| { |
| "best_global_step": 1800, |
| "best_metric": 0.8691588785046729, |
| "best_model_checkpoint": "haryos_model_loras_large/xlm-roberta-large_massive_lora_ru-RU/checkpoint-1800", |
| "epoch": 5.0, |
| "eval_steps": 500, |
| "global_step": 1800, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.1388888888888889, |
| "grad_norm": 3.6620867252349854, |
| "learning_rate": 0.00019455555555555556, |
| "loss": 3.822177734375, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.2777777777777778, |
| "grad_norm": 7.815118789672852, |
| "learning_rate": 0.00018899999999999999, |
| "loss": 3.00836181640625, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.4166666666666667, |
| "grad_norm": 7.057448387145996, |
| "learning_rate": 0.00018344444444444446, |
| "loss": 1.7329855346679688, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.5555555555555556, |
| "grad_norm": 5.466603755950928, |
| "learning_rate": 0.00017788888888888892, |
| "loss": 1.2191082000732423, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.6944444444444444, |
| "grad_norm": 4.801546573638916, |
| "learning_rate": 0.00017233333333333334, |
| "loss": 1.0881301879882812, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.8333333333333334, |
| "grad_norm": 4.551077842712402, |
| "learning_rate": 0.0001667777777777778, |
| "loss": 0.9190669250488281, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.9722222222222222, |
| "grad_norm": 13.638528823852539, |
| "learning_rate": 0.00016122222222222224, |
| "loss": 0.755425033569336, |
| "step": 350 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.8052139695031972, |
| "eval_f1": 0.7951606942706597, |
| "eval_loss": 0.6754380464553833, |
| "eval_runtime": 1.7908, |
| "eval_samples_per_second": 1135.245, |
| "eval_steps_per_second": 17.869, |
| "step": 360 |
| }, |
| { |
| "epoch": 1.1111111111111112, |
| "grad_norm": 6.151719570159912, |
| "learning_rate": 0.00015566666666666666, |
| "loss": 0.7096215057373046, |
| "step": 400 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 5.832527160644531, |
| "learning_rate": 0.00015011111111111112, |
| "loss": 0.6258441543579102, |
| "step": 450 |
| }, |
| { |
| "epoch": 1.3888888888888888, |
| "grad_norm": 3.742814540863037, |
| "learning_rate": 0.00014455555555555557, |
| "loss": 0.663236312866211, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.5277777777777777, |
| "grad_norm": 4.150186538696289, |
| "learning_rate": 0.000139, |
| "loss": 0.6828049468994141, |
| "step": 550 |
| }, |
| { |
| "epoch": 1.6666666666666665, |
| "grad_norm": 3.1718549728393555, |
| "learning_rate": 0.00013344444444444447, |
| "loss": 0.6133577728271484, |
| "step": 600 |
| }, |
| { |
| "epoch": 1.8055555555555556, |
| "grad_norm": 4.844176292419434, |
| "learning_rate": 0.0001278888888888889, |
| "loss": 0.6224682998657226, |
| "step": 650 |
| }, |
| { |
| "epoch": 1.9444444444444444, |
| "grad_norm": 4.165023326873779, |
| "learning_rate": 0.00012233333333333334, |
| "loss": 0.5294166564941406, |
| "step": 700 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.8504672897196262, |
| "eval_f1": 0.8474851881678009, |
| "eval_loss": 0.5576382279396057, |
| "eval_runtime": 1.786, |
| "eval_samples_per_second": 1138.294, |
| "eval_steps_per_second": 17.917, |
| "step": 720 |
| }, |
| { |
| "epoch": 2.0833333333333335, |
| "grad_norm": 3.8788259029388428, |
| "learning_rate": 0.00011677777777777778, |
| "loss": 0.633831787109375, |
| "step": 750 |
| }, |
| { |
| "epoch": 2.2222222222222223, |
| "grad_norm": 6.07531213760376, |
| "learning_rate": 0.00011122222222222223, |
| "loss": 0.47904563903808595, |
| "step": 800 |
| }, |
| { |
| "epoch": 2.361111111111111, |
| "grad_norm": 2.9574525356292725, |
| "learning_rate": 0.00010566666666666667, |
| "loss": 0.534382095336914, |
| "step": 850 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 3.9753854274749756, |
| "learning_rate": 0.0001001111111111111, |
| "loss": 0.5087989044189453, |
| "step": 900 |
| }, |
| { |
| "epoch": 2.638888888888889, |
| "grad_norm": 5.035774230957031, |
| "learning_rate": 9.455555555555556e-05, |
| "loss": 0.4885675048828125, |
| "step": 950 |
| }, |
| { |
| "epoch": 2.7777777777777777, |
| "grad_norm": 6.545742034912109, |
| "learning_rate": 8.900000000000001e-05, |
| "loss": 0.5159788131713867, |
| "step": 1000 |
| }, |
| { |
| "epoch": 2.9166666666666665, |
| "grad_norm": 4.568363666534424, |
| "learning_rate": 8.344444444444445e-05, |
| "loss": 0.4497852325439453, |
| "step": 1050 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.8627643876045253, |
| "eval_f1": 0.8603424259807203, |
| "eval_loss": 0.5127748847007751, |
| "eval_runtime": 1.7832, |
| "eval_samples_per_second": 1140.056, |
| "eval_steps_per_second": 17.945, |
| "step": 1080 |
| }, |
| { |
| "epoch": 3.0555555555555554, |
| "grad_norm": 2.7474329471588135, |
| "learning_rate": 7.788888888888888e-05, |
| "loss": 0.46715248107910157, |
| "step": 1100 |
| }, |
| { |
| "epoch": 3.1944444444444446, |
| "grad_norm": 4.932474613189697, |
| "learning_rate": 7.233333333333335e-05, |
| "loss": 0.42240074157714846, |
| "step": 1150 |
| }, |
| { |
| "epoch": 3.3333333333333335, |
| "grad_norm": 3.918285846710205, |
| "learning_rate": 6.677777777777779e-05, |
| "loss": 0.4302996826171875, |
| "step": 1200 |
| }, |
| { |
| "epoch": 3.4722222222222223, |
| "grad_norm": 2.998631715774536, |
| "learning_rate": 6.122222222222222e-05, |
| "loss": 0.4148902893066406, |
| "step": 1250 |
| }, |
| { |
| "epoch": 3.611111111111111, |
| "grad_norm": 5.161650657653809, |
| "learning_rate": 5.566666666666667e-05, |
| "loss": 0.428927001953125, |
| "step": 1300 |
| }, |
| { |
| "epoch": 3.75, |
| "grad_norm": 3.606273651123047, |
| "learning_rate": 5.011111111111111e-05, |
| "loss": 0.4201693344116211, |
| "step": 1350 |
| }, |
| { |
| "epoch": 3.888888888888889, |
| "grad_norm": 4.365043640136719, |
| "learning_rate": 4.4555555555555555e-05, |
| "loss": 0.4280767059326172, |
| "step": 1400 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.8657156910969012, |
| "eval_f1": 0.8636310959273443, |
| "eval_loss": 0.4980275332927704, |
| "eval_runtime": 1.7858, |
| "eval_samples_per_second": 1138.44, |
| "eval_steps_per_second": 17.919, |
| "step": 1440 |
| }, |
| { |
| "epoch": 4.027777777777778, |
| "grad_norm": 3.231497049331665, |
| "learning_rate": 3.9000000000000006e-05, |
| "loss": 0.39546192169189454, |
| "step": 1450 |
| }, |
| { |
| "epoch": 4.166666666666667, |
| "grad_norm": 6.7287702560424805, |
| "learning_rate": 3.3444444444444443e-05, |
| "loss": 0.372986946105957, |
| "step": 1500 |
| }, |
| { |
| "epoch": 4.305555555555555, |
| "grad_norm": 2.7220969200134277, |
| "learning_rate": 2.788888888888889e-05, |
| "loss": 0.34643054962158204, |
| "step": 1550 |
| }, |
| { |
| "epoch": 4.444444444444445, |
| "grad_norm": 1.9666249752044678, |
| "learning_rate": 2.2333333333333335e-05, |
| "loss": 0.3881946563720703, |
| "step": 1600 |
| }, |
| { |
| "epoch": 4.583333333333333, |
| "grad_norm": 4.047429084777832, |
| "learning_rate": 1.677777777777778e-05, |
| "loss": 0.37857887268066404, |
| "step": 1650 |
| }, |
| { |
| "epoch": 4.722222222222222, |
| "grad_norm": 4.3982744216918945, |
| "learning_rate": 1.1222222222222224e-05, |
| "loss": 0.3813508605957031, |
| "step": 1700 |
| }, |
| { |
| "epoch": 4.861111111111111, |
| "grad_norm": 2.6614363193511963, |
| "learning_rate": 5.666666666666667e-06, |
| "loss": 0.37185771942138673, |
| "step": 1750 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 2.6081504821777344, |
| "learning_rate": 1.1111111111111112e-07, |
| "loss": 0.3751992416381836, |
| "step": 1800 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 0.8691588785046729, |
| "eval_f1": 0.8679303391344835, |
| "eval_loss": 0.4859911799430847, |
| "eval_runtime": 1.7905, |
| "eval_samples_per_second": 1135.431, |
| "eval_steps_per_second": 17.872, |
| "step": 1800 |
| }, |
| { |
| "epoch": 5.0, |
| "step": 1800, |
| "total_flos": 1.353411879487488e+16, |
| "train_loss": 0.7395658906300863, |
| "train_runtime": 140.5322, |
| "train_samples_per_second": 409.657, |
| "train_steps_per_second": 12.808 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 1800, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "EarlyStoppingCallback": { |
| "args": { |
| "early_stopping_patience": 3, |
| "early_stopping_threshold": 0.0 |
| }, |
| "attributes": { |
| "early_stopping_patience_counter": 0 |
| } |
| }, |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.353411879487488e+16, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
