| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.005182152666217547, | |
| "eval_steps": 25, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00010364305332435094, | |
| "grad_norm": 15.462331771850586, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 2.2941, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00010364305332435094, | |
| "eval_loss": 0.42322224378585815, | |
| "eval_runtime": 1541.1253, | |
| "eval_samples_per_second": 5.273, | |
| "eval_steps_per_second": 2.636, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00020728610664870188, | |
| "grad_norm": 10.223081588745117, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 1.6476, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0003109291599730528, | |
| "grad_norm": 5.350977897644043, | |
| "learning_rate": 0.0001, | |
| "loss": 1.594, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00041457221329740376, | |
| "grad_norm": 4.774840831756592, | |
| "learning_rate": 9.99524110790929e-05, | |
| "loss": 1.167, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0005182152666217547, | |
| "grad_norm": 3.6830101013183594, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.7528, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0006218583199461056, | |
| "grad_norm": 5.286176681518555, | |
| "learning_rate": 9.957224306869053e-05, | |
| "loss": 0.9711, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0007255013732704566, | |
| "grad_norm": 3.7488551139831543, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.8706, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0008291444265948075, | |
| "grad_norm": 2.7980217933654785, | |
| "learning_rate": 9.881480035599667e-05, | |
| "loss": 0.8608, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0009327874799191584, | |
| "grad_norm": 4.999605655670166, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 1.1006, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0010364305332435093, | |
| "grad_norm": 2.628920555114746, | |
| "learning_rate": 9.768584753741134e-05, | |
| "loss": 1.1324, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0011400735865678604, | |
| "grad_norm": 2.1676831245422363, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.7149, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0012437166398922112, | |
| "grad_norm": 2.7991559505462646, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.774, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.001347359693216562, | |
| "grad_norm": 2.4429965019226074, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.6822, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0014510027465409131, | |
| "grad_norm": 3.050663948059082, | |
| "learning_rate": 9.435054165891109e-05, | |
| "loss": 0.959, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.001554645799865264, | |
| "grad_norm": 2.318708896636963, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.7727, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.001658288853189615, | |
| "grad_norm": 2.0071775913238525, | |
| "learning_rate": 9.21695722906443e-05, | |
| "loss": 0.7567, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0017619319065139659, | |
| "grad_norm": 2.5710930824279785, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 0.7541, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0018655749598383167, | |
| "grad_norm": 2.7265591621398926, | |
| "learning_rate": 8.966766701456177e-05, | |
| "loss": 0.7756, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0019692180131626676, | |
| "grad_norm": 2.997084140777588, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 0.6538, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0020728610664870186, | |
| "grad_norm": 2.4743478298187256, | |
| "learning_rate": 8.68638668405062e-05, | |
| "loss": 0.6616, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0021765041198113697, | |
| "grad_norm": 2.2632076740264893, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.6122, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0022801471731357208, | |
| "grad_norm": 2.802534580230713, | |
| "learning_rate": 8.377951038078302e-05, | |
| "loss": 0.9774, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0023837902264600714, | |
| "grad_norm": 2.707726001739502, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 0.7128, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0024874332797844225, | |
| "grad_norm": 3.9078145027160645, | |
| "learning_rate": 8.043807145043604e-05, | |
| "loss": 0.9834, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0025910763331087735, | |
| "grad_norm": 3.39276385307312, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 1.0492, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0025910763331087735, | |
| "eval_loss": 0.10315833985805511, | |
| "eval_runtime": 1540.5199, | |
| "eval_samples_per_second": 5.275, | |
| "eval_steps_per_second": 2.637, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.002694719386433124, | |
| "grad_norm": 2.8625752925872803, | |
| "learning_rate": 7.68649804173412e-05, | |
| "loss": 0.6604, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.002798362439757475, | |
| "grad_norm": 2.1166653633117676, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.8203, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0029020054930818263, | |
| "grad_norm": 2.829029083251953, | |
| "learning_rate": 7.308743066175172e-05, | |
| "loss": 0.7229, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.0030056485464061773, | |
| "grad_norm": 2.9142374992370605, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 1.1018, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.003109291599730528, | |
| "grad_norm": 2.367736339569092, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.7547, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003212934653054879, | |
| "grad_norm": 2.9502367973327637, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 1.0028, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.00331657770637923, | |
| "grad_norm": 3.18487811088562, | |
| "learning_rate": 6.503528997521366e-05, | |
| "loss": 0.8339, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0034202207597035807, | |
| "grad_norm": 3.829874277114868, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 0.9514, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0035238638130279318, | |
| "grad_norm": 2.686802864074707, | |
| "learning_rate": 6.0821980696905146e-05, | |
| "loss": 0.8758, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.003627506866352283, | |
| "grad_norm": 3.39548921585083, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 0.7906, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0037311499196766335, | |
| "grad_norm": 4.071652412414551, | |
| "learning_rate": 5.6526309611002594e-05, | |
| "loss": 0.8588, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0038347929730009845, | |
| "grad_norm": 5.348927974700928, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 1.3739, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.003938436026325335, | |
| "grad_norm": 4.617519378662109, | |
| "learning_rate": 5.218096936826681e-05, | |
| "loss": 1.1862, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.004042079079649687, | |
| "grad_norm": 4.1698150634765625, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0598, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.004145722132974037, | |
| "grad_norm": 4.958194255828857, | |
| "learning_rate": 4.781903063173321e-05, | |
| "loss": 0.8871, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.004249365186298389, | |
| "grad_norm": 3.777562141418457, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 0.8924, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.004353008239622739, | |
| "grad_norm": 2.7251391410827637, | |
| "learning_rate": 4.347369038899744e-05, | |
| "loss": 0.5079, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.00445665129294709, | |
| "grad_norm": 3.8885464668273926, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 0.8651, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0045602943462714415, | |
| "grad_norm": 3.1904706954956055, | |
| "learning_rate": 3.917801930309486e-05, | |
| "loss": 0.4951, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.004663937399595792, | |
| "grad_norm": 2.1530394554138184, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 0.3895, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.004767580452920143, | |
| "grad_norm": 2.2201197147369385, | |
| "learning_rate": 3.4964710024786354e-05, | |
| "loss": 0.3857, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.004871223506244494, | |
| "grad_norm": 2.258897304534912, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 0.2666, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.004974866559568845, | |
| "grad_norm": 2.838489294052124, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.4107, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0050785096128931955, | |
| "grad_norm": 4.822262287139893, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 0.485, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.005182152666217547, | |
| "grad_norm": 14.949175834655762, | |
| "learning_rate": 2.6912569338248315e-05, | |
| "loss": 1.2498, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.005182152666217547, | |
| "eval_loss": 0.10206064581871033, | |
| "eval_runtime": 1542.114, | |
| "eval_samples_per_second": 5.269, | |
| "eval_steps_per_second": 2.635, | |
| "step": 50 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 75, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.44383432317993e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |