| { | |
| "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": 2.77743332230532, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 2.1182, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 2.50025816437926, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 2.02, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 4.263857919701068, | |
| "learning_rate": 6.060606060606061e-05, | |
| "loss": 1.434, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.2770280865100368, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.1461, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0325189453594692, | |
| "learning_rate": 0.00012121212121212122, | |
| "loss": 0.0037, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.014426942936471324, | |
| "learning_rate": 0.00015151515151515152, | |
| "loss": 0.0001, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.009014230050117448, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.005869394562969202, | |
| "learning_rate": 0.00019997685019798912, | |
| "loss": 0.0001, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.07549895461350618, | |
| "learning_rate": 0.0001997165380022878, | |
| "loss": 0.0002, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.005627069022101588, | |
| "learning_rate": 0.000199167731989929, | |
| "loss": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.008254135808213726, | |
| "learning_rate": 0.0001983320199330545, | |
| "loss": 0.0001, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.005212629414025529, | |
| "learning_rate": 0.00019721181966290613, | |
| "loss": 0.0001, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.004089859449975345, | |
| "learning_rate": 0.00019581037207470382, | |
| "loss": 0.0001, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.004282429012831237, | |
| "learning_rate": 0.00019413173175128473, | |
| "loss": 0.0001, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.004670275932686161, | |
| "learning_rate": 0.00019218075523263104, | |
| "loss": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.01607388068076669, | |
| "learning_rate": 0.00018996308696522433, | |
| "loss": 0.0384, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.006214309395278606, | |
| "learning_rate": 0.00018748514297187648, | |
| "loss": 0.0004, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.00888958260744134, | |
| "learning_rate": 0.00018475409228928312, | |
| "loss": 0.0001, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.0054672192077601664, | |
| "learning_rate": 0.00018177783622700327, | |
| "loss": 0.0001, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.004208054445057372, | |
| "learning_rate": 0.00017856498550787144, | |
| "loss": 0.0001, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.003471247852347678, | |
| "learning_rate": 0.00017512483535597867, | |
| "loss": 0.0001, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.0041578302953389655, | |
| "learning_rate": 0.00017146733860429612, | |
| "loss": 0.0001, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.004371488630944069, | |
| "learning_rate": 0.0001676030768997445, | |
| "loss": 0.0001, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.0031624863947064662, | |
| "learning_rate": 0.00016354323008901776, | |
| "loss": 0.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.0041366037870679655, | |
| "learning_rate": 0.00015929954387373103, | |
| "loss": 0.0001, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.007463102073942976, | |
| "learning_rate": 0.00015488429582847192, | |
| "loss": 0.0205, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.004091219704181507, | |
| "learning_rate": 0.00015031025988006936, | |
| "loss": 0.0002, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.0035397410165078835, | |
| "learning_rate": 0.00014559066935084588, | |
| "loss": 0.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.003466344781174203, | |
| "learning_rate": 0.00014073917867277557, | |
| "loss": 0.0001, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.0035356522209539923, | |
| "learning_rate": 0.0001357698238833126, | |
| "loss": 0.0001, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.002910323336169199, | |
| "learning_rate": 0.000130696982017182, | |
| "loss": 0.0001, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.0028248709609393647, | |
| "learning_rate": 0.0001255353295116187, | |
| "loss": 0.0001, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.003059403330795667, | |
| "learning_rate": 0.00012029979974539234, | |
| "loss": 0.0001, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.003940653934284354, | |
| "learning_rate": 0.00011500553983446527, | |
| "loss": 0.0001, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.0025607923792118335, | |
| "learning_rate": 0.00010966786680927874, | |
| "loss": 0.0001, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0029630683076141157, | |
| "learning_rate": 0.00010430222330045304, | |
| "loss": 0.0001, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.01274417763198295, | |
| "learning_rate": 9.892413286110886e-05, | |
| "loss": 0.0201, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0026421572879934616, | |
| "learning_rate": 9.354915505506839e-05, | |
| "loss": 0.0003, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.003018848235058481, | |
| "learning_rate": 8.81928404408726e-05, | |
| "loss": 0.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.0027839642805129973, | |
| "learning_rate": 8.287068558185225e-05, | |
| "loss": 0.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.005195726503619321, | |
| "learning_rate": 7.759808821241406e-05, | |
| "loss": 0.0122, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.004534848892475077, | |
| "learning_rate": 7.239030269025311e-05, | |
| "loss": 0.0002, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.0027408262178367257, | |
| "learning_rate": 6.726239586337408e-05, | |
| "loss": 0.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.002654161342750701, | |
| "learning_rate": 6.22292034796035e-05, | |
| "loss": 0.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.002293539973873259, | |
| "learning_rate": 5.730528726470792e-05, | |
| "loss": 0.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.004409012304460454, | |
| "learning_rate": 5.2504892793295e-05, | |
| "loss": 0.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.0027637928722407086, | |
| "learning_rate": 4.7841908274384616e-05, | |
| "loss": 0.0001, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.003785176205158752, | |
| "learning_rate": 4.332982437088825e-05, | |
| "loss": 0.0001, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.003078607123626658, | |
| "learning_rate": 3.898169516924398e-05, | |
| "loss": 0.0001, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.0021769919676504494, | |
| "learning_rate": 3.4810100412128747e-05, | |
| "loss": 0.0001, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.0030869887324349193, | |
| "learning_rate": 3.0827109103512643e-05, | |
| "loss": 0.0001, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.0034257125076202414, | |
| "learning_rate": 2.7044244591351232e-05, | |
| "loss": 0.0001, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.004634548598914417, | |
| "learning_rate": 2.3472451228937253e-05, | |
| "loss": 0.0001, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.002936488614426748, | |
| "learning_rate": 2.0122062711363532e-05, | |
| "loss": 0.0001, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.0021650629608059805, | |
| "learning_rate": 1.7002772178705716e-05, | |
| "loss": 0.0001, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.0025609854777170577, | |
| "learning_rate": 1.4123604172419713e-05, | |
| "loss": 0.0001, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.0024555677739051917, | |
| "learning_rate": 1.149288852608743e-05, | |
| "loss": 0.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.00307921034872769, | |
| "learning_rate": 9.118236266049707e-06, | |
| "loss": 0.0001, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.0025023073547329236, | |
| "learning_rate": 7.0065175916482095e-06, | |
| "loss": 0.0001, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.0027324061007535385, | |
| "learning_rate": 5.163841998782837e-06, | |
| "loss": 0.0001, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.002415645169583888, | |
| "learning_rate": 3.595540604290437e-06, | |
| "loss": 0.0001, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.0030523000925782163, | |
| "learning_rate": 2.30615072228183e-06, | |
| "loss": 0.0001, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.0024719719554582575, | |
| "learning_rate": 1.2994027370611173e-06, | |
| "loss": 0.0001, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.004495220838516153, | |
| "learning_rate": 5.782093106048159e-07, | |
| "loss": 0.0001, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.0023313114504467775, | |
| "learning_rate": 1.446569558255395e-07, | |
| "loss": 0.0071, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.0027508255197871145, | |
| "learning_rate": 0.0, | |
| "loss": 0.0001, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 743.2888, | |
| "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": 0.057327271424807036, | |
| "train_runtime": 5928.0915, | |
| "train_samples_per_second": 1.753, | |
| "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 | |
| } | |