| { | |
| "best_metric": 1.4602320194244385, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-800", | |
| "epoch": 0.8365492344080667, | |
| "eval_steps": 100, | |
| "global_step": 840, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.000995891945723889, | |
| "grad_norm": 16.56658172607422, | |
| "learning_rate": 2.5e-05, | |
| "loss": 27.5358, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.000995891945723889, | |
| "eval_loss": 3.4175312519073486, | |
| "eval_runtime": 287.5927, | |
| "eval_samples_per_second": 4.656, | |
| "eval_steps_per_second": 1.165, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.001991783891447778, | |
| "grad_norm": 16.847103118896484, | |
| "learning_rate": 5e-05, | |
| "loss": 28.6136, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002987675837171667, | |
| "grad_norm": 14.910314559936523, | |
| "learning_rate": 7.5e-05, | |
| "loss": 26.2465, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.003983567782895556, | |
| "grad_norm": 14.995889663696289, | |
| "learning_rate": 0.0001, | |
| "loss": 23.7459, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.004979459728619445, | |
| "grad_norm": 18.5001220703125, | |
| "learning_rate": 0.000125, | |
| "loss": 24.4332, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005975351674343334, | |
| "grad_norm": 13.621232986450195, | |
| "learning_rate": 0.00015, | |
| "loss": 22.1308, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.006971243620067223, | |
| "grad_norm": 12.984804153442383, | |
| "learning_rate": 0.000175, | |
| "loss": 19.3733, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.007967135565791112, | |
| "grad_norm": 13.488744735717773, | |
| "learning_rate": 0.0002, | |
| "loss": 16.8823, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.008963027511515, | |
| "grad_norm": 16.1417293548584, | |
| "learning_rate": 0.00022500000000000002, | |
| "loss": 16.2773, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.00995891945723889, | |
| "grad_norm": 13.883444786071777, | |
| "learning_rate": 0.00025, | |
| "loss": 13.4328, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.010954811402962778, | |
| "grad_norm": 21.88507843017578, | |
| "learning_rate": 0.00024999910458769255, | |
| "loss": 15.1014, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.011950703348686668, | |
| "grad_norm": 18.264280319213867, | |
| "learning_rate": 0.0002499964183635983, | |
| "loss": 13.3915, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.012946595294410557, | |
| "grad_norm": 21.488842010498047, | |
| "learning_rate": 0.0002499919413662018, | |
| "loss": 14.5928, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.013942487240134445, | |
| "grad_norm": 11.669401168823242, | |
| "learning_rate": 0.00024998567365964314, | |
| "loss": 13.6025, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.014938379185858334, | |
| "grad_norm": 12.027138710021973, | |
| "learning_rate": 0.0002499776153337172, | |
| "loss": 14.022, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.015934271131582224, | |
| "grad_norm": 13.013602256774902, | |
| "learning_rate": 0.00024996776650387245, | |
| "loss": 13.9936, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01693016307730611, | |
| "grad_norm": 11.030466079711914, | |
| "learning_rate": 0.000249956127311209, | |
| "loss": 13.786, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.01792605502303, | |
| "grad_norm": 10.779586791992188, | |
| "learning_rate": 0.000249942697922477, | |
| "loss": 12.499, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.01892194696875389, | |
| "grad_norm": 11.274680137634277, | |
| "learning_rate": 0.00024992747853007374, | |
| "loss": 12.219, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.01991783891447778, | |
| "grad_norm": 9.957880020141602, | |
| "learning_rate": 0.00024991046935204144, | |
| "loss": 13.1162, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02091373086020167, | |
| "grad_norm": 8.227347373962402, | |
| "learning_rate": 0.0002498916706320637, | |
| "loss": 13.0072, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.021909622805925556, | |
| "grad_norm": 7.3595051765441895, | |
| "learning_rate": 0.00024987108263946215, | |
| "loss": 12.4328, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.022905514751649446, | |
| "grad_norm": 10.374775886535645, | |
| "learning_rate": 0.00024984870566919273, | |
| "loss": 13.644, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.023901406697373336, | |
| "grad_norm": 7.763638496398926, | |
| "learning_rate": 0.00024982454004184127, | |
| "loss": 12.0825, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.024897298643097223, | |
| "grad_norm": 8.642582893371582, | |
| "learning_rate": 0.0002497985861036189, | |
| "loss": 13.2585, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.025893190588821113, | |
| "grad_norm": 7.242372035980225, | |
| "learning_rate": 0.0002497708442263573, | |
| "loss": 11.6953, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.026889082534545, | |
| "grad_norm": 7.458141326904297, | |
| "learning_rate": 0.0002497413148075032, | |
| "loss": 12.0056, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02788497448026889, | |
| "grad_norm": 7.2677836418151855, | |
| "learning_rate": 0.0002497099982701126, | |
| "loss": 11.1914, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.02888086642599278, | |
| "grad_norm": 7.0768022537231445, | |
| "learning_rate": 0.0002496768950628449, | |
| "loss": 12.0455, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.029876758371716668, | |
| "grad_norm": 7.336495399475098, | |
| "learning_rate": 0.0002496420056599565, | |
| "loss": 13.927, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.030872650317440558, | |
| "grad_norm": 6.7654829025268555, | |
| "learning_rate": 0.00024960533056129374, | |
| "loss": 11.885, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.03186854226316445, | |
| "grad_norm": 7.940896511077881, | |
| "learning_rate": 0.000249566870292286, | |
| "loss": 13.0075, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.032864434208888335, | |
| "grad_norm": 6.816346168518066, | |
| "learning_rate": 0.000249526625403938, | |
| "loss": 12.771, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.03386032615461222, | |
| "grad_norm": 6.953782081604004, | |
| "learning_rate": 0.0002494845964728221, | |
| "loss": 13.2487, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.034856218100336116, | |
| "grad_norm": 8.73217487335205, | |
| "learning_rate": 0.0002494407841010699, | |
| "loss": 13.9326, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03585211004606, | |
| "grad_norm": 6.887628078460693, | |
| "learning_rate": 0.0002493951889163634, | |
| "loss": 11.7964, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.03684800199178389, | |
| "grad_norm": 6.188755035400391, | |
| "learning_rate": 0.00024934781157192666, | |
| "loss": 12.3109, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03784389393750778, | |
| "grad_norm": 6.760220527648926, | |
| "learning_rate": 0.00024929865274651565, | |
| "loss": 12.6023, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03883978588323167, | |
| "grad_norm": 7.21213436126709, | |
| "learning_rate": 0.0002492477131444091, | |
| "loss": 13.9952, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.03983567782895556, | |
| "grad_norm": 7.4104485511779785, | |
| "learning_rate": 0.00024919499349539815, | |
| "loss": 13.8991, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04083156977467945, | |
| "grad_norm": 6.497596740722656, | |
| "learning_rate": 0.0002491404945547759, | |
| "loss": 12.7215, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.04182746172040334, | |
| "grad_norm": 7.254849433898926, | |
| "learning_rate": 0.0002490842171033268, | |
| "loss": 12.7342, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.042823353666127224, | |
| "grad_norm": 6.134991645812988, | |
| "learning_rate": 0.0002490261619473152, | |
| "loss": 11.3577, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.04381924561185111, | |
| "grad_norm": 6.685878276824951, | |
| "learning_rate": 0.0002489663299184738, | |
| "loss": 12.3739, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.044815137557575005, | |
| "grad_norm": 7.548392295837402, | |
| "learning_rate": 0.00024890472187399216, | |
| "loss": 12.5839, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.04581102950329889, | |
| "grad_norm": 6.434916973114014, | |
| "learning_rate": 0.00024884133869650376, | |
| "loss": 12.3762, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.04680692144902278, | |
| "grad_norm": 7.479133605957031, | |
| "learning_rate": 0.00024877618129407386, | |
| "loss": 13.4078, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.04780281339474667, | |
| "grad_norm": 6.191986083984375, | |
| "learning_rate": 0.00024870925060018633, | |
| "loss": 11.6391, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04879870534047056, | |
| "grad_norm": 6.074136257171631, | |
| "learning_rate": 0.00024864054757373023, | |
| "loss": 11.4169, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.049794597286194446, | |
| "grad_norm": 6.535557270050049, | |
| "learning_rate": 0.0002485700731989861, | |
| "loss": 12.7648, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05079048923191834, | |
| "grad_norm": 7.5086517333984375, | |
| "learning_rate": 0.0002484978284856119, | |
| "loss": 13.1785, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.05178638117764223, | |
| "grad_norm": 6.3925395011901855, | |
| "learning_rate": 0.00024842381446862856, | |
| "loss": 12.2323, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.052782273123366114, | |
| "grad_norm": 6.091940879821777, | |
| "learning_rate": 0.000248348032208405, | |
| "loss": 11.4875, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.05377816506909, | |
| "grad_norm": 6.079466342926025, | |
| "learning_rate": 0.0002482704827906432, | |
| "loss": 11.7412, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.054774057014813894, | |
| "grad_norm": 7.269871234893799, | |
| "learning_rate": 0.0002481911673263624, | |
| "loss": 13.4444, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05576994896053778, | |
| "grad_norm": 6.833920955657959, | |
| "learning_rate": 0.00024811008695188326, | |
| "loss": 13.2974, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.05676584090626167, | |
| "grad_norm": 6.509270668029785, | |
| "learning_rate": 0.0002480272428288116, | |
| "loss": 13.3969, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.05776173285198556, | |
| "grad_norm": 6.10461950302124, | |
| "learning_rate": 0.00024794263614402176, | |
| "loss": 13.0233, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.05875762479770945, | |
| "grad_norm": 6.248282432556152, | |
| "learning_rate": 0.0002478562681096397, | |
| "loss": 11.664, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.059753516743433335, | |
| "grad_norm": 5.46411657333374, | |
| "learning_rate": 0.0002477681399630253, | |
| "loss": 10.8694, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06074940868915723, | |
| "grad_norm": 6.836920738220215, | |
| "learning_rate": 0.00024767825296675516, | |
| "loss": 12.7564, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.061745300634881116, | |
| "grad_norm": 8.421375274658203, | |
| "learning_rate": 0.000247586608408604, | |
| "loss": 12.9136, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.06274119258060501, | |
| "grad_norm": 5.612268924713135, | |
| "learning_rate": 0.00024749320760152635, | |
| "loss": 11.2727, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.0637370845263289, | |
| "grad_norm": 7.760491371154785, | |
| "learning_rate": 0.00024739805188363803, | |
| "loss": 13.7674, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.06473297647205278, | |
| "grad_norm": 8.422126770019531, | |
| "learning_rate": 0.00024730114261819656, | |
| "loss": 12.9104, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06572886841777667, | |
| "grad_norm": 7.5595831871032715, | |
| "learning_rate": 0.0002472024811935821, | |
| "loss": 12.0906, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.06672476036350056, | |
| "grad_norm": 7.558265209197998, | |
| "learning_rate": 0.0002471020690232769, | |
| "loss": 11.7225, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.06772065230922444, | |
| "grad_norm": 8.700068473815918, | |
| "learning_rate": 0.00024699990754584584, | |
| "loss": 12.7639, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.06871654425494834, | |
| "grad_norm": 6.340944766998291, | |
| "learning_rate": 0.0002468959982249151, | |
| "loss": 12.6408, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.06971243620067223, | |
| "grad_norm": 5.689936637878418, | |
| "learning_rate": 0.0002467903425491517, | |
| "loss": 11.4365, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07070832814639612, | |
| "grad_norm": 6.519106864929199, | |
| "learning_rate": 0.00024668294203224184, | |
| "loss": 11.9108, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.07170422009212, | |
| "grad_norm": 5.775708198547363, | |
| "learning_rate": 0.00024657379821286954, | |
| "loss": 11.2813, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.07270011203784389, | |
| "grad_norm": 6.3621296882629395, | |
| "learning_rate": 0.00024646291265469425, | |
| "loss": 11.5422, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.07369600398356778, | |
| "grad_norm": 6.622462749481201, | |
| "learning_rate": 0.0002463502869463287, | |
| "loss": 12.9495, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.07469189592929167, | |
| "grad_norm": 7.1500163078308105, | |
| "learning_rate": 0.0002462359227013159, | |
| "loss": 13.1289, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.07568778787501557, | |
| "grad_norm": 6.552374362945557, | |
| "learning_rate": 0.0002461198215581064, | |
| "loss": 11.7408, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.07668367982073945, | |
| "grad_norm": 6.086798191070557, | |
| "learning_rate": 0.00024600198518003453, | |
| "loss": 11.8846, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.07767957176646334, | |
| "grad_norm": 7.924241065979004, | |
| "learning_rate": 0.00024588241525529445, | |
| "loss": 13.0962, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.07867546371218723, | |
| "grad_norm": 6.5235090255737305, | |
| "learning_rate": 0.0002457611134969164, | |
| "loss": 11.482, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.07967135565791111, | |
| "grad_norm": 6.921808242797852, | |
| "learning_rate": 0.0002456380816427417, | |
| "loss": 12.5718, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.080667247603635, | |
| "grad_norm": 6.6802077293396, | |
| "learning_rate": 0.0002455133214553981, | |
| "loss": 11.822, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.0816631395493589, | |
| "grad_norm": 6.431028366088867, | |
| "learning_rate": 0.0002453868347222746, | |
| "loss": 13.067, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.08265903149508279, | |
| "grad_norm": 6.637655735015869, | |
| "learning_rate": 0.0002452586232554956, | |
| "loss": 13.5535, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.08365492344080667, | |
| "grad_norm": 6.012635231018066, | |
| "learning_rate": 0.0002451286888918951, | |
| "loss": 12.0727, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.08465081538653056, | |
| "grad_norm": 6.7169508934021, | |
| "learning_rate": 0.00024499703349299034, | |
| "loss": 12.3954, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.08564670733225445, | |
| "grad_norm": 6.19224214553833, | |
| "learning_rate": 0.0002448636589449552, | |
| "loss": 12.1299, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.08664259927797834, | |
| "grad_norm": 5.419031143188477, | |
| "learning_rate": 0.0002447285671585931, | |
| "loss": 10.0219, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.08763849122370222, | |
| "grad_norm": 6.183884143829346, | |
| "learning_rate": 0.00024459176006930947, | |
| "loss": 13.1559, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.08863438316942612, | |
| "grad_norm": 5.832536697387695, | |
| "learning_rate": 0.0002444532396370844, | |
| "loss": 11.4276, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.08963027511515001, | |
| "grad_norm": 6.158980846405029, | |
| "learning_rate": 0.0002443130078464444, | |
| "loss": 12.6743, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0906261670608739, | |
| "grad_norm": 7.851455211639404, | |
| "learning_rate": 0.0002441710667064337, | |
| "loss": 15.0347, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.09162205900659778, | |
| "grad_norm": 6.618966102600098, | |
| "learning_rate": 0.00024402741825058576, | |
| "loss": 12.8038, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.09261795095232167, | |
| "grad_norm": 7.711861610412598, | |
| "learning_rate": 0.0002438820645368942, | |
| "loss": 14.378, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.09361384289804556, | |
| "grad_norm": 6.327939510345459, | |
| "learning_rate": 0.00024373500764778307, | |
| "loss": 11.8801, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.09460973484376944, | |
| "grad_norm": 6.685515880584717, | |
| "learning_rate": 0.00024358624969007705, | |
| "loss": 12.5321, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.09560562678949334, | |
| "grad_norm": 6.379256725311279, | |
| "learning_rate": 0.0002434357927949716, | |
| "loss": 12.7246, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.09660151873521723, | |
| "grad_norm": 5.819457054138184, | |
| "learning_rate": 0.00024328363911800183, | |
| "loss": 11.6533, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.09759741068094112, | |
| "grad_norm": 6.651020050048828, | |
| "learning_rate": 0.00024312979083901227, | |
| "loss": 13.6643, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.098593302626665, | |
| "grad_norm": 6.834366321563721, | |
| "learning_rate": 0.00024297425016212517, | |
| "loss": 12.5779, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.09958919457238889, | |
| "grad_norm": 5.578615188598633, | |
| "learning_rate": 0.0002428170193157091, | |
| "loss": 12.1336, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09958919457238889, | |
| "eval_loss": 1.5528450012207031, | |
| "eval_runtime": 289.8128, | |
| "eval_samples_per_second": 4.62, | |
| "eval_steps_per_second": 1.156, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.10058508651811278, | |
| "grad_norm": 6.1246747970581055, | |
| "learning_rate": 0.000242658100552347, | |
| "loss": 12.9692, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.10158097846383668, | |
| "grad_norm": 6.364856719970703, | |
| "learning_rate": 0.00024249749614880397, | |
| "loss": 10.5347, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.10257687040956057, | |
| "grad_norm": 6.769341945648193, | |
| "learning_rate": 0.00024233520840599457, | |
| "loss": 12.3651, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.10357276235528445, | |
| "grad_norm": 7.491860389709473, | |
| "learning_rate": 0.00024217123964894986, | |
| "loss": 12.3008, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.10456865430100834, | |
| "grad_norm": 6.12730073928833, | |
| "learning_rate": 0.00024200559222678408, | |
| "loss": 11.3047, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.10556454624673223, | |
| "grad_norm": 7.921679496765137, | |
| "learning_rate": 0.00024183826851266116, | |
| "loss": 13.0155, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.10656043819245611, | |
| "grad_norm": 6.370965480804443, | |
| "learning_rate": 0.00024166927090376052, | |
| "loss": 12.2079, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.10755633013818, | |
| "grad_norm": 7.240699768066406, | |
| "learning_rate": 0.00024149860182124267, | |
| "loss": 13.1831, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.1085522220839039, | |
| "grad_norm": 6.247211456298828, | |
| "learning_rate": 0.0002413262637102148, | |
| "loss": 12.2088, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.10954811402962779, | |
| "grad_norm": 6.4018168449401855, | |
| "learning_rate": 0.00024115225903969568, | |
| "loss": 12.0962, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11054400597535168, | |
| "grad_norm": 6.562201976776123, | |
| "learning_rate": 0.00024097659030257993, | |
| "loss": 11.6653, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.11153989792107556, | |
| "grad_norm": 7.297281742095947, | |
| "learning_rate": 0.0002407992600156028, | |
| "loss": 12.4297, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.11253578986679945, | |
| "grad_norm": 6.0937299728393555, | |
| "learning_rate": 0.00024062027071930386, | |
| "loss": 12.2878, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.11353168181252334, | |
| "grad_norm": 6.412242889404297, | |
| "learning_rate": 0.0002404396249779906, | |
| "loss": 12.9066, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.11452757375824724, | |
| "grad_norm": 6.969156742095947, | |
| "learning_rate": 0.00024025732537970168, | |
| "loss": 11.7735, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.11552346570397112, | |
| "grad_norm": 6.48976993560791, | |
| "learning_rate": 0.00024007337453617005, | |
| "loss": 11.5464, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.11651935764969501, | |
| "grad_norm": 6.79917049407959, | |
| "learning_rate": 0.00023988777508278524, | |
| "loss": 14.0385, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.1175152495954189, | |
| "grad_norm": 5.524596691131592, | |
| "learning_rate": 0.00023970052967855587, | |
| "loss": 12.4122, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.11851114154114278, | |
| "grad_norm": 8.637755393981934, | |
| "learning_rate": 0.00023951164100607128, | |
| "loss": 12.9614, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.11950703348686667, | |
| "grad_norm": 7.299292087554932, | |
| "learning_rate": 0.00023932111177146342, | |
| "loss": 12.6715, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.12050292543259056, | |
| "grad_norm": 6.199335098266602, | |
| "learning_rate": 0.0002391289447043678, | |
| "loss": 11.9946, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.12149881737831446, | |
| "grad_norm": 5.948755741119385, | |
| "learning_rate": 0.0002389351425578845, | |
| "loss": 12.3719, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.12249470932403834, | |
| "grad_norm": 6.023370265960693, | |
| "learning_rate": 0.00023873970810853884, | |
| "loss": 11.4807, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.12349060126976223, | |
| "grad_norm": 5.60730504989624, | |
| "learning_rate": 0.00023854264415624135, | |
| "loss": 11.9841, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.12448649321548612, | |
| "grad_norm": 5.656817436218262, | |
| "learning_rate": 0.0002383439535242478, | |
| "loss": 11.3724, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.12548238516121002, | |
| "grad_norm": 6.920670509338379, | |
| "learning_rate": 0.0002381436390591189, | |
| "loss": 12.0686, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.1264782771069339, | |
| "grad_norm": 7.383021831512451, | |
| "learning_rate": 0.00023794170363067914, | |
| "loss": 14.0681, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1274741690526578, | |
| "grad_norm": 8.584571838378906, | |
| "learning_rate": 0.00023773815013197608, | |
| "loss": 15.4667, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.12847006099838168, | |
| "grad_norm": 12.736078262329102, | |
| "learning_rate": 0.00023753298147923858, | |
| "loss": 12.1753, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.12946595294410557, | |
| "grad_norm": 6.229617118835449, | |
| "learning_rate": 0.0002373262006118353, | |
| "loss": 13.8618, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.13046184488982945, | |
| "grad_norm": 5.858534336090088, | |
| "learning_rate": 0.0002371178104922323, | |
| "loss": 12.3115, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.13145773683555334, | |
| "grad_norm": 6.3285603523254395, | |
| "learning_rate": 0.00023690781410595085, | |
| "loss": 12.8249, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.13245362878127723, | |
| "grad_norm": 5.9630632400512695, | |
| "learning_rate": 0.00023669621446152463, | |
| "loss": 13.0259, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.13344952072700111, | |
| "grad_norm": 6.730586528778076, | |
| "learning_rate": 0.0002364830145904563, | |
| "loss": 12.6343, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.134445412672725, | |
| "grad_norm": 7.248769760131836, | |
| "learning_rate": 0.0002362682175471746, | |
| "loss": 11.8677, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.1354413046184489, | |
| "grad_norm": 6.6009979248046875, | |
| "learning_rate": 0.0002360518264089901, | |
| "loss": 12.3887, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.13643719656417277, | |
| "grad_norm": 5.64854097366333, | |
| "learning_rate": 0.00023583384427605146, | |
| "loss": 10.898, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.1374330885098967, | |
| "grad_norm": 7.965289115905762, | |
| "learning_rate": 0.00023561427427130083, | |
| "loss": 13.6869, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.13842898045562058, | |
| "grad_norm": 8.878471374511719, | |
| "learning_rate": 0.00023539311954042912, | |
| "loss": 12.4927, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.13942487240134446, | |
| "grad_norm": 6.698742866516113, | |
| "learning_rate": 0.0002351703832518311, | |
| "loss": 11.4779, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14042076434706835, | |
| "grad_norm": 8.585515975952148, | |
| "learning_rate": 0.00023494606859655976, | |
| "loss": 13.2428, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.14141665629279224, | |
| "grad_norm": 6.803221225738525, | |
| "learning_rate": 0.00023472017878828073, | |
| "loss": 11.7988, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.14241254823851612, | |
| "grad_norm": 7.021427154541016, | |
| "learning_rate": 0.00023449271706322622, | |
| "loss": 12.4274, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.14340844018424, | |
| "grad_norm": 5.856128692626953, | |
| "learning_rate": 0.00023426368668014874, | |
| "loss": 12.2856, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.1444043321299639, | |
| "grad_norm": 6.2399210929870605, | |
| "learning_rate": 0.00023403309092027424, | |
| "loss": 12.307, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.14540022407568778, | |
| "grad_norm": 6.5773210525512695, | |
| "learning_rate": 0.00023380093308725518, | |
| "loss": 12.4102, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.14639611602141167, | |
| "grad_norm": 7.773036479949951, | |
| "learning_rate": 0.00023356721650712338, | |
| "loss": 14.0709, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.14739200796713556, | |
| "grad_norm": 5.980135917663574, | |
| "learning_rate": 0.00023333194452824195, | |
| "loss": 11.8756, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.14838789991285944, | |
| "grad_norm": 6.181486129760742, | |
| "learning_rate": 0.00023309512052125775, | |
| "loss": 12.3727, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.14938379185858333, | |
| "grad_norm": 13.033953666687012, | |
| "learning_rate": 0.00023285674787905286, | |
| "loss": 13.3292, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15037968380430725, | |
| "grad_norm": 6.594436168670654, | |
| "learning_rate": 0.00023261683001669602, | |
| "loss": 13.3059, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.15137557575003113, | |
| "grad_norm": 7.247442722320557, | |
| "learning_rate": 0.00023237537037139384, | |
| "loss": 12.0958, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.15237146769575502, | |
| "grad_norm": 6.844114303588867, | |
| "learning_rate": 0.0002321323724024412, | |
| "loss": 12.2067, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.1533673596414789, | |
| "grad_norm": 5.828722953796387, | |
| "learning_rate": 0.0002318878395911721, | |
| "loss": 11.2265, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.1543632515872028, | |
| "grad_norm": 7.751287460327148, | |
| "learning_rate": 0.00023164177544090958, | |
| "loss": 14.7071, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.15535914353292668, | |
| "grad_norm": 6.924627780914307, | |
| "learning_rate": 0.00023139418347691555, | |
| "loss": 11.9454, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.15635503547865057, | |
| "grad_norm": 6.240747451782227, | |
| "learning_rate": 0.0002311450672463402, | |
| "loss": 12.9397, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.15735092742437445, | |
| "grad_norm": 6.814810752868652, | |
| "learning_rate": 0.00023089443031817147, | |
| "loss": 11.8282, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.15834681937009834, | |
| "grad_norm": 7.39363956451416, | |
| "learning_rate": 0.0002306422762831835, | |
| "loss": 14.9512, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.15934271131582223, | |
| "grad_norm": 6.947161674499512, | |
| "learning_rate": 0.00023038860875388556, | |
| "loss": 12.2441, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.16033860326154611, | |
| "grad_norm": 6.200173854827881, | |
| "learning_rate": 0.00023013343136447006, | |
| "loss": 11.8266, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.16133449520727, | |
| "grad_norm": 6.987513065338135, | |
| "learning_rate": 0.00022987674777076068, | |
| "loss": 12.7667, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.1623303871529939, | |
| "grad_norm": 6.824809551239014, | |
| "learning_rate": 0.0002296185616501597, | |
| "loss": 12.4046, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.1633262790987178, | |
| "grad_norm": 6.129730701446533, | |
| "learning_rate": 0.00022935887670159566, | |
| "loss": 12.8159, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.1643221710444417, | |
| "grad_norm": 7.575254917144775, | |
| "learning_rate": 0.00022909769664547014, | |
| "loss": 11.9914, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.16531806299016558, | |
| "grad_norm": 6.266209602355957, | |
| "learning_rate": 0.0002288350252236045, | |
| "loss": 11.6365, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.16631395493588946, | |
| "grad_norm": 7.802381992340088, | |
| "learning_rate": 0.00022857086619918634, | |
| "loss": 12.7036, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.16730984688161335, | |
| "grad_norm": 7.222281455993652, | |
| "learning_rate": 0.00022830522335671555, | |
| "loss": 11.8082, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.16830573882733724, | |
| "grad_norm": 7.55634069442749, | |
| "learning_rate": 0.00022803810050195004, | |
| "loss": 12.8177, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.16930163077306112, | |
| "grad_norm": 6.894867420196533, | |
| "learning_rate": 0.00022776950146185127, | |
| "loss": 12.4942, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.170297522718785, | |
| "grad_norm": 7.815672874450684, | |
| "learning_rate": 0.0002274994300845294, | |
| "loss": 13.8819, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.1712934146645089, | |
| "grad_norm": 6.255967140197754, | |
| "learning_rate": 0.00022722789023918823, | |
| "loss": 11.8453, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.17228930661023278, | |
| "grad_norm": 7.725053310394287, | |
| "learning_rate": 0.0002269548858160697, | |
| "loss": 12.0805, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.17328519855595667, | |
| "grad_norm": 8.110404968261719, | |
| "learning_rate": 0.00022668042072639805, | |
| "loss": 11.1405, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.17428109050168056, | |
| "grad_norm": 6.437393665313721, | |
| "learning_rate": 0.00022640449890232403, | |
| "loss": 11.7487, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.17527698244740444, | |
| "grad_norm": 6.381730556488037, | |
| "learning_rate": 0.00022612712429686844, | |
| "loss": 12.7643, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.17627287439312836, | |
| "grad_norm": 5.764153480529785, | |
| "learning_rate": 0.00022584830088386539, | |
| "loss": 11.6039, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.17726876633885225, | |
| "grad_norm": 5.821381568908691, | |
| "learning_rate": 0.00022556803265790553, | |
| "loss": 12.7432, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.17826465828457613, | |
| "grad_norm": 5.812039375305176, | |
| "learning_rate": 0.00022528632363427882, | |
| "loss": 11.2202, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.17926055023030002, | |
| "grad_norm": 6.043585300445557, | |
| "learning_rate": 0.00022500317784891684, | |
| "loss": 11.7178, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1802564421760239, | |
| "grad_norm": 6.763357639312744, | |
| "learning_rate": 0.0002247185993583351, | |
| "loss": 13.0897, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.1812523341217478, | |
| "grad_norm": 6.222050666809082, | |
| "learning_rate": 0.00022443259223957498, | |
| "loss": 12.9561, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.18224822606747168, | |
| "grad_norm": 6.555530071258545, | |
| "learning_rate": 0.00022414516059014516, | |
| "loss": 13.3252, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.18324411801319557, | |
| "grad_norm": 6.633552551269531, | |
| "learning_rate": 0.00022385630852796306, | |
| "loss": 12.5151, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.18424000995891945, | |
| "grad_norm": 5.674712657928467, | |
| "learning_rate": 0.00022356604019129573, | |
| "loss": 10.8516, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.18523590190464334, | |
| "grad_norm": 6.16930627822876, | |
| "learning_rate": 0.00022327435973870058, | |
| "loss": 12.2844, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.18623179385036723, | |
| "grad_norm": 6.338482856750488, | |
| "learning_rate": 0.00022298127134896595, | |
| "loss": 13.021, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.18722768579609111, | |
| "grad_norm": 5.933238506317139, | |
| "learning_rate": 0.000222686779221051, | |
| "loss": 11.8773, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.188223577741815, | |
| "grad_norm": 6.159973621368408, | |
| "learning_rate": 0.00022239088757402582, | |
| "loss": 11.4913, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.1892194696875389, | |
| "grad_norm": 7.25203800201416, | |
| "learning_rate": 0.0002220936006470107, | |
| "loss": 13.7217, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1902153616332628, | |
| "grad_norm": 6.028827667236328, | |
| "learning_rate": 0.00022179492269911564, | |
| "loss": 11.364, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.1912112535789867, | |
| "grad_norm": 6.040252685546875, | |
| "learning_rate": 0.00022149485800937918, | |
| "loss": 12.5145, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.19220714552471058, | |
| "grad_norm": 5.860878944396973, | |
| "learning_rate": 0.00022119341087670723, | |
| "loss": 11.5606, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.19320303747043446, | |
| "grad_norm": 6.734287261962891, | |
| "learning_rate": 0.00022089058561981128, | |
| "loss": 12.0882, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.19419892941615835, | |
| "grad_norm": 6.507574558258057, | |
| "learning_rate": 0.00022058638657714683, | |
| "loss": 13.0287, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.19519482136188224, | |
| "grad_norm": 5.8352179527282715, | |
| "learning_rate": 0.00022028081810685084, | |
| "loss": 11.9273, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.19619071330760612, | |
| "grad_norm": 6.327966690063477, | |
| "learning_rate": 0.00021997388458667972, | |
| "loss": 11.9534, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.19718660525333, | |
| "grad_norm": 7.161978721618652, | |
| "learning_rate": 0.00021966559041394619, | |
| "loss": 14.3134, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.1981824971990539, | |
| "grad_norm": 7.602406978607178, | |
| "learning_rate": 0.00021935594000545663, | |
| "loss": 13.8131, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.19917838914477778, | |
| "grad_norm": 6.5593767166137695, | |
| "learning_rate": 0.00021904493779744766, | |
| "loss": 12.1681, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.19917838914477778, | |
| "eval_loss": 1.545896291732788, | |
| "eval_runtime": 289.9141, | |
| "eval_samples_per_second": 4.619, | |
| "eval_steps_per_second": 1.156, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.20017428109050167, | |
| "grad_norm": 7.11292839050293, | |
| "learning_rate": 0.00021873258824552257, | |
| "loss": 11.3805, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.20117017303622556, | |
| "grad_norm": 5.791740417480469, | |
| "learning_rate": 0.0002184188958245874, | |
| "loss": 11.4427, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.20216606498194944, | |
| "grad_norm": 6.359834671020508, | |
| "learning_rate": 0.0002181038650287871, | |
| "loss": 12.8047, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.20316195692767336, | |
| "grad_norm": 6.810841083526611, | |
| "learning_rate": 0.00021778750037144086, | |
| "loss": 12.9342, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.20415784887339725, | |
| "grad_norm": 6.595353126525879, | |
| "learning_rate": 0.0002174698063849776, | |
| "loss": 11.8879, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.20515374081912113, | |
| "grad_norm": 6.707181453704834, | |
| "learning_rate": 0.00021715078762087108, | |
| "loss": 12.4952, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.20614963276484502, | |
| "grad_norm": 5.755362033843994, | |
| "learning_rate": 0.00021683044864957444, | |
| "loss": 12.742, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2071455247105689, | |
| "grad_norm": 5.790565490722656, | |
| "learning_rate": 0.00021650879406045508, | |
| "loss": 10.6683, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.2081414166562928, | |
| "grad_norm": 6.435606002807617, | |
| "learning_rate": 0.0002161858284617286, | |
| "loss": 12.8725, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.20913730860201668, | |
| "grad_norm": 6.840058326721191, | |
| "learning_rate": 0.00021586155648039296, | |
| "loss": 11.7874, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.21013320054774057, | |
| "grad_norm": 6.608837604522705, | |
| "learning_rate": 0.00021553598276216217, | |
| "loss": 13.3114, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.21112909249346445, | |
| "grad_norm": 5.996092319488525, | |
| "learning_rate": 0.00021520911197139958, | |
| "loss": 11.8956, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.21212498443918834, | |
| "grad_norm": 5.224332332611084, | |
| "learning_rate": 0.00021488094879105134, | |
| "loss": 10.4107, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.21312087638491223, | |
| "grad_norm": 5.718519687652588, | |
| "learning_rate": 0.000214551497922579, | |
| "loss": 11.4343, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.21411676833063611, | |
| "grad_norm": 6.161787033081055, | |
| "learning_rate": 0.00021422076408589237, | |
| "loss": 11.9857, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.21511266027636, | |
| "grad_norm": 6.545485019683838, | |
| "learning_rate": 0.00021388875201928183, | |
| "loss": 13.6983, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.21610855222208392, | |
| "grad_norm": 6.377624988555908, | |
| "learning_rate": 0.0002135554664793504, | |
| "loss": 12.8974, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.2171044441678078, | |
| "grad_norm": 6.514633655548096, | |
| "learning_rate": 0.0002132209122409457, | |
| "loss": 12.2978, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.2181003361135317, | |
| "grad_norm": 5.409609794616699, | |
| "learning_rate": 0.00021288509409709148, | |
| "loss": 10.8475, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.21909622805925558, | |
| "grad_norm": 5.895532131195068, | |
| "learning_rate": 0.00021254801685891887, | |
| "loss": 10.9371, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.22009212000497946, | |
| "grad_norm": 5.818058013916016, | |
| "learning_rate": 0.0002122096853555976, | |
| "loss": 12.6627, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.22108801195070335, | |
| "grad_norm": 5.7109375, | |
| "learning_rate": 0.00021187010443426675, | |
| "loss": 12.3313, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.22208390389642724, | |
| "grad_norm": 6.082381248474121, | |
| "learning_rate": 0.0002115292789599653, | |
| "loss": 11.4319, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.22307979584215112, | |
| "grad_norm": 5.8958048820495605, | |
| "learning_rate": 0.00021118721381556245, | |
| "loss": 11.6646, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.224075687787875, | |
| "grad_norm": 5.592798233032227, | |
| "learning_rate": 0.00021084391390168764, | |
| "loss": 11.1113, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2250715797335989, | |
| "grad_norm": 5.9306182861328125, | |
| "learning_rate": 0.00021049938413666037, | |
| "loss": 12.4602, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.22606747167932278, | |
| "grad_norm": 5.919130325317383, | |
| "learning_rate": 0.0002101536294564197, | |
| "loss": 11.9998, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.22706336362504667, | |
| "grad_norm": 6.082193374633789, | |
| "learning_rate": 0.00020980665481445355, | |
| "loss": 11.6682, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.22805925557077056, | |
| "grad_norm": 6.038330078125, | |
| "learning_rate": 0.00020945846518172776, | |
| "loss": 11.8126, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.22905514751649447, | |
| "grad_norm": 6.054876327514648, | |
| "learning_rate": 0.00020910906554661484, | |
| "loss": 11.6183, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.23005103946221836, | |
| "grad_norm": 6.542704105377197, | |
| "learning_rate": 0.0002087584609148226, | |
| "loss": 12.4262, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.23104693140794225, | |
| "grad_norm": 5.895127296447754, | |
| "learning_rate": 0.00020840665630932225, | |
| "loss": 11.9887, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.23204282335366613, | |
| "grad_norm": 6.9951863288879395, | |
| "learning_rate": 0.00020805365677027646, | |
| "loss": 12.2263, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.23303871529939002, | |
| "grad_norm": 6.330830097198486, | |
| "learning_rate": 0.0002076994673549675, | |
| "loss": 13.8628, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.2340346072451139, | |
| "grad_norm": 5.970463752746582, | |
| "learning_rate": 0.00020734409313772424, | |
| "loss": 11.4963, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2350304991908378, | |
| "grad_norm": 6.236823081970215, | |
| "learning_rate": 0.00020698753920984987, | |
| "loss": 12.8625, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.23602639113656168, | |
| "grad_norm": 5.732664108276367, | |
| "learning_rate": 0.00020662981067954883, | |
| "loss": 11.9416, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.23702228308228557, | |
| "grad_norm": 6.58101224899292, | |
| "learning_rate": 0.00020627091267185355, | |
| "loss": 11.9936, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.23801817502800945, | |
| "grad_norm": 6.880969047546387, | |
| "learning_rate": 0.0002059108503285511, | |
| "loss": 13.5075, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.23901406697373334, | |
| "grad_norm": 5.696409702301025, | |
| "learning_rate": 0.00020554962880810963, | |
| "loss": 11.9267, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.24000995891945723, | |
| "grad_norm": 5.385843276977539, | |
| "learning_rate": 0.00020518725328560417, | |
| "loss": 10.0967, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.24100585086518111, | |
| "grad_norm": 6.039820671081543, | |
| "learning_rate": 0.00020482372895264282, | |
| "loss": 10.6372, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.24200174281090503, | |
| "grad_norm": 5.8607916831970215, | |
| "learning_rate": 0.00020445906101729212, | |
| "loss": 11.0905, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.24299763475662892, | |
| "grad_norm": 6.985530376434326, | |
| "learning_rate": 0.00020409325470400263, | |
| "loss": 11.455, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2439935267023528, | |
| "grad_norm": 5.882106304168701, | |
| "learning_rate": 0.0002037263152535339, | |
| "loss": 12.6405, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2449894186480767, | |
| "grad_norm": 5.3988776206970215, | |
| "learning_rate": 0.0002033582479228796, | |
| "loss": 11.5435, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.24598531059380058, | |
| "grad_norm": 7.481232166290283, | |
| "learning_rate": 0.00020298905798519197, | |
| "loss": 13.4326, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.24698120253952446, | |
| "grad_norm": 6.147839546203613, | |
| "learning_rate": 0.00020261875072970643, | |
| "loss": 11.5964, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.24797709448524835, | |
| "grad_norm": 8.947158813476562, | |
| "learning_rate": 0.0002022473314616658, | |
| "loss": 12.0333, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.24897298643097224, | |
| "grad_norm": 6.649617671966553, | |
| "learning_rate": 0.00020187480550224422, | |
| "loss": 13.5379, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.24996887837669612, | |
| "grad_norm": 6.397286891937256, | |
| "learning_rate": 0.00020150117818847088, | |
| "loss": 11.1719, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.25096477032242004, | |
| "grad_norm": 5.593772888183594, | |
| "learning_rate": 0.0002011264548731538, | |
| "loss": 11.7092, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2519606622681439, | |
| "grad_norm": 6.753810405731201, | |
| "learning_rate": 0.00020075064092480284, | |
| "loss": 13.9164, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.2529565542138678, | |
| "grad_norm": 5.851145267486572, | |
| "learning_rate": 0.0002003737417275529, | |
| "loss": 12.3499, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.25395244615959167, | |
| "grad_norm": 5.922454833984375, | |
| "learning_rate": 0.00019999576268108694, | |
| "loss": 12.029, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2549483381053156, | |
| "grad_norm": 5.876878261566162, | |
| "learning_rate": 0.0001996167092005584, | |
| "loss": 11.8283, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.25594423005103945, | |
| "grad_norm": 6.563518047332764, | |
| "learning_rate": 0.00019923658671651363, | |
| "loss": 13.6434, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.25694012199676336, | |
| "grad_norm": 5.781143665313721, | |
| "learning_rate": 0.00019885540067481427, | |
| "loss": 11.871, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.2579360139424872, | |
| "grad_norm": 6.5753912925720215, | |
| "learning_rate": 0.00019847315653655914, | |
| "loss": 12.0575, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.25893190588821113, | |
| "grad_norm": 5.838126182556152, | |
| "learning_rate": 0.0001980898597780059, | |
| "loss": 12.2294, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.259927797833935, | |
| "grad_norm": 5.333262920379639, | |
| "learning_rate": 0.00019770551589049268, | |
| "loss": 11.7186, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.2609236897796589, | |
| "grad_norm": 6.755197525024414, | |
| "learning_rate": 0.00019732013038035952, | |
| "loss": 12.5457, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.2619195817253828, | |
| "grad_norm": 6.212986946105957, | |
| "learning_rate": 0.00019693370876886916, | |
| "loss": 12.6501, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.2629154736711067, | |
| "grad_norm": 5.702981948852539, | |
| "learning_rate": 0.00019654625659212835, | |
| "loss": 12.1024, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.2639113656168306, | |
| "grad_norm": 5.915462970733643, | |
| "learning_rate": 0.00019615777940100825, | |
| "loss": 11.2701, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.26490725756255445, | |
| "grad_norm": 6.289284706115723, | |
| "learning_rate": 0.00019576828276106497, | |
| "loss": 12.4175, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.26590314950827837, | |
| "grad_norm": 6.77176570892334, | |
| "learning_rate": 0.0001953777722524599, | |
| "loss": 12.4483, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.26689904145400223, | |
| "grad_norm": 6.126654148101807, | |
| "learning_rate": 0.00019498625346987963, | |
| "loss": 11.3449, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.26789493339972614, | |
| "grad_norm": 5.859162330627441, | |
| "learning_rate": 0.000194593732022456, | |
| "loss": 11.2103, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.26889082534545, | |
| "grad_norm": 6.313859939575195, | |
| "learning_rate": 0.00019420021353368556, | |
| "loss": 11.3182, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2698867172911739, | |
| "grad_norm": 6.366731643676758, | |
| "learning_rate": 0.0001938057036413491, | |
| "loss": 12.3075, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.2708826092368978, | |
| "grad_norm": 6.110815525054932, | |
| "learning_rate": 0.00019341020799743075, | |
| "loss": 12.6064, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.2718785011826217, | |
| "grad_norm": 6.0649094581604, | |
| "learning_rate": 0.00019301373226803727, | |
| "loss": 12.5147, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.27287439312834555, | |
| "grad_norm": 7.83512020111084, | |
| "learning_rate": 0.00019261628213331655, | |
| "loss": 12.0182, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.27387028507406946, | |
| "grad_norm": 6.209272861480713, | |
| "learning_rate": 0.00019221786328737651, | |
| "loss": 13.2367, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2748661770197934, | |
| "grad_norm": 5.425859451293945, | |
| "learning_rate": 0.00019181848143820336, | |
| "loss": 11.2941, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 6.430884838104248, | |
| "learning_rate": 0.00019141814230757987, | |
| "loss": 11.9494, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.27685796091124115, | |
| "grad_norm": 5.499374866485596, | |
| "learning_rate": 0.00019101685163100342, | |
| "loss": 11.3679, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.277853852856965, | |
| "grad_norm": 6.815554141998291, | |
| "learning_rate": 0.00019061461515760368, | |
| "loss": 12.3956, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2788497448026889, | |
| "grad_norm": 5.373319149017334, | |
| "learning_rate": 0.00019021143865006058, | |
| "loss": 11.1815, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2798456367484128, | |
| "grad_norm": 5.54528284072876, | |
| "learning_rate": 0.00018980732788452138, | |
| "loss": 11.2036, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.2808415286941367, | |
| "grad_norm": 5.694971084594727, | |
| "learning_rate": 0.00018940228865051812, | |
| "loss": 11.9744, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.28183742063986056, | |
| "grad_norm": 6.034758567810059, | |
| "learning_rate": 0.0001889963267508847, | |
| "loss": 12.2473, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.2828333125855845, | |
| "grad_norm": 5.718782901763916, | |
| "learning_rate": 0.0001885894480016736, | |
| "loss": 11.7797, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.28382920453130833, | |
| "grad_norm": 5.70759391784668, | |
| "learning_rate": 0.00018818165823207263, | |
| "loss": 12.6039, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.28482509647703225, | |
| "grad_norm": 5.922822952270508, | |
| "learning_rate": 0.00018777296328432143, | |
| "loss": 12.0609, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.2858209884227561, | |
| "grad_norm": 6.213382720947266, | |
| "learning_rate": 0.00018736336901362783, | |
| "loss": 12.721, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.28681688036848, | |
| "grad_norm": 5.525908946990967, | |
| "learning_rate": 0.00018695288128808376, | |
| "loss": 10.897, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.28781277231420394, | |
| "grad_norm": 5.966858386993408, | |
| "learning_rate": 0.00018654150598858152, | |
| "loss": 11.5072, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.2888086642599278, | |
| "grad_norm": 6.100443363189697, | |
| "learning_rate": 0.00018612924900872916, | |
| "loss": 12.3079, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2898045562056517, | |
| "grad_norm": 5.905527591705322, | |
| "learning_rate": 0.00018571611625476625, | |
| "loss": 11.6011, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.29080044815137557, | |
| "grad_norm": 18.4097900390625, | |
| "learning_rate": 0.0001853021136454792, | |
| "loss": 12.0392, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.2917963400970995, | |
| "grad_norm": 6.539964199066162, | |
| "learning_rate": 0.0001848872471121166, | |
| "loss": 12.8208, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.29279223204282334, | |
| "grad_norm": 5.856533527374268, | |
| "learning_rate": 0.00018447152259830398, | |
| "loss": 11.9306, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.29378812398854726, | |
| "grad_norm": 6.079041957855225, | |
| "learning_rate": 0.00018405494605995887, | |
| "loss": 11.4677, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.2947840159342711, | |
| "grad_norm": 7.1472039222717285, | |
| "learning_rate": 0.00018363752346520548, | |
| "loss": 12.2707, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.29577990787999503, | |
| "grad_norm": 6.571793556213379, | |
| "learning_rate": 0.00018321926079428903, | |
| "loss": 12.2654, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.2967757998257189, | |
| "grad_norm": 6.160271644592285, | |
| "learning_rate": 0.00018280016403949024, | |
| "loss": 12.8305, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.2977716917714428, | |
| "grad_norm": 6.2567572593688965, | |
| "learning_rate": 0.00018238023920503935, | |
| "loss": 12.1013, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.29876758371716666, | |
| "grad_norm": 7.734886646270752, | |
| "learning_rate": 0.00018195949230703022, | |
| "loss": 13.1276, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.29876758371716666, | |
| "eval_loss": 1.516517162322998, | |
| "eval_runtime": 289.8399, | |
| "eval_samples_per_second": 4.62, | |
| "eval_steps_per_second": 1.156, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2997634756628906, | |
| "grad_norm": 7.519866943359375, | |
| "learning_rate": 0.00018153792937333405, | |
| "loss": 12.56, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.3007593676086145, | |
| "grad_norm": 6.709288120269775, | |
| "learning_rate": 0.00018111555644351307, | |
| "loss": 12.5308, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.30175525955433835, | |
| "grad_norm": 5.9070281982421875, | |
| "learning_rate": 0.000180692379568734, | |
| "loss": 10.9037, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.30275115150006227, | |
| "grad_norm": 7.148046493530273, | |
| "learning_rate": 0.00018026840481168138, | |
| "loss": 12.2679, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3037470434457861, | |
| "grad_norm": 5.894509315490723, | |
| "learning_rate": 0.00017984363824647065, | |
| "loss": 11.3378, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.30474293539151004, | |
| "grad_norm": 5.6303229331970215, | |
| "learning_rate": 0.00017941808595856113, | |
| "loss": 10.7154, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.3057388273372339, | |
| "grad_norm": 5.428180694580078, | |
| "learning_rate": 0.00017899175404466897, | |
| "loss": 10.747, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.3067347192829578, | |
| "grad_norm": 5.827486038208008, | |
| "learning_rate": 0.0001785646486126796, | |
| "loss": 12.2957, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.30773061122868167, | |
| "grad_norm": 6.739930152893066, | |
| "learning_rate": 0.0001781367757815604, | |
| "loss": 13.2347, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.3087265031744056, | |
| "grad_norm": 6.382988929748535, | |
| "learning_rate": 0.000177708141681273, | |
| "loss": 12.367, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.30972239512012945, | |
| "grad_norm": 6.265737533569336, | |
| "learning_rate": 0.00017727875245268534, | |
| "loss": 12.4424, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.31071828706585336, | |
| "grad_norm": 6.454976558685303, | |
| "learning_rate": 0.00017684861424748386, | |
| "loss": 12.2459, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.3117141790115772, | |
| "grad_norm": 5.488022804260254, | |
| "learning_rate": 0.00017641773322808518, | |
| "loss": 10.8191, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.31271007095730113, | |
| "grad_norm": 9.193635940551758, | |
| "learning_rate": 0.00017598611556754804, | |
| "loss": 13.3586, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.31370596290302505, | |
| "grad_norm": 6.968887805938721, | |
| "learning_rate": 0.0001755537674494846, | |
| "loss": 13.9658, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3147018548487489, | |
| "grad_norm": 5.535203456878662, | |
| "learning_rate": 0.00017512069506797224, | |
| "loss": 12.2494, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.3156977467944728, | |
| "grad_norm": 5.723720073699951, | |
| "learning_rate": 0.00017468690462746426, | |
| "loss": 12.0819, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.3166936387401967, | |
| "grad_norm": 6.4840850830078125, | |
| "learning_rate": 0.00017425240234270148, | |
| "loss": 11.7859, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.3176895306859206, | |
| "grad_norm": 6.271676063537598, | |
| "learning_rate": 0.00017381719443862305, | |
| "loss": 13.0953, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.31868542263164445, | |
| "grad_norm": 6.508022785186768, | |
| "learning_rate": 0.00017338128715027717, | |
| "loss": 11.8154, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.31968131457736837, | |
| "grad_norm": 5.159334659576416, | |
| "learning_rate": 0.00017294468672273178, | |
| "loss": 11.3996, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.32067720652309223, | |
| "grad_norm": 5.714421272277832, | |
| "learning_rate": 0.00017250739941098532, | |
| "loss": 12.8778, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.32167309846881614, | |
| "grad_norm": 6.049591064453125, | |
| "learning_rate": 0.00017206943147987677, | |
| "loss": 12.7078, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.32266899041454, | |
| "grad_norm": 5.64936637878418, | |
| "learning_rate": 0.00017163078920399616, | |
| "loss": 12.0689, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.3236648823602639, | |
| "grad_norm": 5.128944396972656, | |
| "learning_rate": 0.00017119147886759462, | |
| "loss": 11.2092, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3246607743059878, | |
| "grad_norm": 5.202014923095703, | |
| "learning_rate": 0.00017075150676449418, | |
| "loss": 10.8206, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.3256566662517117, | |
| "grad_norm": 5.825450420379639, | |
| "learning_rate": 0.00017031087919799792, | |
| "loss": 11.1645, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.3266525581974356, | |
| "grad_norm": 6.169241428375244, | |
| "learning_rate": 0.0001698696024807993, | |
| "loss": 12.5059, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.32764845014315946, | |
| "grad_norm": 5.669341087341309, | |
| "learning_rate": 0.00016942768293489198, | |
| "loss": 12.112, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.3286443420888834, | |
| "grad_norm": 5.91945219039917, | |
| "learning_rate": 0.00016898512689147912, | |
| "loss": 12.5704, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.32964023403460724, | |
| "grad_norm": 5.3270134925842285, | |
| "learning_rate": 0.0001685419406908829, | |
| "loss": 11.1185, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.33063612598033115, | |
| "grad_norm": 6.189324378967285, | |
| "learning_rate": 0.0001680981306824533, | |
| "loss": 13.594, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.331632017926055, | |
| "grad_norm": 5.268672466278076, | |
| "learning_rate": 0.00016765370322447738, | |
| "loss": 11.815, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.3326279098717789, | |
| "grad_norm": 6.1424407958984375, | |
| "learning_rate": 0.0001672086646840883, | |
| "loss": 12.4271, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.3336238018175028, | |
| "grad_norm": 6.57133674621582, | |
| "learning_rate": 0.00016676302143717376, | |
| "loss": 11.8458, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3346196937632267, | |
| "grad_norm": 6.405190467834473, | |
| "learning_rate": 0.0001663167798682849, | |
| "loss": 13.4903, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.33561558570895056, | |
| "grad_norm": 5.354515075683594, | |
| "learning_rate": 0.00016586994637054486, | |
| "loss": 10.9491, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.3366114776546745, | |
| "grad_norm": 5.805549144744873, | |
| "learning_rate": 0.00016542252734555706, | |
| "loss": 12.0369, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.33760736960039833, | |
| "grad_norm": 7.639823913574219, | |
| "learning_rate": 0.0001649745292033135, | |
| "loss": 13.4881, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.33860326154612225, | |
| "grad_norm": 6.16525936126709, | |
| "learning_rate": 0.000164525958362103, | |
| "loss": 11.7652, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.33959915349184616, | |
| "grad_norm": 6.666656017303467, | |
| "learning_rate": 0.00016407682124841916, | |
| "loss": 12.8449, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.34059504543757, | |
| "grad_norm": 7.354873180389404, | |
| "learning_rate": 0.00016362712429686844, | |
| "loss": 10.8402, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.34159093738329394, | |
| "grad_norm": 6.7231526374816895, | |
| "learning_rate": 0.00016317687395007774, | |
| "loss": 13.0444, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.3425868293290178, | |
| "grad_norm": 6.6920485496521, | |
| "learning_rate": 0.0001627260766586023, | |
| "loss": 12.295, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.3435827212747417, | |
| "grad_norm": 5.398272514343262, | |
| "learning_rate": 0.00016227473888083318, | |
| "loss": 11.9999, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.34457861322046557, | |
| "grad_norm": 6.592508316040039, | |
| "learning_rate": 0.00016182286708290485, | |
| "loss": 11.0985, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.3455745051661895, | |
| "grad_norm": 6.421080589294434, | |
| "learning_rate": 0.00016137046773860242, | |
| "loss": 12.2588, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.34657039711191334, | |
| "grad_norm": 7.451521396636963, | |
| "learning_rate": 0.0001609175473292689, | |
| "loss": 11.8329, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.34756628905763726, | |
| "grad_norm": 7.403144836425781, | |
| "learning_rate": 0.00016046411234371249, | |
| "loss": 12.3238, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.3485621810033611, | |
| "grad_norm": 6.073176383972168, | |
| "learning_rate": 0.0001600101692781134, | |
| "loss": 11.3786, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.34955807294908503, | |
| "grad_norm": 5.982321262359619, | |
| "learning_rate": 0.00015955572463593093, | |
| "loss": 12.0612, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.3505539648948089, | |
| "grad_norm": 6.197265148162842, | |
| "learning_rate": 0.00015910078492781038, | |
| "loss": 13.0901, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.3515498568405328, | |
| "grad_norm": 5.7054443359375, | |
| "learning_rate": 0.00015864535667148953, | |
| "loss": 11.8859, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.3525457487862567, | |
| "grad_norm": 6.113707065582275, | |
| "learning_rate": 0.00015818944639170538, | |
| "loss": 12.4638, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.3535416407319806, | |
| "grad_norm": 5.500970840454102, | |
| "learning_rate": 0.00015773306062010083, | |
| "loss": 12.5954, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3545375326777045, | |
| "grad_norm": 6.778495788574219, | |
| "learning_rate": 0.00015727620589513084, | |
| "loss": 12.1178, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.35553342462342835, | |
| "grad_norm": 5.658638954162598, | |
| "learning_rate": 0.0001568188887619689, | |
| "loss": 12.3659, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.35652931656915227, | |
| "grad_norm": 5.579876899719238, | |
| "learning_rate": 0.0001563611157724132, | |
| "loss": 11.3356, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.3575252085148761, | |
| "grad_norm": 5.567670822143555, | |
| "learning_rate": 0.0001559028934847929, | |
| "loss": 11.3109, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.35852110046060004, | |
| "grad_norm": 5.965104579925537, | |
| "learning_rate": 0.00015544422846387398, | |
| "loss": 11.5413, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3595169924063239, | |
| "grad_norm": 5.479038715362549, | |
| "learning_rate": 0.00015498512728076536, | |
| "loss": 11.1271, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.3605128843520478, | |
| "grad_norm": 6.937322616577148, | |
| "learning_rate": 0.0001545255965128246, | |
| "loss": 12.4041, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.3615087762977717, | |
| "grad_norm": 6.604974269866943, | |
| "learning_rate": 0.00015406564274356377, | |
| "loss": 13.2388, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.3625046682434956, | |
| "grad_norm": 5.381167411804199, | |
| "learning_rate": 0.00015360527256255517, | |
| "loss": 12.4161, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.36350056018921945, | |
| "grad_norm": 6.176576137542725, | |
| "learning_rate": 0.00015314449256533677, | |
| "loss": 11.5011, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.36449645213494336, | |
| "grad_norm": 6.568368434906006, | |
| "learning_rate": 0.00015268330935331787, | |
| "loss": 12.6016, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.3654923440806673, | |
| "grad_norm": 6.896481513977051, | |
| "learning_rate": 0.00015222172953368446, | |
| "loss": 13.5208, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.36648823602639113, | |
| "grad_norm": 5.2163262367248535, | |
| "learning_rate": 0.0001517597597193046, | |
| "loss": 10.7047, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.36748412797211505, | |
| "grad_norm": 5.657341480255127, | |
| "learning_rate": 0.00015129740652863354, | |
| "loss": 11.5574, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.3684800199178389, | |
| "grad_norm": 6.786725997924805, | |
| "learning_rate": 0.0001508346765856191, | |
| "loss": 12.2539, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3694759118635628, | |
| "grad_norm": 5.714908599853516, | |
| "learning_rate": 0.00015037157651960677, | |
| "loss": 11.4831, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.3704718038092867, | |
| "grad_norm": 6.115590572357178, | |
| "learning_rate": 0.0001499081129652443, | |
| "loss": 11.983, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.3714676957550106, | |
| "grad_norm": 5.770305633544922, | |
| "learning_rate": 0.00014944429256238742, | |
| "loss": 12.0888, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.37246358770073446, | |
| "grad_norm": 6.408126354217529, | |
| "learning_rate": 0.000148980121956004, | |
| "loss": 12.551, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.37345947964645837, | |
| "grad_norm": 5.341275215148926, | |
| "learning_rate": 0.00014851560779607922, | |
| "loss": 10.8883, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.37445537159218223, | |
| "grad_norm": 5.945551872253418, | |
| "learning_rate": 0.00014805075673752022, | |
| "loss": 11.5723, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.37545126353790614, | |
| "grad_norm": 5.617563724517822, | |
| "learning_rate": 0.0001475855754400608, | |
| "loss": 11.3891, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.37644715548363, | |
| "grad_norm": 5.866842269897461, | |
| "learning_rate": 0.00014712007056816583, | |
| "loss": 11.773, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.3774430474293539, | |
| "grad_norm": 5.4498796463012695, | |
| "learning_rate": 0.00014665424879093598, | |
| "loss": 11.6255, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.3784389393750778, | |
| "grad_norm": 6.97362756729126, | |
| "learning_rate": 0.0001461881167820121, | |
| "loss": 11.3264, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3794348313208017, | |
| "grad_norm": 6.620206832885742, | |
| "learning_rate": 0.0001457216812194796, | |
| "loss": 12.2684, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.3804307232665256, | |
| "grad_norm": 6.0084228515625, | |
| "learning_rate": 0.00014525494878577278, | |
| "loss": 12.1192, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.38142661521224946, | |
| "grad_norm": 6.702908515930176, | |
| "learning_rate": 0.00014478792616757908, | |
| "loss": 11.2421, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.3824225071579734, | |
| "grad_norm": 5.771916389465332, | |
| "learning_rate": 0.00014432062005574332, | |
| "loss": 12.0659, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.38341839910369724, | |
| "grad_norm": 6.246331691741943, | |
| "learning_rate": 0.00014385303714517175, | |
| "loss": 12.3364, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.38441429104942115, | |
| "grad_norm": 6.237067699432373, | |
| "learning_rate": 0.00014338518413473632, | |
| "loss": 10.9562, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.385410182995145, | |
| "grad_norm": 5.690719127655029, | |
| "learning_rate": 0.00014291706772717847, | |
| "loss": 11.8223, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.3864060749408689, | |
| "grad_norm": 6.914798259735107, | |
| "learning_rate": 0.00014244869462901331, | |
| "loss": 12.0813, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.3874019668865928, | |
| "grad_norm": 6.975170612335205, | |
| "learning_rate": 0.00014198007155043343, | |
| "loss": 10.7795, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.3883978588323167, | |
| "grad_norm": 5.881531715393066, | |
| "learning_rate": 0.00014151120520521283, | |
| "loss": 11.8511, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.38939375077804056, | |
| "grad_norm": 5.874497413635254, | |
| "learning_rate": 0.00014104210231061053, | |
| "loss": 11.0519, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.3903896427237645, | |
| "grad_norm": 6.757877826690674, | |
| "learning_rate": 0.00014057276958727468, | |
| "loss": 12.7317, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.39138553466948833, | |
| "grad_norm": 7.4160637855529785, | |
| "learning_rate": 0.0001401032137591461, | |
| "loss": 11.9433, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.39238142661521225, | |
| "grad_norm": 5.124035835266113, | |
| "learning_rate": 0.00013963344155336178, | |
| "loss": 10.2948, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.39337731856093616, | |
| "grad_norm": 6.715900897979736, | |
| "learning_rate": 0.00013916345970015875, | |
| "loss": 12.0971, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.39437321050666, | |
| "grad_norm": 6.183508396148682, | |
| "learning_rate": 0.00013869327493277762, | |
| "loss": 12.2012, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.39536910245238394, | |
| "grad_norm": 6.416903018951416, | |
| "learning_rate": 0.000138222893987366, | |
| "loss": 12.4187, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.3963649943981078, | |
| "grad_norm": 6.140800476074219, | |
| "learning_rate": 0.00013775232360288214, | |
| "loss": 11.6096, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.3973608863438317, | |
| "grad_norm": 6.360043048858643, | |
| "learning_rate": 0.00013728157052099823, | |
| "loss": 12.4352, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.39835677828955557, | |
| "grad_norm": 5.971815586090088, | |
| "learning_rate": 0.00013681064148600392, | |
| "loss": 10.6359, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.39835677828955557, | |
| "eval_loss": 1.5056407451629639, | |
| "eval_runtime": 289.8388, | |
| "eval_samples_per_second": 4.62, | |
| "eval_steps_per_second": 1.156, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3993526702352795, | |
| "grad_norm": 6.695412635803223, | |
| "learning_rate": 0.00013633954324470968, | |
| "loss": 11.4601, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.40034856218100334, | |
| "grad_norm": 6.6551337242126465, | |
| "learning_rate": 0.00013586828254635015, | |
| "loss": 12.2329, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.40134445412672726, | |
| "grad_norm": 6.626735687255859, | |
| "learning_rate": 0.0001353968661424873, | |
| "loss": 12.1324, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.4023403460724511, | |
| "grad_norm": 6.322345733642578, | |
| "learning_rate": 0.00013492530078691403, | |
| "loss": 12.0786, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.40333623801817503, | |
| "grad_norm": 4.97235631942749, | |
| "learning_rate": 0.00013445359323555712, | |
| "loss": 10.4874, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.4043321299638989, | |
| "grad_norm": 6.183226108551025, | |
| "learning_rate": 0.0001339817502463804, | |
| "loss": 12.173, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.4053280219096228, | |
| "grad_norm": 6.04211950302124, | |
| "learning_rate": 0.00013350977857928836, | |
| "loss": 11.1317, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.4063239138553467, | |
| "grad_norm": 5.817178726196289, | |
| "learning_rate": 0.0001330376849960287, | |
| "loss": 12.1543, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.4073198058010706, | |
| "grad_norm": 7.050534248352051, | |
| "learning_rate": 0.0001325654762600959, | |
| "loss": 12.6231, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.4083156977467945, | |
| "grad_norm": 6.723251819610596, | |
| "learning_rate": 0.00013209315913663427, | |
| "loss": 12.7058, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.40931158969251835, | |
| "grad_norm": 4.917559623718262, | |
| "learning_rate": 0.0001316207403923408, | |
| "loss": 10.9655, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.41030748163824227, | |
| "grad_norm": 6.4542107582092285, | |
| "learning_rate": 0.00013114822679536836, | |
| "loss": 13.6992, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.4113033735839661, | |
| "grad_norm": 6.2309346199035645, | |
| "learning_rate": 0.0001306756251152289, | |
| "loss": 12.3255, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.41229926552969004, | |
| "grad_norm": 6.741308689117432, | |
| "learning_rate": 0.00013020294212269615, | |
| "loss": 13.0598, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.4132951574754139, | |
| "grad_norm": 6.19504976272583, | |
| "learning_rate": 0.0001297301845897088, | |
| "loss": 14.3294, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4142910494211378, | |
| "grad_norm": 6.095429420471191, | |
| "learning_rate": 0.0001292573592892735, | |
| "loss": 12.5742, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.4152869413668617, | |
| "grad_norm": 5.67078161239624, | |
| "learning_rate": 0.00012878447299536768, | |
| "loss": 12.8696, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.4162828333125856, | |
| "grad_norm": 5.176650524139404, | |
| "learning_rate": 0.00012831153248284272, | |
| "loss": 12.2691, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.41727872525830945, | |
| "grad_norm": 6.270726680755615, | |
| "learning_rate": 0.00012783854452732668, | |
| "loss": 11.5467, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.41827461720403336, | |
| "grad_norm": 6.120121002197266, | |
| "learning_rate": 0.00012736551590512737, | |
| "loss": 12.1084, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4192705091497573, | |
| "grad_norm": 5.2254862785339355, | |
| "learning_rate": 0.00012689245339313521, | |
| "loss": 12.1979, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.42026640109548113, | |
| "grad_norm": 5.770168781280518, | |
| "learning_rate": 0.00012641936376872606, | |
| "loss": 12.6844, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.42126229304120505, | |
| "grad_norm": 6.377243995666504, | |
| "learning_rate": 0.00012594625380966436, | |
| "loss": 12.6859, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.4222581849869289, | |
| "grad_norm": 5.389707565307617, | |
| "learning_rate": 0.00012547313029400567, | |
| "loss": 11.773, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.4232540769326528, | |
| "grad_norm": 5.801501750946045, | |
| "learning_rate": 0.000125, | |
| "loss": 12.0606, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.4242499688783767, | |
| "grad_norm": 6.649625301361084, | |
| "learning_rate": 0.0001245268697059943, | |
| "loss": 12.1181, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.4252458608241006, | |
| "grad_norm": 6.166293144226074, | |
| "learning_rate": 0.0001240537461903357, | |
| "loss": 12.4382, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.42624175276982446, | |
| "grad_norm": 5.566735744476318, | |
| "learning_rate": 0.00012358063623127394, | |
| "loss": 11.7602, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.42723764471554837, | |
| "grad_norm": 6.631081581115723, | |
| "learning_rate": 0.00012310754660686482, | |
| "loss": 12.4029, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.42823353666127223, | |
| "grad_norm": 5.083254814147949, | |
| "learning_rate": 0.00012263448409487266, | |
| "loss": 11.5772, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.42922942860699614, | |
| "grad_norm": 5.555343151092529, | |
| "learning_rate": 0.00012216145547267333, | |
| "loss": 10.4996, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.43022532055272, | |
| "grad_norm": 4.766176223754883, | |
| "learning_rate": 0.00012168846751715729, | |
| "loss": 11.4977, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.4312212124984439, | |
| "grad_norm": 5.567246437072754, | |
| "learning_rate": 0.00012121552700463235, | |
| "loss": 11.6256, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.43221710444416783, | |
| "grad_norm": 6.13475227355957, | |
| "learning_rate": 0.00012074264071072653, | |
| "loss": 12.4283, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.4332129963898917, | |
| "grad_norm": 5.974978923797607, | |
| "learning_rate": 0.00012026981541029122, | |
| "loss": 12.2817, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4342088883356156, | |
| "grad_norm": 5.512807846069336, | |
| "learning_rate": 0.00011979705787730388, | |
| "loss": 11.7294, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.43520478028133947, | |
| "grad_norm": 5.102151393890381, | |
| "learning_rate": 0.00011932437488477113, | |
| "loss": 9.8982, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.4362006722270634, | |
| "grad_norm": 6.0491623878479, | |
| "learning_rate": 0.00011885177320463165, | |
| "loss": 11.6918, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.43719656417278724, | |
| "grad_norm": 6.371288299560547, | |
| "learning_rate": 0.00011837925960765928, | |
| "loss": 12.9814, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.43819245611851115, | |
| "grad_norm": 5.732490062713623, | |
| "learning_rate": 0.00011790684086336576, | |
| "loss": 12.1104, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.439188348064235, | |
| "grad_norm": 6.025941848754883, | |
| "learning_rate": 0.00011743452373990408, | |
| "loss": 12.2868, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.4401842400099589, | |
| "grad_norm": 7.7239603996276855, | |
| "learning_rate": 0.00011696231500397135, | |
| "loss": 12.834, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.4411801319556828, | |
| "grad_norm": 6.518691062927246, | |
| "learning_rate": 0.00011649022142071167, | |
| "loss": 12.9192, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.4421760239014067, | |
| "grad_norm": 5.347179889678955, | |
| "learning_rate": 0.00011601824975361959, | |
| "loss": 10.4923, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.44317191584713056, | |
| "grad_norm": 6.918272495269775, | |
| "learning_rate": 0.00011554640676444295, | |
| "loss": 11.6119, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4441678077928545, | |
| "grad_norm": 6.190727710723877, | |
| "learning_rate": 0.00011507469921308598, | |
| "loss": 13.9115, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.4451636997385784, | |
| "grad_norm": 6.001166343688965, | |
| "learning_rate": 0.0001146031338575127, | |
| "loss": 11.5416, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.44615959168430225, | |
| "grad_norm": 9.23908519744873, | |
| "learning_rate": 0.00011413171745364992, | |
| "loss": 13.9503, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.44715548363002616, | |
| "grad_norm": 6.2685723304748535, | |
| "learning_rate": 0.00011366045675529033, | |
| "loss": 10.9264, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.44815137557575, | |
| "grad_norm": 5.3125901222229, | |
| "learning_rate": 0.0001131893585139961, | |
| "loss": 11.8189, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.44914726752147394, | |
| "grad_norm": 5.640747547149658, | |
| "learning_rate": 0.00011271842947900179, | |
| "loss": 10.8639, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.4501431594671978, | |
| "grad_norm": 6.120314598083496, | |
| "learning_rate": 0.00011224767639711789, | |
| "loss": 13.9796, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.4511390514129217, | |
| "grad_norm": 5.786348342895508, | |
| "learning_rate": 0.000111777106012634, | |
| "loss": 12.8427, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.45213494335864557, | |
| "grad_norm": 5.406647682189941, | |
| "learning_rate": 0.00011130672506722242, | |
| "loss": 11.3131, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.4531308353043695, | |
| "grad_norm": 6.1720099449157715, | |
| "learning_rate": 0.00011083654029984128, | |
| "loss": 12.1748, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.45412672725009334, | |
| "grad_norm": 6.570352554321289, | |
| "learning_rate": 0.00011036655844663824, | |
| "loss": 12.7982, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.45512261919581726, | |
| "grad_norm": 5.541633129119873, | |
| "learning_rate": 0.00010989678624085394, | |
| "loss": 11.4866, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.4561185111415411, | |
| "grad_norm": 6.466024875640869, | |
| "learning_rate": 0.00010942723041272531, | |
| "loss": 12.9477, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.45711440308726503, | |
| "grad_norm": 4.921136379241943, | |
| "learning_rate": 0.00010895789768938948, | |
| "loss": 10.5616, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.45811029503298895, | |
| "grad_norm": 7.542169094085693, | |
| "learning_rate": 0.00010848879479478724, | |
| "loss": 13.1838, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.4591061869787128, | |
| "grad_norm": 6.347965240478516, | |
| "learning_rate": 0.00010801992844956659, | |
| "loss": 12.3061, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.4601020789244367, | |
| "grad_norm": 5.8513689041137695, | |
| "learning_rate": 0.00010755130537098669, | |
| "loss": 11.7932, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.4610979708701606, | |
| "grad_norm": 6.4326677322387695, | |
| "learning_rate": 0.00010708293227282158, | |
| "loss": 11.9086, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.4620938628158845, | |
| "grad_norm": 5.559072494506836, | |
| "learning_rate": 0.00010661481586526371, | |
| "loss": 11.826, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.46308975476160835, | |
| "grad_norm": 5.315805435180664, | |
| "learning_rate": 0.00010614696285482828, | |
| "loss": 12.2405, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.46408564670733227, | |
| "grad_norm": 5.958090305328369, | |
| "learning_rate": 0.00010567937994425675, | |
| "loss": 10.8293, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.4650815386530561, | |
| "grad_norm": 5.258885860443115, | |
| "learning_rate": 0.00010521207383242094, | |
| "loss": 10.2966, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.46607743059878004, | |
| "grad_norm": 5.901886940002441, | |
| "learning_rate": 0.00010474505121422722, | |
| "loss": 11.8841, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.4670733225445039, | |
| "grad_norm": 5.9542341232299805, | |
| "learning_rate": 0.00010427831878052043, | |
| "loss": 13.1809, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.4680692144902278, | |
| "grad_norm": 5.833163738250732, | |
| "learning_rate": 0.00010381188321798792, | |
| "loss": 11.8099, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4690651064359517, | |
| "grad_norm": 5.479732513427734, | |
| "learning_rate": 0.00010334575120906404, | |
| "loss": 11.7125, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.4700609983816756, | |
| "grad_norm": 5.714720726013184, | |
| "learning_rate": 0.00010287992943183422, | |
| "loss": 11.6581, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.4710568903273995, | |
| "grad_norm": 5.256033420562744, | |
| "learning_rate": 0.00010241442455993925, | |
| "loss": 10.8862, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.47205278227312336, | |
| "grad_norm": 5.788590908050537, | |
| "learning_rate": 0.00010194924326247976, | |
| "loss": 12.4129, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.4730486742188473, | |
| "grad_norm": 6.205699443817139, | |
| "learning_rate": 0.00010148439220392081, | |
| "loss": 12.2752, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.47404456616457114, | |
| "grad_norm": 5.725231170654297, | |
| "learning_rate": 0.00010101987804399601, | |
| "loss": 12.1739, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.47504045811029505, | |
| "grad_norm": 5.984923362731934, | |
| "learning_rate": 0.00010055570743761256, | |
| "loss": 10.9013, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.4760363500560189, | |
| "grad_norm": 6.258663654327393, | |
| "learning_rate": 0.00010009188703475571, | |
| "loss": 11.7148, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.4770322420017428, | |
| "grad_norm": 5.729332447052002, | |
| "learning_rate": 9.962842348039328e-05, | |
| "loss": 11.2752, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.4780281339474667, | |
| "grad_norm": 5.4574995040893555, | |
| "learning_rate": 9.916532341438088e-05, | |
| "loss": 12.89, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4790240258931906, | |
| "grad_norm": 6.573443412780762, | |
| "learning_rate": 9.87025934713665e-05, | |
| "loss": 12.5011, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.48001991783891446, | |
| "grad_norm": 6.030612468719482, | |
| "learning_rate": 9.824024028069541e-05, | |
| "loss": 12.3289, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.48101580978463837, | |
| "grad_norm": 6.0058698654174805, | |
| "learning_rate": 9.777827046631553e-05, | |
| "loss": 13.7869, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.48201170173036223, | |
| "grad_norm": 5.5297040939331055, | |
| "learning_rate": 9.731669064668217e-05, | |
| "loss": 12.0592, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.48300759367608614, | |
| "grad_norm": 5.423988342285156, | |
| "learning_rate": 9.685550743466325e-05, | |
| "loss": 10.2873, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.48400348562181006, | |
| "grad_norm": 5.130181789398193, | |
| "learning_rate": 9.639472743744486e-05, | |
| "loss": 9.5329, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.4849993775675339, | |
| "grad_norm": 6.554098606109619, | |
| "learning_rate": 9.593435725643623e-05, | |
| "loss": 13.0928, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.48599526951325783, | |
| "grad_norm": 5.409817218780518, | |
| "learning_rate": 9.547440348717542e-05, | |
| "loss": 11.0439, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.4869911614589817, | |
| "grad_norm": 5.266180038452148, | |
| "learning_rate": 9.501487271923463e-05, | |
| "loss": 10.6856, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.4879870534047056, | |
| "grad_norm": 6.0554633140563965, | |
| "learning_rate": 9.455577153612602e-05, | |
| "loss": 11.6784, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.48898294535042947, | |
| "grad_norm": 5.316808223724365, | |
| "learning_rate": 9.40971065152071e-05, | |
| "loss": 9.9632, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.4899788372961534, | |
| "grad_norm": 5.960623264312744, | |
| "learning_rate": 9.363888422758678e-05, | |
| "loss": 11.169, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.49097472924187724, | |
| "grad_norm": 6.7030253410339355, | |
| "learning_rate": 9.318111123803113e-05, | |
| "loss": 12.5777, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.49197062118760115, | |
| "grad_norm": 5.716718673706055, | |
| "learning_rate": 9.272379410486917e-05, | |
| "loss": 12.1007, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.492966513133325, | |
| "grad_norm": 5.541021823883057, | |
| "learning_rate": 9.226693937989916e-05, | |
| "loss": 11.0471, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.4939624050790489, | |
| "grad_norm": 6.103461742401123, | |
| "learning_rate": 9.181055360829463e-05, | |
| "loss": 11.6358, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.4949582970247728, | |
| "grad_norm": 5.171212196350098, | |
| "learning_rate": 9.135464332851049e-05, | |
| "loss": 11.737, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.4959541889704967, | |
| "grad_norm": 6.258893013000488, | |
| "learning_rate": 9.089921507218962e-05, | |
| "loss": 12.2711, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.4969500809162206, | |
| "grad_norm": 5.942646026611328, | |
| "learning_rate": 9.044427536406909e-05, | |
| "loss": 12.4533, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.4979459728619445, | |
| "grad_norm": 5.429832458496094, | |
| "learning_rate": 8.998983072188663e-05, | |
| "loss": 11.4921, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4979459728619445, | |
| "eval_loss": 1.483846664428711, | |
| "eval_runtime": 289.886, | |
| "eval_samples_per_second": 4.619, | |
| "eval_steps_per_second": 1.156, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4989418648076684, | |
| "grad_norm": 6.279896259307861, | |
| "learning_rate": 8.953588765628753e-05, | |
| "loss": 12.5236, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.49993775675339225, | |
| "grad_norm": 7.839783668518066, | |
| "learning_rate": 8.90824526707311e-05, | |
| "loss": 11.226, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.5009336486991162, | |
| "grad_norm": 5.284313678741455, | |
| "learning_rate": 8.86295322613976e-05, | |
| "loss": 10.7881, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.5019295406448401, | |
| "grad_norm": 5.617724418640137, | |
| "learning_rate": 8.817713291709513e-05, | |
| "loss": 12.3775, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.5029254325905639, | |
| "grad_norm": 5.214620113372803, | |
| "learning_rate": 8.772526111916685e-05, | |
| "loss": 11.2123, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5039213245362878, | |
| "grad_norm": 6.912374973297119, | |
| "learning_rate": 8.727392334139771e-05, | |
| "loss": 13.137, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.5049172164820117, | |
| "grad_norm": 6.51848840713501, | |
| "learning_rate": 8.682312604992227e-05, | |
| "loss": 11.9434, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.5059131084277356, | |
| "grad_norm": 5.211724758148193, | |
| "learning_rate": 8.637287570313158e-05, | |
| "loss": 10.2712, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.5069090003734594, | |
| "grad_norm": 5.400852203369141, | |
| "learning_rate": 8.592317875158085e-05, | |
| "loss": 10.2705, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.5079048923191833, | |
| "grad_norm": 5.548620223999023, | |
| "learning_rate": 8.5474041637897e-05, | |
| "loss": 10.4555, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5089007842649073, | |
| "grad_norm": 6.38204288482666, | |
| "learning_rate": 8.502547079668653e-05, | |
| "loss": 11.8434, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.5098966762106312, | |
| "grad_norm": 5.735969543457031, | |
| "learning_rate": 8.457747265444296e-05, | |
| "loss": 10.9467, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.5108925681563551, | |
| "grad_norm": 5.540611743927002, | |
| "learning_rate": 8.413005362945512e-05, | |
| "loss": 12.3345, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.5118884601020789, | |
| "grad_norm": 6.0366692543029785, | |
| "learning_rate": 8.368322013171513e-05, | |
| "loss": 11.6241, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.5128843520478028, | |
| "grad_norm": 5.270688056945801, | |
| "learning_rate": 8.323697856282627e-05, | |
| "loss": 10.389, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5138802439935267, | |
| "grad_norm": 6.122136116027832, | |
| "learning_rate": 8.27913353159117e-05, | |
| "loss": 10.98, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.5148761359392506, | |
| "grad_norm": 6.037084579467773, | |
| "learning_rate": 8.234629677552263e-05, | |
| "loss": 12.568, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.5158720278849744, | |
| "grad_norm": 7.1887125968933105, | |
| "learning_rate": 8.190186931754673e-05, | |
| "loss": 12.7303, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.5168679198306984, | |
| "grad_norm": 5.168606758117676, | |
| "learning_rate": 8.14580593091171e-05, | |
| "loss": 11.5019, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.5178638117764223, | |
| "grad_norm": 5.976624965667725, | |
| "learning_rate": 8.101487310852087e-05, | |
| "loss": 12.8172, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5188597037221462, | |
| "grad_norm": 5.563510417938232, | |
| "learning_rate": 8.057231706510807e-05, | |
| "loss": 11.5814, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.51985559566787, | |
| "grad_norm": 5.869482040405273, | |
| "learning_rate": 8.013039751920074e-05, | |
| "loss": 12.1924, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.5208514876135939, | |
| "grad_norm": 6.094331741333008, | |
| "learning_rate": 7.968912080200211e-05, | |
| "loss": 11.5143, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.5218473795593178, | |
| "grad_norm": 5.655307769775391, | |
| "learning_rate": 7.924849323550581e-05, | |
| "loss": 11.5072, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.5228432715050417, | |
| "grad_norm": 5.7834367752075195, | |
| "learning_rate": 7.880852113240539e-05, | |
| "loss": 11.97, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5238391634507656, | |
| "grad_norm": 7.464441776275635, | |
| "learning_rate": 7.836921079600384e-05, | |
| "loss": 12.2958, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.5248350553964894, | |
| "grad_norm": 5.94898796081543, | |
| "learning_rate": 7.793056852012324e-05, | |
| "loss": 12.3025, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.5258309473422134, | |
| "grad_norm": 5.301365852355957, | |
| "learning_rate": 7.749260058901467e-05, | |
| "loss": 10.7913, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.5268268392879373, | |
| "grad_norm": 5.742236137390137, | |
| "learning_rate": 7.705531327726824e-05, | |
| "loss": 11.6252, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.5278227312336612, | |
| "grad_norm": 5.871049404144287, | |
| "learning_rate": 7.661871284972286e-05, | |
| "loss": 12.7432, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.528818623179385, | |
| "grad_norm": 5.6256866455078125, | |
| "learning_rate": 7.618280556137697e-05, | |
| "loss": 11.5549, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.5298145151251089, | |
| "grad_norm": 6.3018903732299805, | |
| "learning_rate": 7.574759765729853e-05, | |
| "loss": 11.8701, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.5308104070708328, | |
| "grad_norm": 5.372267246246338, | |
| "learning_rate": 7.531309537253574e-05, | |
| "loss": 10.7988, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.5318062990165567, | |
| "grad_norm": 5.44589376449585, | |
| "learning_rate": 7.487930493202783e-05, | |
| "loss": 12.7583, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.5328021909622805, | |
| "grad_norm": 6.493985176086426, | |
| "learning_rate": 7.444623255051538e-05, | |
| "loss": 12.9094, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5337980829080045, | |
| "grad_norm": 5.483232021331787, | |
| "learning_rate": 7.4013884432452e-05, | |
| "loss": 12.4428, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.5347939748537284, | |
| "grad_norm": 6.090009689331055, | |
| "learning_rate": 7.358226677191488e-05, | |
| "loss": 13.0753, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.5357898667994523, | |
| "grad_norm": 5.5943522453308105, | |
| "learning_rate": 7.315138575251617e-05, | |
| "loss": 10.7217, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.5367857587451762, | |
| "grad_norm": 5.108747482299805, | |
| "learning_rate": 7.272124754731468e-05, | |
| "loss": 10.47, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.5377816506909, | |
| "grad_norm": 7.087966442108154, | |
| "learning_rate": 7.229185831872701e-05, | |
| "loss": 11.7998, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5387775426366239, | |
| "grad_norm": 5.637360095977783, | |
| "learning_rate": 7.18632242184396e-05, | |
| "loss": 11.9396, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.5397734345823478, | |
| "grad_norm": 5.862568378448486, | |
| "learning_rate": 7.143535138732044e-05, | |
| "loss": 11.211, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.5407693265280717, | |
| "grad_norm": 5.098068714141846, | |
| "learning_rate": 7.100824595533109e-05, | |
| "loss": 11.7145, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.5417652184737956, | |
| "grad_norm": 5.469205856323242, | |
| "learning_rate": 7.05819140414389e-05, | |
| "loss": 11.3163, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.5427611104195195, | |
| "grad_norm": 4.86387825012207, | |
| "learning_rate": 7.015636175352935e-05, | |
| "loss": 12.6151, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5437570023652434, | |
| "grad_norm": 5.209597110748291, | |
| "learning_rate": 6.973159518831865e-05, | |
| "loss": 10.7285, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.5447528943109673, | |
| "grad_norm": 6.904428958892822, | |
| "learning_rate": 6.930762043126598e-05, | |
| "loss": 12.6723, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.5457487862566911, | |
| "grad_norm": 5.574467658996582, | |
| "learning_rate": 6.888444355648694e-05, | |
| "loss": 11.2369, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.546744678202415, | |
| "grad_norm": 5.7822184562683105, | |
| "learning_rate": 6.846207062666598e-05, | |
| "loss": 11.7722, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.5477405701481389, | |
| "grad_norm": 5.367188930511475, | |
| "learning_rate": 6.804050769296982e-05, | |
| "loss": 11.7805, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5487364620938628, | |
| "grad_norm": 4.832945823669434, | |
| "learning_rate": 6.761976079496069e-05, | |
| "loss": 10.9897, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.5497323540395868, | |
| "grad_norm": 6.3430962562561035, | |
| "learning_rate": 6.719983596050979e-05, | |
| "loss": 13.1188, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.5507282459853106, | |
| "grad_norm": 5.310800075531006, | |
| "learning_rate": 6.678073920571101e-05, | |
| "loss": 11.8806, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 5.725086688995361, | |
| "learning_rate": 6.636247653479451e-05, | |
| "loss": 12.9238, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.5527200298767584, | |
| "grad_norm": 6.0460920333862305, | |
| "learning_rate": 6.594505394004116e-05, | |
| "loss": 12.2236, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.5537159218224823, | |
| "grad_norm": 5.62431526184082, | |
| "learning_rate": 6.552847740169603e-05, | |
| "loss": 12.1332, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.5547118137682061, | |
| "grad_norm": 5.583924770355225, | |
| "learning_rate": 6.511275288788341e-05, | |
| "loss": 11.2317, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.55570770571393, | |
| "grad_norm": 5.76798677444458, | |
| "learning_rate": 6.469788635452085e-05, | |
| "loss": 11.5159, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.5567035976596539, | |
| "grad_norm": 4.825177192687988, | |
| "learning_rate": 6.428388374523377e-05, | |
| "loss": 10.9255, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.5576994896053779, | |
| "grad_norm": 5.40618896484375, | |
| "learning_rate": 6.387075099127085e-05, | |
| "loss": 12.1328, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5586953815511017, | |
| "grad_norm": 5.495939254760742, | |
| "learning_rate": 6.345849401141848e-05, | |
| "loss": 12.011, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.5596912734968256, | |
| "grad_norm": 4.913009166717529, | |
| "learning_rate": 6.304711871191624e-05, | |
| "loss": 11.018, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.5606871654425495, | |
| "grad_norm": 6.538085460662842, | |
| "learning_rate": 6.263663098637223e-05, | |
| "loss": 12.6505, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.5616830573882734, | |
| "grad_norm": 5.621396064758301, | |
| "learning_rate": 6.222703671567862e-05, | |
| "loss": 11.4574, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.5626789493339973, | |
| "grad_norm": 5.406931400299072, | |
| "learning_rate": 6.181834176792743e-05, | |
| "loss": 11.9075, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.5636748412797211, | |
| "grad_norm": 5.263576030731201, | |
| "learning_rate": 6.14105519983264e-05, | |
| "loss": 12.0203, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.564670733225445, | |
| "grad_norm": 5.000280857086182, | |
| "learning_rate": 6.100367324911534e-05, | |
| "loss": 11.9682, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.565666625171169, | |
| "grad_norm": 4.999345779418945, | |
| "learning_rate": 6.059771134948186e-05, | |
| "loss": 11.1666, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.5666625171168929, | |
| "grad_norm": 4.500539779663086, | |
| "learning_rate": 6.019267211547863e-05, | |
| "loss": 10.7409, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.5676584090626167, | |
| "grad_norm": 4.902435779571533, | |
| "learning_rate": 5.978856134993944e-05, | |
| "loss": 10.0251, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5686543010083406, | |
| "grad_norm": 4.8086066246032715, | |
| "learning_rate": 5.938538484239635e-05, | |
| "loss": 10.9553, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.5696501929540645, | |
| "grad_norm": 4.757142543792725, | |
| "learning_rate": 5.898314836899664e-05, | |
| "loss": 12.2329, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.5706460848997884, | |
| "grad_norm": 5.331392288208008, | |
| "learning_rate": 5.8581857692420166e-05, | |
| "loss": 11.5873, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.5716419768455122, | |
| "grad_norm": 6.0972723960876465, | |
| "learning_rate": 5.8181518561796684e-05, | |
| "loss": 13.181, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.5726378687912361, | |
| "grad_norm": 5.213257312774658, | |
| "learning_rate": 5.7782136712623484e-05, | |
| "loss": 12.3342, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.57363376073696, | |
| "grad_norm": 5.662869453430176, | |
| "learning_rate": 5.7383717866683506e-05, | |
| "loss": 12.8475, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.574629652682684, | |
| "grad_norm": 5.137444972991943, | |
| "learning_rate": 5.6986267731962766e-05, | |
| "loss": 12.0747, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.5756255446284079, | |
| "grad_norm": 5.8213677406311035, | |
| "learning_rate": 5.6589792002569264e-05, | |
| "loss": 11.4558, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.5766214365741317, | |
| "grad_norm": 5.571473598480225, | |
| "learning_rate": 5.6194296358650935e-05, | |
| "loss": 11.3248, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.5776173285198556, | |
| "grad_norm": 6.315218925476074, | |
| "learning_rate": 5.5799786466314435e-05, | |
| "loss": 14.3275, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5786132204655795, | |
| "grad_norm": 5.514521598815918, | |
| "learning_rate": 5.5406267977544e-05, | |
| "loss": 11.3491, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.5796091124113034, | |
| "grad_norm": 5.521172523498535, | |
| "learning_rate": 5.501374653012038e-05, | |
| "loss": 12.0774, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.5806050043570272, | |
| "grad_norm": 4.984793663024902, | |
| "learning_rate": 5.462222774754014e-05, | |
| "loss": 11.1853, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.5816008963027511, | |
| "grad_norm": 5.918780326843262, | |
| "learning_rate": 5.423171723893501e-05, | |
| "loss": 11.6143, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.582596788248475, | |
| "grad_norm": 5.09220027923584, | |
| "learning_rate": 5.384222059899178e-05, | |
| "loss": 11.6102, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.583592680194199, | |
| "grad_norm": 5.182462692260742, | |
| "learning_rate": 5.345374340787168e-05, | |
| "loss": 11.4592, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.5845885721399228, | |
| "grad_norm": 5.29030179977417, | |
| "learning_rate": 5.306629123113084e-05, | |
| "loss": 10.9399, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.5855844640856467, | |
| "grad_norm": 5.2069807052612305, | |
| "learning_rate": 5.2679869619640555e-05, | |
| "loss": 11.6267, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.5865803560313706, | |
| "grad_norm": 5.4248809814453125, | |
| "learning_rate": 5.229448410950732e-05, | |
| "loss": 10.9395, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.5875762479770945, | |
| "grad_norm": 5.220003128051758, | |
| "learning_rate": 5.1910140221994114e-05, | |
| "loss": 12.5726, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5885721399228184, | |
| "grad_norm": 5.750555038452148, | |
| "learning_rate": 5.152684346344087e-05, | |
| "loss": 11.9805, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.5895680318685422, | |
| "grad_norm": 5.346360683441162, | |
| "learning_rate": 5.114459932518573e-05, | |
| "loss": 11.6031, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.5905639238142661, | |
| "grad_norm": 5.487764358520508, | |
| "learning_rate": 5.0763413283486394e-05, | |
| "loss": 11.916, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.5915598157599901, | |
| "grad_norm": 6.088346004486084, | |
| "learning_rate": 5.038329079944165e-05, | |
| "loss": 11.9875, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.592555707705714, | |
| "grad_norm": 5.190619468688965, | |
| "learning_rate": 5.000423731891307e-05, | |
| "loss": 12.271, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.5935515996514378, | |
| "grad_norm": 5.797019004821777, | |
| "learning_rate": 4.962625827244707e-05, | |
| "loss": 12.3821, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.5945474915971617, | |
| "grad_norm": 5.294320106506348, | |
| "learning_rate": 4.9249359075197204e-05, | |
| "loss": 11.1323, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.5955433835428856, | |
| "grad_norm": 5.5578413009643555, | |
| "learning_rate": 4.8873545126846195e-05, | |
| "loss": 12.2285, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.5965392754886095, | |
| "grad_norm": 6.470888137817383, | |
| "learning_rate": 4.849882181152911e-05, | |
| "loss": 13.0034, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.5975351674343333, | |
| "grad_norm": 5.115138053894043, | |
| "learning_rate": 4.81251944977558e-05, | |
| "loss": 11.1868, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5975351674343333, | |
| "eval_loss": 1.469951868057251, | |
| "eval_runtime": 289.9264, | |
| "eval_samples_per_second": 4.618, | |
| "eval_steps_per_second": 1.155, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5985310593800572, | |
| "grad_norm": 6.162230014801025, | |
| "learning_rate": 4.775266853833421e-05, | |
| "loss": 12.9121, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.5995269513257812, | |
| "grad_norm": 5.4764275550842285, | |
| "learning_rate": 4.738124927029358e-05, | |
| "loss": 10.9161, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.6005228432715051, | |
| "grad_norm": 5.016139507293701, | |
| "learning_rate": 4.7010942014808056e-05, | |
| "loss": 11.407, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.601518735217229, | |
| "grad_norm": 5.546393394470215, | |
| "learning_rate": 4.664175207712043e-05, | |
| "loss": 12.3953, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.6025146271629528, | |
| "grad_norm": 5.607834815979004, | |
| "learning_rate": 4.627368474646608e-05, | |
| "loss": 13.2932, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6035105191086767, | |
| "grad_norm": 5.625347137451172, | |
| "learning_rate": 4.590674529599742e-05, | |
| "loss": 10.9166, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.6045064110544006, | |
| "grad_norm": 5.0566301345825195, | |
| "learning_rate": 4.554093898270788e-05, | |
| "loss": 12.1924, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.6055023030001245, | |
| "grad_norm": 4.960422515869141, | |
| "learning_rate": 4.5176271047357196e-05, | |
| "loss": 12.0646, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.6064981949458483, | |
| "grad_norm": 5.206432819366455, | |
| "learning_rate": 4.4812746714395866e-05, | |
| "loss": 11.2001, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.6074940868915722, | |
| "grad_norm": 6.366247177124023, | |
| "learning_rate": 4.4450371191890365e-05, | |
| "loss": 12.0904, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6084899788372962, | |
| "grad_norm": 6.073464393615723, | |
| "learning_rate": 4.408914967144888e-05, | |
| "loss": 11.4838, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.6094858707830201, | |
| "grad_norm": 5.903442859649658, | |
| "learning_rate": 4.372908732814647e-05, | |
| "loss": 11.7165, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.6104817627287439, | |
| "grad_norm": 6.564481735229492, | |
| "learning_rate": 4.3370189320451195e-05, | |
| "loss": 12.9723, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.6114776546744678, | |
| "grad_norm": 6.052420139312744, | |
| "learning_rate": 4.301246079015016e-05, | |
| "loss": 12.628, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.6124735466201917, | |
| "grad_norm": 5.082505226135254, | |
| "learning_rate": 4.26559068622758e-05, | |
| "loss": 11.0174, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6134694385659156, | |
| "grad_norm": 6.022928714752197, | |
| "learning_rate": 4.230053264503256e-05, | |
| "loss": 13.0683, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.6144653305116395, | |
| "grad_norm": 5.392704963684082, | |
| "learning_rate": 4.1946343229723514e-05, | |
| "loss": 12.6063, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.6154612224573633, | |
| "grad_norm": 6.246056079864502, | |
| "learning_rate": 4.159334369067781e-05, | |
| "loss": 13.3403, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.6164571144030873, | |
| "grad_norm": 7.040495872497559, | |
| "learning_rate": 4.124153908517739e-05, | |
| "loss": 11.5163, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.6174530063488112, | |
| "grad_norm": 5.689401626586914, | |
| "learning_rate": 4.089093445338514e-05, | |
| "loss": 12.6618, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6184488982945351, | |
| "grad_norm": 5.199860572814941, | |
| "learning_rate": 4.054153481827226e-05, | |
| "loss": 11.5894, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.6194447902402589, | |
| "grad_norm": 4.949602127075195, | |
| "learning_rate": 4.019334518554649e-05, | |
| "loss": 11.6394, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.6204406821859828, | |
| "grad_norm": 5.066486358642578, | |
| "learning_rate": 3.984637054358034e-05, | |
| "loss": 10.7524, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.6214365741317067, | |
| "grad_norm": 5.321683406829834, | |
| "learning_rate": 3.950061586333967e-05, | |
| "loss": 12.0487, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.6224324660774306, | |
| "grad_norm": 5.394808769226074, | |
| "learning_rate": 3.9156086098312395e-05, | |
| "loss": 11.734, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6234283580231544, | |
| "grad_norm": 5.522308826446533, | |
| "learning_rate": 3.881278618443754e-05, | |
| "loss": 10.2733, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.6244242499688784, | |
| "grad_norm": 4.8424787521362305, | |
| "learning_rate": 3.847072104003474e-05, | |
| "loss": 11.1419, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.6254201419146023, | |
| "grad_norm": 5.075899600982666, | |
| "learning_rate": 3.812989556573327e-05, | |
| "loss": 11.4723, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.6264160338603262, | |
| "grad_norm": 5.3835835456848145, | |
| "learning_rate": 3.779031464440241e-05, | |
| "loss": 11.468, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.6274119258060501, | |
| "grad_norm": 6.307840347290039, | |
| "learning_rate": 3.7451983141081184e-05, | |
| "loss": 10.6013, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6284078177517739, | |
| "grad_norm": 5.412378311157227, | |
| "learning_rate": 3.711490590290853e-05, | |
| "loss": 11.0776, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.6294037096974978, | |
| "grad_norm": 5.071283340454102, | |
| "learning_rate": 3.67790877590543e-05, | |
| "loss": 11.9251, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.6303996016432217, | |
| "grad_norm": 5.967265605926514, | |
| "learning_rate": 3.6444533520649594e-05, | |
| "loss": 11.9519, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.6313954935889456, | |
| "grad_norm": 5.716352939605713, | |
| "learning_rate": 3.611124798071819e-05, | |
| "loss": 11.6726, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.6323913855346694, | |
| "grad_norm": 5.153800964355469, | |
| "learning_rate": 3.5779235914107616e-05, | |
| "loss": 11.6163, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6333872774803934, | |
| "grad_norm": 4.7233710289001465, | |
| "learning_rate": 3.544850207742104e-05, | |
| "loss": 10.8191, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.6343831694261173, | |
| "grad_norm": 5.4549384117126465, | |
| "learning_rate": 3.511905120894869e-05, | |
| "loss": 12.1209, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.6353790613718412, | |
| "grad_norm": 4.882791996002197, | |
| "learning_rate": 3.4790888028600406e-05, | |
| "loss": 11.3777, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.636374953317565, | |
| "grad_norm": 6.203968524932861, | |
| "learning_rate": 3.4464017237837874e-05, | |
| "loss": 12.2292, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.6373708452632889, | |
| "grad_norm": 5.759111404418945, | |
| "learning_rate": 3.413844351960703e-05, | |
| "loss": 11.9604, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6383667372090128, | |
| "grad_norm": 5.6707763671875, | |
| "learning_rate": 3.3814171538271406e-05, | |
| "loss": 11.5694, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.6393626291547367, | |
| "grad_norm": 5.256777286529541, | |
| "learning_rate": 3.349120593954494e-05, | |
| "loss": 11.3768, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.6403585211004607, | |
| "grad_norm": 4.985710620880127, | |
| "learning_rate": 3.316955135042557e-05, | |
| "loss": 11.9468, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.6413544130461845, | |
| "grad_norm": 6.197348117828369, | |
| "learning_rate": 3.284921237912897e-05, | |
| "loss": 11.4821, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.6423503049919084, | |
| "grad_norm": 5.89221715927124, | |
| "learning_rate": 3.253019361502242e-05, | |
| "loss": 11.8649, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6433461969376323, | |
| "grad_norm": 4.958759784698486, | |
| "learning_rate": 3.2212499628559183e-05, | |
| "loss": 11.3049, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.6443420888833562, | |
| "grad_norm": 6.130683422088623, | |
| "learning_rate": 3.189613497121291e-05, | |
| "loss": 11.6303, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.64533798082908, | |
| "grad_norm": 5.429010391235352, | |
| "learning_rate": 3.158110417541264e-05, | |
| "loss": 11.1867, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.6463338727748039, | |
| "grad_norm": 5.844196796417236, | |
| "learning_rate": 3.126741175447746e-05, | |
| "loss": 11.8687, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.6473297647205278, | |
| "grad_norm": 5.631326675415039, | |
| "learning_rate": 3.0955062202552344e-05, | |
| "loss": 12.6221, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6483256566662517, | |
| "grad_norm": 4.835573673248291, | |
| "learning_rate": 3.0644059994543364e-05, | |
| "loss": 11.0964, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.6493215486119756, | |
| "grad_norm": 5.213573455810547, | |
| "learning_rate": 3.0334409586053822e-05, | |
| "loss": 11.2015, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.6503174405576995, | |
| "grad_norm": 5.156842231750488, | |
| "learning_rate": 3.0026115413320316e-05, | |
| "loss": 11.8143, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.6513133325034234, | |
| "grad_norm": 5.973851680755615, | |
| "learning_rate": 2.971918189314915e-05, | |
| "loss": 10.897, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.6523092244491473, | |
| "grad_norm": 4.918872356414795, | |
| "learning_rate": 2.9413613422853187e-05, | |
| "loss": 11.468, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.6533051163948712, | |
| "grad_norm": 5.078589916229248, | |
| "learning_rate": 2.9109414380188673e-05, | |
| "loss": 11.5197, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.654301008340595, | |
| "grad_norm": 5.085513114929199, | |
| "learning_rate": 2.88065891232928e-05, | |
| "loss": 11.2995, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.6552969002863189, | |
| "grad_norm": 5.631070613861084, | |
| "learning_rate": 2.850514199062085e-05, | |
| "loss": 10.8211, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.6562927922320428, | |
| "grad_norm": 6.195810794830322, | |
| "learning_rate": 2.8205077300884367e-05, | |
| "loss": 11.3694, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.6572886841777668, | |
| "grad_norm": 5.141655445098877, | |
| "learning_rate": 2.790639935298933e-05, | |
| "loss": 11.9514, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6582845761234906, | |
| "grad_norm": 4.97012186050415, | |
| "learning_rate": 2.760911242597418e-05, | |
| "loss": 10.6117, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.6592804680692145, | |
| "grad_norm": 5.144041061401367, | |
| "learning_rate": 2.7313220778948974e-05, | |
| "loss": 10.5493, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.6602763600149384, | |
| "grad_norm": 5.334214210510254, | |
| "learning_rate": 2.7018728651034052e-05, | |
| "loss": 11.3385, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.6612722519606623, | |
| "grad_norm": 5.036880016326904, | |
| "learning_rate": 2.6725640261299408e-05, | |
| "loss": 12.6018, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.6622681439063861, | |
| "grad_norm": 5.173761367797852, | |
| "learning_rate": 2.643395980870428e-05, | |
| "loss": 10.3298, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.66326403585211, | |
| "grad_norm": 5.325448513031006, | |
| "learning_rate": 2.6143691472036934e-05, | |
| "loss": 11.6448, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.6642599277978339, | |
| "grad_norm": 5.270651340484619, | |
| "learning_rate": 2.5854839409854837e-05, | |
| "loss": 10.9496, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.6652558197435579, | |
| "grad_norm": 5.0304999351501465, | |
| "learning_rate": 2.5567407760425002e-05, | |
| "loss": 12.0285, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.6662517116892818, | |
| "grad_norm": 5.503153324127197, | |
| "learning_rate": 2.5281400641664925e-05, | |
| "loss": 11.4235, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.6672476036350056, | |
| "grad_norm": 6.423144340515137, | |
| "learning_rate": 2.499682215108319e-05, | |
| "loss": 12.1289, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6682434955807295, | |
| "grad_norm": 6.0244879722595215, | |
| "learning_rate": 2.4713676365721194e-05, | |
| "loss": 12.4105, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.6692393875264534, | |
| "grad_norm": 5.558983325958252, | |
| "learning_rate": 2.4431967342094465e-05, | |
| "loss": 11.8495, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.6702352794721773, | |
| "grad_norm": 4.985186576843262, | |
| "learning_rate": 2.415169911613463e-05, | |
| "loss": 11.2055, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.6712311714179011, | |
| "grad_norm": 6.300296306610107, | |
| "learning_rate": 2.3872875703131582e-05, | |
| "loss": 11.8234, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.672227063363625, | |
| "grad_norm": 5.401225566864014, | |
| "learning_rate": 2.359550109767597e-05, | |
| "loss": 11.3489, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.673222955309349, | |
| "grad_norm": 5.588122367858887, | |
| "learning_rate": 2.331957927360198e-05, | |
| "loss": 11.3187, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.6742188472550729, | |
| "grad_norm": 5.249821662902832, | |
| "learning_rate": 2.3045114183930307e-05, | |
| "loss": 12.5959, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.6752147392007967, | |
| "grad_norm": 6.340845584869385, | |
| "learning_rate": 2.2772109760811786e-05, | |
| "loss": 11.3688, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.6762106311465206, | |
| "grad_norm": 4.663110256195068, | |
| "learning_rate": 2.2500569915470588e-05, | |
| "loss": 11.021, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.6772065230922445, | |
| "grad_norm": 6.123586654663086, | |
| "learning_rate": 2.223049853814875e-05, | |
| "loss": 11.7617, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6782024150379684, | |
| "grad_norm": 5.436185836791992, | |
| "learning_rate": 2.1961899498049997e-05, | |
| "loss": 11.8642, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.6791983069836923, | |
| "grad_norm": 5.2976765632629395, | |
| "learning_rate": 2.1694776643284453e-05, | |
| "loss": 11.2544, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.6801941989294161, | |
| "grad_norm": 5.813897609710693, | |
| "learning_rate": 2.1429133800813654e-05, | |
| "loss": 11.6754, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.68119009087514, | |
| "grad_norm": 5.2720184326171875, | |
| "learning_rate": 2.116497477639552e-05, | |
| "loss": 12.3032, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.682185982820864, | |
| "grad_norm": 4.996230125427246, | |
| "learning_rate": 2.090230335452989e-05, | |
| "loss": 11.7477, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.6831818747665879, | |
| "grad_norm": 6.202296733856201, | |
| "learning_rate": 2.064112329840437e-05, | |
| "loss": 11.789, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.6841777667123117, | |
| "grad_norm": 5.982651710510254, | |
| "learning_rate": 2.0381438349840326e-05, | |
| "loss": 11.3074, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.6851736586580356, | |
| "grad_norm": 5.608039379119873, | |
| "learning_rate": 2.0123252229239357e-05, | |
| "loss": 11.9267, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.6861695506037595, | |
| "grad_norm": 5.310582637786865, | |
| "learning_rate": 1.986656863552992e-05, | |
| "loss": 12.5416, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.6871654425494834, | |
| "grad_norm": 5.351050853729248, | |
| "learning_rate": 1.9611391246114468e-05, | |
| "loss": 12.0263, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.6881613344952072, | |
| "grad_norm": 5.318876266479492, | |
| "learning_rate": 1.93577237168165e-05, | |
| "loss": 11.4462, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.6891572264409311, | |
| "grad_norm": 7.177599906921387, | |
| "learning_rate": 1.910556968182854e-05, | |
| "loss": 12.3857, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.690153118386655, | |
| "grad_norm": 4.861364841461182, | |
| "learning_rate": 1.8854932753659778e-05, | |
| "loss": 10.3558, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.691149010332379, | |
| "grad_norm": 4.699554443359375, | |
| "learning_rate": 1.8605816523084462e-05, | |
| "loss": 11.7694, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.6921449022781029, | |
| "grad_norm": 5.567951202392578, | |
| "learning_rate": 1.8358224559090418e-05, | |
| "loss": 11.883, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.6931407942238267, | |
| "grad_norm": 7.005056858062744, | |
| "learning_rate": 1.8112160408827906e-05, | |
| "loss": 12.0574, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.6941366861695506, | |
| "grad_norm": 4.946389198303223, | |
| "learning_rate": 1.786762759755882e-05, | |
| "loss": 10.6037, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.6951325781152745, | |
| "grad_norm": 6.21707820892334, | |
| "learning_rate": 1.762462962860617e-05, | |
| "loss": 12.1973, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.6961284700609984, | |
| "grad_norm": 5.530759334564209, | |
| "learning_rate": 1.738316998330397e-05, | |
| "loss": 12.2209, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.6971243620067222, | |
| "grad_norm": 5.81333065032959, | |
| "learning_rate": 1.7143252120947138e-05, | |
| "loss": 11.2614, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6971243620067222, | |
| "eval_loss": 1.4631643295288086, | |
| "eval_runtime": 289.9209, | |
| "eval_samples_per_second": 4.619, | |
| "eval_steps_per_second": 1.155, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6981202539524461, | |
| "grad_norm": 5.98025369644165, | |
| "learning_rate": 1.6904879478742253e-05, | |
| "loss": 12.586, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.6991161458981701, | |
| "grad_norm": 4.925411701202393, | |
| "learning_rate": 1.6668055471758064e-05, | |
| "loss": 11.1688, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.700112037843894, | |
| "grad_norm": 5.7626566886901855, | |
| "learning_rate": 1.6432783492876634e-05, | |
| "loss": 12.3594, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.7011079297896178, | |
| "grad_norm": 6.179996967315674, | |
| "learning_rate": 1.6199066912744793e-05, | |
| "loss": 13.2897, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.7021038217353417, | |
| "grad_norm": 5.263505458831787, | |
| "learning_rate": 1.5966909079725783e-05, | |
| "loss": 11.9709, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7030997136810656, | |
| "grad_norm": 5.473526954650879, | |
| "learning_rate": 1.5736313319851285e-05, | |
| "loss": 12.4221, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.7040956056267895, | |
| "grad_norm": 5.03949499130249, | |
| "learning_rate": 1.5507282936773768e-05, | |
| "loss": 11.0594, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.7050914975725134, | |
| "grad_norm": 5.114665508270264, | |
| "learning_rate": 1.5279821211719307e-05, | |
| "loss": 10.9889, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.7060873895182372, | |
| "grad_norm": 6.260105133056641, | |
| "learning_rate": 1.5053931403440275e-05, | |
| "loss": 12.1907, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.7070832814639612, | |
| "grad_norm": 5.071542263031006, | |
| "learning_rate": 1.4829616748168891e-05, | |
| "loss": 11.6752, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7080791734096851, | |
| "grad_norm": 5.426812171936035, | |
| "learning_rate": 1.4606880459570884e-05, | |
| "loss": 12.0812, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.709075065355409, | |
| "grad_norm": 5.589237689971924, | |
| "learning_rate": 1.4385725728699187e-05, | |
| "loss": 12.6791, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.7100709573011328, | |
| "grad_norm": 5.184196472167969, | |
| "learning_rate": 1.4166155723948554e-05, | |
| "loss": 9.6767, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.7110668492468567, | |
| "grad_norm": 5.003568172454834, | |
| "learning_rate": 1.3948173591009916e-05, | |
| "loss": 10.6604, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.7120627411925806, | |
| "grad_norm": 5.232964038848877, | |
| "learning_rate": 1.3731782452825428e-05, | |
| "loss": 12.3529, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7130586331383045, | |
| "grad_norm": 5.2426629066467285, | |
| "learning_rate": 1.35169854095437e-05, | |
| "loss": 10.9877, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.7140545250840283, | |
| "grad_norm": 5.008747100830078, | |
| "learning_rate": 1.3303785538475403e-05, | |
| "loss": 11.3483, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.7150504170297522, | |
| "grad_norm": 5.193508148193359, | |
| "learning_rate": 1.3092185894049133e-05, | |
| "loss": 11.6414, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.7160463089754762, | |
| "grad_norm": 4.987440586090088, | |
| "learning_rate": 1.2882189507767705e-05, | |
| "loss": 11.1112, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.7170422009212001, | |
| "grad_norm": 4.958714485168457, | |
| "learning_rate": 1.267379938816475e-05, | |
| "loss": 11.5351, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.718038092866924, | |
| "grad_norm": 5.384908676147461, | |
| "learning_rate": 1.2467018520761416e-05, | |
| "loss": 11.5047, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.7190339848126478, | |
| "grad_norm": 5.524514675140381, | |
| "learning_rate": 1.2261849868023936e-05, | |
| "loss": 11.5071, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.7200298767583717, | |
| "grad_norm": 4.934844970703125, | |
| "learning_rate": 1.2058296369320848e-05, | |
| "loss": 11.9821, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.7210257687040956, | |
| "grad_norm": 5.9136457443237305, | |
| "learning_rate": 1.1856360940881106e-05, | |
| "loss": 11.1714, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.7220216606498195, | |
| "grad_norm": 5.626174449920654, | |
| "learning_rate": 1.1656046475752178e-05, | |
| "loss": 11.0057, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7230175525955433, | |
| "grad_norm": 6.411077499389648, | |
| "learning_rate": 1.1457355843758663e-05, | |
| "loss": 11.485, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.7240134445412673, | |
| "grad_norm": 5.232149600982666, | |
| "learning_rate": 1.126029189146116e-05, | |
| "loss": 11.9511, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.7250093364869912, | |
| "grad_norm": 6.392588138580322, | |
| "learning_rate": 1.1064857442115468e-05, | |
| "loss": 12.3394, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.7260052284327151, | |
| "grad_norm": 4.912868499755859, | |
| "learning_rate": 1.0871055295632232e-05, | |
| "loss": 10.2057, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.7270011203784389, | |
| "grad_norm": 4.6938652992248535, | |
| "learning_rate": 1.067888822853659e-05, | |
| "loss": 10.9687, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7279970123241628, | |
| "grad_norm": 7.743927955627441, | |
| "learning_rate": 1.048835899392872e-05, | |
| "loss": 12.2057, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.7289929042698867, | |
| "grad_norm": 7.151386260986328, | |
| "learning_rate": 1.0299470321444168e-05, | |
| "loss": 12.7718, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.7299887962156106, | |
| "grad_norm": 5.987144470214844, | |
| "learning_rate": 1.0112224917214738e-05, | |
| "loss": 11.8501, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.7309846881613346, | |
| "grad_norm": 7.631872653961182, | |
| "learning_rate": 9.926625463829942e-06, | |
| "loss": 12.742, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.7319805801070584, | |
| "grad_norm": 5.5182085037231445, | |
| "learning_rate": 9.742674620298305e-06, | |
| "loss": 11.0, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7329764720527823, | |
| "grad_norm": 4.892227649688721, | |
| "learning_rate": 9.560375022009418e-06, | |
| "loss": 11.4375, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.7339723639985062, | |
| "grad_norm": 5.391102313995361, | |
| "learning_rate": 9.37972928069615e-06, | |
| "loss": 11.0945, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.7349682559442301, | |
| "grad_norm": 5.407810688018799, | |
| "learning_rate": 9.200739984397202e-06, | |
| "loss": 11.0342, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.7359641478899539, | |
| "grad_norm": 5.871718883514404, | |
| "learning_rate": 9.02340969742009e-06, | |
| "loss": 11.9766, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.7369600398356778, | |
| "grad_norm": 5.485849857330322, | |
| "learning_rate": 8.847740960304357e-06, | |
| "loss": 11.4418, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7379559317814017, | |
| "grad_norm": 6.5848164558410645, | |
| "learning_rate": 8.673736289785197e-06, | |
| "loss": 13.1044, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.7389518237271256, | |
| "grad_norm": 6.7905473709106445, | |
| "learning_rate": 8.50139817875735e-06, | |
| "loss": 10.9534, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.7399477156728494, | |
| "grad_norm": 4.94359827041626, | |
| "learning_rate": 8.330729096239539e-06, | |
| "loss": 11.1527, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.7409436076185734, | |
| "grad_norm": 5.690808296203613, | |
| "learning_rate": 8.161731487338827e-06, | |
| "loss": 12.6257, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.7419394995642973, | |
| "grad_norm": 5.709432125091553, | |
| "learning_rate": 7.994407773215903e-06, | |
| "loss": 11.0885, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.7429353915100212, | |
| "grad_norm": 5.193820476531982, | |
| "learning_rate": 7.828760351050165e-06, | |
| "loss": 11.0625, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.7439312834557451, | |
| "grad_norm": 5.115967750549316, | |
| "learning_rate": 7.664791594005433e-06, | |
| "loss": 9.8947, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.7449271754014689, | |
| "grad_norm": 5.30134916305542, | |
| "learning_rate": 7.502503851196024e-06, | |
| "loss": 11.6142, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.7459230673471928, | |
| "grad_norm": 5.2966108322143555, | |
| "learning_rate": 7.341899447652997e-06, | |
| "loss": 11.1821, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.7469189592929167, | |
| "grad_norm": 5.26341438293457, | |
| "learning_rate": 7.182980684290921e-06, | |
| "loss": 12.1355, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7479148512386407, | |
| "grad_norm": 6.888838768005371, | |
| "learning_rate": 7.0257498378748366e-06, | |
| "loss": 13.0227, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.7489107431843645, | |
| "grad_norm": 5.804958343505859, | |
| "learning_rate": 6.87020916098774e-06, | |
| "loss": 11.9386, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.7499066351300884, | |
| "grad_norm": 4.986551284790039, | |
| "learning_rate": 6.716360881998174e-06, | |
| "loss": 10.9869, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.7509025270758123, | |
| "grad_norm": 4.778614044189453, | |
| "learning_rate": 6.564207205028441e-06, | |
| "loss": 11.1464, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.7518984190215362, | |
| "grad_norm": 6.451168060302734, | |
| "learning_rate": 6.413750309922958e-06, | |
| "loss": 12.4462, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.75289431096726, | |
| "grad_norm": 5.491706371307373, | |
| "learning_rate": 6.264992352216961e-06, | |
| "loss": 12.6588, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.7538902029129839, | |
| "grad_norm": 5.205382823944092, | |
| "learning_rate": 6.117935463105809e-06, | |
| "loss": 10.741, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.7548860948587078, | |
| "grad_norm": 5.023211479187012, | |
| "learning_rate": 5.972581749414244e-06, | |
| "loss": 11.0733, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.7558819868044317, | |
| "grad_norm": 5.587502956390381, | |
| "learning_rate": 5.828933293566333e-06, | |
| "loss": 11.1592, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.7568778787501556, | |
| "grad_norm": 5.698174476623535, | |
| "learning_rate": 5.686992153555614e-06, | |
| "loss": 12.2832, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7578737706958795, | |
| "grad_norm": 5.90291166305542, | |
| "learning_rate": 5.546760362915571e-06, | |
| "loss": 12.3435, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.7588696626416034, | |
| "grad_norm": 5.5049591064453125, | |
| "learning_rate": 5.408239930690548e-06, | |
| "loss": 11.5343, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.7598655545873273, | |
| "grad_norm": 5.788873672485352, | |
| "learning_rate": 5.271432841406937e-06, | |
| "loss": 12.3758, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.7608614465330512, | |
| "grad_norm": 5.662258148193359, | |
| "learning_rate": 5.1363410550448045e-06, | |
| "loss": 11.152, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.761857338478775, | |
| "grad_norm": 5.340865612030029, | |
| "learning_rate": 5.0029665070096544e-06, | |
| "loss": 11.2238, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.7628532304244989, | |
| "grad_norm": 5.056994915008545, | |
| "learning_rate": 4.871311108104917e-06, | |
| "loss": 9.9053, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.7638491223702228, | |
| "grad_norm": 5.196466445922852, | |
| "learning_rate": 4.741376744504422e-06, | |
| "loss": 12.2725, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.7648450143159468, | |
| "grad_norm": 5.003063201904297, | |
| "learning_rate": 4.613165277725428e-06, | |
| "loss": 10.418, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.7658409062616706, | |
| "grad_norm": 5.598369598388672, | |
| "learning_rate": 4.486678544601913e-06, | |
| "loss": 10.3517, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.7668367982073945, | |
| "grad_norm": 5.888363361358643, | |
| "learning_rate": 4.36191835725834e-06, | |
| "loss": 11.2754, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7678326901531184, | |
| "grad_norm": 5.854308128356934, | |
| "learning_rate": 4.238886503083628e-06, | |
| "loss": 13.0128, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.7688285820988423, | |
| "grad_norm": 6.681763648986816, | |
| "learning_rate": 4.117584744705527e-06, | |
| "loss": 12.1851, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.7698244740445661, | |
| "grad_norm": 6.164341449737549, | |
| "learning_rate": 3.998014819965479e-06, | |
| "loss": 11.9651, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.77082036599029, | |
| "grad_norm": 5.836161136627197, | |
| "learning_rate": 3.880178441893562e-06, | |
| "loss": 12.3736, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.7718162579360139, | |
| "grad_norm": 5.648519039154053, | |
| "learning_rate": 3.764077298684096e-06, | |
| "loss": 11.6445, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.7728121498817379, | |
| "grad_norm": 5.881044387817383, | |
| "learning_rate": 3.6497130536713716e-06, | |
| "loss": 13.0027, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.7738080418274618, | |
| "grad_norm": 5.143442153930664, | |
| "learning_rate": 3.5370873453057646e-06, | |
| "loss": 11.9685, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.7748039337731856, | |
| "grad_norm": 5.579311370849609, | |
| "learning_rate": 3.426201787130476e-06, | |
| "loss": 11.7632, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.7757998257189095, | |
| "grad_norm": 5.457061290740967, | |
| "learning_rate": 3.3170579677581636e-06, | |
| "loss": 12.535, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.7767957176646334, | |
| "grad_norm": 5.1317644119262695, | |
| "learning_rate": 3.2096574508483357e-06, | |
| "loss": 10.979, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7777916096103573, | |
| "grad_norm": 5.227689266204834, | |
| "learning_rate": 3.1040017750848943e-06, | |
| "loss": 11.2438, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.7787875015560811, | |
| "grad_norm": 6.507405757904053, | |
| "learning_rate": 3.0000924541541687e-06, | |
| "loss": 12.94, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.779783393501805, | |
| "grad_norm": 5.245728492736816, | |
| "learning_rate": 2.8979309767230844e-06, | |
| "loss": 11.8061, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.780779285447529, | |
| "grad_norm": 5.699490547180176, | |
| "learning_rate": 2.7975188064179173e-06, | |
| "loss": 11.486, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.7817751773932529, | |
| "grad_norm": 5.269382476806641, | |
| "learning_rate": 2.6988573818034213e-06, | |
| "loss": 11.2483, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.7827710693389767, | |
| "grad_norm": 5.229419231414795, | |
| "learning_rate": 2.601948116361996e-06, | |
| "loss": 11.6245, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.7837669612847006, | |
| "grad_norm": 5.528886318206787, | |
| "learning_rate": 2.5067923984736647e-06, | |
| "loss": 12.4577, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.7847628532304245, | |
| "grad_norm": 5.435757637023926, | |
| "learning_rate": 2.413391591396044e-06, | |
| "loss": 12.7685, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.7857587451761484, | |
| "grad_norm": 5.185883522033691, | |
| "learning_rate": 2.3217470332448226e-06, | |
| "loss": 11.2059, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.7867546371218723, | |
| "grad_norm": 5.466739177703857, | |
| "learning_rate": 2.2318600369746754e-06, | |
| "loss": 11.3974, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.7877505290675961, | |
| "grad_norm": 5.585619926452637, | |
| "learning_rate": 2.143731890360337e-06, | |
| "loss": 11.8463, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.78874642101332, | |
| "grad_norm": 6.319226264953613, | |
| "learning_rate": 2.057363855978253e-06, | |
| "loss": 12.4982, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.789742312959044, | |
| "grad_norm": 5.821755886077881, | |
| "learning_rate": 1.972757171188444e-06, | |
| "loss": 11.5502, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.7907382049047679, | |
| "grad_norm": 6.314916133880615, | |
| "learning_rate": 1.8899130481167815e-06, | |
| "loss": 12.7203, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.7917340968504917, | |
| "grad_norm": 5.605375289916992, | |
| "learning_rate": 1.8088326736376004e-06, | |
| "loss": 11.8973, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.7927299887962156, | |
| "grad_norm": 5.3746843338012695, | |
| "learning_rate": 1.729517209356782e-06, | |
| "loss": 11.2227, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.7937258807419395, | |
| "grad_norm": 4.950900554656982, | |
| "learning_rate": 1.6519677915949743e-06, | |
| "loss": 10.9039, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.7947217726876634, | |
| "grad_norm": 5.662406921386719, | |
| "learning_rate": 1.5761855313714684e-06, | |
| "loss": 12.4767, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.7957176646333872, | |
| "grad_norm": 6.010452747344971, | |
| "learning_rate": 1.502171514388112e-06, | |
| "loss": 11.6454, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.7967135565791111, | |
| "grad_norm": 6.123356342315674, | |
| "learning_rate": 1.4299268010139339e-06, | |
| "loss": 11.4508, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.7967135565791111, | |
| "eval_loss": 1.4602320194244385, | |
| "eval_runtime": 289.8374, | |
| "eval_samples_per_second": 4.62, | |
| "eval_steps_per_second": 1.156, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.797709448524835, | |
| "grad_norm": 5.147581100463867, | |
| "learning_rate": 1.3594524262698082e-06, | |
| "loss": 11.0485, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.798705340470559, | |
| "grad_norm": 4.97263765335083, | |
| "learning_rate": 1.290749399813676e-06, | |
| "loss": 11.5831, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.7997012324162829, | |
| "grad_norm": 6.502606391906738, | |
| "learning_rate": 1.2238187059261384e-06, | |
| "loss": 10.9787, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.8006971243620067, | |
| "grad_norm": 5.0996527671813965, | |
| "learning_rate": 1.1586613034962333e-06, | |
| "loss": 11.5943, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.8016930163077306, | |
| "grad_norm": 5.570478439331055, | |
| "learning_rate": 1.0952781260078342e-06, | |
| "loss": 12.3992, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8026889082534545, | |
| "grad_norm": 5.16178035736084, | |
| "learning_rate": 1.0336700815261613e-06, | |
| "loss": 11.8013, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.8036848001991784, | |
| "grad_norm": 5.67569637298584, | |
| "learning_rate": 9.738380526848194e-07, | |
| "loss": 12.3158, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.8046806921449022, | |
| "grad_norm": 5.282881259918213, | |
| "learning_rate": 9.157828966731979e-07, | |
| "loss": 12.0702, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.8056765840906261, | |
| "grad_norm": 5.958638668060303, | |
| "learning_rate": 8.595054452241041e-07, | |
| "loss": 12.2277, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.8066724760363501, | |
| "grad_norm": 5.30989408493042, | |
| "learning_rate": 8.050065046018851e-07, | |
| "loss": 11.5156, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.807668367982074, | |
| "grad_norm": 4.585794925689697, | |
| "learning_rate": 7.522868555909223e-07, | |
| "loss": 10.3803, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.8086642599277978, | |
| "grad_norm": 5.959741115570068, | |
| "learning_rate": 7.013472534843635e-07, | |
| "loss": 13.3266, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.8096601518735217, | |
| "grad_norm": 6.225035190582275, | |
| "learning_rate": 6.521884280733526e-07, | |
| "loss": 12.5264, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.8106560438192456, | |
| "grad_norm": 4.888844013214111, | |
| "learning_rate": 6.048110836365666e-07, | |
| "loss": 10.7281, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.8116519357649695, | |
| "grad_norm": 5.534932613372803, | |
| "learning_rate": 5.592158989301405e-07, | |
| "loss": 12.486, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8126478277106934, | |
| "grad_norm": 5.48603630065918, | |
| "learning_rate": 5.154035271778684e-07, | |
| "loss": 11.7151, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.8136437196564172, | |
| "grad_norm": 5.270609378814697, | |
| "learning_rate": 4.733745960619762e-07, | |
| "loss": 11.9809, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.8146396116021412, | |
| "grad_norm": 5.433355331420898, | |
| "learning_rate": 4.331297077140167e-07, | |
| "loss": 12.2935, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.8156355035478651, | |
| "grad_norm": 5.731333255767822, | |
| "learning_rate": 3.94669438706266e-07, | |
| "loss": 12.0509, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.816631395493589, | |
| "grad_norm": 5.320248126983643, | |
| "learning_rate": 3.579943400434937e-07, | |
| "loss": 11.0143, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8176272874393128, | |
| "grad_norm": 5.686704635620117, | |
| "learning_rate": 3.231049371550804e-07, | |
| "loss": 12.9909, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.8186231793850367, | |
| "grad_norm": 5.398179054260254, | |
| "learning_rate": 2.900017298874125e-07, | |
| "loss": 10.8497, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.8196190713307606, | |
| "grad_norm": 6.222070693969727, | |
| "learning_rate": 2.5868519249680475e-07, | |
| "loss": 12.435, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.8206149632764845, | |
| "grad_norm": 5.970067024230957, | |
| "learning_rate": 2.2915577364267247e-07, | |
| "loss": 13.6407, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.8216108552222083, | |
| "grad_norm": 5.729474067687988, | |
| "learning_rate": 2.0141389638109187e-07, | |
| "loss": 12.3141, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8226067471679323, | |
| "grad_norm": 6.03525972366333, | |
| "learning_rate": 1.7545995815876348e-07, | |
| "loss": 11.4593, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.8236026391136562, | |
| "grad_norm": 5.3436360359191895, | |
| "learning_rate": 1.5129433080728062e-07, | |
| "loss": 11.6677, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.8245985310593801, | |
| "grad_norm": 4.991388320922852, | |
| "learning_rate": 1.2891736053785575e-07, | |
| "loss": 10.8385, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.825594423005104, | |
| "grad_norm": 5.2977118492126465, | |
| "learning_rate": 1.0832936793633841e-07, | |
| "loss": 12.2961, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.8265903149508278, | |
| "grad_norm": 5.460627555847168, | |
| "learning_rate": 8.953064795856614e-08, | |
| "loss": 11.1108, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 5.315445423126221, | |
| "learning_rate": 7.252146992625664e-08, | |
| "loss": 10.9681, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.8285820988422756, | |
| "grad_norm": 5.33111572265625, | |
| "learning_rate": 5.730207752302485e-08, | |
| "loss": 11.7059, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.8295779907879995, | |
| "grad_norm": 5.866223335266113, | |
| "learning_rate": 4.387268879098294e-08, | |
| "loss": 13.3463, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.8305738827337233, | |
| "grad_norm": 5.038965225219727, | |
| "learning_rate": 3.223349612756221e-08, | |
| "loss": 11.1416, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.8315697746794473, | |
| "grad_norm": 5.116003036499023, | |
| "learning_rate": 2.2384666282779244e-08, | |
| "loss": 11.2051, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.8325656666251712, | |
| "grad_norm": 5.213040351867676, | |
| "learning_rate": 1.4326340356862754e-08, | |
| "loss": 11.9355, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.8335615585708951, | |
| "grad_norm": 5.077611446380615, | |
| "learning_rate": 8.058633798199711e-09, | |
| "loss": 11.5202, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.8345574505166189, | |
| "grad_norm": 6.297743797302246, | |
| "learning_rate": 3.58163640169773e-09, | |
| "loss": 11.4272, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.8355533424623428, | |
| "grad_norm": 6.544887065887451, | |
| "learning_rate": 8.954123074805809e-10, | |
| "loss": 12.609, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.8365492344080667, | |
| "grad_norm": 5.315311908721924, | |
| "learning_rate": 0.0, | |
| "loss": 12.2755, | |
| "step": 840 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 840, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 2, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.286387212989235e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |