| { | |
| "best_metric": 0.4301711618900299, | |
| "best_model_checkpoint": "data/Llama-31-8B_task-3_120-samples_config-2/checkpoint-49", | |
| "epoch": 16.0, | |
| "eval_steps": 500, | |
| "global_step": 88, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 3.1685640811920166, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.8174, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 2.785682201385498, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 2.5884, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 2.4684979915618896, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.4469, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 2.3539235591888428, | |
| "eval_runtime": 21.0038, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 2.485983371734619, | |
| "learning_rate": 2.4e-05, | |
| "loss": 2.2775, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 3.0354607105255127, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 2.253, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 2.2072792053222656, | |
| "learning_rate": 4e-05, | |
| "loss": 1.8346, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.4921633005142212, | |
| "eval_runtime": 21.0005, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 2.2485885620117188, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.5834, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 1.686313509941101, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.2001, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 1.5054450035095215, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.7576, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 0.7652478218078613, | |
| "eval_runtime": 20.9987, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.7856181263923645, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.53, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 1.1085258722305298, | |
| "learning_rate": 8e-05, | |
| "loss": 0.6365, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.0253838300704956, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.6409, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.5626702308654785, | |
| "eval_runtime": 21.0032, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 1.1476973295211792, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.4895, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.519312858581543, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 0.4304, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 0.5238025784492493, | |
| "eval_runtime": 21.0026, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 2.1630260944366455, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.4678, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.2825535237789154, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.3661, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.2835836708545685, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.3624, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.470536470413208, | |
| "eval_runtime": 20.9986, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 1.6048427820205688, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.3769, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.4334805905818939, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.4372, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.33416426181793213, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.3967, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 0.4451916515827179, | |
| "eval_runtime": 20.9951, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.2907244861125946, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.3551, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.22854046523571014, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.2277, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.2439097911119461, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.3293, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.4328186511993408, | |
| "eval_runtime": 20.9972, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.2829444110393524, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.3193, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.5520598292350769, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.2432, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 0.4301711618900299, | |
| "eval_runtime": 21.2227, | |
| "eval_samples_per_second": 1.131, | |
| "eval_steps_per_second": 1.131, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.37884312868118286, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.3439, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.4332423508167267, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.283, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.306794673204422, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.2102, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.43588265776634216, | |
| "eval_runtime": 21.2262, | |
| "eval_samples_per_second": 1.131, | |
| "eval_steps_per_second": 1.131, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.4354619085788727, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.184, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.3915146291255951, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.2344, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.35923001170158386, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.2004, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 0.4582943022251129, | |
| "eval_runtime": 21.2266, | |
| "eval_samples_per_second": 1.131, | |
| "eval_steps_per_second": 1.131, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.5079478621482849, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.149, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.5917227864265442, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.1587, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.5818120837211609, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.1634, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.47242483496665955, | |
| "eval_runtime": 20.9934, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.4436568319797516, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.0713, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 1.2484676837921143, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.1177, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 0.5530457496643066, | |
| "eval_runtime": 21.002, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.6471790671348572, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.0837, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.5037362575531006, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.0458, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 1.0372614860534668, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.0376, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.7361419796943665, | |
| "eval_runtime": 20.9955, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.181818181818182, | |
| "grad_norm": 0.38053688406944275, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.0366, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 14.545454545454545, | |
| "grad_norm": 0.38170501589775085, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.0225, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "grad_norm": 0.6007455587387085, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.0204, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "eval_loss": 0.7768494486808777, | |
| "eval_runtime": 21.0007, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 15.272727272727273, | |
| "grad_norm": 0.11449413001537323, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.01, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 15.636363636363637, | |
| "grad_norm": 0.4309659004211426, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.0103, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.4790915250778198, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 0.0118, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 0.8608195185661316, | |
| "eval_runtime": 21.0029, | |
| "eval_samples_per_second": 1.143, | |
| "eval_steps_per_second": 1.143, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "step": 88, | |
| "total_flos": 1.098992984302551e+17, | |
| "train_loss": 0.54680911203931, | |
| "train_runtime": 4057.6932, | |
| "train_samples_per_second": 1.084, | |
| "train_steps_per_second": 0.062 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "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": 1.098992984302551e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |