| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 76, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.013157894736842105, | |
| "grad_norm": 13.238144898761828, | |
| "learning_rate": 1.25e-06, | |
| "loss": 0.3431, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02631578947368421, | |
| "grad_norm": 13.138533030251612, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.3429, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.039473684210526314, | |
| "grad_norm": 11.211925298543088, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.2715, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 9.07840716756589, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2923, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06578947368421052, | |
| "grad_norm": 4.255454799591527, | |
| "learning_rate": 6.25e-06, | |
| "loss": 0.1766, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07894736842105263, | |
| "grad_norm": 4.899723911649479, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.1795, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.09210526315789473, | |
| "grad_norm": 21.444343227297015, | |
| "learning_rate": 8.750000000000001e-06, | |
| "loss": 0.2065, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 4.9912226694910995, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1955, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11842105263157894, | |
| "grad_norm": 6.015085827837624, | |
| "learning_rate": 9.994664874011864e-06, | |
| "loss": 0.2171, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.13157894736842105, | |
| "grad_norm": 6.232109397110641, | |
| "learning_rate": 9.978670881475173e-06, | |
| "loss": 0.1812, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14473684210526316, | |
| "grad_norm": 13.260525050150829, | |
| "learning_rate": 9.952052154376027e-06, | |
| "loss": 0.2299, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 2.8474419609274584, | |
| "learning_rate": 9.91486549841951e-06, | |
| "loss": 0.1729, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.17105263157894737, | |
| "grad_norm": 4.441367319332706, | |
| "learning_rate": 9.867190271803466e-06, | |
| "loss": 0.1394, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.18421052631578946, | |
| "grad_norm": 3.9791362793155245, | |
| "learning_rate": 9.809128215864096e-06, | |
| "loss": 0.1596, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.19736842105263158, | |
| "grad_norm": 5.398799668373648, | |
| "learning_rate": 9.74080323795483e-06, | |
| "loss": 0.1883, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 3.801676563771218, | |
| "learning_rate": 9.66236114702178e-06, | |
| "loss": 0.1542, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2236842105263158, | |
| "grad_norm": 1.925322887230293, | |
| "learning_rate": 9.573969342440107e-06, | |
| "loss": 0.1535, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23684210526315788, | |
| "grad_norm": 2.4885347468657666, | |
| "learning_rate": 9.475816456775313e-06, | |
| "loss": 0.1495, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.7097680648804299, | |
| "learning_rate": 9.368111953231849e-06, | |
| "loss": 0.0999, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 1.7070432708248722, | |
| "learning_rate": 9.251085678648072e-06, | |
| "loss": 0.1556, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.27631578947368424, | |
| "grad_norm": 2.2017885079132475, | |
| "learning_rate": 9.124987372991512e-06, | |
| "loss": 0.1323, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2894736842105263, | |
| "grad_norm": 2.2923056410651657, | |
| "learning_rate": 8.990086136401199e-06, | |
| "loss": 0.1442, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.3026315789473684, | |
| "grad_norm": 1.3443143761442609, | |
| "learning_rate": 8.846669854914395e-06, | |
| "loss": 0.1067, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 2.499856741628817, | |
| "learning_rate": 8.695044586103297e-06, | |
| "loss": 0.1428, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32894736842105265, | |
| "grad_norm": 1.6830200645247368, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.1201, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.34210526315789475, | |
| "grad_norm": 1.5831400874663497, | |
| "learning_rate": 8.368478218232787e-06, | |
| "loss": 0.114, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.35526315789473684, | |
| "grad_norm": 1.6535836650289428, | |
| "learning_rate": 8.194234028259806e-06, | |
| "loss": 0.1209, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 1.4967093603851553, | |
| "learning_rate": 8.013173181896283e-06, | |
| "loss": 0.0847, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3815789473684211, | |
| "grad_norm": 1.6642707923715867, | |
| "learning_rate": 7.82568207211296e-06, | |
| "loss": 0.1259, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.39473684210526316, | |
| "grad_norm": 1.1497880560263245, | |
| "learning_rate": 7.63216081438678e-06, | |
| "loss": 0.1029, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.40789473684210525, | |
| "grad_norm": 1.3303308469541215, | |
| "learning_rate": 7.4330223928342814e-06, | |
| "loss": 0.1004, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 2.427667338258928, | |
| "learning_rate": 7.2286917788826926e-06, | |
| "loss": 0.0991, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4342105263157895, | |
| "grad_norm": 1.4112715962724514, | |
| "learning_rate": 7.019605024359475e-06, | |
| "loss": 0.1117, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4473684210526316, | |
| "grad_norm": 1.0945264209924628, | |
| "learning_rate": 6.806208330935766e-06, | |
| "loss": 0.0974, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.4605263157894737, | |
| "grad_norm": 2.222867676262349, | |
| "learning_rate": 6.588957097909509e-06, | |
| "loss": 0.1114, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 1.445820032177189, | |
| "learning_rate": 6.368314950360416e-06, | |
| "loss": 0.1082, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4868421052631579, | |
| "grad_norm": 2.005804185500154, | |
| "learning_rate": 6.144752749750671e-06, | |
| "loss": 0.0932, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.4446629324877371, | |
| "learning_rate": 5.918747589082853e-06, | |
| "loss": 0.11, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.5131578947368421, | |
| "grad_norm": 1.2013902671499292, | |
| "learning_rate": 5.690781774759412e-06, | |
| "loss": 0.088, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 0.9671394427016465, | |
| "learning_rate": 5.46134179731651e-06, | |
| "loss": 0.0965, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5394736842105263, | |
| "grad_norm": 0.874352355591093, | |
| "learning_rate": 5.230917293228699e-06, | |
| "loss": 0.0827, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5526315789473685, | |
| "grad_norm": 1.0990734862760956, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1026, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5657894736842105, | |
| "grad_norm": 1.3629123541548118, | |
| "learning_rate": 4.7690827067713035e-06, | |
| "loss": 0.1379, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 1.1191667397706833, | |
| "learning_rate": 4.53865820268349e-06, | |
| "loss": 0.1006, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5921052631578947, | |
| "grad_norm": 1.1613132302632512, | |
| "learning_rate": 4.309218225240591e-06, | |
| "loss": 0.1196, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6052631578947368, | |
| "grad_norm": 1.4178964529072136, | |
| "learning_rate": 4.081252410917148e-06, | |
| "loss": 0.1107, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.618421052631579, | |
| "grad_norm": 0.7977784874888711, | |
| "learning_rate": 3.855247250249331e-06, | |
| "loss": 0.0797, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 1.0845045774450395, | |
| "learning_rate": 3.6316850496395863e-06, | |
| "loss": 0.1116, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6447368421052632, | |
| "grad_norm": 1.8478183045971928, | |
| "learning_rate": 3.4110429020904924e-06, | |
| "loss": 0.1146, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.6578947368421053, | |
| "grad_norm": 0.7367795913317746, | |
| "learning_rate": 3.1937916690642356e-06, | |
| "loss": 0.0685, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6710526315789473, | |
| "grad_norm": 1.1081708398994705, | |
| "learning_rate": 2.980394975640526e-06, | |
| "loss": 0.0934, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 6.210726159509134, | |
| "learning_rate": 2.771308221117309e-06, | |
| "loss": 0.1012, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6973684210526315, | |
| "grad_norm": 0.987185487363248, | |
| "learning_rate": 2.5669776071657194e-06, | |
| "loss": 0.1173, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.7105263157894737, | |
| "grad_norm": 1.0233786719046054, | |
| "learning_rate": 2.3678391856132203e-06, | |
| "loss": 0.1006, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.7236842105263158, | |
| "grad_norm": 0.8805942584459667, | |
| "learning_rate": 2.174317927887041e-06, | |
| "loss": 0.0773, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 1.4063628412136746, | |
| "learning_rate": 1.9868268181037186e-06, | |
| "loss": 0.1285, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.7609777001959152, | |
| "learning_rate": 1.8057659717401948e-06, | |
| "loss": 0.0827, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7631578947368421, | |
| "grad_norm": 0.699472385650843, | |
| "learning_rate": 1.6315217817672142e-06, | |
| "loss": 0.0728, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7763157894736842, | |
| "grad_norm": 0.8086483434606366, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.0814, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 0.7120681165430872, | |
| "learning_rate": 1.3049554138967052e-06, | |
| "loss": 0.0746, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8026315789473685, | |
| "grad_norm": 0.6530934488185232, | |
| "learning_rate": 1.1533301450856054e-06, | |
| "loss": 0.0759, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.8157894736842105, | |
| "grad_norm": 0.6010174144424382, | |
| "learning_rate": 1.0099138635988026e-06, | |
| "loss": 0.0851, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8289473684210527, | |
| "grad_norm": 0.7632251026621609, | |
| "learning_rate": 8.750126270084891e-07, | |
| "loss": 0.133, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 0.9433341471827504, | |
| "learning_rate": 7.489143213519301e-07, | |
| "loss": 0.0887, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.8552631578947368, | |
| "grad_norm": 0.6626834135219665, | |
| "learning_rate": 6.318880467681527e-07, | |
| "loss": 0.0612, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.868421052631579, | |
| "grad_norm": 0.6498581718557976, | |
| "learning_rate": 5.241835432246888e-07, | |
| "loss": 0.0807, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.881578947368421, | |
| "grad_norm": 0.7771761015493923, | |
| "learning_rate": 4.2603065755989493e-07, | |
| "loss": 0.1067, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 0.7594547340399601, | |
| "learning_rate": 3.3763885297822153e-07, | |
| "loss": 0.0803, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.9078947368421053, | |
| "grad_norm": 0.980352554631326, | |
| "learning_rate": 2.5919676204517073e-07, | |
| "loss": 0.0956, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.9210526315789473, | |
| "grad_norm": 0.8650676858446684, | |
| "learning_rate": 1.908717841359048e-07, | |
| "loss": 0.0798, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9342105263157895, | |
| "grad_norm": 0.6440476763274703, | |
| "learning_rate": 1.328097281965357e-07, | |
| "loss": 0.0724, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 0.989220896843026, | |
| "learning_rate": 8.513450158049109e-08, | |
| "loss": 0.1021, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9605263157894737, | |
| "grad_norm": 0.6855314313706689, | |
| "learning_rate": 4.794784562397459e-08, | |
| "loss": 0.0656, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9736842105263158, | |
| "grad_norm": 0.8621391032787515, | |
| "learning_rate": 2.1329118524827662e-08, | |
| "loss": 0.0734, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.9868421052631579, | |
| "grad_norm": 0.7312318211226774, | |
| "learning_rate": 5.3351259881379016e-09, | |
| "loss": 0.0711, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.8656756829991957, | |
| "learning_rate": 0.0, | |
| "loss": 0.1306, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.09319383651018143, | |
| "eval_runtime": 140.3661, | |
| "eval_samples_per_second": 36.369, | |
| "eval_steps_per_second": 1.14, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 76, | |
| "total_flos": 2.244031216091136e+16, | |
| "train_loss": 0.12603967764267796, | |
| "train_runtime": 1049.273, | |
| "train_samples_per_second": 9.243, | |
| "train_steps_per_second": 0.072 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 76, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.244031216091136e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |