| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.955555555555556, | |
| "eval_steps": 1000, | |
| "global_step": 840, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.6363710165023804, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.9339, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.7148603796958923, | |
| "learning_rate": 5e-05, | |
| "loss": 2.8737, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.973531186580658, | |
| "learning_rate": 4.998165452627025e-05, | |
| "loss": 2.7756, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.0926612615585327, | |
| "learning_rate": 4.993379398069312e-05, | |
| "loss": 2.6328, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.2680305242538452, | |
| "learning_rate": 4.9845854359471646e-05, | |
| "loss": 2.5491, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.4471039772033691, | |
| "learning_rate": 4.972145002077457e-05, | |
| "loss": 2.5006, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.439332365989685, | |
| "learning_rate": 4.956076354512411e-05, | |
| "loss": 2.4194, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.5911873579025269, | |
| "learning_rate": 4.9364030762081674e-05, | |
| "loss": 2.4886, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 1.2575428485870361, | |
| "learning_rate": 4.913154040413551e-05, | |
| "loss": 2.3729, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 1.4824930429458618, | |
| "learning_rate": 4.8863633682945956e-05, | |
| "loss": 2.3385, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 1.5880047082901, | |
| "learning_rate": 4.856070378857025e-05, | |
| "loss": 2.2935, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 2.0501768589019775, | |
| "learning_rate": 4.8223195312401935e-05, | |
| "loss": 2.2132, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 1.5746724605560303, | |
| "learning_rate": 4.7851603594671653e-05, | |
| "loss": 2.3534, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 1.7051725387573242, | |
| "learning_rate": 4.744647399746709e-05, | |
| "loss": 2.3176, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.5449475049972534, | |
| "learning_rate": 4.700840110433883e-05, | |
| "loss": 2.3376, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 3.40295147895813, | |
| "learning_rate": 4.65380278476671e-05, | |
| "loss": 2.3002, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 1.291429877281189, | |
| "learning_rate": 4.603604456506977e-05, | |
| "loss": 2.3059, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 1.512455701828003, | |
| "learning_rate": 4.550318798623673e-05, | |
| "loss": 2.2229, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 1.698573112487793, | |
| "learning_rate": 4.4940240151677495e-05, | |
| "loss": 2.2402, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 1.7456746101379395, | |
| "learning_rate": 4.434802726496884e-05, | |
| "loss": 2.2043, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 1.8384778499603271, | |
| "learning_rate": 4.372741848018724e-05, | |
| "loss": 2.2061, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 1.8294689655303955, | |
| "learning_rate": 4.3079324626305294e-05, | |
| "loss": 2.1849, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 2.516619920730591, | |
| "learning_rate": 4.240469687042467e-05, | |
| "loss": 2.1896, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 2.209205150604248, | |
| "learning_rate": 4.17045253218072e-05, | |
| "loss": 2.1794, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 2.061786651611328, | |
| "learning_rate": 4.0979837578752986e-05, | |
| "loss": 2.2459, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 2.4832069873809814, | |
| "learning_rate": 4.023169722045824e-05, | |
| "loss": 2.1904, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 1.6924707889556885, | |
| "learning_rate": 3.946120224606611e-05, | |
| "loss": 2.1928, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 1.4714900255203247, | |
| "learning_rate": 3.8669483463201505e-05, | |
| "loss": 2.1379, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 1.8269611597061157, | |
| "learning_rate": 3.785770282835504e-05, | |
| "loss": 2.2067, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 2.2453367710113525, | |
| "learning_rate": 3.702705174155156e-05, | |
| "loss": 2.1465, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 2.802905797958374, | |
| "learning_rate": 3.617874929780641e-05, | |
| "loss": 2.2089, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 1.8734266757965088, | |
| "learning_rate": 3.5314040497935315e-05, | |
| "loss": 2.2424, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 1.577213168144226, | |
| "learning_rate": 3.443419442134402e-05, | |
| "loss": 2.1432, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 2.034276008605957, | |
| "learning_rate": 3.3540502363479274e-05, | |
| "loss": 2.1789, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "grad_norm": 2.3561315536499023, | |
| "learning_rate": 3.263427594067468e-05, | |
| "loss": 2.1967, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 3.1146442890167236, | |
| "learning_rate": 3.171684516517287e-05, | |
| "loss": 2.1886, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 1.9383080005645752, | |
| "learning_rate": 3.0789556493149094e-05, | |
| "loss": 2.1295, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 2.770754814147949, | |
| "learning_rate": 2.9853770848601164e-05, | |
| "loss": 2.12, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 2.3625290393829346, | |
| "learning_rate": 2.8910861626005776e-05, | |
| "loss": 2.0874, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "grad_norm": 2.091344118118286, | |
| "learning_rate": 2.7962212674672738e-05, | |
| "loss": 2.1171, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.86, | |
| "grad_norm": 2.646287679672241, | |
| "learning_rate": 2.7009216267755293e-05, | |
| "loss": 2.2145, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.98, | |
| "grad_norm": 2.7173330783843994, | |
| "learning_rate": 2.6053271058897244e-05, | |
| "loss": 2.1575, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 2.383789300918579, | |
| "learning_rate": 2.5095780029515896e-05, | |
| "loss": 2.1603, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 1.9369804859161377, | |
| "learning_rate": 2.4138148429733323e-05, | |
| "loss": 2.2005, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "grad_norm": 4.025020122528076, | |
| "learning_rate": 2.3277321440960733e-05, | |
| "loss": 2.1382, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 5.45, | |
| "grad_norm": 2.6233372688293457, | |
| "learning_rate": 2.2323293298151814e-05, | |
| "loss": 2.1584, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "grad_norm": 1.9993385076522827, | |
| "learning_rate": 2.1373193591541378e-05, | |
| "loss": 2.135, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 5.69, | |
| "grad_norm": 2.5544209480285645, | |
| "learning_rate": 2.042841672346608e-05, | |
| "loss": 2.1118, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.81, | |
| "grad_norm": 2.0512688159942627, | |
| "learning_rate": 1.9490349284263033e-05, | |
| "loss": 2.0913, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 5.93, | |
| "grad_norm": 2.2882919311523438, | |
| "learning_rate": 1.8560368017257228e-05, | |
| "loss": 2.0777, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 2.6118929386138916, | |
| "learning_rate": 1.7639837798200923e-05, | |
| "loss": 2.1112, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 6.16, | |
| "grad_norm": 2.71341609954834, | |
| "learning_rate": 1.6730109632130198e-05, | |
| "loss": 2.1188, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 6.28, | |
| "grad_norm": 2.200782299041748, | |
| "learning_rate": 1.5832518670578804e-05, | |
| "loss": 2.1232, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 2.1322591304779053, | |
| "learning_rate": 1.4948382252059156e-05, | |
| "loss": 2.1321, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 6.52, | |
| "grad_norm": 2.813279867172241, | |
| "learning_rate": 1.4078997968686425e-05, | |
| "loss": 2.13, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 6.64, | |
| "grad_norm": 2.4658961296081543, | |
| "learning_rate": 1.3225641761783125e-05, | |
| "loss": 2.1187, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 6.76, | |
| "grad_norm": 2.3500850200653076, | |
| "learning_rate": 1.2389566049259338e-05, | |
| "loss": 2.1212, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 6.87, | |
| "grad_norm": 2.4541094303131104, | |
| "learning_rate": 1.1571997887516672e-05, | |
| "loss": 2.0698, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 6.99, | |
| "grad_norm": 2.5647811889648438, | |
| "learning_rate": 1.0774137170573825e-05, | |
| "loss": 2.0031, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 7.11, | |
| "grad_norm": 2.4078140258789062, | |
| "learning_rate": 9.997154869056589e-06, | |
| "loss": 2.1054, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "grad_norm": 2.722888946533203, | |
| "learning_rate": 9.24219131163705e-06, | |
| "loss": 2.0498, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 7.35, | |
| "grad_norm": 2.400818347930908, | |
| "learning_rate": 8.510354511443974e-06, | |
| "loss": 2.0875, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "grad_norm": 2.3817691802978516, | |
| "learning_rate": 7.80271853990076e-06, | |
| "loss": 2.0564, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 7.59, | |
| "grad_norm": 2.6328699588775635, | |
| "learning_rate": 7.1203219503774875e-06, | |
| "loss": 2.1291, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 2.3237149715423584, | |
| "learning_rate": 6.464166253970671e-06, | |
| "loss": 2.0729, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 7.82, | |
| "grad_norm": 2.3584659099578857, | |
| "learning_rate": 5.835214449647602e-06, | |
| "loss": 2.095, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 7.94, | |
| "grad_norm": 2.5625736713409424, | |
| "learning_rate": 5.234389610912551e-06, | |
| "loss": 2.1288, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 8.06, | |
| "grad_norm": 2.6210007667541504, | |
| "learning_rate": 4.662573531069139e-06, | |
| "loss": 2.1966, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 8.18, | |
| "grad_norm": 2.7513816356658936, | |
| "learning_rate": 4.120605429067054e-06, | |
| "loss": 2.0396, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 3.0069968700408936, | |
| "learning_rate": 3.6092807178324887e-06, | |
| "loss": 2.1039, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "grad_norm": 2.35355806350708, | |
| "learning_rate": 3.1293498368900417e-06, | |
| "loss": 2.1174, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "grad_norm": 2.615985155105591, | |
| "learning_rate": 2.6815171509891852e-06, | |
| "loss": 2.0799, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "grad_norm": 2.601727247238159, | |
| "learning_rate": 2.2664399163518786e-06, | |
| "loss": 2.0811, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "grad_norm": 2.4444971084594727, | |
| "learning_rate": 1.8847273160584378e-06, | |
| "loss": 2.0565, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "grad_norm": 3.528599977493286, | |
| "learning_rate": 1.5369395659873303e-06, | |
| "loss": 2.0009, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "grad_norm": 2.0438976287841797, | |
| "learning_rate": 1.2235870926211619e-06, | |
| "loss": 2.1047, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "grad_norm": 2.847148895263672, | |
| "learning_rate": 9.451297839253914e-07, | |
| "loss": 2.0975, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 9.24, | |
| "grad_norm": 2.4994664192199707, | |
| "learning_rate": 7.019763143993441e-07, | |
| "loss": 2.0354, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 9.36, | |
| "grad_norm": 2.7058956623077393, | |
| "learning_rate": 4.944835452900199e-07, | |
| "loss": 2.0424, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 9.48, | |
| "grad_norm": 2.4226083755493164, | |
| "learning_rate": 3.229560008490007e-07, | |
| "loss": 2.1076, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 2.2777678966522217, | |
| "learning_rate": 1.876454214011253e-07, | |
| "loss": 2.1511, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 9.72, | |
| "grad_norm": 2.749253273010254, | |
| "learning_rate": 8.875039388084316e-08, | |
| "loss": 2.0381, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 9.84, | |
| "grad_norm": 2.434145450592041, | |
| "learning_rate": 2.6416060378503527e-08, | |
| "loss": 2.1139, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 9.96, | |
| "grad_norm": 2.8431334495544434, | |
| "learning_rate": 7.339051243254735e-10, | |
| "loss": 2.1265, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 9.96, | |
| "step": 840, | |
| "total_flos": 1.0731730608083436e+18, | |
| "train_loss": 2.1995077201298305, | |
| "train_runtime": 13749.9632, | |
| "train_samples_per_second": 1.964, | |
| "train_steps_per_second": 0.061 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 840, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 1000, | |
| "total_flos": 1.0731730608083436e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |