,
"loss": 0.13058799505233765,
"step": 994
},
{
"epoch": 0.5309498399146211,
"grad_norm": 0.88671875,
"learning_rate": 4.690197660147622e-06,
"loss": 0.14014580845832825,
"step": 995
},
{
"epoch": 0.5314834578441836,
"grad_norm": 1.765625,
"learning_rate": 4.683463026374103e-06,
"loss": 0.12197675555944443,
"step": 996
},
{
"epoch": 0.532017075773746,
"grad_norm": 2.109375,
"learning_rate": 4.676727394245279e-06,
"loss": 0.13193392753601074,
"step": 997
},
{
"epoch": 0.5325506937033084,
"grad_norm": 0.79296875,
"learning_rate": 4.669990787484012e-06,
"loss": 0.1408708542585373,
"step": 998
},
{
"epoch": 0.5330843116328708,
"grad_norm": 1.078125,
"learning_rate": 4.663253229816594e-06,
"loss": 0.1418258249759674,
"step": 999
},
{
"epoch": 0.5336179295624333,
"grad_norm": 5.6875,
"learning_rate": 4.656514744972669e-06,
"loss": 0.12787938117980957,
"step": 1000
},
{
"epoch": 0.5341515474919958,
"grad_norm": 1.203125,
"learning_rate": 4.649775356685146e-06,
"loss": 0.1461375653743744,
"step": 1001
},
{
"epoch": 0.5346851654215582,
"grad_norm": 2.015625,
"learning_rate": 4.643035088690115e-06,
"loss": 0.13402216136455536,
"step": 1002
},
{
"epoch": 0.5352187833511206,
"grad_norm": 1.390625,
"learning_rate": 4.636293964726766e-06,
"loss": 0.12698495388031006,
"step": 1003
},
{
"epoch": 0.535752401280683,
"grad_norm": 1.1015625,
"learning_rate": 4.6295520085373e-06,
"loss": 0.1333812028169632,
"step": 1004
},
{
"epoch": 0.5362860192102454,
"grad_norm": 1.03125,
"learning_rate": 4.622809243866855e-06,
"loss": 0.13612286746501923,
"step": 1005
},
{
"epoch": 0.5368196371398078,
"grad_norm": 2.0,
"learning_rate": 4.61606569446341e-06,
"loss": 0.1351289451122284,
"step": 1006
},
{
"epoch": 0.5373532550693704,
"grad_norm": 0.8359375,
"learning_rate": 4.609321384077713e-06,
"loss": 0.12808330357074738,
"step": 1007
},
{
"epoch": 0.5378868729989328,
"grad_norm": 4.34375,
"learning_rate": 4.60257633646319e-06,
"loss": 0.12295901030302048,
"step": 1008
},
{
"epoch": 0.5384204909284952,
"grad_norm": 1.1171875,
"learning_rate": 4.595830575375861e-06,
"loss": 0.141440749168396,
"step": 1009
},
{
"epoch": 0.5389541088580576,
"grad_norm": 1.21875,
"learning_rate": 4.589084124574266e-06,
"loss": 0.1321246176958084,
"step": 1010
},
{
"epoch": 0.53948772678762,
"grad_norm": 1.09375,
"learning_rate": 4.5823370078193666e-06,
"loss": 0.13147391378879547,
"step": 1011
},
{
"epoch": 0.5400213447171825,
"grad_norm": 1.015625,
"learning_rate": 4.575589248874472e-06,
"loss": 0.15272164344787598,
"step": 1012
},
{
"epoch": 0.540554962646745,
"grad_norm": 0.84765625,
"learning_rate": 4.568840871505158e-06,
"loss": 0.13812553882598877,
"step": 1013
},
{
"epoch": 0.5410885805763074,
"grad_norm": 0.828125,
"learning_rate": 4.56209189947917e-06,
"loss": 0.14311441779136658,
"step": 1014
},
{
"epoch": 0.5416221985058698,
"grad_norm": 6.375,
"learning_rate": 4.555342356566355e-06,
"loss": 0.11868622899055481,
"step": 1015
},
{
"epoch": 0.5421558164354322,
"grad_norm": 1.2109375,
"learning_rate": 4.548592266538569e-06,
"loss": 0.13087981939315796,
"step": 1016
},
{
"epoch": 0.5426894343649946,
"grad_norm": 1.0078125,
"learning_rate": 4.541841653169592e-06,
"loss": 0.1373177170753479,
"step": 1017
},
{
"epoch": 0.5432230522945571,
"grad_norm": 1.03125,
"learning_rate": 4.53509054023505e-06,
"loss": 0.14374572038650513,
"step": 1018
},
{
"epoch": 0.5437566702241196,
"grad_norm": 0.75,
"learning_rate": 4.528338951512328e-06,
"loss": 0.1430709958076477,
"step": 1019
},
{
"epoch": 0.544290288153682,
"grad_norm": 5.53125,
"learning_rate": 4.521586910780486e-06,
"loss": 0.14049461483955383,
"step": 1020
},
{
"epoch": 0.5448239060832444,
"grad_norm": 2.328125,
"learning_rate": 4.5148344418201755e-06,
"loss": 0.13056530058383942,
"step": 1021
},
{
"epoch": 0.5453575240128068,
"grad_norm": 2.046875,
"learning_rate": 4.508081568413557e-06,
"loss": 0.12585791945457458,
"step": 1022
},
{
"epoch": 0.5458911419423693,
"grad_norm": 1.9609375,
"learning_rate": 4.501328314344218e-06,
"loss": 0.13284999132156372,
"step": 1023
},
{
"epoch": 0.5464247598719317,
"grad_norm": 1.3359375,
"learning_rate": 4.494574703397083e-06,
"loss": 0.12665511667728424,
"step": 1024
},
{
"epoch": 0.5469583778014941,
"grad_norm": 1.1796875,
"learning_rate": 4.487820759358333e-06,
"loss": 0.13225844502449036,
"step": 1025
},
{
"epoch": 0.5474919957310566,
"grad_norm": 1.1484375,
"learning_rate": 4.481066506015326e-06,
"loss": 0.1345164030790329,
"step": 1026
},
{
"epoch": 0.548025613660619,
"grad_norm": 1.515625,
"learning_rate": 4.474311967156506e-06,
"loss": 0.13486480712890625,
"step": 1027
},
{
"epoch": 0.5485592315901814,
"grad_norm": 1.5546875,
"learning_rate": 4.467557166571324e-06,
"loss": 0.12519966065883636,
"step": 1028
},
{
"epoch": 0.5490928495197439,
"grad_norm": 1.140625,
"learning_rate": 4.460802128050152e-06,
"loss": 0.1261989176273346,
"step": 1029
},
{
"epoch": 0.5496264674493063,
"grad_norm": 1.0,
"learning_rate": 4.454046875384202e-06,
"loss": 0.13525183498859406,
"step": 1030
},
{
"epoch": 0.5501600853788687,
"grad_norm": 1.921875,
"learning_rate": 4.447291432365436e-06,
"loss": 0.1374928206205368,
"step": 1031
},
{
"epoch": 0.5506937033084311,
"grad_norm": 1.0546875,
"learning_rate": 4.440535822786493e-06,
"loss": 0.12927362322807312,
"step": 1032
},
{
"epoch": 0.5512273212379936,
"grad_norm": 0.7109375,
"learning_rate": 4.433780070440593e-06,
"loss": 0.13270899653434753,
"step": 1033
},
{
"epoch": 0.551760939167556,
"grad_norm": 0.66796875,
"learning_rate": 4.42702419912146e-06,
"loss": 0.1262863129377365,
"step": 1034
},
{
"epoch": 0.5522945570971185,
"grad_norm": 0.890625,
"learning_rate": 4.420268232623238e-06,
"loss": 0.13539151847362518,
"step": 1035
},
{
"epoch": 0.5528281750266809,
"grad_norm": 1.421875,
"learning_rate": 4.413512194740405e-06,
"loss": 0.12227917462587357,
"step": 1036
},
{
"epoch": 0.5533617929562433,
"grad_norm": 0.859375,
"learning_rate": 4.4067561092676936e-06,
"loss": 0.13151752948760986,
"step": 1037
},
{
"epoch": 0.5538954108858057,
"grad_norm": 0.984375,
"learning_rate": 4.4e-06,
"loss": 0.1390916109085083,
"step": 1038
},
{
"epoch": 0.5544290288153682,
"grad_norm": 1.25,
"learning_rate": 4.393243890732306e-06,
"loss": 0.12831014394760132,
"step": 1039
},
{
"epoch": 0.5549626467449307,
"grad_norm": 1.59375,
"learning_rate": 4.386487805259594e-06,
"loss": 0.12630639970302582,
"step": 1040
},
{
"epoch": 0.5554962646744931,
"grad_norm": 1.46875,
"learning_rate": 4.379731767376763e-06,
"loss": 0.13487984240055084,
"step": 1041
},
{
"epoch": 0.5560298826040555,
"grad_norm": 0.71484375,
"learning_rate": 4.3729758008785415e-06,
"loss": 0.13328854739665985,
"step": 1042
},
{
"epoch": 0.5565635005336179,
"grad_norm": 1.59375,
"learning_rate": 4.366219929559407e-06,
"loss": 0.11905941367149353,
"step": 1043
},
{
"epoch": 0.5570971184631803,
"grad_norm": 1.3828125,
"learning_rate": 4.359464177213507e-06,
"loss": 0.1352916806936264,
"step": 1044
},
{
"epoch": 0.5576307363927427,
"grad_norm": 0.890625,
"learning_rate": 4.352708567634564e-06,
"loss": 0.12840241193771362,
"step": 1045
},
{
"epoch": 0.5581643543223053,
"grad_norm": 1.3125,
"learning_rate": 4.345953124615799e-06,
"loss": 0.13541565835475922,
"step": 1046
},
{
"epoch": 0.5586979722518677,
"grad_norm": 4.0625,
"learning_rate": 4.339197871949849e-06,
"loss": 0.14854499697685242,
"step": 1047
},
{
"epoch": 0.5592315901814301,
"grad_norm": 1.890625,
"learning_rate": 4.332442833428677e-06,
"loss": 0.1303594410419464,
"step": 1048
},
{
"epoch": 0.5597652081109925,
"grad_norm": 0.81640625,
"learning_rate": 4.325688032843494e-06,
"loss": 0.13286375999450684,
"step": 1049
},
{
"epoch": 0.5602988260405549,
"grad_norm": 1.4296875,
"learning_rate": 4.318933493984675e-06,
"loss": 0.13165423274040222,
"step": 1050
},
{
"epoch": 0.5608324439701174,
"grad_norm": 0.71875,
"learning_rate": 4.3121792406416665e-06,
"loss": 0.13963672518730164,
"step": 1051
},
{
"epoch": 0.5613660618996799,
"grad_norm": 0.7890625,
"learning_rate": 4.305425296602918e-06,
"loss": 0.13965660333633423,
"step": 1052
},
{
"epoch": 0.5618996798292423,
"grad_norm": 0.8828125,
"learning_rate": 4.29867168565578e-06,
"loss": 0.13566473126411438,
"step": 1053
},
{
"epoch": 0.5624332977588047,
"grad_norm": 1.4765625,
"learning_rate": 4.2919184315864416e-06,
"loss": 0.14009982347488403,
"step": 1054
},
{
"epoch": 0.5629669156883671,
"grad_norm": 1.7265625,
"learning_rate": 4.285165558179825e-06,
"loss": 0.1320953071117401,
"step": 1055
},
{
"epoch": 0.5635005336179295,
"grad_norm": 1.203125,
"learning_rate": 4.2784130892195144e-06,
"loss": 0.13880932331085205,
"step": 1056
},
{
"epoch": 0.564034151547492,
"grad_norm": 0.859375,
"learning_rate": 4.271661048487672e-06,
"loss": 0.13258154690265656,
"step": 1057
},
{
"epoch": 0.5645677694770544,
"grad_norm": 0.984375,
"learning_rate": 4.26490945976495e-06,
"loss": 0.14223544299602509,
"step": 1058
},
{
"epoch": 0.5651013874066169,
"grad_norm": 1.0546875,
"learning_rate": 4.258158346830409e-06,
"loss": 0.14419615268707275,
"step": 1059
},
{
"epoch": 0.5656350053361793,
"grad_norm": 2.171875,
"learning_rate": 4.251407733461431e-06,
"loss": 0.1304776966571808,
"step": 1060
},
{
"epoch": 0.5661686232657417,
"grad_norm": 1.171875,
"learning_rate": 4.244657643433645e-06,
"loss": 0.14619165658950806,
"step": 1061
},
{
"epoch": 0.5667022411953042,
"grad_norm": 0.92578125,
"learning_rate": 4.237908100520829e-06,
"loss": 0.12266339361667633,
"step": 1062
},
{
"epoch": 0.5672358591248666,
"grad_norm": 0.7109375,
"learning_rate": 4.231159128494843e-06,
"loss": 0.1430162787437439,
"step": 1063
},
{
"epoch": 0.567769477054429,
"grad_norm": 0.921875,
"learning_rate": 4.224410751125529e-06,
"loss": 0.137426495552063,
"step": 1064
},
{
"epoch": 0.5683030949839915,
"grad_norm": 0.7734375,
"learning_rate": 4.217662992180634e-06,
"loss": 0.11834961920976639,
"step": 1065
},
{
"epoch": 0.5688367129135539,
"grad_norm": 0.87890625,
"learning_rate": 4.210915875425735e-06,
"loss": 0.13567617535591125,
"step": 1066
},
{
"epoch": 0.5693703308431163,
"grad_norm": 0.97265625,
"learning_rate": 4.204169424624138e-06,
"loss": 0.12750878930091858,
"step": 1067
},
{
"epoch": 0.5699039487726788,
"grad_norm": 1.0078125,
"learning_rate": 4.197423663536812e-06,
"loss": 0.13065484166145325,
"step": 1068
},
{
"epoch": 0.5704375667022412,
"grad_norm": 0.98046875,
"learning_rate": 4.190678615922287e-06,
"loss": 0.13887737691402435,
"step": 1069
},
{
"epoch": 0.5709711846318036,
"grad_norm": 0.87890625,
"learning_rate": 4.18393430553659e-06,
"loss": 0.1474212259054184,
"step": 1070
},
{
"epoch": 0.571504802561366,
"grad_norm": 0.71484375,
"learning_rate": 4.1771907561331455e-06,
"loss": 0.13273531198501587,
"step": 1071
},
{
"epoch": 0.5720384204909285,
"grad_norm": 0.93359375,
"learning_rate": 4.170447991462699e-06,
"loss": 0.14379942417144775,
"step": 1072
},
{
"epoch": 0.572572038420491,
"grad_norm": 1.7890625,
"learning_rate": 4.1637060352732355e-06,
"loss": 0.13873887062072754,
"step": 1073
},
{
"epoch": 0.5731056563500534,
"grad_norm": 0.80078125,
"learning_rate": 4.156964911309886e-06,
"loss": 0.13690324127674103,
"step": 1074
},
{
"epoch": 0.5736392742796158,
"grad_norm": 0.7421875,
"learning_rate": 4.150224643314854e-06,
"loss": 0.13565830886363983,
"step": 1075
},
{
"epoch": 0.5741728922091782,
"grad_norm": 0.91015625,
"learning_rate": 4.143485255027331e-06,
"loss": 0.11935442686080933,
"step": 1076
},
{
"epoch": 0.5747065101387406,
"grad_norm": 0.71484375,
"learning_rate": 4.136746770183405e-06,
"loss": 0.13762709498405457,
"step": 1077
},
{
"epoch": 0.575240128068303,
"grad_norm": 0.6875,
"learning_rate": 4.13000921251599e-06,
"loss": 0.1333584487438202,
"step": 1078
},
{
"epoch": 0.5757737459978656,
"grad_norm": 1.3984375,
"learning_rate": 4.123272605754721e-06,
"loss": 0.13309098780155182,
"step": 1079
},
{
"epoch": 0.576307363927428,
"grad_norm": 0.91796875,
"learning_rate": 4.116536973625898e-06,
"loss": 0.12749740481376648,
"step": 1080
},
{
"epoch": 0.5768409818569904,
"grad_norm": 0.78125,
"learning_rate": 4.109802339852379e-06,
"loss": 0.1302868276834488,
"step": 1081
},
{
"epoch": 0.5773745997865528,
"grad_norm": 0.88671875,
"learning_rate": 4.103068728153511e-06,
"loss": 0.1220366433262825,
"step": 1082
},
{
"epoch": 0.5779082177161152,
"grad_norm": 1.390625,
"learning_rate": 4.09633616224504e-06,
"loss": 0.1462421864271164,
"step": 1083
},
{
"epoch": 0.5784418356456777,
"grad_norm": 1.9921875,
"learning_rate": 4.089604665839025e-06,
"loss": 0.13313287496566772,
"step": 1084
},
{
"epoch": 0.5789754535752402,
"grad_norm": 1.1875,
"learning_rate": 4.082874262643764e-06,
"loss": 0.1224430501461029,
"step": 1085
},
{
"epoch": 0.5795090715048026,
"grad_norm": 1.2109375,
"learning_rate": 4.0761449763637e-06,
"loss": 0.13885769248008728,
"step": 1086
},
{
"epoch": 0.580042689434365,
"grad_norm": 2.46875,
"learning_rate": 4.069416830699348e-06,
"loss": 0.13863904774188995,
"step": 1087
},
{
"epoch": 0.5805763073639274,
"grad_norm": 1.0625,
"learning_rate": 4.0626898493472e-06,
"loss": 0.13306480646133423,
"step": 1088
},
{
"epoch": 0.5811099252934898,
"grad_norm": 0.98046875,
"learning_rate": 4.055964055999647e-06,
"loss": 0.12625133991241455,
"step": 1089
},
{
"epoch": 0.5816435432230523,
"grad_norm": 1.3515625,
"learning_rate": 4.049239474344901e-06,
"loss": 0.14368270337581635,
"step": 1090
},
{
"epoch": 0.5821771611526148,
"grad_norm": 0.84375,
"learning_rate": 4.042516128066904e-06,
"loss": 0.12683819234371185,
"step": 1091
},
{
"epoch": 0.5827107790821772,
"grad_norm": 2.34375,
"learning_rate": 4.035794040845245e-06,
"loss": 0.14313079416751862,
"step": 1092
},
{
"epoch": 0.5832443970117396,
"grad_norm": 1.125,
"learning_rate": 4.0290732363550804e-06,
"loss": 0.12667354941368103,
"step": 1093
},
{
"epoch": 0.583778014941302,
"grad_norm": 1.0625,
"learning_rate": 4.022353738267049e-06,
"loss": 0.13862380385398865,
"step": 1094
},
{
"epoch": 0.5843116328708644,
"grad_norm": 0.91796875,
"learning_rate": 4.015635570247188e-06,
"loss": 0.12905526161193848,
"step": 1095
},
{
"epoch": 0.5848452508004269,
"grad_norm": 0.81640625,
"learning_rate": 4.008918755956852e-06,
"loss": 0.11734165251255035,
"step": 1096
},
{
"epoch": 0.5853788687299893,
"grad_norm": 1.21875,
"learning_rate": 4.002203319052623e-06,
"loss": 0.13108983635902405,
"step": 1097
},
{
"epoch": 0.5859124866595518,
"grad_norm": 0.86328125,
"learning_rate": 3.995489283186237e-06,
"loss": 0.14626872539520264,
"step": 1098
},
{
"epoch": 0.5864461045891142,
"grad_norm": 1.6640625,
"learning_rate": 3.988776672004492e-06,
"loss": 0.12972193956375122,
"step": 1099
},
{
"epoch": 0.5869797225186766,
"grad_norm": 0.73046875,
"learning_rate": 3.98206550914917e-06,
"loss": 0.12787331640720367,
"step": 1100
},
{
"epoch": 0.5875133404482391,
"grad_norm": 1.4375,
"learning_rate": 3.975355818256955e-06,
"loss": 0.13627779483795166,
"step": 1101
},
{
"epoch": 0.5880469583778015,
"grad_norm": 1.28125,
"learning_rate": 3.968647622959338e-06,
"loss": 0.1409071981906891,
"step": 1102
},
{
"epoch": 0.5885805763073639,
"grad_norm": 3.0,
"learning_rate": 3.961940946882551e-06,
"loss": 0.14202764630317688,
"step": 1103
},
{
"epoch": 0.5891141942369263,
"grad_norm": 0.6328125,
"learning_rate": 3.955235813647473e-06,
"loss": 0.11367058008909225,
"step": 1104
},
{
"epoch": 0.5896478121664888,
"grad_norm": 0.96484375,
"learning_rate": 3.948532246869546e-06,
"loss": 0.11737130582332611,
"step": 1105
},
{
"epoch": 0.5901814300960512,
"grad_norm": 2.140625,
"learning_rate": 3.9418302701587e-06,
"loss": 0.1399267017841339,
"step": 1106
},
{
"epoch": 0.5907150480256137,
"grad_norm": 1.09375,
"learning_rate": 3.93512990711926e-06,
"loss": 0.13245609402656555,
"step": 1107
},
{
"epoch": 0.5912486659551761,
"grad_norm": 0.8828125,
"learning_rate": 3.928431181349869e-06,
"loss": 0.12998540699481964,
"step": 1108
},
{
"epoch": 0.5917822838847385,
"grad_norm": 1.0078125,
"learning_rate": 3.921734116443406e-06,
"loss": 0.1393250823020935,
"step": 1109
},
{
"epoch": 0.5923159018143009,
"grad_norm": 2.65625,
"learning_rate": 3.915038735986901e-06,
"loss": 0.13024061918258667,
"step": 1110
},
{
"epoch": 0.5928495197438634,
"grad_norm": 0.8671875,
"learning_rate": 3.908345063561446e-06,
"loss": 0.14949244260787964,
"step": 1111
},
{
"epoch": 0.5933831376734259,
"grad_norm": 0.6875,
"learning_rate": 3.901653122742119e-06,
"loss": 0.13749602437019348,
"step": 1112
},
{
"epoch": 0.5939167556029883,
"grad_norm": 1.703125,
"learning_rate": 3.894962937097905e-06,
"loss": 0.13145318627357483,
"step": 1113
},
{
"epoch": 0.5944503735325507,
"grad_norm": 3.203125,
"learning_rate": 3.888274530191602e-06,
"loss": 0.12795287370681763,
"step": 1114
},
{
"epoch": 0.5949839914621131,
"grad_norm": 4.0625,
"learning_rate": 3.881587925579741e-06,
"loss": 0.13613222539424896,
"step": 1115
},
{
"epoch": 0.5955176093916755,
"grad_norm": 1.0234375,
"learning_rate": 3.874903146812513e-06,
"loss": 0.13701140880584717,
"step": 1116
},
{
"epoch": 0.5960512273212379,
"grad_norm": 1.109375,
"learning_rate": 3.868220217433672e-06,
"loss": 0.1258457601070404,
"step": 1117
},
{
"epoch": 0.5965848452508005,
"grad_norm": 1.0,
"learning_rate": 3.861539160980459e-06,
"loss": 0.12153815478086472,
"step": 1118
},
{
"epoch": 0.5971184631803629,
"grad_norm": 0.77734375,
"learning_rate": 3.8548600009835246e-06,
"loss": 0.14082077145576477,
"step": 1119
},
{
"epoch": 0.5976520811099253,
"grad_norm": 0.72265625,
"learning_rate": 3.848182760966831e-06,
"loss": 0.12793788313865662,
"step": 1120
},
{
"epoch": 0.5981856990394877,
"grad_norm": 1.390625,
"learning_rate": 3.841507464447585e-06,
"loss": 0.12715861201286316,
"step": 1121
},
{
"epoch": 0.5987193169690501,
"grad_norm": 1.0859375,
"learning_rate": 3.8348341349361455e-06,
"loss": 0.13636955618858337,
"step": 1122
},
{
"epoch": 0.5992529348986126,
"grad_norm": 0.75390625,
"learning_rate": 3.828162795935945e-06,
"loss": 0.14104749262332916,
"step": 1123
},
{
"epoch": 0.5997865528281751,
"grad_norm": 1.4375,
"learning_rate": 3.821493470943406e-06,
"loss": 0.13936856389045715,
"step": 1124
},
{
"epoch": 0.6003201707577375,
"grad_norm": 0.9140625,
"learning_rate": 3.8148261834478535e-06,
"loss": 0.13057279586791992,
"step": 1125
},
{
"epoch": 0.6008537886872999,
"grad_norm": 0.8984375,
"learning_rate": 3.8081609569314407e-06,
"loss": 0.13948747515678406,
"step": 1126
},
{
"epoch": 0.6013874066168623,
"grad_norm": 1.234375,
"learning_rate": 3.8014978148690603e-06,
"loss": 0.13593068718910217,
"step": 1127
},
{
"epoch": 0.6019210245464247,
"grad_norm": 0.890625,
"learning_rate": 3.7948367807282667e-06,
"loss": 0.126223623752594,
"step": 1128
},
{
"epoch": 0.6024546424759872,
"grad_norm": 1.9140625,
"learning_rate": 3.7881778779691845e-06,
"loss": 0.12878422439098358,
"step": 1129
},
{
"epoch": 0.6029882604055496,
"grad_norm": 1.3046875,
"learning_rate": 3.781521130044436e-06,
"loss": 0.12819713354110718,
"step": 1130
},
{
"epoch": 0.6035218783351121,
"grad_norm": 0.90625,
"learning_rate": 3.7748665603990507e-06,
"loss": 0.1334487795829773,
"step": 1131
},
{
"epoch": 0.6040554962646745,
"grad_norm": 1.671875,
"learning_rate": 3.76821419247039e-06,
"loss": 0.13485948741436005,
"step": 1132
},
{
"epoch": 0.6045891141942369,
"grad_norm": 0.953125,
"learning_rate": 3.7615640496880614e-06,
"loss": 0.12891358137130737,
"step": 1133
},
{
"epoch": 0.6051227321237994,
"grad_norm": 1.046875,
"learning_rate": 3.75491615547383e-06,
"loss": 0.11793476343154907,
"step": 1134
},
{
"epoch": 0.6056563500533618,
"grad_norm": 0.9140625,
"learning_rate": 3.748270533241544e-06,
"loss": 0.13812406361103058,
"step": 1135
},
{
"epoch": 0.6061899679829242,
"grad_norm": 0.9609375,
"learning_rate": 3.7416272063970523e-06,
"loss": 0.13613979518413544,
"step": 1136
},
{
"epoch": 0.6067235859124867,
"grad_norm": 1.03125,
"learning_rate": 3.7349861983381194e-06,
"loss": 0.15167632699012756,
"step": 1137
},
{
"epoch": 0.6072572038420491,
"grad_norm": 1.28125,
"learning_rate": 3.7283475324543366e-06,
"loss": 0.148903027176857,
"step": 1138
},
{
"epoch": 0.6077908217716115,
"grad_norm": 1.328125,
"learning_rate": 3.721711232127053e-06,
"loss": 0.13803043961524963,
"step": 1139
},
{
"epoch": 0.608324439701174,
"grad_norm": 0.78515625,
"learning_rate": 3.7150773207292836e-06,
"loss": 0.1254650354385376,
"step": 1140
},
{
"epoch": 0.6088580576307364,
"grad_norm": 0.921875,
"learning_rate": 3.7084458216256317e-06,
"loss": 0.12935568392276764,
"step": 1141
},
{
"epoch": 0.6093916755602988,
"grad_norm": 2.03125,
"learning_rate": 3.7018167581722004e-06,
"loss": 0.15041708946228027,
"step": 1142
},
{
"epoch": 0.6099252934898612,
"grad_norm": 1.5234375,
"learning_rate": 3.695190153716517e-06,
"loss": 0.14515599608421326,
"step": 1143
},
{
"epoch": 0.6104589114194237,
"grad_norm": 0.78515625,
"learning_rate": 3.6885660315974478e-06,
"loss": 0.1452225148677826,
"step": 1144
},
{
"epoch": 0.6109925293489862,
"grad_norm": 0.9765625,
"learning_rate": 3.6819444151451168e-06,
"loss": 0.1351228654384613,
"step": 1145
},
{
"epoch": 0.6115261472785486,
"grad_norm": 0.765625,
"learning_rate": 3.675325327680823e-06,
"loss": 0.14078357815742493,
"step": 1146
},
{
"epoch": 0.612059765208111,
"grad_norm": 0.84765625,
"learning_rate": 3.66870879251696e-06,
"loss": 0.13935934007167816,
"step": 1147
},
{
"epoch": 0.6125933831376734,
"grad_norm": 1.625,
"learning_rate": 3.6620948329569253e-06,
"loss": 0.13806846737861633,
"step": 1148
},
{
"epoch": 0.6131270010672358,
"grad_norm": 0.7578125,
"learning_rate": 3.655483472295054e-06,
"loss": 0.14086604118347168,
"step": 1149
},
{
"epoch": 0.6136606189967982,
"grad_norm": 0.71875,
"learning_rate": 3.648874733816522e-06,
"loss": 0.1346871554851532,
"step": 1150
},
{
"epoch": 0.6141942369263608,
"grad_norm": 1.15625,
"learning_rate": 3.6422686407972753e-06,
"loss": 0.13968732953071594,
"step": 1151
},
{
"epoch": 0.6147278548559232,
"grad_norm": 0.8828125,
"learning_rate": 3.635665216503935e-06,
"loss": 0.1290009319782257,
"step": 1152
},
{
"epoch": 0.6152614727854856,
"grad_norm": 1.46875,
"learning_rate": 3.6290644841937283e-06,
"loss": 0.1250561773777008,
"step": 1153
},
{
"epoch": 0.615795090715048,
"grad_norm": 0.82421875,
"learning_rate": 3.6224664671144016e-06,
"loss": 0.14070561528205872,
"step": 1154
},
{
"epoch": 0.6163287086446104,
"grad_norm": 0.90234375,
"learning_rate": 3.615871188504134e-06,
"loss": 0.13460339605808258,
"step": 1155
},
{
"epoch": 0.6168623265741728,
"grad_norm": 1.078125,
"learning_rate": 3.6092786715914663e-06,
"loss": 0.1325269341468811,
"step": 1156
},
{
"epoch": 0.6173959445037354,
"grad_norm": 2.9375,
"learning_rate": 3.6026889395952054e-06,
"loss": 0.12819279730319977,
"step": 1157
},
{
"epoch": 0.6179295624332978,
"grad_norm": 1.140625,
"learning_rate": 3.596102015724354e-06,
"loss": 0.13029253482818604,
"step": 1158
},
{
"epoch": 0.6184631803628602,
"grad_norm": 2.171875,
"learning_rate": 3.589517923178025e-06,
"loss": 0.14932042360305786,
"step": 1159
},
{
"epoch": 0.6189967982924226,
"grad_norm": 7.15625,
"learning_rate": 3.582936685145358e-06,
"loss": 0.13518399000167847,
"step": 1160
},
{
"epoch": 0.619530416221985,
"grad_norm": 2.796875,
"learning_rate": 3.5763583248054383e-06,
"loss": 0.12480396777391434,
"step": 1161
},
{
"epoch": 0.6200640341515475,
"grad_norm": 1.0390625,
"learning_rate": 3.569782865327215e-06,
"loss": 0.11909421533346176,
"step": 1162
},
{
"epoch": 0.62059765208111,
"grad_norm": 1.0859375,
"learning_rate": 3.5632103298694257e-06,
"loss": 0.1231987401843071,
"step": 1163
},
{
"epoch": 0.6211312700106724,
"grad_norm": 1.0390625,
"learning_rate": 3.556640741580504e-06,
"loss": 0.1268862932920456,
"step": 1164
},
{
"epoch": 0.6216648879402348,
"grad_norm": 1.2421875,
"learning_rate": 3.5500741235985075e-06,
"loss": 0.1386156678199768,
"step": 1165
},
{
"epoch": 0.6221985058697972,
"grad_norm": 0.8359375,
"learning_rate": 3.5435104990510275e-06,
"loss": 0.11890492588281631,
"step": 1166
},
{
"epoch": 0.6227321237993596,
"grad_norm": 0.69140625,
"learning_rate": 3.5369498910551183e-06,
"loss": 0.1361309438943863,
"step": 1167
},
{
"epoch": 0.6232657417289221,
"grad_norm": 0.80859375,
"learning_rate": 3.5303923227172065e-06,
"loss": 0.13088691234588623,
"step": 1168
},
{
"epoch": 0.6237993596584845,
"grad_norm": 1.5234375,
"learning_rate": 3.523837817133012e-06,
"loss": 0.13206958770751953,
"step": 1169
},
{
"epoch": 0.624332977588047,
"grad_norm": 1.2578125,
"learning_rate": 3.5172863973874734e-06,
"loss": 0.12555518746376038,
"step": 1170
},
{
"epoch": 0.6248665955176094,
"grad_norm": 1.4140625,
"learning_rate": 3.510738086554653e-06,
"loss": 0.13994286954402924,
"step": 1171
},
{
"epoch": 0.6254002134471718,
"grad_norm": 0.765625,
"learning_rate": 3.5041929076976676e-06,
"loss": 0.13059481978416443,
"step": 1172
},
{
"epoch": 0.6259338313767343,
"grad_norm": 0.69140625,
"learning_rate": 3.497650883868606e-06,
"loss": 0.13617053627967834,
"step": 1173
},
{
"epoch": 0.6264674493062967,
"grad_norm": 1.515625,
"learning_rate": 3.4911120381084407e-06,
"loss": 0.1335827112197876,
"step": 1174
},
{
"epoch": 0.6270010672358591,
"grad_norm": 0.890625,
"learning_rate": 3.484576393446952e-06,
"loss": 0.12766879796981812,
"step": 1175
},
{
"epoch": 0.6275346851654215,
"grad_norm": 0.9609375,
"learning_rate": 3.478043972902646e-06,
"loss": 0.13389284908771515,
"step": 1176
},
{
"epoch": 0.628068303094984,
"grad_norm": 1.375,
"learning_rate": 3.4715147994826755e-06,
"loss": 0.1349835991859436,
"step": 1177
},
{
"epoch": 0.6286019210245464,
"grad_norm": 1.3046875,
"learning_rate": 3.4649888961827535e-06,
"loss": 0.14413315057754517,
"step": 1178
},
{
"epoch": 0.6291355389541089,
"grad_norm": 1.25,
"learning_rate": 3.4584662859870797e-06,
"loss": 0.11854719370603561,
"step": 1179
},
{
"epoch": 0.6296691568836713,
"grad_norm": 0.765625,
"learning_rate": 3.4519469918682526e-06,
"loss": 0.12459492683410645,
"step": 1180
},
{
"epoch": 0.6302027748132337,
"grad_norm": 1.109375,
"learning_rate": 3.44543103678719e-06,
"loss": 0.12003341317176819,
"step": 1181
},
{
"epoch": 0.6307363927427961,
"grad_norm": 1.46875,
"learning_rate": 3.4389184436930537e-06,
"loss": 0.1370922476053238,
"step": 1182
},
{
"epoch": 0.6312700106723586,
"grad_norm": 1.7578125,
"learning_rate": 3.432409235523164e-06,
"loss": 0.12402217090129852,
"step": 1183
},
{
"epoch": 0.6318036286019211,
"grad_norm": 1.296875,
"learning_rate": 3.425903435202915e-06,
"loss": 0.15181377530097961,
"step": 1184
},
{
"epoch": 0.6323372465314835,
"grad_norm": 1.0703125,
"learning_rate": 3.4194010656457044e-06,
"loss": 0.13342611491680145,
"step": 1185
},
{
"epoch": 0.6328708644610459,
"grad_norm": 0.91015625,
"learning_rate": 3.4129021497528416e-06,
"loss": 0.12351104617118835,
"step": 1186
},
{
"epoch": 0.6334044823906083,
"grad_norm": 1.25,
"learning_rate": 3.4064067104134775e-06,
"loss": 0.14127784967422485,
"step": 1187
},
{
"epoch": 0.6339381003201707,
"grad_norm": 1.0703125,
"learning_rate": 3.399914770504515e-06,
"loss": 0.10422603785991669,
"step": 1188
},
{
"epoch": 0.6344717182497331,
"grad_norm": 1.421875,
"learning_rate": 3.3934263528905296e-06,
"loss": 0.1303495466709137,
"step": 1189
},
{
"epoch": 0.6350053361792957,
"grad_norm": 0.8046875,
"learning_rate": 3.3869414804236984e-06,
"loss": 0.14190465211868286,
"step": 1190
},
{
"epoch": 0.6355389541088581,
"grad_norm": 1.1875,
"learning_rate": 3.380460175943705e-06,
"loss": 0.14931124448776245,
"step": 1191
},
{
"epoch": 0.6360725720384205,
"grad_norm": 0.9609375,
"learning_rate": 3.3739824622776726e-06,
"loss": 0.12776312232017517,
"step": 1192
},
{
"epoch": 0.6366061899679829,
"grad_norm": 0.93359375,
"learning_rate": 3.367508362240076e-06,
"loss": 0.13060502707958221,
"step": 1193
},
{
"epoch": 0.6371398078975453,
"grad_norm": 1.5625,
"learning_rate": 3.3610378986326584e-06,
"loss": 0.13593590259552002,
"step": 1194
},
{
"epoch": 0.6376734258271078,
"grad_norm": 1.9296875,
"learning_rate": 3.3545710942443605e-06,
"loss": 0.14686843752861023,
"step": 1195
},
{
"epoch": 0.6382070437566703,
"grad_norm": 1.03125,
"learning_rate": 3.3481079718512346e-06,
"loss": 0.1271577924489975,
"step": 1196
},
{
"epoch": 0.6387406616862327,
"grad_norm": 2.125,
"learning_rate": 3.3416485542163645e-06,
"loss": 0.14371488988399506,
"step": 1197
},
{
"epoch": 0.6392742796157951,
"grad_norm": 1.34375,
"learning_rate": 3.3351928640897827e-06,
"loss": 0.12314917147159576,
"step": 1198
},
{
"epoch": 0.6398078975453575,
"grad_norm": 0.8203125,
"learning_rate": 3.328740924208399e-06,
"loss": 0.1370874047279358,
"step": 1199
},
{
"epoch": 0.6403415154749199,
"grad_norm": 0.8828125,
"learning_rate": 3.322292757295911e-06,
"loss": 0.13661031424999237,
"step": 1200
},
{
"epoch": 0.6408751334044824,
"grad_norm": 1.0078125,
"learning_rate": 3.31584838606273e-06,
"loss": 0.1359814703464508,
"step": 1201
},
{
"epoch": 0.6414087513340448,
"grad_norm": 0.80078125,
"learning_rate": 3.309407833205897e-06,
"loss": 0.12317250669002533,
"step": 1202
},
{
"epoch": 0.6419423692636073,
"grad_norm": 1.0,
"learning_rate": 3.3029711214090065e-06,
"loss": 0.1275547444820404,
"step": 1203
},
{
"epoch": 0.6424759871931697,
"grad_norm": 0.8671875,
"learning_rate": 3.296538273342123e-06,
"loss": 0.1242324560880661,
"step": 1204
},
{
"epoch": 0.6430096051227321,
"grad_norm": 0.9765625,
"learning_rate": 3.2901093116617043e-06,
"loss": 0.12946733832359314,
"step": 1205
},
{
"epoch": 0.6435432230522946,
"grad_norm": 0.70703125,
"learning_rate": 3.283684259010521e-06,
"loss": 0.13204079866409302,
"step": 1206
},
{
"epoch": 0.644076840981857,
"grad_norm": 4.15625,
"learning_rate": 3.2772631380175737e-06,
"loss": 0.14818671345710754,
"step": 1207
},
{
"epoch": 0.6446104589114194,
"grad_norm": 0.6796875,
"learning_rate": 3.2708459712980176e-06,
"loss": 0.13285815715789795,
"step": 1208
},
{
"epoch": 0.6451440768409819,
"grad_norm": 0.81640625,
"learning_rate": 3.2644327814530803e-06,
"loss": 0.14059065282344818,
"step": 1209
},
{
"epoch": 0.6456776947705443,
"grad_norm": 0.8125,
"learning_rate": 3.2580235910699837e-06,
"loss": 0.1406671702861786,
"step": 1210
},
{
"epoch": 0.6462113127001067,
"grad_norm": 1.5703125,
"learning_rate": 3.251618422721864e-06,
"loss": 0.1333598792552948,
"step": 1211
},
{
"epoch": 0.6467449306296692,
"grad_norm": 1.203125,
"learning_rate": 3.245217298967689e-06,
"loss": 0.13585197925567627,
"step": 1212
},
{
"epoch": 0.6472785485592316,
"grad_norm": 0.796875,
"learning_rate": 3.238820242352183e-06,
"loss": 0.13474589586257935,
"step": 1213
},
{
"epoch": 0.647812166488794,
"grad_norm": 1.6171875,
"learning_rate": 3.2324272754057475e-06,
"loss": 0.14317402243614197,
"step": 1214
},
{
"epoch": 0.6483457844183564,
"grad_norm": 0.76953125,
"learning_rate": 3.2260384206443788e-06,
"loss": 0.1448487639427185,
"step": 1215
},
{
"epoch": 0.6488794023479189,
"grad_norm": 1.4921875,
"learning_rate": 3.21965370056959e-06,
"loss": 0.13826508820056915,
"step": 1216
},
{
"epoch": 0.6494130202774814,
"grad_norm": 0.67578125,
"learning_rate": 3.2132731376683307e-06,
"loss": 0.1429232358932495,
"step": 1217
},
{
"epoch": 0.6499466382070438,
"grad_norm": 0.95703125,
"learning_rate": 3.2068967544129112e-06,
"loss": 0.12072071433067322,
"step": 1218
},
{
"epoch": 0.6504802561366062,
"grad_norm": 1.25,
"learning_rate": 3.200524573260921e-06,
"loss": 0.1188340038061142,
"step": 1219
},
{
"epoch": 0.6510138740661686,
"grad_norm": 0.84375,
"learning_rate": 3.1941566166551487e-06,
"loss": 0.13041101396083832,
"step": 1220
},
{
"epoch": 0.651547491995731,
"grad_norm": 1.390625,
"learning_rate": 3.1877929070235043e-06,
"loss": 0.1433843970298767,
"step": 1221
},
{
"epoch": 0.6520811099252934,
"grad_norm": 1.3359375,
"learning_rate": 3.1814334667789372e-06,
"loss": 0.1304730474948883,
"step": 1222
},
{
"epoch": 0.652614727854856,
"grad_norm": 1.296875,
"learning_rate": 3.1750783183193672e-06,
"loss": 0.13810792565345764,
"step": 1223
},
{
"epoch": 0.6531483457844184,
"grad_norm": 1.234375,
"learning_rate": 3.1687274840275925e-06,
"loss": 0.13427948951721191,
"step": 1224
},
{
"epoch": 0.6536819637139808,
"grad_norm": 1.96875,
"learning_rate": 3.162380986271218e-06,
"loss": 0.13642597198486328,
"step": 1225
},
{
"epoch": 0.6542155816435432,
"grad_norm": 0.82421875,
"learning_rate": 3.156038847402576e-06,
"loss": 0.1398303508758545,
"step": 1226
},
{
"epoch": 0.6547491995731056,
"grad_norm": 1.5078125,
"learning_rate": 3.1497010897586477e-06,
"loss": 0.13202576339244843,
"step": 1227
},
{
"epoch": 0.655282817502668,
"grad_norm": 1.546875,
"learning_rate": 3.1433677356609794e-06,
"loss": 0.1400923728942871,
"step": 1228
},
{
"epoch": 0.6558164354322306,
"grad_norm": 1.140625,
"learning_rate": 3.1370388074156162e-06,
"loss": 0.14118880033493042,
"step": 1229
},
{
"epoch": 0.656350053361793,
"grad_norm": 0.97265625,
"learning_rate": 3.130714327313006e-06,
"loss": 0.15352116525173187,
"step": 1230
},
{
"epoch": 0.6568836712913554,
"grad_norm": 0.80078125,
"learning_rate": 3.124394317627937e-06,
"loss": 0.13610461354255676,
"step": 1231
},
{
"epoch": 0.6574172892209178,
"grad_norm": 3.640625,
"learning_rate": 3.118078800619449e-06,
"loss": 0.14302591979503632,
"step": 1232
},
{
"epoch": 0.6579509071504802,
"grad_norm": 0.78515625,
"learning_rate": 3.1117677985307626e-06,
"loss": 0.13993985950946808,
"step": 1233
},
{
"epoch": 0.6584845250800427,
"grad_norm": 0.859375,
"learning_rate": 3.1054613335891936e-06,
"loss": 0.11312638223171234,
"step": 1234
},
{
"epoch": 0.6590181430096052,
"grad_norm": 0.97265625,
"learning_rate": 3.0991594280060767e-06,
"loss": 0.14367996156215668,
"step": 1235
},
{
"epoch": 0.6595517609391676,
"grad_norm": 1.4140625,
"learning_rate": 3.092862103976691e-06,
"loss": 0.12292146682739258,
"step": 1236
},
{
"epoch": 0.66008537886873,
"grad_norm": 0.90234375,
"learning_rate": 3.0865693836801825e-06,
"loss": 0.12010985612869263,
"step": 1237
},
{
"epoch": 0.6606189967982924,
"grad_norm": 0.69140625,
"learning_rate": 3.0802812892794745e-06,
"loss": 0.14190517365932465,
"step": 1238
},
{
"epoch": 0.6611526147278548,
"grad_norm": 1.4375,
"learning_rate": 3.073997842921207e-06,
"loss": 0.12481766194105148,
"step": 1239
},
{
"epoch": 0.6616862326574173,
"grad_norm": 1.0078125,
"learning_rate": 3.067719066735642e-06,
"loss": 0.13497541844844818,
"step": 1240
},
{
"epoch": 0.6622198505869797,
"grad_norm": 1.4765625,
"learning_rate": 3.0614449828365983e-06,
"loss": 0.13327732682228088,
"step": 1241
},
{
"epoch": 0.6627534685165422,
"grad_norm": 0.77734375,
"learning_rate": 3.055175613321367e-06,
"loss": 0.1439843773841858,
"step": 1242
},
{
"epoch": 0.6632870864461046,
"grad_norm": 0.99609375,
"learning_rate": 3.0489109802706352e-06,
"loss": 0.13124597072601318,
"step": 1243
},
{
"epoch": 0.663820704375667,
"grad_norm": 4.125,
"learning_rate": 3.0426511057484066e-06,
"loss": 0.1177617609500885,
"step": 1244
},
{
"epoch": 0.6643543223052295,
"grad_norm": 1.3515625,
"learning_rate": 3.036396011801926e-06,
"loss": 0.1474757194519043,
"step": 1245
},
{
"epoch": 0.6648879402347919,
"grad_norm": 1.234375,
"learning_rate": 3.030145720461605e-06,
"loss": 0.1434563845396042,
"step": 1246
},
{
"epoch": 0.6654215581643543,
"grad_norm": 1.0546875,
"learning_rate": 3.023900253740934e-06,
"loss": 0.14083580672740936,
"step": 1247
},
{
"epoch": 0.6659551760939167,
"grad_norm": 0.5859375,
"learning_rate": 3.017659633636417e-06,
"loss": 0.11781337857246399,
"step": 1248
},
{
"epoch": 0.6664887940234792,
"grad_norm": 0.78125,
"learning_rate": 3.0114238821274836e-06,
"loss": 0.1182209849357605,
"step": 1249
},
{
"epoch": 0.6670224119530416,
"grad_norm": 0.74609375,
"learning_rate": 3.0051930211764188e-06,
"loss": 0.1384398341178894,
"step": 1250
},
{
"epoch": 0.6675560298826041,
"grad_norm": 1.1171875,
"learning_rate": 2.998967072728283e-06,
"loss": 0.14124858379364014,
"step": 1251
},
{
"epoch": 0.6680896478121665,
"grad_norm": 2.140625,
"learning_rate": 2.9927460587108328e-06,
"loss": 0.13800565898418427,
"step": 1252
},
{
"epoch": 0.6686232657417289,
"grad_norm": 0.9296875,
"learning_rate": 2.9865300010344497e-06,
"loss": 0.13455000519752502,
"step": 1253
},
{
"epoch": 0.6691568836712913,
"grad_norm": 0.74609375,
"learning_rate": 2.9803189215920536e-06,
"loss": 0.14690835773944855,
"step": 1254
},
{
"epoch": 0.6696905016008537,
"grad_norm": 0.83203125,
"learning_rate": 2.9741128422590356e-06,
"loss": 0.1337839663028717,
"step": 1255
},
{
"epoch": 0.6702241195304163,
"grad_norm": 1.203125,
"learning_rate": 2.967911784893174e-06,
"loss": 0.13191694021224976,
"step": 1256
},
{
"epoch": 0.6707577374599787,
"grad_norm": 1.984375,
"learning_rate": 2.961715771334564e-06,
"loss": 0.12304560840129852,
"step": 1257
},
{
"epoch": 0.6712913553895411,
"grad_norm": 2.1875,
"learning_rate": 2.9555248234055283e-06,
"loss": 0.13382478058338165,
"step": 1258
},
{
"epoch": 0.6718249733191035,
"grad_norm": 0.68359375,
"learning_rate": 2.9493389629105557e-06,
"loss": 0.1367797702550888,
"step": 1259
},
{
"epoch": 0.6723585912486659,
"grad_norm": 1.5078125,
"learning_rate": 2.943158211636214e-06,
"loss": 0.13932141661643982,
"step": 1260
},
{
"epoch": 0.6728922091782283,
"grad_norm": 1.0234375,
"learning_rate": 2.9369825913510787e-06,
"loss": 0.1375121772289276,
"step": 1261
},
{
"epoch": 0.6734258271077909,
"grad_norm": 13.25,
"learning_rate": 2.9308121238056503e-06,
"loss": 0.12229888141155243,
"step": 1262
},
{
"epoch": 0.6739594450373533,
"grad_norm": 0.81640625,
"learning_rate": 2.9246468307322856e-06,
"loss": 0.13100147247314453,
"step": 1263
},
{
"epoch": 0.6744930629669157,
"grad_norm": 0.8984375,
"learning_rate": 2.9184867338451124e-06,
"loss": 0.13381949067115784,
"step": 1264
},
{
"epoch": 0.6750266808964781,
"grad_norm": 0.80859375,
"learning_rate": 2.9123318548399604e-06,
"loss": 0.13163337111473083,
"step": 1265
},
{
"epoch": 0.6755602988260405,
"grad_norm": 0.80859375,
"learning_rate": 2.906182215394284e-06,
"loss": 0.12382838875055313,
"step": 1266
},
{
"epoch": 0.676093916755603,
"grad_norm": 1.8046875,
"learning_rate": 2.900037837167077e-06,
"loss": 0.12427783012390137,
"step": 1267
},
{
"epoch": 0.6766275346851655,
"grad_norm": 1.359375,
"learning_rate": 2.89389874179881e-06,
"loss": 0.15265081822872162,
"step": 1268
},
{
"epoch": 0.6771611526147279,
"grad_norm": 1.0625,
"learning_rate": 2.8877649509113467e-06,
"loss": 0.1261935830116272,
"step": 1269
},
{
"epoch": 0.6776947705442903,
"grad_norm": 0.98828125,
"learning_rate": 2.8816364861078642e-06,
"loss": 0.12687397003173828,
"step": 1270
},
{
"epoch": 0.6782283884738527,
"grad_norm": 0.82421875,
"learning_rate": 2.8755133689727843e-06,
"loss": 0.13381341099739075,
"step": 1271
},
{
"epoch": 0.6787620064034151,
"grad_norm": 0.70703125,
"learning_rate": 2.8693956210716942e-06,
"loss": 0.12962324917316437,
"step": 1272
},
{
"epoch": 0.6792956243329776,
"grad_norm": 1.671875,
"learning_rate": 2.8632832639512685e-06,
"loss": 0.13303682208061218,
"step": 1273
},
{
"epoch": 0.67982924226254,
"grad_norm": 0.859375,
"learning_rate": 2.8571763191391994e-06,
"loss": 0.14044779539108276,
"step": 1274
},
{
"epoch": 0.6803628601921025,
"grad_norm": 1.953125,
"learning_rate": 2.851074808144115e-06,
"loss": 0.14610075950622559,
"step": 1275
},
{
"epoch": 0.6808964781216649,
"grad_norm": 0.6953125,
"learning_rate": 2.8449787524555033e-06,
"loss": 0.13521666824817657,
"step": 1276
},
{
"epoch": 0.6814300960512273,
"grad_norm": 0.95703125,
"learning_rate": 2.8388881735436404e-06,
"loss": 0.1302107721567154,
"step": 1277
},
{
"epoch": 0.6819637139807898,
"grad_norm": 0.81640625,
"learning_rate": 2.8328030928595165e-06,
"loss": 0.13115301728248596,
"step": 1278
},
{
"epoch": 0.6824973319103522,
"grad_norm": 0.84375,
"learning_rate": 2.8267235318347526e-06,
"loss": 0.13243651390075684,
"step": 1279
},
{
"epoch": 0.6830309498399146,
"grad_norm": 1.1328125,
"learning_rate": 2.8206495118815303e-06,
"loss": 0.13630351424217224,
"step": 1280
},
{
"epoch": 0.683564567769477,
"grad_norm": 1.15625,
"learning_rate": 2.8145810543925166e-06,
"loss": 0.1435500979423523,
"step": 1281
},
{
"epoch": 0.6840981856990395,
"grad_norm": 1.4765625,
"learning_rate": 2.8085181807407877e-06,
"loss": 0.13336801528930664,
"step": 1282
},
{
"epoch": 0.6846318036286019,
"grad_norm": 0.890625,
"learning_rate": 2.8024609122797537e-06,
"loss": 0.12938794493675232,
"step": 1283
},
{
"epoch": 0.6851654215581644,
"grad_norm": 1.5390625,
"learning_rate": 2.79640927034308e-06,
"loss": 0.1579258143901825,
"step": 1284
},
{
"epoch": 0.6856990394877268,
"grad_norm": 1.015625,
"learning_rate": 2.790363276244623e-06,
"loss": 0.1352907419204712,
"step": 1285
},
{
"epoch": 0.6862326574172892,
"grad_norm": 0.96875,
"learning_rate": 2.7843229512783364e-06,
"loss": 0.12520363926887512,
"step": 1286
},
{
"epoch": 0.6867662753468516,
"grad_norm": 0.64453125,
"learning_rate": 2.7782883167182176e-06,
"loss": 0.12929406762123108,
"step": 1287
},
{
"epoch": 0.687299893276414,
"grad_norm": 0.9296875,
"learning_rate": 2.7722593938182153e-06,
"loss": 0.1288771629333496,
"step": 1288
},
{
"epoch": 0.6878335112059765,
"grad_norm": 0.734375,
"learning_rate": 2.7662362038121695e-06,
"loss": 0.12036430835723877,
"step": 1289
},
{
"epoch": 0.688367129135539,
"grad_norm": 2.546875,
"learning_rate": 2.7602187679137176e-06,
"loss": 0.14297065138816833,
"step": 1290
},
{
"epoch": 0.6889007470651014,
"grad_norm": 0.890625,
"learning_rate": 2.754207107316242e-06,
"loss": 0.12484479695558548,
"step": 1291
},
{
"epoch": 0.6894343649946638,
"grad_norm": 1.1796875,
"learning_rate": 2.7482012431927785e-06,
"loss": 0.12934032082557678,
"step": 1292
},
{
"epoch": 0.6899679829242262,
"grad_norm": 1.03125,
"learning_rate": 2.742201196695949e-06,
"loss": 0.1202348992228508,
"step": 1293
},
{
"epoch": 0.6905016008537886,
"grad_norm": 0.94921875,
"learning_rate": 2.736206988957888e-06,
"loss": 0.13583780825138092,
"step": 1294
},
{
"epoch": 0.6910352187833512,
"grad_norm": 0.921875,
"learning_rate": 2.730218641090162e-06,
"loss": 0.1329081654548645,
"step": 1295
},
{
"epoch": 0.6915688367129136,
"grad_norm": 0.64453125,
"learning_rate": 2.724236174183702e-06,
"loss": 0.14423075318336487,
"step": 1296
},
{
"epoch": 0.692102454642476,
"grad_norm": 1.8203125,
"learning_rate": 2.718259609308724e-06,
"loss": 0.12285824120044708,
"step": 1297
},
{
"epoch": 0.6926360725720384,
"grad_norm": 0.72265625,
"learning_rate": 2.712288967514662e-06,
"loss": 0.14845716953277588,
"step": 1298
},
{
"epoch": 0.6931696905016008,
"grad_norm": 3.0625,
"learning_rate": 2.7063242698300805e-06,
"loss": 0.1326768547296524,
"step": 1299
},
{
"epoch": 0.6937033084311632,
"grad_norm": 0.9921875,
"learning_rate": 2.700365537262617e-06,
"loss": 0.1288241744041443,
"step": 1300
},
{
"epoch": 0.6942369263607258,
"grad_norm": 1.671875,
"learning_rate": 2.694412790798894e-06,
"loss": 0.12808841466903687,
"step": 1301
},
{
"epoch": 0.6947705442902882,
"grad_norm": 0.99609375,
"learning_rate": 2.688466051404457e-06,
"loss": 0.13787710666656494,
"step": 1302
},
{
"epoch": 0.6953041622198506,
"grad_norm": 1.046875,
"learning_rate": 2.68252534002369e-06,
"loss": 0.1105039045214653,
"step": 1303
},
{
"epoch": 0.695837780149413,
"grad_norm": 0.75,
"learning_rate": 2.6765906775797465e-06,
"loss": 0.13059714436531067,
"step": 1304
},
{
"epoch": 0.6963713980789754,
"grad_norm": 0.7890625,
"learning_rate": 2.670662084974478e-06,
"loss": 0.13293643295764923,
"step": 1305
},
{
"epoch": 0.6969050160085379,
"grad_norm": 1.09375,
"learning_rate": 2.6647395830883542e-06,
"loss": 0.13132569193840027,
"step": 1306
},
{
"epoch": 0.6974386339381003,
"grad_norm": 1.0546875,
"learning_rate": 2.6588231927803995e-06,
"loss": 0.1320025473833084,
"step": 1307
},
{
"epoch": 0.6979722518676628,
"grad_norm": 1.921875,
"learning_rate": 2.6529129348881087e-06,
"loss": 0.12939175963401794,
"step": 1308
},
{
"epoch": 0.6985058697972252,
"grad_norm": 1.2734375,
"learning_rate": 2.6470088302273787e-06,
"loss": 0.12219016999006271,
"step": 1309
},
{
"epoch": 0.6990394877267876,
"grad_norm": 1.1328125,
"learning_rate": 2.641110899592435e-06,
"loss": 0.14250534772872925,
"step": 1310
},
{
"epoch": 0.69957310565635,
"grad_norm": 0.96875,
"learning_rate": 2.63521916375576e-06,
"loss": 0.12662728130817413,
"step": 1311
},
{
"epoch": 0.7001067235859125,
"grad_norm": 1.09375,
"learning_rate": 2.6293336434680163e-06,
"loss": 0.1324494332075119,
"step": 1312
},
{
"epoch": 0.7006403415154749,
"grad_norm": 0.88671875,
"learning_rate": 2.6234543594579756e-06,
"loss": 0.12897473573684692,
"step": 1313
},
{
"epoch": 0.7011739594450374,
"grad_norm": 2.5625,
"learning_rate": 2.6175813324324435e-06,
"loss": 0.12493091821670532,
"step": 1314
},
{
"epoch": 0.7017075773745998,
"grad_norm": 1.15625,
"learning_rate": 2.6117145830761944e-06,
"loss": 0.12946166098117828,
"step": 1315
},
{
"epoch": 0.7022411953041622,
"grad_norm": 0.68359375,
"learning_rate": 2.605854132051888e-06,
"loss": 0.11994671821594238,
"step": 1316
},
{
"epoch": 0.7027748132337247,
"grad_norm": 0.80859375,
"learning_rate": 2.600000000000001e-06,
"loss": 0.13359403610229492,
"step": 1317
},
{
"epoch": 0.7033084311632871,
"grad_norm": 0.96484375,
"learning_rate": 2.5941522075387566e-06,
"loss": 0.1256130337715149,
"step": 1318
},
{
"epoch": 0.7038420490928495,
"grad_norm": 0.74609375,
"learning_rate": 2.588310775264048e-06,
"loss": 0.13067632913589478,
"step": 1319
},
{
"epoch": 0.7043756670224119,
"grad_norm": 0.80859375,
"learning_rate": 2.582475723749373e-06,
"loss": 0.1421496570110321,
"step": 1320
},
{
"epoch": 0.7049092849519744,
"grad_norm": 1.7421875,
"learning_rate": 2.57664707354575e-06,
"loss": 0.11577002704143524,
"step": 1321
},
{
"epoch": 0.7054429028815368,
"grad_norm": 0.91015625,
"learning_rate": 2.5708248451816554e-06,
"loss": 0.14334681630134583,
"step": 1322
},
{
"epoch": 0.7059765208110993,
"grad_norm": 2.9375,
"learning_rate": 2.5650090591629454e-06,
"loss": 0.12631277740001678,
"step": 1323
},
{
"epoch": 0.7065101387406617,
"grad_norm": 1.1171875,
"learning_rate": 2.5591997359727915e-06,
"loss": 0.13312454521656036,
"step": 1324
},
{
"epoch": 0.7070437566702241,
"grad_norm": 0.86328125,
"learning_rate": 2.553396896071597e-06,
"loss": 0.1253337562084198,
"step": 1325
},
{
"epoch": 0.7075773745997865,
"grad_norm": 0.75390625,
"learning_rate": 2.547600559896937e-06,
"loss": 0.1347123235464096,
"step": 1326
},
{
"epoch": 0.708110992529349,
"grad_norm": 0.78515625,
"learning_rate": 2.541810747863473e-06,
"loss": 0.13334760069847107,
"step": 1327
},
{
"epoch": 0.7086446104589115,
"grad_norm": 1.2890625,
"learning_rate": 2.5360274803628965e-06,
"loss": 0.1294020414352417,
"step": 1328
},
{
"epoch": 0.7091782283884739,
"grad_norm": 1.71875,
"learning_rate": 2.5302507777638448e-06,
"loss": 0.12742722034454346,
"step": 1329
},
{
"epoch": 0.7097118463180363,
"grad_norm": 0.859375,
"learning_rate": 2.5244806604118324e-06,
"loss": 0.1390266716480255,
"step": 1330
},
{
"epoch": 0.7102454642475987,
"grad_norm": 1.3828125,
"learning_rate": 2.5187171486291867e-06,
"loss": 0.1451367884874344,
"step": 1331
},
{
"epoch": 0.7107790821771611,
"grad_norm": 1.328125,
"learning_rate": 2.5129602627149606e-06,
"loss": 0.1402292251586914,
"step": 1332
},
{
"epoch": 0.7113127001067235,
"grad_norm": 0.73828125,
"learning_rate": 2.507210022944881e-06,
"loss": 0.12348093092441559,
"step": 1333
},
{
"epoch": 0.7118463180362861,
"grad_norm": 1.0546875,
"learning_rate": 2.50146644957126e-06,
"loss": 0.13808563351631165,
"step": 1334
},
{
"epoch": 0.7123799359658485,
"grad_norm": 0.71875,
"learning_rate": 2.4957295628229365e-06,
"loss": 0.12552477419376373,
"step": 1335
},
{
"epoch": 0.7129135538954109,
"grad_norm": 1.90625,
"learning_rate": 2.48999938290519e-06,
"loss": 0.12992532551288605,
"step": 1336
},
{
"epoch": 0.7134471718249733,
"grad_norm": 1.53125,
"learning_rate": 2.48427592999969e-06,
"loss": 0.13032697141170502,
"step": 1337
},
{
"epoch": 0.7139807897545357,
"grad_norm": 2.8125,
"learning_rate": 2.4785592242644045e-06,
"loss": 0.13915517926216125,
"step": 1338
},
{
"epoch": 0.7145144076840982,
"grad_norm": 0.75390625,
"learning_rate": 2.4728492858335433e-06,
"loss": 0.11959775537252426,
"step": 1339
},
{
"epoch": 0.7150480256136607,
"grad_norm": 0.9609375,
"learning_rate": 2.467146134817479e-06,
"loss": 0.13928765058517456,
"step": 1340
},
{
"epoch": 0.7155816435432231,
"grad_norm": 0.875,
"learning_rate": 2.46144979130268e-06,
"loss": 0.11964480578899384,
"step": 1341
},
{
"epoch": 0.7161152614727855,
"grad_norm": 0.93359375,
"learning_rate": 2.455760275351639e-06,
"loss": 0.12196454405784607,
"step": 1342
},
{
"epoch": 0.7166488794023479,
"grad_norm": 2.40625,
"learning_rate": 2.4500776070027993e-06,
"loss": 0.14471834897994995,
"step": 1343
},
{
"epoch": 0.7171824973319103,
"grad_norm": 0.89453125,
"learning_rate": 2.444401806270494e-06,
"loss": 0.12269604951143265,
"step": 1344
},
{
"epoch": 0.7177161152614728,
"grad_norm": 1.5390625,
"learning_rate": 2.4387328931448573e-06,
"loss": 0.12529435753822327,
"step": 1345
},
{
"epoch": 0.7182497331910352,
"grad_norm": 1.7109375,
"learning_rate": 2.4330708875917775e-06,
"loss": 0.13360175490379333,
"step": 1346
},
{
"epoch": 0.7187833511205977,
"grad_norm": 1.1328125,
"learning_rate": 2.427415809552805e-06,
"loss": 0.14393943548202515,
"step": 1347
},
{
"epoch": 0.7193169690501601,
"grad_norm": 0.921875,
"learning_rate": 2.4217676789450983e-06,
"loss": 0.12287233769893646,
"step": 1348
},
{
"epoch": 0.7198505869797225,
"grad_norm": 0.8046875,
"learning_rate": 2.4161265156613423e-06,
"loss": 0.14484211802482605,
"step": 1349
},
{
"epoch": 0.7203842049092849,
"grad_norm": 0.80078125,
"learning_rate": 2.4104923395696846e-06,
"loss": 0.14987465739250183,
"step": 1350
},
{
"epoch": 0.7209178228388474,
"grad_norm": 0.8671875,
"learning_rate": 2.4048651705136633e-06,
"loss": 0.13077110052108765,
"step": 1351
},
{
"epoch": 0.7214514407684098,
"grad_norm": 0.84375,
"learning_rate": 2.399245028312142e-06,
"loss": 0.13799089193344116,
"step": 1352
},
{
"epoch": 0.7219850586979722,
"grad_norm": 1.1171875,
"learning_rate": 2.3936319327592287e-06,
"loss": 0.13428080081939697,
"step": 1353
},
{
"epoch": 0.7225186766275347,
"grad_norm": 0.625,
"learning_rate": 2.388025903624219e-06,
"loss": 0.13322536647319794,
"step": 1354
},
{
"epoch": 0.7230522945570971,
"grad_norm": 1.203125,
"learning_rate": 2.3824269606515176e-06,
"loss": 0.1374453604221344,
"step": 1355
},
{
"epoch": 0.7235859124866596,
"grad_norm": 1.6484375,
"learning_rate": 2.3768351235605705e-06,
"loss": 0.12982240319252014,
"step": 1356
},
{
"epoch": 0.724119530416222,
"grad_norm": 0.91796875,
"learning_rate": 2.3712504120458026e-06,
"loss": 0.1454874724149704,
"step": 1357
},
{
"epoch": 0.7246531483457844,
"grad_norm": 1.28125,
"learning_rate": 2.3656728457765366e-06,
"loss": 0.13524863123893738,
"step": 1358
},
{
"epoch": 0.7251867662753468,
"grad_norm": 1.1875,
"learning_rate": 2.360102444396933e-06,
"loss": 0.1337810754776001,
"step": 1359
},
{
"epoch": 0.7257203842049093,
"grad_norm": 1.078125,
"learning_rate": 2.3545392275259136e-06,
"loss": 0.1390707939863205,
"step": 1360
},
{
"epoch": 0.7262540021344717,
"grad_norm": 0.96875,
"learning_rate": 2.3489832147571032e-06,
"loss": 0.1369370073080063,
"step": 1361
},
{
"epoch": 0.7267876200640342,
"grad_norm": 0.9765625,
"learning_rate": 2.3434344256587465e-06,
"loss": 0.15307995676994324,
"step": 1362
},
{
"epoch": 0.7273212379935966,
"grad_norm": 1.34375,
"learning_rate": 2.3378928797736547e-06,
"loss": 0.135157510638237,
"step": 1363
},
{
"epoch": 0.727854855923159,
"grad_norm": 1.5078125,
"learning_rate": 2.332358596619118e-06,
"loss": 0.13804444670677185,
"step": 1364
},
{
"epoch": 0.7283884738527214,
"grad_norm": 0.9453125,
"learning_rate": 2.3268315956868563e-06,
"loss": 0.1507582664489746,
"step": 1365
},
{
"epoch": 0.7289220917822838,
"grad_norm": 2.03125,
"learning_rate": 2.321311896442938e-06,
"loss": 0.13264447450637817,
"step": 1366
},
{
"epoch": 0.7294557097118464,
"grad_norm": 1.671875,
"learning_rate": 2.3157995183277133e-06,
"loss": 0.13714554905891418,
"step": 1367
},
{
"epoch": 0.7299893276414088,
"grad_norm": 1.5078125,
"learning_rate": 2.3102944807557508e-06,
"loss": 0.12926441431045532,
"step": 1368
},
{
"epoch": 0.7305229455709712,
"grad_norm": 0.984375,
"learning_rate": 2.3047968031157626e-06,
"loss": 0.16150528192520142,
"step": 1369
},
{
"epoch": 0.7310565635005336,
"grad_norm": 0.94921875,
"learning_rate": 2.299306504770543e-06,
"loss": 0.1350216567516327,
"step": 1370
},
{
"epoch": 0.731590181430096,
"grad_norm": 1.859375,
"learning_rate": 2.293823605056892e-06,
"loss": 0.13649573922157288,
"step": 1371
},
{
"epoch": 0.7321237993596584,
"grad_norm": 1.640625,
"learning_rate": 2.2883481232855576e-06,
"loss": 0.13735274970531464,
"step": 1372
},
{
"epoch": 0.732657417289221,
"grad_norm": 2.28125,
"learning_rate": 2.2828800787411527e-06,
"loss": 0.1144658774137497,
"step": 1373
},
{
"epoch": 0.7331910352187834,
"grad_norm": 1.0,
"learning_rate": 2.277419490682106e-06,
"loss": 0.15584929287433624,
"step": 1374
},
{
"epoch": 0.7337246531483458,
"grad_norm": 1.4140625,
"learning_rate": 2.2719663783405778e-06,
"loss": 0.12320087850093842,
"step": 1375
},
{
"epoch": 0.7342582710779082,
"grad_norm": 0.80859375,
"learning_rate": 2.266520760922404e-06,
"loss": 0.1273914873600006,
"step": 1376
},
{
"epoch": 0.7347918890074706,
"grad_norm": 0.94921875,
"learning_rate": 2.2610826576070185e-06,
"loss": 0.1400269716978073,
"step": 1377
},
{
"epoch": 0.7353255069370331,
"grad_norm": 0.76953125,
"learning_rate": 2.2556520875473944e-06,
"loss": 0.13784019649028778,
"step": 1378
},
{
"epoch": 0.7358591248665955,
"grad_norm": 1.5,
"learning_rate": 2.2502290698699706e-06,
"loss": 0.12822571396827698,
"step": 1379
},
{
"epoch": 0.736392742796158,
"grad_norm": 1.1484375,
"learning_rate": 2.2448136236745856e-06,
"loss": 0.14050519466400146,
"step": 1380
},
{
"epoch": 0.7369263607257204,
"grad_norm": 0.8828125,
"learning_rate": 2.239405768034417e-06,
"loss": 0.12620662152767181,
"step": 1381
},
{
"epoch": 0.7374599786552828,
"grad_norm": 1.640625,
"learning_rate": 2.2340055219958987e-06,
"loss": 0.13421779870986938,
"step": 1382
},
{
"epoch": 0.7379935965848452,
"grad_norm": 0.76953125,
"learning_rate": 2.2286129045786734e-06,
"loss": 0.13223791122436523,
"step": 1383
},
{
"epoch": 0.7385272145144077,
"grad_norm": 1.0078125,
"learning_rate": 2.223227934775509e-06,
"loss": 0.13489994406700134,
"step": 1384
},
{
"epoch": 0.7390608324439701,
"grad_norm": 2.28125,
"learning_rate": 2.2178506315522436e-06,
"loss": 0.15167149901390076,
"step": 1385
},
{
"epoch": 0.7395944503735326,
"grad_norm": 0.84765625,
"learning_rate": 2.2124810138477098e-06,
"loss": 0.13452233374118805,
"step": 1386
},
{
"epoch": 0.740128068303095,
"grad_norm": 1.1640625,
"learning_rate": 2.2071191005736737e-06,
"loss": 0.13255596160888672,
"step": 1387
},
{
"epoch": 0.7406616862326574,
"grad_norm": 1.015625,
"learning_rate": 2.2017649106147645e-06,
"loss": 0.12334451824426651,
"step": 1388
},
{
"epoch": 0.7411953041622199,
"grad_norm": 3.125,
"learning_rate": 2.196418462828413e-06,
"loss": 0.12104532122612,
"step": 1389
},
{
"epoch": 0.7417289220917823,
"grad_norm": 1.140625,
"learning_rate": 2.1910797760447813e-06,
"loss": 0.12886758148670197,
"step": 1390
},
{
"epoch": 0.7422625400213447,
"grad_norm": 1.265625,
"learning_rate": 2.185748869066695e-06,
"loss": 0.13216081261634827,
"step": 1391
},
{
"epoch": 0.7427961579509071,
"grad_norm": 12.125,
"learning_rate": 2.180425760669581e-06,
"loss": 0.12087278813123703,
"step": 1392
},
{
"epoch": 0.7433297758804696,
"grad_norm": 0.91796875,
"learning_rate": 2.175110469601398e-06,
"loss": 0.12328571081161499,
"step": 1393
},
{
"epoch": 0.743863393810032,
"grad_norm": 5.25,
"learning_rate": 2.169803014582576e-06,
"loss": 0.136301189661026,
"step": 1394
},
{
"epoch": 0.7443970117395945,
"grad_norm": 1.4609375,
"learning_rate": 2.164503414305943e-06,
"loss": 0.1343958079814911,
"step": 1395
},
{
"epoch": 0.7449306296691569,
"grad_norm": 0.73046875,
"learning_rate": 2.159211687436664e-06,
"loss": 0.12515291571617126,
"step": 1396
},
{
"epoch": 0.7454642475987193,
"grad_norm": 1.0625,
"learning_rate": 2.153927852612172e-06,
"loss": 0.14262662827968597,
"step": 1397
},
{
"epoch": 0.7459978655282817,
"grad_norm": 1.015625,
"learning_rate": 2.1486519284421075e-06,
"loss": 0.14405196905136108,
"step": 1398
},
{
"epoch": 0.7465314834578441,
"grad_norm": 0.796875,
"learning_rate": 2.1433839335082466e-06,
"loss": 0.13885879516601562,
"step": 1399
},
{
"epoch": 0.7470651013874067,
"grad_norm": 0.94921875,
"learning_rate": 2.1381238863644435e-06,
"loss": 0.12077438831329346,
"step": 1400
},
{
"epoch": 0.7475987193169691,
"grad_norm": 1.3671875,
"learning_rate": 2.132871805536551e-06,
"loss": 0.13963539898395538,
"step": 1401
},
{
"epoch": 0.7481323372465315,
"grad_norm": 0.59375,
"learning_rate": 2.127627709522375e-06,
"loss": 0.13025550544261932,
"step": 1402
},
{
"epoch": 0.7486659551760939,
"grad_norm": 0.91015625,
"learning_rate": 2.122391616791594e-06,
"loss": 0.12093956768512726,
"step": 1403
},
{
"epoch": 0.7491995731056563,
"grad_norm": 0.79296875,
"learning_rate": 2.117163545785698e-06,
"loss": 0.1370830237865448,
"step": 1404
},
{
"epoch": 0.7497331910352187,
"grad_norm": 0.9140625,
"learning_rate": 2.111943514917926e-06,
"loss": 0.1320624053478241,
"step": 1405
},
{
"epoch": 0.7502668089647813,
"grad_norm": 0.9296875,
"learning_rate": 2.1067315425732e-06,
"loss": 0.1295088678598404,
"step": 1406
},
{
"epoch": 0.7508004268943437,
"grad_norm": 1.1015625,
"learning_rate": 2.1015276471080595e-06,
"loss": 0.12421064078807831,
"step": 1407
},
{
"epoch": 0.7513340448239061,
"grad_norm": 0.74609375,
"learning_rate": 2.0963318468505967e-06,
"loss": 0.13148941099643707,
"step": 1408
},
{
"epoch": 0.7518676627534685,
"grad_norm": 4.0625,
"learning_rate": 2.091144160100396e-06,
"loss": 0.1256633698940277,
"step": 1409
},
{
"epoch": 0.7524012806830309,
"grad_norm": 0.9140625,
"learning_rate": 2.085964605128458e-06,
"loss": 0.13601458072662354,
"step": 1410
},
{
"epoch": 0.7529348986125933,
"grad_norm": 1.171875,
"learning_rate": 2.080793200177153e-06,
"loss": 0.1359083652496338,
"step": 1411
},
{
"epoch": 0.7534685165421559,
"grad_norm": 0.9140625,
"learning_rate": 2.0756299634601392e-06,
"loss": 0.1323002576828003,
"step": 1412
},
{
"epoch": 0.7540021344717183,
"grad_norm": 2.25,
"learning_rate": 2.070474913162313e-06,
"loss": 0.13984021544456482,
"step": 1413
},
{
"epoch": 0.7545357524012807,
"grad_norm": 1.3515625,
"learning_rate": 2.0653280674397306e-06,
"loss": 0.1385597437620163,
"step": 1414
},
{
"epoch": 0.7550693703308431,
"grad_norm": 0.9609375,
"learning_rate": 2.060189444419559e-06,
"loss": 0.12163829058408737,
"step": 1415
},
{
"epoch": 0.7556029882604055,
"grad_norm": 0.79296875,
"learning_rate": 2.0550590622000005e-06,
"loss": 0.1338130086660385,
"step": 1416
},
{
"epoch": 0.756136606189968,
"grad_norm": 4.96875,
"learning_rate": 2.0499369388502335e-06,
"loss": 0.13578353822231293,
"step": 1417
},
{
"epoch": 0.7566702241195304,
"grad_norm": 1.5234375,
"learning_rate": 2.044823092410354e-06,
"loss": 0.1328636258840561,
"step": 1418
},
{
"epoch": 0.7572038420490929,
"grad_norm": 0.67578125,
"learning_rate": 2.0397175408912966e-06,
"loss": 0.13831090927124023,
"step": 1419
},
{
"epoch": 0.7577374599786553,
"grad_norm": 0.9453125,
"learning_rate": 2.034620302274791e-06,
"loss": 0.12291224300861359,
"step": 1420
},
{
"epoch": 0.7582710779082177,
"grad_norm": 1.4453125,
"learning_rate": 2.029531394513284e-06,
"loss": 0.1286519169807434,
"step": 1421
},
{
"epoch": 0.7588046958377801,
"grad_norm": 2.359375,
"learning_rate": 2.0244508355298842e-06,
"loss": 0.1528690755367279,
"step": 1422
},
{
"epoch": 0.7593383137673426,
"grad_norm": 1.1640625,
"learning_rate": 2.0193786432182916e-06,
"loss": 0.13557785749435425,
"step": 1423
},
{
"epoch": 0.759871931696905,
"grad_norm": 1.34375,
"learning_rate": 2.014314835442742e-06,
"loss": 0.13548403978347778,
"step": 1424
},
{
"epoch": 0.7604055496264674,
"grad_norm": 1.5234375,
"learning_rate": 2.0092594300379386e-06,
"loss": 0.12891140580177307,
"step": 1425
},
{
"epoch": 0.7609391675560299,
"grad_norm": 1.9609375,
"learning_rate": 2.004212444808996e-06,
"loss": 0.13795675337314606,
"step": 1426
},
{
"epoch": 0.7614727854855923,
"grad_norm": 1.3125,
"learning_rate": 1.9991738975313683e-06,
"loss": 0.1275990605354309,
"step": 1427
},
{
"epoch": 0.7620064034151548,
"grad_norm": 0.87109375,
"learning_rate": 1.994143805950793e-06,
"loss": 0.11925936490297318,
"step": 1428
},
{
"epoch": 0.7625400213447172,
"grad_norm": 4.125,
"learning_rate": 1.989122187783226e-06,
"loss": 0.13084454834461212,
"step": 1429
},
{
"epoch": 0.7630736392742796,
"grad_norm": 2.640625,
"learning_rate": 1.9841090607147794e-06,
"loss": 0.14653316140174866,
"step": 1430
},
{
"epoch": 0.763607257203842,
"grad_norm": 1.25,
"learning_rate": 1.9791044424016627e-06,
"loss": 0.13554765284061432,
"step": 1431
},
{
"epoch": 0.7641408751334045,
"grad_norm": 1.0546875,
"learning_rate": 1.9741083504701154e-06,
"loss": 0.13005059957504272,
"step": 1432
},
{
"epoch": 0.7646744930629669,
"grad_norm": 2.421875,
"learning_rate": 1.9691208025163464e-06,
"loss": 0.13782243430614471,
"step": 1433
},
{
"epoch": 0.7652081109925294,
"grad_norm": 0.86328125,
"learning_rate": 1.964141816106473e-06,
"loss": 0.11705458164215088,
"step": 1434
},
{
"epoch": 0.7657417289220918,
"grad_norm": 1.5859375,
"learning_rate": 1.959171408776462e-06,
"loss": 0.12835600972175598,
"step": 1435
},
{
"epoch": 0.7662753468516542,
"grad_norm": 1.1875,
"learning_rate": 1.95420959803206e-06,
"loss": 0.13009797036647797,
"step": 1436
},
{
"epoch": 0.7668089647812166,
"grad_norm": 1.1640625,
"learning_rate": 1.9492564013487394e-06,
"loss": 0.12874095141887665,
"step": 1437
},
{
"epoch": 0.767342582710779,
"grad_norm": 0.71875,
"learning_rate": 1.9443118361716325e-06,
"loss": 0.13046720623970032,
"step": 1438
},
{
"epoch": 0.7678762006403416,
"grad_norm": 1.1171875,
"learning_rate": 1.9393759199154736e-06,
"loss": 0.12460866570472717,
"step": 1439
},
{
"epoch": 0.768409818569904,
"grad_norm": 0.828125,
"learning_rate": 1.934448669964533e-06,
"loss": 0.14308089017868042,
"step": 1440
},
{
"epoch": 0.7689434364994664,
"grad_norm": 1.265625,
"learning_rate": 1.9295301036725594e-06,
"loss": 0.13541656732559204,
"step": 1441
},
{
"epoch": 0.7694770544290288,
"grad_norm": 0.69140625,
"learning_rate": 1.9246202383627182e-06,
"loss": 0.1290501058101654,
"step": 1442
},
{
"epoch": 0.7700106723585912,
"grad_norm": 0.91015625,
"learning_rate": 1.919719091327528e-06,
"loss": 0.14724613726139069,
"step": 1443
},
{
"epoch": 0.7705442902881536,
"grad_norm": 1.6015625,
"learning_rate": 1.914826679828806e-06,
"loss": 0.12952086329460144,
"step": 1444
},
{
"epoch": 0.7710779082177162,
"grad_norm": 1.0390625,
"learning_rate": 1.909943021097597e-06,
"loss": 0.15102215111255646,
"step": 1445
},
{
"epoch": 0.7716115261472786,
"grad_norm": 0.8125,
"learning_rate": 1.9050681323341253e-06,
"loss": 0.1296108067035675,
"step": 1446
},
{
"epoch": 0.772145144076841,
"grad_norm": 1.9609375,
"learning_rate": 1.9002020307077193e-06,
"loss": 0.133722722530365,
"step": 1447
},
{
"epoch": 0.7726787620064034,
"grad_norm": 1.4921875,
"learning_rate": 1.8953447333567674e-06,
"loss": 0.13981425762176514,
"step": 1448
},
{
"epoch": 0.7732123799359658,
"grad_norm": 2.203125,
"learning_rate": 1.8904962573886437e-06,
"loss": 0.1374809443950653,
"step": 1449
},
{
"epoch": 0.7737459978655283,
"grad_norm": 1.59375,
"learning_rate": 1.8856566198796559e-06,
"loss": 0.1250973641872406,
"step": 1450
},
{
"epoch": 0.7742796157950907,
"grad_norm": 0.79296875,
"learning_rate": 1.8808258378749813e-06,
"loss": 0.136212557554245,
"step": 1451
},
{
"epoch": 0.7748132337246532,
"grad_norm": 0.765625,
"learning_rate": 1.8760039283886106e-06,
"loss": 0.1279085874557495,
"step": 1452
},
{
"epoch": 0.7753468516542156,
"grad_norm": 0.81640625,
"learning_rate": 1.8711909084032832e-06,
"loss": 0.13408775627613068,
"step": 1453
},
{
"epoch": 0.775880469583778,
"grad_norm": 0.7734375,
"learning_rate": 1.8663867948704293e-06,
"loss": 0.13363221287727356,
"step": 1454
},
{
"epoch": 0.7764140875133404,
"grad_norm": 0.83984375,
"learning_rate": 1.861591604710115e-06,
"loss": 0.12866361439228058,
"step": 1455
},
{
"epoch": 0.7769477054429029,
"grad_norm": 1.546875,
"learning_rate": 1.856805354810971e-06,
"loss": 0.12831313908100128,
"step": 1456
},
{
"epoch": 0.7774813233724653,
"grad_norm": 0.7734375,
"learning_rate": 1.8520280620301473e-06,
"loss": 0.13136044144630432,
"step": 1457
},
{
"epoch": 0.7780149413020278,
"grad_norm": 1.3125,
"learning_rate": 1.8472597431932428e-06,
"loss": 0.131224125623703,
"step": 1458
},
{
"epoch": 0.7785485592315902,
"grad_norm": 1.109375,
"learning_rate": 1.8425004150942545e-06,
"loss": 0.13651737570762634,
"step": 1459
},
{
"epoch": 0.7790821771611526,
"grad_norm": 1.2109375,
"learning_rate": 1.8377500944955067e-06,
"loss": 0.14164122939109802,
"step": 1460
},
{
"epoch": 0.7796157950907151,
"grad_norm": 0.76953125,
"learning_rate": 1.8330087981276072e-06,
"loss": 0.1353672742843628,
"step": 1461
},
{
"epoch": 0.7801494130202775,
"grad_norm": 1.0234375,
"learning_rate": 1.8282765426893766e-06,
"loss": 0.1324540376663208,
"step": 1462
},
{
"epoch": 0.7806830309498399,
"grad_norm": 1.65625,
"learning_rate": 1.823553344847792e-06,
"loss": 0.13273343443870544,
"step": 1463
},
{
"epoch": 0.7812166488794023,
"grad_norm": 1.390625,
"learning_rate": 1.8188392212379344e-06,
"loss": 0.12248252332210541,
"step": 1464
},
{
"epoch": 0.7817502668089648,
"grad_norm": 0.8515625,
"learning_rate": 1.8141341884629216e-06,
"loss": 0.13564269244670868,
"step": 1465
},
{
"epoch": 0.7822838847385272,
"grad_norm": 1.421875,
"learning_rate": 1.8094382630938542e-06,
"loss": 0.13905777037143707,
"step": 1466
},
{
"epoch": 0.7828175026680897,
"grad_norm": 3.0625,
"learning_rate": 1.804751461669756e-06,
"loss": 0.13549892604351044,
"step": 1467
},
{
"epoch": 0.7833511205976521,
"grad_norm": 1.0546875,
"learning_rate": 1.800073800697521e-06,
"loss": 0.14187997579574585,
"step": 1468
},
{
"epoch": 0.7838847385272145,
"grad_norm": 0.8046875,
"learning_rate": 1.7954052966518413e-06,
"loss": 0.12620186805725098,
"step": 1469
},
{
"epoch": 0.7844183564567769,
"grad_norm": 0.8203125,
"learning_rate": 1.790745965975169e-06,
"loss": 0.14258602261543274,
"step": 1470
},
{
"epoch": 0.7849519743863393,
"grad_norm": 1.0625,
"learning_rate": 1.7860958250776381e-06,
"loss": 0.1372625231742859,
"step": 1471
},
{
"epoch": 0.7854855923159018,
"grad_norm": 1.421875,
"learning_rate": 1.7814548903370245e-06,
"loss": 0.12988245487213135,
"step": 1472
},
{
"epoch": 0.7860192102454643,
"grad_norm": 1.1796875,
"learning_rate": 1.7768231780986746e-06,
"loss": 0.14955168962478638,
"step": 1473
},
{
"epoch": 0.7865528281750267,
"grad_norm": 1.1171875,
"learning_rate": 1.772200704675455e-06,
"loss": 0.1341627836227417,
"step": 1474
},
{
"epoch": 0.7870864461045891,
"grad_norm": 1.53125,
"learning_rate": 1.767587486347693e-06,
"loss": 0.1261909455060959,
"step": 1475
},
{
"epoch": 0.7876200640341515,
"grad_norm": 0.78515625,
"learning_rate": 1.7629835393631177e-06,
"loss": 0.12421926856040955,
"step": 1476
},
{
"epoch": 0.7881536819637139,
"grad_norm": 0.98828125,
"learning_rate": 1.75838887993681e-06,
"loss": 0.13347256183624268,
"step": 1477
},
{
"epoch": 0.7886872998932765,
"grad_norm": 0.91015625,
"learning_rate": 1.7538035242511344e-06,
"loss": 0.13310089707374573,
"step": 1478
},
{
"epoch": 0.7892209178228389,
"grad_norm": 0.77734375,
"learning_rate": 1.7492274884556903e-06,
"loss": 0.13861967623233795,
"step": 1479
},
{
"epoch": 0.7897545357524013,
"grad_norm": 3.0,
"learning_rate": 1.7446607886672507e-06,
"loss": 0.1375465989112854,
"step": 1480
},
{
"epoch": 0.7902881536819637,
"grad_norm": 1.1171875,
"learning_rate": 1.740103440969711e-06,
"loss": 0.12484005093574524,
"step": 1481
},
{
"epoch": 0.7908217716115261,
"grad_norm": 0.87890625,
"learning_rate": 1.735555461414024e-06,
"loss": 0.13937249779701233,
"step": 1482
},
{
"epoch": 0.7913553895410885,
"grad_norm": 1.203125,
"learning_rate": 1.7310168660181513e-06,
"loss": 0.14172214269638062,
"step": 1483
},
{
"epoch": 0.791889007470651,
"grad_norm": 0.6875,
"learning_rate": 1.7264876707670007e-06,
"loss": 0.13021104037761688,
"step": 1484
},
{
"epoch": 0.7924226254002135,
"grad_norm": 1.6484375,
"learning_rate": 1.7219678916123763e-06,
"loss": 0.11556795984506607,
"step": 1485
},
{
"epoch": 0.7929562433297759,
"grad_norm": 1.7265625,
"learning_rate": 1.717457544472916e-06,
"loss": 0.12306451052427292,
"step": 1486
},
{
"epoch": 0.7934898612593383,
"grad_norm": 0.86328125,
"learning_rate": 1.7129566452340387e-06,
"loss": 0.1400870978832245,
"step": 1487
},
{
"epoch": 0.7940234791889007,
"grad_norm": 1.0078125,
"learning_rate": 1.7084652097478887e-06,
"loss": 0.1360422968864441,
"step": 1488
},
{
"epoch": 0.7945570971184632,
"grad_norm": 1.6640625,
"learning_rate": 1.7039832538332765e-06,
"loss": 0.14386096596717834,
"step": 1489
},
{
"epoch": 0.7950907150480256,
"grad_norm": 0.80078125,
"learning_rate": 1.6995107932756305e-06,
"loss": 0.13437721133232117,
"step": 1490
},
{
"epoch": 0.7956243329775881,
"grad_norm": 1.90625,
"learning_rate": 1.695047843826932e-06,
"loss": 0.1314612478017807,
"step": 1491
},
{
"epoch": 0.7961579509071505,
"grad_norm": 1.1484375,
"learning_rate": 1.6905944212056653e-06,
"loss": 0.12197023630142212,
"step": 1492
},
{
"epoch": 0.7966915688367129,
"grad_norm": 3.140625,
"learning_rate": 1.6861505410967618e-06,
"loss": 0.14306682348251343,
"step": 1493
},
{
"epoch": 0.7972251867662753,
"grad_norm": 0.81640625,
"learning_rate": 1.6817162191515454e-06,
"loss": 0.12597501277923584,
"step": 1494
},
{
"epoch": 0.7977588046958378,
"grad_norm": 0.7421875,
"learning_rate": 1.677291470987673e-06,
"loss": 0.1234937310218811,
"step": 1495
},
{
"epoch": 0.7982924226254002,
"grad_norm": 0.703125,
"learning_rate": 1.6728763121890884e-06,
"loss": 0.1416109800338745,
"step": 1496
},
{
"epoch": 0.7988260405549626,
"grad_norm": 0.828125,
"learning_rate": 1.6684707583059528e-06,
"loss": 0.14253093302249908,
"step": 1497
},
{
"epoch": 0.7993596584845251,
"grad_norm": 1.640625,
"learning_rate": 1.6640748248546076e-06,
"loss": 0.12052501738071442,
"step": 1498
},
{
"epoch": 0.7998932764140875,
"grad_norm": 0.87109375,
"learning_rate": 1.6596885273175064e-06,
"loss": 0.1342918872833252,
"step": 1499
},
{
"epoch": 0.80042689434365,
"grad_norm": 4.40625,
"learning_rate": 1.6553118811431656e-06,
"loss": 0.12489061057567596,
"step": 1500
},
{
"epoch": 0.8009605122732124,
"grad_norm": 0.76171875,
"learning_rate": 1.6509449017461137e-06,
"loss": 0.1385369896888733,
"step": 1501
},
{
"epoch": 0.8014941302027748,
"grad_norm": 0.890625,
"learning_rate": 1.6465876045068242e-06,
"loss": 0.13919919729232788,
"step": 1502
},
{
"epoch": 0.8020277481323372,
"grad_norm": 1.125,
"learning_rate": 1.6422400047716796e-06,
"loss": 0.13582268357276917,
"step": 1503
},
{
"epoch": 0.8025613660618997,
"grad_norm": 0.93359375,
"learning_rate": 1.6379021178529007e-06,
"loss": 0.13229192793369293,
"step": 1504
},
{
"epoch": 0.8030949839914621,
"grad_norm": 0.84765625,
"learning_rate": 1.633573959028506e-06,
"loss": 0.13842912018299103,
"step": 1505
},
{
"epoch": 0.8036286019210246,
"grad_norm": 0.83203125,
"learning_rate": 1.6292555435422438e-06,
"loss": 0.13777904212474823,
"step": 1506
},
{
"epoch": 0.804162219850587,
"grad_norm": 0.98046875,
"learning_rate": 1.6249468866035554e-06,
"loss": 0.13319657742977142,
"step": 1507
},
{
"epoch": 0.8046958377801494,
"grad_norm": 1.296875,
"learning_rate": 1.6206480033875046e-06,
"loss": 0.1248672753572464,
"step": 1508
},
{
"epoch": 0.8052294557097118,
"grad_norm": 1.71875,
"learning_rate": 1.6163589090347387e-06,
"loss": 0.12971623241901398,
"step": 1509
},
{
"epoch": 0.8057630736392742,
"grad_norm": 0.78125,
"learning_rate": 1.612079618651424e-06,
"loss": 0.14820817112922668,
"step": 1510
},
{
"epoch": 0.8062966915688368,
"grad_norm": 0.97265625,
"learning_rate": 1.6078101473092e-06,
"loss": 0.13331134617328644,
"step": 1511
},
{
"epoch": 0.8068303094983992,
"grad_norm": 0.90625,
"learning_rate": 1.603550510045122e-06,
"loss": 0.12173394858837128,
"step": 1512
},
{
"epoch": 0.8073639274279616,
"grad_norm": 0.8671875,
"learning_rate": 1.5993007218616085e-06,
"loss": 0.13103632628917694,
"step": 1513
},
{
"epoch": 0.807897545357524,
"grad_norm": 0.78515625,
"learning_rate": 1.5950607977263957e-06,
"loss": 0.14954793453216553,
"step": 1514
},
{
"epoch": 0.8084311632870864,
"grad_norm": 0.875,
"learning_rate": 1.5908307525724698e-06,
"loss": 0.13069681823253632,
"step": 1515
},
{
"epoch": 0.8089647812166488,
"grad_norm": 1.0078125,
"learning_rate": 1.5866106012980307e-06,
"loss": 0.13540677726268768,
"step": 1516
},
{
"epoch": 0.8094983991462114,
"grad_norm": 1.0625,
"learning_rate": 1.582400358766428e-06,
"loss": 0.13573959469795227,
"step": 1517
},
{
"epoch": 0.8100320170757738,
"grad_norm": 0.83984375,
"learning_rate": 1.5782000398061145e-06,
"loss": 0.12955842912197113,
"step": 1518
},
{
"epoch": 0.8105656350053362,
"grad_norm": 2.390625,
"learning_rate": 1.5740096592105924e-06,
"loss": 0.13925619423389435,
"step": 1519
},
{
"epoch": 0.8110992529348986,
"grad_norm": 1.953125,
"learning_rate": 1.5698292317383595e-06,
"loss": 0.13408657908439636,
"step": 1520
},
{
"epoch": 0.811632870864461,
"grad_norm": 1.1640625,
"learning_rate": 1.5656587721128585e-06,
"loss": 0.12406972795724869,
"step": 1521
},
{
"epoch": 0.8121664887940235,
"grad_norm": 1.234375,
"learning_rate": 1.5614982950224293e-06,
"loss": 0.1366141438484192,
"step": 1522
},
{
"epoch": 0.812700106723586,
"grad_norm": 0.82421875,
"learning_rate": 1.5573478151202484e-06,
"loss": 0.14020280539989471,
"step": 1523
},
{
"epoch": 0.8132337246531484,
"grad_norm": 1.625,
"learning_rate": 1.5532073470242856e-06,
"loss": 0.13411226868629456,
"step": 1524
},
{
"epoch": 0.8137673425827108,
"grad_norm": 0.8125,
"learning_rate": 1.5490769053172472e-06,
"loss": 0.14651578664779663,
"step": 1525
},
{
"epoch": 0.8143009605122732,
"grad_norm": 0.77734375,
"learning_rate": 1.544956504546527e-06,
"loss": 0.1234428882598877,
"step": 1526
},
{
"epoch": 0.8148345784418356,
"grad_norm": 1.421875,
"learning_rate": 1.5408461592241567e-06,
"loss": 0.12729822099208832,
"step": 1527
},
{
"epoch": 0.8153681963713981,
"grad_norm": 3.03125,
"learning_rate": 1.5367458838267507e-06,
"loss": 0.13210302591323853,
"step": 1528
},
{
"epoch": 0.8159018143009605,
"grad_norm": 0.765625,
"learning_rate": 1.5326556927954566e-06,
"loss": 0.1251302808523178,
"step": 1529
},
{
"epoch": 0.816435432230523,
"grad_norm": 1.171875,
"learning_rate": 1.5285756005359064e-06,
"loss": 0.1211385577917099,
"step": 1530
},
{
"epoch": 0.8169690501600854,
"grad_norm": 0.70703125,
"learning_rate": 1.5245056214181644e-06,
"loss": 0.132429301738739,
"step": 1531
},
{
"epoch": 0.8175026680896478,
"grad_norm": 0.91015625,
"learning_rate": 1.5204457697766744e-06,
"loss": 0.13539186120033264,
"step": 1532
},
{
"epoch": 0.8180362860192102,
"grad_norm": 0.8515625,
"learning_rate": 1.516396059910216e-06,
"loss": 0.12651708722114563,
"step": 1533
},
{
"epoch": 0.8185699039487727,
"grad_norm": 0.64453125,
"learning_rate": 1.5123565060818421e-06,
"loss": 0.1310691088438034,
"step": 1534
},
{
"epoch": 0.8191035218783351,
"grad_norm": 1.1796875,
"learning_rate": 1.5083271225188434e-06,
"loss": 0.1312372386455536,
"step": 1535
},
{
"epoch": 0.8196371398078975,
"grad_norm": 0.77734375,
"learning_rate": 1.5043079234126874e-06,
"loss": 0.11605694890022278,
"step": 1536
},
{
"epoch": 0.82017075773746,
"grad_norm": 0.7578125,
"learning_rate": 1.500298922918971e-06,
"loss": 0.13226571679115295,
"step": 1537
},
{
"epoch": 0.8207043756670224,
"grad_norm": 1.1015625,
"learning_rate": 1.4963001351573744e-06,
"loss": 0.1395004242658615,
"step": 1538
},
{
"epoch": 0.8212379935965849,
"grad_norm": 1.015625,
"learning_rate": 1.4923115742116061e-06,
"loss": 0.14467453956604004,
"step": 1539
},
{
"epoch": 0.8217716115261473,
"grad_norm": 0.84375,
"learning_rate": 1.4883332541293588e-06,
"loss": 0.1377476155757904,
"step": 1540
},
{
"epoch": 0.8223052294557097,
"grad_norm": 1.4296875,
"learning_rate": 1.484365188922253e-06,
"loss": 0.13889798521995544,
"step": 1541
},
{
"epoch": 0.8228388473852721,
"grad_norm": 0.88671875,
"learning_rate": 1.4804073925657966e-06,
"loss": 0.13588564097881317,
"step": 1542
},
{
"epoch": 0.8233724653148345,
"grad_norm": 1.046875,
"learning_rate": 1.4764598789993245e-06,
"loss": 0.1392589509487152,
"step": 1543
},
{
"epoch": 0.823906083244397,
"grad_norm": 1.109375,
"learning_rate": 1.472522662125961e-06,
"loss": 0.13790416717529297,
"step": 1544
},
{
"epoch": 0.8244397011739595,
"grad_norm": 0.93359375,
"learning_rate": 1.4685957558125625e-06,
"loss": 0.1263730525970459,
"step": 1545
},
{
"epoch": 0.8249733191035219,
"grad_norm": 0.83984375,
"learning_rate": 1.4646791738896737e-06,
"loss": 0.14427456259727478,
"step": 1546
},
{
"epoch": 0.8255069370330843,
"grad_norm": 0.84375,
"learning_rate": 1.4607729301514757e-06,
"loss": 0.13653981685638428,
"step": 1547
},
{
"epoch": 0.8260405549626467,
"grad_norm": 0.71875,
"learning_rate": 1.4568770383557376e-06,
"loss": 0.12189627438783646,
"step": 1548
},
{
"epoch": 0.8265741728922091,
"grad_norm": 1.7578125,
"learning_rate": 1.4529915122237713e-06,
"loss": 0.13752180337905884,
"step": 1549
},
{
"epoch": 0.8271077908217717,
"grad_norm": 1.015625,
"learning_rate": 1.4491163654403775e-06,
"loss": 0.1484394073486328,
"step": 1550
},
{
"epoch": 0.8276414087513341,
"grad_norm": 1.546875,
"learning_rate": 1.445251611653806e-06,
"loss": 0.13767069578170776,
"step": 1551
},
{
"epoch": 0.8281750266808965,
"grad_norm": 0.83203125,
"learning_rate": 1.4413972644756966e-06,
"loss": 0.14050151407718658,
"step": 1552
},
{
"epoch": 0.8287086446104589,
"grad_norm": 0.7421875,
"learning_rate": 1.4375533374810425e-06,
"loss": 0.12382489442825317,
"step": 1553
},
{
"epoch": 0.8292422625400213,
"grad_norm": 1.1875,
"learning_rate": 1.433719844208132e-06,
"loss": 0.1450960338115692,
"step": 1554
},
{
"epoch": 0.8297758804695837,
"grad_norm": 0.79296875,
"learning_rate": 1.429896798158511e-06,
"loss": 0.12364508956670761,
"step": 1555
},
{
"epoch": 0.8303094983991463,
"grad_norm": 1.4375,
"learning_rate": 1.4260842127969275e-06,
"loss": 0.12293899804353714,
"step": 1556
},
{
"epoch": 0.8308431163287087,
"grad_norm": 0.94921875,
"learning_rate": 1.4222821015512856e-06,
"loss": 0.13613364100456238,
"step": 1557
},
{
"epoch": 0.8313767342582711,
"grad_norm": 1.6875,
"learning_rate": 1.418490477812602e-06,
"loss": 0.13770051300525665,
"step": 1558
},
{
"epoch": 0.8319103521878335,
"grad_norm": 1.046875,
"learning_rate": 1.4147093549349573e-06,
"loss": 0.1287848800420761,
"step": 1559
},
{
"epoch": 0.8324439701173959,
"grad_norm": 1.03125,
"learning_rate": 1.4109387462354454e-06,
"loss": 0.15285789966583252,
"step": 1560
},
{
"epoch": 0.8329775880469584,
"grad_norm": 1.328125,
"learning_rate": 1.4071786649941303e-06,
"loss": 0.1319538950920105,
"step": 1561
},
{
"epoch": 0.8335112059765208,
"grad_norm": 0.69140625,
"learning_rate": 1.403429124453999e-06,
"loss": 0.13582220673561096,
"step": 1562
},
{
"epoch": 0.8340448239060833,
"grad_norm": 1.578125,
"learning_rate": 1.399690137820913e-06,
"loss": 0.1222437173128128,
"step": 1563
},
{
"epoch": 0.8345784418356457,
"grad_norm": 1.1953125,
"learning_rate": 1.3959617182635652e-06,
"loss": 0.13752412796020508,
"step": 1564
},
{
"epoch": 0.8351120597652081,
"grad_norm": 1.0546875,
"learning_rate": 1.3922438789134294e-06,
"loss": 0.13985900580883026,
"step": 1565
},
{
"epoch": 0.8356456776947705,
"grad_norm": 0.953125,
"learning_rate": 1.3885366328647158e-06,
"loss": 0.1466849148273468,
"step": 1566
},
{
"epoch": 0.836179295624333,
"grad_norm": 0.9921875,
"learning_rate": 1.3848399931743257e-06,
"loss": 0.12854596972465515,
"step": 1567
},
{
"epoch": 0.8367129135538954,
"grad_norm": 15.625,
"learning_rate": 1.3811539728618057e-06,
"loss": 0.14100824296474457,
"step": 1568
},
{
"epoch": 0.8372465314834578,
"grad_norm": 1.640625,
"learning_rate": 1.3774785849092996e-06,
"loss": 0.12274054437875748,
"step": 1569
},
{
"epoch": 0.8377801494130203,
"grad_norm": 0.93359375,
"learning_rate": 1.3738138422615065e-06,
"loss": 0.12491217255592346,
"step": 1570
},
{
"epoch": 0.8383137673425827,
"grad_norm": 1.125,
"learning_rate": 1.3701597578256272e-06,
"loss": 0.13382726907730103,
"step": 1571
},
{
"epoch": 0.8388473852721452,
"grad_norm": 1.1015625,
"learning_rate": 1.366516344471331e-06,
"loss": 0.12828874588012695,
"step": 1572
},
{
"epoch": 0.8393810032017076,
"grad_norm": 0.875,
"learning_rate": 1.3628836150307006e-06,
"loss": 0.14455264806747437,
"step": 1573
},
{
"epoch": 0.83991462113127,
"grad_norm": 1.2890625,
"learning_rate": 1.3592615822981889e-06,
"loss": 0.12032708525657654,
"step": 1574
},
{
"epoch": 0.8404482390608324,
"grad_norm": 0.74609375,
"learning_rate": 1.3556502590305776e-06,
"loss": 0.11596512794494629,
"step": 1575
},
{
"epoch": 0.8409818569903948,
"grad_norm": 0.76953125,
"learning_rate": 1.3520496579469267e-06,
"loss": 0.12437745183706284,
"step": 1576
},
{
"epoch": 0.8415154749199573,
"grad_norm": 1.140625,
"learning_rate": 1.3484597917285365e-06,
"loss": 0.12841296195983887,
"step": 1577
},
{
"epoch": 0.8420490928495198,
"grad_norm": 1.2890625,
"learning_rate": 1.3448806730188964e-06,
"loss": 0.13422322273254395,
"step": 1578
},
{
"epoch": 0.8425827107790822,
"grad_norm": 1.25,
"learning_rate": 1.3413123144236448e-06,
"loss": 0.13106584548950195,
"step": 1579
},
{
"epoch": 0.8431163287086446,
"grad_norm": 0.83203125,
"learning_rate": 1.3377547285105197e-06,
"loss": 0.12615007162094116,
"step": 1580
},
{
"epoch": 0.843649946638207,
"grad_norm": 0.85546875,
"learning_rate": 1.3342079278093228e-06,
"loss": 0.12829062342643738,
"step": 1581
},
{
"epoch": 0.8441835645677694,
"grad_norm": 0.921875,
"learning_rate": 1.3306719248118652e-06,
"loss": 0.14391165971755981,
"step": 1582
},
{
"epoch": 0.844717182497332,
"grad_norm": 2.859375,
"learning_rate": 1.327146731971935e-06,
"loss": 0.11162194609642029,
"step": 1583
},
{
"epoch": 0.8452508004268944,
"grad_norm": 0.91015625,
"learning_rate": 1.323632361705238e-06,
"loss": 0.1360490471124649,
"step": 1584
},
{
"epoch": 0.8457844183564568,
"grad_norm": 1.140625,
"learning_rate": 1.320128826389372e-06,
"loss": 0.13658775389194489,
"step": 1585
},
{
"epoch": 0.8463180362860192,
"grad_norm": 0.96875,
"learning_rate": 1.3166361383637685e-06,
"loss": 0.12432929128408432,
"step": 1586
},
{
"epoch": 0.8468516542155816,
"grad_norm": 1.046875,
"learning_rate": 1.313154309929656e-06,
"loss": 0.14409440755844116,
"step": 1587
},
{
"epoch": 0.847385272145144,
"grad_norm": 0.78125,
"learning_rate": 1.3096833533500192e-06,
"loss": 0.12185705453157425,
"step": 1588
},
{
"epoch": 0.8479188900747066,
"grad_norm": 0.88671875,
"learning_rate": 1.3062232808495442e-06,
"loss": 0.12867087125778198,
"step": 1589
},
{
"epoch": 0.848452508004269,
"grad_norm": 0.75390625,
"learning_rate": 1.3027741046145926e-06,
"loss": 0.13508981466293335,
"step": 1590
},
{
"epoch": 0.8489861259338314,
"grad_norm": 1.390625,
"learning_rate": 1.2993358367931421e-06,
"loss": 0.13000202178955078,
"step": 1591
},
{
"epoch": 0.8495197438633938,
"grad_norm": 0.78125,
"learning_rate": 1.2959084894947574e-06,
"loss": 0.13356545567512512,
"step": 1592
},
{
"epoch": 0.8500533617929562,
"grad_norm": 1.2109375,
"learning_rate": 1.2924920747905356e-06,
"loss": 0.1299126148223877,
"step": 1593
},
{
"epoch": 0.8505869797225186,
"grad_norm": 1.015625,
"learning_rate": 1.289086604713072e-06,
"loss": 0.13546958565711975,
"step": 1594
},
{
"epoch": 0.8511205976520811,
"grad_norm": 0.97265625,
"learning_rate": 1.2856920912564136e-06,
"loss": 0.13522681593894958,
"step": 1595
},
{
"epoch": 0.8516542155816436,
"grad_norm": 1.859375,
"learning_rate": 1.2823085463760207e-06,
"loss": 0.15748058259487152,
"step": 1596
},
{
"epoch": 0.852187833511206,
"grad_norm": 1.1875,
"learning_rate": 1.2789359819887205e-06,
"loss": 0.14872127771377563,
"step": 1597
},
{
"epoch": 0.8527214514407684,
"grad_norm": 1.046875,
"learning_rate": 1.2755744099726667e-06,
"loss": 0.15079641342163086,
"step": 1598
},
{
"epoch": 0.8532550693703308,
"grad_norm": 0.98828125,
"learning_rate": 1.272223842167299e-06,
"loss": 0.1323232352733612,
"step": 1599
},
{
"epoch": 0.8537886872998933,
"grad_norm": 1.6484375,
"learning_rate": 1.2688842903732982e-06,
"loss": 0.1307537704706192,
"step": 1600
},
{
"epoch": 0.8543223052294557,
"grad_norm": 3.109375,
"learning_rate": 1.2655557663525506e-06,
"loss": 0.11972157657146454,
"step": 1601
},
{
"epoch": 0.8548559231590181,
"grad_norm": 1.1328125,
"learning_rate": 1.262238281828099e-06,
"loss": 0.1408715546131134,
"step": 1602
},
{
"epoch": 0.8553895410885806,
"grad_norm": 1.1171875,
"learning_rate": 1.2589318484841077e-06,
"loss": 0.1208139955997467,
"step": 1603
},
{
"epoch": 0.855923159018143,
"grad_norm": 4.875,
"learning_rate": 1.2556364779658162e-06,
"loss": 0.15541815757751465,
"step": 1604
},
{
"epoch": 0.8564567769477054,
"grad_norm": 1.4296875,
"learning_rate": 1.2523521818795045e-06,
"loss": 0.12846364080905914,
"step": 1605
},
{
"epoch": 0.8569903948772679,
"grad_norm": 1.515625,
"learning_rate": 1.2490789717924447e-06,
"loss": 0.1414722502231598,
"step": 1606
},
{
"epoch": 0.8575240128068303,
"grad_norm": 2.171875,
"learning_rate": 1.2458168592328661e-06,
"loss": 0.12406115233898163,
"step": 1607
},
{
"epoch": 0.8580576307363927,
"grad_norm": 1.0,
"learning_rate": 1.242565855689912e-06,
"loss": 0.14140132069587708,
"step": 1608
},
{
"epoch": 0.8585912486659552,
"grad_norm": 4.0,
"learning_rate": 1.239325972613601e-06,
"loss": 0.12680977582931519,
"step": 1609
},
{
"epoch": 0.8591248665955176,
"grad_norm": 0.8984375,
"learning_rate": 1.2360972214147834e-06,
"loss": 0.1218956932425499,
"step": 1610
},
{
"epoch": 0.8596584845250801,
"grad_norm": 3.359375,
"learning_rate": 1.232879613465104e-06,
"loss": 0.13928741216659546,
"step": 1611
},
{
"epoch": 0.8601921024546425,
"grad_norm": 0.98828125,
"learning_rate": 1.2296731600969623e-06,
"loss": 0.13581764698028564,
"step": 1612
},
{
"epoch": 0.8607257203842049,
"grad_norm": 0.91796875,
"learning_rate": 1.2264778726034689e-06,
"loss": 0.15143683552742004,
"step": 1613
},
{
"epoch": 0.8612593383137673,
"grad_norm": 0.91796875,
"learning_rate": 1.2232937622384114e-06,
"loss": 0.12860378623008728,
"step": 1614
},
{
"epoch": 0.8617929562433297,
"grad_norm": 3.375,
"learning_rate": 1.220120840216208e-06,
"loss": 0.13705119490623474,
"step": 1615
},
{
"epoch": 0.8623265741728922,
"grad_norm": 2.40625,
"learning_rate": 1.2169591177118768e-06,
"loss": 0.13764588534832,
"step": 1616
},
{
"epoch": 0.8628601921024547,
"grad_norm": 1.375,
"learning_rate": 1.2138086058609843e-06,
"loss": 0.14480401575565338,
"step": 1617
},
{
"epoch": 0.8633938100320171,
"grad_norm": 1.109375,
"learning_rate": 1.2106693157596194e-06,
"loss": 0.13437631726264954,
"step": 1618
},
{
"epoch": 0.8639274279615795,
"grad_norm": 0.79296875,
"learning_rate": 1.2075412584643448e-06,
"loss": 0.1355127990245819,
"step": 1619
},
{
"epoch": 0.8644610458911419,
"grad_norm": 0.984375,
"learning_rate": 1.2044244449921615e-06,
"loss": 0.14481469988822937,
"step": 1620
},
{
"epoch": 0.8649946638207043,
"grad_norm": 0.8515625,
"learning_rate": 1.2013188863204702e-06,
"loss": 0.13199275732040405,
"step": 1621
},
{
"epoch": 0.8655282817502669,
"grad_norm": 1.15625,
"learning_rate": 1.1982245933870337e-06,
"loss": 0.13204403221607208,
"step": 1622
},
{
"epoch": 0.8660618996798293,
"grad_norm": 1.7890625,
"learning_rate": 1.1951415770899345e-06,
"loss": 0.1343294382095337,
"step": 1623
},
{
"epoch": 0.8665955176093917,
"grad_norm": 0.6640625,
"learning_rate": 1.19206984828754e-06,
"loss": 0.12739582359790802,
"step": 1624
},
{
"epoch": 0.8671291355389541,
"grad_norm": 4.96875,
"learning_rate": 1.1890094177984646e-06,
"loss": 0.1375499665737152,
"step": 1625
},
{
"epoch": 0.8676627534685165,
"grad_norm": 0.875,
"learning_rate": 1.1859602964015264e-06,
"loss": 0.12971892952919006,
"step": 1626
},
{
"epoch": 0.8681963713980789,
"grad_norm": 2.09375,
"learning_rate": 1.1829224948357162e-06,
"loss": 0.13460181653499603,
"step": 1627
},
{
"epoch": 0.8687299893276414,
"grad_norm": 0.80859375,
"learning_rate": 1.1798960238001545e-06,
"loss": 0.13045546412467957,
"step": 1628
},
{
"epoch": 0.8692636072572039,
"grad_norm": 0.79296875,
"learning_rate": 1.1768808939540589e-06,
"loss": 0.12583422660827637,
"step": 1629
},
{
"epoch": 0.8697972251867663,
"grad_norm": 1.375,
"learning_rate": 1.1738771159166984e-06,
"loss": 0.1304159164428711,
"step": 1630
},
{
"epoch": 0.8703308431163287,
"grad_norm": 1.078125,
"learning_rate": 1.170884700267365e-06,
"loss": 0.11638733744621277,
"step": 1631
},
{
"epoch": 0.8708644610458911,
"grad_norm": 2.671875,
"learning_rate": 1.1679036575453303e-06,
"loss": 0.14425690472126007,
"step": 1632
},
{
"epoch": 0.8713980789754536,
"grad_norm": 1.1484375,
"learning_rate": 1.1649339982498113e-06,
"loss": 0.1339808702468872,
"step": 1633
},
{
"epoch": 0.871931696905016,
"grad_norm": 1.03125,
"learning_rate": 1.161975732839934e-06,
"loss": 0.13750818371772766,
"step": 1634
},
{
"epoch": 0.8724653148345785,
"grad_norm": 1.640625,
"learning_rate": 1.1590288717346923e-06,
"loss": 0.13689792156219482,
"step": 1635
},
{
"epoch": 0.8729989327641409,
"grad_norm": 0.93359375,
"learning_rate": 1.1560934253129164e-06,
"loss": 0.14501652121543884,
"step": 1636
},
{
"epoch": 0.8735325506937033,
"grad_norm": 1.578125,
"learning_rate": 1.1531694039132325e-06,
"loss": 0.12719105184078217,
"step": 1637
},
{
"epoch": 0.8740661686232657,
"grad_norm": 1.046875,
"learning_rate": 1.1502568178340297e-06,
"loss": 0.13025647401809692,
"step": 1638
},
{
"epoch": 0.8745997865528282,
"grad_norm": 1.8046875,
"learning_rate": 1.1473556773334208e-06,
"loss": 0.13237416744232178,
"step": 1639
},
{
"epoch": 0.8751334044823906,
"grad_norm": 1.90625,
"learning_rate": 1.1444659926292073e-06,
"loss": 0.13917690515518188,
"step": 1640
},
{
"epoch": 0.875667022411953,
"grad_norm": 0.79296875,
"learning_rate": 1.1415877738988434e-06,
"loss": 0.1287122219800949,
"step": 1641
},
{
"epoch": 0.8762006403415155,
"grad_norm": 0.65625,
"learning_rate": 1.138721031279402e-06,
"loss": 0.13589254021644592,
"step": 1642
},
{
"epoch": 0.8767342582710779,
"grad_norm": 0.86328125,
"learning_rate": 1.1358657748675343e-06,
"loss": 0.1326775848865509,
"step": 1643
},
{
"epoch": 0.8772678762006404,
"grad_norm": 3.328125,
"learning_rate": 1.1330220147194387e-06,
"loss": 0.11686262488365173,
"step": 1644
},
{
"epoch": 0.8778014941302028,
"grad_norm": 1.0078125,
"learning_rate": 1.130189760850824e-06,
"loss": 0.14186236262321472,
"step": 1645
},
{
"epoch": 0.8783351120597652,
"grad_norm": 1.421875,
"learning_rate": 1.127369023236873e-06,
"loss": 0.11880980432033539,
"step": 1646
},
{
"epoch": 0.8788687299893276,
"grad_norm": 2.140625,
"learning_rate": 1.12455981181221e-06,
"loss": 0.12937670946121216,
"step": 1647
},
{
"epoch": 0.87940234791889,
"grad_norm": 1.1875,
"learning_rate": 1.1217621364708623e-06,
"loss": 0.13842421770095825,
"step": 1648
},
{
"epoch": 0.8799359658484525,
"grad_norm": 0.94140625,
"learning_rate": 1.1189760070662281e-06,
"loss": 0.12946456670761108,
"step": 1649
},
{
"epoch": 0.880469583778015,
"grad_norm": 0.80859375,
"learning_rate": 1.11620143341104e-06,
"loss": 0.13728994131088257,
"step": 1650
},
{
"epoch": 0.8810032017075774,
"grad_norm": 1.140625,
"learning_rate": 1.1134384252773331e-06,
"loss": 0.13778895139694214,
"step": 1651
},
{
"epoch": 0.8815368196371398,
"grad_norm": 1.21875,
"learning_rate": 1.1106869923964072e-06,
"loss": 0.13748079538345337,
"step": 1652
},
{
"epoch": 0.8820704375667022,
"grad_norm": 0.77734375,
"learning_rate": 1.107947144458795e-06,
"loss": 0.12813305854797363,
"step": 1653
},
{
"epoch": 0.8826040554962646,
"grad_norm": 0.9609375,
"learning_rate": 1.1052188911142248e-06,
"loss": 0.15039172768592834,
"step": 1654
},
{
"epoch": 0.8831376734258272,
"grad_norm": 0.984375,
"learning_rate": 1.1025022419715932e-06,
"loss": 0.12616077065467834,
"step": 1655
},
{
"epoch": 0.8836712913553896,
"grad_norm": 1.3515625,
"learning_rate": 1.0997972065989235e-06,
"loss": 0.12917748093605042,
"step": 1656
},
{
"epoch": 0.884204909284952,
"grad_norm": 0.875,
"learning_rate": 1.0971037945233364e-06,
"loss": 0.12897878885269165,
"step": 1657
},
{
"epoch": 0.8847385272145144,
"grad_norm": 1.046875,
"learning_rate": 1.094422015231014e-06,
"loss": 0.16346164047718048,
"step": 1658
},
{
"epoch": 0.8852721451440768,
"grad_norm": 0.76171875,
"learning_rate": 1.09175187816717e-06,
"loss": 0.1254361867904663,
"step": 1659
},
{
"epoch": 0.8858057630736392,
"grad_norm": 0.9140625,
"learning_rate": 1.089093392736013e-06,
"loss": 0.14035779237747192,
"step": 1660
},
{
"epoch": 0.8863393810032018,
"grad_norm": 1.03125,
"learning_rate": 1.0864465683007154e-06,
"loss": 0.1295059323310852,
"step": 1661
},
{
"epoch": 0.8868729989327642,
"grad_norm": 1.1796875,
"learning_rate": 1.0838114141833793e-06,
"loss": 0.12817421555519104,
"step": 1662
},
{
"epoch": 0.8874066168623266,
"grad_norm": 1.78125,
"learning_rate": 1.081187939665002e-06,
"loss": 0.14354227483272552,
"step": 1663
},
{
"epoch": 0.887940234791889,
"grad_norm": 1.1328125,
"learning_rate": 1.0785761539854493e-06,
"loss": 0.133744478225708,
"step": 1664
},
{
"epoch": 0.8884738527214514,
"grad_norm": 1.0859375,
"learning_rate": 1.075976066343415e-06,
"loss": 0.1302565038204193,
"step": 1665
},
{
"epoch": 0.8890074706510138,
"grad_norm": 0.84375,
"learning_rate": 1.0733876858963966e-06,
"loss": 0.13001035153865814,
"step": 1666
},
{
"epoch": 0.8895410885805763,
"grad_norm": 1.0546875,
"learning_rate": 1.0708110217606536e-06,
"loss": 0.14882463216781616,
"step": 1667
},
{
"epoch": 0.8900747065101388,
"grad_norm": 0.75,
"learning_rate": 1.068246083011186e-06,
"loss": 0.13776758313179016,
"step": 1668
},
{
"epoch": 0.8906083244397012,
"grad_norm": 1.1484375,
"learning_rate": 1.0656928786816934e-06,
"loss": 0.1487037092447281,
"step": 1669
},
{
"epoch": 0.8911419423692636,
"grad_norm": 1.828125,
"learning_rate": 1.0631514177645481e-06,
"loss": 0.12689588963985443,
"step": 1670
},
{
"epoch": 0.891675560298826,
"grad_norm": 1.0390625,
"learning_rate": 1.060621709210765e-06,
"loss": 0.12347482889890671,
"step": 1671
},
{
"epoch": 0.8922091782283885,
"grad_norm": 1.6328125,
"learning_rate": 1.0581037619299603e-06,
"loss": 0.133344829082489,
"step": 1672
},
{
"epoch": 0.8927427961579509,
"grad_norm": 1.484375,
"learning_rate": 1.055597584790334e-06,
"loss": 0.1300434172153473,
"step": 1673
},
{
"epoch": 0.8932764140875133,
"grad_norm": 2.015625,
"learning_rate": 1.0531031866186274e-06,
"loss": 0.139629065990448,
"step": 1674
},
{
"epoch": 0.8938100320170758,
"grad_norm": 0.69921875,
"learning_rate": 1.0506205762001e-06,
"loss": 0.1281210482120514,
"step": 1675
},
{
"epoch": 0.8943436499466382,
"grad_norm": 1.0859375,
"learning_rate": 1.04814976227849e-06,
"loss": 0.14286497235298157,
"step": 1676
},
{
"epoch": 0.8948772678762006,
"grad_norm": 1.671875,
"learning_rate": 1.0456907535559935e-06,
"loss": 0.14387962222099304,
"step": 1677
},
{
"epoch": 0.8954108858057631,
"grad_norm": 1.265625,
"learning_rate": 1.0432435586932254e-06,
"loss": 0.13069745898246765,
"step": 1678
},
{
"epoch": 0.8959445037353255,
"grad_norm": 1.09375,
"learning_rate": 1.0408081863091938e-06,
"loss": 0.1324278861284256,
"step": 1679
},
{
"epoch": 0.8964781216648879,
"grad_norm": 4.75,
"learning_rate": 1.038384644981267e-06,
"loss": 0.12640777230262756,
"step": 1680
},
{
"epoch": 0.8970117395944504,
"grad_norm": 0.80078125,
"learning_rate": 1.0359729432451456e-06,
"loss": 0.13706810772418976,
"step": 1681
},
{
"epoch": 0.8975453575240128,
"grad_norm": 1.0078125,
"learning_rate": 1.0335730895948298e-06,
"loss": 0.12800201773643494,
"step": 1682
},
{
"epoch": 0.8980789754535753,
"grad_norm": 7.25,
"learning_rate": 1.0311850924825918e-06,
"loss": 0.15117138624191284,
"step": 1683
},
{
"epoch": 0.8986125933831377,
"grad_norm": 0.88671875,
"learning_rate": 1.028808960318946e-06,
"loss": 0.13372406363487244,
"step": 1684
},
{
"epoch": 0.8991462113127001,
"grad_norm": 1.296875,
"learning_rate": 1.0264447014726163e-06,
"loss": 0.1341048926115036,
"step": 1685
},
{
"epoch": 0.8996798292422625,
"grad_norm": 0.8046875,
"learning_rate": 1.0240923242705107e-06,
"loss": 0.12325557321310043,
"step": 1686
},
{
"epoch": 0.9002134471718249,
"grad_norm": 0.83984375,
"learning_rate": 1.0217518369976889e-06,
"loss": 0.1388193517923355,
"step": 1687
},
{
"epoch": 0.9007470651013874,
"grad_norm": 0.83984375,
"learning_rate": 1.0194232478973362e-06,
"loss": 0.1348731815814972,
"step": 1688
},
{
"epoch": 0.9012806830309499,
"grad_norm": 1.703125,
"learning_rate": 1.01710656517073e-06,
"loss": 0.15366414189338684,
"step": 1689
},
{
"epoch": 0.9018143009605123,
"grad_norm": 0.83984375,
"learning_rate": 1.0148017969772153e-06,
"loss": 0.13599376380443573,
"step": 1690
},
{
"epoch": 0.9023479188900747,
"grad_norm": 0.9609375,
"learning_rate": 1.0125089514341736e-06,
"loss": 0.11989589780569077,
"step": 1691
},
{
"epoch": 0.9028815368196371,
"grad_norm": 1.0234375,
"learning_rate": 1.0102280366169952e-06,
"loss": 0.15231576561927795,
"step": 1692
},
{
"epoch": 0.9034151547491995,
"grad_norm": 0.73046875,
"learning_rate": 1.0079590605590505e-06,
"loss": 0.134813129901886,
"step": 1693
},
{
"epoch": 0.9039487726787621,
"grad_norm": 2.6875,
"learning_rate": 1.0057020312516607e-06,
"loss": 0.128554105758667,
"step": 1694
},
{
"epoch": 0.9044823906083245,
"grad_norm": 1.671875,
"learning_rate": 1.0034569566440715e-06,
"loss": 0.14225754141807556,
"step": 1695
},
{
"epoch": 0.9050160085378869,
"grad_norm": 0.8125,
"learning_rate": 1.0012238446434237e-06,
"loss": 0.13177025318145752,
"step": 1696
},
{
"epoch": 0.9055496264674493,
"grad_norm": 0.78125,
"learning_rate": 9.990027031147273e-07,
"loss": 0.14222922921180725,
"step": 1697
},
{
"epoch": 0.9060832443970117,
"grad_norm": 1.4296875,
"learning_rate": 9.967935398808299e-07,
"loss": 0.1329667568206787,
"step": 1698
},
{
"epoch": 0.9066168623265741,
"grad_norm": 0.91015625,
"learning_rate": 9.945963627223936e-07,
"loss": 0.13734778761863708,
"step": 1699
},
{
"epoch": 0.9071504802561366,
"grad_norm": 4.125,
"learning_rate": 9.924111793778644e-07,
"loss": 0.13520969450473785,
"step": 1700
},
{
"epoch": 0.9076840981856991,
"grad_norm": 1.0234375,
"learning_rate": 9.90237997543448e-07,
"loss": 0.13347968459129333,
"step": 1701
},
{
"epoch": 0.9082177161152615,
"grad_norm": 1.171875,
"learning_rate": 9.880768248730772e-07,
"loss": 0.11313894391059875,
"step": 1702
},
{
"epoch": 0.9087513340448239,
"grad_norm": 1.015625,
"learning_rate": 9.859276689783942e-07,
"loss": 0.1418248414993286,
"step": 1703
},
{
"epoch": 0.9092849519743863,
"grad_norm": 1.6484375,
"learning_rate": 9.837905374287114e-07,
"loss": 0.11847983300685883,
"step": 1704
},
{
"epoch": 0.9098185699039488,
"grad_norm": 2.03125,
"learning_rate": 9.816654377509974e-07,
"loss": 0.12662900984287262,
"step": 1705
},
{
"epoch": 0.9103521878335112,
"grad_norm": 0.89453125,
"learning_rate": 9.795523774298413e-07,
"loss": 0.12179495394229889,
"step": 1706
},
{
"epoch": 0.9108858057630737,
"grad_norm": 6.9375,
"learning_rate": 9.7745136390743e-07,
"loss": 0.12566936016082764,
"step": 1707
},
{
"epoch": 0.9114194236926361,
"grad_norm": 0.765625,
"learning_rate": 9.753624045835244e-07,
"loss": 0.1257477104663849,
"step": 1708
},
{
"epoch": 0.9119530416221985,
"grad_norm": 1.1875,
"learning_rate": 9.732855068154253e-07,
"loss": 0.12025701254606247,
"step": 1709
},
{
"epoch": 0.9124866595517609,
"grad_norm": 0.86328125,
"learning_rate": 9.712206779179585e-07,
"loss": 0.12734198570251465,
"step": 1710
},
{
"epoch": 0.9130202774813234,
"grad_norm": 3.796875,
"learning_rate": 9.691679251634386e-07,
"loss": 0.1333707869052887,
"step": 1711
},
{
"epoch": 0.9135538954108858,
"grad_norm": 1.1875,
"learning_rate": 9.67127255781652e-07,
"loss": 0.1363159716129303,
"step": 1712
},
{
"epoch": 0.9140875133404482,
"grad_norm": 1.8828125,
"learning_rate": 9.650986769598242e-07,
"loss": 0.13390953838825226,
"step": 1713
},
{
"epoch": 0.9146211312700107,
"grad_norm": 0.6953125,
"learning_rate": 9.630821958426003e-07,
"loss": 0.1299741566181183,
"step": 1714
},
{
"epoch": 0.9151547491995731,
"grad_norm": 1.0234375,
"learning_rate": 9.610778195320147e-07,
"loss": 0.14529259502887726,
"step": 1715
},
{
"epoch": 0.9156883671291356,
"grad_norm": 0.89453125,
"learning_rate": 9.590855550874725e-07,
"loss": 0.13616861402988434,
"step": 1716
},
{
"epoch": 0.916221985058698,
"grad_norm": 1.8828125,
"learning_rate": 9.57105409525718e-07,
"loss": 0.12502923607826233,
"step": 1717
},
{
"epoch": 0.9167556029882604,
"grad_norm": 0.9296875,
"learning_rate": 9.551373898208124e-07,
"loss": 0.13610002398490906,
"step": 1718
},
{
"epoch": 0.9172892209178228,
"grad_norm": 0.78515625,
"learning_rate": 9.531815029041122e-07,
"loss": 0.13290883600711823,
"step": 1719
},
{
"epoch": 0.9178228388473852,
"grad_norm": 0.9921875,
"learning_rate": 9.512377556642399e-07,
"loss": 0.1424621343612671,
"step": 1720
},
{
"epoch": 0.9183564567769477,
"grad_norm": 0.875,
"learning_rate": 9.493061549470649e-07,
"loss": 0.1423891931772232,
"step": 1721
},
{
"epoch": 0.9188900747065102,
"grad_norm": 0.984375,
"learning_rate": 9.473867075556724e-07,
"loss": 0.15676197409629822,
"step": 1722
},
{
"epoch": 0.9194236926360726,
"grad_norm": 1.0,
"learning_rate": 9.454794202503483e-07,
"loss": 0.1283886879682541,
"step": 1723
},
{
"epoch": 0.919957310565635,
"grad_norm": 2.765625,
"learning_rate": 9.435842997485473e-07,
"loss": 0.14085377752780914,
"step": 1724
},
{
"epoch": 0.9204909284951974,
"grad_norm": 0.76171875,
"learning_rate": 9.417013527248748e-07,
"loss": 0.12868866324424744,
"step": 1725
},
{
"epoch": 0.9210245464247598,
"grad_norm": 1.046875,
"learning_rate": 9.3983058581106e-07,
"loss": 0.134757861495018,
"step": 1726
},
{
"epoch": 0.9215581643543223,
"grad_norm": 1.0234375,
"learning_rate": 9.379720055959339e-07,
"loss": 0.12237986922264099,
"step": 1727
},
{
"epoch": 0.9220917822838848,
"grad_norm": 1.1796875,
"learning_rate": 9.36125618625407e-07,
"loss": 0.13216546177864075,
"step": 1728
},
{
"epoch": 0.9226254002134472,
"grad_norm": 1.984375,
"learning_rate": 9.342914314024442e-07,
"loss": 0.13459865748882294,
"step": 1729
},
{
"epoch": 0.9231590181430096,
"grad_norm": 0.953125,
"learning_rate": 9.324694503870439e-07,
"loss": 0.15165108442306519,
"step": 1730
},
{
"epoch": 0.923692636072572,
"grad_norm": 1.0234375,
"learning_rate": 9.306596819962129e-07,
"loss": 0.13417083024978638,
"step": 1731
},
{
"epoch": 0.9242262540021344,
"grad_norm": 0.88671875,
"learning_rate": 9.288621326039471e-07,
"loss": 0.15639328956604004,
"step": 1732
},
{
"epoch": 0.924759871931697,
"grad_norm": 1.234375,
"learning_rate": 9.270768085412047e-07,
"loss": 0.11828254163265228,
"step": 1733
},
{
"epoch": 0.9252934898612594,
"grad_norm": 1.5,
"learning_rate": 9.253037160958893e-07,
"loss": 0.1382131576538086,
"step": 1734
},
{
"epoch": 0.9258271077908218,
"grad_norm": 0.69921875,
"learning_rate": 9.235428615128227e-07,
"loss": 0.1357635259628296,
"step": 1735
},
{
"epoch": 0.9263607257203842,
"grad_norm": 1.296875,
"learning_rate": 9.217942509937251e-07,
"loss": 0.15630367398262024,
"step": 1736
},
{
"epoch": 0.9268943436499466,
"grad_norm": 0.71484375,
"learning_rate": 9.20057890697194e-07,
"loss": 0.12460669130086899,
"step": 1737
},
{
"epoch": 0.927427961579509,
"grad_norm": 0.953125,
"learning_rate": 9.183337867386817e-07,
"loss": 0.1225997731089592,
"step": 1738
},
{
"epoch": 0.9279615795090715,
"grad_norm": 1.4765625,
"learning_rate": 9.166219451904725e-07,
"loss": 0.12175506353378296,
"step": 1739
},
{
"epoch": 0.928495197438634,
"grad_norm": 0.70703125,
"learning_rate": 9.149223720816651e-07,
"loss": 0.13386204838752747,
"step": 1740
},
{
"epoch": 0.9290288153681964,
"grad_norm": 0.90234375,
"learning_rate": 9.132350733981447e-07,
"loss": 0.11104894429445267,
"step": 1741
},
{
"epoch": 0.9295624332977588,
"grad_norm": 6.21875,
"learning_rate": 9.115600550825707e-07,
"loss": 0.1413423717021942,
"step": 1742
},
{
"epoch": 0.9300960512273212,
"grad_norm": 0.7578125,
"learning_rate": 9.098973230343478e-07,
"loss": 0.13780491054058075,
"step": 1743
},
{
"epoch": 0.9306296691568837,
"grad_norm": 2.8125,
"learning_rate": 9.082468831096099e-07,
"loss": 0.12261441349983215,
"step": 1744
},
{
"epoch": 0.9311632870864461,
"grad_norm": 0.88671875,
"learning_rate": 9.066087411211978e-07,
"loss": 0.12203633040189743,
"step": 1745
},
{
"epoch": 0.9316969050160085,
"grad_norm": 1.375,
"learning_rate": 9.04982902838639e-07,
"loss": 0.1383974403142929,
"step": 1746
},
{
"epoch": 0.932230522945571,
"grad_norm": 0.984375,
"learning_rate": 9.033693739881274e-07,
"loss": 0.12611787021160126,
"step": 1747
},
{
"epoch": 0.9327641408751334,
"grad_norm": 1.421875,
"learning_rate": 9.017681602525034e-07,
"loss": 0.14889508485794067,
"step": 1748
},
{
"epoch": 0.9332977588046958,
"grad_norm": 0.765625,
"learning_rate": 9.00179267271234e-07,
"loss": 0.128648579120636,
"step": 1749
},
{
"epoch": 0.9338313767342583,
"grad_norm": 0.6875,
"learning_rate": 8.98602700640391e-07,
"loss": 0.14330706000328064,
"step": 1750
},
{
"epoch": 0.9343649946638207,
"grad_norm": 1.5390625,
"learning_rate": 8.970384659126343e-07,
"loss": 0.14756599068641663,
"step": 1751
},
{
"epoch": 0.9348986125933831,
"grad_norm": 0.8203125,
"learning_rate": 8.954865685971902e-07,
"loss": 0.12900526821613312,
"step": 1752
},
{
"epoch": 0.9354322305229456,
"grad_norm": 1.234375,
"learning_rate": 8.939470141598336e-07,
"loss": 0.12822943925857544,
"step": 1753
},
{
"epoch": 0.935965848452508,
"grad_norm": 1.328125,
"learning_rate": 8.924198080228673e-07,
"loss": 0.12857690453529358,
"step": 1754
},
{
"epoch": 0.9364994663820705,
"grad_norm": 2.546875,
"learning_rate": 8.909049555651027e-07,
"loss": 0.13910698890686035,
"step": 1755
},
{
"epoch": 0.9370330843116329,
"grad_norm": 1.4609375,
"learning_rate": 8.894024621218426e-07,
"loss": 0.12563705444335938,
"step": 1756
},
{
"epoch": 0.9375667022411953,
"grad_norm": 0.76171875,
"learning_rate": 8.879123329848609e-07,
"loss": 0.1285485327243805,
"step": 1757
},
{
"epoch": 0.9381003201707577,
"grad_norm": 1.0625,
"learning_rate": 8.864345734023851e-07,
"loss": 0.14297977089881897,
"step": 1758
},
{
"epoch": 0.9386339381003201,
"grad_norm": 0.921875,
"learning_rate": 8.849691885790762e-07,
"loss": 0.1433422863483429,
"step": 1759
},
{
"epoch": 0.9391675560298826,
"grad_norm": 0.83984375,
"learning_rate": 8.835161836760124e-07,
"loss": 0.12399041652679443,
"step": 1760
},
{
"epoch": 0.9397011739594451,
"grad_norm": 0.796875,
"learning_rate": 8.82075563810669e-07,
"loss": 0.1285131275653839,
"step": 1761
},
{
"epoch": 0.9402347918890075,
"grad_norm": 0.8984375,
"learning_rate": 8.80647334056902e-07,
"loss": 0.14258256554603577,
"step": 1762
},
{
"epoch": 0.9407684098185699,
"grad_norm": 3.09375,
"learning_rate": 8.792314994449293e-07,
"loss": 0.1322297751903534,
"step": 1763
},
{
"epoch": 0.9413020277481323,
"grad_norm": 1.1015625,
"learning_rate": 8.77828064961313e-07,
"loss": 0.1252509355545044,
"step": 1764
},
{
"epoch": 0.9418356456776947,
"grad_norm": 0.73046875,
"learning_rate": 8.764370355489414e-07,
"loss": 0.12242335081100464,
"step": 1765
},
{
"epoch": 0.9423692636072573,
"grad_norm": 0.76953125,
"learning_rate": 8.750584161070139e-07,
"loss": 0.1380864381790161,
"step": 1766
},
{
"epoch": 0.9429028815368197,
"grad_norm": 1.1484375,
"learning_rate": 8.736922114910199e-07,
"loss": 0.12974321842193604,
"step": 1767
},
{
"epoch": 0.9434364994663821,
"grad_norm": 4.78125,
"learning_rate": 8.723384265127258e-07,
"loss": 0.12368044257164001,
"step": 1768
},
{
"epoch": 0.9439701173959445,
"grad_norm": 0.75390625,
"learning_rate": 8.709970659401548e-07,
"loss": 0.13815104961395264,
"step": 1769
},
{
"epoch": 0.9445037353255069,
"grad_norm": 1.0546875,
"learning_rate": 8.696681344975723e-07,
"loss": 0.13595116138458252,
"step": 1770
},
{
"epoch": 0.9450373532550693,
"grad_norm": 1.0078125,
"learning_rate": 8.683516368654673e-07,
"loss": 0.13592040538787842,
"step": 1771
},
{
"epoch": 0.9455709711846318,
"grad_norm": 0.71484375,
"learning_rate": 8.670475776805382e-07,
"loss": 0.1378227025270462,
"step": 1772
},
{
"epoch": 0.9461045891141943,
"grad_norm": 0.91796875,
"learning_rate": 8.657559615356746e-07,
"loss": 0.13306660950183868,
"step": 1773
},
{
"epoch": 0.9466382070437567,
"grad_norm": 0.75390625,
"learning_rate": 8.644767929799412e-07,
"loss": 0.13247272372245789,
"step": 1774
},
{
"epoch": 0.9471718249733191,
"grad_norm": 0.8046875,
"learning_rate": 8.632100765185643e-07,
"loss": 0.12949666380882263,
"step": 1775
},
{
"epoch": 0.9477054429028815,
"grad_norm": 0.79296875,
"learning_rate": 8.619558166129115e-07,
"loss": 0.12929758429527283,
"step": 1776
},
{
"epoch": 0.948239060832444,
"grad_norm": 0.8359375,
"learning_rate": 8.607140176804799e-07,
"loss": 0.14128342270851135,
"step": 1777
},
{
"epoch": 0.9487726787620064,
"grad_norm": 2.234375,
"learning_rate": 8.594846840948791e-07,
"loss": 0.14850196242332458,
"step": 1778
},
{
"epoch": 0.9493062966915689,
"grad_norm": 1.4921875,
"learning_rate": 8.582678201858157e-07,
"loss": 0.12694458663463593,
"step": 1779
},
{
"epoch": 0.9498399146211313,
"grad_norm": 0.734375,
"learning_rate": 8.570634302390777e-07,
"loss": 0.14477261900901794,
"step": 1780
},
{
"epoch": 0.9503735325506937,
"grad_norm": 0.63671875,
"learning_rate": 8.558715184965204e-07,
"loss": 0.13382896780967712,
"step": 1781
},
{
"epoch": 0.9509071504802561,
"grad_norm": 1.046875,
"learning_rate": 8.546920891560511e-07,
"loss": 0.12475122511386871,
"step": 1782
},
{
"epoch": 0.9514407684098186,
"grad_norm": 1.03125,
"learning_rate": 8.535251463716133e-07,
"loss": 0.13719260692596436,
"step": 1783
},
{
"epoch": 0.951974386339381,
"grad_norm": 1.0078125,
"learning_rate": 8.523706942531736e-07,
"loss": 0.13573943078517914,
"step": 1784
},
{
"epoch": 0.9525080042689434,
"grad_norm": 0.625,
"learning_rate": 8.512287368667064e-07,
"loss": 0.12442819774150848,
"step": 1785
},
{
"epoch": 0.9530416221985059,
"grad_norm": 1.046875,
"learning_rate": 8.500992782341798e-07,
"loss": 0.151163250207901,
"step": 1786
},
{
"epoch": 0.9535752401280683,
"grad_norm": 0.77734375,
"learning_rate": 8.48982322333541e-07,
"loss": 0.11478054523468018,
"step": 1787
},
{
"epoch": 0.9541088580576307,
"grad_norm": 2.359375,
"learning_rate": 8.478778730987028e-07,
"loss": 0.1267850697040558,
"step": 1788
},
{
"epoch": 0.9546424759871932,
"grad_norm": 1.265625,
"learning_rate": 8.4678593441953e-07,
"loss": 0.13507798314094543,
"step": 1789
},
{
"epoch": 0.9551760939167556,
"grad_norm": 1.09375,
"learning_rate": 8.457065101418253e-07,
"loss": 0.11745525151491165,
"step": 1790
},
{
"epoch": 0.955709711846318,
"grad_norm": 0.8515625,
"learning_rate": 8.44639604067314e-07,
"loss": 0.13406552374362946,
"step": 1791
},
{
"epoch": 0.9562433297758804,
"grad_norm": 1.0625,
"learning_rate": 8.43585219953635e-07,
"loss": 0.13691429793834686,
"step": 1792
},
{
"epoch": 0.9567769477054429,
"grad_norm": 1.3515625,
"learning_rate": 8.425433615143233e-07,
"loss": 0.13233208656311035,
"step": 1793
},
{
"epoch": 0.9573105656350054,
"grad_norm": 0.91796875,
"learning_rate": 8.41514032418799e-07,
"loss": 0.13404051959514618,
"step": 1794
},
{
"epoch": 0.9578441835645678,
"grad_norm": 0.84765625,
"learning_rate": 8.40497236292354e-07,
"loss": 0.14244693517684937,
"step": 1795
},
{
"epoch": 0.9583778014941302,
"grad_norm": 0.7578125,
"learning_rate": 8.394929767161381e-07,
"loss": 0.14010155200958252,
"step": 1796
},
{
"epoch": 0.9589114194236926,
"grad_norm": 1.3515625,
"learning_rate": 8.385012572271493e-07,
"loss": 0.14121106266975403,
"step": 1797
},
{
"epoch": 0.959445037353255,
"grad_norm": 0.9453125,
"learning_rate": 8.375220813182173e-07,
"loss": 0.1285974383354187,
"step": 1798
},
{
"epoch": 0.9599786552828175,
"grad_norm": 0.7578125,
"learning_rate": 8.365554524379963e-07,
"loss": 0.13701464235782623,
"step": 1799
},
{
"epoch": 0.96051227321238,
"grad_norm": 0.92578125,
"learning_rate": 8.356013739909469e-07,
"loss": 0.14377588033676147,
"step": 1800
},
{
"epoch": 0.9610458911419424,
"grad_norm": 0.859375,
"learning_rate": 8.346598493373289e-07,
"loss": 0.13290400803089142,
"step": 1801
},
{
"epoch": 0.9615795090715048,
"grad_norm": 1.921875,
"learning_rate": 8.337308817931872e-07,
"loss": 0.12635719776153564,
"step": 1802
},
{
"epoch": 0.9621131270010672,
"grad_norm": 0.94140625,
"learning_rate": 8.328144746303402e-07,
"loss": 0.13557040691375732,
"step": 1803
},
{
"epoch": 0.9626467449306296,
"grad_norm": 1.1640625,
"learning_rate": 8.319106310763696e-07,
"loss": 0.13522247970104218,
"step": 1804
},
{
"epoch": 0.9631803628601922,
"grad_norm": 1.3671875,
"learning_rate": 8.310193543146078e-07,
"loss": 0.13603797554969788,
"step": 1805
},
{
"epoch": 0.9637139807897546,
"grad_norm": 0.75,
"learning_rate": 8.301406474841267e-07,
"loss": 0.1409311294555664,
"step": 1806
},
{
"epoch": 0.964247598719317,
"grad_norm": 0.859375,
"learning_rate": 8.292745136797271e-07,
"loss": 0.14464767277240753,
"step": 1807
},
{
"epoch": 0.9647812166488794,
"grad_norm": 1.640625,
"learning_rate": 8.28420955951928e-07,
"loss": 0.13562503457069397,
"step": 1808
},
{
"epoch": 0.9653148345784418,
"grad_norm": 0.76953125,
"learning_rate": 8.275799773069553e-07,
"loss": 0.13538064062595367,
"step": 1809
},
{
"epoch": 0.9658484525080042,
"grad_norm": 0.79296875,
"learning_rate": 8.267515807067313e-07,
"loss": 0.13306298851966858,
"step": 1810
},
{
"epoch": 0.9663820704375667,
"grad_norm": 1.4765625,
"learning_rate": 8.259357690688641e-07,
"loss": 0.1385498046875,
"step": 1811
},
{
"epoch": 0.9669156883671292,
"grad_norm": 1.7734375,
"learning_rate": 8.25132545266639e-07,
"loss": 0.13380414247512817,
"step": 1812
},
{
"epoch": 0.9674493062966916,
"grad_norm": 1.28125,
"learning_rate": 8.243419121290052e-07,
"loss": 0.13307860493659973,
"step": 1813
},
{
"epoch": 0.967982924226254,
"grad_norm": 0.8125,
"learning_rate": 8.235638724405691e-07,
"loss": 0.13511045277118683,
"step": 1814
},
{
"epoch": 0.9685165421558164,
"grad_norm": 1.1015625,
"learning_rate": 8.227984289415822e-07,
"loss": 0.12149843573570251,
"step": 1815
},
{
"epoch": 0.9690501600853789,
"grad_norm": 1.4609375,
"learning_rate": 8.220455843279318e-07,
"loss": 0.12214652448892593,
"step": 1816
},
{
"epoch": 0.9695837780149413,
"grad_norm": 0.9609375,
"learning_rate": 8.213053412511338e-07,
"loss": 0.12955230474472046,
"step": 1817
},
{
"epoch": 0.9701173959445037,
"grad_norm": 0.953125,
"learning_rate": 8.205777023183204e-07,
"loss": 0.13512593507766724,
"step": 1818
},
{
"epoch": 0.9706510138740662,
"grad_norm": 0.98828125,
"learning_rate": 8.198626700922315e-07,
"loss": 0.12321390956640244,
"step": 1819
},
{
"epoch": 0.9711846318036286,
"grad_norm": 0.7578125,
"learning_rate": 8.191602470912072e-07,
"loss": 0.12847599387168884,
"step": 1820
},
{
"epoch": 0.971718249733191,
"grad_norm": 0.91015625,
"learning_rate": 8.184704357891774e-07,
"loss": 0.1245943158864975,
"step": 1821
},
{
"epoch": 0.9722518676627535,
"grad_norm": 1.09375,
"learning_rate": 8.177932386156543e-07,
"loss": 0.1346108615398407,
"step": 1822
},
{
"epoch": 0.9727854855923159,
"grad_norm": 1.2578125,
"learning_rate": 8.17128657955723e-07,
"loss": 0.13684573769569397,
"step": 1823
},
{
"epoch": 0.9733191035218783,
"grad_norm": 0.875,
"learning_rate": 8.164766961500324e-07,
"loss": 0.1380728930234909,
"step": 1824
},
{
"epoch": 0.9738527214514408,
"grad_norm": 0.91015625,
"learning_rate": 8.158373554947895e-07,
"loss": 0.14562927186489105,
"step": 1825
},
{
"epoch": 0.9743863393810032,
"grad_norm": 1.1796875,
"learning_rate": 8.15210638241748e-07,
"loss": 0.11774437874555588,
"step": 1826
},
{
"epoch": 0.9749199573105657,
"grad_norm": 0.91015625,
"learning_rate": 8.145965465982032e-07,
"loss": 0.12457894533872604,
"step": 1827
},
{
"epoch": 0.9754535752401281,
"grad_norm": 1.078125,
"learning_rate": 8.139950827269822e-07,
"loss": 0.14075064659118652,
"step": 1828
},
{
"epoch": 0.9759871931696905,
"grad_norm": 1.265625,
"learning_rate": 8.134062487464382e-07,
"loss": 0.1353759914636612,
"step": 1829
},
{
"epoch": 0.9765208110992529,
"grad_norm": 1.109375,
"learning_rate": 8.128300467304412e-07,
"loss": 0.1447814404964447,
"step": 1830
},
{
"epoch": 0.9770544290288153,
"grad_norm": 1.21875,
"learning_rate": 8.122664787083717e-07,
"loss": 0.14289940893650055,
"step": 1831
},
{
"epoch": 0.9775880469583778,
"grad_norm": 1.1171875,
"learning_rate": 8.117155466651135e-07,
"loss": 0.12737734615802765,
"step": 1832
},
{
"epoch": 0.9781216648879403,
"grad_norm": 0.82421875,
"learning_rate": 8.111772525410462e-07,
"loss": 0.1438935101032257,
"step": 1833
},
{
"epoch": 0.9786552828175027,
"grad_norm": 1.1171875,
"learning_rate": 8.10651598232039e-07,
"loss": 0.1311788558959961,
"step": 1834
},
{
"epoch": 0.9791889007470651,
"grad_norm": 2.265625,
"learning_rate": 8.10138585589443e-07,
"loss": 0.13595178723335266,
"step": 1835
},
{
"epoch": 0.9797225186766275,
"grad_norm": 0.64453125,
"learning_rate": 8.096382164200874e-07,
"loss": 0.12458772957324982,
"step": 1836
},
{
"epoch": 0.9802561366061899,
"grad_norm": 0.9140625,
"learning_rate": 8.091504924862686e-07,
"loss": 0.13194678723812103,
"step": 1837
},
{
"epoch": 0.9807897545357525,
"grad_norm": 1.65625,
"learning_rate": 8.086754155057487e-07,
"loss": 0.1387510448694229,
"step": 1838
},
{
"epoch": 0.9813233724653149,
"grad_norm": 1.0625,
"learning_rate": 8.082129871517454e-07,
"loss": 0.12657923996448517,
"step": 1839
},
{
"epoch": 0.9818569903948773,
"grad_norm": 0.87890625,
"learning_rate": 8.077632090529295e-07,
"loss": 0.13320772349834442,
"step": 1840
},
{
"epoch": 0.9823906083244397,
"grad_norm": 1.1796875,
"learning_rate": 8.073260827934174e-07,
"loss": 0.12741559743881226,
"step": 1841
},
{
"epoch": 0.9829242262540021,
"grad_norm": 1.25,
"learning_rate": 8.069016099127651e-07,
"loss": 0.13130024075508118,
"step": 1842
},
{
"epoch": 0.9834578441835645,
"grad_norm": 0.91015625,
"learning_rate": 8.064897919059637e-07,
"loss": 0.14356930553913116,
"step": 1843
},
{
"epoch": 0.983991462113127,
"grad_norm": 1.2109375,
"learning_rate": 8.060906302234346e-07,
"loss": 0.13707491755485535,
"step": 1844
},
{
"epoch": 0.9845250800426895,
"grad_norm": 1.8359375,
"learning_rate": 8.057041262710227e-07,
"loss": 0.13387662172317505,
"step": 1845
},
{
"epoch": 0.9850586979722519,
"grad_norm": 1.703125,
"learning_rate": 8.05330281409993e-07,
"loss": 0.11836963891983032,
"step": 1846
},
{
"epoch": 0.9855923159018143,
"grad_norm": 1.90625,
"learning_rate": 8.049690969570257e-07,
"loss": 0.1521456241607666,
"step": 1847
},
{
"epoch": 0.9861259338313767,
"grad_norm": 1.1796875,
"learning_rate": 8.046205741842101e-07,
"loss": 0.13790878653526306,
"step": 1848
},
{
"epoch": 0.9866595517609391,
"grad_norm": 0.66796875,
"learning_rate": 8.042847143190412e-07,
"loss": 0.12888917326927185,
"step": 1849
},
{
"epoch": 0.9871931696905016,
"grad_norm": 1.140625,
"learning_rate": 8.039615185444167e-07,
"loss": 0.13533660769462585,
"step": 1850
},
{
"epoch": 0.987726787620064,
"grad_norm": 1.9296875,
"learning_rate": 8.036509879986293e-07,
"loss": 0.12285204231739044,
"step": 1851
},
{
"epoch": 0.9882604055496265,
"grad_norm": 1.328125,
"learning_rate": 8.033531237753671e-07,
"loss": 0.1325324922800064,
"step": 1852
},
{
"epoch": 0.9887940234791889,
"grad_norm": 2.328125,
"learning_rate": 8.030679269237055e-07,
"loss": 0.11829686164855957,
"step": 1853
},
{
"epoch": 0.9893276414087513,
"grad_norm": 0.7734375,
"learning_rate": 8.027953984481064e-07,
"loss": 0.1387133002281189,
"step": 1854
},
{
"epoch": 0.9898612593383138,
"grad_norm": 0.78515625,
"learning_rate": 8.025355393084141e-07,
"loss": 0.1255006194114685,
"step": 1855
},
{
"epoch": 0.9903948772678762,
"grad_norm": 1.2109375,
"learning_rate": 8.02288350419851e-07,
"loss": 0.13844719529151917,
"step": 1856
},
{
"epoch": 0.9909284951974386,
"grad_norm": 0.6796875,
"learning_rate": 8.020538326530152e-07,
"loss": 0.14026233553886414,
"step": 1857
},
{
"epoch": 0.991462113127001,
"grad_norm": 0.84375,
"learning_rate": 8.018319868338766e-07,
"loss": 0.13939185440540314,
"step": 1858
},
{
"epoch": 0.9919957310565635,
"grad_norm": 1.078125,
"learning_rate": 8.016228137437754e-07,
"loss": 0.13194409012794495,
"step": 1859
},
{
"epoch": 0.9925293489861259,
"grad_norm": 0.74609375,
"learning_rate": 8.014263141194179e-07,
"loss": 0.1324710249900818,
"step": 1860
},
{
"epoch": 0.9930629669156884,
"grad_norm": 3.296875,
"learning_rate": 8.012424886528747e-07,
"loss": 0.11792932450771332,
"step": 1861
},
{
"epoch": 0.9935965848452508,
"grad_norm": 1.0703125,
"learning_rate": 8.010713379915785e-07,
"loss": 0.1291143298149109,
"step": 1862
},
{
"epoch": 0.9941302027748132,
"grad_norm": 0.75,
"learning_rate": 8.009128627383211e-07,
"loss": 0.13740311563014984,
"step": 1863
},
{
"epoch": 0.9946638207043756,
"grad_norm": 0.9296875,
"learning_rate": 8.007670634512514e-07,
"loss": 0.13769245147705078,
"step": 1864
},
{
"epoch": 0.9951974386339381,
"grad_norm": 2.890625,
"learning_rate": 8.006339406438729e-07,
"loss": 0.11990561336278915,
"step": 1865
},
{
"epoch": 0.9957310565635006,
"grad_norm": 2.390625,
"learning_rate": 8.005134947850447e-07,
"loss": 0.13896150887012482,
"step": 1866
},
{
"epoch": 0.996264674493063,
"grad_norm": 1.21875,
"learning_rate": 8.004057262989756e-07,
"loss": 0.1304696798324585,
"step": 1867
},
{
"epoch": 0.9967982924226254,
"grad_norm": 0.91015625,
"learning_rate": 8.003106355652261e-07,
"loss": 0.11913923919200897,
"step": 1868
},
{
"epoch": 0.9973319103521878,
"grad_norm": 0.90625,
"learning_rate": 8.002282229187054e-07,
"loss": 0.13869155943393707,
"step": 1869
},
{
"epoch": 0.9978655282817502,
"grad_norm": 0.9375,
"learning_rate": 8.001584886496698e-07,
"loss": 0.14089763164520264,
"step": 1870
},
{
"epoch": 0.9983991462113126,
"grad_norm": 1.125,
"learning_rate": 8.001014330037235e-07,
"loss": 0.1332048773765564,
"step": 1871
},
{
"epoch": 0.9989327641408752,
"grad_norm": 1.65625,
"learning_rate": 8.000570561818164e-07,
"loss": 0.12420602142810822,
"step": 1872
},
{
"epoch": 0.9994663820704376,
"grad_norm": 1.7890625,
"learning_rate": 8.000253583402429e-07,
"loss": 0.12362730503082275,
"step": 1873
},
{
"epoch": 1.0,
"grad_norm": 1.265625,
"learning_rate": 8.000063395906429e-07,
"loss": 0.12663546204566956,
"step": 1874
},
{
"epoch": 1.0,
"eval_loss": 0.13393093645572662,
"eval_runtime": 5.1942,
"eval_samples_per_second": 11.551,
"eval_steps_per_second": 0.385,
"step": 1874
}
],
"logging_steps": 1,
"max_steps": 1874,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 937,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.657241807105137e+19,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}