| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9993805492463349, | |
| "eval_steps": 500, | |
| "global_step": 1210, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0008259343382201115, | |
| "grad_norm": 23.24616747351659, | |
| "learning_rate": 8.264462809917357e-08, | |
| "loss": 1.4263, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0041296716911005575, | |
| "grad_norm": 24.213699866017716, | |
| "learning_rate": 4.132231404958678e-07, | |
| "loss": 1.4129, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.008259343382201115, | |
| "grad_norm": 8.707304164936993, | |
| "learning_rate": 8.264462809917356e-07, | |
| "loss": 1.3289, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.012389015073301672, | |
| "grad_norm": 10.918691302163335, | |
| "learning_rate": 1.2396694214876035e-06, | |
| "loss": 1.1721, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01651868676440223, | |
| "grad_norm": 2.87174085951705, | |
| "learning_rate": 1.6528925619834712e-06, | |
| "loss": 1.0219, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.020648358455502787, | |
| "grad_norm": 3.071874166303048, | |
| "learning_rate": 2.066115702479339e-06, | |
| "loss": 0.9724, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.024778030146603344, | |
| "grad_norm": 2.4133173682706213, | |
| "learning_rate": 2.479338842975207e-06, | |
| "loss": 0.9426, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.028907701837703904, | |
| "grad_norm": 2.236585592513117, | |
| "learning_rate": 2.8925619834710743e-06, | |
| "loss": 0.921, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03303737352880446, | |
| "grad_norm": 2.385522667127509, | |
| "learning_rate": 3.3057851239669424e-06, | |
| "loss": 0.9121, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03716704521990502, | |
| "grad_norm": 2.380746269584164, | |
| "learning_rate": 3.71900826446281e-06, | |
| "loss": 0.9001, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.041296716911005574, | |
| "grad_norm": 2.569197332815026, | |
| "learning_rate": 4.132231404958678e-06, | |
| "loss": 0.8867, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.045426388602106134, | |
| "grad_norm": 2.3723293989848213, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 0.8707, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04955606029320669, | |
| "grad_norm": 2.3355520193379142, | |
| "learning_rate": 4.958677685950414e-06, | |
| "loss": 0.8698, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05368573198430725, | |
| "grad_norm": 2.375730929200025, | |
| "learning_rate": 5.371900826446281e-06, | |
| "loss": 0.859, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.05781540367540781, | |
| "grad_norm": 2.3481213761418838, | |
| "learning_rate": 5.785123966942149e-06, | |
| "loss": 0.8632, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06194507536650836, | |
| "grad_norm": 2.3161171161194254, | |
| "learning_rate": 6.198347107438017e-06, | |
| "loss": 0.8462, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06607474705760892, | |
| "grad_norm": 2.273903126512892, | |
| "learning_rate": 6.611570247933885e-06, | |
| "loss": 0.8593, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07020441874870947, | |
| "grad_norm": 2.3651082830117787, | |
| "learning_rate": 7.0247933884297525e-06, | |
| "loss": 0.8388, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07433409043981004, | |
| "grad_norm": 2.5780580351650584, | |
| "learning_rate": 7.43801652892562e-06, | |
| "loss": 0.8354, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0784637621309106, | |
| "grad_norm": 2.5881003929304165, | |
| "learning_rate": 7.851239669421489e-06, | |
| "loss": 0.8177, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.08259343382201115, | |
| "grad_norm": 2.3662580429191262, | |
| "learning_rate": 8.264462809917356e-06, | |
| "loss": 0.8397, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0867231055131117, | |
| "grad_norm": 2.4549245207596173, | |
| "learning_rate": 8.677685950413224e-06, | |
| "loss": 0.8209, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.09085277720421227, | |
| "grad_norm": 2.6986718605229454, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 0.8076, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.09498244889531282, | |
| "grad_norm": 2.6597400001383136, | |
| "learning_rate": 9.50413223140496e-06, | |
| "loss": 0.812, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.09911212058641337, | |
| "grad_norm": 2.6969489551453703, | |
| "learning_rate": 9.917355371900828e-06, | |
| "loss": 0.8064, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.10324179227751394, | |
| "grad_norm": 2.5801588887536204, | |
| "learning_rate": 9.999667111219573e-06, | |
| "loss": 0.7993, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.1073714639686145, | |
| "grad_norm": 2.5177823470227314, | |
| "learning_rate": 9.998314826517564e-06, | |
| "loss": 0.8123, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11150113565971505, | |
| "grad_norm": 2.3846369879591767, | |
| "learning_rate": 9.995922621477252e-06, | |
| "loss": 0.8064, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.11563080735081561, | |
| "grad_norm": 2.37259551970217, | |
| "learning_rate": 9.99249099380692e-06, | |
| "loss": 0.7961, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11976047904191617, | |
| "grad_norm": 2.1621998588898346, | |
| "learning_rate": 9.988020657471078e-06, | |
| "loss": 0.7924, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.12389015073301672, | |
| "grad_norm": 2.3342111944519623, | |
| "learning_rate": 9.98251254254193e-06, | |
| "loss": 0.7785, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12801982242411727, | |
| "grad_norm": 2.3550905722953823, | |
| "learning_rate": 9.97596779500586e-06, | |
| "loss": 0.7695, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.13214949411521784, | |
| "grad_norm": 2.4721774789219344, | |
| "learning_rate": 9.968387776525009e-06, | |
| "loss": 0.7777, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1362791658063184, | |
| "grad_norm": 2.390765566890305, | |
| "learning_rate": 9.959774064153977e-06, | |
| "loss": 0.7865, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.14040883749741895, | |
| "grad_norm": 2.392413139550164, | |
| "learning_rate": 9.950128450011706e-06, | |
| "loss": 0.7723, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14453850918851952, | |
| "grad_norm": 2.24515023866806, | |
| "learning_rate": 9.939452940908627e-06, | |
| "loss": 0.755, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.14866818087962008, | |
| "grad_norm": 2.546933963480071, | |
| "learning_rate": 9.927749757929125e-06, | |
| "loss": 0.7538, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15279785257072062, | |
| "grad_norm": 2.5390977331021136, | |
| "learning_rate": 9.915021335969452e-06, | |
| "loss": 0.7436, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.1569275242618212, | |
| "grad_norm": 2.435644472976208, | |
| "learning_rate": 9.901270323231114e-06, | |
| "loss": 0.7423, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.16105719595292176, | |
| "grad_norm": 2.3962427953474794, | |
| "learning_rate": 9.886499580669917e-06, | |
| "loss": 0.7402, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.1651868676440223, | |
| "grad_norm": 2.293192651317426, | |
| "learning_rate": 9.870712181400726e-06, | |
| "loss": 0.7328, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.16931653933512286, | |
| "grad_norm": 2.3625939065789523, | |
| "learning_rate": 9.853911410058097e-06, | |
| "loss": 0.7414, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.1734462110262234, | |
| "grad_norm": 2.1188762141799633, | |
| "learning_rate": 9.836100762112887e-06, | |
| "loss": 0.7324, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.17757588271732397, | |
| "grad_norm": 2.3459257676868885, | |
| "learning_rate": 9.817283943145014e-06, | |
| "loss": 0.7216, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.18170555440842454, | |
| "grad_norm": 2.353468899573234, | |
| "learning_rate": 9.797464868072489e-06, | |
| "loss": 0.7244, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18583522609952507, | |
| "grad_norm": 2.204966096546786, | |
| "learning_rate": 9.776647660336905e-06, | |
| "loss": 0.7193, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.18996489779062564, | |
| "grad_norm": 2.310006498983248, | |
| "learning_rate": 9.754836651045538e-06, | |
| "loss": 0.7161, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.1940945694817262, | |
| "grad_norm": 2.5104871936969277, | |
| "learning_rate": 9.732036378070243e-06, | |
| "loss": 0.7044, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.19822424117282675, | |
| "grad_norm": 2.366571213288955, | |
| "learning_rate": 9.708251585103324e-06, | |
| "loss": 0.7004, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.20235391286392732, | |
| "grad_norm": 2.570331809825131, | |
| "learning_rate": 9.683487220670595e-06, | |
| "loss": 0.7158, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.20648358455502788, | |
| "grad_norm": 2.4620380694723147, | |
| "learning_rate": 9.657748437101819e-06, | |
| "loss": 0.7019, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.21061325624612842, | |
| "grad_norm": 2.3314886422942513, | |
| "learning_rate": 9.631040589458742e-06, | |
| "loss": 0.6989, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.214742927937229, | |
| "grad_norm": 2.2564618217324743, | |
| "learning_rate": 9.603369234420944e-06, | |
| "loss": 0.6947, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.21887259962832956, | |
| "grad_norm": 2.2447803221031903, | |
| "learning_rate": 9.574740129129767e-06, | |
| "loss": 0.6756, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2230022713194301, | |
| "grad_norm": 2.590402933040932, | |
| "learning_rate": 9.545159229990493e-06, | |
| "loss": 0.6795, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.22713194301053066, | |
| "grad_norm": 2.3017458472761634, | |
| "learning_rate": 9.514632691433108e-06, | |
| "loss": 0.6905, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.23126161470163123, | |
| "grad_norm": 2.363184146884425, | |
| "learning_rate": 9.483166864631837e-06, | |
| "loss": 0.6572, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23539128639273177, | |
| "grad_norm": 2.531694092570608, | |
| "learning_rate": 9.450768296183764e-06, | |
| "loss": 0.6694, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.23952095808383234, | |
| "grad_norm": 2.3202550588633017, | |
| "learning_rate": 9.417443726746776e-06, | |
| "loss": 0.6704, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2436506297749329, | |
| "grad_norm": 2.3120309970976685, | |
| "learning_rate": 9.383200089637143e-06, | |
| "loss": 0.6572, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.24778030146603344, | |
| "grad_norm": 2.3498499961336687, | |
| "learning_rate": 9.348044509387022e-06, | |
| "loss": 0.6655, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.251909973157134, | |
| "grad_norm": 2.5440021629079665, | |
| "learning_rate": 9.311984300262151e-06, | |
| "loss": 0.6612, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.25603964484823455, | |
| "grad_norm": 2.358901339481344, | |
| "learning_rate": 9.275026964740101e-06, | |
| "loss": 0.6444, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2601693165393351, | |
| "grad_norm": 2.206937338630286, | |
| "learning_rate": 9.237180191949347e-06, | |
| "loss": 0.6445, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.2642989882304357, | |
| "grad_norm": 2.226748345548228, | |
| "learning_rate": 9.198451856069514e-06, | |
| "loss": 0.6474, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.26842865992153625, | |
| "grad_norm": 2.366603978699467, | |
| "learning_rate": 9.158850014693123e-06, | |
| "loss": 0.6361, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.2725583316126368, | |
| "grad_norm": 2.2830136404441657, | |
| "learning_rate": 9.118382907149164e-06, | |
| "loss": 0.6395, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.27668800330373733, | |
| "grad_norm": 2.4029340921426727, | |
| "learning_rate": 9.077058952788888e-06, | |
| "loss": 0.6379, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.2808176749948379, | |
| "grad_norm": 2.4477206203283512, | |
| "learning_rate": 9.034886749234112e-06, | |
| "loss": 0.6272, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.28494734668593846, | |
| "grad_norm": 2.3054439086007243, | |
| "learning_rate": 8.991875070588449e-06, | |
| "loss": 0.6279, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.28907701837703903, | |
| "grad_norm": 2.2372770463589196, | |
| "learning_rate": 8.948032865611823e-06, | |
| "loss": 0.6276, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2932066900681396, | |
| "grad_norm": 2.2567729960946803, | |
| "learning_rate": 8.90336925585864e-06, | |
| "loss": 0.6107, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.29733636175924016, | |
| "grad_norm": 2.2634696566360697, | |
| "learning_rate": 8.857893533780015e-06, | |
| "loss": 0.6222, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3014660334503407, | |
| "grad_norm": 2.2158775605445573, | |
| "learning_rate": 8.811615160790428e-06, | |
| "loss": 0.6099, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.30559570514144124, | |
| "grad_norm": 2.359375453911521, | |
| "learning_rate": 8.764543765299245e-06, | |
| "loss": 0.6088, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3097253768325418, | |
| "grad_norm": 2.2574003131115017, | |
| "learning_rate": 8.716689140707488e-06, | |
| "loss": 0.5978, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.3138550485236424, | |
| "grad_norm": 2.305475140301284, | |
| "learning_rate": 8.668061243370273e-06, | |
| "loss": 0.5955, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.31798472021474294, | |
| "grad_norm": 2.512743372685998, | |
| "learning_rate": 8.61867019052535e-06, | |
| "loss": 0.6035, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3221143919058435, | |
| "grad_norm": 2.407216931553165, | |
| "learning_rate": 8.568526258188172e-06, | |
| "loss": 0.5836, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.326244063596944, | |
| "grad_norm": 2.2192688797123066, | |
| "learning_rate": 8.517639879013918e-06, | |
| "loss": 0.5888, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3303737352880446, | |
| "grad_norm": 2.1358480481396267, | |
| "learning_rate": 8.466021640126946e-06, | |
| "loss": 0.5916, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.33450340697914516, | |
| "grad_norm": 2.1925271239115034, | |
| "learning_rate": 8.413682280918094e-06, | |
| "loss": 0.577, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.3386330786702457, | |
| "grad_norm": 2.2104260296514613, | |
| "learning_rate": 8.3606326908103e-06, | |
| "loss": 0.5921, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3427627503613463, | |
| "grad_norm": 2.3351708112925946, | |
| "learning_rate": 8.306883906993022e-06, | |
| "loss": 0.5664, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.3468924220524468, | |
| "grad_norm": 2.129763886062084, | |
| "learning_rate": 8.25244711212589e-06, | |
| "loss": 0.575, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.35102209374354737, | |
| "grad_norm": 2.210580574947652, | |
| "learning_rate": 8.197333632012123e-06, | |
| "loss": 0.5797, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.35515176543464794, | |
| "grad_norm": 2.3295885927377387, | |
| "learning_rate": 8.141554933242135e-06, | |
| "loss": 0.569, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3592814371257485, | |
| "grad_norm": 2.332203260771466, | |
| "learning_rate": 8.08512262080787e-06, | |
| "loss": 0.573, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.36341110881684907, | |
| "grad_norm": 2.13902847391916, | |
| "learning_rate": 8.028048435688333e-06, | |
| "loss": 0.5788, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.36754078050794964, | |
| "grad_norm": 2.3592904701751145, | |
| "learning_rate": 7.970344252406832e-06, | |
| "loss": 0.569, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.37167045219905015, | |
| "grad_norm": 2.4900138472719617, | |
| "learning_rate": 7.912022076560426e-06, | |
| "loss": 0.5697, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3758001238901507, | |
| "grad_norm": 2.1966947004630826, | |
| "learning_rate": 7.853094042322121e-06, | |
| "loss": 0.5577, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.3799297955812513, | |
| "grad_norm": 2.4924992867127, | |
| "learning_rate": 7.7935724099163e-06, | |
| "loss": 0.5564, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.38405946727235185, | |
| "grad_norm": 2.3168755091791775, | |
| "learning_rate": 7.733469563067928e-06, | |
| "loss": 0.558, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.3881891389634524, | |
| "grad_norm": 2.090377992450065, | |
| "learning_rate": 7.67279800642607e-06, | |
| "loss": 0.5472, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.392318810654553, | |
| "grad_norm": 2.19832848896829, | |
| "learning_rate": 7.611570362962247e-06, | |
| "loss": 0.547, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.3964484823456535, | |
| "grad_norm": 2.139436532548822, | |
| "learning_rate": 7.549799371344175e-06, | |
| "loss": 0.5376, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.40057815403675406, | |
| "grad_norm": 2.1648898845336304, | |
| "learning_rate": 7.487497883285428e-06, | |
| "loss": 0.5435, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.40470782572785463, | |
| "grad_norm": 2.1434615200482403, | |
| "learning_rate": 7.424678860871584e-06, | |
| "loss": 0.5347, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.4088374974189552, | |
| "grad_norm": 2.028970353253071, | |
| "learning_rate": 7.361355373863415e-06, | |
| "loss": 0.5349, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.41296716911005577, | |
| "grad_norm": 2.1772432202945624, | |
| "learning_rate": 7.297540596977663e-06, | |
| "loss": 0.532, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41709684080115633, | |
| "grad_norm": 2.2430530701967832, | |
| "learning_rate": 7.233247807145989e-06, | |
| "loss": 0.5385, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.42122651249225684, | |
| "grad_norm": 2.0619754497698035, | |
| "learning_rate": 7.168490380752648e-06, | |
| "loss": 0.5349, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4253561841833574, | |
| "grad_norm": 2.1855473382543993, | |
| "learning_rate": 7.103281790851483e-06, | |
| "loss": 0.5233, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.429485855874458, | |
| "grad_norm": 2.2538669231669712, | |
| "learning_rate": 7.037635604362786e-06, | |
| "loss": 0.5283, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.43361552756555855, | |
| "grad_norm": 2.2135151111234794, | |
| "learning_rate": 6.971565479250659e-06, | |
| "loss": 0.5167, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.4377451992566591, | |
| "grad_norm": 2.383590881907924, | |
| "learning_rate": 6.905085161681408e-06, | |
| "loss": 0.5219, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4418748709477597, | |
| "grad_norm": 2.3721647738464258, | |
| "learning_rate": 6.838208483163601e-06, | |
| "loss": 0.5174, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.4460045426388602, | |
| "grad_norm": 2.338909227744761, | |
| "learning_rate": 6.770949357670358e-06, | |
| "loss": 0.5273, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.45013421432996076, | |
| "grad_norm": 2.1495131685935163, | |
| "learning_rate": 6.703321778744495e-06, | |
| "loss": 0.5186, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.4542638860210613, | |
| "grad_norm": 2.010655198436141, | |
| "learning_rate": 6.635339816587109e-06, | |
| "loss": 0.5164, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4583935577121619, | |
| "grad_norm": 2.1095925263837385, | |
| "learning_rate": 6.5670176151302136e-06, | |
| "loss": 0.4991, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.46252322940326246, | |
| "grad_norm": 2.2264319027764383, | |
| "learning_rate": 6.4983693890940335e-06, | |
| "loss": 0.5195, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.46665290109436297, | |
| "grad_norm": 2.1349760132456246, | |
| "learning_rate": 6.4294094210295725e-06, | |
| "loss": 0.504, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.47078257278546354, | |
| "grad_norm": 2.2896775214795033, | |
| "learning_rate": 6.360152058347068e-06, | |
| "loss": 0.5078, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.4749122444765641, | |
| "grad_norm": 2.1221923536777747, | |
| "learning_rate": 6.290611710330957e-06, | |
| "loss": 0.5059, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.47904191616766467, | |
| "grad_norm": 2.1068836258621406, | |
| "learning_rate": 6.2208028451419575e-06, | |
| "loss": 0.5083, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.48317158785876524, | |
| "grad_norm": 2.0972952063729475, | |
| "learning_rate": 6.150739986806911e-06, | |
| "loss": 0.4923, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.4873012595498658, | |
| "grad_norm": 2.233487162462348, | |
| "learning_rate": 6.0804377121969985e-06, | |
| "loss": 0.4953, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.4914309312409663, | |
| "grad_norm": 2.108166108174955, | |
| "learning_rate": 6.009910647994956e-06, | |
| "loss": 0.4918, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.4955606029320669, | |
| "grad_norm": 2.055777194538188, | |
| "learning_rate": 5.939173467651942e-06, | |
| "loss": 0.483, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.49969027462316745, | |
| "grad_norm": 2.1437965393063374, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.4847, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.503819946314268, | |
| "grad_norm": 2.019446515641709, | |
| "learning_rate": 5.7971276678633625e-06, | |
| "loss": 0.4935, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5079496180053685, | |
| "grad_norm": 2.1084072043678894, | |
| "learning_rate": 5.725848601641492e-06, | |
| "loss": 0.4736, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.5120792896964691, | |
| "grad_norm": 2.185992748799692, | |
| "learning_rate": 5.654418519577369e-06, | |
| "loss": 0.4743, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5162089613875697, | |
| "grad_norm": 2.0909394304010416, | |
| "learning_rate": 5.5828522829987965e-06, | |
| "loss": 0.4808, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.5203386330786702, | |
| "grad_norm": 2.058129072039577, | |
| "learning_rate": 5.511164781561096e-06, | |
| "loss": 0.4727, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5244683047697708, | |
| "grad_norm": 2.1781635883341837, | |
| "learning_rate": 5.439370930149252e-06, | |
| "loss": 0.4665, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.5285979764608714, | |
| "grad_norm": 2.132119418871688, | |
| "learning_rate": 5.367485665774802e-06, | |
| "loss": 0.4647, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5327276481519719, | |
| "grad_norm": 2.0923424452918953, | |
| "learning_rate": 5.295523944468137e-06, | |
| "loss": 0.4752, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.5368573198430725, | |
| "grad_norm": 2.1500817775070704, | |
| "learning_rate": 5.223500738166837e-06, | |
| "loss": 0.4653, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5409869915341731, | |
| "grad_norm": 2.3057702591417764, | |
| "learning_rate": 5.1514310316006835e-06, | |
| "loss": 0.459, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.5451166632252736, | |
| "grad_norm": 2.027442727638895, | |
| "learning_rate": 5.07932981917404e-06, | |
| "loss": 0.467, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5492463349163742, | |
| "grad_norm": 2.003781400613107, | |
| "learning_rate": 5.007212101846194e-06, | |
| "loss": 0.4534, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.5533760066074747, | |
| "grad_norm": 2.1173198648186236, | |
| "learning_rate": 4.935092884010347e-06, | |
| "loss": 0.4582, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5575056782985752, | |
| "grad_norm": 2.1197258530702303, | |
| "learning_rate": 4.8629871703718844e-06, | |
| "loss": 0.4484, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.5616353499896758, | |
| "grad_norm": 2.112938054969618, | |
| "learning_rate": 4.7909099628265946e-06, | |
| "loss": 0.4504, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5657650216807764, | |
| "grad_norm": 2.1008936035245056, | |
| "learning_rate": 4.718876257339444e-06, | |
| "loss": 0.4485, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.5698946933718769, | |
| "grad_norm": 2.925269693235509, | |
| "learning_rate": 4.646901040824622e-06, | |
| "loss": 0.4523, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5740243650629775, | |
| "grad_norm": 2.013226878250545, | |
| "learning_rate": 4.574999288027437e-06, | |
| "loss": 0.4544, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.5781540367540781, | |
| "grad_norm": 2.059493099572322, | |
| "learning_rate": 4.503185958408767e-06, | |
| "loss": 0.4467, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5822837084451786, | |
| "grad_norm": 2.0049953659179716, | |
| "learning_rate": 4.431475993032673e-06, | |
| "loss": 0.441, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.5864133801362792, | |
| "grad_norm": 2.1512212806706255, | |
| "learning_rate": 4.359884311457857e-06, | |
| "loss": 0.4337, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.5905430518273798, | |
| "grad_norm": 2.1177615430048755, | |
| "learning_rate": 4.2884258086335755e-06, | |
| "loss": 0.4458, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.5946727235184803, | |
| "grad_norm": 2.008192166381139, | |
| "learning_rate": 4.217115351800693e-06, | |
| "loss": 0.4413, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.5988023952095808, | |
| "grad_norm": 2.063297374873059, | |
| "learning_rate": 4.145967777398481e-06, | |
| "loss": 0.4344, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.6029320669006814, | |
| "grad_norm": 2.004714733933455, | |
| "learning_rate": 4.074997887977843e-06, | |
| "loss": 0.4358, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6070617385917819, | |
| "grad_norm": 2.025598614826924, | |
| "learning_rate": 4.004220449121574e-06, | |
| "loss": 0.4482, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.6111914102828825, | |
| "grad_norm": 1.938185383839704, | |
| "learning_rate": 3.933650186372329e-06, | |
| "loss": 0.4337, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.615321081973983, | |
| "grad_norm": 2.0495762719124104, | |
| "learning_rate": 3.863301782168896e-06, | |
| "loss": 0.4412, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.6194507536650836, | |
| "grad_norm": 2.028634933360134, | |
| "learning_rate": 3.7931898727914723e-06, | |
| "loss": 0.4293, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6235804253561842, | |
| "grad_norm": 2.0029138326822404, | |
| "learning_rate": 3.7233290453165127e-06, | |
| "loss": 0.4304, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.6277100970472848, | |
| "grad_norm": 2.048375432552439, | |
| "learning_rate": 3.6537338345818273e-06, | |
| "loss": 0.4213, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6318397687383853, | |
| "grad_norm": 2.0326809490611852, | |
| "learning_rate": 3.5844187201625567e-06, | |
| "loss": 0.424, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.6359694404294859, | |
| "grad_norm": 2.019159008104501, | |
| "learning_rate": 3.5153981233586277e-06, | |
| "loss": 0.4228, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6400991121205865, | |
| "grad_norm": 2.09985714316485, | |
| "learning_rate": 3.446686404194337e-06, | |
| "loss": 0.4206, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.644228783811687, | |
| "grad_norm": 1.9414886074513862, | |
| "learning_rate": 3.3782978584307035e-06, | |
| "loss": 0.4173, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6483584555027875, | |
| "grad_norm": 2.3066100655713755, | |
| "learning_rate": 3.310246714591162e-06, | |
| "loss": 0.4248, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.652488127193888, | |
| "grad_norm": 2.106953928933072, | |
| "learning_rate": 3.2425471310012645e-06, | |
| "loss": 0.4159, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6566177988849886, | |
| "grad_norm": 2.197357084281967, | |
| "learning_rate": 3.1752131928429787e-06, | |
| "loss": 0.425, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.6607474705760892, | |
| "grad_norm": 1.9729477591043592, | |
| "learning_rate": 3.1082589092242116e-06, | |
| "loss": 0.4241, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6648771422671897, | |
| "grad_norm": 2.0442009757093875, | |
| "learning_rate": 3.041698210264149e-06, | |
| "loss": 0.4238, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.6690068139582903, | |
| "grad_norm": 2.031799210389413, | |
| "learning_rate": 2.9755449441950434e-06, | |
| "loss": 0.4131, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6731364856493909, | |
| "grad_norm": 2.070029206474213, | |
| "learning_rate": 2.9098128744810245e-06, | |
| "loss": 0.4159, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.6772661573404914, | |
| "grad_norm": 2.0394445933724414, | |
| "learning_rate": 2.84451567695456e-06, | |
| "loss": 0.42, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.681395829031592, | |
| "grad_norm": 2.0377377910207657, | |
| "learning_rate": 2.7796669369711294e-06, | |
| "loss": 0.4095, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.6855255007226926, | |
| "grad_norm": 2.0626337564803965, | |
| "learning_rate": 2.715280146582752e-06, | |
| "loss": 0.4069, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 2.0040834447649454, | |
| "learning_rate": 2.651368701730889e-06, | |
| "loss": 0.41, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.6937848441048936, | |
| "grad_norm": 1.917041220777535, | |
| "learning_rate": 2.5879458994593786e-06, | |
| "loss": 0.4074, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.6979145157959942, | |
| "grad_norm": 2.0269668889490515, | |
| "learning_rate": 2.5250249351479206e-06, | |
| "loss": 0.4081, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.7020441874870947, | |
| "grad_norm": 1.9753046952791828, | |
| "learning_rate": 2.4626188997667224e-06, | |
| "loss": 0.4018, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7061738591781953, | |
| "grad_norm": 2.014280374102551, | |
| "learning_rate": 2.400740777152874e-06, | |
| "loss": 0.4063, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.7103035308692959, | |
| "grad_norm": 2.026632120733398, | |
| "learning_rate": 2.3394034413090015e-06, | |
| "loss": 0.3955, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7144332025603964, | |
| "grad_norm": 2.088872002071019, | |
| "learning_rate": 2.278619653724781e-06, | |
| "loss": 0.4038, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.718562874251497, | |
| "grad_norm": 2.096103572784026, | |
| "learning_rate": 2.218402060721845e-06, | |
| "loss": 0.3921, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7226925459425976, | |
| "grad_norm": 1.9266065409655262, | |
| "learning_rate": 2.1587631908226812e-06, | |
| "loss": 0.4085, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.7268222176336981, | |
| "grad_norm": 1.9536655952796587, | |
| "learning_rate": 2.09971545214401e-06, | |
| "loss": 0.3965, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.7309518893247987, | |
| "grad_norm": 2.061222200669705, | |
| "learning_rate": 2.04127112981522e-06, | |
| "loss": 0.3941, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.7350815610158993, | |
| "grad_norm": 2.010203957849096, | |
| "learning_rate": 1.9834423834224014e-06, | |
| "loss": 0.3982, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.7392112327069998, | |
| "grad_norm": 1.9380026785117144, | |
| "learning_rate": 1.926241244478496e-06, | |
| "loss": 0.3965, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.7433409043981003, | |
| "grad_norm": 2.0416911208044244, | |
| "learning_rate": 1.8696796139200852e-06, | |
| "loss": 0.4015, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7474705760892009, | |
| "grad_norm": 2.040772561659114, | |
| "learning_rate": 1.8137692596313528e-06, | |
| "loss": 0.3827, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.7516002477803014, | |
| "grad_norm": 1.9118935892662992, | |
| "learning_rate": 1.7585218139957205e-06, | |
| "loss": 0.3895, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.755729919471402, | |
| "grad_norm": 1.927575457582737, | |
| "learning_rate": 1.7039487714756953e-06, | |
| "loss": 0.3965, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.7598595911625026, | |
| "grad_norm": 1.9797748596618283, | |
| "learning_rate": 1.6500614862213866e-06, | |
| "loss": 0.3937, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7639892628536031, | |
| "grad_norm": 2.05432833461262, | |
| "learning_rate": 1.596871169708235e-06, | |
| "loss": 0.3891, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.7681189345447037, | |
| "grad_norm": 1.9501310111794177, | |
| "learning_rate": 1.54438888840441e-06, | |
| "loss": 0.3801, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7722486062358043, | |
| "grad_norm": 1.970421056222239, | |
| "learning_rate": 1.4926255614683931e-06, | |
| "loss": 0.3806, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.7763782779269048, | |
| "grad_norm": 2.1277167725190598, | |
| "learning_rate": 1.4415919584771999e-06, | |
| "loss": 0.3843, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.7805079496180054, | |
| "grad_norm": 1.9784981959186678, | |
| "learning_rate": 1.3912986971857168e-06, | |
| "loss": 0.3828, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.784637621309106, | |
| "grad_norm": 2.0003697611347975, | |
| "learning_rate": 1.3417562413176405e-06, | |
| "loss": 0.3859, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.7887672930002065, | |
| "grad_norm": 2.0785112893490485, | |
| "learning_rate": 1.292974898388456e-06, | |
| "loss": 0.3812, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.792896964691307, | |
| "grad_norm": 2.045681460940008, | |
| "learning_rate": 1.2449648175609115e-06, | |
| "loss": 0.3794, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.7970266363824076, | |
| "grad_norm": 2.006461021136274, | |
| "learning_rate": 1.1977359875334454e-06, | |
| "loss": 0.3867, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.8011563080735081, | |
| "grad_norm": 2.0404335899757187, | |
| "learning_rate": 1.1512982344619904e-06, | |
| "loss": 0.3845, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.8052859797646087, | |
| "grad_norm": 2.1247685897806186, | |
| "learning_rate": 1.1056612199156093e-06, | |
| "loss": 0.3838, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.8094156514557093, | |
| "grad_norm": 2.059455987291636, | |
| "learning_rate": 1.0608344388663576e-06, | |
| "loss": 0.3769, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.8135453231468098, | |
| "grad_norm": 2.010190102670995, | |
| "learning_rate": 1.01682721771382e-06, | |
| "loss": 0.38, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.8176749948379104, | |
| "grad_norm": 2.1310463662484755, | |
| "learning_rate": 9.73648712344707e-07, | |
| "loss": 0.3727, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.821804666529011, | |
| "grad_norm": 2.026722500514339, | |
| "learning_rate": 9.313079062279429e-07, | |
| "loss": 0.3775, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.8259343382201115, | |
| "grad_norm": 1.98494077041973, | |
| "learning_rate": 8.898136085456127e-07, | |
| "loss": 0.3795, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8300640099112121, | |
| "grad_norm": 2.0710248547928267, | |
| "learning_rate": 8.491744523601741e-07, | |
| "loss": 0.3789, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.8341936816023127, | |
| "grad_norm": 2.052056544033471, | |
| "learning_rate": 8.093988928183127e-07, | |
| "loss": 0.376, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.8383233532934131, | |
| "grad_norm": 2.0064833870293612, | |
| "learning_rate": 7.70495205391818e-07, | |
| "loss": 0.3728, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.8424530249845137, | |
| "grad_norm": 2.032309328318966, | |
| "learning_rate": 7.324714841558322e-07, | |
| "loss": 0.3701, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.8465826966756143, | |
| "grad_norm": 1.976437679829882, | |
| "learning_rate": 6.953356401048466e-07, | |
| "loss": 0.3796, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.8507123683667148, | |
| "grad_norm": 2.0095046825383927, | |
| "learning_rate": 6.590953995067812e-07, | |
| "loss": 0.3718, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.8548420400578154, | |
| "grad_norm": 2.1362491401601598, | |
| "learning_rate": 6.237583022955079e-07, | |
| "loss": 0.3669, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.858971711748916, | |
| "grad_norm": 2.030103317522303, | |
| "learning_rate": 5.89331700502136e-07, | |
| "loss": 0.3674, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.8631013834400165, | |
| "grad_norm": 2.0320527842019316, | |
| "learning_rate": 5.558227567253832e-07, | |
| "loss": 0.3679, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.8672310551311171, | |
| "grad_norm": 2.1203806417251214, | |
| "learning_rate": 5.232384426413772e-07, | |
| "loss": 0.3697, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8713607268222177, | |
| "grad_norm": 1.9029660119177207, | |
| "learning_rate": 4.915855375531642e-07, | |
| "loss": 0.3708, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.8754903985133182, | |
| "grad_norm": 1.9598889356044642, | |
| "learning_rate": 4.608706269802471e-07, | |
| "loss": 0.3641, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.8796200702044188, | |
| "grad_norm": 2.015130495239755, | |
| "learning_rate": 4.3110010128843747e-07, | |
| "loss": 0.3659, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.8837497418955194, | |
| "grad_norm": 1.9290620704097432, | |
| "learning_rate": 4.022801543603194e-07, | |
| "loss": 0.3648, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.8878794135866198, | |
| "grad_norm": 2.131063629467497, | |
| "learning_rate": 3.744167823065814e-07, | |
| "loss": 0.3619, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.8920090852777204, | |
| "grad_norm": 2.007895538821572, | |
| "learning_rate": 3.4751578221850377e-07, | |
| "loss": 0.3729, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.896138756968821, | |
| "grad_norm": 2.0860597695484113, | |
| "learning_rate": 3.2158275096184744e-07, | |
| "loss": 0.3674, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.9002684286599215, | |
| "grad_norm": 2.0200760244926044, | |
| "learning_rate": 2.966230840124007e-07, | |
| "loss": 0.3677, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.9043981003510221, | |
| "grad_norm": 2.0288061050580053, | |
| "learning_rate": 2.72641974333433e-07, | |
| "loss": 0.3698, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.9085277720421226, | |
| "grad_norm": 2.0237726744767, | |
| "learning_rate": 2.4964441129527337e-07, | |
| "loss": 0.3622, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9126574437332232, | |
| "grad_norm": 1.9537324233727318, | |
| "learning_rate": 2.2763517963725169e-07, | |
| "loss": 0.3678, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.9167871154243238, | |
| "grad_norm": 2.0620765585117837, | |
| "learning_rate": 2.0661885847221398e-07, | |
| "loss": 0.3698, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9209167871154244, | |
| "grad_norm": 1.9666263090601819, | |
| "learning_rate": 1.8659982033381928e-07, | |
| "loss": 0.3615, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.9250464588065249, | |
| "grad_norm": 2.0277290304841142, | |
| "learning_rate": 1.6758223026681507e-07, | |
| "loss": 0.3613, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9291761304976255, | |
| "grad_norm": 2.106138721077295, | |
| "learning_rate": 1.4957004496048256e-07, | |
| "loss": 0.3621, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.9333058021887259, | |
| "grad_norm": 1.9942405326271708, | |
| "learning_rate": 1.3256701192542853e-07, | |
| "loss": 0.3659, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.9374354738798265, | |
| "grad_norm": 1.9945899501208246, | |
| "learning_rate": 1.1657666871390471e-07, | |
| "loss": 0.3671, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.9415651455709271, | |
| "grad_norm": 1.8785796566381439, | |
| "learning_rate": 1.0160234218380305e-07, | |
| "loss": 0.3593, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.9456948172620276, | |
| "grad_norm": 2.025291462525694, | |
| "learning_rate": 8.76471478064872e-08, | |
| "loss": 0.3723, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.9498244889531282, | |
| "grad_norm": 1.9718773659356506, | |
| "learning_rate": 7.471398901860772e-08, | |
| "loss": 0.3665, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.9539541606442288, | |
| "grad_norm": 1.9769770127109265, | |
| "learning_rate": 6.280555661802857e-08, | |
| "loss": 0.3646, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.9580838323353293, | |
| "grad_norm": 2.0267147785284716, | |
| "learning_rate": 5.192432820399718e-08, | |
| "loss": 0.3657, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.9622135040264299, | |
| "grad_norm": 2.07581128095638, | |
| "learning_rate": 4.207256766166845e-08, | |
| "loss": 0.3565, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.9663431757175305, | |
| "grad_norm": 2.0255359776120065, | |
| "learning_rate": 3.3252324691093185e-08, | |
| "loss": 0.3616, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.970472847408631, | |
| "grad_norm": 2.0868262303852396, | |
| "learning_rate": 2.546543438077087e-08, | |
| "loss": 0.3683, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.9746025190997316, | |
| "grad_norm": 2.0003015972698504, | |
| "learning_rate": 1.8713516825851207e-08, | |
| "loss": 0.371, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.9787321907908322, | |
| "grad_norm": 1.9827933568645475, | |
| "learning_rate": 1.2997976791065403e-08, | |
| "loss": 0.3549, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.9828618624819326, | |
| "grad_norm": 2.020348792034948, | |
| "learning_rate": 8.32000341846162e-09, | |
| "loss": 0.3645, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.9869915341730332, | |
| "grad_norm": 1.9998961679855551, | |
| "learning_rate": 4.6805699799967744e-09, | |
| "loss": 0.3624, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.9911212058641338, | |
| "grad_norm": 1.952421429252492, | |
| "learning_rate": 2.0804336750429588e-09, | |
| "loss": 0.3733, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.9952508775552343, | |
| "grad_norm": 2.101729255399416, | |
| "learning_rate": 5.201354728517905e-10, | |
| "loss": 0.3664, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.9993805492463349, | |
| "grad_norm": 2.1247280097277037, | |
| "learning_rate": 0.0, | |
| "loss": 0.3624, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.9993805492463349, | |
| "eval_loss": 0.3181818723678589, | |
| "eval_runtime": 1.1842, | |
| "eval_samples_per_second": 2.533, | |
| "eval_steps_per_second": 0.844, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.9993805492463349, | |
| "step": 1210, | |
| "total_flos": 253297038458880.0, | |
| "train_loss": 0.5432274522860188, | |
| "train_runtime": 27235.9072, | |
| "train_samples_per_second": 1.422, | |
| "train_steps_per_second": 0.044 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1210, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 253297038458880.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |