| { | |
| "best_metric": 0.9358716011047363, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-3_120-samples_config-2_full_auto/checkpoint-60", | |
| "epoch": 18.0, | |
| "eval_steps": 500, | |
| "global_step": 99, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.4636595845222473, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.3988, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.4026557207107544, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.3426, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.4368878901004791, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.4057, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "eval_loss": 1.37501060962677, | |
| "eval_runtime": 7.687, | |
| "eval_samples_per_second": 3.122, | |
| "eval_steps_per_second": 3.122, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 0.4618467390537262, | |
| "learning_rate": 2.4e-05, | |
| "loss": 1.393, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 0.4304371774196625, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 1.3613, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 0.3771096467971802, | |
| "learning_rate": 4e-05, | |
| "loss": 1.3708, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.2888946533203125, | |
| "eval_runtime": 7.494, | |
| "eval_samples_per_second": 3.203, | |
| "eval_steps_per_second": 3.203, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 0.3073701858520508, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.2637, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.2576550543308258, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 1.2369, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.2462034523487091, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 1.2158, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "eval_loss": 1.194419264793396, | |
| "eval_runtime": 7.4989, | |
| "eval_samples_per_second": 3.2, | |
| "eval_steps_per_second": 3.2, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.24131014943122864, | |
| "learning_rate": 7.2e-05, | |
| "loss": 1.1875, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.2574351131916046, | |
| "learning_rate": 8e-05, | |
| "loss": 1.1281, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.24738362431526184, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 1.1179, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.1043965816497803, | |
| "eval_runtime": 7.4982, | |
| "eval_samples_per_second": 3.201, | |
| "eval_steps_per_second": 3.201, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.2551267445087433, | |
| "learning_rate": 9.6e-05, | |
| "loss": 1.0771, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.39611178636550903, | |
| "learning_rate": 9.999512620046522e-05, | |
| "loss": 1.0198, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "eval_loss": 1.0250672101974487, | |
| "eval_runtime": 7.4968, | |
| "eval_samples_per_second": 3.201, | |
| "eval_steps_per_second": 3.201, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.21375422179698944, | |
| "learning_rate": 9.995614150494293e-05, | |
| "loss": 0.9995, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.218486949801445, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.9668, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.17936424911022186, | |
| "learning_rate": 9.976136999909156e-05, | |
| "loss": 0.9467, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.9803627133369446, | |
| "eval_runtime": 7.4983, | |
| "eval_samples_per_second": 3.201, | |
| "eval_steps_per_second": 3.201, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.18685294687747955, | |
| "learning_rate": 9.96057350657239e-05, | |
| "loss": 0.9313, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.18190714716911316, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 0.9125, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.14362622797489166, | |
| "learning_rate": 9.917857354066931e-05, | |
| "loss": 0.9148, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "eval_loss": 0.9614717364311218, | |
| "eval_runtime": 7.5039, | |
| "eval_samples_per_second": 3.198, | |
| "eval_steps_per_second": 3.198, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 0.14666882157325745, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.9208, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.13798680901527405, | |
| "learning_rate": 9.859805002892732e-05, | |
| "loss": 0.8708, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.13614502549171448, | |
| "learning_rate": 9.825082472361557e-05, | |
| "loss": 0.8909, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.9480839371681213, | |
| "eval_runtime": 7.4992, | |
| "eval_samples_per_second": 3.2, | |
| "eval_steps_per_second": 3.2, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.13590388000011444, | |
| "learning_rate": 9.786597487660337e-05, | |
| "loss": 0.8601, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.16521644592285156, | |
| "learning_rate": 9.744380058222483e-05, | |
| "loss": 0.8645, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "eval_loss": 0.9406649470329285, | |
| "eval_runtime": 7.5056, | |
| "eval_samples_per_second": 3.198, | |
| "eval_steps_per_second": 3.198, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 0.1455041617155075, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.869, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 0.14802001416683197, | |
| "learning_rate": 9.648882429441257e-05, | |
| "loss": 0.8245, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 0.14117573201656342, | |
| "learning_rate": 9.595676696276172e-05, | |
| "loss": 0.867, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.9362649321556091, | |
| "eval_runtime": 7.5008, | |
| "eval_samples_per_second": 3.2, | |
| "eval_steps_per_second": 3.2, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.1531136929988861, | |
| "learning_rate": 9.538887392664544e-05, | |
| "loss": 0.8142, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.15327757596969604, | |
| "learning_rate": 9.478558801197065e-05, | |
| "loss": 0.8096, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.1597648411989212, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.8193, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "eval_loss": 0.9358716011047363, | |
| "eval_runtime": 7.5003, | |
| "eval_samples_per_second": 3.2, | |
| "eval_steps_per_second": 3.2, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 0.15637877583503723, | |
| "learning_rate": 9.347474647526095e-05, | |
| "loss": 0.8015, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.16304054856300354, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 0.8046, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.20407624542713165, | |
| "learning_rate": 9.202833017478422e-05, | |
| "loss": 0.7911, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.9404296278953552, | |
| "eval_runtime": 7.4973, | |
| "eval_samples_per_second": 3.201, | |
| "eval_steps_per_second": 3.201, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 12.363636363636363, | |
| "grad_norm": 0.20418952405452728, | |
| "learning_rate": 9.125567491391476e-05, | |
| "loss": 0.7724, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 12.727272727272727, | |
| "grad_norm": 0.2167651504278183, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.7557, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 12.909090909090908, | |
| "eval_loss": 0.9464240074157715, | |
| "eval_runtime": 7.534, | |
| "eval_samples_per_second": 3.186, | |
| "eval_steps_per_second": 3.186, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 13.090909090909092, | |
| "grad_norm": 0.23100724816322327, | |
| "learning_rate": 8.961448216775954e-05, | |
| "loss": 0.75, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 13.454545454545455, | |
| "grad_norm": 0.22398324310779572, | |
| "learning_rate": 8.874722443520899e-05, | |
| "loss": 0.7782, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 13.818181818181818, | |
| "grad_norm": 0.23406745493412018, | |
| "learning_rate": 8.784975278258783e-05, | |
| "loss": 0.7439, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.961434543132782, | |
| "eval_runtime": 7.4991, | |
| "eval_samples_per_second": 3.2, | |
| "eval_steps_per_second": 3.2, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 14.181818181818182, | |
| "grad_norm": 0.2717793881893158, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 0.6948, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 14.545454545454545, | |
| "grad_norm": 0.29807981848716736, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 0.7433, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "grad_norm": 0.38755056262016296, | |
| "learning_rate": 8.498316702566828e-05, | |
| "loss": 0.6475, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 14.909090909090908, | |
| "eval_loss": 0.9827837944030762, | |
| "eval_runtime": 7.4993, | |
| "eval_samples_per_second": 3.2, | |
| "eval_steps_per_second": 3.2, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 15.272727272727273, | |
| "grad_norm": 0.35523751378059387, | |
| "learning_rate": 8.397206521307584e-05, | |
| "loss": 0.6913, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 15.636363636363637, | |
| "grad_norm": 0.3585447072982788, | |
| "learning_rate": 8.293447300593402e-05, | |
| "loss": 0.6525, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.35638245940208435, | |
| "learning_rate": 8.18711994874345e-05, | |
| "loss": 0.6577, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.0100159645080566, | |
| "eval_runtime": 7.5095, | |
| "eval_samples_per_second": 3.196, | |
| "eval_steps_per_second": 3.196, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 16.363636363636363, | |
| "grad_norm": 0.4268643856048584, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 0.687, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 16.727272727272727, | |
| "grad_norm": 0.4350644648075104, | |
| "learning_rate": 7.967094433018508e-05, | |
| "loss": 0.5958, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 16.90909090909091, | |
| "eval_loss": 1.0415138006210327, | |
| "eval_runtime": 7.4981, | |
| "eval_samples_per_second": 3.201, | |
| "eval_steps_per_second": 3.201, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 17.09090909090909, | |
| "grad_norm": 0.4909497797489166, | |
| "learning_rate": 7.85356783842216e-05, | |
| "loss": 0.5815, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 17.454545454545453, | |
| "grad_norm": 0.45181789994239807, | |
| "learning_rate": 7.737816117462752e-05, | |
| "loss": 0.5734, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 17.818181818181817, | |
| "grad_norm": 0.5492954850196838, | |
| "learning_rate": 7.619929529850397e-05, | |
| "loss": 0.5391, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 1.0821270942687988, | |
| "eval_runtime": 7.505, | |
| "eval_samples_per_second": 3.198, | |
| "eval_steps_per_second": 3.198, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "step": 99, | |
| "total_flos": 4.287246251746918e+16, | |
| "train_loss": 0.9050446465761974, | |
| "train_runtime": 1813.5478, | |
| "train_samples_per_second": 2.426, | |
| "train_steps_per_second": 0.138 | |
| } | |
| ], | |
| "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": 4.287246251746918e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |