| { | |
| "best_metric": 0.3037484288215637, | |
| "best_model_checkpoint": "./ryan_model3272024/checkpoint-250", | |
| "epoch": 1.4627659574468086, | |
| "eval_steps": 25, | |
| "global_step": 275, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.5357832312583923, | |
| "learning_rate": 0.00019867021276595746, | |
| "loss": 0.4838, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.3301275372505188, | |
| "learning_rate": 0.00019734042553191489, | |
| "loss": 0.3457, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.829723060131073, | |
| "learning_rate": 0.00019601063829787234, | |
| "loss": 0.4508, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.43481916189193726, | |
| "learning_rate": 0.00019468085106382982, | |
| "loss": 0.3205, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.9304245114326477, | |
| "learning_rate": 0.00019335106382978724, | |
| "loss": 0.4062, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_loss": 0.3798711597919464, | |
| "eval_na_accuracy": 0.6216216087341309, | |
| "eval_ordinal_accuracy": 0.23954372107982635, | |
| "eval_ordinal_mae": 0.9244347810745239, | |
| "eval_runtime": 31.9197, | |
| "eval_samples_per_second": 9.399, | |
| "eval_steps_per_second": 1.19, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.8552330136299133, | |
| "learning_rate": 0.0001920212765957447, | |
| "loss": 0.3029, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.43845850229263306, | |
| "learning_rate": 0.00019069148936170214, | |
| "loss": 0.3261, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.18216854333877563, | |
| "learning_rate": 0.00018936170212765957, | |
| "loss": 0.2895, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.7537259459495544, | |
| "learning_rate": 0.00018829787234042554, | |
| "loss": 0.4943, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.4183604419231415, | |
| "learning_rate": 0.000186968085106383, | |
| "loss": 0.3536, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_loss": 0.3699643015861511, | |
| "eval_na_accuracy": 0.6756756901741028, | |
| "eval_ordinal_accuracy": 0.38403043150901794, | |
| "eval_ordinal_mae": 0.9066693782806396, | |
| "eval_runtime": 12.0769, | |
| "eval_samples_per_second": 24.841, | |
| "eval_steps_per_second": 3.147, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.8079000115394592, | |
| "learning_rate": 0.00018563829787234044, | |
| "loss": 0.3601, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.9893763065338135, | |
| "learning_rate": 0.0001843085106382979, | |
| "loss": 0.4152, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 1.3993595838546753, | |
| "learning_rate": 0.00018297872340425532, | |
| "loss": 0.448, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.45827603340148926, | |
| "learning_rate": 0.00018164893617021277, | |
| "loss": 0.3132, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.4677202701568604, | |
| "learning_rate": 0.00018031914893617022, | |
| "loss": 0.4295, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.34051504731178284, | |
| "eval_na_accuracy": 0.7837837934494019, | |
| "eval_ordinal_accuracy": 0.29657796025276184, | |
| "eval_ordinal_mae": 0.8797782063484192, | |
| "eval_runtime": 11.5094, | |
| "eval_samples_per_second": 26.066, | |
| "eval_steps_per_second": 3.302, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 1.1136772632598877, | |
| "learning_rate": 0.00017898936170212767, | |
| "loss": 0.2444, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 3.5127644538879395, | |
| "learning_rate": 0.00017765957446808512, | |
| "loss": 0.4542, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.6393898725509644, | |
| "learning_rate": 0.00017632978723404257, | |
| "loss": 0.3411, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.8198608160018921, | |
| "learning_rate": 0.000175, | |
| "loss": 0.3184, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.1194807291030884, | |
| "learning_rate": 0.00017367021276595745, | |
| "loss": 0.4114, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_loss": 0.39057785272598267, | |
| "eval_na_accuracy": 0.7297297120094299, | |
| "eval_ordinal_accuracy": 0.35361215472221375, | |
| "eval_ordinal_mae": 0.8806185126304626, | |
| "eval_runtime": 12.2664, | |
| "eval_samples_per_second": 24.457, | |
| "eval_steps_per_second": 3.098, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 3.365068197250366, | |
| "learning_rate": 0.0001723404255319149, | |
| "loss": 0.469, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.6197894215583801, | |
| "learning_rate": 0.00017101063829787233, | |
| "loss": 0.2738, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.5341880917549133, | |
| "learning_rate": 0.0001696808510638298, | |
| "loss": 0.2867, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.0934885740280151, | |
| "learning_rate": 0.00016835106382978726, | |
| "loss": 0.4105, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.5764520764350891, | |
| "learning_rate": 0.00016702127659574468, | |
| "loss": 0.3521, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "eval_loss": 0.35300251841545105, | |
| "eval_na_accuracy": 0.8108108043670654, | |
| "eval_ordinal_accuracy": 0.42585551738739014, | |
| "eval_ordinal_mae": 0.8441764116287231, | |
| "eval_runtime": 12.3015, | |
| "eval_samples_per_second": 24.387, | |
| "eval_steps_per_second": 3.089, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.2090584933757782, | |
| "learning_rate": 0.00016569148936170213, | |
| "loss": 0.294, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.2995198369026184, | |
| "learning_rate": 0.00016436170212765958, | |
| "loss": 0.3099, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.41820451617240906, | |
| "learning_rate": 0.00016303191489361703, | |
| "loss": 0.4392, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.1886084079742432, | |
| "learning_rate": 0.00016170212765957446, | |
| "loss": 0.3725, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.8490511178970337, | |
| "learning_rate": 0.00016037234042553194, | |
| "loss": 0.3349, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.34123122692108154, | |
| "eval_na_accuracy": 0.7297297120094299, | |
| "eval_ordinal_accuracy": 0.4752851724624634, | |
| "eval_ordinal_mae": 0.8015652298927307, | |
| "eval_runtime": 12.509, | |
| "eval_samples_per_second": 23.983, | |
| "eval_steps_per_second": 3.038, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.4098907709121704, | |
| "learning_rate": 0.00015904255319148936, | |
| "loss": 0.2306, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 1.5454349517822266, | |
| "learning_rate": 0.0001577127659574468, | |
| "loss": 0.2382, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.61043381690979, | |
| "learning_rate": 0.00015638297872340426, | |
| "loss": 0.3448, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.7741652727127075, | |
| "learning_rate": 0.00015505319148936171, | |
| "loss": 0.2037, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.5108156204223633, | |
| "learning_rate": 0.00015372340425531914, | |
| "loss": 0.4612, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_loss": 0.36386463046073914, | |
| "eval_na_accuracy": 0.5405405163764954, | |
| "eval_ordinal_accuracy": 0.46768060326576233, | |
| "eval_ordinal_mae": 0.7603853940963745, | |
| "eval_runtime": 12.3807, | |
| "eval_samples_per_second": 24.231, | |
| "eval_steps_per_second": 3.069, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.8889250159263611, | |
| "learning_rate": 0.00015239361702127662, | |
| "loss": 0.4401, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.047706127166748, | |
| "learning_rate": 0.00015106382978723407, | |
| "loss": 0.2771, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 0.6727350354194641, | |
| "learning_rate": 0.0001497340425531915, | |
| "loss": 0.3726, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.4043642282485962, | |
| "learning_rate": 0.00014840425531914894, | |
| "loss": 0.3504, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 1.089250922203064, | |
| "learning_rate": 0.0001470744680851064, | |
| "loss": 0.2424, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "eval_loss": 0.32972484827041626, | |
| "eval_na_accuracy": 0.7027027010917664, | |
| "eval_ordinal_accuracy": 0.48669201135635376, | |
| "eval_ordinal_mae": 0.7117426991462708, | |
| "eval_runtime": 11.8333, | |
| "eval_samples_per_second": 25.352, | |
| "eval_steps_per_second": 3.211, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 0.6453298926353455, | |
| "learning_rate": 0.00014574468085106382, | |
| "loss": 0.2115, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.7021524310112, | |
| "learning_rate": 0.00014441489361702127, | |
| "loss": 0.2281, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.7665510773658752, | |
| "learning_rate": 0.00014308510638297875, | |
| "loss": 0.2048, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 1.2339574098587036, | |
| "learning_rate": 0.00014175531914893617, | |
| "loss": 0.2344, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 2.540107011795044, | |
| "learning_rate": 0.00014042553191489363, | |
| "loss": 0.2928, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.3493916690349579, | |
| "eval_na_accuracy": 0.6756756901741028, | |
| "eval_ordinal_accuracy": 0.5285171270370483, | |
| "eval_ordinal_mae": 0.6955077052116394, | |
| "eval_runtime": 10.8667, | |
| "eval_samples_per_second": 27.607, | |
| "eval_steps_per_second": 3.497, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.7262032628059387, | |
| "learning_rate": 0.00013909574468085108, | |
| "loss": 0.3973, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.28402724862098694, | |
| "learning_rate": 0.0001377659574468085, | |
| "loss": 0.218, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.5170676708221436, | |
| "learning_rate": 0.00013643617021276595, | |
| "loss": 0.227, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.35739636421203613, | |
| "learning_rate": 0.0001351063829787234, | |
| "loss": 0.2275, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.5471745133399963, | |
| "learning_rate": 0.00013377659574468086, | |
| "loss": 0.2436, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "eval_loss": 0.3037484288215637, | |
| "eval_na_accuracy": 0.7297297120094299, | |
| "eval_ordinal_accuracy": 0.5285171270370483, | |
| "eval_ordinal_mae": 0.6723113059997559, | |
| "eval_runtime": 12.062, | |
| "eval_samples_per_second": 24.871, | |
| "eval_steps_per_second": 3.15, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 0.41173043847084045, | |
| "learning_rate": 0.0001324468085106383, | |
| "loss": 0.2408, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.8529615998268127, | |
| "learning_rate": 0.00013111702127659576, | |
| "loss": 0.2935, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 1.3896653652191162, | |
| "learning_rate": 0.00012978723404255318, | |
| "loss": 0.2547, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.30819597840309143, | |
| "learning_rate": 0.00012845744680851063, | |
| "loss": 0.3372, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 2.0342652797698975, | |
| "learning_rate": 0.00012712765957446809, | |
| "loss": 0.2776, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "eval_loss": 0.3365646004676819, | |
| "eval_na_accuracy": 0.5945945978164673, | |
| "eval_ordinal_accuracy": 0.517110288143158, | |
| "eval_ordinal_mae": 0.672748863697052, | |
| "eval_runtime": 11.4536, | |
| "eval_samples_per_second": 26.193, | |
| "eval_steps_per_second": 3.318, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "step": 275, | |
| "total_flos": 3.403570199991091e+17, | |
| "train_loss": 0.32911812565543436, | |
| "train_runtime": 761.2747, | |
| "train_samples_per_second": 15.763, | |
| "train_steps_per_second": 0.988 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 752, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 25, | |
| "total_flos": 3.403570199991091e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |