| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 19.999535243996903, | |
| "eval_steps": 500, | |
| "global_step": 32260, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.061967467079783116, | |
| "grad_norm": 20.777441024780273, | |
| "learning_rate": 9.97179169249845e-06, | |
| "loss": 5.1314, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12393493415956623, | |
| "grad_norm": 7.313554763793945, | |
| "learning_rate": 9.940793552386857e-06, | |
| "loss": 2.5079, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18590240123934934, | |
| "grad_norm": 6.049108982086182, | |
| "learning_rate": 9.909795412275263e-06, | |
| "loss": 2.262, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.24786986831913246, | |
| "grad_norm": 7.533471584320068, | |
| "learning_rate": 9.878797272163671e-06, | |
| "loss": 2.1505, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.30983733539891556, | |
| "grad_norm": 7.321562767028809, | |
| "learning_rate": 9.847799132052077e-06, | |
| "loss": 2.0719, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3718048024786987, | |
| "grad_norm": 9.049120903015137, | |
| "learning_rate": 9.816800991940483e-06, | |
| "loss": 2.0476, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4337722695584818, | |
| "grad_norm": 4.262355804443359, | |
| "learning_rate": 9.785802851828891e-06, | |
| "loss": 2.0214, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.4957397366382649, | |
| "grad_norm": 4.2409491539001465, | |
| "learning_rate": 9.754804711717297e-06, | |
| "loss": 2.0108, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.557707203718048, | |
| "grad_norm": 4.823696136474609, | |
| "learning_rate": 9.723806571605703e-06, | |
| "loss": 1.9636, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6196746707978311, | |
| "grad_norm": 16.929576873779297, | |
| "learning_rate": 9.692808431494111e-06, | |
| "loss": 1.9787, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6816421378776143, | |
| "grad_norm": 4.100649833679199, | |
| "learning_rate": 9.661810291382517e-06, | |
| "loss": 1.9076, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7436096049573974, | |
| "grad_norm": 8.428403854370117, | |
| "learning_rate": 9.630812151270923e-06, | |
| "loss": 1.9312, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8055770720371804, | |
| "grad_norm": 21.134408950805664, | |
| "learning_rate": 9.599814011159331e-06, | |
| "loss": 1.9262, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.8675445391169636, | |
| "grad_norm": 4.445871829986572, | |
| "learning_rate": 9.568815871047737e-06, | |
| "loss": 1.9038, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9295120061967467, | |
| "grad_norm": 10.70151424407959, | |
| "learning_rate": 9.537817730936143e-06, | |
| "loss": 1.909, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9914794732765299, | |
| "grad_norm": 4.220454692840576, | |
| "learning_rate": 9.506819590824551e-06, | |
| "loss": 1.9011, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.9995352439969016, | |
| "eval_loss": 1.478387713432312, | |
| "eval_runtime": 71.2201, | |
| "eval_samples_per_second": 45.31, | |
| "eval_steps_per_second": 22.662, | |
| "step": 1613 | |
| }, | |
| { | |
| "epoch": 1.0539116963594113, | |
| "grad_norm": 5.241837024688721, | |
| "learning_rate": 9.475821450712957e-06, | |
| "loss": 1.8493, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.1158791634391945, | |
| "grad_norm": 8.490580558776855, | |
| "learning_rate": 9.444823310601364e-06, | |
| "loss": 1.8844, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.1778466305189776, | |
| "grad_norm": 4.5867743492126465, | |
| "learning_rate": 9.413825170489771e-06, | |
| "loss": 1.8588, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.2398140975987606, | |
| "grad_norm": 5.750781536102295, | |
| "learning_rate": 9.382827030378178e-06, | |
| "loss": 1.8462, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.3017815646785438, | |
| "grad_norm": 3.4414772987365723, | |
| "learning_rate": 9.351828890266584e-06, | |
| "loss": 1.8476, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.363749031758327, | |
| "grad_norm": 3.991649627685547, | |
| "learning_rate": 9.320830750154991e-06, | |
| "loss": 1.8111, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.42571649883811, | |
| "grad_norm": 3.774946689605713, | |
| "learning_rate": 9.289832610043398e-06, | |
| "loss": 1.8224, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.487683965917893, | |
| "grad_norm": 11.512676239013672, | |
| "learning_rate": 9.258834469931804e-06, | |
| "loss": 1.8164, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.5496514329976763, | |
| "grad_norm": 4.229243755340576, | |
| "learning_rate": 9.227836329820212e-06, | |
| "loss": 1.8052, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.6116189000774592, | |
| "grad_norm": 3.6015431880950928, | |
| "learning_rate": 9.196838189708618e-06, | |
| "loss": 1.8253, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.6735863671572424, | |
| "grad_norm": 3.8703105449676514, | |
| "learning_rate": 9.165840049597024e-06, | |
| "loss": 1.7702, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.7355538342370256, | |
| "grad_norm": 3.6859114170074463, | |
| "learning_rate": 9.134841909485432e-06, | |
| "loss": 1.7925, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.7975213013168085, | |
| "grad_norm": 6.744151592254639, | |
| "learning_rate": 9.103843769373838e-06, | |
| "loss": 1.7995, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.859488768396592, | |
| "grad_norm": 5.791450500488281, | |
| "learning_rate": 9.072845629262244e-06, | |
| "loss": 1.7894, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.921456235476375, | |
| "grad_norm": 5.028300762176514, | |
| "learning_rate": 9.041847489150652e-06, | |
| "loss": 1.7955, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.9834237025561579, | |
| "grad_norm": 8.930970191955566, | |
| "learning_rate": 9.010849349039058e-06, | |
| "loss": 1.7856, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.9995352439969016, | |
| "eval_loss": 1.438106894493103, | |
| "eval_runtime": 71.4662, | |
| "eval_samples_per_second": 45.154, | |
| "eval_steps_per_second": 22.584, | |
| "step": 3226 | |
| }, | |
| { | |
| "epoch": 2.0458559256390396, | |
| "grad_norm": 4.4157915115356445, | |
| "learning_rate": 8.979851208927464e-06, | |
| "loss": 1.7661, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.1078233927188226, | |
| "grad_norm": 5.970052242279053, | |
| "learning_rate": 8.948853068815872e-06, | |
| "loss": 1.7539, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.1697908597986055, | |
| "grad_norm": 6.610071659088135, | |
| "learning_rate": 8.917854928704278e-06, | |
| "loss": 1.7629, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.231758326878389, | |
| "grad_norm": 5.124378204345703, | |
| "learning_rate": 8.886856788592684e-06, | |
| "loss": 1.7512, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.293725793958172, | |
| "grad_norm": 3.7913496494293213, | |
| "learning_rate": 8.855858648481092e-06, | |
| "loss": 1.7574, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.3556932610379553, | |
| "grad_norm": 3.5346639156341553, | |
| "learning_rate": 8.824860508369498e-06, | |
| "loss": 1.7298, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.4176607281177382, | |
| "grad_norm": 3.6782634258270264, | |
| "learning_rate": 8.793862368257904e-06, | |
| "loss": 1.756, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.479628195197521, | |
| "grad_norm": 3.087649345397949, | |
| "learning_rate": 8.762864228146312e-06, | |
| "loss": 1.7557, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.541595662277304, | |
| "grad_norm": 3.854729413986206, | |
| "learning_rate": 8.731866088034718e-06, | |
| "loss": 1.7425, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.6035631293570876, | |
| "grad_norm": 3.7198565006256104, | |
| "learning_rate": 8.700867947923124e-06, | |
| "loss": 1.7379, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.6655305964368705, | |
| "grad_norm": 4.5739593505859375, | |
| "learning_rate": 8.669869807811532e-06, | |
| "loss": 1.7379, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.727498063516654, | |
| "grad_norm": 4.533520698547363, | |
| "learning_rate": 8.638871667699938e-06, | |
| "loss": 1.7475, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.789465530596437, | |
| "grad_norm": 3.113757610321045, | |
| "learning_rate": 8.607873527588344e-06, | |
| "loss": 1.7243, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.85143299767622, | |
| "grad_norm": 3.061246871948242, | |
| "learning_rate": 8.576875387476752e-06, | |
| "loss": 1.7114, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 2.9134004647560032, | |
| "grad_norm": 9.025867462158203, | |
| "learning_rate": 8.545877247365158e-06, | |
| "loss": 1.7184, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 2.975367931835786, | |
| "grad_norm": 18.252887725830078, | |
| "learning_rate": 8.514879107253565e-06, | |
| "loss": 1.7165, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.9995352439969016, | |
| "eval_loss": 1.4203578233718872, | |
| "eval_runtime": 71.3513, | |
| "eval_samples_per_second": 45.227, | |
| "eval_steps_per_second": 22.62, | |
| "step": 4839 | |
| }, | |
| { | |
| "epoch": 3.0378001549186675, | |
| "grad_norm": 3.237386703491211, | |
| "learning_rate": 8.483880967141972e-06, | |
| "loss": 1.7036, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 3.099767621998451, | |
| "grad_norm": 3.4343390464782715, | |
| "learning_rate": 8.452882827030378e-06, | |
| "loss": 1.6937, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.161735089078234, | |
| "grad_norm": 4.684260845184326, | |
| "learning_rate": 8.421884686918785e-06, | |
| "loss": 1.6825, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 3.223702556158017, | |
| "grad_norm": 4.4700236320495605, | |
| "learning_rate": 8.390886546807192e-06, | |
| "loss": 1.7013, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 3.2856700232378, | |
| "grad_norm": 3.5046918392181396, | |
| "learning_rate": 8.359888406695599e-06, | |
| "loss": 1.6819, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 3.347637490317583, | |
| "grad_norm": 3.9169533252716064, | |
| "learning_rate": 8.328890266584005e-06, | |
| "loss": 1.6939, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 3.4096049573973666, | |
| "grad_norm": 3.2398970127105713, | |
| "learning_rate": 8.297892126472413e-06, | |
| "loss": 1.689, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.4715724244771495, | |
| "grad_norm": 5.543476104736328, | |
| "learning_rate": 8.266893986360819e-06, | |
| "loss": 1.6819, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 3.5335398915569325, | |
| "grad_norm": 3.900956153869629, | |
| "learning_rate": 8.235895846249225e-06, | |
| "loss": 1.7153, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 3.595507358636716, | |
| "grad_norm": 3.7085001468658447, | |
| "learning_rate": 8.204897706137633e-06, | |
| "loss": 1.6663, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 3.657474825716499, | |
| "grad_norm": 3.672489881515503, | |
| "learning_rate": 8.173899566026039e-06, | |
| "loss": 1.6961, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 3.719442292796282, | |
| "grad_norm": 5.308550834655762, | |
| "learning_rate": 8.142901425914445e-06, | |
| "loss": 1.6702, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.781409759876065, | |
| "grad_norm": 6.25948429107666, | |
| "learning_rate": 8.111903285802853e-06, | |
| "loss": 1.7, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 3.843377226955848, | |
| "grad_norm": 4.889005661010742, | |
| "learning_rate": 8.080905145691259e-06, | |
| "loss": 1.6783, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 3.905344694035631, | |
| "grad_norm": 3.8252604007720947, | |
| "learning_rate": 8.049907005579665e-06, | |
| "loss": 1.6679, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 3.9673121611154145, | |
| "grad_norm": 5.56712532043457, | |
| "learning_rate": 8.018908865468073e-06, | |
| "loss": 1.6778, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 3.9995352439969016, | |
| "eval_loss": 1.4068138599395752, | |
| "eval_runtime": 71.8216, | |
| "eval_samples_per_second": 44.931, | |
| "eval_steps_per_second": 22.472, | |
| "step": 6452 | |
| }, | |
| { | |
| "epoch": 4.029744384198296, | |
| "grad_norm": 4.865719318389893, | |
| "learning_rate": 7.987910725356479e-06, | |
| "loss": 1.6718, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 4.091711851278079, | |
| "grad_norm": 3.5504400730133057, | |
| "learning_rate": 7.956912585244885e-06, | |
| "loss": 1.6252, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 4.153679318357862, | |
| "grad_norm": 2.8009276390075684, | |
| "learning_rate": 7.925914445133293e-06, | |
| "loss": 1.6529, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 4.215646785437645, | |
| "grad_norm": 3.3487560749053955, | |
| "learning_rate": 7.894916305021699e-06, | |
| "loss": 1.6402, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 4.277614252517428, | |
| "grad_norm": 3.037754535675049, | |
| "learning_rate": 7.863918164910105e-06, | |
| "loss": 1.6659, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 4.339581719597211, | |
| "grad_norm": 3.340179204940796, | |
| "learning_rate": 7.832920024798513e-06, | |
| "loss": 1.646, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.401549186676995, | |
| "grad_norm": 4.411381244659424, | |
| "learning_rate": 7.801921884686919e-06, | |
| "loss": 1.6583, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 4.463516653756778, | |
| "grad_norm": 4.7763352394104, | |
| "learning_rate": 7.770923744575325e-06, | |
| "loss": 1.6524, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 4.525484120836561, | |
| "grad_norm": 3.343606472015381, | |
| "learning_rate": 7.739925604463733e-06, | |
| "loss": 1.6454, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 4.587451587916344, | |
| "grad_norm": 5.252141952514648, | |
| "learning_rate": 7.70892746435214e-06, | |
| "loss": 1.6555, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 4.649419054996127, | |
| "grad_norm": 2.8249170780181885, | |
| "learning_rate": 7.677929324240545e-06, | |
| "loss": 1.6506, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 4.711386522075911, | |
| "grad_norm": 6.525687217712402, | |
| "learning_rate": 7.646931184128953e-06, | |
| "loss": 1.6215, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 4.7733539891556935, | |
| "grad_norm": 4.455771446228027, | |
| "learning_rate": 7.615933044017359e-06, | |
| "loss": 1.6465, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 4.8353214562354765, | |
| "grad_norm": 2.8847739696502686, | |
| "learning_rate": 7.584934903905766e-06, | |
| "loss": 1.6411, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 4.8972889233152594, | |
| "grad_norm": 2.8771958351135254, | |
| "learning_rate": 7.553936763794173e-06, | |
| "loss": 1.6353, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 4.959256390395042, | |
| "grad_norm": 3.5082240104675293, | |
| "learning_rate": 7.522938623682579e-06, | |
| "loss": 1.6679, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.999535243996902, | |
| "eval_loss": 1.3975656032562256, | |
| "eval_runtime": 71.9374, | |
| "eval_samples_per_second": 44.858, | |
| "eval_steps_per_second": 22.436, | |
| "step": 8065 | |
| }, | |
| { | |
| "epoch": 5.021688613477924, | |
| "grad_norm": 3.450159788131714, | |
| "learning_rate": 7.491940483570986e-06, | |
| "loss": 1.6245, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 5.083656080557708, | |
| "grad_norm": 3.078734874725342, | |
| "learning_rate": 7.460942343459393e-06, | |
| "loss": 1.5949, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 5.1456235476374905, | |
| "grad_norm": 3.123811721801758, | |
| "learning_rate": 7.4299442033477995e-06, | |
| "loss": 1.5804, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 5.2075910147172735, | |
| "grad_norm": 2.746354341506958, | |
| "learning_rate": 7.3989460632362065e-06, | |
| "loss": 1.6264, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 5.269558481797056, | |
| "grad_norm": 3.969360113143921, | |
| "learning_rate": 7.3679479231246135e-06, | |
| "loss": 1.6285, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 5.331525948876839, | |
| "grad_norm": 11.645624160766602, | |
| "learning_rate": 7.33694978301302e-06, | |
| "loss": 1.6196, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 5.393493415956623, | |
| "grad_norm": 4.237279415130615, | |
| "learning_rate": 7.305951642901427e-06, | |
| "loss": 1.6271, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 5.455460883036406, | |
| "grad_norm": 5.038291931152344, | |
| "learning_rate": 7.2749535027898336e-06, | |
| "loss": 1.6097, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 5.517428350116189, | |
| "grad_norm": 6.474031448364258, | |
| "learning_rate": 7.24395536267824e-06, | |
| "loss": 1.6148, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 5.579395817195972, | |
| "grad_norm": 3.735469341278076, | |
| "learning_rate": 7.212957222566647e-06, | |
| "loss": 1.608, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 5.641363284275755, | |
| "grad_norm": 4.774326801300049, | |
| "learning_rate": 7.181959082455054e-06, | |
| "loss": 1.6362, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 5.703330751355538, | |
| "grad_norm": 6.094056606292725, | |
| "learning_rate": 7.15096094234346e-06, | |
| "loss": 1.5946, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 5.765298218435321, | |
| "grad_norm": 2.613210439682007, | |
| "learning_rate": 7.119962802231867e-06, | |
| "loss": 1.6237, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 5.827265685515105, | |
| "grad_norm": 5.291254043579102, | |
| "learning_rate": 7.088964662120274e-06, | |
| "loss": 1.6225, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 5.889233152594888, | |
| "grad_norm": 3.139016628265381, | |
| "learning_rate": 7.05796652200868e-06, | |
| "loss": 1.6229, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 5.951200619674671, | |
| "grad_norm": 2.9346985816955566, | |
| "learning_rate": 7.026968381897087e-06, | |
| "loss": 1.6091, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 5.999535243996902, | |
| "eval_loss": 1.3893319368362427, | |
| "eval_runtime": 71.9002, | |
| "eval_samples_per_second": 44.882, | |
| "eval_steps_per_second": 22.448, | |
| "step": 9678 | |
| }, | |
| { | |
| "epoch": 6.013632842757552, | |
| "grad_norm": 2.6308019161224365, | |
| "learning_rate": 6.995970241785494e-06, | |
| "loss": 1.6145, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 6.075600309837335, | |
| "grad_norm": 6.885835647583008, | |
| "learning_rate": 6.9649721016739e-06, | |
| "loss": 1.575, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 6.137567776917119, | |
| "grad_norm": 8.84401798248291, | |
| "learning_rate": 6.933973961562307e-06, | |
| "loss": 1.5861, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 6.199535243996902, | |
| "grad_norm": 3.830970287322998, | |
| "learning_rate": 6.902975821450714e-06, | |
| "loss": 1.6048, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.261502711076685, | |
| "grad_norm": 2.8187596797943115, | |
| "learning_rate": 6.87197768133912e-06, | |
| "loss": 1.5923, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 6.323470178156468, | |
| "grad_norm": 3.0165576934814453, | |
| "learning_rate": 6.840979541227527e-06, | |
| "loss": 1.5773, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 6.385437645236251, | |
| "grad_norm": 8.133129119873047, | |
| "learning_rate": 6.809981401115934e-06, | |
| "loss": 1.576, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 6.447405112316034, | |
| "grad_norm": 3.153198480606079, | |
| "learning_rate": 6.77898326100434e-06, | |
| "loss": 1.5922, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 6.5093725793958175, | |
| "grad_norm": 3.5953714847564697, | |
| "learning_rate": 6.7482951022938625e-06, | |
| "loss": 1.5954, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 6.5713400464756, | |
| "grad_norm": 3.0549609661102295, | |
| "learning_rate": 6.7172969621822695e-06, | |
| "loss": 1.5751, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 6.633307513555383, | |
| "grad_norm": 3.2885892391204834, | |
| "learning_rate": 6.686298822070676e-06, | |
| "loss": 1.6103, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 6.695274980635166, | |
| "grad_norm": 3.2449259757995605, | |
| "learning_rate": 6.6553006819590826e-06, | |
| "loss": 1.5879, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 6.757242447714949, | |
| "grad_norm": 3.855558395385742, | |
| "learning_rate": 6.6243025418474895e-06, | |
| "loss": 1.5863, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 6.819209914794733, | |
| "grad_norm": 7.185797214508057, | |
| "learning_rate": 6.593304401735896e-06, | |
| "loss": 1.5594, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 6.881177381874516, | |
| "grad_norm": 5.2492804527282715, | |
| "learning_rate": 6.562306261624303e-06, | |
| "loss": 1.5716, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 6.943144848954299, | |
| "grad_norm": 3.938108444213867, | |
| "learning_rate": 6.53130812151271e-06, | |
| "loss": 1.6033, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 6.999535243996902, | |
| "eval_loss": 1.3868727684020996, | |
| "eval_runtime": 71.7471, | |
| "eval_samples_per_second": 44.977, | |
| "eval_steps_per_second": 22.496, | |
| "step": 11291 | |
| }, | |
| { | |
| "epoch": 7.00557707203718, | |
| "grad_norm": 3.1100411415100098, | |
| "learning_rate": 6.500309981401116e-06, | |
| "loss": 1.5961, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 7.067544539116963, | |
| "grad_norm": 4.577524185180664, | |
| "learning_rate": 6.469311841289523e-06, | |
| "loss": 1.5562, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 7.129512006196746, | |
| "grad_norm": 3.86492919921875, | |
| "learning_rate": 6.43831370117793e-06, | |
| "loss": 1.5757, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 7.19147947327653, | |
| "grad_norm": 6.450658321380615, | |
| "learning_rate": 6.407315561066336e-06, | |
| "loss": 1.5467, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 7.253446940356313, | |
| "grad_norm": 3.469329357147217, | |
| "learning_rate": 6.376317420954743e-06, | |
| "loss": 1.5676, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 7.315414407436096, | |
| "grad_norm": 3.4291296005249023, | |
| "learning_rate": 6.34531928084315e-06, | |
| "loss": 1.5693, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 7.377381874515879, | |
| "grad_norm": 2.776810884475708, | |
| "learning_rate": 6.314321140731556e-06, | |
| "loss": 1.5761, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 7.439349341595662, | |
| "grad_norm": 3.687276840209961, | |
| "learning_rate": 6.283323000619963e-06, | |
| "loss": 1.5582, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 7.501316808675446, | |
| "grad_norm": 3.623704195022583, | |
| "learning_rate": 6.25232486050837e-06, | |
| "loss": 1.557, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 7.563284275755229, | |
| "grad_norm": 2.607255220413208, | |
| "learning_rate": 6.221326720396776e-06, | |
| "loss": 1.5467, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 7.625251742835012, | |
| "grad_norm": 4.738313674926758, | |
| "learning_rate": 6.190328580285183e-06, | |
| "loss": 1.5648, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 7.687219209914795, | |
| "grad_norm": 3.4459781646728516, | |
| "learning_rate": 6.15933044017359e-06, | |
| "loss": 1.57, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 7.749186676994578, | |
| "grad_norm": 2.6644325256347656, | |
| "learning_rate": 6.128332300061996e-06, | |
| "loss": 1.5757, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 7.811154144074361, | |
| "grad_norm": 6.151153564453125, | |
| "learning_rate": 6.097334159950403e-06, | |
| "loss": 1.5653, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 7.873121611154144, | |
| "grad_norm": 4.14183235168457, | |
| "learning_rate": 6.06633601983881e-06, | |
| "loss": 1.5708, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 7.935089078233927, | |
| "grad_norm": 11.848958015441895, | |
| "learning_rate": 6.035337879727216e-06, | |
| "loss": 1.5633, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 7.99705654531371, | |
| "grad_norm": 3.1628572940826416, | |
| "learning_rate": 6.004339739615623e-06, | |
| "loss": 1.548, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 7.999535243996902, | |
| "eval_loss": 1.3799071311950684, | |
| "eval_runtime": 71.5231, | |
| "eval_samples_per_second": 45.118, | |
| "eval_steps_per_second": 22.566, | |
| "step": 12904 | |
| }, | |
| { | |
| "epoch": 8.059488768396593, | |
| "grad_norm": 2.565005302429199, | |
| "learning_rate": 5.97334159950403e-06, | |
| "loss": 1.5559, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 8.121456235476375, | |
| "grad_norm": 3.772648334503174, | |
| "learning_rate": 5.942343459392436e-06, | |
| "loss": 1.5516, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 8.183423702556158, | |
| "grad_norm": 3.1010353565216064, | |
| "learning_rate": 5.911345319280843e-06, | |
| "loss": 1.5494, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 8.24539116963594, | |
| "grad_norm": 4.525482177734375, | |
| "learning_rate": 5.88034717916925e-06, | |
| "loss": 1.5391, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 8.307358636715724, | |
| "grad_norm": 4.108111381530762, | |
| "learning_rate": 5.8493490390576564e-06, | |
| "loss": 1.5798, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 8.369326103795508, | |
| "grad_norm": 3.5569262504577637, | |
| "learning_rate": 5.818350898946063e-06, | |
| "loss": 1.5332, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 8.43129357087529, | |
| "grad_norm": 2.829259157180786, | |
| "learning_rate": 5.78735275883447e-06, | |
| "loss": 1.5564, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 8.493261037955074, | |
| "grad_norm": 9.748035430908203, | |
| "learning_rate": 5.7563546187228765e-06, | |
| "loss": 1.5082, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 8.555228505034856, | |
| "grad_norm": 5.017675399780273, | |
| "learning_rate": 5.7253564786112835e-06, | |
| "loss": 1.5086, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 8.61719597211464, | |
| "grad_norm": 3.3384227752685547, | |
| "learning_rate": 5.6943583384996905e-06, | |
| "loss": 1.5682, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 8.679163439194422, | |
| "grad_norm": 3.0591983795166016, | |
| "learning_rate": 5.663670179789214e-06, | |
| "loss": 1.5359, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 8.741130906274206, | |
| "grad_norm": 3.4246435165405273, | |
| "learning_rate": 5.632672039677621e-06, | |
| "loss": 1.5611, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 8.80309837335399, | |
| "grad_norm": 3.443187713623047, | |
| "learning_rate": 5.601673899566027e-06, | |
| "loss": 1.5592, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 8.865065840433772, | |
| "grad_norm": 2.6613450050354004, | |
| "learning_rate": 5.570675759454434e-06, | |
| "loss": 1.5484, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 8.927033307513556, | |
| "grad_norm": 3.1063573360443115, | |
| "learning_rate": 5.539677619342841e-06, | |
| "loss": 1.5146, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 8.989000774593338, | |
| "grad_norm": 3.6589224338531494, | |
| "learning_rate": 5.508679479231247e-06, | |
| "loss": 1.5641, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 8.999535243996903, | |
| "eval_loss": 1.3773841857910156, | |
| "eval_runtime": 71.387, | |
| "eval_samples_per_second": 45.204, | |
| "eval_steps_per_second": 22.609, | |
| "step": 14517 | |
| }, | |
| { | |
| "epoch": 9.05143299767622, | |
| "grad_norm": 3.909100294113159, | |
| "learning_rate": 5.477681339119654e-06, | |
| "loss": 1.5137, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 9.113400464756003, | |
| "grad_norm": 2.70833683013916, | |
| "learning_rate": 5.446683199008061e-06, | |
| "loss": 1.519, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 9.175367931835787, | |
| "grad_norm": 2.6168839931488037, | |
| "learning_rate": 5.415685058896467e-06, | |
| "loss": 1.5082, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 9.237335398915569, | |
| "grad_norm": 2.4784512519836426, | |
| "learning_rate": 5.384686918784874e-06, | |
| "loss": 1.5098, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 9.299302865995353, | |
| "grad_norm": 2.7963085174560547, | |
| "learning_rate": 5.353688778673281e-06, | |
| "loss": 1.5257, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 9.361270333075135, | |
| "grad_norm": 2.6264543533325195, | |
| "learning_rate": 5.322690638561687e-06, | |
| "loss": 1.5122, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 9.423237800154919, | |
| "grad_norm": 2.796173095703125, | |
| "learning_rate": 5.291692498450094e-06, | |
| "loss": 1.5629, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 9.485205267234702, | |
| "grad_norm": 2.661559820175171, | |
| "learning_rate": 5.260694358338501e-06, | |
| "loss": 1.5337, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 9.547172734314485, | |
| "grad_norm": 2.721785068511963, | |
| "learning_rate": 5.229696218226907e-06, | |
| "loss": 1.5443, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 9.609140201394268, | |
| "grad_norm": 3.5480453968048096, | |
| "learning_rate": 5.198698078115314e-06, | |
| "loss": 1.5475, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 9.67110766847405, | |
| "grad_norm": 4.556975841522217, | |
| "learning_rate": 5.167699938003721e-06, | |
| "loss": 1.5032, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 9.733075135553834, | |
| "grad_norm": 3.4796106815338135, | |
| "learning_rate": 5.136701797892127e-06, | |
| "loss": 1.5328, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 9.795042602633618, | |
| "grad_norm": 2.3589675426483154, | |
| "learning_rate": 5.105703657780534e-06, | |
| "loss": 1.5333, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 9.8570100697134, | |
| "grad_norm": 8.50188159942627, | |
| "learning_rate": 5.074705517668941e-06, | |
| "loss": 1.5401, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 9.918977536793184, | |
| "grad_norm": 2.818558931350708, | |
| "learning_rate": 5.043707377557347e-06, | |
| "loss": 1.5305, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 9.980945003872966, | |
| "grad_norm": 7.929882526397705, | |
| "learning_rate": 5.01301921884687e-06, | |
| "loss": 1.5037, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 9.999535243996903, | |
| "eval_loss": 1.3749291896820068, | |
| "eval_runtime": 71.4165, | |
| "eval_samples_per_second": 45.186, | |
| "eval_steps_per_second": 22.6, | |
| "step": 16130 | |
| }, | |
| { | |
| "epoch": 10.043377226955847, | |
| "grad_norm": 2.9387943744659424, | |
| "learning_rate": 4.982021078735277e-06, | |
| "loss": 1.5233, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 10.105344694035631, | |
| "grad_norm": 3.994661331176758, | |
| "learning_rate": 4.951022938623683e-06, | |
| "loss": 1.4957, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 10.167312161115415, | |
| "grad_norm": 5.1318840980529785, | |
| "learning_rate": 4.92002479851209e-06, | |
| "loss": 1.5198, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 10.229279628195197, | |
| "grad_norm": 5.07450008392334, | |
| "learning_rate": 4.889026658400497e-06, | |
| "loss": 1.5168, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 10.291247095274981, | |
| "grad_norm": 3.5865488052368164, | |
| "learning_rate": 4.858028518288903e-06, | |
| "loss": 1.5136, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 10.353214562354763, | |
| "grad_norm": 3.8123815059661865, | |
| "learning_rate": 4.82703037817731e-06, | |
| "loss": 1.5095, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 10.415182029434547, | |
| "grad_norm": 3.0475172996520996, | |
| "learning_rate": 4.796032238065717e-06, | |
| "loss": 1.5005, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 10.47714949651433, | |
| "grad_norm": 3.0361108779907227, | |
| "learning_rate": 4.765034097954123e-06, | |
| "loss": 1.5169, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 10.539116963594113, | |
| "grad_norm": 2.7775065898895264, | |
| "learning_rate": 4.73403595784253e-06, | |
| "loss": 1.5025, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 10.601084430673897, | |
| "grad_norm": 4.8608598709106445, | |
| "learning_rate": 4.703037817730937e-06, | |
| "loss": 1.5216, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 10.663051897753679, | |
| "grad_norm": 3.024165630340576, | |
| "learning_rate": 4.672039677619343e-06, | |
| "loss": 1.5107, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 10.725019364833463, | |
| "grad_norm": 3.055216073989868, | |
| "learning_rate": 4.64104153750775e-06, | |
| "loss": 1.5083, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 10.786986831913246, | |
| "grad_norm": 2.7835144996643066, | |
| "learning_rate": 4.610043397396157e-06, | |
| "loss": 1.501, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 10.848954298993029, | |
| "grad_norm": 3.1248793601989746, | |
| "learning_rate": 4.579045257284563e-06, | |
| "loss": 1.519, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 10.910921766072812, | |
| "grad_norm": 3.8737070560455322, | |
| "learning_rate": 4.54804711717297e-06, | |
| "loss": 1.5087, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 10.972889233152594, | |
| "grad_norm": 2.474802017211914, | |
| "learning_rate": 4.517048977061377e-06, | |
| "loss": 1.4803, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 10.999535243996903, | |
| "eval_loss": 1.3729863166809082, | |
| "eval_runtime": 71.8366, | |
| "eval_samples_per_second": 44.921, | |
| "eval_steps_per_second": 22.468, | |
| "step": 17743 | |
| }, | |
| { | |
| "epoch": 11.035321456235476, | |
| "grad_norm": 2.705595016479492, | |
| "learning_rate": 4.486050836949783e-06, | |
| "loss": 1.5091, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 11.09728892331526, | |
| "grad_norm": 2.9070913791656494, | |
| "learning_rate": 4.45505269683819e-06, | |
| "loss": 1.5065, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 11.159256390395043, | |
| "grad_norm": 2.360722780227661, | |
| "learning_rate": 4.424054556726597e-06, | |
| "loss": 1.4999, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 11.221223857474826, | |
| "grad_norm": 2.7821319103240967, | |
| "learning_rate": 4.393056416615003e-06, | |
| "loss": 1.4863, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 11.28319132455461, | |
| "grad_norm": 3.3408045768737793, | |
| "learning_rate": 4.36205827650341e-06, | |
| "loss": 1.4846, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 11.345158791634391, | |
| "grad_norm": 2.658116579055786, | |
| "learning_rate": 4.331060136391817e-06, | |
| "loss": 1.4911, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 11.407126258714175, | |
| "grad_norm": 7.110032081604004, | |
| "learning_rate": 4.300061996280223e-06, | |
| "loss": 1.5037, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 11.469093725793957, | |
| "grad_norm": 3.009939193725586, | |
| "learning_rate": 4.26906385616863e-06, | |
| "loss": 1.481, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 11.531061192873741, | |
| "grad_norm": 3.2190310955047607, | |
| "learning_rate": 4.238065716057037e-06, | |
| "loss": 1.5163, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 11.593028659953525, | |
| "grad_norm": 3.240391969680786, | |
| "learning_rate": 4.2070675759454435e-06, | |
| "loss": 1.4934, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 11.654996127033307, | |
| "grad_norm": 2.4078099727630615, | |
| "learning_rate": 4.1760694358338505e-06, | |
| "loss": 1.5076, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 11.716963594113091, | |
| "grad_norm": 2.757641077041626, | |
| "learning_rate": 4.1450712957222575e-06, | |
| "loss": 1.5011, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 11.778931061192873, | |
| "grad_norm": 3.859056234359741, | |
| "learning_rate": 4.11438313701178e-06, | |
| "loss": 1.5009, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 11.840898528272657, | |
| "grad_norm": 2.826043128967285, | |
| "learning_rate": 4.083384996900187e-06, | |
| "loss": 1.5103, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 11.90286599535244, | |
| "grad_norm": 3.0263473987579346, | |
| "learning_rate": 4.052386856788593e-06, | |
| "loss": 1.4945, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 11.964833462432223, | |
| "grad_norm": 3.322974920272827, | |
| "learning_rate": 4.021388716677e-06, | |
| "loss": 1.4653, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 11.999535243996903, | |
| "eval_loss": 1.3713266849517822, | |
| "eval_runtime": 71.4255, | |
| "eval_samples_per_second": 45.18, | |
| "eval_steps_per_second": 22.597, | |
| "step": 19356 | |
| }, | |
| { | |
| "epoch": 12.027265685515104, | |
| "grad_norm": 3.6630640029907227, | |
| "learning_rate": 3.990390576565407e-06, | |
| "loss": 1.5061, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 12.089233152594888, | |
| "grad_norm": 2.810619831085205, | |
| "learning_rate": 3.959392436453813e-06, | |
| "loss": 1.5, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 12.15120061967467, | |
| "grad_norm": 2.4999682903289795, | |
| "learning_rate": 3.92839429634222e-06, | |
| "loss": 1.4863, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 12.213168086754454, | |
| "grad_norm": 3.3250415325164795, | |
| "learning_rate": 3.897396156230627e-06, | |
| "loss": 1.4719, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 12.275135553834238, | |
| "grad_norm": 3.2047810554504395, | |
| "learning_rate": 3.866398016119033e-06, | |
| "loss": 1.492, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 12.33710302091402, | |
| "grad_norm": 2.8707590103149414, | |
| "learning_rate": 3.83539987600744e-06, | |
| "loss": 1.4772, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 12.399070487993804, | |
| "grad_norm": 5.695312023162842, | |
| "learning_rate": 3.804401735895847e-06, | |
| "loss": 1.4785, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 12.461037955073586, | |
| "grad_norm": 3.48818039894104, | |
| "learning_rate": 3.7734035957842536e-06, | |
| "loss": 1.4789, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 12.52300542215337, | |
| "grad_norm": 2.8195247650146484, | |
| "learning_rate": 3.7424054556726597e-06, | |
| "loss": 1.4859, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 12.584972889233153, | |
| "grad_norm": 3.1309738159179688, | |
| "learning_rate": 3.7114073155610663e-06, | |
| "loss": 1.4914, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 12.646940356312935, | |
| "grad_norm": 2.7474892139434814, | |
| "learning_rate": 3.6804091754494733e-06, | |
| "loss": 1.4776, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 12.70890782339272, | |
| "grad_norm": 3.022172451019287, | |
| "learning_rate": 3.64941103533788e-06, | |
| "loss": 1.498, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 12.770875290472501, | |
| "grad_norm": 2.747607946395874, | |
| "learning_rate": 3.6184128952262864e-06, | |
| "loss": 1.4901, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 12.832842757552285, | |
| "grad_norm": 3.077296733856201, | |
| "learning_rate": 3.5874147551146933e-06, | |
| "loss": 1.4773, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 12.894810224632067, | |
| "grad_norm": 4.3411078453063965, | |
| "learning_rate": 3.5564166150031e-06, | |
| "loss": 1.4909, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 12.956777691711851, | |
| "grad_norm": 3.112104654312134, | |
| "learning_rate": 3.5254184748915065e-06, | |
| "loss": 1.5027, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 12.999535243996903, | |
| "eval_loss": 1.3685630559921265, | |
| "eval_runtime": 70.8923, | |
| "eval_samples_per_second": 45.52, | |
| "eval_steps_per_second": 22.767, | |
| "step": 20969 | |
| }, | |
| { | |
| "epoch": 13.019209914794732, | |
| "grad_norm": 2.8162424564361572, | |
| "learning_rate": 3.4944203347799134e-06, | |
| "loss": 1.4868, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 13.081177381874516, | |
| "grad_norm": 2.862400770187378, | |
| "learning_rate": 3.4637321760694358e-06, | |
| "loss": 1.4565, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 13.143144848954298, | |
| "grad_norm": 9.166791915893555, | |
| "learning_rate": 3.4327340359578427e-06, | |
| "loss": 1.4968, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 13.205112316034082, | |
| "grad_norm": 2.9333505630493164, | |
| "learning_rate": 3.4017358958462493e-06, | |
| "loss": 1.4707, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 13.267079783113866, | |
| "grad_norm": 3.7271652221679688, | |
| "learning_rate": 3.370737755734656e-06, | |
| "loss": 1.4769, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 13.329047250193648, | |
| "grad_norm": 4.895938396453857, | |
| "learning_rate": 3.339739615623063e-06, | |
| "loss": 1.4825, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 13.391014717273432, | |
| "grad_norm": 4.098718643188477, | |
| "learning_rate": 3.3087414755114694e-06, | |
| "loss": 1.4473, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 13.452982184353214, | |
| "grad_norm": 2.393143892288208, | |
| "learning_rate": 3.277743335399876e-06, | |
| "loss": 1.4547, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 13.514949651432998, | |
| "grad_norm": 2.8273305892944336, | |
| "learning_rate": 3.246745195288283e-06, | |
| "loss": 1.4937, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 13.576917118512782, | |
| "grad_norm": 2.560944080352783, | |
| "learning_rate": 3.2157470551766895e-06, | |
| "loss": 1.4617, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 13.638884585592564, | |
| "grad_norm": 2.7508301734924316, | |
| "learning_rate": 3.184748915065096e-06, | |
| "loss": 1.4681, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 13.700852052672348, | |
| "grad_norm": 3.0896778106689453, | |
| "learning_rate": 3.153750774953503e-06, | |
| "loss": 1.4863, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 13.76281951975213, | |
| "grad_norm": 3.0587971210479736, | |
| "learning_rate": 3.1227526348419096e-06, | |
| "loss": 1.4872, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 13.824786986831914, | |
| "grad_norm": 5.594135284423828, | |
| "learning_rate": 3.091754494730316e-06, | |
| "loss": 1.4551, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 13.886754453911696, | |
| "grad_norm": 4.74953031539917, | |
| "learning_rate": 3.060756354618723e-06, | |
| "loss": 1.458, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 13.94872192099148, | |
| "grad_norm": 2.3097078800201416, | |
| "learning_rate": 3.0297582145071297e-06, | |
| "loss": 1.4707, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 13.999535243996903, | |
| "eval_loss": 1.367042899131775, | |
| "eval_runtime": 71.4468, | |
| "eval_samples_per_second": 45.167, | |
| "eval_steps_per_second": 22.59, | |
| "step": 22582 | |
| }, | |
| { | |
| "epoch": 14.01115414407436, | |
| "grad_norm": 2.73579478263855, | |
| "learning_rate": 2.9987600743955362e-06, | |
| "loss": 1.4815, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 14.073121611154145, | |
| "grad_norm": 2.643583297729492, | |
| "learning_rate": 2.967761934283943e-06, | |
| "loss": 1.4623, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 14.135089078233927, | |
| "grad_norm": 4.030912399291992, | |
| "learning_rate": 2.9367637941723498e-06, | |
| "loss": 1.4406, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 14.19705654531371, | |
| "grad_norm": 3.95310378074646, | |
| "learning_rate": 2.9057656540607563e-06, | |
| "loss": 1.4533, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 14.259024012393493, | |
| "grad_norm": 2.382369041442871, | |
| "learning_rate": 2.8747675139491633e-06, | |
| "loss": 1.4729, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 14.320991479473276, | |
| "grad_norm": 3.3369767665863037, | |
| "learning_rate": 2.84376937383757e-06, | |
| "loss": 1.447, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 14.38295894655306, | |
| "grad_norm": 3.532909393310547, | |
| "learning_rate": 2.8127712337259764e-06, | |
| "loss": 1.4646, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 14.444926413632842, | |
| "grad_norm": 2.707247257232666, | |
| "learning_rate": 2.7817730936143834e-06, | |
| "loss": 1.4782, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 14.506893880712626, | |
| "grad_norm": 2.511441707611084, | |
| "learning_rate": 2.75077495350279e-06, | |
| "loss": 1.4732, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 14.568861347792408, | |
| "grad_norm": 6.0035576820373535, | |
| "learning_rate": 2.7197768133911965e-06, | |
| "loss": 1.4625, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 14.630828814872192, | |
| "grad_norm": 3.396369695663452, | |
| "learning_rate": 2.6887786732796035e-06, | |
| "loss": 1.4794, | |
| "step": 23600 | |
| }, | |
| { | |
| "epoch": 14.692796281951976, | |
| "grad_norm": 2.5729544162750244, | |
| "learning_rate": 2.65778053316801e-06, | |
| "loss": 1.4666, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 14.754763749031758, | |
| "grad_norm": 2.8480477333068848, | |
| "learning_rate": 2.6270923744575328e-06, | |
| "loss": 1.4644, | |
| "step": 23800 | |
| }, | |
| { | |
| "epoch": 14.816731216111542, | |
| "grad_norm": 2.4807331562042236, | |
| "learning_rate": 2.5960942343459393e-06, | |
| "loss": 1.4752, | |
| "step": 23900 | |
| }, | |
| { | |
| "epoch": 14.878698683191324, | |
| "grad_norm": 2.850637435913086, | |
| "learning_rate": 2.5650960942343463e-06, | |
| "loss": 1.4769, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 14.940666150271108, | |
| "grad_norm": 2.480095148086548, | |
| "learning_rate": 2.534097954122753e-06, | |
| "loss": 1.4791, | |
| "step": 24100 | |
| }, | |
| { | |
| "epoch": 14.999535243996903, | |
| "eval_loss": 1.3670425415039062, | |
| "eval_runtime": 71.2802, | |
| "eval_samples_per_second": 45.272, | |
| "eval_steps_per_second": 22.643, | |
| "step": 24195 | |
| }, | |
| { | |
| "epoch": 15.003098373353989, | |
| "grad_norm": 2.5988998413085938, | |
| "learning_rate": 2.5030998140111594e-06, | |
| "loss": 1.4693, | |
| "step": 24200 | |
| }, | |
| { | |
| "epoch": 15.065065840433773, | |
| "grad_norm": 2.674956798553467, | |
| "learning_rate": 2.4721016738995664e-06, | |
| "loss": 1.4584, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 15.127033307513555, | |
| "grad_norm": 20.858081817626953, | |
| "learning_rate": 2.441103533787973e-06, | |
| "loss": 1.4593, | |
| "step": 24400 | |
| }, | |
| { | |
| "epoch": 15.189000774593339, | |
| "grad_norm": 2.9068284034729004, | |
| "learning_rate": 2.4101053936763795e-06, | |
| "loss": 1.437, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 15.250968241673121, | |
| "grad_norm": 2.5827035903930664, | |
| "learning_rate": 2.3791072535647865e-06, | |
| "loss": 1.4371, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 15.312935708752905, | |
| "grad_norm": 2.852088451385498, | |
| "learning_rate": 2.348109113453193e-06, | |
| "loss": 1.4434, | |
| "step": 24700 | |
| }, | |
| { | |
| "epoch": 15.374903175832689, | |
| "grad_norm": 3.28821063041687, | |
| "learning_rate": 2.3171109733415996e-06, | |
| "loss": 1.4606, | |
| "step": 24800 | |
| }, | |
| { | |
| "epoch": 15.43687064291247, | |
| "grad_norm": 3.762103319168091, | |
| "learning_rate": 2.2861128332300066e-06, | |
| "loss": 1.4584, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 15.498838109992255, | |
| "grad_norm": 2.845301866531372, | |
| "learning_rate": 2.255114693118413e-06, | |
| "loss": 1.4781, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 15.560805577072037, | |
| "grad_norm": 3.243618965148926, | |
| "learning_rate": 2.2241165530068197e-06, | |
| "loss": 1.4489, | |
| "step": 25100 | |
| }, | |
| { | |
| "epoch": 15.62277304415182, | |
| "grad_norm": 2.8567991256713867, | |
| "learning_rate": 2.1931184128952267e-06, | |
| "loss": 1.4768, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 15.684740511231603, | |
| "grad_norm": 2.683873414993286, | |
| "learning_rate": 2.162430254184749e-06, | |
| "loss": 1.4578, | |
| "step": 25300 | |
| }, | |
| { | |
| "epoch": 15.746707978311386, | |
| "grad_norm": 3.3914294242858887, | |
| "learning_rate": 2.131432114073156e-06, | |
| "loss": 1.4564, | |
| "step": 25400 | |
| }, | |
| { | |
| "epoch": 15.80867544539117, | |
| "grad_norm": 2.8761868476867676, | |
| "learning_rate": 2.1004339739615625e-06, | |
| "loss": 1.4648, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 15.870642912470952, | |
| "grad_norm": 3.788562059402466, | |
| "learning_rate": 2.069435833849969e-06, | |
| "loss": 1.4568, | |
| "step": 25600 | |
| }, | |
| { | |
| "epoch": 15.932610379550736, | |
| "grad_norm": 3.973054885864258, | |
| "learning_rate": 2.038437693738376e-06, | |
| "loss": 1.4785, | |
| "step": 25700 | |
| }, | |
| { | |
| "epoch": 15.994577846630518, | |
| "grad_norm": 4.635127544403076, | |
| "learning_rate": 2.0074395536267826e-06, | |
| "loss": 1.4741, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 15.999535243996903, | |
| "eval_loss": 1.3650240898132324, | |
| "eval_runtime": 71.1811, | |
| "eval_samples_per_second": 45.335, | |
| "eval_steps_per_second": 22.675, | |
| "step": 25808 | |
| }, | |
| { | |
| "epoch": 16.0570100697134, | |
| "grad_norm": 3.8094494342803955, | |
| "learning_rate": 1.976441413515189e-06, | |
| "loss": 1.4585, | |
| "step": 25900 | |
| }, | |
| { | |
| "epoch": 16.118977536793185, | |
| "grad_norm": 2.649723529815674, | |
| "learning_rate": 1.945443273403596e-06, | |
| "loss": 1.4624, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 16.180945003872967, | |
| "grad_norm": 2.537461757659912, | |
| "learning_rate": 1.9144451332920027e-06, | |
| "loss": 1.4576, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 16.24291247095275, | |
| "grad_norm": 3.1259541511535645, | |
| "learning_rate": 1.8834469931804095e-06, | |
| "loss": 1.4657, | |
| "step": 26200 | |
| }, | |
| { | |
| "epoch": 16.30487993803253, | |
| "grad_norm": 3.396057367324829, | |
| "learning_rate": 1.8524488530688158e-06, | |
| "loss": 1.4532, | |
| "step": 26300 | |
| }, | |
| { | |
| "epoch": 16.366847405112317, | |
| "grad_norm": 2.7809932231903076, | |
| "learning_rate": 1.8214507129572226e-06, | |
| "loss": 1.4483, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 16.4288148721921, | |
| "grad_norm": 4.369870662689209, | |
| "learning_rate": 1.7904525728456294e-06, | |
| "loss": 1.4457, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 16.49078233927188, | |
| "grad_norm": 2.7700088024139404, | |
| "learning_rate": 1.759454432734036e-06, | |
| "loss": 1.4722, | |
| "step": 26600 | |
| }, | |
| { | |
| "epoch": 16.552749806351667, | |
| "grad_norm": 2.55792236328125, | |
| "learning_rate": 1.7284562926224427e-06, | |
| "loss": 1.4242, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 16.61471727343145, | |
| "grad_norm": 3.0416712760925293, | |
| "learning_rate": 1.6974581525108495e-06, | |
| "loss": 1.4516, | |
| "step": 26800 | |
| }, | |
| { | |
| "epoch": 16.67668474051123, | |
| "grad_norm": 2.6362993717193604, | |
| "learning_rate": 1.666460012399256e-06, | |
| "loss": 1.4263, | |
| "step": 26900 | |
| }, | |
| { | |
| "epoch": 16.738652207591016, | |
| "grad_norm": 2.5275797843933105, | |
| "learning_rate": 1.6354618722876628e-06, | |
| "loss": 1.4513, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 16.8006196746708, | |
| "grad_norm": 3.758063793182373, | |
| "learning_rate": 1.6044637321760696e-06, | |
| "loss": 1.4629, | |
| "step": 27100 | |
| }, | |
| { | |
| "epoch": 16.86258714175058, | |
| "grad_norm": 2.653982162475586, | |
| "learning_rate": 1.5734655920644761e-06, | |
| "loss": 1.4549, | |
| "step": 27200 | |
| }, | |
| { | |
| "epoch": 16.924554608830363, | |
| "grad_norm": 3.276362180709839, | |
| "learning_rate": 1.5424674519528829e-06, | |
| "loss": 1.4645, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 16.98652207591015, | |
| "grad_norm": 4.8964056968688965, | |
| "learning_rate": 1.5114693118412897e-06, | |
| "loss": 1.4292, | |
| "step": 27400 | |
| }, | |
| { | |
| "epoch": 16.999535243996903, | |
| "eval_loss": 1.3658363819122314, | |
| "eval_runtime": 71.5249, | |
| "eval_samples_per_second": 45.117, | |
| "eval_steps_per_second": 22.566, | |
| "step": 27421 | |
| }, | |
| { | |
| "epoch": 17.048954298993028, | |
| "grad_norm": 2.632962703704834, | |
| "learning_rate": 1.4804711717296962e-06, | |
| "loss": 1.4485, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 17.110921766072813, | |
| "grad_norm": 2.65366792678833, | |
| "learning_rate": 1.449473031618103e-06, | |
| "loss": 1.4335, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 17.172889233152596, | |
| "grad_norm": 7.086144924163818, | |
| "learning_rate": 1.4184748915065097e-06, | |
| "loss": 1.4483, | |
| "step": 27700 | |
| }, | |
| { | |
| "epoch": 17.234856700232378, | |
| "grad_norm": 4.501598834991455, | |
| "learning_rate": 1.3874767513949163e-06, | |
| "loss": 1.4328, | |
| "step": 27800 | |
| }, | |
| { | |
| "epoch": 17.29682416731216, | |
| "grad_norm": 3.042269706726074, | |
| "learning_rate": 1.356478611283323e-06, | |
| "loss": 1.4476, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 17.358791634391945, | |
| "grad_norm": 2.902914047241211, | |
| "learning_rate": 1.3254804711717298e-06, | |
| "loss": 1.4551, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 17.420759101471727, | |
| "grad_norm": 4.015540599822998, | |
| "learning_rate": 1.2944823310601364e-06, | |
| "loss": 1.4408, | |
| "step": 28100 | |
| }, | |
| { | |
| "epoch": 17.48272656855151, | |
| "grad_norm": 2.1433181762695312, | |
| "learning_rate": 1.2634841909485432e-06, | |
| "loss": 1.4363, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 17.544694035631295, | |
| "grad_norm": 3.0657026767730713, | |
| "learning_rate": 1.23248605083695e-06, | |
| "loss": 1.4519, | |
| "step": 28300 | |
| }, | |
| { | |
| "epoch": 17.606661502711077, | |
| "grad_norm": 2.260270357131958, | |
| "learning_rate": 1.2014879107253565e-06, | |
| "loss": 1.4609, | |
| "step": 28400 | |
| }, | |
| { | |
| "epoch": 17.66862896979086, | |
| "grad_norm": 3.2962841987609863, | |
| "learning_rate": 1.1704897706137633e-06, | |
| "loss": 1.4267, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 17.73059643687064, | |
| "grad_norm": 2.8596572875976562, | |
| "learning_rate": 1.13949163050217e-06, | |
| "loss": 1.4485, | |
| "step": 28600 | |
| }, | |
| { | |
| "epoch": 17.792563903950427, | |
| "grad_norm": 2.8049371242523193, | |
| "learning_rate": 1.1084934903905766e-06, | |
| "loss": 1.4272, | |
| "step": 28700 | |
| }, | |
| { | |
| "epoch": 17.85453137103021, | |
| "grad_norm": 3.8641581535339355, | |
| "learning_rate": 1.0774953502789834e-06, | |
| "loss": 1.4426, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 17.91649883810999, | |
| "grad_norm": 2.718754529953003, | |
| "learning_rate": 1.0464972101673901e-06, | |
| "loss": 1.4696, | |
| "step": 28900 | |
| }, | |
| { | |
| "epoch": 17.978466305189777, | |
| "grad_norm": 2.8387906551361084, | |
| "learning_rate": 1.0154990700557969e-06, | |
| "loss": 1.437, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 17.999535243996903, | |
| "eval_loss": 1.3650579452514648, | |
| "eval_runtime": 71.088, | |
| "eval_samples_per_second": 45.394, | |
| "eval_steps_per_second": 22.704, | |
| "step": 29034 | |
| }, | |
| { | |
| "epoch": 18.040898528272656, | |
| "grad_norm": 7.15321159362793, | |
| "learning_rate": 9.845009299442034e-07, | |
| "loss": 1.4728, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 18.10286599535244, | |
| "grad_norm": 3.0295169353485107, | |
| "learning_rate": 9.535027898326102e-07, | |
| "loss": 1.444, | |
| "step": 29200 | |
| }, | |
| { | |
| "epoch": 18.164833462432224, | |
| "grad_norm": 4.30358362197876, | |
| "learning_rate": 9.225046497210168e-07, | |
| "loss": 1.4373, | |
| "step": 29300 | |
| }, | |
| { | |
| "epoch": 18.226800929512006, | |
| "grad_norm": 2.6451058387756348, | |
| "learning_rate": 8.915065096094234e-07, | |
| "loss": 1.4633, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 18.288768396591788, | |
| "grad_norm": 2.5222158432006836, | |
| "learning_rate": 8.605083694978301e-07, | |
| "loss": 1.4361, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 18.350735863671574, | |
| "grad_norm": 3.099586009979248, | |
| "learning_rate": 8.295102293862369e-07, | |
| "loss": 1.4451, | |
| "step": 29600 | |
| }, | |
| { | |
| "epoch": 18.412703330751356, | |
| "grad_norm": 2.569441795349121, | |
| "learning_rate": 7.985120892746435e-07, | |
| "loss": 1.4167, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 18.474670797831138, | |
| "grad_norm": 3.3484814167022705, | |
| "learning_rate": 7.678239305641662e-07, | |
| "loss": 1.4573, | |
| "step": 29800 | |
| }, | |
| { | |
| "epoch": 18.536638264910923, | |
| "grad_norm": 3.530923843383789, | |
| "learning_rate": 7.368257904525729e-07, | |
| "loss": 1.4653, | |
| "step": 29900 | |
| }, | |
| { | |
| "epoch": 18.598605731990705, | |
| "grad_norm": 2.9912829399108887, | |
| "learning_rate": 7.058276503409796e-07, | |
| "loss": 1.4376, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 18.660573199070488, | |
| "grad_norm": 4.254590034484863, | |
| "learning_rate": 6.748295102293862e-07, | |
| "loss": 1.4159, | |
| "step": 30100 | |
| }, | |
| { | |
| "epoch": 18.72254066615027, | |
| "grad_norm": 2.5701520442962646, | |
| "learning_rate": 6.43831370117793e-07, | |
| "loss": 1.4456, | |
| "step": 30200 | |
| }, | |
| { | |
| "epoch": 18.784508133230055, | |
| "grad_norm": 2.5587334632873535, | |
| "learning_rate": 6.131432114073156e-07, | |
| "loss": 1.4398, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 18.846475600309837, | |
| "grad_norm": 2.7531826496124268, | |
| "learning_rate": 5.821450712957223e-07, | |
| "loss": 1.4373, | |
| "step": 30400 | |
| }, | |
| { | |
| "epoch": 18.90844306738962, | |
| "grad_norm": 2.606565237045288, | |
| "learning_rate": 5.51146931184129e-07, | |
| "loss": 1.4322, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 18.970410534469405, | |
| "grad_norm": 2.6392128467559814, | |
| "learning_rate": 5.201487910725357e-07, | |
| "loss": 1.4442, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 18.999535243996903, | |
| "eval_loss": 1.3643269538879395, | |
| "eval_runtime": 71.2663, | |
| "eval_samples_per_second": 45.281, | |
| "eval_steps_per_second": 22.647, | |
| "step": 30647 | |
| }, | |
| { | |
| "epoch": 19.032842757552284, | |
| "grad_norm": 2.5503151416778564, | |
| "learning_rate": 4.891506509609424e-07, | |
| "loss": 1.4546, | |
| "step": 30700 | |
| }, | |
| { | |
| "epoch": 19.094810224632067, | |
| "grad_norm": 3.5481629371643066, | |
| "learning_rate": 4.5815251084934906e-07, | |
| "loss": 1.4389, | |
| "step": 30800 | |
| }, | |
| { | |
| "epoch": 19.156777691711852, | |
| "grad_norm": 2.849660634994507, | |
| "learning_rate": 4.271543707377557e-07, | |
| "loss": 1.4311, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 19.218745158791634, | |
| "grad_norm": 2.581972599029541, | |
| "learning_rate": 3.9615623062616244e-07, | |
| "loss": 1.4772, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 19.280712625871416, | |
| "grad_norm": 4.163836479187012, | |
| "learning_rate": 3.6515809051456916e-07, | |
| "loss": 1.4237, | |
| "step": 31100 | |
| }, | |
| { | |
| "epoch": 19.342680092951202, | |
| "grad_norm": 2.7274580001831055, | |
| "learning_rate": 3.341599504029758e-07, | |
| "loss": 1.4404, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 19.404647560030984, | |
| "grad_norm": 3.983724355697632, | |
| "learning_rate": 3.0316181029138253e-07, | |
| "loss": 1.4473, | |
| "step": 31300 | |
| }, | |
| { | |
| "epoch": 19.466615027110766, | |
| "grad_norm": 2.6508560180664062, | |
| "learning_rate": 2.7216367017978925e-07, | |
| "loss": 1.4563, | |
| "step": 31400 | |
| }, | |
| { | |
| "epoch": 19.52858249419055, | |
| "grad_norm": 2.608508348464966, | |
| "learning_rate": 2.4116553006819597e-07, | |
| "loss": 1.4381, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 19.590549961270334, | |
| "grad_norm": 3.2333600521087646, | |
| "learning_rate": 2.101673899566026e-07, | |
| "loss": 1.4418, | |
| "step": 31600 | |
| }, | |
| { | |
| "epoch": 19.652517428350116, | |
| "grad_norm": 5.008152008056641, | |
| "learning_rate": 1.7916924984500932e-07, | |
| "loss": 1.447, | |
| "step": 31700 | |
| }, | |
| { | |
| "epoch": 19.714484895429898, | |
| "grad_norm": 3.073624610900879, | |
| "learning_rate": 1.48171109733416e-07, | |
| "loss": 1.4241, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 19.776452362509684, | |
| "grad_norm": 5.159261226654053, | |
| "learning_rate": 1.171729696218227e-07, | |
| "loss": 1.424, | |
| "step": 31900 | |
| }, | |
| { | |
| "epoch": 19.838419829589466, | |
| "grad_norm": 3.225008249282837, | |
| "learning_rate": 8.61748295102294e-08, | |
| "loss": 1.4327, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 19.900387296669248, | |
| "grad_norm": 3.629281520843506, | |
| "learning_rate": 5.517668939863608e-08, | |
| "loss": 1.437, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 19.962354763749033, | |
| "grad_norm": 2.323373556137085, | |
| "learning_rate": 2.417854928704278e-08, | |
| "loss": 1.4345, | |
| "step": 32200 | |
| }, | |
| { | |
| "epoch": 19.999535243996903, | |
| "eval_loss": 1.3642308712005615, | |
| "eval_runtime": 70.2626, | |
| "eval_samples_per_second": 45.928, | |
| "eval_steps_per_second": 22.971, | |
| "step": 32260 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 32260, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.2226940169519104e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |