| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 325, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 5.4355961871184695, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 3.2405, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.909094332293041, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 3.0858, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.112464378906454, | |
| "learning_rate": 6.060606060606061e-05, | |
| "loss": 1.9599, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.3661963814900562, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.1188, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.06921527944926209, | |
| "learning_rate": 0.00012121212121212122, | |
| "loss": 0.0046, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.040536599266940046, | |
| "learning_rate": 0.00015151515151515152, | |
| "loss": 0.0005, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.009045738723501581, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 0.0002, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.005949166769229894, | |
| "learning_rate": 0.00019997685019798912, | |
| "loss": 0.0001, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.2264354893455914, | |
| "learning_rate": 0.0001997165380022878, | |
| "loss": 0.0001, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.0016974513785695078, | |
| "learning_rate": 0.000199167731989929, | |
| "loss": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.001311791625613364, | |
| "learning_rate": 0.0001983320199330545, | |
| "loss": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.0009269384203782724, | |
| "learning_rate": 0.00019721181966290613, | |
| "loss": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.0005037438705437563, | |
| "learning_rate": 0.00019581037207470382, | |
| "loss": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.00038243685210402487, | |
| "learning_rate": 0.00019413173175128473, | |
| "loss": 0.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.00029852007055196047, | |
| "learning_rate": 0.00019218075523263104, | |
| "loss": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.00024917847674092806, | |
| "learning_rate": 0.00018996308696522433, | |
| "loss": 0.0411, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.0005842645787073277, | |
| "learning_rate": 0.00018748514297187648, | |
| "loss": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.000982758385218839, | |
| "learning_rate": 0.00018475409228928312, | |
| "loss": 0.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.0015145443761257185, | |
| "learning_rate": 0.00018177783622700327, | |
| "loss": 0.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.0014210913401211155, | |
| "learning_rate": 0.00017856498550787144, | |
| "loss": 0.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.0016652844653558673, | |
| "learning_rate": 0.00017512483535597867, | |
| "loss": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.0012566962692868043, | |
| "learning_rate": 0.00017146733860429612, | |
| "loss": 0.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.001071493530031031, | |
| "learning_rate": 0.0001676030768997445, | |
| "loss": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.0008304219904860013, | |
| "learning_rate": 0.00016354323008901776, | |
| "loss": 0.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.0007519983627498121, | |
| "learning_rate": 0.00015929954387373103, | |
| "loss": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.0008553710391757099, | |
| "learning_rate": 0.00015488429582847192, | |
| "loss": 0.0243, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.0011485573169575547, | |
| "learning_rate": 0.00015031025988006936, | |
| "loss": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.001488839923205427, | |
| "learning_rate": 0.00014559066935084588, | |
| "loss": 0.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.001433398295827094, | |
| "learning_rate": 0.00014073917867277557, | |
| "loss": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.0007962292096287589, | |
| "learning_rate": 0.0001357698238833126, | |
| "loss": 0.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.0009535283111172765, | |
| "learning_rate": 0.000130696982017182, | |
| "loss": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.0007182907824322434, | |
| "learning_rate": 0.0001255353295116187, | |
| "loss": 0.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.0005856905396976327, | |
| "learning_rate": 0.00012029979974539234, | |
| "loss": 0.0001, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.0006597356088929785, | |
| "learning_rate": 0.00011500553983446527, | |
| "loss": 0.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.00046921969944600897, | |
| "learning_rate": 0.00010966786680927874, | |
| "loss": 0.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0004747157984707723, | |
| "learning_rate": 0.00010430222330045304, | |
| "loss": 0.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.0005057109670404985, | |
| "learning_rate": 9.892413286110886e-05, | |
| "loss": 0.0247, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0011112012256732618, | |
| "learning_rate": 9.354915505506839e-05, | |
| "loss": 0.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.0017849176040784312, | |
| "learning_rate": 8.81928404408726e-05, | |
| "loss": 0.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.0018843191660111102, | |
| "learning_rate": 8.287068558185225e-05, | |
| "loss": 0.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.012001828038146962, | |
| "learning_rate": 7.759808821241406e-05, | |
| "loss": 0.0159, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.025457948501307993, | |
| "learning_rate": 7.239030269025311e-05, | |
| "loss": 0.0002, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.006148971890773547, | |
| "learning_rate": 6.726239586337408e-05, | |
| "loss": 0.0001, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.0012801225296810595, | |
| "learning_rate": 6.22292034796035e-05, | |
| "loss": 0.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.0006558300127264007, | |
| "learning_rate": 5.730528726470792e-05, | |
| "loss": 0.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.0009452296531015114, | |
| "learning_rate": 5.2504892793295e-05, | |
| "loss": 0.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.0005350663526509846, | |
| "learning_rate": 4.7841908274384616e-05, | |
| "loss": 0.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.0005255386736926693, | |
| "learning_rate": 4.332982437088825e-05, | |
| "loss": 0.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.00047811823805356946, | |
| "learning_rate": 3.898169516924398e-05, | |
| "loss": 0.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.0004586707641668245, | |
| "learning_rate": 3.4810100412128747e-05, | |
| "loss": 0.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.0005340910278924554, | |
| "learning_rate": 3.0827109103512643e-05, | |
| "loss": 0.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.0004995001428771688, | |
| "learning_rate": 2.7044244591351232e-05, | |
| "loss": 0.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.0005545357128251603, | |
| "learning_rate": 2.3472451228937253e-05, | |
| "loss": 0.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.0004629723700238154, | |
| "learning_rate": 2.0122062711363532e-05, | |
| "loss": 0.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.0004321445477887726, | |
| "learning_rate": 1.7002772178705716e-05, | |
| "loss": 0.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.00037493112860377136, | |
| "learning_rate": 1.4123604172419713e-05, | |
| "loss": 0.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.00040990047440936877, | |
| "learning_rate": 1.149288852608743e-05, | |
| "loss": 0.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.0004512586760278399, | |
| "learning_rate": 9.118236266049707e-06, | |
| "loss": 0.0, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.00042807750826434474, | |
| "learning_rate": 7.0065175916482095e-06, | |
| "loss": 0.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.000439904459704477, | |
| "learning_rate": 5.163841998782837e-06, | |
| "loss": 0.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.0003964306028862484, | |
| "learning_rate": 3.595540604290437e-06, | |
| "loss": 0.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.00041726866247303416, | |
| "learning_rate": 2.30615072228183e-06, | |
| "loss": 0.0, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.00039289982459953004, | |
| "learning_rate": 1.2994027370611173e-06, | |
| "loss": 0.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.0004343633608887479, | |
| "learning_rate": 5.782093106048159e-07, | |
| "loss": 0.0, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.0004005418615495667, | |
| "learning_rate": 1.446569558255395e-07, | |
| "loss": 0.0111, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.00038582080807370717, | |
| "learning_rate": 0.0, | |
| "loss": 0.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 743.4758, | |
| "eval_samples_per_second": 1.555, | |
| "eval_steps_per_second": 0.389, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 325, | |
| "total_flos": 5242797101678592.0, | |
| "train_loss": 0.0818373912579521, | |
| "train_runtime": 5942.6057, | |
| "train_samples_per_second": 1.749, | |
| "train_steps_per_second": 0.055 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 325, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 40, | |
| "total_flos": 5242797101678592.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |