| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 23.44766505636071, | |
| "eval_steps": 100, | |
| "global_step": 900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2576489533011272, | |
| "grad_norm": 5.974159240722656, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 8.6521, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5152979066022544, | |
| "grad_norm": 3.2873778343200684, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 8.3497, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 2.522794485092163, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 7.9896, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0418679549114331, | |
| "grad_norm": 1.5867993831634521, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 8.1084, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.2995169082125604, | |
| "grad_norm": 1.300277829170227, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 7.3393, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.5571658615136876, | |
| "grad_norm": 1.1117545366287231, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 7.1266, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.8148148148148149, | |
| "grad_norm": 0.9784226417541504, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 6.9662, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.0837359098228663, | |
| "grad_norm": 0.9447280764579773, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 7.2558, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.3413848631239937, | |
| "grad_norm": 1.1698005199432373, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 6.6472, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.5990338164251208, | |
| "grad_norm": 1.544875144958496, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 6.392, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.5990338164251208, | |
| "eval_loss": 6.279848575592041, | |
| "eval_runtime": 8.7699, | |
| "eval_samples_per_second": 183.012, | |
| "eval_steps_per_second": 5.815, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.8566827697262482, | |
| "grad_norm": 1.4024465084075928, | |
| "learning_rate": 1.1578947368421053e-05, | |
| "loss": 6.1704, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 3.1256038647342996, | |
| "grad_norm": 1.2708696126937866, | |
| "learning_rate": 1.263157894736842e-05, | |
| "loss": 6.369, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.3832528180354267, | |
| "grad_norm": 0.7357354760169983, | |
| "learning_rate": 1.3684210526315791e-05, | |
| "loss": 5.8538, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.640901771336554, | |
| "grad_norm": 0.5254281163215637, | |
| "learning_rate": 1.4736842105263159e-05, | |
| "loss": 5.7556, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.898550724637681, | |
| "grad_norm": 0.8618260025978088, | |
| "learning_rate": 1.578947368421053e-05, | |
| "loss": 5.6495, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 4.1674718196457325, | |
| "grad_norm": 1.0980093479156494, | |
| "learning_rate": 1.6842105263157896e-05, | |
| "loss": 5.9024, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 4.42512077294686, | |
| "grad_norm": 1.3449251651763916, | |
| "learning_rate": 1.7894736842105264e-05, | |
| "loss": 5.4768, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.6827697262479875, | |
| "grad_norm": 1.2047125101089478, | |
| "learning_rate": 1.894736842105263e-05, | |
| "loss": 5.3661, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.940418679549114, | |
| "grad_norm": 1.6401121616363525, | |
| "learning_rate": 2e-05, | |
| "loss": 5.2836, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 5.209339774557166, | |
| "grad_norm": 1.8774281740188599, | |
| "learning_rate": 1.999831241633323e-05, | |
| "loss": 5.5386, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.209339774557166, | |
| "eval_loss": 5.1901116371154785, | |
| "eval_runtime": 8.0613, | |
| "eval_samples_per_second": 199.099, | |
| "eval_steps_per_second": 6.326, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.466988727858293, | |
| "grad_norm": 2.0122838020324707, | |
| "learning_rate": 1.9993250234920638e-05, | |
| "loss": 5.145, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.72463768115942, | |
| "grad_norm": 1.5243710279464722, | |
| "learning_rate": 1.9984815164333163e-05, | |
| "loss": 5.0726, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 5.982286634460547, | |
| "grad_norm": 1.613467812538147, | |
| "learning_rate": 1.9973010051548274e-05, | |
| "loss": 5.1139, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 6.251207729468599, | |
| "grad_norm": 1.169236183166504, | |
| "learning_rate": 1.9957838880989076e-05, | |
| "loss": 5.1586, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 6.508856682769727, | |
| "grad_norm": 1.2285501956939697, | |
| "learning_rate": 1.9939306773179498e-05, | |
| "loss": 4.9011, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 6.766505636070853, | |
| "grad_norm": 1.4358283281326294, | |
| "learning_rate": 1.9917419983016025e-05, | |
| "loss": 4.8479, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 7.035426731078905, | |
| "grad_norm": 1.3988443613052368, | |
| "learning_rate": 1.989218589765658e-05, | |
| "loss": 5.097, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 7.293075684380033, | |
| "grad_norm": 1.1683136224746704, | |
| "learning_rate": 1.9863613034027224e-05, | |
| "loss": 4.7351, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 7.550724637681159, | |
| "grad_norm": 1.2952402830123901, | |
| "learning_rate": 1.9831711035947552e-05, | |
| "loss": 4.6976, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 7.808373590982287, | |
| "grad_norm": 1.4266207218170166, | |
| "learning_rate": 1.979649067087574e-05, | |
| "loss": 4.6391, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.808373590982287, | |
| "eval_loss": 4.603694438934326, | |
| "eval_runtime": 8.3521, | |
| "eval_samples_per_second": 192.167, | |
| "eval_steps_per_second": 6.106, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 8.077294685990339, | |
| "grad_norm": 2.173689126968384, | |
| "learning_rate": 1.9757963826274357e-05, | |
| "loss": 4.8403, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 8.334943639291465, | |
| "grad_norm": 1.874500036239624, | |
| "learning_rate": 1.971614350559814e-05, | |
| "loss": 4.4785, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 8.592592592592592, | |
| "grad_norm": 1.9841059446334839, | |
| "learning_rate": 1.967104382390511e-05, | |
| "loss": 4.4265, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 8.85024154589372, | |
| "grad_norm": 2.523819923400879, | |
| "learning_rate": 1.9622680003092503e-05, | |
| "loss": 4.3497, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 9.119162640901772, | |
| "grad_norm": 2.2171430587768555, | |
| "learning_rate": 1.9571068366759143e-05, | |
| "loss": 4.5573, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 9.376811594202898, | |
| "grad_norm": 2.516089916229248, | |
| "learning_rate": 1.951622633469592e-05, | |
| "loss": 4.2197, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 9.634460547504025, | |
| "grad_norm": 3.2976479530334473, | |
| "learning_rate": 1.9458172417006347e-05, | |
| "loss": 4.1526, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 9.892109500805153, | |
| "grad_norm": 2.310131311416626, | |
| "learning_rate": 1.9396926207859085e-05, | |
| "loss": 4.1294, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 10.161030595813205, | |
| "grad_norm": 2.445627450942993, | |
| "learning_rate": 1.933250837887457e-05, | |
| "loss": 4.3354, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 10.418679549114332, | |
| "grad_norm": 2.188768148422241, | |
| "learning_rate": 1.9264940672148018e-05, | |
| "loss": 4.0545, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 10.418679549114332, | |
| "eval_loss": 4.0533294677734375, | |
| "eval_runtime": 8.3508, | |
| "eval_samples_per_second": 192.197, | |
| "eval_steps_per_second": 6.107, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 10.676328502415458, | |
| "grad_norm": 3.4046390056610107, | |
| "learning_rate": 1.9194245892911077e-05, | |
| "loss": 4.0121, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 10.933977455716587, | |
| "grad_norm": 2.701680898666382, | |
| "learning_rate": 1.9120447901834708e-05, | |
| "loss": 3.9847, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 11.202898550724637, | |
| "grad_norm": 2.2474334239959717, | |
| "learning_rate": 1.9043571606975776e-05, | |
| "loss": 4.1875, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 11.460547504025765, | |
| "grad_norm": 2.2355682849884033, | |
| "learning_rate": 1.8963642955370203e-05, | |
| "loss": 3.9325, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 11.718196457326892, | |
| "grad_norm": 2.977954626083374, | |
| "learning_rate": 1.888068892427538e-05, | |
| "loss": 3.8992, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 11.97584541062802, | |
| "grad_norm": 2.2921981811523438, | |
| "learning_rate": 1.879473751206489e-05, | |
| "loss": 3.858, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 12.24476650563607, | |
| "grad_norm": 1.8335373401641846, | |
| "learning_rate": 1.8705817728778626e-05, | |
| "loss": 4.0718, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 12.502415458937199, | |
| "grad_norm": 2.2911853790283203, | |
| "learning_rate": 1.8613959586331364e-05, | |
| "loss": 3.8305, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 12.760064412238325, | |
| "grad_norm": 2.170738697052002, | |
| "learning_rate": 1.851919408838327e-05, | |
| "loss": 3.7923, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 13.028985507246377, | |
| "grad_norm": 1.8763988018035889, | |
| "learning_rate": 1.842155321987566e-05, | |
| "loss": 4.014, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 13.028985507246377, | |
| "eval_loss": 3.779900550842285, | |
| "eval_runtime": 8.1613, | |
| "eval_samples_per_second": 196.661, | |
| "eval_steps_per_second": 6.249, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 13.286634460547504, | |
| "grad_norm": 2.2329862117767334, | |
| "learning_rate": 1.8321069936235503e-05, | |
| "loss": 3.753, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 13.544283413848632, | |
| "grad_norm": 2.3311359882354736, | |
| "learning_rate": 1.821777815225245e-05, | |
| "loss": 3.7099, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 13.801932367149758, | |
| "grad_norm": 2.5093774795532227, | |
| "learning_rate": 1.8111712730632024e-05, | |
| "loss": 3.6944, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 14.07085346215781, | |
| "grad_norm": 2.2526538372039795, | |
| "learning_rate": 1.800290947022884e-05, | |
| "loss": 3.9225, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 14.328502415458937, | |
| "grad_norm": 2.5449490547180176, | |
| "learning_rate": 1.789140509396394e-05, | |
| "loss": 3.6411, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 14.586151368760065, | |
| "grad_norm": 2.722266435623169, | |
| "learning_rate": 1.777723723643014e-05, | |
| "loss": 3.6571, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 14.843800322061192, | |
| "grad_norm": 2.3767805099487305, | |
| "learning_rate": 1.766044443118978e-05, | |
| "loss": 3.6036, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 15.112721417069244, | |
| "grad_norm": 2.4693193435668945, | |
| "learning_rate": 1.7541066097768965e-05, | |
| "loss": 3.8201, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 15.37037037037037, | |
| "grad_norm": 2.3329803943634033, | |
| "learning_rate": 1.7419142528352815e-05, | |
| "loss": 3.5643, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 15.628019323671497, | |
| "grad_norm": 2.396742343902588, | |
| "learning_rate": 1.729471487418621e-05, | |
| "loss": 3.5476, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 15.628019323671497, | |
| "eval_loss": 3.5582447052001953, | |
| "eval_runtime": 8.3607, | |
| "eval_samples_per_second": 191.969, | |
| "eval_steps_per_second": 6.1, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 15.885668276972625, | |
| "grad_norm": 2.3643219470977783, | |
| "learning_rate": 1.7167825131684516e-05, | |
| "loss": 3.5246, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 16.154589371980677, | |
| "grad_norm": 3.809258460998535, | |
| "learning_rate": 1.7038516128259118e-05, | |
| "loss": 3.7499, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 16.412238325281802, | |
| "grad_norm": 2.6895406246185303, | |
| "learning_rate": 1.6906831507862446e-05, | |
| "loss": 3.499, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 16.66988727858293, | |
| "grad_norm": 2.7662012577056885, | |
| "learning_rate": 1.6772815716257414e-05, | |
| "loss": 3.4918, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 16.92753623188406, | |
| "grad_norm": 2.217256784439087, | |
| "learning_rate": 1.6636513986016215e-05, | |
| "loss": 3.4541, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 17.19645732689211, | |
| "grad_norm": 3.010216474533081, | |
| "learning_rate": 1.64979723212536e-05, | |
| "loss": 3.6718, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 17.454106280193237, | |
| "grad_norm": 3.430988073348999, | |
| "learning_rate": 1.6357237482099682e-05, | |
| "loss": 3.4517, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 17.711755233494365, | |
| "grad_norm": 2.378178119659424, | |
| "learning_rate": 1.621435696891765e-05, | |
| "loss": 3.4259, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 17.96940418679549, | |
| "grad_norm": 2.877561569213867, | |
| "learning_rate": 1.606937900627157e-05, | |
| "loss": 3.4104, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 18.238325281803544, | |
| "grad_norm": 2.5511293411254883, | |
| "learning_rate": 1.5922352526649803e-05, | |
| "loss": 3.6157, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 18.238325281803544, | |
| "eval_loss": 3.421673059463501, | |
| "eval_runtime": 8.3529, | |
| "eval_samples_per_second": 192.149, | |
| "eval_steps_per_second": 6.106, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 18.49597423510467, | |
| "grad_norm": 2.8884575366973877, | |
| "learning_rate": 1.5773327153949465e-05, | |
| "loss": 3.4023, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 18.753623188405797, | |
| "grad_norm": 2.9396398067474365, | |
| "learning_rate": 1.5622353186727542e-05, | |
| "loss": 3.3875, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 19.022544283413847, | |
| "grad_norm": 3.390272855758667, | |
| "learning_rate": 1.5469481581224274e-05, | |
| "loss": 3.596, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 19.280193236714975, | |
| "grad_norm": 2.9179751873016357, | |
| "learning_rate": 1.531476393416456e-05, | |
| "loss": 3.3614, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 19.537842190016104, | |
| "grad_norm": 4.342045307159424, | |
| "learning_rate": 1.5158252465343242e-05, | |
| "loss": 3.3597, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 19.79549114331723, | |
| "grad_norm": 2.955822706222534, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 3.3584, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 20.064412238325282, | |
| "grad_norm": 2.6151533126831055, | |
| "learning_rate": 1.4840059950989992e-05, | |
| "loss": 3.5532, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 20.32206119162641, | |
| "grad_norm": 3.1554718017578125, | |
| "learning_rate": 1.467848630075608e-05, | |
| "loss": 3.3281, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 20.579710144927535, | |
| "grad_norm": 2.572737455368042, | |
| "learning_rate": 1.4515333583108896e-05, | |
| "loss": 3.3162, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 20.837359098228664, | |
| "grad_norm": 3.3176584243774414, | |
| "learning_rate": 1.4350656864820733e-05, | |
| "loss": 3.3161, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 20.837359098228664, | |
| "eval_loss": 3.3334078788757324, | |
| "eval_runtime": 8.7701, | |
| "eval_samples_per_second": 183.009, | |
| "eval_steps_per_second": 5.815, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 21.106280193236714, | |
| "grad_norm": 3.9218039512634277, | |
| "learning_rate": 1.4184511727039612e-05, | |
| "loss": 3.5253, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 21.363929146537842, | |
| "grad_norm": 3.392868995666504, | |
| "learning_rate": 1.4016954246529697e-05, | |
| "loss": 3.2947, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 21.62157809983897, | |
| "grad_norm": 3.4435510635375977, | |
| "learning_rate": 1.3848040976744459e-05, | |
| "loss": 3.3021, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 21.879227053140095, | |
| "grad_norm": 3.319772243499756, | |
| "learning_rate": 1.3677828928738934e-05, | |
| "loss": 3.3022, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 22.14814814814815, | |
| "grad_norm": 3.0634384155273438, | |
| "learning_rate": 1.3506375551927546e-05, | |
| "loss": 3.4676, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 22.405797101449274, | |
| "grad_norm": 4.093255996704102, | |
| "learning_rate": 1.3333738714693958e-05, | |
| "loss": 3.2705, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 22.663446054750402, | |
| "grad_norm": 3.1328248977661133, | |
| "learning_rate": 1.3159976684859528e-05, | |
| "loss": 3.2759, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 22.92109500805153, | |
| "grad_norm": 3.3607850074768066, | |
| "learning_rate": 1.2985148110016947e-05, | |
| "loss": 3.26, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 23.19001610305958, | |
| "grad_norm": 4.250182628631592, | |
| "learning_rate": 1.2809311997735697e-05, | |
| "loss": 3.4547, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 23.44766505636071, | |
| "grad_norm": 3.2897439002990723, | |
| "learning_rate": 1.2632527695645993e-05, | |
| "loss": 3.2482, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 23.44766505636071, | |
| "eval_loss": 3.2658839225769043, | |
| "eval_runtime": 8.3079, | |
| "eval_samples_per_second": 193.189, | |
| "eval_steps_per_second": 6.139, | |
| "step": 900 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1900, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.7940579890639667e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |