| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.041235696367822414, | |
| "eval_steps": 500, | |
| "global_step": 600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00048108312429126146, | |
| "grad_norm": 8.154312133789062, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.8456, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0009621662485825229, | |
| "grad_norm": 2.8053348064422607, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 1.0127, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0014432493728737843, | |
| "grad_norm": 5.31929349899292, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.791, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0019243324971650458, | |
| "grad_norm": 5.091348171234131, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.8497, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.002405415621456307, | |
| "grad_norm": 3.4551076889038086, | |
| "learning_rate": 2.2666666666666668e-05, | |
| "loss": 0.5847, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0028864987457475687, | |
| "grad_norm": 2.0914766788482666, | |
| "learning_rate": 2.733333333333333e-05, | |
| "loss": 0.6838, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.00336758187003883, | |
| "grad_norm": 4.122292518615723, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.5328, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.0038486649943300917, | |
| "grad_norm": 4.441967964172363, | |
| "learning_rate": 3.6666666666666666e-05, | |
| "loss": 0.7609, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.004329748118621354, | |
| "grad_norm": 2.408400535583496, | |
| "learning_rate": 4.133333333333333e-05, | |
| "loss": 0.7353, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.004810831242912614, | |
| "grad_norm": 4.386236190795898, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.5822, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.005291914367203876, | |
| "grad_norm": 2.6810431480407715, | |
| "learning_rate": 5.0666666666666674e-05, | |
| "loss": 0.5084, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.005772997491495137, | |
| "grad_norm": 10.199173927307129, | |
| "learning_rate": 5.5333333333333334e-05, | |
| "loss": 1.0656, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.006254080615786399, | |
| "grad_norm": 2.62958025932312, | |
| "learning_rate": 6e-05, | |
| "loss": 0.9499, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.00673516374007766, | |
| "grad_norm": 3.928619861602783, | |
| "learning_rate": 6.466666666666666e-05, | |
| "loss": 0.7763, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.007216246864368922, | |
| "grad_norm": 7.499195098876953, | |
| "learning_rate": 6.933333333333334e-05, | |
| "loss": 1.0431, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.007697329988660183, | |
| "grad_norm": 1.4210686683654785, | |
| "learning_rate": 7.4e-05, | |
| "loss": 0.5441, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.008178413112951446, | |
| "grad_norm": 3.6974828243255615, | |
| "learning_rate": 7.866666666666666e-05, | |
| "loss": 0.6841, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.008659496237242707, | |
| "grad_norm": 7.652034282684326, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 0.8487, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.009140579361533967, | |
| "grad_norm": 11.174522399902344, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.9727, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.009621662485825229, | |
| "grad_norm": 11.040456771850586, | |
| "learning_rate": 9.266666666666666e-05, | |
| "loss": 1.122, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.01010274561011649, | |
| "grad_norm": 8.87497329711914, | |
| "learning_rate": 9.733333333333335e-05, | |
| "loss": 1.4102, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.010583828734407752, | |
| "grad_norm": 9.45786190032959, | |
| "learning_rate": 9.999878153526974e-05, | |
| "loss": 1.2933, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.011064911858699013, | |
| "grad_norm": 8.866625785827637, | |
| "learning_rate": 9.998646205897309e-05, | |
| "loss": 1.2747, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.011545994982990275, | |
| "grad_norm": 6.202242374420166, | |
| "learning_rate": 9.996087868740243e-05, | |
| "loss": 1.7031, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.012027078107281536, | |
| "grad_norm": 10.355546951293945, | |
| "learning_rate": 9.992203820909906e-05, | |
| "loss": 1.8137, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.012508161231572798, | |
| "grad_norm": 8.716798782348633, | |
| "learning_rate": 9.986995093037421e-05, | |
| "loss": 1.2634, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.01298924435586406, | |
| "grad_norm": 11.727725982666016, | |
| "learning_rate": 9.980463067257437e-05, | |
| "loss": 0.8589, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.01347032748015532, | |
| "grad_norm": 7.918148517608643, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.4014, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.013951410604446582, | |
| "grad_norm": 28.750410079956055, | |
| "learning_rate": 9.963436405737476e-05, | |
| "loss": 1.2997, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.014432493728737844, | |
| "grad_norm": 15.347841262817383, | |
| "learning_rate": 9.952946288017899e-05, | |
| "loss": 1.766, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.014913576853029105, | |
| "grad_norm": 7.8105950355529785, | |
| "learning_rate": 9.941141907232765e-05, | |
| "loss": 1.4616, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.015394659977320367, | |
| "grad_norm": 16.68589973449707, | |
| "learning_rate": 9.928026395671576e-05, | |
| "loss": 1.0106, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.01587574310161163, | |
| "grad_norm": 14.563543319702148, | |
| "learning_rate": 9.913603233532067e-05, | |
| "loss": 1.2966, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.01635682622590289, | |
| "grad_norm": 11.76324462890625, | |
| "learning_rate": 9.89787624799672e-05, | |
| "loss": 1.4801, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.016837909350194153, | |
| "grad_norm": 10.999220848083496, | |
| "learning_rate": 9.880849612217238e-05, | |
| "loss": 1.7505, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.017318992474485415, | |
| "grad_norm": 17.89532470703125, | |
| "learning_rate": 9.862527844207189e-05, | |
| "loss": 0.8774, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.017800075598776673, | |
| "grad_norm": 7.980606555938721, | |
| "learning_rate": 9.842915805643155e-05, | |
| "loss": 1.7281, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.018281158723067934, | |
| "grad_norm": 17.765323638916016, | |
| "learning_rate": 9.822018700574695e-05, | |
| "loss": 1.2403, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.018762241847359196, | |
| "grad_norm": 18.817676544189453, | |
| "learning_rate": 9.799842074043439e-05, | |
| "loss": 1.6112, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.019243324971650457, | |
| "grad_norm": 15.211094856262207, | |
| "learning_rate": 9.776391810611718e-05, | |
| "loss": 0.784, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.01972440809594172, | |
| "grad_norm": 13.791271209716797, | |
| "learning_rate": 9.751674132801107e-05, | |
| "loss": 1.5379, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.02020549122023298, | |
| "grad_norm": 12.567949295043945, | |
| "learning_rate": 9.725695599441258e-05, | |
| "loss": 1.7221, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.02068657434452424, | |
| "grad_norm": 119.533447265625, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 2.089, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.021167657468815503, | |
| "grad_norm": 11.758546829223633, | |
| "learning_rate": 9.669983872401867e-05, | |
| "loss": 1.3068, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.021648740593106765, | |
| "grad_norm": 14.528461456298828, | |
| "learning_rate": 9.640265461815234e-05, | |
| "loss": 1.6841, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.022129823717398026, | |
| "grad_norm": 12.57625675201416, | |
| "learning_rate": 9.609315757942503e-05, | |
| "loss": 1.2772, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.022610906841689288, | |
| "grad_norm": 20.562517166137695, | |
| "learning_rate": 9.577142973279896e-05, | |
| "loss": 1.1584, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.02309198996598055, | |
| "grad_norm": 15.014174461364746, | |
| "learning_rate": 9.543755644867822e-05, | |
| "loss": 1.9943, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.02357307309027181, | |
| "grad_norm": 3.206474542617798, | |
| "learning_rate": 9.50916263202557e-05, | |
| "loss": 1.1005, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.024054156214563072, | |
| "grad_norm": 20.237760543823242, | |
| "learning_rate": 9.473373114000492e-05, | |
| "loss": 1.1736, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.024535239338854334, | |
| "grad_norm": 23.357084274291992, | |
| "learning_rate": 9.436396587532296e-05, | |
| "loss": 1.6269, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.025016322463145595, | |
| "grad_norm": 25.333106994628906, | |
| "learning_rate": 9.398242864333083e-05, | |
| "loss": 1.1378, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.025497405587436857, | |
| "grad_norm": 28.834636688232422, | |
| "learning_rate": 9.358922068483812e-05, | |
| "loss": 0.9558, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.02597848871172812, | |
| "grad_norm": 10.581332206726074, | |
| "learning_rate": 9.318444633747883e-05, | |
| "loss": 1.4568, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.02645957183601938, | |
| "grad_norm": 22.025737762451172, | |
| "learning_rate": 9.276821300802534e-05, | |
| "loss": 1.2215, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.02694065496031064, | |
| "grad_norm": 23.992765426635742, | |
| "learning_rate": 9.234063114388809e-05, | |
| "loss": 2.0521, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.027421738084601903, | |
| "grad_norm": 18.102312088012695, | |
| "learning_rate": 9.190181420380836e-05, | |
| "loss": 1.1839, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.027902821208893164, | |
| "grad_norm": 21.294204711914062, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 1.0416, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.028383904333184426, | |
| "grad_norm": 11.440023422241211, | |
| "learning_rate": 9.099094380601244e-05, | |
| "loss": 0.9043, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.028864987457475687, | |
| "grad_norm": 18.921781539916992, | |
| "learning_rate": 9.051913204752972e-05, | |
| "loss": 1.373, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.02934607058176695, | |
| "grad_norm": 27.478322982788086, | |
| "learning_rate": 9.003656854743667e-05, | |
| "loss": 1.8896, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.02982715370605821, | |
| "grad_norm": 24.056333541870117, | |
| "learning_rate": 8.954338135383804e-05, | |
| "loss": 1.8207, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.030308236830349472, | |
| "grad_norm": 29.121618270874023, | |
| "learning_rate": 8.903970133383297e-05, | |
| "loss": 2.3176, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.030789319954640734, | |
| "grad_norm": 24.770221710205078, | |
| "learning_rate": 8.852566213878947e-05, | |
| "loss": 1.4976, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.031270403078931995, | |
| "grad_norm": 12.233966827392578, | |
| "learning_rate": 8.800140016888009e-05, | |
| "loss": 1.1954, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.03175148620322326, | |
| "grad_norm": 28.35350799560547, | |
| "learning_rate": 8.746705453688814e-05, | |
| "loss": 1.8075, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.03223256932751452, | |
| "grad_norm": 15.999966621398926, | |
| "learning_rate": 8.692276703129421e-05, | |
| "loss": 1.3143, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.03271365245180578, | |
| "grad_norm": 10.898785591125488, | |
| "learning_rate": 8.636868207865244e-05, | |
| "loss": 1.2549, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.03319473557609704, | |
| "grad_norm": 26.46807861328125, | |
| "learning_rate": 8.580494670526725e-05, | |
| "loss": 1.5415, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.033675818700388306, | |
| "grad_norm": 33.8633918762207, | |
| "learning_rate": 8.523171049817974e-05, | |
| "loss": 2.2099, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.034156901824679564, | |
| "grad_norm": 22.85337257385254, | |
| "learning_rate": 8.464912556547486e-05, | |
| "loss": 1.7322, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.03463798494897083, | |
| "grad_norm": 39.19854736328125, | |
| "learning_rate": 8.405734649591963e-05, | |
| "loss": 1.7439, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.03511906807326209, | |
| "grad_norm": 24.207155227661133, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.3688, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.035600151197553345, | |
| "grad_norm": 13.661985397338867, | |
| "learning_rate": 8.284683645796813e-05, | |
| "loss": 1.2163, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.03608123432184461, | |
| "grad_norm": 13.821810722351074, | |
| "learning_rate": 8.222842669810935e-05, | |
| "loss": 2.1381, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.03656231744613587, | |
| "grad_norm": 17.461071014404297, | |
| "learning_rate": 8.160146513324254e-05, | |
| "loss": 1.834, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.03704340057042713, | |
| "grad_norm": 27.663408279418945, | |
| "learning_rate": 8.096611812746301e-05, | |
| "loss": 1.7051, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.03752448369471839, | |
| "grad_norm": 20.866432189941406, | |
| "learning_rate": 8.032255426994069e-05, | |
| "loss": 1.5578, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.038005566819009656, | |
| "grad_norm": 18.067535400390625, | |
| "learning_rate": 7.967094433018508e-05, | |
| "loss": 1.9054, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.038486649943300914, | |
| "grad_norm": 19.965730667114258, | |
| "learning_rate": 7.901146121273164e-05, | |
| "loss": 1.2985, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.03896773306759218, | |
| "grad_norm": 13.658997535705566, | |
| "learning_rate": 7.834427991126155e-05, | |
| "loss": 1.419, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.03944881619188344, | |
| "grad_norm": 53.16463088989258, | |
| "learning_rate": 7.766957746216721e-05, | |
| "loss": 1.7695, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.0399298993161747, | |
| "grad_norm": 17.96440887451172, | |
| "learning_rate": 7.698753289757565e-05, | |
| "loss": 2.298, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.04041098244046596, | |
| "grad_norm": 34.66753387451172, | |
| "learning_rate": 7.629832719784245e-05, | |
| "loss": 1.3731, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.040892065564757225, | |
| "grad_norm": 15.050975799560547, | |
| "learning_rate": 7.560214324352858e-05, | |
| "loss": 1.7391, | |
| "step": 595 | |
| } | |
| ], | |
| "logging_steps": 7, | |
| "max_steps": 1500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 300, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.259314668896256e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |