| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9985734664764622, | |
| "eval_steps": 500, | |
| "global_step": 175, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005706134094151213, | |
| "grad_norm": 0.19783145189285278, | |
| "learning_rate": 8e-07, | |
| "loss": 1.7061, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.011412268188302425, | |
| "grad_norm": 0.07535335421562195, | |
| "learning_rate": 1.6e-06, | |
| "loss": 1.7548, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.017118402282453638, | |
| "grad_norm": 0.06703274697065353, | |
| "learning_rate": 2.4e-06, | |
| "loss": 1.7272, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.02282453637660485, | |
| "grad_norm": 0.07223883271217346, | |
| "learning_rate": 3.2e-06, | |
| "loss": 1.7647, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.028530670470756064, | |
| "grad_norm": 0.0680837407708168, | |
| "learning_rate": 4e-06, | |
| "loss": 1.7567, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.034236804564907276, | |
| "grad_norm": 0.07516428083181381, | |
| "learning_rate": 4.8e-06, | |
| "loss": 1.7508, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.039942938659058486, | |
| "grad_norm": 0.0698549747467041, | |
| "learning_rate": 5.6e-06, | |
| "loss": 1.6912, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0456490727532097, | |
| "grad_norm": 0.0676363930106163, | |
| "learning_rate": 6.4e-06, | |
| "loss": 1.7945, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.05135520684736091, | |
| "grad_norm": 0.07524928450584412, | |
| "learning_rate": 7.2e-06, | |
| "loss": 1.7098, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.05706134094151213, | |
| "grad_norm": 2.2318921089172363, | |
| "learning_rate": 8e-06, | |
| "loss": 1.753, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06276747503566334, | |
| "grad_norm": 0.08612281829118729, | |
| "learning_rate": 7.999274982094103e-06, | |
| "loss": 1.7445, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.06847360912981455, | |
| "grad_norm": 0.10080026090145111, | |
| "learning_rate": 7.997100191201894e-06, | |
| "loss": 1.7603, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07417974322396577, | |
| "grad_norm": 0.0728079155087471, | |
| "learning_rate": 7.993476415704541e-06, | |
| "loss": 1.7145, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.07988587731811697, | |
| "grad_norm": 0.09083904325962067, | |
| "learning_rate": 7.988404969253109e-06, | |
| "loss": 1.7196, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08559201141226819, | |
| "grad_norm": 0.11588940024375916, | |
| "learning_rate": 7.981887690292338e-06, | |
| "loss": 1.7525, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0912981455064194, | |
| "grad_norm": 0.09483379870653152, | |
| "learning_rate": 7.973926941394201e-06, | |
| "loss": 1.7095, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.09700427960057062, | |
| "grad_norm": 0.07794108241796494, | |
| "learning_rate": 7.964525608401445e-06, | |
| "loss": 1.7198, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.10271041369472182, | |
| "grad_norm": 0.08914346247911453, | |
| "learning_rate": 7.953687099381448e-06, | |
| "loss": 1.7009, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.10841654778887304, | |
| "grad_norm": 0.10453473776578903, | |
| "learning_rate": 7.941415343390771e-06, | |
| "loss": 1.6877, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.11412268188302425, | |
| "grad_norm": 0.08587995171546936, | |
| "learning_rate": 7.927714789050827e-06, | |
| "loss": 1.7668, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11982881597717546, | |
| "grad_norm": 0.14489293098449707, | |
| "learning_rate": 7.912590402935222e-06, | |
| "loss": 1.7715, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12553495007132667, | |
| "grad_norm": 0.09673446416854858, | |
| "learning_rate": 7.896047667769334e-06, | |
| "loss": 1.7606, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1312410841654779, | |
| "grad_norm": 0.09498561918735504, | |
| "learning_rate": 7.878092580442764e-06, | |
| "loss": 1.6759, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1369472182596291, | |
| "grad_norm": 2.831144332885742, | |
| "learning_rate": 7.858731649835423e-06, | |
| "loss": 1.7195, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.14265335235378032, | |
| "grad_norm": 0.10375729948282242, | |
| "learning_rate": 7.837971894457989e-06, | |
| "loss": 1.7335, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.14835948644793154, | |
| "grad_norm": 0.10433638095855713, | |
| "learning_rate": 7.81582083990765e-06, | |
| "loss": 1.7296, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.15406562054208273, | |
| "grad_norm": 0.15234149992465973, | |
| "learning_rate": 7.792286516139997e-06, | |
| "loss": 1.7611, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.15977175463623394, | |
| "grad_norm": 0.10203755646944046, | |
| "learning_rate": 7.767377454558098e-06, | |
| "loss": 1.7692, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.16547788873038516, | |
| "grad_norm": 0.11315208673477173, | |
| "learning_rate": 7.741102684919786e-06, | |
| "loss": 1.785, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.17118402282453637, | |
| "grad_norm": 0.6384971141815186, | |
| "learning_rate": 7.71347173206429e-06, | |
| "loss": 1.7158, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1768901569186876, | |
| "grad_norm": 0.09646636992692947, | |
| "learning_rate": 7.6844946124594e-06, | |
| "loss": 1.7082, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1825962910128388, | |
| "grad_norm": 0.10170239955186844, | |
| "learning_rate": 7.654181830570403e-06, | |
| "loss": 1.7142, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.18830242510699002, | |
| "grad_norm": 0.1125100702047348, | |
| "learning_rate": 7.622544375052123e-06, | |
| "loss": 1.7012, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.19400855920114124, | |
| "grad_norm": 0.09590566158294678, | |
| "learning_rate": 7.589593714765433e-06, | |
| "loss": 1.6863, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.19971469329529243, | |
| "grad_norm": 0.12389321625232697, | |
| "learning_rate": 7.555341794619694e-06, | |
| "loss": 1.761, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.20542082738944364, | |
| "grad_norm": 0.1353520005941391, | |
| "learning_rate": 7.519801031242613e-06, | |
| "loss": 1.7201, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.21112696148359486, | |
| "grad_norm": 0.13579188287258148, | |
| "learning_rate": 7.4829843084791085e-06, | |
| "loss": 1.6785, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.21683309557774608, | |
| "grad_norm": 0.10585342347621918, | |
| "learning_rate": 7.4449049727208025e-06, | |
| "loss": 1.7489, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.2225392296718973, | |
| "grad_norm": 0.10238084942102432, | |
| "learning_rate": 7.405576828067827e-06, | |
| "loss": 1.6709, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.2282453637660485, | |
| "grad_norm": 0.15183645486831665, | |
| "learning_rate": 7.365014131324725e-06, | |
| "loss": 1.6628, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.23395149786019973, | |
| "grad_norm": 0.18305522203445435, | |
| "learning_rate": 7.323231586832218e-06, | |
| "loss": 1.694, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.2396576319543509, | |
| "grad_norm": 0.08764177560806274, | |
| "learning_rate": 7.2802443411367645e-06, | |
| "loss": 1.7025, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.24536376604850213, | |
| "grad_norm": 0.11116138845682144, | |
| "learning_rate": 7.2360679774997895e-06, | |
| "loss": 1.7481, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.25106990014265335, | |
| "grad_norm": 0.09760347753763199, | |
| "learning_rate": 7.190718510248621e-06, | |
| "loss": 1.7106, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.25677603423680456, | |
| "grad_norm": 0.1078905463218689, | |
| "learning_rate": 7.1442123789711495e-06, | |
| "loss": 1.7066, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2624821683309558, | |
| "grad_norm": 0.11371272057294846, | |
| "learning_rate": 7.09656644255633e-06, | |
| "loss": 1.7185, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.268188302425107, | |
| "grad_norm": 0.11446070671081543, | |
| "learning_rate": 7.047797973082684e-06, | |
| "loss": 1.6853, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2738944365192582, | |
| "grad_norm": 0.09420523792505264, | |
| "learning_rate": 6.997924649557016e-06, | |
| "loss": 1.6615, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.2796005706134094, | |
| "grad_norm": 0.15625618398189545, | |
| "learning_rate": 6.946964551505619e-06, | |
| "loss": 1.7103, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.28530670470756064, | |
| "grad_norm": 0.18687541782855988, | |
| "learning_rate": 6.89493615242028e-06, | |
| "loss": 1.6903, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.29101283880171186, | |
| "grad_norm": 0.1630401462316513, | |
| "learning_rate": 6.8418583130614755e-06, | |
| "loss": 1.7384, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.2967189728958631, | |
| "grad_norm": 0.10917814075946808, | |
| "learning_rate": 6.7877502746211744e-06, | |
| "loss": 1.6806, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3024251069900143, | |
| "grad_norm": 0.1082145944237709, | |
| "learning_rate": 6.732631651747738e-06, | |
| "loss": 1.767, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.30813124108416545, | |
| "grad_norm": 0.09101668745279312, | |
| "learning_rate": 6.676522425435432e-06, | |
| "loss": 1.7022, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.31383737517831667, | |
| "grad_norm": 0.1459859013557434, | |
| "learning_rate": 6.619442935781141e-06, | |
| "loss": 1.6798, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3195435092724679, | |
| "grad_norm": 0.17180761694908142, | |
| "learning_rate": 6.561413874610889e-06, | |
| "loss": 1.675, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3252496433666191, | |
| "grad_norm": 0.1387602835893631, | |
| "learning_rate": 6.502456277978886e-06, | |
| "loss": 1.6825, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3309557774607703, | |
| "grad_norm": 0.1187388226389885, | |
| "learning_rate": 6.442591518541753e-06, | |
| "loss": 1.7759, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.33666191155492153, | |
| "grad_norm": 0.31210440397262573, | |
| "learning_rate": 6.381841297810752e-06, | |
| "loss": 1.647, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.34236804564907275, | |
| "grad_norm": 0.16231843829154968, | |
| "learning_rate": 6.3202276382847925e-06, | |
| "loss": 1.7275, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.34807417974322397, | |
| "grad_norm": 0.13726073503494263, | |
| "learning_rate": 6.257772875467077e-06, | |
| "loss": 1.7107, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3537803138373752, | |
| "grad_norm": 0.11954201012849808, | |
| "learning_rate": 6.1944996497682805e-06, | |
| "loss": 1.7106, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.3594864479315264, | |
| "grad_norm": 0.17120261490345, | |
| "learning_rate": 6.130430898299199e-06, | |
| "loss": 1.7048, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.3651925820256776, | |
| "grad_norm": 0.2885434627532959, | |
| "learning_rate": 6.065589846555847e-06, | |
| "loss": 1.734, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.37089871611982883, | |
| "grad_norm": 0.11576446890830994, | |
| "learning_rate": 6e-06, | |
| "loss": 1.7462, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.37660485021398005, | |
| "grad_norm": 0.1531599462032318, | |
| "learning_rate": 5.933685135538254e-06, | |
| "loss": 1.7104, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.38231098430813126, | |
| "grad_norm": 0.12773863971233368, | |
| "learning_rate": 5.866669292902695e-06, | |
| "loss": 1.719, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.3880171184022825, | |
| "grad_norm": 0.0994916781783104, | |
| "learning_rate": 5.798976765936263e-06, | |
| "loss": 1.6403, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.39372325249643364, | |
| "grad_norm": 0.09449557214975357, | |
| "learning_rate": 5.730632093786033e-06, | |
| "loss": 1.6675, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.39942938659058486, | |
| "grad_norm": 0.18067103624343872, | |
| "learning_rate": 5.661660052007546e-06, | |
| "loss": 1.7129, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4051355206847361, | |
| "grad_norm": 0.0928979218006134, | |
| "learning_rate": 5.59208564358345e-06, | |
| "loss": 1.6611, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4108416547788873, | |
| "grad_norm": 0.11206196993589401, | |
| "learning_rate": 5.521934089859692e-06, | |
| "loss": 1.6997, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.4165477888730385, | |
| "grad_norm": 0.1811038702726364, | |
| "learning_rate": 5.451230821402563e-06, | |
| "loss": 1.7485, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.4222539229671897, | |
| "grad_norm": 0.24817204475402832, | |
| "learning_rate": 5.380001468779882e-06, | |
| "loss": 1.6898, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.42796005706134094, | |
| "grad_norm": 0.13622242212295532, | |
| "learning_rate": 5.308271853269687e-06, | |
| "loss": 1.7142, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.43366619115549215, | |
| "grad_norm": 0.13679026067256927, | |
| "learning_rate": 5.236067977499789e-06, | |
| "loss": 1.743, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.43937232524964337, | |
| "grad_norm": 0.15019570291042328, | |
| "learning_rate": 5.163416016021596e-06, | |
| "loss": 1.6775, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4450784593437946, | |
| "grad_norm": 0.10396311432123184, | |
| "learning_rate": 5.090342305821591e-06, | |
| "loss": 1.6702, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4507845934379458, | |
| "grad_norm": 0.2947867512702942, | |
| "learning_rate": 5.0168733367739484e-06, | |
| "loss": 1.7389, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.456490727532097, | |
| "grad_norm": 0.23940597474575043, | |
| "learning_rate": 4.943035742037709e-06, | |
| "loss": 1.7602, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.46219686162624823, | |
| "grad_norm": 0.17485539615154266, | |
| "learning_rate": 4.868856288402031e-06, | |
| "loss": 1.7057, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.46790299572039945, | |
| "grad_norm": 0.15298104286193848, | |
| "learning_rate": 4.794361866582981e-06, | |
| "loss": 1.6806, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.47360912981455067, | |
| "grad_norm": 0.18328501284122467, | |
| "learning_rate": 4.719579481475415e-06, | |
| "loss": 1.6852, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4793152639087018, | |
| "grad_norm": 0.13269121944904327, | |
| "learning_rate": 4.644536242363457e-06, | |
| "loss": 1.6662, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.48502139800285304, | |
| "grad_norm": 0.1813397854566574, | |
| "learning_rate": 4.569259353093141e-06, | |
| "loss": 1.7516, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.49072753209700426, | |
| "grad_norm": 0.10082633793354034, | |
| "learning_rate": 4.493776102210779e-06, | |
| "loss": 1.7438, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.4964336661911555, | |
| "grad_norm": 0.10862607508897781, | |
| "learning_rate": 4.418113853070614e-06, | |
| "loss": 1.6985, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5021398002853067, | |
| "grad_norm": 0.1371043622493744, | |
| "learning_rate": 4.342300033915359e-06, | |
| "loss": 1.6684, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5078459343794579, | |
| "grad_norm": 0.19980739057064056, | |
| "learning_rate": 4.266362127933216e-06, | |
| "loss": 1.6687, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5135520684736091, | |
| "grad_norm": 0.1500827819108963, | |
| "learning_rate": 4.1903276632949695e-06, | |
| "loss": 1.7018, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5192582025677603, | |
| "grad_norm": 0.19625936448574066, | |
| "learning_rate": 4.1142242031747846e-06, | |
| "loss": 1.6426, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5249643366619116, | |
| "grad_norm": 0.11851513385772705, | |
| "learning_rate": 4.038079335758307e-06, | |
| "loss": 1.6565, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5306704707560628, | |
| "grad_norm": 0.10231585800647736, | |
| "learning_rate": 3.961920664241693e-06, | |
| "loss": 1.6705, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.536376604850214, | |
| "grad_norm": 0.15432794392108917, | |
| "learning_rate": 3.885775796825215e-06, | |
| "loss": 1.6802, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5420827389443652, | |
| "grad_norm": 0.10955602675676346, | |
| "learning_rate": 3.8096723367050306e-06, | |
| "loss": 1.7885, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5477888730385164, | |
| "grad_norm": 0.11839420348405838, | |
| "learning_rate": 3.733637872066784e-06, | |
| "loss": 1.7251, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5534950071326676, | |
| "grad_norm": 0.19860365986824036, | |
| "learning_rate": 3.657699966084642e-06, | |
| "loss": 1.7, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5592011412268189, | |
| "grad_norm": 0.09971334040164948, | |
| "learning_rate": 3.5818861469293865e-06, | |
| "loss": 1.6763, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5649072753209701, | |
| "grad_norm": 0.1208500787615776, | |
| "learning_rate": 3.506223897789221e-06, | |
| "loss": 1.6746, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5706134094151213, | |
| "grad_norm": 0.3567933142185211, | |
| "learning_rate": 3.4307406469068596e-06, | |
| "loss": 1.6917, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5763195435092725, | |
| "grad_norm": 0.13815858960151672, | |
| "learning_rate": 3.355463757636544e-06, | |
| "loss": 1.6664, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.5820256776034237, | |
| "grad_norm": 0.14547164738178253, | |
| "learning_rate": 3.2804205185245845e-06, | |
| "loss": 1.6911, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5877318116975749, | |
| "grad_norm": 0.12211696803569794, | |
| "learning_rate": 3.205638133417019e-06, | |
| "loss": 1.7104, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5934379457917262, | |
| "grad_norm": 0.09823016822338104, | |
| "learning_rate": 3.1311437115979695e-06, | |
| "loss": 1.6655, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5991440798858774, | |
| "grad_norm": 0.2253655046224594, | |
| "learning_rate": 3.0569642579622904e-06, | |
| "loss": 1.6848, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6048502139800286, | |
| "grad_norm": 0.16188812255859375, | |
| "learning_rate": 2.983126663226053e-06, | |
| "loss": 1.7024, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6105563480741797, | |
| "grad_norm": 0.20521730184555054, | |
| "learning_rate": 2.909657694178409e-06, | |
| "loss": 1.6877, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6162624821683309, | |
| "grad_norm": 0.11846901476383209, | |
| "learning_rate": 2.8365839839784036e-06, | |
| "loss": 1.6708, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6219686162624821, | |
| "grad_norm": 0.1269894540309906, | |
| "learning_rate": 2.7639320225002105e-06, | |
| "loss": 1.7331, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6276747503566333, | |
| "grad_norm": 0.12369225174188614, | |
| "learning_rate": 2.6917281467303133e-06, | |
| "loss": 1.6938, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6333808844507846, | |
| "grad_norm": 0.2189532369375229, | |
| "learning_rate": 2.6199985312201182e-06, | |
| "loss": 1.6623, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6390870185449358, | |
| "grad_norm": 0.12445096671581268, | |
| "learning_rate": 2.5487691785974363e-06, | |
| "loss": 1.72, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.644793152639087, | |
| "grad_norm": 0.3778502941131592, | |
| "learning_rate": 2.478065910140308e-06, | |
| "loss": 1.7012, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6504992867332382, | |
| "grad_norm": 0.12095467001199722, | |
| "learning_rate": 2.407914356416552e-06, | |
| "loss": 1.6972, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6562054208273894, | |
| "grad_norm": 0.1321328580379486, | |
| "learning_rate": 2.3383399479924544e-06, | |
| "loss": 1.702, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6619115549215406, | |
| "grad_norm": 0.3205236792564392, | |
| "learning_rate": 2.269367906213966e-06, | |
| "loss": 1.6933, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6676176890156919, | |
| "grad_norm": 0.24388933181762695, | |
| "learning_rate": 2.2010232340637375e-06, | |
| "loss": 1.6686, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6733238231098431, | |
| "grad_norm": 0.1540047526359558, | |
| "learning_rate": 2.133330707097305e-06, | |
| "loss": 1.699, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.6790299572039943, | |
| "grad_norm": 0.20498210191726685, | |
| "learning_rate": 2.066314864461744e-06, | |
| "loss": 1.7129, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.6847360912981455, | |
| "grad_norm": 0.13441228866577148, | |
| "learning_rate": 2.0000000000000008e-06, | |
| "loss": 1.6787, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6904422253922967, | |
| "grad_norm": 0.16007143259048462, | |
| "learning_rate": 1.9344101534441527e-06, | |
| "loss": 1.7257, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6961483594864479, | |
| "grad_norm": 0.11582008004188538, | |
| "learning_rate": 1.8695691017008004e-06, | |
| "loss": 1.7686, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7018544935805991, | |
| "grad_norm": 0.15291719138622284, | |
| "learning_rate": 1.8055003502317212e-06, | |
| "loss": 1.7115, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7075606276747504, | |
| "grad_norm": 0.13494698703289032, | |
| "learning_rate": 1.742227124532924e-06, | |
| "loss": 1.6658, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7132667617689016, | |
| "grad_norm": 0.1385360211133957, | |
| "learning_rate": 1.6797723617152077e-06, | |
| "loss": 1.7218, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7189728958630528, | |
| "grad_norm": 0.11533591896295547, | |
| "learning_rate": 1.6181587021892483e-06, | |
| "loss": 1.6812, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.724679029957204, | |
| "grad_norm": 0.14457185566425323, | |
| "learning_rate": 1.557408481458247e-06, | |
| "loss": 1.6951, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7303851640513552, | |
| "grad_norm": 0.11555743962526321, | |
| "learning_rate": 1.4975437220211138e-06, | |
| "loss": 1.7104, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.7360912981455064, | |
| "grad_norm": 0.23432128131389618, | |
| "learning_rate": 1.4385861253891109e-06, | |
| "loss": 1.7321, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7417974322396577, | |
| "grad_norm": 0.16473281383514404, | |
| "learning_rate": 1.38055706421886e-06, | |
| "loss": 1.6835, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7475035663338089, | |
| "grad_norm": 0.13933511078357697, | |
| "learning_rate": 1.3234775745645683e-06, | |
| "loss": 1.6737, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7532097004279601, | |
| "grad_norm": 0.1289213001728058, | |
| "learning_rate": 1.2673683482522628e-06, | |
| "loss": 1.6766, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.7589158345221113, | |
| "grad_norm": 0.13357162475585938, | |
| "learning_rate": 1.2122497253788267e-06, | |
| "loss": 1.6895, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.7646219686162625, | |
| "grad_norm": 0.24594484269618988, | |
| "learning_rate": 1.1581416869385252e-06, | |
| "loss": 1.6795, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7703281027104137, | |
| "grad_norm": 0.1374809890985489, | |
| "learning_rate": 1.1050638475797191e-06, | |
| "loss": 1.6351, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.776034236804565, | |
| "grad_norm": 0.12060262262821198, | |
| "learning_rate": 1.0530354484943798e-06, | |
| "loss": 1.7299, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.7817403708987162, | |
| "grad_norm": 0.16807512938976288, | |
| "learning_rate": 1.0020753504429843e-06, | |
| "loss": 1.6747, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.7874465049928673, | |
| "grad_norm": 0.12208617478609085, | |
| "learning_rate": 9.52202026917317e-07, | |
| "loss": 1.7106, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.7931526390870185, | |
| "grad_norm": 0.2812899351119995, | |
| "learning_rate": 9.0343355744367e-07, | |
| "loss": 1.7116, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.7988587731811697, | |
| "grad_norm": 0.15467038750648499, | |
| "learning_rate": 8.557876210288508e-07, | |
| "loss": 1.6687, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8045649072753209, | |
| "grad_norm": 0.24310462176799774, | |
| "learning_rate": 8.092814897513789e-07, | |
| "loss": 1.69, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.8102710413694721, | |
| "grad_norm": 0.36173245310783386, | |
| "learning_rate": 7.639320225002106e-07, | |
| "loss": 1.7107, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.8159771754636234, | |
| "grad_norm": 0.11579599976539612, | |
| "learning_rate": 7.197556588632357e-07, | |
| "loss": 1.7039, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.8216833095577746, | |
| "grad_norm": 0.1441017985343933, | |
| "learning_rate": 6.767684131677813e-07, | |
| "loss": 1.7099, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.8273894436519258, | |
| "grad_norm": 0.6187875270843506, | |
| "learning_rate": 6.349858686752747e-07, | |
| "loss": 1.6932, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.833095577746077, | |
| "grad_norm": 0.20695286989212036, | |
| "learning_rate": 5.944231719321724e-07, | |
| "loss": 1.6552, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.8388017118402282, | |
| "grad_norm": 0.11275847256183624, | |
| "learning_rate": 5.550950272791977e-07, | |
| "loss": 1.67, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8445078459343794, | |
| "grad_norm": 0.1475992500782013, | |
| "learning_rate": 5.170156915208919e-07, | |
| "loss": 1.6698, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8502139800285307, | |
| "grad_norm": 0.12389583885669708, | |
| "learning_rate": 4.801989687573878e-07, | |
| "loss": 1.7556, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8559201141226819, | |
| "grad_norm": 0.11729484051465988, | |
| "learning_rate": 4.4465820538030653e-07, | |
| "loss": 1.6801, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8616262482168331, | |
| "grad_norm": 0.7728609442710876, | |
| "learning_rate": 4.104062852345671e-07, | |
| "loss": 1.7103, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8673323823109843, | |
| "grad_norm": 0.10821712017059326, | |
| "learning_rate": 3.774556249478764e-07, | |
| "loss": 1.6808, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.8730385164051355, | |
| "grad_norm": 0.11125564575195312, | |
| "learning_rate": 3.458181694295961e-07, | |
| "loss": 1.6624, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.8787446504992867, | |
| "grad_norm": 0.32291027903556824, | |
| "learning_rate": 3.155053875406e-07, | |
| "loss": 1.6835, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.884450784593438, | |
| "grad_norm": 0.13799256086349487, | |
| "learning_rate": 2.865282679357097e-07, | |
| "loss": 1.6339, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.8901569186875892, | |
| "grad_norm": 0.13891300559043884, | |
| "learning_rate": 2.588973150802136e-07, | |
| "loss": 1.6816, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.8958630527817404, | |
| "grad_norm": 0.4576370120048523, | |
| "learning_rate": 2.3262254544190153e-07, | |
| "loss": 1.6876, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.9015691868758916, | |
| "grad_norm": 0.10818008333444595, | |
| "learning_rate": 2.0771348386000233e-07, | |
| "loss": 1.7076, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.9072753209700428, | |
| "grad_norm": 0.13714520633220673, | |
| "learning_rate": 1.8417916009234946e-07, | |
| "loss": 1.6874, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.912981455064194, | |
| "grad_norm": 0.2243780493736267, | |
| "learning_rate": 1.6202810554201097e-07, | |
| "loss": 1.6801, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.9186875891583453, | |
| "grad_norm": 0.10469643771648407, | |
| "learning_rate": 1.4126835016457749e-07, | |
| "loss": 1.6569, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.9243937232524965, | |
| "grad_norm": 0.14431771636009216, | |
| "learning_rate": 1.2190741955723494e-07, | |
| "loss": 1.7185, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.9300998573466477, | |
| "grad_norm": 0.15075361728668213, | |
| "learning_rate": 1.0395233223066613e-07, | |
| "loss": 1.6624, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.9358059914407989, | |
| "grad_norm": 0.37998688220977783, | |
| "learning_rate": 8.740959706477724e-08, | |
| "loss": 1.7068, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.9415121255349501, | |
| "grad_norm": 0.4298974275588989, | |
| "learning_rate": 7.228521094917317e-08, | |
| "loss": 1.6794, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.9472182596291013, | |
| "grad_norm": 0.18983392417430878, | |
| "learning_rate": 5.8584656609228065e-08, | |
| "loss": 1.6797, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9529243937232525, | |
| "grad_norm": 0.1515531986951828, | |
| "learning_rate": 4.631290061855031e-08, | |
| "loss": 1.7324, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.9586305278174037, | |
| "grad_norm": 0.18322157859802246, | |
| "learning_rate": 3.547439159855559e-08, | |
| "loss": 1.6838, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9643366619115549, | |
| "grad_norm": 0.15800349414348602, | |
| "learning_rate": 2.6073058605799115e-08, | |
| "loss": 1.669, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.9700427960057061, | |
| "grad_norm": 0.17917828261852264, | |
| "learning_rate": 1.8112309707661643e-08, | |
| "loss": 1.6658, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9757489300998573, | |
| "grad_norm": 0.2858942747116089, | |
| "learning_rate": 1.1595030746890166e-08, | |
| "loss": 1.6626, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.9814550641940085, | |
| "grad_norm": 0.1326640099287033, | |
| "learning_rate": 6.523584295457407e-09, | |
| "loss": 1.689, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.9871611982881597, | |
| "grad_norm": 0.17796051502227783, | |
| "learning_rate": 2.8998087981055853e-09, | |
| "loss": 1.7692, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.992867332382311, | |
| "grad_norm": 0.20098765194416046, | |
| "learning_rate": 7.25017905896852e-10, | |
| "loss": 1.7088, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.9985734664764622, | |
| "grad_norm": 0.16359874606132507, | |
| "learning_rate": 0.0, | |
| "loss": 1.692, | |
| "step": 175 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 175, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.111205269274624e+19, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |