| { |
| "best_global_step": 904, |
| "best_metric": 2.465582847595215, |
| "best_model_checkpoint": "./mt5_ossetian_translator/checkpoint-904", |
| "epoch": 8.0, |
| "eval_steps": 500, |
| "global_step": 904, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.17699115044247787, |
| "grad_norm": 328.7373352050781, |
| "learning_rate": 0.0004915929203539823, |
| "loss": 18.003941345214844, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.35398230088495575, |
| "grad_norm": 16.042678833007812, |
| "learning_rate": 0.0004827433628318584, |
| "loss": 8.6413818359375, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.5309734513274337, |
| "grad_norm": 7.103315353393555, |
| "learning_rate": 0.0004738938053097345, |
| "loss": 6.339987182617188, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.7079646017699115, |
| "grad_norm": 4.68532133102417, |
| "learning_rate": 0.0004650442477876106, |
| "loss": 5.369053649902344, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.8849557522123894, |
| "grad_norm": 3.539532423019409, |
| "learning_rate": 0.0004561946902654867, |
| "loss": 4.760232543945312, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_loss": 3.66064453125, |
| "eval_runtime": 0.9181, |
| "eval_samples_per_second": 110.006, |
| "eval_steps_per_second": 14.159, |
| "step": 113 |
| }, |
| { |
| "epoch": 1.0619469026548674, |
| "grad_norm": 3.4077095985412598, |
| "learning_rate": 0.0004473451327433629, |
| "loss": 4.4729164123535154, |
| "step": 120 |
| }, |
| { |
| "epoch": 1.238938053097345, |
| "grad_norm": 2.3176043033599854, |
| "learning_rate": 0.000438495575221239, |
| "loss": 4.149492263793945, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.415929203539823, |
| "grad_norm": 2.055542230606079, |
| "learning_rate": 0.00042964601769911507, |
| "loss": 3.9551513671875, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.592920353982301, |
| "grad_norm": 3.381129026412964, |
| "learning_rate": 0.00042079646017699116, |
| "loss": 3.8981708526611327, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.7699115044247788, |
| "grad_norm": 2.6095192432403564, |
| "learning_rate": 0.0004119469026548673, |
| "loss": 3.908449172973633, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.9469026548672566, |
| "grad_norm": 2.930842399597168, |
| "learning_rate": 0.0004030973451327434, |
| "loss": 3.665481185913086, |
| "step": 220 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_loss": 3.1441268920898438, |
| "eval_runtime": 0.8098, |
| "eval_samples_per_second": 124.721, |
| "eval_steps_per_second": 16.053, |
| "step": 226 |
| }, |
| { |
| "epoch": 2.1238938053097347, |
| "grad_norm": 2.2758448123931885, |
| "learning_rate": 0.0003942477876106195, |
| "loss": 3.4207077026367188, |
| "step": 240 |
| }, |
| { |
| "epoch": 2.3008849557522124, |
| "grad_norm": 2.5473992824554443, |
| "learning_rate": 0.0003853982300884956, |
| "loss": 3.3584686279296876, |
| "step": 260 |
| }, |
| { |
| "epoch": 2.47787610619469, |
| "grad_norm": 1.703884243965149, |
| "learning_rate": 0.0003765486725663717, |
| "loss": 3.4501358032226563, |
| "step": 280 |
| }, |
| { |
| "epoch": 2.6548672566371683, |
| "grad_norm": 2.2273247241973877, |
| "learning_rate": 0.0003676991150442478, |
| "loss": 3.2697242736816405, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.831858407079646, |
| "grad_norm": 2.3433310985565186, |
| "learning_rate": 0.0003588495575221239, |
| "loss": 3.2843551635742188, |
| "step": 320 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_loss": 2.8717918395996094, |
| "eval_runtime": 0.8009, |
| "eval_samples_per_second": 126.116, |
| "eval_steps_per_second": 16.233, |
| "step": 339 |
| }, |
| { |
| "epoch": 3.0088495575221237, |
| "grad_norm": 2.4545156955718994, |
| "learning_rate": 0.00035, |
| "loss": 3.2385746002197267, |
| "step": 340 |
| }, |
| { |
| "epoch": 3.185840707964602, |
| "grad_norm": 2.047863483428955, |
| "learning_rate": 0.00034115044247787614, |
| "loss": 3.073255157470703, |
| "step": 360 |
| }, |
| { |
| "epoch": 3.3628318584070795, |
| "grad_norm": 3.017425537109375, |
| "learning_rate": 0.00033230088495575223, |
| "loss": 2.9876617431640624, |
| "step": 380 |
| }, |
| { |
| "epoch": 3.5398230088495577, |
| "grad_norm": 2.587419271469116, |
| "learning_rate": 0.0003234513274336283, |
| "loss": 2.9378360748291015, |
| "step": 400 |
| }, |
| { |
| "epoch": 3.7168141592920354, |
| "grad_norm": 2.2757842540740967, |
| "learning_rate": 0.0003146017699115044, |
| "loss": 2.997949409484863, |
| "step": 420 |
| }, |
| { |
| "epoch": 3.893805309734513, |
| "grad_norm": 2.942039728164673, |
| "learning_rate": 0.00030575221238938056, |
| "loss": 2.910443115234375, |
| "step": 440 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_loss": 2.6989588737487793, |
| "eval_runtime": 0.8141, |
| "eval_samples_per_second": 124.061, |
| "eval_steps_per_second": 15.968, |
| "step": 452 |
| }, |
| { |
| "epoch": 4.070796460176991, |
| "grad_norm": 2.663364887237549, |
| "learning_rate": 0.00029690265486725665, |
| "loss": 2.8674571990966795, |
| "step": 460 |
| }, |
| { |
| "epoch": 4.247787610619469, |
| "grad_norm": 2.9044437408447266, |
| "learning_rate": 0.00028805309734513274, |
| "loss": 2.7972084045410157, |
| "step": 480 |
| }, |
| { |
| "epoch": 4.424778761061947, |
| "grad_norm": 2.2075090408325195, |
| "learning_rate": 0.00027920353982300883, |
| "loss": 2.7429454803466795, |
| "step": 500 |
| }, |
| { |
| "epoch": 4.601769911504425, |
| "grad_norm": 4.215453147888184, |
| "learning_rate": 0.000270353982300885, |
| "loss": 2.767496109008789, |
| "step": 520 |
| }, |
| { |
| "epoch": 4.778761061946903, |
| "grad_norm": 2.807522773742676, |
| "learning_rate": 0.00026150442477876107, |
| "loss": 2.7617130279541016, |
| "step": 540 |
| }, |
| { |
| "epoch": 4.95575221238938, |
| "grad_norm": 2.4849319458007812, |
| "learning_rate": 0.00025265486725663716, |
| "loss": 2.7167985916137694, |
| "step": 560 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_loss": 2.6085963249206543, |
| "eval_runtime": 0.8712, |
| "eval_samples_per_second": 115.935, |
| "eval_steps_per_second": 14.922, |
| "step": 565 |
| }, |
| { |
| "epoch": 5.132743362831858, |
| "grad_norm": 2.3556389808654785, |
| "learning_rate": 0.00024380530973451328, |
| "loss": 2.6242326736450194, |
| "step": 580 |
| }, |
| { |
| "epoch": 5.3097345132743365, |
| "grad_norm": 3.390760898590088, |
| "learning_rate": 0.00023495575221238937, |
| "loss": 2.6653488159179686, |
| "step": 600 |
| }, |
| { |
| "epoch": 5.486725663716814, |
| "grad_norm": 2.506216526031494, |
| "learning_rate": 0.0002261061946902655, |
| "loss": 2.572140121459961, |
| "step": 620 |
| }, |
| { |
| "epoch": 5.663716814159292, |
| "grad_norm": 2.8693790435791016, |
| "learning_rate": 0.00021725663716814158, |
| "loss": 2.493195343017578, |
| "step": 640 |
| }, |
| { |
| "epoch": 5.84070796460177, |
| "grad_norm": 2.4581332206726074, |
| "learning_rate": 0.0002084070796460177, |
| "loss": 2.567514991760254, |
| "step": 660 |
| }, |
| { |
| "epoch": 6.0, |
| "eval_loss": 2.5541763305664062, |
| "eval_runtime": 0.8956, |
| "eval_samples_per_second": 112.775, |
| "eval_steps_per_second": 14.516, |
| "step": 678 |
| }, |
| { |
| "epoch": 6.017699115044247, |
| "grad_norm": 3.3094470500946045, |
| "learning_rate": 0.0001995575221238938, |
| "loss": 2.548930358886719, |
| "step": 680 |
| }, |
| { |
| "epoch": 6.1946902654867255, |
| "grad_norm": 2.6823666095733643, |
| "learning_rate": 0.00019070796460176994, |
| "loss": 2.4545686721801756, |
| "step": 700 |
| }, |
| { |
| "epoch": 6.371681415929204, |
| "grad_norm": 2.289599895477295, |
| "learning_rate": 0.00018185840707964603, |
| "loss": 2.381149673461914, |
| "step": 720 |
| }, |
| { |
| "epoch": 6.548672566371682, |
| "grad_norm": 2.9119653701782227, |
| "learning_rate": 0.00017300884955752214, |
| "loss": 2.358357620239258, |
| "step": 740 |
| }, |
| { |
| "epoch": 6.725663716814159, |
| "grad_norm": 2.473071575164795, |
| "learning_rate": 0.00016415929203539824, |
| "loss": 2.446194648742676, |
| "step": 760 |
| }, |
| { |
| "epoch": 6.902654867256637, |
| "grad_norm": 1.9398255348205566, |
| "learning_rate": 0.00015530973451327435, |
| "loss": 2.3346189498901366, |
| "step": 780 |
| }, |
| { |
| "epoch": 7.0, |
| "eval_loss": 2.507996082305908, |
| "eval_runtime": 0.9073, |
| "eval_samples_per_second": 111.319, |
| "eval_steps_per_second": 14.328, |
| "step": 791 |
| }, |
| { |
| "epoch": 7.079646017699115, |
| "grad_norm": 2.9123404026031494, |
| "learning_rate": 0.00014646017699115044, |
| "loss": 2.482748794555664, |
| "step": 800 |
| }, |
| { |
| "epoch": 7.256637168141593, |
| "grad_norm": 2.646937131881714, |
| "learning_rate": 0.00013761061946902656, |
| "loss": 2.3216293334960936, |
| "step": 820 |
| }, |
| { |
| "epoch": 7.433628318584071, |
| "grad_norm": 4.007692337036133, |
| "learning_rate": 0.00012876106194690265, |
| "loss": 2.234068489074707, |
| "step": 840 |
| }, |
| { |
| "epoch": 7.610619469026549, |
| "grad_norm": 2.953052043914795, |
| "learning_rate": 0.00011991150442477876, |
| "loss": 2.2527175903320313, |
| "step": 860 |
| }, |
| { |
| "epoch": 7.787610619469026, |
| "grad_norm": 2.417543649673462, |
| "learning_rate": 0.00011106194690265486, |
| "loss": 2.412649154663086, |
| "step": 880 |
| }, |
| { |
| "epoch": 7.964601769911504, |
| "grad_norm": 2.309840440750122, |
| "learning_rate": 0.00010221238938053097, |
| "loss": 2.326849365234375, |
| "step": 900 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_loss": 2.465582847595215, |
| "eval_runtime": 1.3915, |
| "eval_samples_per_second": 72.581, |
| "eval_steps_per_second": 9.342, |
| "step": 904 |
| } |
| ], |
| "logging_steps": 20, |
| "max_steps": 1130, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 10, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 332418321285120.0, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
