| { | |
| "best_metric": 0.5465434193611145, | |
| "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/google_t5/t5_small_ledgar/checkpoint-2800", | |
| "epoch": 3.0, | |
| "eval_steps": 100, | |
| "global_step": 2814, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 1.7876489162445068, | |
| "learning_rate": 0.0004955579246624023, | |
| "loss": 4.3484, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 1.8848645687103271, | |
| "learning_rate": 0.0004911158493248046, | |
| "loss": 3.6753, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.908879041671753, | |
| "learning_rate": 0.00048667377398720687, | |
| "loss": 2.9728, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.578331708908081, | |
| "learning_rate": 0.0004822316986496091, | |
| "loss": 2.3898, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "eval_accuracy": 0.6083, | |
| "eval_f1_macro": 0.3304955770365372, | |
| "eval_f1_micro": 0.6083, | |
| "eval_loss": 1.8530884981155396, | |
| "eval_runtime": 8.4672, | |
| "eval_samples_per_second": 1181.029, | |
| "eval_steps_per_second": 18.542, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 2.0771608352661133, | |
| "learning_rate": 0.0004777896233120114, | |
| "loss": 1.8696, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.535646438598633, | |
| "learning_rate": 0.00047334754797441367, | |
| "loss": 1.6221, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.7007129192352295, | |
| "learning_rate": 0.00046890547263681595, | |
| "loss": 1.3656, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.9887003898620605, | |
| "learning_rate": 0.00046446339729921824, | |
| "loss": 1.1887, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_accuracy": 0.7307, | |
| "eval_f1_macro": 0.5339813428887715, | |
| "eval_f1_micro": 0.7307, | |
| "eval_loss": 1.0730310678482056, | |
| "eval_runtime": 8.4036, | |
| "eval_samples_per_second": 1189.963, | |
| "eval_steps_per_second": 18.682, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.7746949195861816, | |
| "learning_rate": 0.0004600213219616205, | |
| "loss": 1.1038, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 2.2641148567199707, | |
| "learning_rate": 0.00045557924662402275, | |
| "loss": 1.0577, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.5648245811462402, | |
| "learning_rate": 0.000451137171286425, | |
| "loss": 1.0154, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.4375596046447754, | |
| "learning_rate": 0.00044669509594882727, | |
| "loss": 0.946, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.77, | |
| "eval_f1_macro": 0.6068459597214834, | |
| "eval_f1_micro": 0.77, | |
| "eval_loss": 0.8826236128807068, | |
| "eval_runtime": 8.4296, | |
| "eval_samples_per_second": 1186.302, | |
| "eval_steps_per_second": 18.625, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.7988697290420532, | |
| "learning_rate": 0.00044225302061122956, | |
| "loss": 0.9106, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 2.5426275730133057, | |
| "learning_rate": 0.00043781094527363184, | |
| "loss": 0.8913, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 2.555204391479492, | |
| "learning_rate": 0.0004333688699360341, | |
| "loss": 0.8176, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 2.390054941177368, | |
| "learning_rate": 0.0004289267945984364, | |
| "loss": 0.8383, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "eval_accuracy": 0.7851, | |
| "eval_f1_macro": 0.6350877431295691, | |
| "eval_f1_micro": 0.7851, | |
| "eval_loss": 0.8015705943107605, | |
| "eval_runtime": 8.4118, | |
| "eval_samples_per_second": 1188.809, | |
| "eval_steps_per_second": 18.664, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.5415456295013428, | |
| "learning_rate": 0.00042448471926083864, | |
| "loss": 0.8728, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.9339377880096436, | |
| "learning_rate": 0.00042004264392324093, | |
| "loss": 0.9204, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 2.0690979957580566, | |
| "learning_rate": 0.0004156005685856432, | |
| "loss": 0.859, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 2.4634995460510254, | |
| "learning_rate": 0.0004111584932480455, | |
| "loss": 0.8559, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_accuracy": 0.8011, | |
| "eval_f1_macro": 0.674742730667306, | |
| "eval_f1_micro": 0.8011, | |
| "eval_loss": 0.7436844110488892, | |
| "eval_runtime": 8.4108, | |
| "eval_samples_per_second": 1188.952, | |
| "eval_steps_per_second": 18.667, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.8953908681869507, | |
| "learning_rate": 0.0004067164179104478, | |
| "loss": 0.8096, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.3915468454360962, | |
| "learning_rate": 0.00040227434257285007, | |
| "loss": 0.7702, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 3.767996072769165, | |
| "learning_rate": 0.00039783226723525235, | |
| "loss": 0.7342, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 2.76263165473938, | |
| "learning_rate": 0.0003933901918976546, | |
| "loss": 0.7944, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.8091, | |
| "eval_f1_macro": 0.6932776242562073, | |
| "eval_f1_micro": 0.8091, | |
| "eval_loss": 0.706779956817627, | |
| "eval_runtime": 8.4071, | |
| "eval_samples_per_second": 1189.465, | |
| "eval_steps_per_second": 18.675, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 2.232637882232666, | |
| "learning_rate": 0.00038894811656005687, | |
| "loss": 0.7901, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 2.9547877311706543, | |
| "learning_rate": 0.00038450604122245916, | |
| "loss": 0.7668, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.981079578399658, | |
| "learning_rate": 0.00038006396588486144, | |
| "loss": 0.7579, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.8093345165252686, | |
| "learning_rate": 0.0003756218905472637, | |
| "loss": 0.7151, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_accuracy": 0.8191, | |
| "eval_f1_macro": 0.6983413957630255, | |
| "eval_f1_micro": 0.8191, | |
| "eval_loss": 0.6853256821632385, | |
| "eval_runtime": 8.3986, | |
| "eval_samples_per_second": 1190.672, | |
| "eval_steps_per_second": 18.694, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 2.0441739559173584, | |
| "learning_rate": 0.00037117981520966596, | |
| "loss": 0.6854, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.1465673446655273, | |
| "learning_rate": 0.00036673773987206824, | |
| "loss": 0.7541, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.6907949447631836, | |
| "learning_rate": 0.0003622956645344705, | |
| "loss": 0.7349, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 2.384601354598999, | |
| "learning_rate": 0.00035785358919687276, | |
| "loss": 0.7077, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_accuracy": 0.8187, | |
| "eval_f1_macro": 0.7119750119742265, | |
| "eval_f1_micro": 0.8187, | |
| "eval_loss": 0.6665953993797302, | |
| "eval_runtime": 8.3427, | |
| "eval_samples_per_second": 1198.652, | |
| "eval_steps_per_second": 18.819, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 2.5287020206451416, | |
| "learning_rate": 0.00035341151385927504, | |
| "loss": 0.756, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 2.267557382583618, | |
| "learning_rate": 0.00034896943852167733, | |
| "loss": 0.6752, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 2.61407208442688, | |
| "learning_rate": 0.0003445273631840796, | |
| "loss": 0.6903, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 2.658088445663452, | |
| "learning_rate": 0.0003400852878464819, | |
| "loss": 0.6645, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_accuracy": 0.8196, | |
| "eval_f1_macro": 0.7210971149577416, | |
| "eval_f1_micro": 0.8196, | |
| "eval_loss": 0.6476125121116638, | |
| "eval_runtime": 8.3951, | |
| "eval_samples_per_second": 1191.177, | |
| "eval_steps_per_second": 18.701, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 2.093398332595825, | |
| "learning_rate": 0.00033564321250888413, | |
| "loss": 0.6929, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 1.9347296953201294, | |
| "learning_rate": 0.0003312011371712864, | |
| "loss": 0.6415, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 2.2153849601745605, | |
| "learning_rate": 0.0003267590618336887, | |
| "loss": 0.6228, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.8041683435440063, | |
| "learning_rate": 0.000322316986496091, | |
| "loss": 0.5918, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_accuracy": 0.8297, | |
| "eval_f1_macro": 0.7261577527278286, | |
| "eval_f1_micro": 0.8297, | |
| "eval_loss": 0.646891713142395, | |
| "eval_runtime": 8.3945, | |
| "eval_samples_per_second": 1191.262, | |
| "eval_steps_per_second": 18.703, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 1.6319005489349365, | |
| "learning_rate": 0.00031787491115849327, | |
| "loss": 0.5817, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 2.7899856567382812, | |
| "learning_rate": 0.00031343283582089556, | |
| "loss": 0.581, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 2.230139970779419, | |
| "learning_rate": 0.00030899076048329784, | |
| "loss": 0.642, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 2.1724562644958496, | |
| "learning_rate": 0.0003045486851457001, | |
| "loss": 0.5866, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "eval_accuracy": 0.8288, | |
| "eval_f1_macro": 0.7285961959729647, | |
| "eval_f1_micro": 0.8288, | |
| "eval_loss": 0.6308562755584717, | |
| "eval_runtime": 8.3833, | |
| "eval_samples_per_second": 1192.854, | |
| "eval_steps_per_second": 18.728, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.5955884456634521, | |
| "learning_rate": 0.00030010660980810236, | |
| "loss": 0.605, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 1.6877374649047852, | |
| "learning_rate": 0.00029566453447050464, | |
| "loss": 0.5427, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.8254977464675903, | |
| "learning_rate": 0.0002912224591329069, | |
| "loss": 0.5278, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 2.1816537380218506, | |
| "learning_rate": 0.00028678038379530916, | |
| "loss": 0.6665, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.8363, | |
| "eval_f1_macro": 0.7473206082782838, | |
| "eval_f1_micro": 0.8363, | |
| "eval_loss": 0.6188435554504395, | |
| "eval_runtime": 8.3807, | |
| "eval_samples_per_second": 1193.214, | |
| "eval_steps_per_second": 18.733, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 2.975969076156616, | |
| "learning_rate": 0.00028233830845771145, | |
| "loss": 0.6379, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 1.7407147884368896, | |
| "learning_rate": 0.00027789623312011373, | |
| "loss": 0.6198, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 1.3424392938613892, | |
| "learning_rate": 0.00027345415778251596, | |
| "loss": 0.5808, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 3.161609172821045, | |
| "learning_rate": 0.00026901208244491825, | |
| "loss": 0.5684, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_accuracy": 0.837, | |
| "eval_f1_macro": 0.7455872865987888, | |
| "eval_f1_micro": 0.837, | |
| "eval_loss": 0.6117735505104065, | |
| "eval_runtime": 8.3705, | |
| "eval_samples_per_second": 1194.666, | |
| "eval_steps_per_second": 18.756, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 2.2285666465759277, | |
| "learning_rate": 0.00026457000710732053, | |
| "loss": 0.5964, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.9442161321640015, | |
| "learning_rate": 0.0002601279317697228, | |
| "loss": 0.5932, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 2.2172343730926514, | |
| "learning_rate": 0.0002556858564321251, | |
| "loss": 0.5752, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 1.6125428676605225, | |
| "learning_rate": 0.0002512437810945274, | |
| "loss": 0.4986, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "eval_accuracy": 0.8374, | |
| "eval_f1_macro": 0.7519546979290682, | |
| "eval_f1_micro": 0.8374, | |
| "eval_loss": 0.6116703152656555, | |
| "eval_runtime": 8.3786, | |
| "eval_samples_per_second": 1193.512, | |
| "eval_steps_per_second": 18.738, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 2.1590497493743896, | |
| "learning_rate": 0.0002468017057569296, | |
| "loss": 0.5876, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 2.2718493938446045, | |
| "learning_rate": 0.00024235963041933193, | |
| "loss": 0.5973, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.4332152605056763, | |
| "learning_rate": 0.0002379175550817342, | |
| "loss": 0.5829, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.6546821594238281, | |
| "learning_rate": 0.00023347547974413648, | |
| "loss": 0.5786, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_accuracy": 0.8363, | |
| "eval_f1_macro": 0.7461921136413558, | |
| "eval_f1_micro": 0.8363, | |
| "eval_loss": 0.6103752851486206, | |
| "eval_runtime": 8.3693, | |
| "eval_samples_per_second": 1194.837, | |
| "eval_steps_per_second": 18.759, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 1.9171111583709717, | |
| "learning_rate": 0.00022903340440653876, | |
| "loss": 0.5505, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 2.0956099033355713, | |
| "learning_rate": 0.000224591329068941, | |
| "loss": 0.535, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 2.201850414276123, | |
| "learning_rate": 0.00022014925373134328, | |
| "loss": 0.5165, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 1.9713943004608154, | |
| "learning_rate": 0.00021570717839374556, | |
| "loss": 0.5956, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "eval_accuracy": 0.8365, | |
| "eval_f1_macro": 0.7455143804937282, | |
| "eval_f1_micro": 0.8365, | |
| "eval_loss": 0.5964870452880859, | |
| "eval_runtime": 8.3692, | |
| "eval_samples_per_second": 1194.851, | |
| "eval_steps_per_second": 18.759, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 2.684727907180786, | |
| "learning_rate": 0.00021126510305614785, | |
| "loss": 0.5603, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 2.506880283355713, | |
| "learning_rate": 0.0002068230277185501, | |
| "loss": 0.5434, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 2.2245657444000244, | |
| "learning_rate": 0.0002023809523809524, | |
| "loss": 0.5225, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 2.1831839084625244, | |
| "learning_rate": 0.00019793887704335468, | |
| "loss": 0.5653, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "eval_accuracy": 0.8425, | |
| "eval_f1_macro": 0.75879689124166, | |
| "eval_f1_micro": 0.8425, | |
| "eval_loss": 0.5817448496818542, | |
| "eval_runtime": 8.2668, | |
| "eval_samples_per_second": 1209.66, | |
| "eval_steps_per_second": 18.992, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 3.0693764686584473, | |
| "learning_rate": 0.00019349680170575694, | |
| "loss": 0.5874, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 2.015434741973877, | |
| "learning_rate": 0.00018905472636815922, | |
| "loss": 0.5276, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 1.8584706783294678, | |
| "learning_rate": 0.00018461265103056148, | |
| "loss": 0.5956, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 2.165771007537842, | |
| "learning_rate": 0.00018017057569296374, | |
| "loss": 0.5292, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_accuracy": 0.842, | |
| "eval_f1_macro": 0.7515987736281642, | |
| "eval_f1_micro": 0.842, | |
| "eval_loss": 0.5732225775718689, | |
| "eval_runtime": 8.3577, | |
| "eval_samples_per_second": 1196.496, | |
| "eval_steps_per_second": 18.785, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.7540485858917236, | |
| "learning_rate": 0.00017572850035536602, | |
| "loss": 0.6377, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 2.1525564193725586, | |
| "learning_rate": 0.0001712864250177683, | |
| "loss": 0.5995, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.8227964639663696, | |
| "learning_rate": 0.0001668443496801706, | |
| "loss": 0.4946, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 1.2243927717208862, | |
| "learning_rate": 0.00016240227434257285, | |
| "loss": 0.4674, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "eval_accuracy": 0.8456, | |
| "eval_f1_macro": 0.7544341963792838, | |
| "eval_f1_micro": 0.8456, | |
| "eval_loss": 0.5669511556625366, | |
| "eval_runtime": 8.3607, | |
| "eval_samples_per_second": 1196.075, | |
| "eval_steps_per_second": 18.778, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 1.6851863861083984, | |
| "learning_rate": 0.00015796019900497514, | |
| "loss": 0.4851, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.5451288223266602, | |
| "learning_rate": 0.00015351812366737742, | |
| "loss": 0.4883, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "grad_norm": 1.6870522499084473, | |
| "learning_rate": 0.00014907604832977968, | |
| "loss": 0.4975, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 2.1712546348571777, | |
| "learning_rate": 0.00014463397299218194, | |
| "loss": 0.452, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "eval_accuracy": 0.847, | |
| "eval_f1_macro": 0.7615345380474722, | |
| "eval_f1_micro": 0.847, | |
| "eval_loss": 0.5686009526252747, | |
| "eval_runtime": 8.3632, | |
| "eval_samples_per_second": 1195.713, | |
| "eval_steps_per_second": 18.773, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 2.1659963130950928, | |
| "learning_rate": 0.00014019189765458422, | |
| "loss": 0.4517, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 1.7023169994354248, | |
| "learning_rate": 0.00013574982231698648, | |
| "loss": 0.4938, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 1.8300412893295288, | |
| "learning_rate": 0.00013130774697938877, | |
| "loss": 0.4468, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 2.360961675643921, | |
| "learning_rate": 0.00012686567164179105, | |
| "loss": 0.4827, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "eval_accuracy": 0.8461, | |
| "eval_f1_macro": 0.7715708187088293, | |
| "eval_f1_micro": 0.8461, | |
| "eval_loss": 0.5636329054832458, | |
| "eval_runtime": 8.3161, | |
| "eval_samples_per_second": 1202.485, | |
| "eval_steps_per_second": 18.879, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 1.511915683746338, | |
| "learning_rate": 0.0001224235963041933, | |
| "loss": 0.4353, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 1.1389654874801636, | |
| "learning_rate": 0.0001179815209665956, | |
| "loss": 0.4304, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 1.280794620513916, | |
| "learning_rate": 0.00011353944562899787, | |
| "loss": 0.3848, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 1.9939156770706177, | |
| "learning_rate": 0.00010909737029140014, | |
| "loss": 0.4617, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "eval_accuracy": 0.8491, | |
| "eval_f1_macro": 0.7612720077781023, | |
| "eval_f1_micro": 0.8491, | |
| "eval_loss": 0.5610827803611755, | |
| "eval_runtime": 8.3647, | |
| "eval_samples_per_second": 1195.499, | |
| "eval_steps_per_second": 18.769, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 1.6452387571334839, | |
| "learning_rate": 0.00010465529495380242, | |
| "loss": 0.41, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.8779040575027466, | |
| "learning_rate": 0.0001002132196162047, | |
| "loss": 0.4365, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 2.088947296142578, | |
| "learning_rate": 9.577114427860697e-05, | |
| "loss": 0.4681, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 2.114861011505127, | |
| "learning_rate": 9.132906894100924e-05, | |
| "loss": 0.4508, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "eval_accuracy": 0.8499, | |
| "eval_f1_macro": 0.7610245022013704, | |
| "eval_f1_micro": 0.8499, | |
| "eval_loss": 0.5593724846839905, | |
| "eval_runtime": 8.3635, | |
| "eval_samples_per_second": 1195.672, | |
| "eval_steps_per_second": 18.772, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 2.641166925430298, | |
| "learning_rate": 8.688699360341151e-05, | |
| "loss": 0.4928, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 2.3455002307891846, | |
| "learning_rate": 8.24449182658138e-05, | |
| "loss": 0.477, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 2.1197574138641357, | |
| "learning_rate": 7.800284292821607e-05, | |
| "loss": 0.4401, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 2.29925537109375, | |
| "learning_rate": 7.356076759061834e-05, | |
| "loss": 0.432, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_accuracy": 0.85, | |
| "eval_f1_macro": 0.7653751726212001, | |
| "eval_f1_micro": 0.85, | |
| "eval_loss": 0.5532112121582031, | |
| "eval_runtime": 8.3602, | |
| "eval_samples_per_second": 1196.146, | |
| "eval_steps_per_second": 18.779, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 1.8475215435028076, | |
| "learning_rate": 6.911869225302061e-05, | |
| "loss": 0.53, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 2.600374221801758, | |
| "learning_rate": 6.467661691542288e-05, | |
| "loss": 0.4721, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 2.111037492752075, | |
| "learning_rate": 6.023454157782516e-05, | |
| "loss": 0.4328, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 2.5330514907836914, | |
| "learning_rate": 5.579246624022743e-05, | |
| "loss": 0.4298, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_accuracy": 0.8503, | |
| "eval_f1_macro": 0.7666150685407888, | |
| "eval_f1_micro": 0.8503, | |
| "eval_loss": 0.5520814061164856, | |
| "eval_runtime": 8.2724, | |
| "eval_samples_per_second": 1208.838, | |
| "eval_steps_per_second": 18.979, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 2.078260898590088, | |
| "learning_rate": 5.135039090262971e-05, | |
| "loss": 0.473, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 2.062211751937866, | |
| "learning_rate": 4.690831556503199e-05, | |
| "loss": 0.4516, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 2.3843843936920166, | |
| "learning_rate": 4.2466240227434255e-05, | |
| "loss": 0.4763, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "grad_norm": 1.875730037689209, | |
| "learning_rate": 3.802416488983653e-05, | |
| "loss": 0.4627, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_accuracy": 0.85, | |
| "eval_f1_macro": 0.7661219116228528, | |
| "eval_f1_micro": 0.85, | |
| "eval_loss": 0.5510755181312561, | |
| "eval_runtime": 8.3637, | |
| "eval_samples_per_second": 1195.644, | |
| "eval_steps_per_second": 18.772, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.708553671836853, | |
| "learning_rate": 3.3582089552238805e-05, | |
| "loss": 0.4443, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 2.3728208541870117, | |
| "learning_rate": 2.9140014214641083e-05, | |
| "loss": 0.4784, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 2.382185459136963, | |
| "learning_rate": 2.4697938877043355e-05, | |
| "loss": 0.4554, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.8948814868927002, | |
| "learning_rate": 2.025586353944563e-05, | |
| "loss": 0.4353, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_accuracy": 0.8532, | |
| "eval_f1_macro": 0.7705762474067852, | |
| "eval_f1_micro": 0.8532, | |
| "eval_loss": 0.5466225743293762, | |
| "eval_runtime": 8.3561, | |
| "eval_samples_per_second": 1196.726, | |
| "eval_steps_per_second": 18.789, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 1.1043734550476074, | |
| "learning_rate": 1.5813788201847902e-05, | |
| "loss": 0.4379, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 2.514373302459717, | |
| "learning_rate": 1.1371712864250177e-05, | |
| "loss": 0.442, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 2.475358724594116, | |
| "learning_rate": 6.929637526652452e-06, | |
| "loss": 0.4166, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 1.829185128211975, | |
| "learning_rate": 2.4875621890547264e-06, | |
| "loss": 0.4371, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_accuracy": 0.8527, | |
| "eval_f1_macro": 0.7698066205317469, | |
| "eval_f1_micro": 0.8527, | |
| "eval_loss": 0.5465434193611145, | |
| "eval_runtime": 8.2631, | |
| "eval_samples_per_second": 1210.206, | |
| "eval_steps_per_second": 19.0, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 2814, | |
| "total_flos": 6137053555195904.0, | |
| "train_loss": 0.7291999783343089, | |
| "train_runtime": 614.0283, | |
| "train_samples_per_second": 293.146, | |
| "train_steps_per_second": 4.583 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2814, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "total_flos": 6137053555195904.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |