| { | |
| "best_metric": 0.5962514877319336, | |
| "best_model_checkpoint": "./ttm_finetuned_models/output/checkpoint-60", | |
| "epoch": 30.0, | |
| "eval_steps": 500, | |
| "global_step": 60, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7443529367446899, | |
| "learning_rate": 4.732772046676698e-05, | |
| "loss": 0.8431, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6096403002738953, | |
| "eval_runtime": 0.3627, | |
| "eval_samples_per_second": 192.988, | |
| "eval_steps_per_second": 2.757, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8946840763092041, | |
| "learning_rate": 4.809321690693591e-05, | |
| "loss": 0.8393, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.60591721534729, | |
| "eval_runtime": 0.36, | |
| "eval_samples_per_second": 194.421, | |
| "eval_steps_per_second": 2.777, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.7669185996055603, | |
| "learning_rate": 4.936827562256963e-05, | |
| "loss": 0.8184, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.6019852757453918, | |
| "eval_runtime": 0.3543, | |
| "eval_samples_per_second": 197.593, | |
| "eval_steps_per_second": 2.823, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.7224565744400024, | |
| "learning_rate": 5.115174431611605e-05, | |
| "loss": 0.8012, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.6119012832641602, | |
| "eval_runtime": 0.3476, | |
| "eval_samples_per_second": 201.402, | |
| "eval_steps_per_second": 2.877, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5908293128013611, | |
| "learning_rate": 5.344201122916398e-05, | |
| "loss": 0.7874, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.6097100973129272, | |
| "eval_runtime": 0.3479, | |
| "eval_samples_per_second": 201.225, | |
| "eval_steps_per_second": 2.875, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6504310369491577, | |
| "learning_rate": 5.623700659902384e-05, | |
| "loss": 0.8027, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.6008864045143127, | |
| "eval_runtime": 0.3626, | |
| "eval_samples_per_second": 193.042, | |
| "eval_steps_per_second": 2.758, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.4692302346229553, | |
| "learning_rate": 5.9534204529214504e-05, | |
| "loss": 0.7916, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.6066852807998657, | |
| "eval_runtime": 0.3625, | |
| "eval_samples_per_second": 193.12, | |
| "eval_steps_per_second": 2.759, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.273665189743042, | |
| "learning_rate": 6.333062527217074e-05, | |
| "loss": 0.79, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.6052869558334351, | |
| "eval_runtime": 0.3656, | |
| "eval_samples_per_second": 191.441, | |
| "eval_steps_per_second": 2.735, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.6313425302505493, | |
| "learning_rate": 6.762283792210435e-05, | |
| "loss": 0.7741, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.6067262887954712, | |
| "eval_runtime": 0.3465, | |
| "eval_samples_per_second": 201.993, | |
| "eval_steps_per_second": 2.886, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.0959845781326294, | |
| "learning_rate": 7.240696351558515e-05, | |
| "loss": 0.7734, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.6191284656524658, | |
| "eval_runtime": 0.7504, | |
| "eval_samples_per_second": 93.288, | |
| "eval_steps_per_second": 1.333, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.440314382314682, | |
| "learning_rate": 7.767867853704304e-05, | |
| "loss": 0.7732, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.6069210171699524, | |
| "eval_runtime": 0.3562, | |
| "eval_samples_per_second": 196.544, | |
| "eval_steps_per_second": 2.808, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.5499612092971802, | |
| "learning_rate": 8.343321882602062e-05, | |
| "loss": 0.7621, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.6009647250175476, | |
| "eval_runtime": 0.3571, | |
| "eval_samples_per_second": 196.025, | |
| "eval_steps_per_second": 2.8, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.5899090766906738, | |
| "learning_rate": 8.966538388264578e-05, | |
| "loss": 0.7535, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 0.6042256355285645, | |
| "eval_runtime": 0.3477, | |
| "eval_samples_per_second": 201.309, | |
| "eval_steps_per_second": 2.876, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.5770731568336487, | |
| "learning_rate": 9.636954156743349e-05, | |
| "loss": 0.7807, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.6018546223640442, | |
| "eval_runtime": 0.355, | |
| "eval_samples_per_second": 197.191, | |
| "eval_steps_per_second": 2.817, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.423060804605484, | |
| "learning_rate": 0.00010353963319116892, | |
| "loss": 0.7376, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 0.6036869287490845, | |
| "eval_runtime": 0.3443, | |
| "eval_samples_per_second": 203.313, | |
| "eval_steps_per_second": 2.904, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.444051593542099, | |
| "learning_rate": 0.00011116917899027415, | |
| "loss": 0.7431, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.5995596051216125, | |
| "eval_runtime": 0.4433, | |
| "eval_samples_per_second": 157.902, | |
| "eval_steps_per_second": 2.256, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.7095910906791687, | |
| "learning_rate": 0.00011925128398270726, | |
| "loss": 0.7551, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 0.6076721549034119, | |
| "eval_runtime": 0.3753, | |
| "eval_samples_per_second": 186.526, | |
| "eval_steps_per_second": 2.665, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.938398540019989, | |
| "learning_rate": 0.0001277786441991034, | |
| "loss": 0.741, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 0.6111529469490051, | |
| "eval_runtime": 0.3616, | |
| "eval_samples_per_second": 193.562, | |
| "eval_steps_per_second": 2.765, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.45169830322265625, | |
| "learning_rate": 0.00013674355328352613, | |
| "loss": 0.7409, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 0.5968735814094543, | |
| "eval_runtime": 0.3676, | |
| "eval_samples_per_second": 190.435, | |
| "eval_steps_per_second": 2.72, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.3840300440788269, | |
| "learning_rate": 0.0001461379094578651, | |
| "loss": 0.7308, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 0.6049572229385376, | |
| "eval_runtime": 0.7422, | |
| "eval_samples_per_second": 94.316, | |
| "eval_steps_per_second": 1.347, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 1.2706289291381836, | |
| "learning_rate": 0.00015595322284358464, | |
| "loss": 0.7472, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "eval_loss": 0.6062923669815063, | |
| "eval_runtime": 0.3735, | |
| "eval_samples_per_second": 187.409, | |
| "eval_steps_per_second": 2.677, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.5400757193565369, | |
| "learning_rate": 0.00016618062313420584, | |
| "loss": 0.7378, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "eval_loss": 0.5963739156723022, | |
| "eval_runtime": 0.3635, | |
| "eval_samples_per_second": 192.591, | |
| "eval_steps_per_second": 2.751, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.42090097069740295, | |
| "learning_rate": 0.00017681086761159098, | |
| "loss": 0.7316, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "eval_loss": 0.6015997529029846, | |
| "eval_runtime": 0.3627, | |
| "eval_samples_per_second": 192.989, | |
| "eval_steps_per_second": 2.757, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.47120270133018494, | |
| "learning_rate": 0.00018783434949878235, | |
| "loss": 0.7405, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "eval_loss": 0.5992478728294373, | |
| "eval_runtime": 0.3755, | |
| "eval_samples_per_second": 186.405, | |
| "eval_steps_per_second": 2.663, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.3660564422607422, | |
| "learning_rate": 0.00019924110664185264, | |
| "loss": 0.713, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_loss": 0.5973361134529114, | |
| "eval_runtime": 0.371, | |
| "eval_samples_per_second": 188.686, | |
| "eval_steps_per_second": 2.696, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.4799431562423706, | |
| "learning_rate": 0.00021102083051291085, | |
| "loss": 0.7258, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "eval_loss": 0.5997373461723328, | |
| "eval_runtime": 0.3657, | |
| "eval_samples_per_second": 191.402, | |
| "eval_steps_per_second": 2.734, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.30694785714149475, | |
| "learning_rate": 0.0002231628755261397, | |
| "loss": 0.7112, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "eval_loss": 0.5968739986419678, | |
| "eval_runtime": 0.3648, | |
| "eval_samples_per_second": 191.882, | |
| "eval_steps_per_second": 2.741, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.22820888459682465, | |
| "learning_rate": 0.0002356562686584362, | |
| "loss": 0.7121, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "eval_loss": 0.5964872241020203, | |
| "eval_runtime": 0.3458, | |
| "eval_samples_per_second": 202.446, | |
| "eval_steps_per_second": 2.892, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.55999755859375, | |
| "learning_rate": 0.0002484897193659642, | |
| "loss": 0.717, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "eval_loss": 0.597991943359375, | |
| "eval_runtime": 0.3373, | |
| "eval_samples_per_second": 207.519, | |
| "eval_steps_per_second": 2.965, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.25612083077430725, | |
| "learning_rate": 0.0002616516297876592, | |
| "loss": 0.7091, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_loss": 0.5962514877319336, | |
| "eval_runtime": 0.3376, | |
| "eval_samples_per_second": 207.368, | |
| "eval_steps_per_second": 2.962, | |
| "step": 60 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 10, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 15922917144000.0, | |
| "train_batch_size": 288, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |