| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 3.5397530864197533, |
| "eval_steps": 1024, |
| "global_step": 7168, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.06320987654320988, |
| "grad_norm": 0.724014937877655, |
| "learning_rate": 0.000248046875, |
| "loss": 0.838069498538971, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.12641975308641976, |
| "grad_norm": 0.704939603805542, |
| "learning_rate": 0.000498046875, |
| "loss": 0.4357231557369232, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.18962962962962962, |
| "grad_norm": 1.2203071117401123, |
| "learning_rate": 0.000748046875, |
| "loss": 0.26281777024269104, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.2528395061728395, |
| "grad_norm": 1.0112401247024536, |
| "learning_rate": 0.000998046875, |
| "loss": 0.20087672770023346, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.3160493827160494, |
| "grad_norm": 0.7569617629051208, |
| "learning_rate": 0.0009993089770195807, |
| "loss": 0.18235813081264496, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.37925925925925924, |
| "grad_norm": 0.507542073726654, |
| "learning_rate": 0.0009972160460972733, |
| "loss": 0.1611708402633667, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.44246913580246916, |
| "grad_norm": 0.17671959102153778, |
| "learning_rate": 0.0009937270408736224, |
| "loss": 0.151814267039299, |
| "step": 896 |
| }, |
| { |
| "epoch": 0.505679012345679, |
| "grad_norm": 0.1602219045162201, |
| "learning_rate": 0.0009888517577149526, |
| "loss": 0.15154045820236206, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.505679012345679, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03877834157972801, |
| "eval_loss": 0.14963696024766782, |
| "eval_mse_loss": 0.14963696024766782, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.505679012345679, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03877834157972801, |
| "eval_loss": 0.14963696024766782, |
| "eval_mse_loss": 0.14963696024766782, |
| "eval_runtime": 17.2641, |
| "eval_samples_per_second": 151.644, |
| "eval_steps_per_second": 2.375, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.5688888888888889, |
| "grad_norm": 0.08139930665493011, |
| "learning_rate": 0.0009826038853539248, |
| "loss": 0.15169958770275116, |
| "step": 1152 |
| }, |
| { |
| "epoch": 0.6320987654320988, |
| "grad_norm": 0.09654489159584045, |
| "learning_rate": 0.0009750009664545572, |
| "loss": 0.14849385619163513, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.6953086419753086, |
| "grad_norm": 0.05821019783616066, |
| "learning_rate": 0.0009660643483562486, |
| "loss": 0.15167970955371857, |
| "step": 1408 |
| }, |
| { |
| "epoch": 0.7585185185185185, |
| "grad_norm": 0.09790857881307602, |
| "learning_rate": 0.0009558191231351013, |
| "loss": 0.15071730315685272, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.8217283950617283, |
| "grad_norm": 0.09251409024000168, |
| "learning_rate": 0.0009442940571508399, |
| "loss": 0.1502133160829544, |
| "step": 1664 |
| }, |
| { |
| "epoch": 0.8849382716049383, |
| "grad_norm": 0.051930081099271774, |
| "learning_rate": 0.0009315215102771411, |
| "loss": 0.14858268201351166, |
| "step": 1792 |
| }, |
| { |
| "epoch": 0.9481481481481482, |
| "grad_norm": 0.08928167074918747, |
| "learning_rate": 0.0009175373450421618, |
| "loss": 0.15144191682338715, |
| "step": 1920 |
| }, |
| { |
| "epoch": 1.011358024691358, |
| "grad_norm": 0.056316301226615906, |
| "learning_rate": 0.0009023808259343743, |
| "loss": 0.14895020425319672, |
| "step": 2048 |
| }, |
| { |
| "epoch": 1.011358024691358, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03770822227546355, |
| "eval_loss": 0.1449529039787083, |
| "eval_mse_loss": 0.1449529039787083, |
| "step": 2048 |
| }, |
| { |
| "epoch": 1.011358024691358, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03770822227546355, |
| "eval_loss": 0.1449529039787083, |
| "eval_mse_loss": 0.1449529039787083, |
| "eval_runtime": 17.1022, |
| "eval_samples_per_second": 153.08, |
| "eval_steps_per_second": 2.397, |
| "step": 2048 |
| }, |
| { |
| "epoch": 1.074567901234568, |
| "grad_norm": 0.047452766448259354, |
| "learning_rate": 0.0008860945091564399, |
| "loss": 0.1469879448413849, |
| "step": 2176 |
| }, |
| { |
| "epoch": 1.1377777777777778, |
| "grad_norm": 0.07337811589241028, |
| "learning_rate": 0.0008687241231366662, |
| "loss": 0.14955241978168488, |
| "step": 2304 |
| }, |
| { |
| "epoch": 1.2009876543209876, |
| "grad_norm": 0.0511033795773983, |
| "learning_rate": 0.0008503184401335448, |
| "loss": 0.15107223391532898, |
| "step": 2432 |
| }, |
| { |
| "epoch": 1.2641975308641975, |
| "grad_norm": 0.12204128503799438, |
| "learning_rate": 0.0008309291392938795, |
| "loss": 0.1506786197423935, |
| "step": 2560 |
| }, |
| { |
| "epoch": 1.3274074074074074, |
| "grad_norm": 0.08734283596277237, |
| "learning_rate": 0.0008106106615490032, |
| "loss": 0.15213392674922943, |
| "step": 2688 |
| }, |
| { |
| "epoch": 1.3906172839506172, |
| "grad_norm": 0.10540172457695007, |
| "learning_rate": 0.0007894200567565075, |
| "loss": 0.15196450054645538, |
| "step": 2816 |
| }, |
| { |
| "epoch": 1.453827160493827, |
| "grad_norm": 0.0695379450917244, |
| "learning_rate": 0.0007674168235166747, |
| "loss": 0.14914581179618835, |
| "step": 2944 |
| }, |
| { |
| "epoch": 1.5170370370370372, |
| "grad_norm": 0.07360924035310745, |
| "learning_rate": 0.000744662742113375, |
| "loss": 0.15044452250003815, |
| "step": 3072 |
| }, |
| { |
| "epoch": 1.5170370370370372, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03918857636248193, |
| "eval_loss": 0.15137197622438756, |
| "eval_mse_loss": 0.15137197622438756, |
| "step": 3072 |
| }, |
| { |
| "epoch": 1.5170370370370372, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03918857636248193, |
| "eval_loss": 0.15137197622438756, |
| "eval_mse_loss": 0.15137197622438756, |
| "eval_runtime": 17.0045, |
| "eval_samples_per_second": 153.96, |
| "eval_steps_per_second": 2.411, |
| "step": 3072 |
| }, |
| { |
| "epoch": 1.5802469135802468, |
| "grad_norm": 0.10566578805446625, |
| "learning_rate": 0.0007212217010484917, |
| "loss": 0.15180513262748718, |
| "step": 3200 |
| }, |
| { |
| "epoch": 1.643456790123457, |
| "grad_norm": 0.056859929114580154, |
| "learning_rate": 0.0006971595176569332, |
| "loss": 0.14987537264823914, |
| "step": 3328 |
| }, |
| { |
| "epoch": 1.7066666666666666, |
| "grad_norm": 0.042681753635406494, |
| "learning_rate": 0.0006725437533058971, |
| "loss": 0.14968286454677582, |
| "step": 3456 |
| }, |
| { |
| "epoch": 1.7698765432098766, |
| "grad_norm": 0.0382704995572567, |
| "learning_rate": 0.0006474435236972767, |
| "loss": 0.15063193440437317, |
| "step": 3584 |
| }, |
| { |
| "epoch": 1.8330864197530863, |
| "grad_norm": 0.04770047217607498, |
| "learning_rate": 0.0006219293048058301, |
| "loss": 0.14851045608520508, |
| "step": 3712 |
| }, |
| { |
| "epoch": 1.8962962962962964, |
| "grad_norm": 0.05498025193810463, |
| "learning_rate": 0.0005960727349980042, |
| "loss": 0.14856384694576263, |
| "step": 3840 |
| }, |
| { |
| "epoch": 1.959506172839506, |
| "grad_norm": 0.03446037694811821, |
| "learning_rate": 0.0005699464138870106, |
| "loss": 0.14929655194282532, |
| "step": 3968 |
| }, |
| { |
| "epoch": 2.022716049382716, |
| "grad_norm": 0.03936333954334259, |
| "learning_rate": 0.0005436236984889356, |
| "loss": 0.14908051490783691, |
| "step": 4096 |
| }, |
| { |
| "epoch": 2.022716049382716, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.0382816258181886, |
| "eval_loss": 0.14783194661140442, |
| "eval_mse_loss": 0.14783194661140442, |
| "step": 4096 |
| }, |
| { |
| "epoch": 2.022716049382716, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.0382816258181886, |
| "eval_loss": 0.14783194661140442, |
| "eval_mse_loss": 0.14783194661140442, |
| "eval_runtime": 16.9592, |
| "eval_samples_per_second": 154.37, |
| "eval_steps_per_second": 2.418, |
| "step": 4096 |
| }, |
| { |
| "epoch": 2.0859259259259257, |
| "grad_norm": 0.044177353382110596, |
| "learning_rate": 0.0005171784972522235, |
| "loss": 0.15089942514896393, |
| "step": 4224 |
| }, |
| { |
| "epoch": 2.149135802469136, |
| "grad_norm": 0.0455547496676445, |
| "learning_rate": 0.0004906850625388614, |
| "loss": 0.15073473751544952, |
| "step": 4352 |
| }, |
| { |
| "epoch": 2.212345679012346, |
| "grad_norm": 0.03271918743848801, |
| "learning_rate": 0.0004642177821399269, |
| "loss": 0.1500038355588913, |
| "step": 4480 |
| }, |
| { |
| "epoch": 2.2755555555555556, |
| "grad_norm": 0.03396615758538246, |
| "learning_rate": 0.00043785097041088293, |
| "loss": 0.1482856571674347, |
| "step": 4608 |
| }, |
| { |
| "epoch": 2.338765432098765, |
| "grad_norm": 0.023670442402362823, |
| "learning_rate": 0.00041165865961306135, |
| "loss": 0.1509697139263153, |
| "step": 4736 |
| }, |
| { |
| "epoch": 2.4019753086419753, |
| "grad_norm": 0.02988285757601261, |
| "learning_rate": 0.000385714392047208, |
| "loss": 0.149857297539711, |
| "step": 4864 |
| }, |
| { |
| "epoch": 2.4651851851851854, |
| "grad_norm": 0.01895061321556568, |
| "learning_rate": 0.00036009101356272814, |
| "loss": 0.1492118537425995, |
| "step": 4992 |
| }, |
| { |
| "epoch": 2.528395061728395, |
| "grad_norm": 0.06466109305620193, |
| "learning_rate": 0.00033486046902241664, |
| "loss": 0.1494925320148468, |
| "step": 5120 |
| }, |
| { |
| "epoch": 2.528395061728395, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03705163218262719, |
| "eval_loss": 0.14231552746964665, |
| "eval_mse_loss": 0.14231552746964665, |
| "step": 5120 |
| }, |
| { |
| "epoch": 2.528395061728395, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03705163218262719, |
| "eval_loss": 0.14231552746964665, |
| "eval_mse_loss": 0.14231552746964665, |
| "eval_runtime": 17.0116, |
| "eval_samples_per_second": 153.895, |
| "eval_steps_per_second": 2.41, |
| "step": 5120 |
| }, |
| { |
| "epoch": 2.591604938271605, |
| "grad_norm": 0.03147516027092934, |
| "learning_rate": 0.00031009360029696107, |
| "loss": 0.1494501531124115, |
| "step": 5248 |
| }, |
| { |
| "epoch": 2.6548148148148147, |
| "grad_norm": 0.01734239235520363, |
| "learning_rate": 0.00028585994735640823, |
| "loss": 0.1486799716949463, |
| "step": 5376 |
| }, |
| { |
| "epoch": 2.718024691358025, |
| "grad_norm": 0.03119756653904915, |
| "learning_rate": 0.0002622275530170825, |
| "loss": 0.1488998383283615, |
| "step": 5504 |
| }, |
| { |
| "epoch": 2.7812345679012345, |
| "grad_norm": 0.029683848842978477, |
| "learning_rate": 0.0002392627718921852, |
| "loss": 0.14848105609416962, |
| "step": 5632 |
| }, |
| { |
| "epoch": 2.8444444444444446, |
| "grad_norm": 0.029232973232865334, |
| "learning_rate": 0.00021703008408250187, |
| "loss": 0.14881552755832672, |
| "step": 5760 |
| }, |
| { |
| "epoch": 2.907654320987654, |
| "grad_norm": 0.025576096028089523, |
| "learning_rate": 0.00019559191413033017, |
| "loss": 0.14923110604286194, |
| "step": 5888 |
| }, |
| { |
| "epoch": 2.9708641975308643, |
| "grad_norm": 0.05523916333913803, |
| "learning_rate": 0.00017500845574497032, |
| "loss": 0.1484050154685974, |
| "step": 6016 |
| }, |
| { |
| "epoch": 3.034074074074074, |
| "grad_norm": 0.02349747158586979, |
| "learning_rate": 0.00015533750279190617, |
| "loss": 0.14911824464797974, |
| "step": 6144 |
| }, |
| { |
| "epoch": 3.034074074074074, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03870651475721743, |
| "eval_loss": 0.14876795905392345, |
| "eval_mse_loss": 0.14876795905392345, |
| "step": 6144 |
| }, |
| { |
| "epoch": 3.034074074074074, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03870651475721743, |
| "eval_loss": 0.14876795905392345, |
| "eval_mse_loss": 0.14876795905392345, |
| "eval_runtime": 18.1257, |
| "eval_samples_per_second": 144.436, |
| "eval_steps_per_second": 2.262, |
| "step": 6144 |
| }, |
| { |
| "epoch": 3.097283950617284, |
| "grad_norm": 0.04030594229698181, |
| "learning_rate": 0.000136634287020226, |
| "loss": 0.14769518375396729, |
| "step": 6272 |
| }, |
| { |
| "epoch": 3.1604938271604937, |
| "grad_norm": 0.027767272666096687, |
| "learning_rate": 0.00011895132298390743, |
| "loss": 0.1493716984987259, |
| "step": 6400 |
| }, |
| { |
| "epoch": 3.2237037037037037, |
| "grad_norm": 0.04019133746623993, |
| "learning_rate": 0.00010233826059239426, |
| "loss": 0.15157844126224518, |
| "step": 6528 |
| }, |
| { |
| "epoch": 3.286913580246914, |
| "grad_norm": 0.01638754829764366, |
| "learning_rate": 8.68417457044704e-05, |
| "loss": 0.14995069801807404, |
| "step": 6656 |
| }, |
| { |
| "epoch": 3.3501234567901235, |
| "grad_norm": 0.038517311215400696, |
| "learning_rate": 7.250528915685422e-05, |
| "loss": 0.15167345106601715, |
| "step": 6784 |
| }, |
| { |
| "epoch": 3.413333333333333, |
| "grad_norm": 0.01678868569433689, |
| "learning_rate": 5.93691445952505e-05, |
| "loss": 0.14608222246170044, |
| "step": 6912 |
| }, |
| { |
| "epoch": 3.476543209876543, |
| "grad_norm": 0.015662306919693947, |
| "learning_rate": 4.7470195450886376e-05, |
| "loss": 0.1496647447347641, |
| "step": 7040 |
| }, |
| { |
| "epoch": 3.5397530864197533, |
| "grad_norm": 0.014813782647252083, |
| "learning_rate": 3.684185137987378e-05, |
| "loss": 0.14931762218475342, |
| "step": 7168 |
| }, |
| { |
| "epoch": 3.5397530864197533, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03843161172983123, |
| "eval_loss": 0.14863424693665853, |
| "eval_mse_loss": 0.14863424693665853, |
| "step": 7168 |
| }, |
| { |
| "epoch": 3.5397530864197533, |
| "eval_bleu": 0.0, |
| "eval_cos_loss": 0.03843161172983123, |
| "eval_loss": 0.14863424693665853, |
| "eval_mse_loss": 0.14863424693665853, |
| "eval_runtime": 16.9602, |
| "eval_samples_per_second": 154.362, |
| "eval_steps_per_second": 2.417, |
| "step": 7168 |
| } |
| ], |
| "logging_steps": 128, |
| "max_steps": 8100, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 4, |
| "save_steps": 1024, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 64, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
