| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9997159897756319, | |
| "eval_steps": 500, | |
| "global_step": 440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01136040897472309, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 4.99840706414248e-06, | |
| "loss": 1.227, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02272081794944618, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 4.993630286525634e-06, | |
| "loss": 1.1593, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03408122692416927, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 4.985675754429744e-06, | |
| "loss": 1.2643, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04544163589889236, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 4.974553604702332e-06, | |
| "loss": 1.1069, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05680204487361545, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 4.96027801084029e-06, | |
| "loss": 1.1399, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06816245384833854, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 4.942867164927899e-06, | |
| "loss": 1.1407, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07952286282306163, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 4.922343254453769e-06, | |
| "loss": 1.1242, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09088327179778471, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 4.8987324340362445e-06, | |
| "loss": 1.2389, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10224368077250781, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 4.8720647920932995e-06, | |
| "loss": 1.1767, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1136040897472309, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 4.842374312499405e-06, | |
| "loss": 1.1298, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12496449872195399, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 4.809698831278217e-06, | |
| "loss": 1.1761, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.13632490769667707, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 4.7740799883862966e-06, | |
| "loss": 1.1133, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14768531667140017, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 4.735563174649278e-06, | |
| "loss": 1.2019, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15904572564612326, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 4.694197473918139e-06, | |
| "loss": 1.064, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17040613462084636, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 4.6500356005192514e-06, | |
| "loss": 1.224, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.18176654359556943, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 4.603133832077953e-06, | |
| "loss": 1.1449, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19312695257029253, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 4.553551937801229e-06, | |
| "loss": 1.1504, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.20448736154501562, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 4.501353102310901e-06, | |
| "loss": 1.1483, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.21584777051973872, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 4.446603845124388e-06, | |
| "loss": 1.2041, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2272081794944618, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 4.3893739358856465e-06, | |
| "loss": 1.1619, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23856858846918488, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 4.329736305454314e-06, | |
| "loss": 1.1075, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.24992899744390798, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 4.267766952966369e-06, | |
| "loss": 1.2116, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2612894064186311, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 4.203544848984729e-06, | |
| "loss": 1.0938, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.27264981539335414, | |
| "grad_norm": 0.625, | |
| "learning_rate": 4.137151834863213e-06, | |
| "loss": 1.2049, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.28401022436807727, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 4.06867251845213e-06, | |
| "loss": 1.1644, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.29537063334280034, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 3.9981941662783675e-06, | |
| "loss": 1.1084, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3067310423175234, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 3.9258065923374104e-06, | |
| "loss": 1.2247, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.31809145129224653, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 3.8516020436389945e-06, | |
| "loss": 1.2347, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3294518602669696, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 3.77567508265225e-06, | |
| "loss": 1.133, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3408122692416927, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 3.6981224668001427e-06, | |
| "loss": 1.1363, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3521726782164158, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 3.6190430251567817e-06, | |
| "loss": 1.1839, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.36353308719113886, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 3.5385375325047167e-06, | |
| "loss": 1.1871, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.374893496165862, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 3.4567085809127247e-06, | |
| "loss": 1.0985, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.38625390514058505, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 3.3736604489977465e-06, | |
| "loss": 1.1271, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3976143141153082, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 3.289498969037563e-06, | |
| "loss": 1.1251, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.40897472309003124, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 3.2043313921035747e-06, | |
| "loss": 1.2161, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4203351320647543, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 3.118266251385539e-06, | |
| "loss": 1.2708, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.43169554103947744, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 3.0314132238824416e-06, | |
| "loss": 1.1735, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4430559500142005, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 2.943882990635759e-06, | |
| "loss": 1.1582, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4544163589889236, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 2.8557870956832135e-06, | |
| "loss": 1.1468, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4657767679636467, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 2.767237803912783e-06, | |
| "loss": 1.1469, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.47713717693836977, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 2.678347957998081e-06, | |
| "loss": 1.1409, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4884975859130929, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 2.5892308345974517e-06, | |
| "loss": 1.1462, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.49985799488781596, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.2359, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.511218403862539, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 2.410769165402549e-06, | |
| "loss": 1.2733, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5225788128372622, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 2.3216520420019194e-06, | |
| "loss": 1.258, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5339392218119853, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 2.2327621960872187e-06, | |
| "loss": 1.1889, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5452996307867083, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 2.1442129043167877e-06, | |
| "loss": 1.1496, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5566600397614314, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 2.0561170093642422e-06, | |
| "loss": 1.1756, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5680204487361545, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 1.9685867761175584e-06, | |
| "loss": 1.2382, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5793808577108776, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 1.8817337486144614e-06, | |
| "loss": 1.1796, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5907412666856007, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 1.7956686078964257e-06, | |
| "loss": 1.1422, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6021016756603238, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 1.7105010309624381e-06, | |
| "loss": 1.2503, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6134620846350468, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 1.6263395510022546e-06, | |
| "loss": 1.2315, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6248224936097699, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 1.5432914190872757e-06, | |
| "loss": 1.1293, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6361829025844931, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 1.4614624674952843e-06, | |
| "loss": 1.2278, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6475433115592162, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 1.380956974843219e-06, | |
| "loss": 1.2107, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6589037205339392, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 1.301877533199859e-06, | |
| "loss": 1.1506, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6702641295086623, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 1.2243249173477514e-06, | |
| "loss": 1.1529, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6816245384833854, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 1.148397956361007e-06, | |
| "loss": 1.116, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6929849474581085, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 1.0741934076625895e-06, | |
| "loss": 1.1773, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7043453564328316, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 1.0018058337216327e-06, | |
| "loss": 1.2332, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7157057654075547, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 9.313274815478698e-07, | |
| "loss": 1.1661, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7270661743822777, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 8.628481651367876e-07, | |
| "loss": 1.1957, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7384265833570008, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 7.964551510152721e-07, | |
| "loss": 1.1944, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.749786992331724, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 7.322330470336314e-07, | |
| "loss": 1.1843, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.761147401306447, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 6.70263694545687e-07, | |
| "loss": 1.329, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7725078102811701, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 6.106260641143547e-07, | |
| "loss": 1.0995, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7838682192558932, | |
| "grad_norm": 0.5, | |
| "learning_rate": 5.533961548756128e-07, | |
| "loss": 1.0897, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7952286282306164, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 4.986468976890993e-07, | |
| "loss": 1.1155, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8065890372053394, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 4.4644806219877183e-07, | |
| "loss": 1.1304, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8179494461800625, | |
| "grad_norm": 0.5, | |
| "learning_rate": 3.9686616792204677e-07, | |
| "loss": 1.1847, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8293098551547856, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 3.499643994807486e-07, | |
| "loss": 1.1612, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8406702641295086, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 3.058025260818609e-07, | |
| "loss": 1.1129, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8520306731042318, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 2.644368253507218e-07, | |
| "loss": 1.1347, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8633910820789549, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 2.2592001161370392e-07, | |
| "loss": 1.2145, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8747514910536779, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 1.9030116872178317e-07, | |
| "loss": 1.069, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.886111900028401, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 1.5762568750059604e-07, | |
| "loss": 1.1762, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8974723090031241, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 1.2793520790670117e-07, | |
| "loss": 1.1969, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9088327179778471, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 1.0126756596375687e-07, | |
| "loss": 1.1673, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9201931269525703, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 7.765674554623182e-08, | |
| "loss": 1.2625, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9315535359272934, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 5.7132835072101486e-08, | |
| "loss": 1.1271, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9429139449020165, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 3.972198915970976e-08, | |
| "loss": 1.2224, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9542743538767395, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 2.544639529766829e-08, | |
| "loss": 1.152, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9656347628514627, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 1.4324245570256634e-08, | |
| "loss": 1.2343, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9769951718261858, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 6.369713474366213e-09, | |
| "loss": 1.2062, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9883555808009088, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 1.5929358575206277e-09, | |
| "loss": 1.1676, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9997159897756319, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.0, | |
| "loss": 1.2424, | |
| "step": 440 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 440, | |
| "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": 1.2268647632338944e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |