gkd-542 / trainer_state.json
iamPi's picture
Add files using upload-large-folder tool
1ac3c7c verified
Raw
History Blame Contribute Delete
92.8 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.2000092255177822,
"eval_steps": 500,
"global_step": 542,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.000369020711287421,
"grad_norm": 32.204185485839844,
"learning_rate": 0.0,
"loss": 6.701247215270996,
"step": 1
},
{
"epoch": 0.000738041422574842,
"grad_norm": 32.250003814697266,
"learning_rate": 2.0000000000000003e-06,
"loss": 6.400952339172363,
"step": 2
},
{
"epoch": 0.001107062133862263,
"grad_norm": 24.697145462036133,
"learning_rate": 4.000000000000001e-06,
"loss": 6.332036972045898,
"step": 3
},
{
"epoch": 0.001476082845149684,
"grad_norm": 18.79606819152832,
"learning_rate": 6e-06,
"loss": 6.286121845245361,
"step": 4
},
{
"epoch": 0.001845103556437105,
"grad_norm": 18.390621185302734,
"learning_rate": 8.000000000000001e-06,
"loss": 6.1857757568359375,
"step": 5
},
{
"epoch": 0.002214124267724526,
"grad_norm": 15.014426231384277,
"learning_rate": 1e-05,
"loss": 6.110501766204834,
"step": 6
},
{
"epoch": 0.002583144979011947,
"grad_norm": 15.987100601196289,
"learning_rate": 1.2e-05,
"loss": 5.976349830627441,
"step": 7
},
{
"epoch": 0.002952165690299368,
"grad_norm": 18.343597412109375,
"learning_rate": 1.4000000000000001e-05,
"loss": 6.035942077636719,
"step": 8
},
{
"epoch": 0.0033211864015867893,
"grad_norm": 13.626900672912598,
"learning_rate": 1.6000000000000003e-05,
"loss": 6.0868072509765625,
"step": 9
},
{
"epoch": 0.00369020711287421,
"grad_norm": 11.279500007629395,
"learning_rate": 1.8e-05,
"loss": 6.021449089050293,
"step": 10
},
{
"epoch": 0.004059227824161631,
"grad_norm": 12.686036109924316,
"learning_rate": 2e-05,
"loss": 6.056863784790039,
"step": 11
},
{
"epoch": 0.004428248535449052,
"grad_norm": 10.581238746643066,
"learning_rate": 2.2000000000000003e-05,
"loss": 5.860020637512207,
"step": 12
},
{
"epoch": 0.004797269246736473,
"grad_norm": 8.909687042236328,
"learning_rate": 2.4e-05,
"loss": 5.828363418579102,
"step": 13
},
{
"epoch": 0.005166289958023894,
"grad_norm": 6.024884223937988,
"learning_rate": 2.6000000000000002e-05,
"loss": 5.932955741882324,
"step": 14
},
{
"epoch": 0.0055353106693113155,
"grad_norm": 7.050922393798828,
"learning_rate": 2.8000000000000003e-05,
"loss": 5.747481822967529,
"step": 15
},
{
"epoch": 0.005904331380598736,
"grad_norm": 5.779654502868652,
"learning_rate": 3e-05,
"loss": 5.645102024078369,
"step": 16
},
{
"epoch": 0.006273352091886157,
"grad_norm": 6.848283290863037,
"learning_rate": 3.2000000000000005e-05,
"loss": 6.164463996887207,
"step": 17
},
{
"epoch": 0.0066423728031735785,
"grad_norm": 5.797471046447754,
"learning_rate": 3.4000000000000007e-05,
"loss": 5.728422164916992,
"step": 18
},
{
"epoch": 0.007011393514460999,
"grad_norm": 6.957800388336182,
"learning_rate": 3.6e-05,
"loss": 5.617619514465332,
"step": 19
},
{
"epoch": 0.00738041422574842,
"grad_norm": 6.550650596618652,
"learning_rate": 3.8e-05,
"loss": 5.684273719787598,
"step": 20
},
{
"epoch": 0.007749434937035841,
"grad_norm": 5.633530616760254,
"learning_rate": 4e-05,
"loss": 5.696666240692139,
"step": 21
},
{
"epoch": 0.008118455648323262,
"grad_norm": 5.457925319671631,
"learning_rate": 4.2e-05,
"loss": 5.586109638214111,
"step": 22
},
{
"epoch": 0.008487476359610684,
"grad_norm": 7.086246967315674,
"learning_rate": 4.4000000000000006e-05,
"loss": 5.715998649597168,
"step": 23
},
{
"epoch": 0.008856497070898105,
"grad_norm": 6.937062740325928,
"learning_rate": 4.600000000000001e-05,
"loss": 5.574023246765137,
"step": 24
},
{
"epoch": 0.009225517782185525,
"grad_norm": 6.063978672027588,
"learning_rate": 4.8e-05,
"loss": 5.798689365386963,
"step": 25
},
{
"epoch": 0.009594538493472946,
"grad_norm": 5.2796807289123535,
"learning_rate": 5e-05,
"loss": 5.642449855804443,
"step": 26
},
{
"epoch": 0.009963559204760367,
"grad_norm": 6.248167991638184,
"learning_rate": 5.2000000000000004e-05,
"loss": 5.889494895935059,
"step": 27
},
{
"epoch": 0.010332579916047788,
"grad_norm": 5.35820198059082,
"learning_rate": 5.4000000000000005e-05,
"loss": 5.48094367980957,
"step": 28
},
{
"epoch": 0.010701600627335208,
"grad_norm": 5.648565292358398,
"learning_rate": 5.6000000000000006e-05,
"loss": 5.6747331619262695,
"step": 29
},
{
"epoch": 0.011070621338622631,
"grad_norm": 4.546300888061523,
"learning_rate": 5.8e-05,
"loss": 5.303746223449707,
"step": 30
},
{
"epoch": 0.011439642049910052,
"grad_norm": 4.514847278594971,
"learning_rate": 6e-05,
"loss": 5.3704681396484375,
"step": 31
},
{
"epoch": 0.011808662761197472,
"grad_norm": 3.8123302459716797,
"learning_rate": 6.2e-05,
"loss": 5.293277263641357,
"step": 32
},
{
"epoch": 0.012177683472484893,
"grad_norm": 6.512507438659668,
"learning_rate": 6.400000000000001e-05,
"loss": 5.347484588623047,
"step": 33
},
{
"epoch": 0.012546704183772314,
"grad_norm": 6.077487468719482,
"learning_rate": 6.6e-05,
"loss": 5.459488868713379,
"step": 34
},
{
"epoch": 0.012915724895059735,
"grad_norm": 9.170805931091309,
"learning_rate": 6.800000000000001e-05,
"loss": 5.849956512451172,
"step": 35
},
{
"epoch": 0.013284745606347157,
"grad_norm": 6.619089126586914,
"learning_rate": 7e-05,
"loss": 5.7309160232543945,
"step": 36
},
{
"epoch": 0.013653766317634578,
"grad_norm": 8.339101791381836,
"learning_rate": 7.2e-05,
"loss": 5.443577766418457,
"step": 37
},
{
"epoch": 0.014022787028921999,
"grad_norm": 6.749815464019775,
"learning_rate": 7.4e-05,
"loss": 5.468483924865723,
"step": 38
},
{
"epoch": 0.01439180774020942,
"grad_norm": 6.140146732330322,
"learning_rate": 7.6e-05,
"loss": 5.542171001434326,
"step": 39
},
{
"epoch": 0.01476082845149684,
"grad_norm": 5.653714656829834,
"learning_rate": 7.800000000000001e-05,
"loss": 5.644989490509033,
"step": 40
},
{
"epoch": 0.01512984916278426,
"grad_norm": 5.6543474197387695,
"learning_rate": 8e-05,
"loss": 5.580202102661133,
"step": 41
},
{
"epoch": 0.015498869874071682,
"grad_norm": 6.0401506423950195,
"learning_rate": 8.2e-05,
"loss": 6.081020832061768,
"step": 42
},
{
"epoch": 0.015867890585359104,
"grad_norm": 4.91713809967041,
"learning_rate": 8.4e-05,
"loss": 5.453029155731201,
"step": 43
},
{
"epoch": 0.016236911296646523,
"grad_norm": 4.726991653442383,
"learning_rate": 8.6e-05,
"loss": 5.531949520111084,
"step": 44
},
{
"epoch": 0.016605932007933945,
"grad_norm": 7.298604488372803,
"learning_rate": 8.800000000000001e-05,
"loss": 5.451730728149414,
"step": 45
},
{
"epoch": 0.016974952719221368,
"grad_norm": 5.089509963989258,
"learning_rate": 9e-05,
"loss": 5.463062763214111,
"step": 46
},
{
"epoch": 0.017343973430508787,
"grad_norm": 5.357382774353027,
"learning_rate": 9.200000000000001e-05,
"loss": 5.308757781982422,
"step": 47
},
{
"epoch": 0.01771299414179621,
"grad_norm": 5.978388786315918,
"learning_rate": 9.4e-05,
"loss": 5.5749430656433105,
"step": 48
},
{
"epoch": 0.01808201485308363,
"grad_norm": 5.822226047515869,
"learning_rate": 9.6e-05,
"loss": 5.463890075683594,
"step": 49
},
{
"epoch": 0.01845103556437105,
"grad_norm": 6.441373348236084,
"learning_rate": 9.8e-05,
"loss": 5.169106483459473,
"step": 50
},
{
"epoch": 0.01882005627565847,
"grad_norm": 4.273548603057861,
"learning_rate": 0.0001,
"loss": 5.433943748474121,
"step": 51
},
{
"epoch": 0.019189076986945892,
"grad_norm": 6.303419589996338,
"learning_rate": 0.0001,
"loss": 5.323245048522949,
"step": 52
},
{
"epoch": 0.019558097698233315,
"grad_norm": 6.22584342956543,
"learning_rate": 0.0001,
"loss": 5.212267875671387,
"step": 53
},
{
"epoch": 0.019927118409520734,
"grad_norm": 6.891298770904541,
"learning_rate": 0.0001,
"loss": 5.431396484375,
"step": 54
},
{
"epoch": 0.020296139120808156,
"grad_norm": 7.352054595947266,
"learning_rate": 0.0001,
"loss": 5.629201889038086,
"step": 55
},
{
"epoch": 0.020665159832095575,
"grad_norm": 4.706995487213135,
"learning_rate": 0.0001,
"loss": 5.119016647338867,
"step": 56
},
{
"epoch": 0.021034180543382998,
"grad_norm": 6.846929550170898,
"learning_rate": 0.0001,
"loss": 5.710336685180664,
"step": 57
},
{
"epoch": 0.021403201254670417,
"grad_norm": 5.81605339050293,
"learning_rate": 0.0001,
"loss": 5.506718158721924,
"step": 58
},
{
"epoch": 0.02177222196595784,
"grad_norm": 5.613030910491943,
"learning_rate": 0.0001,
"loss": 5.485965728759766,
"step": 59
},
{
"epoch": 0.022141242677245262,
"grad_norm": 4.727166652679443,
"learning_rate": 0.0001,
"loss": 5.300134181976318,
"step": 60
},
{
"epoch": 0.02251026338853268,
"grad_norm": 6.181882381439209,
"learning_rate": 0.0001,
"loss": 5.327762126922607,
"step": 61
},
{
"epoch": 0.022879284099820103,
"grad_norm": 4.523162364959717,
"learning_rate": 0.0001,
"loss": 5.624878406524658,
"step": 62
},
{
"epoch": 0.023248304811107522,
"grad_norm": 5.7826828956604,
"learning_rate": 0.0001,
"loss": 5.246135711669922,
"step": 63
},
{
"epoch": 0.023617325522394945,
"grad_norm": 3.8491437435150146,
"learning_rate": 0.0001,
"loss": 5.223575592041016,
"step": 64
},
{
"epoch": 0.023986346233682364,
"grad_norm": 6.430501937866211,
"learning_rate": 0.0001,
"loss": 5.349424839019775,
"step": 65
},
{
"epoch": 0.024355366944969786,
"grad_norm": 4.793713092803955,
"learning_rate": 0.0001,
"loss": 5.306467056274414,
"step": 66
},
{
"epoch": 0.02472438765625721,
"grad_norm": 5.871583461761475,
"learning_rate": 0.0001,
"loss": 5.41292142868042,
"step": 67
},
{
"epoch": 0.025093408367544628,
"grad_norm": 3.803302764892578,
"learning_rate": 0.0001,
"loss": 5.451375484466553,
"step": 68
},
{
"epoch": 0.02546242907883205,
"grad_norm": 4.517539978027344,
"learning_rate": 0.0001,
"loss": 5.264144420623779,
"step": 69
},
{
"epoch": 0.02583144979011947,
"grad_norm": 4.541337966918945,
"learning_rate": 0.0001,
"loss": 5.2370219230651855,
"step": 70
},
{
"epoch": 0.02620047050140689,
"grad_norm": 7.2052812576293945,
"learning_rate": 0.0001,
"loss": 5.298012733459473,
"step": 71
},
{
"epoch": 0.026569491212694314,
"grad_norm": 4.4917707443237305,
"learning_rate": 0.0001,
"loss": 5.396561145782471,
"step": 72
},
{
"epoch": 0.026938511923981733,
"grad_norm": 6.187958240509033,
"learning_rate": 0.0001,
"loss": 5.577371597290039,
"step": 73
},
{
"epoch": 0.027307532635269156,
"grad_norm": 5.212502956390381,
"learning_rate": 0.0001,
"loss": 5.232723712921143,
"step": 74
},
{
"epoch": 0.027676553346556575,
"grad_norm": 4.552979946136475,
"learning_rate": 0.0001,
"loss": 5.3533616065979,
"step": 75
},
{
"epoch": 0.028045574057843997,
"grad_norm": 5.144165515899658,
"learning_rate": 0.0001,
"loss": 5.281284809112549,
"step": 76
},
{
"epoch": 0.028414594769131416,
"grad_norm": 4.348478317260742,
"learning_rate": 0.0001,
"loss": 5.15976619720459,
"step": 77
},
{
"epoch": 0.02878361548041884,
"grad_norm": 3.5937416553497314,
"learning_rate": 0.0001,
"loss": 5.613420009613037,
"step": 78
},
{
"epoch": 0.02915263619170626,
"grad_norm": 5.10581111907959,
"learning_rate": 0.0001,
"loss": 5.1962571144104,
"step": 79
},
{
"epoch": 0.02952165690299368,
"grad_norm": 5.265993595123291,
"learning_rate": 0.0001,
"loss": 5.335290431976318,
"step": 80
},
{
"epoch": 0.029890677614281103,
"grad_norm": 3.9692020416259766,
"learning_rate": 0.0001,
"loss": 5.45660924911499,
"step": 81
},
{
"epoch": 0.03025969832556852,
"grad_norm": 3.537494421005249,
"learning_rate": 0.0001,
"loss": 5.203220844268799,
"step": 82
},
{
"epoch": 0.030628719036855944,
"grad_norm": 3.4767425060272217,
"learning_rate": 0.0001,
"loss": 5.189380645751953,
"step": 83
},
{
"epoch": 0.030997739748143363,
"grad_norm": 3.707651138305664,
"learning_rate": 0.0001,
"loss": 5.333460330963135,
"step": 84
},
{
"epoch": 0.031366760459430786,
"grad_norm": 3.983680486679077,
"learning_rate": 0.0001,
"loss": 5.286036491394043,
"step": 85
},
{
"epoch": 0.03173578117071821,
"grad_norm": 5.241593837738037,
"learning_rate": 0.0001,
"loss": 5.400127410888672,
"step": 86
},
{
"epoch": 0.03210480188200563,
"grad_norm": 3.111454963684082,
"learning_rate": 0.0001,
"loss": 5.3072919845581055,
"step": 87
},
{
"epoch": 0.032473822593293046,
"grad_norm": 3.6329469680786133,
"learning_rate": 0.0001,
"loss": 5.42648458480835,
"step": 88
},
{
"epoch": 0.03284284330458047,
"grad_norm": 6.178943634033203,
"learning_rate": 0.0001,
"loss": 5.166748046875,
"step": 89
},
{
"epoch": 0.03321186401586789,
"grad_norm": 4.884125232696533,
"learning_rate": 0.0001,
"loss": 5.268251895904541,
"step": 90
},
{
"epoch": 0.033580884727155313,
"grad_norm": 4.925684452056885,
"learning_rate": 0.0001,
"loss": 5.187631130218506,
"step": 91
},
{
"epoch": 0.033949905438442736,
"grad_norm": 5.778391361236572,
"learning_rate": 0.0001,
"loss": 4.894795894622803,
"step": 92
},
{
"epoch": 0.03431892614973015,
"grad_norm": 4.655902862548828,
"learning_rate": 0.0001,
"loss": 5.2550225257873535,
"step": 93
},
{
"epoch": 0.034687946861017574,
"grad_norm": 5.219376564025879,
"learning_rate": 0.0001,
"loss": 5.362205505371094,
"step": 94
},
{
"epoch": 0.035056967572304996,
"grad_norm": 3.6282198429107666,
"learning_rate": 0.0001,
"loss": 5.384284019470215,
"step": 95
},
{
"epoch": 0.03542598828359242,
"grad_norm": 4.7581787109375,
"learning_rate": 0.0001,
"loss": 5.144690036773682,
"step": 96
},
{
"epoch": 0.035795008994879834,
"grad_norm": 3.4947946071624756,
"learning_rate": 0.0001,
"loss": 5.272378444671631,
"step": 97
},
{
"epoch": 0.03616402970616726,
"grad_norm": 4.599036693572998,
"learning_rate": 0.0001,
"loss": 4.959323406219482,
"step": 98
},
{
"epoch": 0.03653305041745468,
"grad_norm": 4.1109418869018555,
"learning_rate": 0.0001,
"loss": 5.293513774871826,
"step": 99
},
{
"epoch": 0.0369020711287421,
"grad_norm": 4.760090351104736,
"learning_rate": 0.0001,
"loss": 5.218464374542236,
"step": 100
},
{
"epoch": 0.037271091840029524,
"grad_norm": 3.34501314163208,
"learning_rate": 0.0001,
"loss": 5.282923698425293,
"step": 101
},
{
"epoch": 0.03764011255131694,
"grad_norm": 3.4620864391326904,
"learning_rate": 0.0001,
"loss": 5.078286170959473,
"step": 102
},
{
"epoch": 0.03800913326260436,
"grad_norm": 4.38137674331665,
"learning_rate": 0.0001,
"loss": 5.183228969573975,
"step": 103
},
{
"epoch": 0.038378153973891785,
"grad_norm": 4.770188331604004,
"learning_rate": 0.0001,
"loss": 5.21339750289917,
"step": 104
},
{
"epoch": 0.03874717468517921,
"grad_norm": 4.081193923950195,
"learning_rate": 0.0001,
"loss": 5.423886299133301,
"step": 105
},
{
"epoch": 0.03911619539646663,
"grad_norm": 4.0991010665893555,
"learning_rate": 0.0001,
"loss": 5.067556381225586,
"step": 106
},
{
"epoch": 0.039485216107754045,
"grad_norm": 4.759697437286377,
"learning_rate": 0.0001,
"loss": 5.147449016571045,
"step": 107
},
{
"epoch": 0.03985423681904147,
"grad_norm": 4.935500621795654,
"learning_rate": 0.0001,
"loss": 5.575446128845215,
"step": 108
},
{
"epoch": 0.04022325753032889,
"grad_norm": 4.266407012939453,
"learning_rate": 0.0001,
"loss": 5.134807586669922,
"step": 109
},
{
"epoch": 0.04059227824161631,
"grad_norm": 3.929640769958496,
"learning_rate": 0.0001,
"loss": 5.2398786544799805,
"step": 110
},
{
"epoch": 0.04096129895290373,
"grad_norm": 4.449552059173584,
"learning_rate": 0.0001,
"loss": 5.156042575836182,
"step": 111
},
{
"epoch": 0.04133031966419115,
"grad_norm": 3.860306739807129,
"learning_rate": 0.0001,
"loss": 4.952601432800293,
"step": 112
},
{
"epoch": 0.04169934037547857,
"grad_norm": 4.714514255523682,
"learning_rate": 0.0001,
"loss": 5.524988174438477,
"step": 113
},
{
"epoch": 0.042068361086765996,
"grad_norm": 4.664876461029053,
"learning_rate": 0.0001,
"loss": 4.995589256286621,
"step": 114
},
{
"epoch": 0.04243738179805342,
"grad_norm": 4.218993663787842,
"learning_rate": 0.0001,
"loss": 5.228067874908447,
"step": 115
},
{
"epoch": 0.042806402509340834,
"grad_norm": 4.040130138397217,
"learning_rate": 0.0001,
"loss": 5.018726348876953,
"step": 116
},
{
"epoch": 0.043175423220628256,
"grad_norm": 4.016865253448486,
"learning_rate": 0.0001,
"loss": 5.552411079406738,
"step": 117
},
{
"epoch": 0.04354444393191568,
"grad_norm": 4.608650207519531,
"learning_rate": 0.0001,
"loss": 5.315113067626953,
"step": 118
},
{
"epoch": 0.0439134646432031,
"grad_norm": 4.381293296813965,
"learning_rate": 0.0001,
"loss": 5.192525386810303,
"step": 119
},
{
"epoch": 0.044282485354490524,
"grad_norm": 3.6763710975646973,
"learning_rate": 0.0001,
"loss": 5.22324275970459,
"step": 120
},
{
"epoch": 0.04465150606577794,
"grad_norm": 4.947065830230713,
"learning_rate": 0.0001,
"loss": 5.236631870269775,
"step": 121
},
{
"epoch": 0.04502052677706536,
"grad_norm": 3.920776844024658,
"learning_rate": 0.0001,
"loss": 5.249703407287598,
"step": 122
},
{
"epoch": 0.045389547488352784,
"grad_norm": 2.573857545852661,
"learning_rate": 0.0001,
"loss": 4.9618330001831055,
"step": 123
},
{
"epoch": 0.04575856819964021,
"grad_norm": 3.178628921508789,
"learning_rate": 0.0001,
"loss": 5.127320766448975,
"step": 124
},
{
"epoch": 0.04612758891092763,
"grad_norm": 3.242992877960205,
"learning_rate": 0.0001,
"loss": 5.084701061248779,
"step": 125
},
{
"epoch": 0.046496609622215045,
"grad_norm": 3.685932159423828,
"learning_rate": 0.0001,
"loss": 5.189405918121338,
"step": 126
},
{
"epoch": 0.04686563033350247,
"grad_norm": 2.7726705074310303,
"learning_rate": 0.0001,
"loss": 5.2822041511535645,
"step": 127
},
{
"epoch": 0.04723465104478989,
"grad_norm": 3.1517961025238037,
"learning_rate": 0.0001,
"loss": 5.11806583404541,
"step": 128
},
{
"epoch": 0.04760367175607731,
"grad_norm": 3.016364097595215,
"learning_rate": 0.0001,
"loss": 5.419538974761963,
"step": 129
},
{
"epoch": 0.04797269246736473,
"grad_norm": 3.950747489929199,
"learning_rate": 0.0001,
"loss": 5.245120525360107,
"step": 130
},
{
"epoch": 0.04834171317865215,
"grad_norm": 3.8114898204803467,
"learning_rate": 0.0001,
"loss": 4.765509128570557,
"step": 131
},
{
"epoch": 0.04871073388993957,
"grad_norm": 3.9141814708709717,
"learning_rate": 0.0001,
"loss": 5.1021881103515625,
"step": 132
},
{
"epoch": 0.049079754601226995,
"grad_norm": 4.04232931137085,
"learning_rate": 0.0001,
"loss": 4.959789276123047,
"step": 133
},
{
"epoch": 0.04944877531251442,
"grad_norm": 2.6271684169769287,
"learning_rate": 0.0001,
"loss": 5.169219493865967,
"step": 134
},
{
"epoch": 0.04981779602380183,
"grad_norm": 3.3536503314971924,
"learning_rate": 0.0001,
"loss": 5.00529146194458,
"step": 135
},
{
"epoch": 0.050186816735089256,
"grad_norm": 4.068567276000977,
"learning_rate": 0.0001,
"loss": 4.860372543334961,
"step": 136
},
{
"epoch": 0.05055583744637668,
"grad_norm": 4.450196743011475,
"learning_rate": 0.0001,
"loss": 5.2418718338012695,
"step": 137
},
{
"epoch": 0.0509248581576641,
"grad_norm": 3.7781050205230713,
"learning_rate": 0.0001,
"loss": 5.009589672088623,
"step": 138
},
{
"epoch": 0.05129387886895152,
"grad_norm": 4.003530979156494,
"learning_rate": 0.0001,
"loss": 5.156592845916748,
"step": 139
},
{
"epoch": 0.05166289958023894,
"grad_norm": 4.052554130554199,
"learning_rate": 0.0001,
"loss": 5.234455585479736,
"step": 140
},
{
"epoch": 0.05203192029152636,
"grad_norm": 3.1616437435150146,
"learning_rate": 0.0001,
"loss": 4.669750213623047,
"step": 141
},
{
"epoch": 0.05240094100281378,
"grad_norm": 3.804659366607666,
"learning_rate": 0.0001,
"loss": 5.179162979125977,
"step": 142
},
{
"epoch": 0.052769961714101206,
"grad_norm": 3.394329071044922,
"learning_rate": 0.0001,
"loss": 5.047037601470947,
"step": 143
},
{
"epoch": 0.05313898242538863,
"grad_norm": 3.5721750259399414,
"learning_rate": 0.0001,
"loss": 5.083056926727295,
"step": 144
},
{
"epoch": 0.053508003136676044,
"grad_norm": 3.9708645343780518,
"learning_rate": 0.0001,
"loss": 5.124134063720703,
"step": 145
},
{
"epoch": 0.053877023847963466,
"grad_norm": 3.5934338569641113,
"learning_rate": 0.0001,
"loss": 5.215319633483887,
"step": 146
},
{
"epoch": 0.05424604455925089,
"grad_norm": 4.272321701049805,
"learning_rate": 0.0001,
"loss": 5.307074546813965,
"step": 147
},
{
"epoch": 0.05461506527053831,
"grad_norm": 3.845202922821045,
"learning_rate": 0.0001,
"loss": 4.817156791687012,
"step": 148
},
{
"epoch": 0.05498408598182573,
"grad_norm": 5.178013801574707,
"learning_rate": 0.0001,
"loss": 5.578018665313721,
"step": 149
},
{
"epoch": 0.05535310669311315,
"grad_norm": 4.21903133392334,
"learning_rate": 0.0001,
"loss": 5.05971622467041,
"step": 150
},
{
"epoch": 0.05572212740440057,
"grad_norm": 3.9957923889160156,
"learning_rate": 0.0001,
"loss": 5.0507731437683105,
"step": 151
},
{
"epoch": 0.056091148115687994,
"grad_norm": 4.366008281707764,
"learning_rate": 0.0001,
"loss": 4.8964433670043945,
"step": 152
},
{
"epoch": 0.05646016882697542,
"grad_norm": 5.036316871643066,
"learning_rate": 0.0001,
"loss": 4.990871429443359,
"step": 153
},
{
"epoch": 0.05682918953826283,
"grad_norm": 3.921596050262451,
"learning_rate": 0.0001,
"loss": 5.315304756164551,
"step": 154
},
{
"epoch": 0.057198210249550255,
"grad_norm": 3.0889859199523926,
"learning_rate": 0.0001,
"loss": 5.343647003173828,
"step": 155
},
{
"epoch": 0.05756723096083768,
"grad_norm": 3.3512563705444336,
"learning_rate": 0.0001,
"loss": 5.151203155517578,
"step": 156
},
{
"epoch": 0.0579362516721251,
"grad_norm": 3.630702018737793,
"learning_rate": 0.0001,
"loss": 4.823637008666992,
"step": 157
},
{
"epoch": 0.05830527238341252,
"grad_norm": 3.516183853149414,
"learning_rate": 0.0001,
"loss": 4.9256272315979,
"step": 158
},
{
"epoch": 0.05867429309469994,
"grad_norm": 3.8045616149902344,
"learning_rate": 0.0001,
"loss": 4.9611992835998535,
"step": 159
},
{
"epoch": 0.05904331380598736,
"grad_norm": 4.987512111663818,
"learning_rate": 0.0001,
"loss": 5.136240005493164,
"step": 160
},
{
"epoch": 0.05941233451727478,
"grad_norm": 3.295520305633545,
"learning_rate": 0.0001,
"loss": 4.939165115356445,
"step": 161
},
{
"epoch": 0.059781355228562205,
"grad_norm": 3.4286487102508545,
"learning_rate": 0.0001,
"loss": 5.1253533363342285,
"step": 162
},
{
"epoch": 0.06015037593984962,
"grad_norm": 3.3474066257476807,
"learning_rate": 0.0001,
"loss": 4.942944526672363,
"step": 163
},
{
"epoch": 0.06051939665113704,
"grad_norm": 3.4838764667510986,
"learning_rate": 0.0001,
"loss": 4.937647819519043,
"step": 164
},
{
"epoch": 0.060888417362424466,
"grad_norm": 3.3820011615753174,
"learning_rate": 0.0001,
"loss": 4.728400230407715,
"step": 165
},
{
"epoch": 0.06125743807371189,
"grad_norm": 3.039099931716919,
"learning_rate": 0.0001,
"loss": 5.29592227935791,
"step": 166
},
{
"epoch": 0.06162645878499931,
"grad_norm": 3.884432554244995,
"learning_rate": 0.0001,
"loss": 5.097470283508301,
"step": 167
},
{
"epoch": 0.061995479496286726,
"grad_norm": 3.927171468734741,
"learning_rate": 0.0001,
"loss": 5.287883758544922,
"step": 168
},
{
"epoch": 0.06236450020757415,
"grad_norm": 3.2847440242767334,
"learning_rate": 0.0001,
"loss": 4.842876434326172,
"step": 169
},
{
"epoch": 0.06273352091886157,
"grad_norm": 3.7847704887390137,
"learning_rate": 0.0001,
"loss": 5.119341850280762,
"step": 170
},
{
"epoch": 0.06310254163014899,
"grad_norm": 3.4630985260009766,
"learning_rate": 0.0001,
"loss": 5.680130958557129,
"step": 171
},
{
"epoch": 0.06347156234143642,
"grad_norm": 2.9577441215515137,
"learning_rate": 0.0001,
"loss": 5.309061050415039,
"step": 172
},
{
"epoch": 0.06384058305272383,
"grad_norm": 3.408936023712158,
"learning_rate": 0.0001,
"loss": 4.966146945953369,
"step": 173
},
{
"epoch": 0.06420960376401126,
"grad_norm": 2.8536202907562256,
"learning_rate": 0.0001,
"loss": 5.047423362731934,
"step": 174
},
{
"epoch": 0.06457862447529868,
"grad_norm": 2.683049201965332,
"learning_rate": 0.0001,
"loss": 5.021755218505859,
"step": 175
},
{
"epoch": 0.06494764518658609,
"grad_norm": 8.568255424499512,
"learning_rate": 0.0001,
"loss": 5.02105712890625,
"step": 176
},
{
"epoch": 0.06531666589787352,
"grad_norm": 3.26704478263855,
"learning_rate": 0.0001,
"loss": 4.734854221343994,
"step": 177
},
{
"epoch": 0.06568568660916094,
"grad_norm": 3.1402342319488525,
"learning_rate": 0.0001,
"loss": 5.746152400970459,
"step": 178
},
{
"epoch": 0.06605470732044837,
"grad_norm": 3.1304996013641357,
"learning_rate": 0.0001,
"loss": 5.069252014160156,
"step": 179
},
{
"epoch": 0.06642372803173578,
"grad_norm": 3.1906208992004395,
"learning_rate": 0.0001,
"loss": 5.060096263885498,
"step": 180
},
{
"epoch": 0.0667927487430232,
"grad_norm": 3.0051543712615967,
"learning_rate": 0.0001,
"loss": 5.333315849304199,
"step": 181
},
{
"epoch": 0.06716176945431063,
"grad_norm": 2.9292314052581787,
"learning_rate": 0.0001,
"loss": 4.98300838470459,
"step": 182
},
{
"epoch": 0.06753079016559804,
"grad_norm": 3.8607776165008545,
"learning_rate": 0.0001,
"loss": 5.22333288192749,
"step": 183
},
{
"epoch": 0.06789981087688547,
"grad_norm": 5.044845104217529,
"learning_rate": 0.0001,
"loss": 4.9512128829956055,
"step": 184
},
{
"epoch": 0.06826883158817289,
"grad_norm": 3.930791139602661,
"learning_rate": 0.0001,
"loss": 4.994528770446777,
"step": 185
},
{
"epoch": 0.0686378522994603,
"grad_norm": 4.407391548156738,
"learning_rate": 0.0001,
"loss": 5.1336493492126465,
"step": 186
},
{
"epoch": 0.06900687301074773,
"grad_norm": 3.2872698307037354,
"learning_rate": 0.0001,
"loss": 4.903454303741455,
"step": 187
},
{
"epoch": 0.06937589372203515,
"grad_norm": 2.9913930892944336,
"learning_rate": 0.0001,
"loss": 4.907477378845215,
"step": 188
},
{
"epoch": 0.06974491443332256,
"grad_norm": 3.690415382385254,
"learning_rate": 0.0001,
"loss": 5.368701457977295,
"step": 189
},
{
"epoch": 0.07011393514460999,
"grad_norm": 3.2525527477264404,
"learning_rate": 0.0001,
"loss": 4.97088623046875,
"step": 190
},
{
"epoch": 0.07048295585589741,
"grad_norm": 3.62608003616333,
"learning_rate": 0.0001,
"loss": 5.350979328155518,
"step": 191
},
{
"epoch": 0.07085197656718484,
"grad_norm": 2.343998908996582,
"learning_rate": 0.0001,
"loss": 4.678924083709717,
"step": 192
},
{
"epoch": 0.07122099727847225,
"grad_norm": 2.5523736476898193,
"learning_rate": 0.0001,
"loss": 5.1422343254089355,
"step": 193
},
{
"epoch": 0.07159001798975967,
"grad_norm": 3.007969856262207,
"learning_rate": 0.0001,
"loss": 5.103295803070068,
"step": 194
},
{
"epoch": 0.0719590387010471,
"grad_norm": 2.446394205093384,
"learning_rate": 0.0001,
"loss": 4.8293867111206055,
"step": 195
},
{
"epoch": 0.07232805941233451,
"grad_norm": 2.6612651348114014,
"learning_rate": 0.0001,
"loss": 5.177353858947754,
"step": 196
},
{
"epoch": 0.07269708012362194,
"grad_norm": 2.940788507461548,
"learning_rate": 0.0001,
"loss": 4.852959156036377,
"step": 197
},
{
"epoch": 0.07306610083490936,
"grad_norm": 3.0821218490600586,
"learning_rate": 0.0001,
"loss": 4.744450569152832,
"step": 198
},
{
"epoch": 0.07343512154619677,
"grad_norm": 3.23681640625,
"learning_rate": 0.0001,
"loss": 5.011826515197754,
"step": 199
},
{
"epoch": 0.0738041422574842,
"grad_norm": 2.6175684928894043,
"learning_rate": 0.0001,
"loss": 5.107472896575928,
"step": 200
},
{
"epoch": 0.07417316296877162,
"grad_norm": 3.0995991230010986,
"learning_rate": 0.0001,
"loss": 4.885207653045654,
"step": 201
},
{
"epoch": 0.07454218368005905,
"grad_norm": 2.8786182403564453,
"learning_rate": 0.0001,
"loss": 4.845186233520508,
"step": 202
},
{
"epoch": 0.07491120439134646,
"grad_norm": 3.120236396789551,
"learning_rate": 0.0001,
"loss": 4.899354457855225,
"step": 203
},
{
"epoch": 0.07528022510263388,
"grad_norm": 2.4766738414764404,
"learning_rate": 0.0001,
"loss": 5.172772407531738,
"step": 204
},
{
"epoch": 0.07564924581392131,
"grad_norm": 2.8088064193725586,
"learning_rate": 0.0001,
"loss": 4.881832599639893,
"step": 205
},
{
"epoch": 0.07601826652520872,
"grad_norm": 3.229881763458252,
"learning_rate": 0.0001,
"loss": 5.073505401611328,
"step": 206
},
{
"epoch": 0.07638728723649615,
"grad_norm": 3.5004501342773438,
"learning_rate": 0.0001,
"loss": 4.852757930755615,
"step": 207
},
{
"epoch": 0.07675630794778357,
"grad_norm": 3.0608553886413574,
"learning_rate": 0.0001,
"loss": 4.712113380432129,
"step": 208
},
{
"epoch": 0.07712532865907099,
"grad_norm": 3.229390859603882,
"learning_rate": 0.0001,
"loss": 4.8641276359558105,
"step": 209
},
{
"epoch": 0.07749434937035841,
"grad_norm": 2.9614827632904053,
"learning_rate": 0.0001,
"loss": 5.257576942443848,
"step": 210
},
{
"epoch": 0.07786337008164583,
"grad_norm": 2.8287434577941895,
"learning_rate": 0.0001,
"loss": 4.992201805114746,
"step": 211
},
{
"epoch": 0.07823239079293326,
"grad_norm": 2.8483452796936035,
"learning_rate": 0.0001,
"loss": 4.637844562530518,
"step": 212
},
{
"epoch": 0.07860141150422068,
"grad_norm": 2.9546141624450684,
"learning_rate": 0.0001,
"loss": 4.944921016693115,
"step": 213
},
{
"epoch": 0.07897043221550809,
"grad_norm": 4.089045524597168,
"learning_rate": 0.0001,
"loss": 5.075927734375,
"step": 214
},
{
"epoch": 0.07933945292679552,
"grad_norm": 2.8543574810028076,
"learning_rate": 0.0001,
"loss": 4.585363388061523,
"step": 215
},
{
"epoch": 0.07970847363808294,
"grad_norm": 2.697917938232422,
"learning_rate": 0.0001,
"loss": 5.132508277893066,
"step": 216
},
{
"epoch": 0.08007749434937037,
"grad_norm": 2.9316647052764893,
"learning_rate": 0.0001,
"loss": 5.069299221038818,
"step": 217
},
{
"epoch": 0.08044651506065778,
"grad_norm": 2.8044543266296387,
"learning_rate": 0.0001,
"loss": 4.850956439971924,
"step": 218
},
{
"epoch": 0.0808155357719452,
"grad_norm": 2.7793681621551514,
"learning_rate": 0.0001,
"loss": 5.32306432723999,
"step": 219
},
{
"epoch": 0.08118455648323263,
"grad_norm": 2.586214542388916,
"learning_rate": 0.0001,
"loss": 4.600836277008057,
"step": 220
},
{
"epoch": 0.08155357719452004,
"grad_norm": 3.0345284938812256,
"learning_rate": 0.0001,
"loss": 4.788100719451904,
"step": 221
},
{
"epoch": 0.08192259790580746,
"grad_norm": 2.685389518737793,
"learning_rate": 0.0001,
"loss": 4.780126571655273,
"step": 222
},
{
"epoch": 0.08229161861709489,
"grad_norm": 2.843975305557251,
"learning_rate": 0.0001,
"loss": 4.858603000640869,
"step": 223
},
{
"epoch": 0.0826606393283823,
"grad_norm": 3.5730929374694824,
"learning_rate": 0.0001,
"loss": 4.825096130371094,
"step": 224
},
{
"epoch": 0.08302966003966973,
"grad_norm": 3.2879269123077393,
"learning_rate": 0.0001,
"loss": 4.757035255432129,
"step": 225
},
{
"epoch": 0.08339868075095715,
"grad_norm": 3.235781669616699,
"learning_rate": 0.0001,
"loss": 5.257933139801025,
"step": 226
},
{
"epoch": 0.08376770146224456,
"grad_norm": 2.9078941345214844,
"learning_rate": 0.0001,
"loss": 4.914604663848877,
"step": 227
},
{
"epoch": 0.08413672217353199,
"grad_norm": 2.557096004486084,
"learning_rate": 0.0001,
"loss": 4.683157920837402,
"step": 228
},
{
"epoch": 0.08450574288481941,
"grad_norm": 3.4715662002563477,
"learning_rate": 0.0001,
"loss": 4.910725116729736,
"step": 229
},
{
"epoch": 0.08487476359610684,
"grad_norm": 3.880197048187256,
"learning_rate": 0.0001,
"loss": 4.976219177246094,
"step": 230
},
{
"epoch": 0.08524378430739425,
"grad_norm": 3.1803884506225586,
"learning_rate": 0.0001,
"loss": 4.871642112731934,
"step": 231
},
{
"epoch": 0.08561280501868167,
"grad_norm": 5.288760662078857,
"learning_rate": 0.0001,
"loss": 4.9952287673950195,
"step": 232
},
{
"epoch": 0.0859818257299691,
"grad_norm": 3.4574294090270996,
"learning_rate": 0.0001,
"loss": 5.026848793029785,
"step": 233
},
{
"epoch": 0.08635084644125651,
"grad_norm": 2.983919382095337,
"learning_rate": 0.0001,
"loss": 4.823044776916504,
"step": 234
},
{
"epoch": 0.08671986715254394,
"grad_norm": 2.493537664413452,
"learning_rate": 0.0001,
"loss": 4.764354705810547,
"step": 235
},
{
"epoch": 0.08708888786383136,
"grad_norm": 3.7901830673217773,
"learning_rate": 0.0001,
"loss": 5.213089942932129,
"step": 236
},
{
"epoch": 0.08745790857511877,
"grad_norm": 2.74006986618042,
"learning_rate": 0.0001,
"loss": 4.810093879699707,
"step": 237
},
{
"epoch": 0.0878269292864062,
"grad_norm": 2.547126531600952,
"learning_rate": 0.0001,
"loss": 4.611993312835693,
"step": 238
},
{
"epoch": 0.08819594999769362,
"grad_norm": 2.8028151988983154,
"learning_rate": 0.0001,
"loss": 4.800611972808838,
"step": 239
},
{
"epoch": 0.08856497070898105,
"grad_norm": 2.3688666820526123,
"learning_rate": 0.0001,
"loss": 4.928980827331543,
"step": 240
},
{
"epoch": 0.08893399142026846,
"grad_norm": 2.43617844581604,
"learning_rate": 0.0001,
"loss": 5.174886703491211,
"step": 241
},
{
"epoch": 0.08930301213155588,
"grad_norm": 4.133114814758301,
"learning_rate": 0.0001,
"loss": 4.833251476287842,
"step": 242
},
{
"epoch": 0.08967203284284331,
"grad_norm": 2.575796365737915,
"learning_rate": 0.0001,
"loss": 5.019968032836914,
"step": 243
},
{
"epoch": 0.09004105355413072,
"grad_norm": 3.121090888977051,
"learning_rate": 0.0001,
"loss": 4.538406848907471,
"step": 244
},
{
"epoch": 0.09041007426541815,
"grad_norm": 3.4044008255004883,
"learning_rate": 0.0001,
"loss": 5.054136753082275,
"step": 245
},
{
"epoch": 0.09077909497670557,
"grad_norm": 2.7993178367614746,
"learning_rate": 0.0001,
"loss": 4.802661418914795,
"step": 246
},
{
"epoch": 0.09114811568799298,
"grad_norm": 2.870140552520752,
"learning_rate": 0.0001,
"loss": 5.112711429595947,
"step": 247
},
{
"epoch": 0.09151713639928041,
"grad_norm": 3.2912027835845947,
"learning_rate": 0.0001,
"loss": 4.869266510009766,
"step": 248
},
{
"epoch": 0.09188615711056783,
"grad_norm": 2.9518210887908936,
"learning_rate": 0.0001,
"loss": 4.794067859649658,
"step": 249
},
{
"epoch": 0.09225517782185526,
"grad_norm": 2.7578539848327637,
"learning_rate": 0.0001,
"loss": 4.8072967529296875,
"step": 250
},
{
"epoch": 0.09262419853314267,
"grad_norm": 2.4775185585021973,
"learning_rate": 0.0001,
"loss": 4.741370677947998,
"step": 251
},
{
"epoch": 0.09299321924443009,
"grad_norm": 2.850309133529663,
"learning_rate": 0.0001,
"loss": 4.818919658660889,
"step": 252
},
{
"epoch": 0.09336223995571752,
"grad_norm": 3.1108920574188232,
"learning_rate": 0.0001,
"loss": 5.032450199127197,
"step": 253
},
{
"epoch": 0.09373126066700493,
"grad_norm": 2.813636541366577,
"learning_rate": 0.0001,
"loss": 4.874199867248535,
"step": 254
},
{
"epoch": 0.09410028137829235,
"grad_norm": 2.8555026054382324,
"learning_rate": 0.0001,
"loss": 4.985494136810303,
"step": 255
},
{
"epoch": 0.09446930208957978,
"grad_norm": 2.5190069675445557,
"learning_rate": 0.0001,
"loss": 4.894311904907227,
"step": 256
},
{
"epoch": 0.0948383228008672,
"grad_norm": 2.6791582107543945,
"learning_rate": 0.0001,
"loss": 4.945681095123291,
"step": 257
},
{
"epoch": 0.09520734351215462,
"grad_norm": 2.9527032375335693,
"learning_rate": 0.0001,
"loss": 4.8298468589782715,
"step": 258
},
{
"epoch": 0.09557636422344204,
"grad_norm": 2.7807869911193848,
"learning_rate": 0.0001,
"loss": 4.615987777709961,
"step": 259
},
{
"epoch": 0.09594538493472946,
"grad_norm": 2.682555675506592,
"learning_rate": 0.0001,
"loss": 4.74656867980957,
"step": 260
},
{
"epoch": 0.09631440564601688,
"grad_norm": 3.3544092178344727,
"learning_rate": 0.0001,
"loss": 4.863458633422852,
"step": 261
},
{
"epoch": 0.0966834263573043,
"grad_norm": 2.737848997116089,
"learning_rate": 0.0001,
"loss": 4.7740397453308105,
"step": 262
},
{
"epoch": 0.09705244706859173,
"grad_norm": 2.7295894622802734,
"learning_rate": 0.0001,
"loss": 4.383846759796143,
"step": 263
},
{
"epoch": 0.09742146777987915,
"grad_norm": 2.332279682159424,
"learning_rate": 0.0001,
"loss": 4.982947826385498,
"step": 264
},
{
"epoch": 0.09779048849116656,
"grad_norm": 2.687039613723755,
"learning_rate": 0.0001,
"loss": 4.721158981323242,
"step": 265
},
{
"epoch": 0.09815950920245399,
"grad_norm": 2.505675792694092,
"learning_rate": 0.0001,
"loss": 4.891917705535889,
"step": 266
},
{
"epoch": 0.0985285299137414,
"grad_norm": 2.8479607105255127,
"learning_rate": 0.0001,
"loss": 4.618453502655029,
"step": 267
},
{
"epoch": 0.09889755062502883,
"grad_norm": 2.6204135417938232,
"learning_rate": 0.0001,
"loss": 4.758787631988525,
"step": 268
},
{
"epoch": 0.09926657133631625,
"grad_norm": 3.021299362182617,
"learning_rate": 0.0001,
"loss": 4.7189507484436035,
"step": 269
},
{
"epoch": 0.09963559204760367,
"grad_norm": 2.4360415935516357,
"learning_rate": 0.0001,
"loss": 4.794710636138916,
"step": 270
},
{
"epoch": 0.1000046127588911,
"grad_norm": 2.366312026977539,
"learning_rate": 0.0001,
"loss": 4.708616256713867,
"step": 271
},
{
"epoch": 0.10037363347017851,
"grad_norm": 2.5458579063415527,
"learning_rate": 0.0001,
"loss": 4.5073394775390625,
"step": 272
},
{
"epoch": 0.10074265418146594,
"grad_norm": 2.9471662044525146,
"learning_rate": 0.0001,
"loss": 4.495577335357666,
"step": 273
},
{
"epoch": 0.10111167489275336,
"grad_norm": 3.0150680541992188,
"learning_rate": 0.0001,
"loss": 4.821500301361084,
"step": 274
},
{
"epoch": 0.10148069560404077,
"grad_norm": 2.578234910964966,
"learning_rate": 0.0001,
"loss": 4.77611780166626,
"step": 275
},
{
"epoch": 0.1018497163153282,
"grad_norm": 2.579213857650757,
"learning_rate": 0.0001,
"loss": 4.857175350189209,
"step": 276
},
{
"epoch": 0.10221873702661562,
"grad_norm": 2.65431547164917,
"learning_rate": 0.0001,
"loss": 4.517264366149902,
"step": 277
},
{
"epoch": 0.10258775773790305,
"grad_norm": 3.1191048622131348,
"learning_rate": 0.0001,
"loss": 5.008939266204834,
"step": 278
},
{
"epoch": 0.10295677844919046,
"grad_norm": 2.8840596675872803,
"learning_rate": 0.0001,
"loss": 4.874368190765381,
"step": 279
},
{
"epoch": 0.10332579916047788,
"grad_norm": 2.867748975753784,
"learning_rate": 0.0001,
"loss": 4.8059797286987305,
"step": 280
},
{
"epoch": 0.1036948198717653,
"grad_norm": 3.3783462047576904,
"learning_rate": 0.0001,
"loss": 4.704605579376221,
"step": 281
},
{
"epoch": 0.10406384058305272,
"grad_norm": 2.759956121444702,
"learning_rate": 0.0001,
"loss": 4.987379550933838,
"step": 282
},
{
"epoch": 0.10443286129434015,
"grad_norm": 2.6871654987335205,
"learning_rate": 0.0001,
"loss": 4.723685264587402,
"step": 283
},
{
"epoch": 0.10480188200562757,
"grad_norm": 2.8111624717712402,
"learning_rate": 0.0001,
"loss": 4.588558673858643,
"step": 284
},
{
"epoch": 0.10517090271691498,
"grad_norm": 2.244652509689331,
"learning_rate": 0.0001,
"loss": 4.661073207855225,
"step": 285
},
{
"epoch": 0.10553992342820241,
"grad_norm": 2.8253273963928223,
"learning_rate": 0.0001,
"loss": 4.67581033706665,
"step": 286
},
{
"epoch": 0.10590894413948983,
"grad_norm": 2.362382650375366,
"learning_rate": 0.0001,
"loss": 4.8821916580200195,
"step": 287
},
{
"epoch": 0.10627796485077726,
"grad_norm": 2.725433111190796,
"learning_rate": 0.0001,
"loss": 4.793010234832764,
"step": 288
},
{
"epoch": 0.10664698556206467,
"grad_norm": 2.443126678466797,
"learning_rate": 0.0001,
"loss": 4.6530046463012695,
"step": 289
},
{
"epoch": 0.10701600627335209,
"grad_norm": 2.5819406509399414,
"learning_rate": 0.0001,
"loss": 5.03138542175293,
"step": 290
},
{
"epoch": 0.10738502698463952,
"grad_norm": 2.345172166824341,
"learning_rate": 0.0001,
"loss": 4.908970832824707,
"step": 291
},
{
"epoch": 0.10775404769592693,
"grad_norm": 3.7422969341278076,
"learning_rate": 0.0001,
"loss": 4.663451194763184,
"step": 292
},
{
"epoch": 0.10812306840721435,
"grad_norm": 2.690535068511963,
"learning_rate": 0.0001,
"loss": 4.976020812988281,
"step": 293
},
{
"epoch": 0.10849208911850178,
"grad_norm": 3.6050312519073486,
"learning_rate": 0.0001,
"loss": 4.173515319824219,
"step": 294
},
{
"epoch": 0.1088611098297892,
"grad_norm": 2.6177196502685547,
"learning_rate": 0.0001,
"loss": 4.570525169372559,
"step": 295
},
{
"epoch": 0.10923013054107662,
"grad_norm": 3.2142410278320312,
"learning_rate": 0.0001,
"loss": 5.103811740875244,
"step": 296
},
{
"epoch": 0.10959915125236404,
"grad_norm": 4.097013473510742,
"learning_rate": 0.0001,
"loss": 4.417541027069092,
"step": 297
},
{
"epoch": 0.10996817196365145,
"grad_norm": 2.6971688270568848,
"learning_rate": 0.0001,
"loss": 4.503973007202148,
"step": 298
},
{
"epoch": 0.11033719267493888,
"grad_norm": 2.4972405433654785,
"learning_rate": 0.0001,
"loss": 4.668544769287109,
"step": 299
},
{
"epoch": 0.1107062133862263,
"grad_norm": 3.145050287246704,
"learning_rate": 0.0001,
"loss": 4.453399658203125,
"step": 300
},
{
"epoch": 0.11107523409751373,
"grad_norm": 2.5468809604644775,
"learning_rate": 0.0001,
"loss": 4.50095272064209,
"step": 301
},
{
"epoch": 0.11144425480880114,
"grad_norm": 2.6690492630004883,
"learning_rate": 0.0001,
"loss": 4.6230950355529785,
"step": 302
},
{
"epoch": 0.11181327552008856,
"grad_norm": 2.255617141723633,
"learning_rate": 0.0001,
"loss": 4.648458003997803,
"step": 303
},
{
"epoch": 0.11218229623137599,
"grad_norm": 2.83013916015625,
"learning_rate": 0.0001,
"loss": 4.58681583404541,
"step": 304
},
{
"epoch": 0.1125513169426634,
"grad_norm": 2.867711067199707,
"learning_rate": 0.0001,
"loss": 4.499338150024414,
"step": 305
},
{
"epoch": 0.11292033765395083,
"grad_norm": 2.580369234085083,
"learning_rate": 0.0001,
"loss": 4.6251220703125,
"step": 306
},
{
"epoch": 0.11328935836523825,
"grad_norm": 2.3685848712921143,
"learning_rate": 0.0001,
"loss": 4.643671989440918,
"step": 307
},
{
"epoch": 0.11365837907652566,
"grad_norm": 3.4852209091186523,
"learning_rate": 0.0001,
"loss": 4.932046890258789,
"step": 308
},
{
"epoch": 0.1140273997878131,
"grad_norm": 2.830315351486206,
"learning_rate": 0.0001,
"loss": 4.5891923904418945,
"step": 309
},
{
"epoch": 0.11439642049910051,
"grad_norm": 3.2897260189056396,
"learning_rate": 0.0001,
"loss": 4.485490798950195,
"step": 310
},
{
"epoch": 0.11476544121038794,
"grad_norm": 2.5602848529815674,
"learning_rate": 0.0001,
"loss": 4.726596832275391,
"step": 311
},
{
"epoch": 0.11513446192167535,
"grad_norm": 2.8062198162078857,
"learning_rate": 0.0001,
"loss": 4.881383895874023,
"step": 312
},
{
"epoch": 0.11550348263296277,
"grad_norm": 2.9454972743988037,
"learning_rate": 0.0001,
"loss": 4.755986213684082,
"step": 313
},
{
"epoch": 0.1158725033442502,
"grad_norm": 2.8830761909484863,
"learning_rate": 0.0001,
"loss": 5.087071418762207,
"step": 314
},
{
"epoch": 0.11624152405553762,
"grad_norm": 2.050590991973877,
"learning_rate": 0.0001,
"loss": 4.456117630004883,
"step": 315
},
{
"epoch": 0.11661054476682504,
"grad_norm": 2.8796334266662598,
"learning_rate": 0.0001,
"loss": 4.842302322387695,
"step": 316
},
{
"epoch": 0.11697956547811246,
"grad_norm": 2.1899056434631348,
"learning_rate": 0.0001,
"loss": 4.674966335296631,
"step": 317
},
{
"epoch": 0.11734858618939988,
"grad_norm": 2.62785267829895,
"learning_rate": 0.0001,
"loss": 4.86789608001709,
"step": 318
},
{
"epoch": 0.1177176069006873,
"grad_norm": 2.5556700229644775,
"learning_rate": 0.0001,
"loss": 4.4173264503479,
"step": 319
},
{
"epoch": 0.11808662761197472,
"grad_norm": 3.35113525390625,
"learning_rate": 0.0001,
"loss": 4.507681369781494,
"step": 320
},
{
"epoch": 0.11845564832326215,
"grad_norm": 3.0669710636138916,
"learning_rate": 0.0001,
"loss": 4.245882987976074,
"step": 321
},
{
"epoch": 0.11882466903454957,
"grad_norm": 2.4834563732147217,
"learning_rate": 0.0001,
"loss": 4.907164573669434,
"step": 322
},
{
"epoch": 0.11919368974583698,
"grad_norm": 2.5204923152923584,
"learning_rate": 0.0001,
"loss": 4.891724586486816,
"step": 323
},
{
"epoch": 0.11956271045712441,
"grad_norm": 2.367497205734253,
"learning_rate": 0.0001,
"loss": 4.811779022216797,
"step": 324
},
{
"epoch": 0.11993173116841183,
"grad_norm": 2.360835552215576,
"learning_rate": 0.0001,
"loss": 4.670647144317627,
"step": 325
},
{
"epoch": 0.12030075187969924,
"grad_norm": 2.2293601036071777,
"learning_rate": 0.0001,
"loss": 4.589348793029785,
"step": 326
},
{
"epoch": 0.12066977259098667,
"grad_norm": 2.154679775238037,
"learning_rate": 0.0001,
"loss": 4.731776237487793,
"step": 327
},
{
"epoch": 0.12103879330227409,
"grad_norm": 2.914649486541748,
"learning_rate": 0.0001,
"loss": 4.931624889373779,
"step": 328
},
{
"epoch": 0.12140781401356152,
"grad_norm": 2.734717845916748,
"learning_rate": 0.0001,
"loss": 4.975724697113037,
"step": 329
},
{
"epoch": 0.12177683472484893,
"grad_norm": 2.389617443084717,
"learning_rate": 0.0001,
"loss": 4.796389102935791,
"step": 330
},
{
"epoch": 0.12214585543613635,
"grad_norm": 1.979083776473999,
"learning_rate": 0.0001,
"loss": 4.383009433746338,
"step": 331
},
{
"epoch": 0.12251487614742378,
"grad_norm": 2.6853559017181396,
"learning_rate": 0.0001,
"loss": 4.8341898918151855,
"step": 332
},
{
"epoch": 0.12288389685871119,
"grad_norm": 2.5915274620056152,
"learning_rate": 0.0001,
"loss": 4.717322826385498,
"step": 333
},
{
"epoch": 0.12325291756999862,
"grad_norm": 2.812741994857788,
"learning_rate": 0.0001,
"loss": 4.969723701477051,
"step": 334
},
{
"epoch": 0.12362193828128604,
"grad_norm": 2.57147216796875,
"learning_rate": 0.0001,
"loss": 4.582840919494629,
"step": 335
},
{
"epoch": 0.12399095899257345,
"grad_norm": 2.5397863388061523,
"learning_rate": 0.0001,
"loss": 4.691579341888428,
"step": 336
},
{
"epoch": 0.12435997970386088,
"grad_norm": 2.942898988723755,
"learning_rate": 0.0001,
"loss": 4.7678728103637695,
"step": 337
},
{
"epoch": 0.1247290004151483,
"grad_norm": 2.3598756790161133,
"learning_rate": 0.0001,
"loss": 4.659327030181885,
"step": 338
},
{
"epoch": 0.12509802112643573,
"grad_norm": 2.3905725479125977,
"learning_rate": 0.0001,
"loss": 4.716471195220947,
"step": 339
},
{
"epoch": 0.12546704183772314,
"grad_norm": 2.3143739700317383,
"learning_rate": 0.0001,
"loss": 4.569877624511719,
"step": 340
},
{
"epoch": 0.12583606254901056,
"grad_norm": 2.9056665897369385,
"learning_rate": 0.0001,
"loss": 4.622424602508545,
"step": 341
},
{
"epoch": 0.12620508326029797,
"grad_norm": 2.454422950744629,
"learning_rate": 0.0001,
"loss": 4.7714080810546875,
"step": 342
},
{
"epoch": 0.12657410397158542,
"grad_norm": 2.410001516342163,
"learning_rate": 0.0001,
"loss": 4.591516494750977,
"step": 343
},
{
"epoch": 0.12694312468287283,
"grad_norm": 4.945067405700684,
"learning_rate": 0.0001,
"loss": 4.671370983123779,
"step": 344
},
{
"epoch": 0.12731214539416025,
"grad_norm": 2.9925453662872314,
"learning_rate": 0.0001,
"loss": 4.712247848510742,
"step": 345
},
{
"epoch": 0.12768116610544766,
"grad_norm": 2.5714759826660156,
"learning_rate": 0.0001,
"loss": 4.530534744262695,
"step": 346
},
{
"epoch": 0.12805018681673508,
"grad_norm": 2.7953736782073975,
"learning_rate": 0.0001,
"loss": 4.63610315322876,
"step": 347
},
{
"epoch": 0.12841920752802252,
"grad_norm": 2.2027106285095215,
"learning_rate": 0.0001,
"loss": 4.753771781921387,
"step": 348
},
{
"epoch": 0.12878822823930994,
"grad_norm": 2.802049398422241,
"learning_rate": 0.0001,
"loss": 4.764090061187744,
"step": 349
},
{
"epoch": 0.12915724895059735,
"grad_norm": 2.5809788703918457,
"learning_rate": 0.0001,
"loss": 4.627032279968262,
"step": 350
},
{
"epoch": 0.12952626966188477,
"grad_norm": 2.522502899169922,
"learning_rate": 0.0001,
"loss": 4.610208988189697,
"step": 351
},
{
"epoch": 0.12989529037317218,
"grad_norm": 2.322376012802124,
"learning_rate": 0.0001,
"loss": 4.736202239990234,
"step": 352
},
{
"epoch": 0.13026431108445963,
"grad_norm": 2.6052095890045166,
"learning_rate": 0.0001,
"loss": 4.679316997528076,
"step": 353
},
{
"epoch": 0.13063333179574704,
"grad_norm": 2.5171854496002197,
"learning_rate": 0.0001,
"loss": 4.587606906890869,
"step": 354
},
{
"epoch": 0.13100235250703446,
"grad_norm": 2.7908267974853516,
"learning_rate": 0.0001,
"loss": 4.818106651306152,
"step": 355
},
{
"epoch": 0.13137137321832187,
"grad_norm": 2.3033812046051025,
"learning_rate": 0.0001,
"loss": 4.679779529571533,
"step": 356
},
{
"epoch": 0.1317403939296093,
"grad_norm": 2.204928159713745,
"learning_rate": 0.0001,
"loss": 4.498159408569336,
"step": 357
},
{
"epoch": 0.13210941464089673,
"grad_norm": 2.2332544326782227,
"learning_rate": 0.0001,
"loss": 4.56221866607666,
"step": 358
},
{
"epoch": 0.13247843535218415,
"grad_norm": 2.151496648788452,
"learning_rate": 0.0001,
"loss": 4.715498924255371,
"step": 359
},
{
"epoch": 0.13284745606347156,
"grad_norm": 2.1087806224823,
"learning_rate": 0.0001,
"loss": 4.749574184417725,
"step": 360
},
{
"epoch": 0.13321647677475898,
"grad_norm": 2.1617531776428223,
"learning_rate": 0.0001,
"loss": 4.659748077392578,
"step": 361
},
{
"epoch": 0.1335854974860464,
"grad_norm": 2.515841245651245,
"learning_rate": 0.0001,
"loss": 4.4495697021484375,
"step": 362
},
{
"epoch": 0.13395451819733384,
"grad_norm": 2.170407772064209,
"learning_rate": 0.0001,
"loss": 4.438388347625732,
"step": 363
},
{
"epoch": 0.13432353890862125,
"grad_norm": 2.7017438411712646,
"learning_rate": 0.0001,
"loss": 4.622103214263916,
"step": 364
},
{
"epoch": 0.13469255961990867,
"grad_norm": 3.747965097427368,
"learning_rate": 0.0001,
"loss": 4.560365676879883,
"step": 365
},
{
"epoch": 0.13506158033119608,
"grad_norm": 2.727411985397339,
"learning_rate": 0.0001,
"loss": 4.737141132354736,
"step": 366
},
{
"epoch": 0.1354306010424835,
"grad_norm": 2.811324119567871,
"learning_rate": 0.0001,
"loss": 4.603142261505127,
"step": 367
},
{
"epoch": 0.13579962175377094,
"grad_norm": 2.113163948059082,
"learning_rate": 0.0001,
"loss": 4.581015110015869,
"step": 368
},
{
"epoch": 0.13616864246505836,
"grad_norm": 2.3354477882385254,
"learning_rate": 0.0001,
"loss": 4.416581630706787,
"step": 369
},
{
"epoch": 0.13653766317634577,
"grad_norm": 1.9050804376602173,
"learning_rate": 0.0001,
"loss": 4.300870418548584,
"step": 370
},
{
"epoch": 0.1369066838876332,
"grad_norm": 2.382469654083252,
"learning_rate": 0.0001,
"loss": 4.5361223220825195,
"step": 371
},
{
"epoch": 0.1372757045989206,
"grad_norm": 2.4976327419281006,
"learning_rate": 0.0001,
"loss": 4.676791191101074,
"step": 372
},
{
"epoch": 0.13764472531020805,
"grad_norm": 2.061070680618286,
"learning_rate": 0.0001,
"loss": 4.431769371032715,
"step": 373
},
{
"epoch": 0.13801374602149546,
"grad_norm": 2.2746779918670654,
"learning_rate": 0.0001,
"loss": 4.311607837677002,
"step": 374
},
{
"epoch": 0.13838276673278288,
"grad_norm": 2.6687610149383545,
"learning_rate": 0.0001,
"loss": 5.022364139556885,
"step": 375
},
{
"epoch": 0.1387517874440703,
"grad_norm": 2.4456582069396973,
"learning_rate": 0.0001,
"loss": 4.58386754989624,
"step": 376
},
{
"epoch": 0.1391208081553577,
"grad_norm": 3.6035640239715576,
"learning_rate": 0.0001,
"loss": 4.901766777038574,
"step": 377
},
{
"epoch": 0.13948982886664513,
"grad_norm": 2.442969799041748,
"learning_rate": 0.0001,
"loss": 4.6150665283203125,
"step": 378
},
{
"epoch": 0.13985884957793257,
"grad_norm": 2.26698899269104,
"learning_rate": 0.0001,
"loss": 4.721152305603027,
"step": 379
},
{
"epoch": 0.14022787028921999,
"grad_norm": 2.289684534072876,
"learning_rate": 0.0001,
"loss": 4.594067573547363,
"step": 380
},
{
"epoch": 0.1405968910005074,
"grad_norm": 2.3382251262664795,
"learning_rate": 0.0001,
"loss": 4.569343566894531,
"step": 381
},
{
"epoch": 0.14096591171179482,
"grad_norm": 2.277287006378174,
"learning_rate": 0.0001,
"loss": 4.555330276489258,
"step": 382
},
{
"epoch": 0.14133493242308223,
"grad_norm": 2.5999021530151367,
"learning_rate": 0.0001,
"loss": 4.632719993591309,
"step": 383
},
{
"epoch": 0.14170395313436968,
"grad_norm": 2.2470548152923584,
"learning_rate": 0.0001,
"loss": 4.36204719543457,
"step": 384
},
{
"epoch": 0.1420729738456571,
"grad_norm": 2.7499003410339355,
"learning_rate": 0.0001,
"loss": 4.743986129760742,
"step": 385
},
{
"epoch": 0.1424419945569445,
"grad_norm": 2.742394208908081,
"learning_rate": 0.0001,
"loss": 4.636870384216309,
"step": 386
},
{
"epoch": 0.14281101526823192,
"grad_norm": 2.180415391921997,
"learning_rate": 0.0001,
"loss": 4.016098976135254,
"step": 387
},
{
"epoch": 0.14318003597951934,
"grad_norm": 2.76760196685791,
"learning_rate": 0.0001,
"loss": 4.450103759765625,
"step": 388
},
{
"epoch": 0.14354905669080678,
"grad_norm": 2.0674028396606445,
"learning_rate": 0.0001,
"loss": 4.38997745513916,
"step": 389
},
{
"epoch": 0.1439180774020942,
"grad_norm": 2.3727712631225586,
"learning_rate": 0.0001,
"loss": 4.650741100311279,
"step": 390
},
{
"epoch": 0.1442870981133816,
"grad_norm": 3.3699772357940674,
"learning_rate": 0.0001,
"loss": 4.770174026489258,
"step": 391
},
{
"epoch": 0.14465611882466903,
"grad_norm": 2.5660290718078613,
"learning_rate": 0.0001,
"loss": 4.350974082946777,
"step": 392
},
{
"epoch": 0.14502513953595644,
"grad_norm": 2.490234613418579,
"learning_rate": 0.0001,
"loss": 4.715450286865234,
"step": 393
},
{
"epoch": 0.1453941602472439,
"grad_norm": 2.437030076980591,
"learning_rate": 0.0001,
"loss": 4.426624298095703,
"step": 394
},
{
"epoch": 0.1457631809585313,
"grad_norm": 2.3875768184661865,
"learning_rate": 0.0001,
"loss": 4.674262046813965,
"step": 395
},
{
"epoch": 0.14613220166981872,
"grad_norm": 3.1936697959899902,
"learning_rate": 0.0001,
"loss": 4.431836128234863,
"step": 396
},
{
"epoch": 0.14650122238110613,
"grad_norm": 2.755526542663574,
"learning_rate": 0.0001,
"loss": 4.50398588180542,
"step": 397
},
{
"epoch": 0.14687024309239355,
"grad_norm": 2.3031511306762695,
"learning_rate": 0.0001,
"loss": 4.393855094909668,
"step": 398
},
{
"epoch": 0.147239263803681,
"grad_norm": 2.156735420227051,
"learning_rate": 0.0001,
"loss": 4.521570205688477,
"step": 399
},
{
"epoch": 0.1476082845149684,
"grad_norm": 2.1634793281555176,
"learning_rate": 0.0001,
"loss": 4.347168922424316,
"step": 400
},
{
"epoch": 0.14797730522625582,
"grad_norm": 2.3540008068084717,
"learning_rate": 0.0001,
"loss": 4.335336685180664,
"step": 401
},
{
"epoch": 0.14834632593754324,
"grad_norm": 2.099828004837036,
"learning_rate": 0.0001,
"loss": 4.490433216094971,
"step": 402
},
{
"epoch": 0.14871534664883065,
"grad_norm": 2.301783800125122,
"learning_rate": 0.0001,
"loss": 4.512866973876953,
"step": 403
},
{
"epoch": 0.1490843673601181,
"grad_norm": 4.869900226593018,
"learning_rate": 0.0001,
"loss": 4.349788665771484,
"step": 404
},
{
"epoch": 0.1494533880714055,
"grad_norm": 2.6026854515075684,
"learning_rate": 0.0001,
"loss": 4.820313930511475,
"step": 405
},
{
"epoch": 0.14982240878269293,
"grad_norm": 2.8290839195251465,
"learning_rate": 0.0001,
"loss": 4.873462200164795,
"step": 406
},
{
"epoch": 0.15019142949398034,
"grad_norm": 2.684792995452881,
"learning_rate": 0.0001,
"loss": 4.517108917236328,
"step": 407
},
{
"epoch": 0.15056045020526776,
"grad_norm": 2.500267505645752,
"learning_rate": 0.0001,
"loss": 4.553900241851807,
"step": 408
},
{
"epoch": 0.1509294709165552,
"grad_norm": 2.6919009685516357,
"learning_rate": 0.0001,
"loss": 4.5190510749816895,
"step": 409
},
{
"epoch": 0.15129849162784262,
"grad_norm": 2.3876686096191406,
"learning_rate": 0.0001,
"loss": 4.545931339263916,
"step": 410
},
{
"epoch": 0.15166751233913003,
"grad_norm": 2.572681427001953,
"learning_rate": 0.0001,
"loss": 4.756655693054199,
"step": 411
},
{
"epoch": 0.15203653305041745,
"grad_norm": 2.4582159519195557,
"learning_rate": 0.0001,
"loss": 4.263705730438232,
"step": 412
},
{
"epoch": 0.15240555376170487,
"grad_norm": 2.2141013145446777,
"learning_rate": 0.0001,
"loss": 4.258913040161133,
"step": 413
},
{
"epoch": 0.1527745744729923,
"grad_norm": 2.1214869022369385,
"learning_rate": 0.0001,
"loss": 4.433398723602295,
"step": 414
},
{
"epoch": 0.15314359518427972,
"grad_norm": 3.5143837928771973,
"learning_rate": 0.0001,
"loss": 4.403622627258301,
"step": 415
},
{
"epoch": 0.15351261589556714,
"grad_norm": 2.399301767349243,
"learning_rate": 0.0001,
"loss": 4.545055389404297,
"step": 416
},
{
"epoch": 0.15388163660685455,
"grad_norm": 2.3257532119750977,
"learning_rate": 0.0001,
"loss": 4.454653263092041,
"step": 417
},
{
"epoch": 0.15425065731814197,
"grad_norm": 2.069754123687744,
"learning_rate": 0.0001,
"loss": 4.391440391540527,
"step": 418
},
{
"epoch": 0.1546196780294294,
"grad_norm": 2.3238942623138428,
"learning_rate": 0.0001,
"loss": 4.560251235961914,
"step": 419
},
{
"epoch": 0.15498869874071683,
"grad_norm": 2.3678717613220215,
"learning_rate": 0.0001,
"loss": 4.415225028991699,
"step": 420
},
{
"epoch": 0.15535771945200424,
"grad_norm": 2.5803234577178955,
"learning_rate": 0.0001,
"loss": 4.502828598022461,
"step": 421
},
{
"epoch": 0.15572674016329166,
"grad_norm": 2.0687503814697266,
"learning_rate": 0.0001,
"loss": 4.573302268981934,
"step": 422
},
{
"epoch": 0.15609576087457908,
"grad_norm": 2.0890164375305176,
"learning_rate": 0.0001,
"loss": 4.3898725509643555,
"step": 423
},
{
"epoch": 0.15646478158586652,
"grad_norm": 6.507479667663574,
"learning_rate": 0.0001,
"loss": 4.818697452545166,
"step": 424
},
{
"epoch": 0.15683380229715393,
"grad_norm": 2.371757745742798,
"learning_rate": 0.0001,
"loss": 4.283136367797852,
"step": 425
},
{
"epoch": 0.15720282300844135,
"grad_norm": 2.1233596801757812,
"learning_rate": 0.0001,
"loss": 4.306228160858154,
"step": 426
},
{
"epoch": 0.15757184371972877,
"grad_norm": 1.91828191280365,
"learning_rate": 0.0001,
"loss": 4.135748386383057,
"step": 427
},
{
"epoch": 0.15794086443101618,
"grad_norm": 2.200847864151001,
"learning_rate": 0.0001,
"loss": 4.567739486694336,
"step": 428
},
{
"epoch": 0.15830988514230362,
"grad_norm": 2.3396151065826416,
"learning_rate": 0.0001,
"loss": 4.297729015350342,
"step": 429
},
{
"epoch": 0.15867890585359104,
"grad_norm": 2.224308967590332,
"learning_rate": 0.0001,
"loss": 4.248779773712158,
"step": 430
},
{
"epoch": 0.15904792656487846,
"grad_norm": 2.5491580963134766,
"learning_rate": 0.0001,
"loss": 4.218733787536621,
"step": 431
},
{
"epoch": 0.15941694727616587,
"grad_norm": 2.4283812046051025,
"learning_rate": 0.0001,
"loss": 4.502690315246582,
"step": 432
},
{
"epoch": 0.1597859679874533,
"grad_norm": 2.155597448348999,
"learning_rate": 0.0001,
"loss": 4.561605930328369,
"step": 433
},
{
"epoch": 0.16015498869874073,
"grad_norm": 2.9684088230133057,
"learning_rate": 0.0001,
"loss": 4.341606140136719,
"step": 434
},
{
"epoch": 0.16052400941002815,
"grad_norm": 2.1386497020721436,
"learning_rate": 0.0001,
"loss": 4.319955825805664,
"step": 435
},
{
"epoch": 0.16089303012131556,
"grad_norm": 2.0225462913513184,
"learning_rate": 0.0001,
"loss": 4.451445579528809,
"step": 436
},
{
"epoch": 0.16126205083260298,
"grad_norm": 2.4696478843688965,
"learning_rate": 0.0001,
"loss": 4.350612163543701,
"step": 437
},
{
"epoch": 0.1616310715438904,
"grad_norm": 3.6279401779174805,
"learning_rate": 0.0001,
"loss": 4.606565952301025,
"step": 438
},
{
"epoch": 0.16200009225517784,
"grad_norm": 2.111154556274414,
"learning_rate": 0.0001,
"loss": 4.30889368057251,
"step": 439
},
{
"epoch": 0.16236911296646525,
"grad_norm": 3.213124990463257,
"learning_rate": 0.0001,
"loss": 4.794126987457275,
"step": 440
},
{
"epoch": 0.16273813367775267,
"grad_norm": 2.129725933074951,
"learning_rate": 0.0001,
"loss": 4.6410322189331055,
"step": 441
},
{
"epoch": 0.16310715438904008,
"grad_norm": 1.9005858898162842,
"learning_rate": 0.0001,
"loss": 4.66891622543335,
"step": 442
},
{
"epoch": 0.1634761751003275,
"grad_norm": 2.47507381439209,
"learning_rate": 0.0001,
"loss": 3.983210802078247,
"step": 443
},
{
"epoch": 0.1638451958116149,
"grad_norm": 2.970912456512451,
"learning_rate": 0.0001,
"loss": 4.768887996673584,
"step": 444
},
{
"epoch": 0.16421421652290236,
"grad_norm": 2.7908215522766113,
"learning_rate": 0.0001,
"loss": 4.4359517097473145,
"step": 445
},
{
"epoch": 0.16458323723418977,
"grad_norm": 2.090733289718628,
"learning_rate": 0.0001,
"loss": 4.485992908477783,
"step": 446
},
{
"epoch": 0.1649522579454772,
"grad_norm": 2.2121379375457764,
"learning_rate": 0.0001,
"loss": 4.506903648376465,
"step": 447
},
{
"epoch": 0.1653212786567646,
"grad_norm": 2.533843517303467,
"learning_rate": 0.0001,
"loss": 4.350931644439697,
"step": 448
},
{
"epoch": 0.16569029936805202,
"grad_norm": 2.2086493968963623,
"learning_rate": 0.0001,
"loss": 3.9907069206237793,
"step": 449
},
{
"epoch": 0.16605932007933946,
"grad_norm": 2.3507115840911865,
"learning_rate": 0.0001,
"loss": 4.416835308074951,
"step": 450
},
{
"epoch": 0.16642834079062688,
"grad_norm": 2.2729458808898926,
"learning_rate": 0.0001,
"loss": 4.586258411407471,
"step": 451
},
{
"epoch": 0.1667973615019143,
"grad_norm": 2.1393730640411377,
"learning_rate": 0.0001,
"loss": 4.258632183074951,
"step": 452
},
{
"epoch": 0.1671663822132017,
"grad_norm": 2.0558552742004395,
"learning_rate": 0.0001,
"loss": 4.597503662109375,
"step": 453
},
{
"epoch": 0.16753540292448912,
"grad_norm": 2.473813533782959,
"learning_rate": 0.0001,
"loss": 4.395957946777344,
"step": 454
},
{
"epoch": 0.16790442363577657,
"grad_norm": 2.095021963119507,
"learning_rate": 0.0001,
"loss": 4.062214374542236,
"step": 455
},
{
"epoch": 0.16827344434706398,
"grad_norm": 2.6224169731140137,
"learning_rate": 0.0001,
"loss": 4.3939208984375,
"step": 456
},
{
"epoch": 0.1686424650583514,
"grad_norm": 2.2001235485076904,
"learning_rate": 0.0001,
"loss": 4.333123683929443,
"step": 457
},
{
"epoch": 0.16901148576963881,
"grad_norm": 2.21736741065979,
"learning_rate": 0.0001,
"loss": 4.0645246505737305,
"step": 458
},
{
"epoch": 0.16938050648092623,
"grad_norm": 2.2809412479400635,
"learning_rate": 0.0001,
"loss": 4.523746967315674,
"step": 459
},
{
"epoch": 0.16974952719221367,
"grad_norm": 2.6346044540405273,
"learning_rate": 0.0001,
"loss": 4.581881999969482,
"step": 460
},
{
"epoch": 0.1701185479035011,
"grad_norm": 2.1072700023651123,
"learning_rate": 0.0001,
"loss": 4.362997531890869,
"step": 461
},
{
"epoch": 0.1704875686147885,
"grad_norm": 2.147351026535034,
"learning_rate": 0.0001,
"loss": 4.131965637207031,
"step": 462
},
{
"epoch": 0.17085658932607592,
"grad_norm": 2.3557586669921875,
"learning_rate": 0.0001,
"loss": 4.431070327758789,
"step": 463
},
{
"epoch": 0.17122561003736333,
"grad_norm": 2.1692962646484375,
"learning_rate": 0.0001,
"loss": 4.711268901824951,
"step": 464
},
{
"epoch": 0.17159463074865078,
"grad_norm": 2.254636764526367,
"learning_rate": 0.0001,
"loss": 4.4809041023254395,
"step": 465
},
{
"epoch": 0.1719636514599382,
"grad_norm": 2.2716152667999268,
"learning_rate": 0.0001,
"loss": 4.050970077514648,
"step": 466
},
{
"epoch": 0.1723326721712256,
"grad_norm": 2.9789211750030518,
"learning_rate": 0.0001,
"loss": 4.269569396972656,
"step": 467
},
{
"epoch": 0.17270169288251302,
"grad_norm": 2.0386388301849365,
"learning_rate": 0.0001,
"loss": 4.284242630004883,
"step": 468
},
{
"epoch": 0.17307071359380044,
"grad_norm": 1.8008029460906982,
"learning_rate": 0.0001,
"loss": 4.464146137237549,
"step": 469
},
{
"epoch": 0.17343973430508788,
"grad_norm": 1.944584608078003,
"learning_rate": 0.0001,
"loss": 4.559577465057373,
"step": 470
},
{
"epoch": 0.1738087550163753,
"grad_norm": 4.818556785583496,
"learning_rate": 0.0001,
"loss": 4.378731727600098,
"step": 471
},
{
"epoch": 0.17417777572766271,
"grad_norm": 2.3885557651519775,
"learning_rate": 0.0001,
"loss": 4.05840539932251,
"step": 472
},
{
"epoch": 0.17454679643895013,
"grad_norm": 2.1333820819854736,
"learning_rate": 0.0001,
"loss": 4.016430377960205,
"step": 473
},
{
"epoch": 0.17491581715023755,
"grad_norm": 1.8299129009246826,
"learning_rate": 0.0001,
"loss": 4.615049362182617,
"step": 474
},
{
"epoch": 0.175284837861525,
"grad_norm": 2.0134215354919434,
"learning_rate": 0.0001,
"loss": 4.2402191162109375,
"step": 475
},
{
"epoch": 0.1756538585728124,
"grad_norm": 2.5740880966186523,
"learning_rate": 0.0001,
"loss": 4.674981594085693,
"step": 476
},
{
"epoch": 0.17602287928409982,
"grad_norm": 3.1095097064971924,
"learning_rate": 0.0001,
"loss": 4.765364646911621,
"step": 477
},
{
"epoch": 0.17639189999538724,
"grad_norm": 2.1833994388580322,
"learning_rate": 0.0001,
"loss": 4.4260969161987305,
"step": 478
},
{
"epoch": 0.17676092070667465,
"grad_norm": 2.396684408187866,
"learning_rate": 0.0001,
"loss": 4.116750717163086,
"step": 479
},
{
"epoch": 0.1771299414179621,
"grad_norm": 2.2355973720550537,
"learning_rate": 0.0001,
"loss": 4.551278114318848,
"step": 480
},
{
"epoch": 0.1774989621292495,
"grad_norm": 2.0249733924865723,
"learning_rate": 0.0001,
"loss": 4.321774959564209,
"step": 481
},
{
"epoch": 0.17786798284053693,
"grad_norm": 2.2655394077301025,
"learning_rate": 0.0001,
"loss": 4.832188129425049,
"step": 482
},
{
"epoch": 0.17823700355182434,
"grad_norm": 2.384474754333496,
"learning_rate": 0.0001,
"loss": 4.327328681945801,
"step": 483
},
{
"epoch": 0.17860602426311176,
"grad_norm": 2.0592844486236572,
"learning_rate": 0.0001,
"loss": 4.406461238861084,
"step": 484
},
{
"epoch": 0.1789750449743992,
"grad_norm": 2.1442856788635254,
"learning_rate": 0.0001,
"loss": 4.520481109619141,
"step": 485
},
{
"epoch": 0.17934406568568662,
"grad_norm": 2.215745449066162,
"learning_rate": 0.0001,
"loss": 4.652613162994385,
"step": 486
},
{
"epoch": 0.17971308639697403,
"grad_norm": 2.274385452270508,
"learning_rate": 0.0001,
"loss": 4.071110248565674,
"step": 487
},
{
"epoch": 0.18008210710826145,
"grad_norm": 2.5120763778686523,
"learning_rate": 0.0001,
"loss": 4.499899387359619,
"step": 488
},
{
"epoch": 0.18045112781954886,
"grad_norm": 2.1488053798675537,
"learning_rate": 0.0001,
"loss": 4.293880939483643,
"step": 489
},
{
"epoch": 0.1808201485308363,
"grad_norm": 2.8950445652008057,
"learning_rate": 0.0001,
"loss": 4.485057830810547,
"step": 490
},
{
"epoch": 0.18118916924212372,
"grad_norm": 2.100581407546997,
"learning_rate": 0.0001,
"loss": 4.113532066345215,
"step": 491
},
{
"epoch": 0.18155818995341114,
"grad_norm": 2.3395795822143555,
"learning_rate": 0.0001,
"loss": 4.322099208831787,
"step": 492
},
{
"epoch": 0.18192721066469855,
"grad_norm": 2.030890703201294,
"learning_rate": 0.0001,
"loss": 4.442522048950195,
"step": 493
},
{
"epoch": 0.18229623137598597,
"grad_norm": 1.977658987045288,
"learning_rate": 0.0001,
"loss": 4.220512866973877,
"step": 494
},
{
"epoch": 0.1826652520872734,
"grad_norm": 3.0135514736175537,
"learning_rate": 0.0001,
"loss": 4.858924388885498,
"step": 495
},
{
"epoch": 0.18303427279856083,
"grad_norm": 2.4730112552642822,
"learning_rate": 0.0001,
"loss": 4.4013543128967285,
"step": 496
},
{
"epoch": 0.18340329350984824,
"grad_norm": 2.456479549407959,
"learning_rate": 0.0001,
"loss": 4.416372776031494,
"step": 497
},
{
"epoch": 0.18377231422113566,
"grad_norm": 2.1360464096069336,
"learning_rate": 0.0001,
"loss": 4.0488104820251465,
"step": 498
},
{
"epoch": 0.18414133493242307,
"grad_norm": 2.367122173309326,
"learning_rate": 0.0001,
"loss": 4.326754570007324,
"step": 499
},
{
"epoch": 0.18451035564371052,
"grad_norm": 2.3929057121276855,
"learning_rate": 0.0001,
"loss": 4.3895063400268555,
"step": 500
},
{
"epoch": 0.18487937635499793,
"grad_norm": 1.9481403827667236,
"learning_rate": 0.0001,
"loss": 4.406152248382568,
"step": 501
},
{
"epoch": 0.18524839706628535,
"grad_norm": 2.086376190185547,
"learning_rate": 0.0001,
"loss": 4.4221906661987305,
"step": 502
},
{
"epoch": 0.18561741777757276,
"grad_norm": 2.2188632488250732,
"learning_rate": 0.0001,
"loss": 4.660634994506836,
"step": 503
},
{
"epoch": 0.18598643848886018,
"grad_norm": 2.2751147747039795,
"learning_rate": 0.0001,
"loss": 4.23121452331543,
"step": 504
},
{
"epoch": 0.18635545920014762,
"grad_norm": 2.1705162525177,
"learning_rate": 0.0001,
"loss": 4.265094757080078,
"step": 505
},
{
"epoch": 0.18672447991143504,
"grad_norm": 2.121332883834839,
"learning_rate": 0.0001,
"loss": 4.205435752868652,
"step": 506
},
{
"epoch": 0.18709350062272245,
"grad_norm": 1.964434266090393,
"learning_rate": 0.0001,
"loss": 4.287982940673828,
"step": 507
},
{
"epoch": 0.18746252133400987,
"grad_norm": 2.236156463623047,
"learning_rate": 0.0001,
"loss": 4.284115791320801,
"step": 508
},
{
"epoch": 0.18783154204529728,
"grad_norm": 2.2012174129486084,
"learning_rate": 0.0001,
"loss": 4.316831111907959,
"step": 509
},
{
"epoch": 0.1882005627565847,
"grad_norm": 1.9142396450042725,
"learning_rate": 0.0001,
"loss": 4.5977678298950195,
"step": 510
},
{
"epoch": 0.18856958346787214,
"grad_norm": 1.975645899772644,
"learning_rate": 0.0001,
"loss": 4.377478122711182,
"step": 511
},
{
"epoch": 0.18893860417915956,
"grad_norm": 2.210484027862549,
"learning_rate": 0.0001,
"loss": 4.167498588562012,
"step": 512
},
{
"epoch": 0.18930762489044697,
"grad_norm": 1.8951791524887085,
"learning_rate": 0.0001,
"loss": 4.046764373779297,
"step": 513
},
{
"epoch": 0.1896766456017344,
"grad_norm": 2.129709005355835,
"learning_rate": 0.0001,
"loss": 4.1118035316467285,
"step": 514
},
{
"epoch": 0.1900456663130218,
"grad_norm": 1.926705241203308,
"learning_rate": 0.0001,
"loss": 3.9770379066467285,
"step": 515
},
{
"epoch": 0.19041468702430925,
"grad_norm": 2.785149335861206,
"learning_rate": 0.0001,
"loss": 4.119317054748535,
"step": 516
},
{
"epoch": 0.19078370773559666,
"grad_norm": 2.28780198097229,
"learning_rate": 0.0001,
"loss": 3.9503774642944336,
"step": 517
},
{
"epoch": 0.19115272844688408,
"grad_norm": 2.1383399963378906,
"learning_rate": 0.0001,
"loss": 4.462101459503174,
"step": 518
},
{
"epoch": 0.1915217491581715,
"grad_norm": 1.8943592309951782,
"learning_rate": 0.0001,
"loss": 4.395929336547852,
"step": 519
},
{
"epoch": 0.1918907698694589,
"grad_norm": 2.1494009494781494,
"learning_rate": 0.0001,
"loss": 4.108682155609131,
"step": 520
},
{
"epoch": 0.19225979058074635,
"grad_norm": 2.092240810394287,
"learning_rate": 0.0001,
"loss": 4.348626613616943,
"step": 521
},
{
"epoch": 0.19262881129203377,
"grad_norm": 2.151693820953369,
"learning_rate": 0.0001,
"loss": 4.288055896759033,
"step": 522
},
{
"epoch": 0.19299783200332118,
"grad_norm": 1.9059604406356812,
"learning_rate": 0.0001,
"loss": 4.300743103027344,
"step": 523
},
{
"epoch": 0.1933668527146086,
"grad_norm": 2.5519795417785645,
"learning_rate": 0.0001,
"loss": 4.22593355178833,
"step": 524
},
{
"epoch": 0.19373587342589602,
"grad_norm": 1.7208030223846436,
"learning_rate": 0.0001,
"loss": 4.30415153503418,
"step": 525
},
{
"epoch": 0.19410489413718346,
"grad_norm": 1.7663142681121826,
"learning_rate": 0.0001,
"loss": 4.290582180023193,
"step": 526
},
{
"epoch": 0.19447391484847087,
"grad_norm": 2.063027858734131,
"learning_rate": 0.0001,
"loss": 4.227534770965576,
"step": 527
},
{
"epoch": 0.1948429355597583,
"grad_norm": 2.6470508575439453,
"learning_rate": 0.0001,
"loss": 4.313108921051025,
"step": 528
},
{
"epoch": 0.1952119562710457,
"grad_norm": 1.9663456678390503,
"learning_rate": 0.0001,
"loss": 4.123837947845459,
"step": 529
},
{
"epoch": 0.19558097698233312,
"grad_norm": 2.032116413116455,
"learning_rate": 0.0001,
"loss": 3.8263444900512695,
"step": 530
},
{
"epoch": 0.19594999769362056,
"grad_norm": 3.4389142990112305,
"learning_rate": 0.0001,
"loss": 4.467240333557129,
"step": 531
},
{
"epoch": 0.19631901840490798,
"grad_norm": 1.9850727319717407,
"learning_rate": 0.0001,
"loss": 4.335421085357666,
"step": 532
},
{
"epoch": 0.1966880391161954,
"grad_norm": 2.0280203819274902,
"learning_rate": 0.0001,
"loss": 4.235400676727295,
"step": 533
},
{
"epoch": 0.1970570598274828,
"grad_norm": 2.8351972103118896,
"learning_rate": 0.0001,
"loss": 4.130883693695068,
"step": 534
},
{
"epoch": 0.19742608053877023,
"grad_norm": 2.1816749572753906,
"learning_rate": 0.0001,
"loss": 4.412903785705566,
"step": 535
},
{
"epoch": 0.19779510125005767,
"grad_norm": 2.455271005630493,
"learning_rate": 0.0001,
"loss": 4.324191093444824,
"step": 536
},
{
"epoch": 0.19816412196134509,
"grad_norm": 2.081894874572754,
"learning_rate": 0.0001,
"loss": 4.476474761962891,
"step": 537
},
{
"epoch": 0.1985331426726325,
"grad_norm": 2.1476821899414062,
"learning_rate": 0.0001,
"loss": 4.266482830047607,
"step": 538
},
{
"epoch": 0.19890216338391992,
"grad_norm": 1.989898681640625,
"learning_rate": 0.0001,
"loss": 4.313356876373291,
"step": 539
},
{
"epoch": 0.19927118409520733,
"grad_norm": 2.1811554431915283,
"learning_rate": 0.0001,
"loss": 4.685174465179443,
"step": 540
},
{
"epoch": 0.19964020480649478,
"grad_norm": 2.0105929374694824,
"learning_rate": 0.0001,
"loss": 4.369940280914307,
"step": 541
},
{
"epoch": 0.2000092255177822,
"grad_norm": 1.7295829057693481,
"learning_rate": 0.0001,
"loss": 3.995760679244995,
"step": 542
}
],
"logging_steps": 1,
"max_steps": 5420,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 542,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.628137438584914e+17,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}