| { | |
| "best_metric": 0.7551020408163265, | |
| "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-0/checkpoint-240", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 384, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2701194286346436, | |
| "learning_rate": 1.7573936619349767e-06, | |
| "loss": 0.7024, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.218135952949524, | |
| "learning_rate": 3.5147873238699533e-06, | |
| "loss": 0.6932, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.5910921096801758, | |
| "learning_rate": 5.2721809858049295e-06, | |
| "loss": 0.6916, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 1.0592715740203857, | |
| "learning_rate": 7.029574647739907e-06, | |
| "loss": 0.6854, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.132177472114563, | |
| "learning_rate": 8.786968309674883e-06, | |
| "loss": 0.6719, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.9851714372634888, | |
| "learning_rate": 1.0544361971609859e-05, | |
| "loss": 0.6633, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.5592440366744995, | |
| "learning_rate": 1.2301755633544835e-05, | |
| "loss": 0.6288, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.8793619871139526, | |
| "learning_rate": 1.4059149295479813e-05, | |
| "loss": 0.6561, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.987308382987976, | |
| "learning_rate": 1.581654295741479e-05, | |
| "loss": 0.6343, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.6868629455566406, | |
| "eval_runtime": 1.3443, | |
| "eval_samples_per_second": 47.609, | |
| "eval_steps_per_second": 5.951, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 3.1652989387512207, | |
| "learning_rate": 1.6792872769600888e-05, | |
| "loss": 0.6712, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.9990371465682983, | |
| "learning_rate": 1.6597606807163667e-05, | |
| "loss": 0.6657, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.8210965394973755, | |
| "learning_rate": 1.640234084472645e-05, | |
| "loss": 0.476, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.5364782810211182, | |
| "learning_rate": 1.6207074882289228e-05, | |
| "loss": 0.5768, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.6772537231445312, | |
| "learning_rate": 1.6011808919852007e-05, | |
| "loss": 0.5039, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.0453933477401733, | |
| "learning_rate": 1.581654295741479e-05, | |
| "loss": 0.7182, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.621067762374878, | |
| "learning_rate": 1.562127699497757e-05, | |
| "loss": 0.6178, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.7878597974777222, | |
| "learning_rate": 1.542601103254035e-05, | |
| "loss": 0.6192, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.6824157238006592, | |
| "learning_rate": 1.5230745070103131e-05, | |
| "loss": 0.6109, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.7989784479141235, | |
| "learning_rate": 1.5035479107665912e-05, | |
| "loss": 0.6367, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.7258987426757812, | |
| "eval_runtime": 1.3947, | |
| "eval_samples_per_second": 45.888, | |
| "eval_steps_per_second": 5.736, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.6805263757705688, | |
| "learning_rate": 1.484021314522869e-05, | |
| "loss": 0.6271, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 0.6469630002975464, | |
| "learning_rate": 1.4644947182791471e-05, | |
| "loss": 0.5085, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 0.9272030591964722, | |
| "learning_rate": 1.4449681220354252e-05, | |
| "loss": 0.6377, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.7515610456466675, | |
| "learning_rate": 1.4254415257917033e-05, | |
| "loss": 0.655, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.1927835941314697, | |
| "learning_rate": 1.4059149295479813e-05, | |
| "loss": 0.5797, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.0289098024368286, | |
| "learning_rate": 1.3863883333042594e-05, | |
| "loss": 0.5291, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 1.7173566818237305, | |
| "learning_rate": 1.3668617370605374e-05, | |
| "loss": 0.6829, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.8574482202529907, | |
| "learning_rate": 1.3473351408168155e-05, | |
| "loss": 0.4509, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 1.1971725225448608, | |
| "learning_rate": 1.3278085445730936e-05, | |
| "loss": 0.4369, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.7368831634521484, | |
| "eval_runtime": 1.4074, | |
| "eval_samples_per_second": 45.475, | |
| "eval_steps_per_second": 5.684, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 2.6530747413635254, | |
| "learning_rate": 1.3082819483293713e-05, | |
| "loss": 0.6167, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 2.0653493404388428, | |
| "learning_rate": 1.2887553520856493e-05, | |
| "loss": 0.5266, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 1.5351653099060059, | |
| "learning_rate": 1.2692287558419274e-05, | |
| "loss": 0.5224, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 1.5351920127868652, | |
| "learning_rate": 1.2497021595982055e-05, | |
| "loss": 0.622, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 1.5447399616241455, | |
| "learning_rate": 1.2301755633544835e-05, | |
| "loss": 0.4725, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 4.063059329986572, | |
| "learning_rate": 1.2106489671107616e-05, | |
| "loss": 0.4089, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 2.517096519470215, | |
| "learning_rate": 1.1911223708670397e-05, | |
| "loss": 0.579, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 2.847086191177368, | |
| "learning_rate": 1.1755010938720622e-05, | |
| "loss": 0.4687, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 1.7841814756393433, | |
| "learning_rate": 1.1559744976283403e-05, | |
| "loss": 0.4846, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 3.5284550189971924, | |
| "learning_rate": 1.1364479013846184e-05, | |
| "loss": 0.5628, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_f1": 0.74, | |
| "eval_loss": 0.6948537826538086, | |
| "eval_runtime": 1.3673, | |
| "eval_samples_per_second": 46.807, | |
| "eval_steps_per_second": 5.851, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 1.6642357110977173, | |
| "learning_rate": 1.1169213051408962e-05, | |
| "loss": 0.5529, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 3.0695078372955322, | |
| "learning_rate": 1.0973947088971741e-05, | |
| "loss": 0.3534, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 2.5939130783081055, | |
| "learning_rate": 1.0778681126534522e-05, | |
| "loss": 0.4534, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": 9.777613639831543, | |
| "learning_rate": 1.0583415164097303e-05, | |
| "loss": 0.5345, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 4.490750312805176, | |
| "learning_rate": 1.0388149201660083e-05, | |
| "loss": 0.4854, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 3.7475597858428955, | |
| "learning_rate": 1.0192883239222864e-05, | |
| "loss": 0.4174, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 1.7532075643539429, | |
| "learning_rate": 9.997617276785644e-06, | |
| "loss": 0.4379, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 1.4538604021072388, | |
| "learning_rate": 9.802351314348425e-06, | |
| "loss": 0.3269, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 3.058314085006714, | |
| "learning_rate": 9.607085351911206e-06, | |
| "loss": 0.3553, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 5.367821216583252, | |
| "learning_rate": 9.45087258196143e-06, | |
| "loss": 0.4753, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_f1": 0.7551020408163265, | |
| "eval_loss": 0.7308206558227539, | |
| "eval_runtime": 1.3546, | |
| "eval_samples_per_second": 47.246, | |
| "eval_steps_per_second": 5.906, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 1.079416275024414, | |
| "learning_rate": 9.25560661952421e-06, | |
| "loss": 0.2547, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 6.890663146972656, | |
| "learning_rate": 9.060340657086991e-06, | |
| "loss": 0.3043, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "grad_norm": 5.391942977905273, | |
| "learning_rate": 8.86507469464977e-06, | |
| "loss": 0.4958, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "grad_norm": Infinity, | |
| "learning_rate": 8.708861924699996e-06, | |
| "loss": 0.4157, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 2.2877392768859863, | |
| "learning_rate": 8.513595962262776e-06, | |
| "loss": 0.289, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "grad_norm": 6.283646106719971, | |
| "learning_rate": 8.318329999825557e-06, | |
| "loss": 0.4315, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 8.02907657623291, | |
| "learning_rate": 8.123064037388336e-06, | |
| "loss": 0.3901, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 5.0022664070129395, | |
| "learning_rate": 7.927798074951117e-06, | |
| "loss": 0.3458, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "grad_norm": 9.924915313720703, | |
| "learning_rate": 7.732532112513897e-06, | |
| "loss": 0.3239, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_f1": 0.7526881720430108, | |
| "eval_loss": 0.734130859375, | |
| "eval_runtime": 1.3629, | |
| "eval_samples_per_second": 46.958, | |
| "eval_steps_per_second": 5.87, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 6.063592910766602, | |
| "learning_rate": 7.537266150076677e-06, | |
| "loss": 0.4556, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "grad_norm": 5.061245441436768, | |
| "learning_rate": 7.3420001876394575e-06, | |
| "loss": 0.2344, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 1.229814052581787, | |
| "learning_rate": 7.146734225202238e-06, | |
| "loss": 0.2622, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "grad_norm": 10.423860549926758, | |
| "learning_rate": 6.951468262765019e-06, | |
| "loss": 0.3694, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 6.46, | |
| "grad_norm": 9.655594825744629, | |
| "learning_rate": 6.756202300327799e-06, | |
| "loss": 0.3442, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "grad_norm": 3.6673190593719482, | |
| "learning_rate": 6.560936337890579e-06, | |
| "loss": 0.3164, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 11.887277603149414, | |
| "learning_rate": 6.36567037545336e-06, | |
| "loss": 0.2361, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.77, | |
| "grad_norm": 12.80982780456543, | |
| "learning_rate": 6.17040441301614e-06, | |
| "loss": 0.3502, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "grad_norm": 14.678775787353516, | |
| "learning_rate": 5.975138450578921e-06, | |
| "loss": 0.2829, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 6.98, | |
| "grad_norm": 2.002340078353882, | |
| "learning_rate": 5.7798724881417015e-06, | |
| "loss": 0.3153, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_f1": 0.7368421052631577, | |
| "eval_loss": 0.8391213417053223, | |
| "eval_runtime": 1.3632, | |
| "eval_samples_per_second": 46.947, | |
| "eval_steps_per_second": 5.868, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 7.08, | |
| "grad_norm": 14.519176483154297, | |
| "learning_rate": 5.584606525704481e-06, | |
| "loss": 0.1675, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "grad_norm": 1.7075341939926147, | |
| "learning_rate": 5.389340563267261e-06, | |
| "loss": 0.3502, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "grad_norm": 1.1975409984588623, | |
| "learning_rate": 5.194074600830042e-06, | |
| "loss": 0.1761, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 24.773109436035156, | |
| "learning_rate": 4.998808638392822e-06, | |
| "loss": 0.2723, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.3228851556777954, | |
| "learning_rate": 4.803542675955603e-06, | |
| "loss": 0.2997, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 11.504044532775879, | |
| "learning_rate": 4.608276713518383e-06, | |
| "loss": 0.167, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 7.71, | |
| "grad_norm": 4.855327129364014, | |
| "learning_rate": 4.413010751081163e-06, | |
| "loss": 0.1576, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 7.81, | |
| "grad_norm": 3.231921434402466, | |
| "learning_rate": 4.217744788643944e-06, | |
| "loss": 0.3102, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "grad_norm": 11.368690490722656, | |
| "learning_rate": 4.022478826206724e-06, | |
| "loss": 0.3451, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_f1": 0.7333333333333333, | |
| "eval_loss": 0.8304653167724609, | |
| "eval_runtime": 1.3846, | |
| "eval_samples_per_second": 46.222, | |
| "eval_steps_per_second": 5.778, | |
| "step": 384 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 480, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 5808783041309760.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.6870979154575775e-05, | |
| "per_device_train_batch_size": 4 | |
| } | |
| } | |