| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.4786729857819905, | |
| "eval_steps": 6, | |
| "global_step": 78, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.018957345971563982, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 2e-05, | |
| "loss": 2.0442, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.018957345971563982, | |
| "eval_loss": 2.073391914367676, | |
| "eval_runtime": 6.8335, | |
| "eval_samples_per_second": 52.096, | |
| "eval_steps_per_second": 13.024, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.037914691943127965, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 4e-05, | |
| "loss": 2.1817, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.05687203791469194, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 6e-05, | |
| "loss": 2.0671, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.07582938388625593, | |
| "grad_norm": 1.5, | |
| "learning_rate": 8e-05, | |
| "loss": 1.9289, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0947867298578199, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 0.0001, | |
| "loss": 1.7705, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.11374407582938388, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.00012, | |
| "loss": 1.449, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.11374407582938388, | |
| "eval_loss": 1.2773768901824951, | |
| "eval_runtime": 7.1055, | |
| "eval_samples_per_second": 50.102, | |
| "eval_steps_per_second": 12.526, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.13270142180094788, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.00014, | |
| "loss": 1.3112, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.15165876777251186, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 0.00016, | |
| "loss": 1.2196, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.17061611374407584, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 0.00018, | |
| "loss": 1.1166, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.1895734597156398, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.0002, | |
| "loss": 1.0836, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.20853080568720378, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00019997685019798912, | |
| "loss": 0.898, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.22748815165876776, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00019990741151022301, | |
| "loss": 0.8548, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.22748815165876776, | |
| "eval_loss": 0.9005841612815857, | |
| "eval_runtime": 6.5385, | |
| "eval_samples_per_second": 54.447, | |
| "eval_steps_per_second": 13.612, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.24644549763033174, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00019979171608653924, | |
| "loss": 0.8835, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.26540284360189575, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00019962981749346078, | |
| "loss": 0.8374, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.2843601895734597, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 0.0001994217906893952, | |
| "loss": 0.8658, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.3033175355450237, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.000199167731989929, | |
| "loss": 0.8232, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.3222748815165877, | |
| "grad_norm": 0.234375, | |
| "learning_rate": 0.00019886775902323405, | |
| "loss": 0.7541, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.3412322274881517, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 0.00019852201067560606, | |
| "loss": 0.8561, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.3412322274881517, | |
| "eval_loss": 0.7924264669418335, | |
| "eval_runtime": 7.7645, | |
| "eval_samples_per_second": 45.85, | |
| "eval_steps_per_second": 11.462, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.36018957345971564, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 0.00019813064702716094, | |
| "loss": 0.7909, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.3791469194312796, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.0001976938492777182, | |
| "loss": 0.7657, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3981042654028436, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 0.00019721181966290613, | |
| "loss": 0.7214, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.41706161137440756, | |
| "grad_norm": 0.33203125, | |
| "learning_rate": 0.00019668478136052774, | |
| "loss": 0.7511, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.43601895734597157, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 0.0001961129783872301, | |
| "loss": 0.7159, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.4549763033175355, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 0.00019549667548552556, | |
| "loss": 0.744, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4549763033175355, | |
| "eval_loss": 0.717645525932312, | |
| "eval_runtime": 6.6339, | |
| "eval_samples_per_second": 53.664, | |
| "eval_steps_per_second": 13.416, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.47393364928909953, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019483615800121716, | |
| "loss": 0.6753, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.4928909952606635, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 0.00019413173175128473, | |
| "loss": 0.7068, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.5118483412322274, | |
| "grad_norm": 0.263671875, | |
| "learning_rate": 0.0001933837228822925, | |
| "loss": 0.6545, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.5308056872037915, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 0.000192592477719385, | |
| "loss": 0.7141, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.5497630331753555, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 0.00019175836260593938, | |
| "loss": 0.6713, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.5687203791469194, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 0.0001908817637339503, | |
| "loss": 0.6752, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5687203791469194, | |
| "eval_loss": 0.6603197455406189, | |
| "eval_runtime": 7.3719, | |
| "eval_samples_per_second": 48.291, | |
| "eval_steps_per_second": 12.073, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.5876777251184834, | |
| "grad_norm": 0.326171875, | |
| "learning_rate": 0.00018996308696522433, | |
| "loss": 0.6389, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.6066350710900474, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 0.00018900275764346768, | |
| "loss": 0.6711, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.6255924170616114, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.00018800122039735358, | |
| "loss": 0.6097, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.6445497630331753, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 0.0001869589389346611, | |
| "loss": 0.6204, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.6635071090047393, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 0.00018587639582758031, | |
| "loss": 0.6288, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.6824644549763034, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00018475409228928312, | |
| "loss": 0.5908, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.6824644549763034, | |
| "eval_loss": 0.6117041707038879, | |
| "eval_runtime": 7.5585, | |
| "eval_samples_per_second": 47.099, | |
| "eval_steps_per_second": 11.775, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.7014218009478673, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 0.0001835925479418637, | |
| "loss": 0.5851, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.7203791469194313, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00018239230057575542, | |
| "loss": 0.5967, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.7393364928909952, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.0001811539059007361, | |
| "loss": 0.5626, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.7582938388625592, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00017987793728863651, | |
| "loss": 0.5691, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7772511848341233, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00017856498550787144, | |
| "loss": 0.5913, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.7962085308056872, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00017721565844991643, | |
| "loss": 0.5229, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.7962085308056872, | |
| "eval_loss": 0.5701669454574585, | |
| "eval_runtime": 7.5799, | |
| "eval_samples_per_second": 46.966, | |
| "eval_steps_per_second": 11.742, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.8151658767772512, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00017583058084785625, | |
| "loss": 0.5542, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.8341232227488151, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.00017441039398713608, | |
| "loss": 0.5131, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.8530805687203792, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00017295575540864877, | |
| "loss": 0.5479, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.8720379146919431, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00017146733860429612, | |
| "loss": 0.5456, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.8909952606635071, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.0001699458327051647, | |
| "loss": 0.5475, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.909952606635071, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00016839194216246108, | |
| "loss": 0.558, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.909952606635071, | |
| "eval_loss": 0.528121829032898, | |
| "eval_runtime": 7.5358, | |
| "eval_samples_per_second": 47.241, | |
| "eval_steps_per_second": 11.81, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.9289099526066351, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00016680638642135336, | |
| "loss": 0.4805, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.9478672985781991, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 0.00016518989958787126, | |
| "loss": 0.4929, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.966824644549763, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00016354323008901776, | |
| "loss": 0.5415, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.985781990521327, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00016186714032625035, | |
| "loss": 0.5038, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.004739336492891, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.00016016240632249224, | |
| "loss": 0.4915, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.0236966824644549, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00015842981736283686, | |
| "loss": 0.4343, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.0236966824644549, | |
| "eval_loss": 0.4752185344696045, | |
| "eval_runtime": 7.8454, | |
| "eval_samples_per_second": 45.377, | |
| "eval_steps_per_second": 11.344, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.042654028436019, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00015667017562911176, | |
| "loss": 0.4282, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.061611374407583, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00015488429582847192, | |
| "loss": 0.396, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.080568720379147, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00015307300481619333, | |
| "loss": 0.4685, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.099526066350711, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.0001512371412128424, | |
| "loss": 0.428, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.1184834123222749, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00014937755501599772, | |
| "loss": 0.4002, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.1374407582938388, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00014749510720670506, | |
| "loss": 0.4039, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.1374407582938388, | |
| "eval_loss": 0.4151807725429535, | |
| "eval_runtime": 7.7375, | |
| "eval_samples_per_second": 46.01, | |
| "eval_steps_per_second": 11.502, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.1563981042654028, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.00014559066935084588, | |
| "loss": 0.3903, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.1753554502369667, | |
| "grad_norm": 0.2890625, | |
| "learning_rate": 0.0001436651231956064, | |
| "loss": 0.3936, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.1943127962085307, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 0.00014171936026123168, | |
| "loss": 0.3759, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.2132701421800949, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.0001397542814282556, | |
| "loss": 0.3949, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.2322274881516588, | |
| "grad_norm": 2.8125, | |
| "learning_rate": 0.0001377707965203965, | |
| "loss": 0.3979, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.2511848341232228, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.0001357698238833126, | |
| "loss": 0.3744, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.2511848341232228, | |
| "eval_loss": 0.422529935836792, | |
| "eval_runtime": 7.4489, | |
| "eval_samples_per_second": 47.792, | |
| "eval_steps_per_second": 11.948, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.2701421800947867, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 0.00013375228995941133, | |
| "loss": 0.3851, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.2890995260663507, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00013171912885891063, | |
| "loss": 0.3524, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.3080568720379147, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00012967128192734902, | |
| "loss": 0.3688, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.3270142180094786, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 0.00012760969730974694, | |
| "loss": 0.3338, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.3459715639810428, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.0001255353295116187, | |
| "loss": 0.3206, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.3649289099526065, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 0.00012344913895704097, | |
| "loss": 0.3313, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.3649289099526065, | |
| "eval_loss": 0.3852485120296478, | |
| "eval_runtime": 7.69, | |
| "eval_samples_per_second": 46.294, | |
| "eval_steps_per_second": 11.573, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.3838862559241707, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 0.00012135209154397962, | |
| "loss": 0.3514, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.4028436018957346, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 0.000119245158197083, | |
| "loss": 0.3608, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.4218009478672986, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.00011712931441814776, | |
| "loss": 0.3361, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.4407582938388626, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 0.00011500553983446527, | |
| "loss": 0.3095, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.4597156398104265, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 0.0001128748177452581, | |
| "loss": 0.318, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.4786729857819905, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 0.00011073813466641632, | |
| "loss": 0.374, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.4786729857819905, | |
| "eval_loss": 0.3739808201789856, | |
| "eval_runtime": 7.5558, | |
| "eval_samples_per_second": 47.116, | |
| "eval_steps_per_second": 11.779, | |
| "step": 78 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 156, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 26, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.1515674458062848e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |