| { | |
| "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": 5.434233784386483, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 5.3587, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 4.905477378648474, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 5.1051, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 5.270930514694448, | |
| "learning_rate": 6.060606060606061e-05, | |
| "loss": 3.4357, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.3618606789587138, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.3624, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.06941498462106709, | |
| "learning_rate": 0.00012121212121212122, | |
| "loss": 0.0843, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.04064697552867018, | |
| "learning_rate": 0.00015151515151515152, | |
| "loss": 0.0289, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.009288632165411351, | |
| "learning_rate": 0.00018181818181818183, | |
| "loss": 0.0178, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.006293796326369993, | |
| "learning_rate": 0.00019997685019798912, | |
| "loss": 0.0149, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.1906256952245856, | |
| "learning_rate": 0.0001997165380022878, | |
| "loss": 0.0132, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.0017591555466765156, | |
| "learning_rate": 0.000199167731989929, | |
| "loss": 0.01, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.0013071409318138017, | |
| "learning_rate": 0.0001983320199330545, | |
| "loss": 0.0082, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.0009433621765961648, | |
| "learning_rate": 0.00019721181966290613, | |
| "loss": 0.0067, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.0005177333796827652, | |
| "learning_rate": 0.00019581037207470382, | |
| "loss": 0.0057, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.00039250212435773405, | |
| "learning_rate": 0.00019413173175128473, | |
| "loss": 0.0049, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.00029772481292231705, | |
| "learning_rate": 0.00019218075523263104, | |
| "loss": 0.0044, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.0002507402780542911, | |
| "learning_rate": 0.00018996308696522433, | |
| "loss": 0.0729, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.0005904602103990053, | |
| "learning_rate": 0.00018748514297187648, | |
| "loss": 0.0058, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.000979792833537355, | |
| "learning_rate": 0.00018475409228928312, | |
| "loss": 0.0084, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.0015160678160695531, | |
| "learning_rate": 0.00018177783622700327, | |
| "loss": 0.0104, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.001483201059583737, | |
| "learning_rate": 0.00017856498550787144, | |
| "loss": 0.0117, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.0016699269860680358, | |
| "learning_rate": 0.00017512483535597867, | |
| "loss": 0.0117, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.0013065938174500955, | |
| "learning_rate": 0.00017146733860429612, | |
| "loss": 0.0113, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.001108648512762412, | |
| "learning_rate": 0.0001676030768997445, | |
| "loss": 0.0107, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.0008817246126042861, | |
| "learning_rate": 0.00016354323008901776, | |
| "loss": 0.0099, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.0007765931672905603, | |
| "learning_rate": 0.00015929954387373103, | |
| "loss": 0.0092, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.0008313798889983195, | |
| "learning_rate": 0.00015488429582847192, | |
| "loss": 0.0526, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.0011465080758489292, | |
| "learning_rate": 0.00015031025988006936, | |
| "loss": 0.0106, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.0015289659191642599, | |
| "learning_rate": 0.00014559066935084588, | |
| "loss": 0.0113, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.001431377139280294, | |
| "learning_rate": 0.00014073917867277557, | |
| "loss": 0.0119, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.0007764760432382378, | |
| "learning_rate": 0.0001357698238833126, | |
| "loss": 0.0108, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.000929883831003457, | |
| "learning_rate": 0.000130696982017182, | |
| "loss": 0.0107, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.0007351401317411369, | |
| "learning_rate": 0.0001255353295116187, | |
| "loss": 0.0101, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.0005848350734757287, | |
| "learning_rate": 0.00012029979974539234, | |
| "loss": 0.0095, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.0006593756085078484, | |
| "learning_rate": 0.00011500553983446527, | |
| "loss": 0.0093, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.00046311633397912643, | |
| "learning_rate": 0.00010966786680927874, | |
| "loss": 0.0084, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0004798326957832526, | |
| "learning_rate": 0.00010430222330045304, | |
| "loss": 0.0081, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.0005200079230795159, | |
| "learning_rate": 9.892413286110886e-05, | |
| "loss": 0.0524, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0011293522137763314, | |
| "learning_rate": 9.354915505506839e-05, | |
| "loss": 0.0093, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.0020109230697902976, | |
| "learning_rate": 8.81928404408726e-05, | |
| "loss": 0.0106, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.0020295886614259714, | |
| "learning_rate": 8.287068558185225e-05, | |
| "loss": 0.0115, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.013880677459645187, | |
| "learning_rate": 7.759808821241406e-05, | |
| "loss": 0.0429, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.025579316344818064, | |
| "learning_rate": 7.239030269025311e-05, | |
| "loss": 0.0227, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.00587547091852476, | |
| "learning_rate": 6.726239586337408e-05, | |
| "loss": 0.0173, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.001168548414346044, | |
| "learning_rate": 6.22292034796035e-05, | |
| "loss": 0.0124, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.0006254826490570403, | |
| "learning_rate": 5.730528726470792e-05, | |
| "loss": 0.0104, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.0009469247245448829, | |
| "learning_rate": 5.2504892793295e-05, | |
| "loss": 0.0096, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.000538986457963209, | |
| "learning_rate": 4.7841908274384616e-05, | |
| "loss": 0.0088, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.0005138624078734311, | |
| "learning_rate": 4.332982437088825e-05, | |
| "loss": 0.0087, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.0004754475990403913, | |
| "learning_rate": 3.898169516924398e-05, | |
| "loss": 0.0087, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.00044235741477306284, | |
| "learning_rate": 3.4810100412128747e-05, | |
| "loss": 0.0081, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.0005356812279608025, | |
| "learning_rate": 3.0827109103512643e-05, | |
| "loss": 0.0084, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.0004866021061182579, | |
| "learning_rate": 2.7044244591351232e-05, | |
| "loss": 0.0082, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.0005541072776907527, | |
| "learning_rate": 2.3472451228937253e-05, | |
| "loss": 0.0081, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.0004669324029460342, | |
| "learning_rate": 2.0122062711363532e-05, | |
| "loss": 0.008, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.0004222921215236744, | |
| "learning_rate": 1.7002772178705716e-05, | |
| "loss": 0.0081, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.00037935649836293145, | |
| "learning_rate": 1.4123604172419713e-05, | |
| "loss": 0.0078, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.00041870421442444797, | |
| "learning_rate": 1.149288852608743e-05, | |
| "loss": 0.0077, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.00044398149572571745, | |
| "learning_rate": 9.118236266049707e-06, | |
| "loss": 0.0079, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.0004263492971348259, | |
| "learning_rate": 7.0065175916482095e-06, | |
| "loss": 0.0077, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.00044499020775488917, | |
| "learning_rate": 5.163841998782837e-06, | |
| "loss": 0.0079, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.0003979573867652517, | |
| "learning_rate": 3.595540604290437e-06, | |
| "loss": 0.0078, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.0004200122969057789, | |
| "learning_rate": 2.30615072228183e-06, | |
| "loss": 0.0078, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.00039339735065362075, | |
| "learning_rate": 1.2994027370611173e-06, | |
| "loss": 0.0076, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.00043214757348710424, | |
| "learning_rate": 5.782093106048159e-07, | |
| "loss": 0.0077, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.00040109280842680117, | |
| "learning_rate": 1.446569558255395e-07, | |
| "loss": 0.0259, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.00039332444010680133, | |
| "learning_rate": 0.0, | |
| "loss": 0.0076, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 747.8903, | |
| "eval_samples_per_second": 1.546, | |
| "eval_steps_per_second": 0.386, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 325, | |
| "total_flos": 5880639324684288.0, | |
| "train_loss": 0.15147324924285596, | |
| "train_runtime": 6121.0939, | |
| "train_samples_per_second": 1.698, | |
| "train_steps_per_second": 0.053 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 325, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 40, | |
| "total_flos": 5880639324684288.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |