| { | |
| "best_metric": 0.45555031299591064, | |
| "best_model_checkpoint": "data/Gemma-2-2B_task-3_120-samples_config-1/checkpoint-55", | |
| "epoch": 12.0, | |
| "eval_steps": 500, | |
| "global_step": 132, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 10.476847648620605, | |
| "learning_rate": 1.818181818181818e-06, | |
| "loss": 4.4215, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 6.425926208496094, | |
| "learning_rate": 3.636363636363636e-06, | |
| "loss": 3.3941, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 6.461874008178711, | |
| "learning_rate": 7.272727272727272e-06, | |
| "loss": 3.4752, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 5.155916213989258, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 3.2494, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 8.943306922912598, | |
| "learning_rate": 1.4545454545454545e-05, | |
| "loss": 3.2396, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 7.687283039093018, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 2.9782, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.7756738662719727, | |
| "eval_runtime": 7.6912, | |
| "eval_samples_per_second": 3.12, | |
| "eval_steps_per_second": 3.12, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.0909090909090908, | |
| "grad_norm": 5.9325127601623535, | |
| "learning_rate": 2.1818181818181818e-05, | |
| "loss": 2.7883, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.2727272727272727, | |
| "grad_norm": 5.49841833114624, | |
| "learning_rate": 2.5454545454545454e-05, | |
| "loss": 2.351, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.4545454545454546, | |
| "grad_norm": 4.112461566925049, | |
| "learning_rate": 2.909090909090909e-05, | |
| "loss": 2.251, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.6363636363636362, | |
| "grad_norm": 3.527035713195801, | |
| "learning_rate": 3.272727272727273e-05, | |
| "loss": 1.6547, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.8181818181818183, | |
| "grad_norm": 2.4498353004455566, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 1.3783, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.315340280532837, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0655, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.049608588218689, | |
| "eval_runtime": 7.4517, | |
| "eval_samples_per_second": 3.221, | |
| "eval_steps_per_second": 3.221, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.1818181818181817, | |
| "grad_norm": 2.298762321472168, | |
| "learning_rate": 4.3636363636363636e-05, | |
| "loss": 0.8694, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.3636363636363638, | |
| "grad_norm": 1.123829960823059, | |
| "learning_rate": 4.7272727272727275e-05, | |
| "loss": 0.8959, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.5454545454545454, | |
| "grad_norm": 0.9805992841720581, | |
| "learning_rate": 5.090909090909091e-05, | |
| "loss": 0.6095, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.7272727272727275, | |
| "grad_norm": 0.8541581630706787, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 0.5157, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.909090909090909, | |
| "grad_norm": 0.5621142387390137, | |
| "learning_rate": 5.818181818181818e-05, | |
| "loss": 0.3407, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.5811724662780762, | |
| "eval_runtime": 7.4549, | |
| "eval_samples_per_second": 3.219, | |
| "eval_steps_per_second": 3.219, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.090909090909091, | |
| "grad_norm": 0.7350859045982361, | |
| "learning_rate": 6.181818181818182e-05, | |
| "loss": 0.6581, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.2727272727272725, | |
| "grad_norm": 0.35103848576545715, | |
| "learning_rate": 6.545454545454546e-05, | |
| "loss": 0.1479, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.4545454545454546, | |
| "grad_norm": 0.6595249772071838, | |
| "learning_rate": 6.90909090909091e-05, | |
| "loss": 0.4825, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.6363636363636362, | |
| "grad_norm": 0.7842703461647034, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.4647, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.8181818181818183, | |
| "grad_norm": 0.7872003316879272, | |
| "learning_rate": 7.636363636363637e-05, | |
| "loss": 0.5288, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6303789615631104, | |
| "learning_rate": 8e-05, | |
| "loss": 0.4394, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.4946717321872711, | |
| "eval_runtime": 7.4514, | |
| "eval_samples_per_second": 3.221, | |
| "eval_steps_per_second": 3.221, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.181818181818182, | |
| "grad_norm": 0.4205116927623749, | |
| "learning_rate": 8.363636363636364e-05, | |
| "loss": 0.2832, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.363636363636363, | |
| "grad_norm": 0.7606728076934814, | |
| "learning_rate": 8.727272727272727e-05, | |
| "loss": 0.4348, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 0.537746012210846, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.2961, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.7272727272727275, | |
| "grad_norm": 0.5678563714027405, | |
| "learning_rate": 9.454545454545455e-05, | |
| "loss": 0.3423, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.909090909090909, | |
| "grad_norm": 0.6702199578285217, | |
| "learning_rate": 9.818181818181818e-05, | |
| "loss": 0.3251, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.45555031299591064, | |
| "eval_runtime": 7.4522, | |
| "eval_samples_per_second": 3.221, | |
| "eval_steps_per_second": 3.221, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.090909090909091, | |
| "grad_norm": 0.6493393182754517, | |
| "learning_rate": 9.999899300364532e-05, | |
| "loss": 0.3555, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.2727272727272725, | |
| "grad_norm": 0.6028119921684265, | |
| "learning_rate": 9.99909372761763e-05, | |
| "loss": 0.2755, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.454545454545454, | |
| "grad_norm": 0.513911247253418, | |
| "learning_rate": 9.997482711915927e-05, | |
| "loss": 0.193, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.636363636363637, | |
| "grad_norm": 0.38173243403434753, | |
| "learning_rate": 9.99506651282272e-05, | |
| "loss": 0.2908, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 5.818181818181818, | |
| "grad_norm": 0.5847815871238708, | |
| "learning_rate": 9.991845519630678e-05, | |
| "loss": 0.1816, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.7407585978507996, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.2914, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.4621676504611969, | |
| "eval_runtime": 7.4658, | |
| "eval_samples_per_second": 3.215, | |
| "eval_steps_per_second": 3.215, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.181818181818182, | |
| "grad_norm": 0.561898946762085, | |
| "learning_rate": 9.982991356370404e-05, | |
| "loss": 0.177, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.363636363636363, | |
| "grad_norm": 0.5280710458755493, | |
| "learning_rate": 9.977359612865423e-05, | |
| "loss": 0.1615, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.545454545454545, | |
| "grad_norm": 0.905296802520752, | |
| "learning_rate": 9.970925928158274e-05, | |
| "loss": 0.2824, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 6.7272727272727275, | |
| "grad_norm": 0.9980334043502808, | |
| "learning_rate": 9.963691338830044e-05, | |
| "loss": 0.2253, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 6.909090909090909, | |
| "grad_norm": 0.6320990324020386, | |
| "learning_rate": 9.955657010501806e-05, | |
| "loss": 0.1443, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.5257542133331299, | |
| "eval_runtime": 7.4501, | |
| "eval_samples_per_second": 3.221, | |
| "eval_steps_per_second": 3.221, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.090909090909091, | |
| "grad_norm": 0.712302565574646, | |
| "learning_rate": 9.946824237646824e-05, | |
| "loss": 0.098, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.2727272727272725, | |
| "grad_norm": 1.163100004196167, | |
| "learning_rate": 9.937194443381972e-05, | |
| "loss": 0.1656, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.454545454545454, | |
| "grad_norm": 0.5829131603240967, | |
| "learning_rate": 9.926769179238466e-05, | |
| "loss": 0.0639, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 7.636363636363637, | |
| "grad_norm": 0.5312383770942688, | |
| "learning_rate": 9.915550124911866e-05, | |
| "loss": 0.0833, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 7.818181818181818, | |
| "grad_norm": 0.5749314427375793, | |
| "learning_rate": 9.903539087991462e-05, | |
| "loss": 0.1238, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.2859961986541748, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.112, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.5946211218833923, | |
| "eval_runtime": 7.4535, | |
| "eval_samples_per_second": 3.22, | |
| "eval_steps_per_second": 3.22, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.181818181818182, | |
| "grad_norm": 0.5246594548225403, | |
| "learning_rate": 9.877148934427037e-05, | |
| "loss": 0.0732, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.363636363636363, | |
| "grad_norm": 0.901326060295105, | |
| "learning_rate": 9.862774069706346e-05, | |
| "loss": 0.0431, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 8.545454545454545, | |
| "grad_norm": 1.6021242141723633, | |
| "learning_rate": 9.847615725553456e-05, | |
| "loss": 0.0323, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 8.727272727272727, | |
| "grad_norm": 0.8605382442474365, | |
| "learning_rate": 9.831676344247342e-05, | |
| "loss": 0.0508, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 8.909090909090908, | |
| "grad_norm": 1.1480231285095215, | |
| "learning_rate": 9.814958493905963e-05, | |
| "loss": 0.059, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.7451796531677246, | |
| "eval_runtime": 7.4546, | |
| "eval_samples_per_second": 3.22, | |
| "eval_steps_per_second": 3.22, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 9.090909090909092, | |
| "grad_norm": 1.022650122642517, | |
| "learning_rate": 9.797464868072488e-05, | |
| "loss": 0.0546, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.272727272727273, | |
| "grad_norm": 0.11207925528287888, | |
| "learning_rate": 9.779198285281325e-05, | |
| "loss": 0.0134, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 9.454545454545455, | |
| "grad_norm": 1.7919929027557373, | |
| "learning_rate": 9.760161688604008e-05, | |
| "loss": 0.0474, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 9.636363636363637, | |
| "grad_norm": 1.1809546947479248, | |
| "learning_rate": 9.740358145174998e-05, | |
| "loss": 0.0233, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 9.818181818181818, | |
| "grad_norm": 1.2386544942855835, | |
| "learning_rate": 9.719790845697533e-05, | |
| "loss": 0.0228, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.7144274711608887, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0304, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.8397977948188782, | |
| "eval_runtime": 7.4498, | |
| "eval_samples_per_second": 3.222, | |
| "eval_steps_per_second": 3.222, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 10.181818181818182, | |
| "grad_norm": 0.36995142698287964, | |
| "learning_rate": 9.676378356149734e-05, | |
| "loss": 0.007, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 10.363636363636363, | |
| "grad_norm": 1.0517113208770752, | |
| "learning_rate": 9.653540160603956e-05, | |
| "loss": 0.0256, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 10.545454545454545, | |
| "grad_norm": 0.5318811535835266, | |
| "learning_rate": 9.629952196931901e-05, | |
| "loss": 0.01, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 10.727272727272727, | |
| "grad_norm": 1.7974720001220703, | |
| "learning_rate": 9.60561826557425e-05, | |
| "loss": 0.0349, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 10.909090909090908, | |
| "grad_norm": 0.7175325155258179, | |
| "learning_rate": 9.580542287160348e-05, | |
| "loss": 0.0078, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 0.8924921154975891, | |
| "eval_runtime": 7.4516, | |
| "eval_samples_per_second": 3.221, | |
| "eval_steps_per_second": 3.221, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 11.090909090909092, | |
| "grad_norm": 0.5240145921707153, | |
| "learning_rate": 9.554728301876526e-05, | |
| "loss": 0.0187, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 11.272727272727273, | |
| "grad_norm": 1.4186664819717407, | |
| "learning_rate": 9.528180468815155e-05, | |
| "loss": 0.0316, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 11.454545454545455, | |
| "grad_norm": 0.2332746386528015, | |
| "learning_rate": 9.50090306530454e-05, | |
| "loss": 0.0139, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 11.636363636363637, | |
| "grad_norm": 0.6709339022636414, | |
| "learning_rate": 9.472900486219769e-05, | |
| "loss": 0.0127, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 11.818181818181818, | |
| "grad_norm": 0.08679351955652237, | |
| "learning_rate": 9.444177243274618e-05, | |
| "loss": 0.0065, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.8816165328025818, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0205, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.9054031372070312, | |
| "eval_runtime": 7.4621, | |
| "eval_samples_per_second": 3.216, | |
| "eval_steps_per_second": 3.216, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "step": 132, | |
| "total_flos": 2.837694963829965e+16, | |
| "train_loss": 0.6035953224495505, | |
| "train_runtime": 1200.5262, | |
| "train_samples_per_second": 3.665, | |
| "train_steps_per_second": 0.458 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 550, | |
| "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": 2.837694963829965e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |