wz7475's picture
Training in progress, step 304, checkpoint
4c408fa verified
Raw
History Blame Contribute Delete
54.5 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 50,
"global_step": 304,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0032921810699588477,
"grad_norm": 1.8741168975830078,
"learning_rate": 0.0,
"loss": 0.1604,
"step": 1
},
{
"epoch": 0.006584362139917695,
"grad_norm": 1.54278564453125,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.2501,
"step": 2
},
{
"epoch": 0.009876543209876543,
"grad_norm": 1.8454307317733765,
"learning_rate": 4.000000000000001e-06,
"loss": 0.2626,
"step": 3
},
{
"epoch": 0.01316872427983539,
"grad_norm": 1.913107991218567,
"learning_rate": 6e-06,
"loss": 0.2546,
"step": 4
},
{
"epoch": 0.01646090534979424,
"grad_norm": 1.361615538597107,
"learning_rate": 8.000000000000001e-06,
"loss": 0.1564,
"step": 5
},
{
"epoch": 0.019753086419753086,
"grad_norm": 1.5066672563552856,
"learning_rate": 1e-05,
"loss": 0.2574,
"step": 6
},
{
"epoch": 0.023045267489711935,
"grad_norm": 1.3929274082183838,
"learning_rate": 9.966555183946488e-06,
"loss": 0.2793,
"step": 7
},
{
"epoch": 0.02633744855967078,
"grad_norm": 1.124476432800293,
"learning_rate": 9.933110367892978e-06,
"loss": 0.1822,
"step": 8
},
{
"epoch": 0.02962962962962963,
"grad_norm": 1.0770095586776733,
"learning_rate": 9.899665551839465e-06,
"loss": 0.1879,
"step": 9
},
{
"epoch": 0.03292181069958848,
"grad_norm": 1.3482930660247803,
"learning_rate": 9.866220735785954e-06,
"loss": 0.1962,
"step": 10
},
{
"epoch": 0.03621399176954732,
"grad_norm": 1.2692549228668213,
"learning_rate": 9.832775919732442e-06,
"loss": 0.2154,
"step": 11
},
{
"epoch": 0.03950617283950617,
"grad_norm": 1.062791347503662,
"learning_rate": 9.799331103678931e-06,
"loss": 0.1903,
"step": 12
},
{
"epoch": 0.04279835390946502,
"grad_norm": 1.1957647800445557,
"learning_rate": 9.765886287625419e-06,
"loss": 0.2027,
"step": 13
},
{
"epoch": 0.04609053497942387,
"grad_norm": 0.9908819794654846,
"learning_rate": 9.732441471571908e-06,
"loss": 0.1236,
"step": 14
},
{
"epoch": 0.04938271604938271,
"grad_norm": 1.0809367895126343,
"learning_rate": 9.698996655518395e-06,
"loss": 0.1713,
"step": 15
},
{
"epoch": 0.05267489711934156,
"grad_norm": 1.233067512512207,
"learning_rate": 9.665551839464884e-06,
"loss": 0.1802,
"step": 16
},
{
"epoch": 0.05596707818930041,
"grad_norm": 1.327765941619873,
"learning_rate": 9.632107023411372e-06,
"loss": 0.2246,
"step": 17
},
{
"epoch": 0.05925925925925926,
"grad_norm": 1.594014286994934,
"learning_rate": 9.598662207357861e-06,
"loss": 0.318,
"step": 18
},
{
"epoch": 0.06255144032921811,
"grad_norm": 1.3657221794128418,
"learning_rate": 9.565217391304349e-06,
"loss": 0.2717,
"step": 19
},
{
"epoch": 0.06584362139917696,
"grad_norm": 1.087826132774353,
"learning_rate": 9.531772575250838e-06,
"loss": 0.1511,
"step": 20
},
{
"epoch": 0.0691358024691358,
"grad_norm": 1.1943680047988892,
"learning_rate": 9.498327759197325e-06,
"loss": 0.1541,
"step": 21
},
{
"epoch": 0.07242798353909465,
"grad_norm": 1.1076587438583374,
"learning_rate": 9.464882943143815e-06,
"loss": 0.1941,
"step": 22
},
{
"epoch": 0.0757201646090535,
"grad_norm": 1.31613290309906,
"learning_rate": 9.431438127090302e-06,
"loss": 0.2718,
"step": 23
},
{
"epoch": 0.07901234567901234,
"grad_norm": 1.373462438583374,
"learning_rate": 9.39799331103679e-06,
"loss": 0.2126,
"step": 24
},
{
"epoch": 0.0823045267489712,
"grad_norm": 1.1619008779525757,
"learning_rate": 9.364548494983279e-06,
"loss": 0.1676,
"step": 25
},
{
"epoch": 0.08559670781893004,
"grad_norm": 1.2414864301681519,
"learning_rate": 9.331103678929766e-06,
"loss": 0.2515,
"step": 26
},
{
"epoch": 0.08888888888888889,
"grad_norm": 1.2002915143966675,
"learning_rate": 9.297658862876256e-06,
"loss": 0.2194,
"step": 27
},
{
"epoch": 0.09218106995884774,
"grad_norm": 1.189249873161316,
"learning_rate": 9.264214046822743e-06,
"loss": 0.1991,
"step": 28
},
{
"epoch": 0.09547325102880659,
"grad_norm": 1.7674837112426758,
"learning_rate": 9.230769230769232e-06,
"loss": 0.2262,
"step": 29
},
{
"epoch": 0.09876543209876543,
"grad_norm": 1.246118187904358,
"learning_rate": 9.19732441471572e-06,
"loss": 0.21,
"step": 30
},
{
"epoch": 0.10205761316872428,
"grad_norm": 0.9715161323547363,
"learning_rate": 9.163879598662207e-06,
"loss": 0.0922,
"step": 31
},
{
"epoch": 0.10534979423868313,
"grad_norm": 0.7676671743392944,
"learning_rate": 9.130434782608697e-06,
"loss": 0.0998,
"step": 32
},
{
"epoch": 0.10864197530864197,
"grad_norm": 0.9821469783782959,
"learning_rate": 9.096989966555184e-06,
"loss": 0.1289,
"step": 33
},
{
"epoch": 0.11193415637860082,
"grad_norm": 1.1367111206054688,
"learning_rate": 9.063545150501673e-06,
"loss": 0.1944,
"step": 34
},
{
"epoch": 0.11522633744855967,
"grad_norm": 1.1714760065078735,
"learning_rate": 9.03010033444816e-06,
"loss": 0.1824,
"step": 35
},
{
"epoch": 0.11851851851851852,
"grad_norm": 1.2322204113006592,
"learning_rate": 8.996655518394648e-06,
"loss": 0.2045,
"step": 36
},
{
"epoch": 0.12181069958847737,
"grad_norm": 1.5084517002105713,
"learning_rate": 8.963210702341138e-06,
"loss": 0.2557,
"step": 37
},
{
"epoch": 0.12510288065843622,
"grad_norm": 1.052878975868225,
"learning_rate": 8.929765886287625e-06,
"loss": 0.1849,
"step": 38
},
{
"epoch": 0.12839506172839507,
"grad_norm": 1.36421537399292,
"learning_rate": 8.896321070234114e-06,
"loss": 0.224,
"step": 39
},
{
"epoch": 0.13168724279835392,
"grad_norm": 1.2042744159698486,
"learning_rate": 8.862876254180602e-06,
"loss": 0.172,
"step": 40
},
{
"epoch": 0.13497942386831277,
"grad_norm": 1.1954560279846191,
"learning_rate": 8.829431438127091e-06,
"loss": 0.151,
"step": 41
},
{
"epoch": 0.1382716049382716,
"grad_norm": 1.2024387121200562,
"learning_rate": 8.795986622073578e-06,
"loss": 0.1909,
"step": 42
},
{
"epoch": 0.14156378600823044,
"grad_norm": 1.3407139778137207,
"learning_rate": 8.762541806020068e-06,
"loss": 0.2797,
"step": 43
},
{
"epoch": 0.1448559670781893,
"grad_norm": 1.145772933959961,
"learning_rate": 8.729096989966555e-06,
"loss": 0.1652,
"step": 44
},
{
"epoch": 0.14814814814814814,
"grad_norm": 1.3911467790603638,
"learning_rate": 8.695652173913044e-06,
"loss": 0.2215,
"step": 45
},
{
"epoch": 0.151440329218107,
"grad_norm": 1.2512582540512085,
"learning_rate": 8.662207357859532e-06,
"loss": 0.196,
"step": 46
},
{
"epoch": 0.15473251028806584,
"grad_norm": 1.17892324924469,
"learning_rate": 8.628762541806021e-06,
"loss": 0.171,
"step": 47
},
{
"epoch": 0.1580246913580247,
"grad_norm": 1.0726583003997803,
"learning_rate": 8.595317725752509e-06,
"loss": 0.1614,
"step": 48
},
{
"epoch": 0.16131687242798354,
"grad_norm": 1.1015340089797974,
"learning_rate": 8.561872909698998e-06,
"loss": 0.1797,
"step": 49
},
{
"epoch": 0.1646090534979424,
"grad_norm": 1.237147569656372,
"learning_rate": 8.528428093645485e-06,
"loss": 0.1635,
"step": 50
},
{
"epoch": 0.1646090534979424,
"eval_loss": 0.22677525877952576,
"eval_runtime": 21.2479,
"eval_samples_per_second": 25.414,
"eval_steps_per_second": 12.707,
"step": 50
},
{
"epoch": 0.16790123456790124,
"grad_norm": 1.2092516422271729,
"learning_rate": 8.494983277591975e-06,
"loss": 0.1658,
"step": 51
},
{
"epoch": 0.17119341563786009,
"grad_norm": 0.9954591393470764,
"learning_rate": 8.461538461538462e-06,
"loss": 0.1353,
"step": 52
},
{
"epoch": 0.17448559670781894,
"grad_norm": 1.4036070108413696,
"learning_rate": 8.42809364548495e-06,
"loss": 0.1717,
"step": 53
},
{
"epoch": 0.17777777777777778,
"grad_norm": 1.1405463218688965,
"learning_rate": 8.394648829431439e-06,
"loss": 0.1449,
"step": 54
},
{
"epoch": 0.18106995884773663,
"grad_norm": 1.036046028137207,
"learning_rate": 8.361204013377926e-06,
"loss": 0.1448,
"step": 55
},
{
"epoch": 0.18436213991769548,
"grad_norm": 1.202657699584961,
"learning_rate": 8.327759197324416e-06,
"loss": 0.2184,
"step": 56
},
{
"epoch": 0.18765432098765433,
"grad_norm": 1.224108338356018,
"learning_rate": 8.294314381270903e-06,
"loss": 0.1737,
"step": 57
},
{
"epoch": 0.19094650205761318,
"grad_norm": 0.8357282876968384,
"learning_rate": 8.260869565217392e-06,
"loss": 0.0922,
"step": 58
},
{
"epoch": 0.194238683127572,
"grad_norm": 1.122328758239746,
"learning_rate": 8.22742474916388e-06,
"loss": 0.1453,
"step": 59
},
{
"epoch": 0.19753086419753085,
"grad_norm": 1.1722753047943115,
"learning_rate": 8.193979933110369e-06,
"loss": 0.1833,
"step": 60
},
{
"epoch": 0.2008230452674897,
"grad_norm": 1.2879515886306763,
"learning_rate": 8.160535117056857e-06,
"loss": 0.2993,
"step": 61
},
{
"epoch": 0.20411522633744855,
"grad_norm": 1.1054010391235352,
"learning_rate": 8.127090301003346e-06,
"loss": 0.149,
"step": 62
},
{
"epoch": 0.2074074074074074,
"grad_norm": 1.1728150844573975,
"learning_rate": 8.093645484949833e-06,
"loss": 0.1652,
"step": 63
},
{
"epoch": 0.21069958847736625,
"grad_norm": 0.9511211514472961,
"learning_rate": 8.060200668896322e-06,
"loss": 0.1463,
"step": 64
},
{
"epoch": 0.2139917695473251,
"grad_norm": 0.8870574235916138,
"learning_rate": 8.02675585284281e-06,
"loss": 0.0867,
"step": 65
},
{
"epoch": 0.21728395061728395,
"grad_norm": 1.151691198348999,
"learning_rate": 7.9933110367893e-06,
"loss": 0.1839,
"step": 66
},
{
"epoch": 0.2205761316872428,
"grad_norm": 1.1549429893493652,
"learning_rate": 7.959866220735787e-06,
"loss": 0.168,
"step": 67
},
{
"epoch": 0.22386831275720165,
"grad_norm": 0.9350190162658691,
"learning_rate": 7.926421404682276e-06,
"loss": 0.0969,
"step": 68
},
{
"epoch": 0.2271604938271605,
"grad_norm": 1.2745375633239746,
"learning_rate": 7.892976588628763e-06,
"loss": 0.2097,
"step": 69
},
{
"epoch": 0.23045267489711935,
"grad_norm": 1.4003045558929443,
"learning_rate": 7.859531772575253e-06,
"loss": 0.274,
"step": 70
},
{
"epoch": 0.2337448559670782,
"grad_norm": 1.1932967901229858,
"learning_rate": 7.82608695652174e-06,
"loss": 0.1893,
"step": 71
},
{
"epoch": 0.23703703703703705,
"grad_norm": 1.3451650142669678,
"learning_rate": 7.792642140468228e-06,
"loss": 0.2189,
"step": 72
},
{
"epoch": 0.2403292181069959,
"grad_norm": 1.1293561458587646,
"learning_rate": 7.759197324414717e-06,
"loss": 0.1483,
"step": 73
},
{
"epoch": 0.24362139917695474,
"grad_norm": 1.2387956380844116,
"learning_rate": 7.725752508361204e-06,
"loss": 0.1302,
"step": 74
},
{
"epoch": 0.24691358024691357,
"grad_norm": 1.1072285175323486,
"learning_rate": 7.692307692307694e-06,
"loss": 0.1875,
"step": 75
},
{
"epoch": 0.25020576131687244,
"grad_norm": 1.6144697666168213,
"learning_rate": 7.658862876254181e-06,
"loss": 0.2692,
"step": 76
},
{
"epoch": 0.25349794238683127,
"grad_norm": 1.1080658435821533,
"learning_rate": 7.62541806020067e-06,
"loss": 0.1549,
"step": 77
},
{
"epoch": 0.25679012345679014,
"grad_norm": 1.0728422403335571,
"learning_rate": 7.591973244147159e-06,
"loss": 0.1473,
"step": 78
},
{
"epoch": 0.26008230452674896,
"grad_norm": 1.4992091655731201,
"learning_rate": 7.558528428093647e-06,
"loss": 0.2255,
"step": 79
},
{
"epoch": 0.26337448559670784,
"grad_norm": 1.3775638341903687,
"learning_rate": 7.5250836120401346e-06,
"loss": 0.1916,
"step": 80
},
{
"epoch": 0.26666666666666666,
"grad_norm": 1.2366762161254883,
"learning_rate": 7.491638795986622e-06,
"loss": 0.1953,
"step": 81
},
{
"epoch": 0.26995884773662554,
"grad_norm": 1.4844088554382324,
"learning_rate": 7.4581939799331104e-06,
"loss": 0.192,
"step": 82
},
{
"epoch": 0.27325102880658436,
"grad_norm": 1.0697194337844849,
"learning_rate": 7.424749163879599e-06,
"loss": 0.1211,
"step": 83
},
{
"epoch": 0.2765432098765432,
"grad_norm": 1.4621111154556274,
"learning_rate": 7.391304347826087e-06,
"loss": 0.1999,
"step": 84
},
{
"epoch": 0.27983539094650206,
"grad_norm": 1.0858594179153442,
"learning_rate": 7.3578595317725755e-06,
"loss": 0.0881,
"step": 85
},
{
"epoch": 0.2831275720164609,
"grad_norm": 1.0190343856811523,
"learning_rate": 7.324414715719064e-06,
"loss": 0.1362,
"step": 86
},
{
"epoch": 0.28641975308641976,
"grad_norm": 1.049277663230896,
"learning_rate": 7.290969899665552e-06,
"loss": 0.1349,
"step": 87
},
{
"epoch": 0.2897119341563786,
"grad_norm": 1.3304250240325928,
"learning_rate": 7.257525083612041e-06,
"loss": 0.2062,
"step": 88
},
{
"epoch": 0.29300411522633746,
"grad_norm": 1.1579619646072388,
"learning_rate": 7.224080267558529e-06,
"loss": 0.2138,
"step": 89
},
{
"epoch": 0.2962962962962963,
"grad_norm": 1.0575640201568604,
"learning_rate": 7.190635451505017e-06,
"loss": 0.1691,
"step": 90
},
{
"epoch": 0.29958847736625516,
"grad_norm": 0.9682750105857849,
"learning_rate": 7.157190635451506e-06,
"loss": 0.128,
"step": 91
},
{
"epoch": 0.302880658436214,
"grad_norm": 1.2027302980422974,
"learning_rate": 7.123745819397993e-06,
"loss": 0.2106,
"step": 92
},
{
"epoch": 0.30617283950617286,
"grad_norm": 1.0247256755828857,
"learning_rate": 7.0903010033444816e-06,
"loss": 0.1402,
"step": 93
},
{
"epoch": 0.3094650205761317,
"grad_norm": 1.1904380321502686,
"learning_rate": 7.05685618729097e-06,
"loss": 0.2089,
"step": 94
},
{
"epoch": 0.31275720164609055,
"grad_norm": 1.1549211740493774,
"learning_rate": 7.023411371237458e-06,
"loss": 0.161,
"step": 95
},
{
"epoch": 0.3160493827160494,
"grad_norm": 1.0648497343063354,
"learning_rate": 6.989966555183947e-06,
"loss": 0.1863,
"step": 96
},
{
"epoch": 0.31934156378600825,
"grad_norm": 1.2347862720489502,
"learning_rate": 6.956521739130435e-06,
"loss": 0.202,
"step": 97
},
{
"epoch": 0.3226337448559671,
"grad_norm": 1.3978097438812256,
"learning_rate": 6.923076923076923e-06,
"loss": 0.2016,
"step": 98
},
{
"epoch": 0.32592592592592595,
"grad_norm": 1.1399036645889282,
"learning_rate": 6.889632107023412e-06,
"loss": 0.2006,
"step": 99
},
{
"epoch": 0.3292181069958848,
"grad_norm": 0.949056088924408,
"learning_rate": 6.8561872909699e-06,
"loss": 0.1063,
"step": 100
},
{
"epoch": 0.3292181069958848,
"eval_loss": 0.21909213066101074,
"eval_runtime": 20.445,
"eval_samples_per_second": 26.412,
"eval_steps_per_second": 13.206,
"step": 100
},
{
"epoch": 0.3325102880658436,
"grad_norm": 1.2960280179977417,
"learning_rate": 6.8227424749163885e-06,
"loss": 0.2124,
"step": 101
},
{
"epoch": 0.3358024691358025,
"grad_norm": 1.2040042877197266,
"learning_rate": 6.789297658862877e-06,
"loss": 0.1994,
"step": 102
},
{
"epoch": 0.3390946502057613,
"grad_norm": 1.4233226776123047,
"learning_rate": 6.755852842809365e-06,
"loss": 0.1684,
"step": 103
},
{
"epoch": 0.34238683127572017,
"grad_norm": 1.6198140382766724,
"learning_rate": 6.7224080267558536e-06,
"loss": 0.2676,
"step": 104
},
{
"epoch": 0.345679012345679,
"grad_norm": 1.1984881162643433,
"learning_rate": 6.688963210702342e-06,
"loss": 0.1611,
"step": 105
},
{
"epoch": 0.34897119341563787,
"grad_norm": 1.1460095643997192,
"learning_rate": 6.65551839464883e-06,
"loss": 0.1523,
"step": 106
},
{
"epoch": 0.3522633744855967,
"grad_norm": 1.1304082870483398,
"learning_rate": 6.622073578595319e-06,
"loss": 0.179,
"step": 107
},
{
"epoch": 0.35555555555555557,
"grad_norm": 1.4257352352142334,
"learning_rate": 6.588628762541807e-06,
"loss": 0.267,
"step": 108
},
{
"epoch": 0.3588477366255144,
"grad_norm": 1.0096205472946167,
"learning_rate": 6.5551839464882945e-06,
"loss": 0.1537,
"step": 109
},
{
"epoch": 0.36213991769547327,
"grad_norm": 0.9798669219017029,
"learning_rate": 6.521739130434783e-06,
"loss": 0.1354,
"step": 110
},
{
"epoch": 0.3654320987654321,
"grad_norm": 1.4535608291625977,
"learning_rate": 6.488294314381271e-06,
"loss": 0.2711,
"step": 111
},
{
"epoch": 0.36872427983539097,
"grad_norm": 1.2683688402175903,
"learning_rate": 6.45484949832776e-06,
"loss": 0.1787,
"step": 112
},
{
"epoch": 0.3720164609053498,
"grad_norm": 1.420069932937622,
"learning_rate": 6.421404682274248e-06,
"loss": 0.2037,
"step": 113
},
{
"epoch": 0.37530864197530867,
"grad_norm": 1.1771941184997559,
"learning_rate": 6.387959866220736e-06,
"loss": 0.1961,
"step": 114
},
{
"epoch": 0.3786008230452675,
"grad_norm": 1.0671218633651733,
"learning_rate": 6.354515050167225e-06,
"loss": 0.2058,
"step": 115
},
{
"epoch": 0.38189300411522636,
"grad_norm": 1.549080729484558,
"learning_rate": 6.321070234113713e-06,
"loss": 0.2374,
"step": 116
},
{
"epoch": 0.3851851851851852,
"grad_norm": 1.0160949230194092,
"learning_rate": 6.287625418060201e-06,
"loss": 0.1798,
"step": 117
},
{
"epoch": 0.388477366255144,
"grad_norm": 1.6964318752288818,
"learning_rate": 6.25418060200669e-06,
"loss": 0.2037,
"step": 118
},
{
"epoch": 0.3917695473251029,
"grad_norm": 1.384793758392334,
"learning_rate": 6.220735785953178e-06,
"loss": 0.2182,
"step": 119
},
{
"epoch": 0.3950617283950617,
"grad_norm": 1.2188315391540527,
"learning_rate": 6.1872909698996665e-06,
"loss": 0.1777,
"step": 120
},
{
"epoch": 0.3983539094650206,
"grad_norm": 0.8319004774093628,
"learning_rate": 6.153846153846155e-06,
"loss": 0.1077,
"step": 121
},
{
"epoch": 0.4016460905349794,
"grad_norm": 1.340081810951233,
"learning_rate": 6.120401337792643e-06,
"loss": 0.2474,
"step": 122
},
{
"epoch": 0.4049382716049383,
"grad_norm": 1.0518878698349,
"learning_rate": 6.086956521739132e-06,
"loss": 0.1514,
"step": 123
},
{
"epoch": 0.4082304526748971,
"grad_norm": 1.2594196796417236,
"learning_rate": 6.05351170568562e-06,
"loss": 0.1454,
"step": 124
},
{
"epoch": 0.411522633744856,
"grad_norm": 1.8243649005889893,
"learning_rate": 6.020066889632108e-06,
"loss": 0.3147,
"step": 125
},
{
"epoch": 0.4148148148148148,
"grad_norm": 1.1051766872406006,
"learning_rate": 5.986622073578597e-06,
"loss": 0.1223,
"step": 126
},
{
"epoch": 0.4181069958847737,
"grad_norm": 1.2124512195587158,
"learning_rate": 5.953177257525084e-06,
"loss": 0.2044,
"step": 127
},
{
"epoch": 0.4213991769547325,
"grad_norm": 1.2862944602966309,
"learning_rate": 5.9197324414715726e-06,
"loss": 0.2007,
"step": 128
},
{
"epoch": 0.4246913580246914,
"grad_norm": 1.0270161628723145,
"learning_rate": 5.886287625418061e-06,
"loss": 0.1294,
"step": 129
},
{
"epoch": 0.4279835390946502,
"grad_norm": 1.4540200233459473,
"learning_rate": 5.852842809364549e-06,
"loss": 0.198,
"step": 130
},
{
"epoch": 0.4312757201646091,
"grad_norm": 1.0568363666534424,
"learning_rate": 5.819397993311037e-06,
"loss": 0.1995,
"step": 131
},
{
"epoch": 0.4345679012345679,
"grad_norm": 1.393204689025879,
"learning_rate": 5.785953177257525e-06,
"loss": 0.1952,
"step": 132
},
{
"epoch": 0.4378600823045268,
"grad_norm": 0.7753736972808838,
"learning_rate": 5.7525083612040135e-06,
"loss": 0.0882,
"step": 133
},
{
"epoch": 0.4411522633744856,
"grad_norm": 1.2455613613128662,
"learning_rate": 5.719063545150502e-06,
"loss": 0.1819,
"step": 134
},
{
"epoch": 0.4444444444444444,
"grad_norm": 1.3445703983306885,
"learning_rate": 5.68561872909699e-06,
"loss": 0.1347,
"step": 135
},
{
"epoch": 0.4477366255144033,
"grad_norm": 1.4933937788009644,
"learning_rate": 5.652173913043479e-06,
"loss": 0.2149,
"step": 136
},
{
"epoch": 0.4510288065843621,
"grad_norm": 0.9655482172966003,
"learning_rate": 5.618729096989967e-06,
"loss": 0.1274,
"step": 137
},
{
"epoch": 0.454320987654321,
"grad_norm": 1.2997103929519653,
"learning_rate": 5.585284280936455e-06,
"loss": 0.1526,
"step": 138
},
{
"epoch": 0.4576131687242798,
"grad_norm": 1.1393921375274658,
"learning_rate": 5.551839464882943e-06,
"loss": 0.133,
"step": 139
},
{
"epoch": 0.4609053497942387,
"grad_norm": 1.1311509609222412,
"learning_rate": 5.518394648829431e-06,
"loss": 0.1731,
"step": 140
},
{
"epoch": 0.4641975308641975,
"grad_norm": 1.1631413698196411,
"learning_rate": 5.48494983277592e-06,
"loss": 0.1503,
"step": 141
},
{
"epoch": 0.4674897119341564,
"grad_norm": 1.2505308389663696,
"learning_rate": 5.451505016722408e-06,
"loss": 0.1945,
"step": 142
},
{
"epoch": 0.4707818930041152,
"grad_norm": 1.3157411813735962,
"learning_rate": 5.418060200668896e-06,
"loss": 0.1708,
"step": 143
},
{
"epoch": 0.4740740740740741,
"grad_norm": 1.09896719455719,
"learning_rate": 5.384615384615385e-06,
"loss": 0.1571,
"step": 144
},
{
"epoch": 0.4773662551440329,
"grad_norm": 1.4595751762390137,
"learning_rate": 5.351170568561873e-06,
"loss": 0.3103,
"step": 145
},
{
"epoch": 0.4806584362139918,
"grad_norm": 0.8986005783081055,
"learning_rate": 5.317725752508361e-06,
"loss": 0.0904,
"step": 146
},
{
"epoch": 0.4839506172839506,
"grad_norm": 1.0470385551452637,
"learning_rate": 5.28428093645485e-06,
"loss": 0.1463,
"step": 147
},
{
"epoch": 0.4872427983539095,
"grad_norm": 1.01559317111969,
"learning_rate": 5.250836120401338e-06,
"loss": 0.1545,
"step": 148
},
{
"epoch": 0.4905349794238683,
"grad_norm": 1.498091220855713,
"learning_rate": 5.2173913043478265e-06,
"loss": 0.2597,
"step": 149
},
{
"epoch": 0.49382716049382713,
"grad_norm": 1.0352078676223755,
"learning_rate": 5.183946488294315e-06,
"loss": 0.1061,
"step": 150
},
{
"epoch": 0.49382716049382713,
"eval_loss": 0.21452146768569946,
"eval_runtime": 20.423,
"eval_samples_per_second": 26.441,
"eval_steps_per_second": 13.22,
"step": 150
},
{
"epoch": 0.497119341563786,
"grad_norm": 1.227718472480774,
"learning_rate": 5.150501672240803e-06,
"loss": 0.1595,
"step": 151
},
{
"epoch": 0.5004115226337449,
"grad_norm": 1.6983928680419922,
"learning_rate": 5.1170568561872916e-06,
"loss": 0.2058,
"step": 152
},
{
"epoch": 0.5037037037037037,
"grad_norm": 1.1558934450149536,
"learning_rate": 5.08361204013378e-06,
"loss": 0.1978,
"step": 153
},
{
"epoch": 0.5069958847736625,
"grad_norm": 1.1555702686309814,
"learning_rate": 5.050167224080268e-06,
"loss": 0.1722,
"step": 154
},
{
"epoch": 0.5102880658436214,
"grad_norm": 1.0149585008621216,
"learning_rate": 5.016722408026757e-06,
"loss": 0.163,
"step": 155
},
{
"epoch": 0.5135802469135803,
"grad_norm": 1.016308307647705,
"learning_rate": 4.983277591973244e-06,
"loss": 0.1135,
"step": 156
},
{
"epoch": 0.516872427983539,
"grad_norm": 1.2144166231155396,
"learning_rate": 4.9498327759197325e-06,
"loss": 0.2053,
"step": 157
},
{
"epoch": 0.5201646090534979,
"grad_norm": 1.0773507356643677,
"learning_rate": 4.916387959866221e-06,
"loss": 0.1554,
"step": 158
},
{
"epoch": 0.5234567901234568,
"grad_norm": 0.8874961137771606,
"learning_rate": 4.882943143812709e-06,
"loss": 0.1004,
"step": 159
},
{
"epoch": 0.5267489711934157,
"grad_norm": 1.1773099899291992,
"learning_rate": 4.849498327759198e-06,
"loss": 0.159,
"step": 160
},
{
"epoch": 0.5300411522633744,
"grad_norm": 1.0774908065795898,
"learning_rate": 4.816053511705686e-06,
"loss": 0.1703,
"step": 161
},
{
"epoch": 0.5333333333333333,
"grad_norm": 1.155623435974121,
"learning_rate": 4.782608695652174e-06,
"loss": 0.2199,
"step": 162
},
{
"epoch": 0.5366255144032922,
"grad_norm": 1.3065152168273926,
"learning_rate": 4.749163879598663e-06,
"loss": 0.1445,
"step": 163
},
{
"epoch": 0.5399176954732511,
"grad_norm": 1.207419514656067,
"learning_rate": 4.715719063545151e-06,
"loss": 0.2059,
"step": 164
},
{
"epoch": 0.5432098765432098,
"grad_norm": 1.113638162612915,
"learning_rate": 4.6822742474916394e-06,
"loss": 0.1589,
"step": 165
},
{
"epoch": 0.5465020576131687,
"grad_norm": 1.2327063083648682,
"learning_rate": 4.648829431438128e-06,
"loss": 0.1559,
"step": 166
},
{
"epoch": 0.5497942386831276,
"grad_norm": 1.2051098346710205,
"learning_rate": 4.615384615384616e-06,
"loss": 0.1861,
"step": 167
},
{
"epoch": 0.5530864197530864,
"grad_norm": 0.8835117220878601,
"learning_rate": 4.581939799331104e-06,
"loss": 0.1412,
"step": 168
},
{
"epoch": 0.5563786008230452,
"grad_norm": 1.1923123598098755,
"learning_rate": 4.548494983277592e-06,
"loss": 0.1894,
"step": 169
},
{
"epoch": 0.5596707818930041,
"grad_norm": 2.0249183177948,
"learning_rate": 4.51505016722408e-06,
"loss": 0.297,
"step": 170
},
{
"epoch": 0.562962962962963,
"grad_norm": 0.82772296667099,
"learning_rate": 4.481605351170569e-06,
"loss": 0.0873,
"step": 171
},
{
"epoch": 0.5662551440329218,
"grad_norm": 1.1904910802841187,
"learning_rate": 4.448160535117057e-06,
"loss": 0.1094,
"step": 172
},
{
"epoch": 0.5695473251028806,
"grad_norm": 1.7875385284423828,
"learning_rate": 4.4147157190635455e-06,
"loss": 0.307,
"step": 173
},
{
"epoch": 0.5728395061728395,
"grad_norm": 1.0402239561080933,
"learning_rate": 4.381270903010034e-06,
"loss": 0.1387,
"step": 174
},
{
"epoch": 0.5761316872427984,
"grad_norm": 1.3781318664550781,
"learning_rate": 4.347826086956522e-06,
"loss": 0.1528,
"step": 175
},
{
"epoch": 0.5794238683127572,
"grad_norm": 1.369577169418335,
"learning_rate": 4.3143812709030106e-06,
"loss": 0.2203,
"step": 176
},
{
"epoch": 0.582716049382716,
"grad_norm": 1.260014295578003,
"learning_rate": 4.280936454849499e-06,
"loss": 0.1716,
"step": 177
},
{
"epoch": 0.5860082304526749,
"grad_norm": 1.137498378753662,
"learning_rate": 4.247491638795987e-06,
"loss": 0.1357,
"step": 178
},
{
"epoch": 0.5893004115226338,
"grad_norm": 0.8582133650779724,
"learning_rate": 4.214046822742475e-06,
"loss": 0.0948,
"step": 179
},
{
"epoch": 0.5925925925925926,
"grad_norm": 1.0633777379989624,
"learning_rate": 4.180602006688963e-06,
"loss": 0.1865,
"step": 180
},
{
"epoch": 0.5958847736625514,
"grad_norm": 1.183653712272644,
"learning_rate": 4.1471571906354515e-06,
"loss": 0.162,
"step": 181
},
{
"epoch": 0.5991769547325103,
"grad_norm": 1.1974170207977295,
"learning_rate": 4.11371237458194e-06,
"loss": 0.1406,
"step": 182
},
{
"epoch": 0.6024691358024692,
"grad_norm": 1.8192566633224487,
"learning_rate": 4.080267558528428e-06,
"loss": 0.2827,
"step": 183
},
{
"epoch": 0.605761316872428,
"grad_norm": 1.079257845878601,
"learning_rate": 4.046822742474917e-06,
"loss": 0.1582,
"step": 184
},
{
"epoch": 0.6090534979423868,
"grad_norm": 1.2687796354293823,
"learning_rate": 4.013377926421405e-06,
"loss": 0.1898,
"step": 185
},
{
"epoch": 0.6123456790123457,
"grad_norm": 1.1945315599441528,
"learning_rate": 3.979933110367893e-06,
"loss": 0.2028,
"step": 186
},
{
"epoch": 0.6156378600823045,
"grad_norm": 1.1647101640701294,
"learning_rate": 3.946488294314382e-06,
"loss": 0.1261,
"step": 187
},
{
"epoch": 0.6189300411522634,
"grad_norm": 1.15535569190979,
"learning_rate": 3.91304347826087e-06,
"loss": 0.1781,
"step": 188
},
{
"epoch": 0.6222222222222222,
"grad_norm": 1.1469674110412598,
"learning_rate": 3.8795986622073584e-06,
"loss": 0.189,
"step": 189
},
{
"epoch": 0.6255144032921811,
"grad_norm": 1.2798105478286743,
"learning_rate": 3.846153846153847e-06,
"loss": 0.1553,
"step": 190
},
{
"epoch": 0.6288065843621399,
"grad_norm": 1.1819498538970947,
"learning_rate": 3.812709030100335e-06,
"loss": 0.2014,
"step": 191
},
{
"epoch": 0.6320987654320988,
"grad_norm": 0.9453372359275818,
"learning_rate": 3.7792642140468235e-06,
"loss": 0.1253,
"step": 192
},
{
"epoch": 0.6353909465020576,
"grad_norm": 1.1707664728164673,
"learning_rate": 3.745819397993311e-06,
"loss": 0.1072,
"step": 193
},
{
"epoch": 0.6386831275720165,
"grad_norm": 1.3771895170211792,
"learning_rate": 3.7123745819397994e-06,
"loss": 0.1599,
"step": 194
},
{
"epoch": 0.6419753086419753,
"grad_norm": 1.3412706851959229,
"learning_rate": 3.6789297658862878e-06,
"loss": 0.26,
"step": 195
},
{
"epoch": 0.6452674897119342,
"grad_norm": 1.1179051399230957,
"learning_rate": 3.645484949832776e-06,
"loss": 0.1645,
"step": 196
},
{
"epoch": 0.648559670781893,
"grad_norm": 1.3963730335235596,
"learning_rate": 3.6120401337792645e-06,
"loss": 0.2152,
"step": 197
},
{
"epoch": 0.6518518518518519,
"grad_norm": 1.1093635559082031,
"learning_rate": 3.578595317725753e-06,
"loss": 0.2135,
"step": 198
},
{
"epoch": 0.6551440329218107,
"grad_norm": 1.0025739669799805,
"learning_rate": 3.5451505016722408e-06,
"loss": 0.1666,
"step": 199
},
{
"epoch": 0.6584362139917695,
"grad_norm": 1.1125346422195435,
"learning_rate": 3.511705685618729e-06,
"loss": 0.1792,
"step": 200
},
{
"epoch": 0.6584362139917695,
"eval_loss": 0.21066853404045105,
"eval_runtime": 20.3235,
"eval_samples_per_second": 26.57,
"eval_steps_per_second": 13.285,
"step": 200
},
{
"epoch": 0.6617283950617284,
"grad_norm": 0.9261401891708374,
"learning_rate": 3.4782608695652175e-06,
"loss": 0.1346,
"step": 201
},
{
"epoch": 0.6650205761316872,
"grad_norm": 1.05147123336792,
"learning_rate": 3.444816053511706e-06,
"loss": 0.144,
"step": 202
},
{
"epoch": 0.6683127572016461,
"grad_norm": 1.116198182106018,
"learning_rate": 3.4113712374581942e-06,
"loss": 0.1458,
"step": 203
},
{
"epoch": 0.671604938271605,
"grad_norm": 1.1480129957199097,
"learning_rate": 3.3779264214046826e-06,
"loss": 0.2164,
"step": 204
},
{
"epoch": 0.6748971193415638,
"grad_norm": 1.2926453351974487,
"learning_rate": 3.344481605351171e-06,
"loss": 0.1952,
"step": 205
},
{
"epoch": 0.6781893004115226,
"grad_norm": 0.8901944160461426,
"learning_rate": 3.3110367892976593e-06,
"loss": 0.1197,
"step": 206
},
{
"epoch": 0.6814814814814815,
"grad_norm": 1.1148337125778198,
"learning_rate": 3.2775919732441473e-06,
"loss": 0.1705,
"step": 207
},
{
"epoch": 0.6847736625514403,
"grad_norm": 0.9655947089195251,
"learning_rate": 3.2441471571906356e-06,
"loss": 0.1384,
"step": 208
},
{
"epoch": 0.6880658436213992,
"grad_norm": 1.0507655143737793,
"learning_rate": 3.210702341137124e-06,
"loss": 0.1615,
"step": 209
},
{
"epoch": 0.691358024691358,
"grad_norm": 1.174107313156128,
"learning_rate": 3.1772575250836123e-06,
"loss": 0.1376,
"step": 210
},
{
"epoch": 0.6946502057613169,
"grad_norm": 1.0368140935897827,
"learning_rate": 3.1438127090301007e-06,
"loss": 0.1353,
"step": 211
},
{
"epoch": 0.6979423868312757,
"grad_norm": 1.2840222120285034,
"learning_rate": 3.110367892976589e-06,
"loss": 0.1825,
"step": 212
},
{
"epoch": 0.7012345679012346,
"grad_norm": 1.1343237161636353,
"learning_rate": 3.0769230769230774e-06,
"loss": 0.1562,
"step": 213
},
{
"epoch": 0.7045267489711934,
"grad_norm": 1.2657742500305176,
"learning_rate": 3.043478260869566e-06,
"loss": 0.2141,
"step": 214
},
{
"epoch": 0.7078189300411523,
"grad_norm": 0.9019526839256287,
"learning_rate": 3.010033444816054e-06,
"loss": 0.1025,
"step": 215
},
{
"epoch": 0.7111111111111111,
"grad_norm": 1.1106802225112915,
"learning_rate": 2.976588628762542e-06,
"loss": 0.1448,
"step": 216
},
{
"epoch": 0.7144032921810699,
"grad_norm": 0.8560775518417358,
"learning_rate": 2.9431438127090305e-06,
"loss": 0.0734,
"step": 217
},
{
"epoch": 0.7176954732510288,
"grad_norm": 1.1214252710342407,
"learning_rate": 2.9096989966555184e-06,
"loss": 0.1703,
"step": 218
},
{
"epoch": 0.7209876543209877,
"grad_norm": 1.1007462739944458,
"learning_rate": 2.8762541806020068e-06,
"loss": 0.1685,
"step": 219
},
{
"epoch": 0.7242798353909465,
"grad_norm": 1.1149451732635498,
"learning_rate": 2.842809364548495e-06,
"loss": 0.1679,
"step": 220
},
{
"epoch": 0.7275720164609053,
"grad_norm": 1.3119255304336548,
"learning_rate": 2.8093645484949835e-06,
"loss": 0.1707,
"step": 221
},
{
"epoch": 0.7308641975308642,
"grad_norm": 0.918712317943573,
"learning_rate": 2.7759197324414714e-06,
"loss": 0.0984,
"step": 222
},
{
"epoch": 0.7341563786008231,
"grad_norm": 1.0684345960617065,
"learning_rate": 2.74247491638796e-06,
"loss": 0.1501,
"step": 223
},
{
"epoch": 0.7374485596707819,
"grad_norm": 0.9572702050209045,
"learning_rate": 2.709030100334448e-06,
"loss": 0.1344,
"step": 224
},
{
"epoch": 0.7407407407407407,
"grad_norm": 1.1752140522003174,
"learning_rate": 2.6755852842809365e-06,
"loss": 0.1684,
"step": 225
},
{
"epoch": 0.7440329218106996,
"grad_norm": 1.1885508298873901,
"learning_rate": 2.642140468227425e-06,
"loss": 0.1766,
"step": 226
},
{
"epoch": 0.7473251028806585,
"grad_norm": 0.8276930451393127,
"learning_rate": 2.6086956521739132e-06,
"loss": 0.0797,
"step": 227
},
{
"epoch": 0.7506172839506173,
"grad_norm": 1.5419270992279053,
"learning_rate": 2.5752508361204016e-06,
"loss": 0.2712,
"step": 228
},
{
"epoch": 0.7539094650205761,
"grad_norm": 1.3757195472717285,
"learning_rate": 2.54180602006689e-06,
"loss": 0.1659,
"step": 229
},
{
"epoch": 0.757201646090535,
"grad_norm": 1.044273853302002,
"learning_rate": 2.5083612040133783e-06,
"loss": 0.1705,
"step": 230
},
{
"epoch": 0.7604938271604939,
"grad_norm": 1.2797305583953857,
"learning_rate": 2.4749163879598663e-06,
"loss": 0.2069,
"step": 231
},
{
"epoch": 0.7637860082304527,
"grad_norm": 1.4022698402404785,
"learning_rate": 2.4414715719063546e-06,
"loss": 0.2589,
"step": 232
},
{
"epoch": 0.7670781893004115,
"grad_norm": 1.3316810131072998,
"learning_rate": 2.408026755852843e-06,
"loss": 0.1659,
"step": 233
},
{
"epoch": 0.7703703703703704,
"grad_norm": 1.025339126586914,
"learning_rate": 2.3745819397993314e-06,
"loss": 0.1435,
"step": 234
},
{
"epoch": 0.7736625514403292,
"grad_norm": 0.9907081723213196,
"learning_rate": 2.3411371237458197e-06,
"loss": 0.1458,
"step": 235
},
{
"epoch": 0.776954732510288,
"grad_norm": 1.1326748132705688,
"learning_rate": 2.307692307692308e-06,
"loss": 0.2025,
"step": 236
},
{
"epoch": 0.7802469135802469,
"grad_norm": 1.4742684364318848,
"learning_rate": 2.274247491638796e-06,
"loss": 0.2715,
"step": 237
},
{
"epoch": 0.7835390946502058,
"grad_norm": 1.5240503549575806,
"learning_rate": 2.2408026755852844e-06,
"loss": 0.2928,
"step": 238
},
{
"epoch": 0.7868312757201646,
"grad_norm": 1.1958032846450806,
"learning_rate": 2.2073578595317727e-06,
"loss": 0.179,
"step": 239
},
{
"epoch": 0.7901234567901234,
"grad_norm": 1.4034862518310547,
"learning_rate": 2.173913043478261e-06,
"loss": 0.1899,
"step": 240
},
{
"epoch": 0.7934156378600823,
"grad_norm": 0.9086658954620361,
"learning_rate": 2.1404682274247495e-06,
"loss": 0.0979,
"step": 241
},
{
"epoch": 0.7967078189300412,
"grad_norm": 1.3181670904159546,
"learning_rate": 2.1070234113712374e-06,
"loss": 0.1898,
"step": 242
},
{
"epoch": 0.8,
"grad_norm": 1.5995630025863647,
"learning_rate": 2.0735785953177258e-06,
"loss": 0.2548,
"step": 243
},
{
"epoch": 0.8032921810699588,
"grad_norm": 1.7661391496658325,
"learning_rate": 2.040133779264214e-06,
"loss": 0.158,
"step": 244
},
{
"epoch": 0.8065843621399177,
"grad_norm": 1.6756339073181152,
"learning_rate": 2.0066889632107025e-06,
"loss": 0.2855,
"step": 245
},
{
"epoch": 0.8098765432098766,
"grad_norm": 1.165334939956665,
"learning_rate": 1.973244147157191e-06,
"loss": 0.1677,
"step": 246
},
{
"epoch": 0.8131687242798354,
"grad_norm": 1.4782936573028564,
"learning_rate": 1.9397993311036792e-06,
"loss": 0.2611,
"step": 247
},
{
"epoch": 0.8164609053497942,
"grad_norm": 1.2721978425979614,
"learning_rate": 1.9063545150501676e-06,
"loss": 0.2048,
"step": 248
},
{
"epoch": 0.8197530864197531,
"grad_norm": 0.8288710713386536,
"learning_rate": 1.8729096989966555e-06,
"loss": 0.076,
"step": 249
},
{
"epoch": 0.823045267489712,
"grad_norm": 1.2034995555877686,
"learning_rate": 1.8394648829431439e-06,
"loss": 0.1754,
"step": 250
},
{
"epoch": 0.823045267489712,
"eval_loss": 0.20761683583259583,
"eval_runtime": 20.3983,
"eval_samples_per_second": 26.473,
"eval_steps_per_second": 13.236,
"step": 250
},
{
"epoch": 0.8263374485596707,
"grad_norm": 1.3443710803985596,
"learning_rate": 1.8060200668896322e-06,
"loss": 0.2362,
"step": 251
},
{
"epoch": 0.8296296296296296,
"grad_norm": 1.316748857498169,
"learning_rate": 1.7725752508361204e-06,
"loss": 0.1986,
"step": 252
},
{
"epoch": 0.8329218106995885,
"grad_norm": 1.0996335744857788,
"learning_rate": 1.7391304347826088e-06,
"loss": 0.1276,
"step": 253
},
{
"epoch": 0.8362139917695474,
"grad_norm": 1.7422056198120117,
"learning_rate": 1.7056856187290971e-06,
"loss": 0.2009,
"step": 254
},
{
"epoch": 0.8395061728395061,
"grad_norm": 1.1015255451202393,
"learning_rate": 1.6722408026755855e-06,
"loss": 0.2003,
"step": 255
},
{
"epoch": 0.842798353909465,
"grad_norm": 1.652260184288025,
"learning_rate": 1.6387959866220736e-06,
"loss": 0.2481,
"step": 256
},
{
"epoch": 0.8460905349794239,
"grad_norm": 1.003190279006958,
"learning_rate": 1.605351170568562e-06,
"loss": 0.153,
"step": 257
},
{
"epoch": 0.8493827160493828,
"grad_norm": 1.1012665033340454,
"learning_rate": 1.5719063545150504e-06,
"loss": 0.146,
"step": 258
},
{
"epoch": 0.8526748971193415,
"grad_norm": 1.030242919921875,
"learning_rate": 1.5384615384615387e-06,
"loss": 0.1628,
"step": 259
},
{
"epoch": 0.8559670781893004,
"grad_norm": 1.3571093082427979,
"learning_rate": 1.505016722408027e-06,
"loss": 0.1672,
"step": 260
},
{
"epoch": 0.8592592592592593,
"grad_norm": 1.1187517642974854,
"learning_rate": 1.4715719063545152e-06,
"loss": 0.1366,
"step": 261
},
{
"epoch": 0.8625514403292182,
"grad_norm": 1.0086344480514526,
"learning_rate": 1.4381270903010034e-06,
"loss": 0.1287,
"step": 262
},
{
"epoch": 0.8658436213991769,
"grad_norm": 1.2355748414993286,
"learning_rate": 1.4046822742474917e-06,
"loss": 0.162,
"step": 263
},
{
"epoch": 0.8691358024691358,
"grad_norm": 1.1569114923477173,
"learning_rate": 1.37123745819398e-06,
"loss": 0.1677,
"step": 264
},
{
"epoch": 0.8724279835390947,
"grad_norm": 1.2336384057998657,
"learning_rate": 1.3377926421404683e-06,
"loss": 0.1766,
"step": 265
},
{
"epoch": 0.8757201646090536,
"grad_norm": 1.389993667602539,
"learning_rate": 1.3043478260869566e-06,
"loss": 0.176,
"step": 266
},
{
"epoch": 0.8790123456790123,
"grad_norm": 1.5454909801483154,
"learning_rate": 1.270903010033445e-06,
"loss": 0.2657,
"step": 267
},
{
"epoch": 0.8823045267489712,
"grad_norm": 1.0365800857543945,
"learning_rate": 1.2374581939799331e-06,
"loss": 0.1597,
"step": 268
},
{
"epoch": 0.8855967078189301,
"grad_norm": 1.2040349245071411,
"learning_rate": 1.2040133779264215e-06,
"loss": 0.1646,
"step": 269
},
{
"epoch": 0.8888888888888888,
"grad_norm": 1.556441307067871,
"learning_rate": 1.1705685618729099e-06,
"loss": 0.1863,
"step": 270
},
{
"epoch": 0.8921810699588477,
"grad_norm": 0.9683781862258911,
"learning_rate": 1.137123745819398e-06,
"loss": 0.1467,
"step": 271
},
{
"epoch": 0.8954732510288066,
"grad_norm": 1.211626648902893,
"learning_rate": 1.1036789297658864e-06,
"loss": 0.1847,
"step": 272
},
{
"epoch": 0.8987654320987655,
"grad_norm": 1.4814642667770386,
"learning_rate": 1.0702341137123747e-06,
"loss": 0.1531,
"step": 273
},
{
"epoch": 0.9020576131687242,
"grad_norm": 0.9509296417236328,
"learning_rate": 1.0367892976588629e-06,
"loss": 0.1006,
"step": 274
},
{
"epoch": 0.9053497942386831,
"grad_norm": 1.2226829528808594,
"learning_rate": 1.0033444816053512e-06,
"loss": 0.179,
"step": 275
},
{
"epoch": 0.908641975308642,
"grad_norm": 1.361574411392212,
"learning_rate": 9.698996655518396e-07,
"loss": 0.1983,
"step": 276
},
{
"epoch": 0.9119341563786009,
"grad_norm": 0.9766643643379211,
"learning_rate": 9.364548494983278e-07,
"loss": 0.1369,
"step": 277
},
{
"epoch": 0.9152263374485596,
"grad_norm": 1.490809440612793,
"learning_rate": 9.030100334448161e-07,
"loss": 0.2128,
"step": 278
},
{
"epoch": 0.9185185185185185,
"grad_norm": 1.198103666305542,
"learning_rate": 8.695652173913044e-07,
"loss": 0.1491,
"step": 279
},
{
"epoch": 0.9218106995884774,
"grad_norm": 1.5616865158081055,
"learning_rate": 8.361204013377927e-07,
"loss": 0.2354,
"step": 280
},
{
"epoch": 0.9251028806584363,
"grad_norm": 1.4300072193145752,
"learning_rate": 8.02675585284281e-07,
"loss": 0.2043,
"step": 281
},
{
"epoch": 0.928395061728395,
"grad_norm": 1.1989507675170898,
"learning_rate": 7.692307692307694e-07,
"loss": 0.2068,
"step": 282
},
{
"epoch": 0.9316872427983539,
"grad_norm": 1.4325531721115112,
"learning_rate": 7.357859531772576e-07,
"loss": 0.1631,
"step": 283
},
{
"epoch": 0.9349794238683128,
"grad_norm": 0.9682386517524719,
"learning_rate": 7.023411371237459e-07,
"loss": 0.1067,
"step": 284
},
{
"epoch": 0.9382716049382716,
"grad_norm": 1.1140415668487549,
"learning_rate": 6.688963210702341e-07,
"loss": 0.1887,
"step": 285
},
{
"epoch": 0.9415637860082304,
"grad_norm": 1.3551994562149048,
"learning_rate": 6.354515050167225e-07,
"loss": 0.1419,
"step": 286
},
{
"epoch": 0.9448559670781893,
"grad_norm": 0.988887369632721,
"learning_rate": 6.020066889632107e-07,
"loss": 0.1456,
"step": 287
},
{
"epoch": 0.9481481481481482,
"grad_norm": 1.1655179262161255,
"learning_rate": 5.68561872909699e-07,
"loss": 0.1789,
"step": 288
},
{
"epoch": 0.951440329218107,
"grad_norm": 0.9582886099815369,
"learning_rate": 5.351170568561874e-07,
"loss": 0.1455,
"step": 289
},
{
"epoch": 0.9547325102880658,
"grad_norm": 0.7865822315216064,
"learning_rate": 5.016722408026756e-07,
"loss": 0.078,
"step": 290
},
{
"epoch": 0.9580246913580247,
"grad_norm": 1.1264853477478027,
"learning_rate": 4.682274247491639e-07,
"loss": 0.1626,
"step": 291
},
{
"epoch": 0.9613168724279836,
"grad_norm": 1.4673939943313599,
"learning_rate": 4.347826086956522e-07,
"loss": 0.2005,
"step": 292
},
{
"epoch": 0.9646090534979423,
"grad_norm": 1.222092628479004,
"learning_rate": 4.013377926421405e-07,
"loss": 0.1435,
"step": 293
},
{
"epoch": 0.9679012345679012,
"grad_norm": 1.2059528827667236,
"learning_rate": 3.678929765886288e-07,
"loss": 0.226,
"step": 294
},
{
"epoch": 0.9711934156378601,
"grad_norm": 1.1891047954559326,
"learning_rate": 3.3444816053511706e-07,
"loss": 0.1677,
"step": 295
},
{
"epoch": 0.974485596707819,
"grad_norm": 1.0115400552749634,
"learning_rate": 3.010033444816054e-07,
"loss": 0.1125,
"step": 296
},
{
"epoch": 0.9777777777777777,
"grad_norm": 0.9168137907981873,
"learning_rate": 2.675585284280937e-07,
"loss": 0.1239,
"step": 297
},
{
"epoch": 0.9810699588477366,
"grad_norm": 2.377708911895752,
"learning_rate": 2.3411371237458194e-07,
"loss": 0.3888,
"step": 298
},
{
"epoch": 0.9843621399176955,
"grad_norm": 1.1567450761795044,
"learning_rate": 2.0066889632107025e-07,
"loss": 0.1933,
"step": 299
},
{
"epoch": 0.9876543209876543,
"grad_norm": 1.213723063468933,
"learning_rate": 1.6722408026755853e-07,
"loss": 0.1619,
"step": 300
},
{
"epoch": 0.9876543209876543,
"eval_loss": 0.2061951458454132,
"eval_runtime": 20.4112,
"eval_samples_per_second": 26.456,
"eval_steps_per_second": 13.228,
"step": 300
},
{
"epoch": 0.9909465020576131,
"grad_norm": 1.0399972200393677,
"learning_rate": 1.3377926421404684e-07,
"loss": 0.1235,
"step": 301
},
{
"epoch": 0.994238683127572,
"grad_norm": 0.9800129532814026,
"learning_rate": 1.0033444816053512e-07,
"loss": 0.1255,
"step": 302
},
{
"epoch": 0.9975308641975309,
"grad_norm": 1.3416835069656372,
"learning_rate": 6.688963210702342e-08,
"loss": 0.2557,
"step": 303
},
{
"epoch": 1.0,
"grad_norm": 1.3664584159851074,
"learning_rate": 3.344481605351171e-08,
"loss": 0.1413,
"step": 304
}
],
"logging_steps": 1,
"max_steps": 304,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 10000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.334762911328051e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}