| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 555, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009029345372460496, | |
| "grad_norm": 4.1226773140670385, | |
| "learning_rate": 0.0, | |
| "loss": 1.3433, | |
| "num_tokens": 417963.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01805869074492099, | |
| "grad_norm": 3.9462061203782963, | |
| "learning_rate": 5.882352941176471e-07, | |
| "loss": 1.3076, | |
| "num_tokens": 856730.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02708803611738149, | |
| "grad_norm": 4.056538068886665, | |
| "learning_rate": 1.1764705882352942e-06, | |
| "loss": 1.3225, | |
| "num_tokens": 1283679.0, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03611738148984198, | |
| "grad_norm": 3.9035828019837617, | |
| "learning_rate": 1.7647058823529414e-06, | |
| "loss": 1.2971, | |
| "num_tokens": 1722056.0, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.045146726862302484, | |
| "grad_norm": 3.818499985659685, | |
| "learning_rate": 2.3529411764705885e-06, | |
| "loss": 1.3056, | |
| "num_tokens": 2150137.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05417607223476298, | |
| "grad_norm": 3.3482582599671074, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 1.2864, | |
| "num_tokens": 2564649.0, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06320541760722348, | |
| "grad_norm": 2.431710651256278, | |
| "learning_rate": 3.529411764705883e-06, | |
| "loss": 1.1649, | |
| "num_tokens": 2985977.0, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07223476297968397, | |
| "grad_norm": 2.291927554769836, | |
| "learning_rate": 4.11764705882353e-06, | |
| "loss": 1.1279, | |
| "num_tokens": 3413212.0, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.08126410835214447, | |
| "grad_norm": 2.029519481294733, | |
| "learning_rate": 4.705882352941177e-06, | |
| "loss": 0.9535, | |
| "num_tokens": 3843515.0, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.09029345372460497, | |
| "grad_norm": 1.8569383671360136, | |
| "learning_rate": 5.294117647058824e-06, | |
| "loss": 0.8774, | |
| "num_tokens": 4289732.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09932279909706546, | |
| "grad_norm": 1.9074099721896551, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 0.8476, | |
| "num_tokens": 4715096.0, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.10835214446952596, | |
| "grad_norm": 2.2685460452669806, | |
| "learning_rate": 6.470588235294119e-06, | |
| "loss": 0.5251, | |
| "num_tokens": 5128419.0, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.11738148984198646, | |
| "grad_norm": 2.261721528203699, | |
| "learning_rate": 7.058823529411766e-06, | |
| "loss": 0.4533, | |
| "num_tokens": 5540634.0, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.12641083521444696, | |
| "grad_norm": 1.8091458825349906, | |
| "learning_rate": 7.647058823529411e-06, | |
| "loss": 0.3695, | |
| "num_tokens": 5968005.0, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.13544018058690746, | |
| "grad_norm": 1.505512868285292, | |
| "learning_rate": 8.23529411764706e-06, | |
| "loss": 0.2675, | |
| "num_tokens": 6397897.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.14446952595936793, | |
| "grad_norm": 0.8544464295725124, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 0.1612, | |
| "num_tokens": 6815448.0, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.15349887133182843, | |
| "grad_norm": 0.6784675541905167, | |
| "learning_rate": 9.411764705882354e-06, | |
| "loss": 0.1197, | |
| "num_tokens": 7230528.0, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.16252821670428894, | |
| "grad_norm": 0.6379689058024756, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1302, | |
| "num_tokens": 7649044.0, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.17155756207674944, | |
| "grad_norm": 0.25897815088971404, | |
| "learning_rate": 9.999923278607256e-06, | |
| "loss": 0.0892, | |
| "num_tokens": 8067225.0, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.18058690744920994, | |
| "grad_norm": 0.244824125801969, | |
| "learning_rate": 9.999693117045099e-06, | |
| "loss": 0.0876, | |
| "num_tokens": 8500120.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.18961625282167044, | |
| "grad_norm": 0.4453079318767393, | |
| "learning_rate": 9.99930952316167e-06, | |
| "loss": 0.0798, | |
| "num_tokens": 8925262.0, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.1986455981941309, | |
| "grad_norm": 0.18027632385122908, | |
| "learning_rate": 9.998772510036905e-06, | |
| "loss": 0.0776, | |
| "num_tokens": 9354973.0, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2076749435665914, | |
| "grad_norm": 0.1682514733086765, | |
| "learning_rate": 9.998082095982091e-06, | |
| "loss": 0.0913, | |
| "num_tokens": 9795967.0, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.21670428893905191, | |
| "grad_norm": 0.6839625538532086, | |
| "learning_rate": 9.997238304539241e-06, | |
| "loss": 0.0805, | |
| "num_tokens": 10217971.0, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.22573363431151242, | |
| "grad_norm": 0.1801620006143571, | |
| "learning_rate": 9.99624116448029e-06, | |
| "loss": 0.1043, | |
| "num_tokens": 10675155.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.23476297968397292, | |
| "grad_norm": 0.1510431579702823, | |
| "learning_rate": 9.995090709806113e-06, | |
| "loss": 0.0894, | |
| "num_tokens": 11119872.0, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.24379232505643342, | |
| "grad_norm": 0.1552751490698237, | |
| "learning_rate": 9.993786979745374e-06, | |
| "loss": 0.0724, | |
| "num_tokens": 11538902.0, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.2528216704288939, | |
| "grad_norm": 0.13901109032980072, | |
| "learning_rate": 9.992330018753175e-06, | |
| "loss": 0.0592, | |
| "num_tokens": 11961057.0, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2618510158013544, | |
| "grad_norm": 0.12982893321708955, | |
| "learning_rate": 9.990719876509551e-06, | |
| "loss": 0.0575, | |
| "num_tokens": 12406878.0, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2708803611738149, | |
| "grad_norm": 0.13249591564705812, | |
| "learning_rate": 9.98895660791777e-06, | |
| "loss": 0.053, | |
| "num_tokens": 12819692.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2799097065462754, | |
| "grad_norm": 0.13079157117035578, | |
| "learning_rate": 9.987040273102466e-06, | |
| "loss": 0.0567, | |
| "num_tokens": 13240915.0, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.28893905191873587, | |
| "grad_norm": 0.12490016513516743, | |
| "learning_rate": 9.984970937407583e-06, | |
| "loss": 0.0597, | |
| "num_tokens": 13660214.0, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2979683972911964, | |
| "grad_norm": 0.12284775257936209, | |
| "learning_rate": 9.98274867139415e-06, | |
| "loss": 0.0485, | |
| "num_tokens": 14114321.0, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.30699774266365687, | |
| "grad_norm": 0.10605884412119998, | |
| "learning_rate": 9.980373550837877e-06, | |
| "loss": 0.0507, | |
| "num_tokens": 14548964.0, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3160270880361174, | |
| "grad_norm": 0.0996863253036289, | |
| "learning_rate": 9.977845656726565e-06, | |
| "loss": 0.0533, | |
| "num_tokens": 14981191.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.32505643340857787, | |
| "grad_norm": 0.099940697319749, | |
| "learning_rate": 9.975165075257351e-06, | |
| "loss": 0.0456, | |
| "num_tokens": 15397252.0, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3340857787810384, | |
| "grad_norm": 0.09953657791437007, | |
| "learning_rate": 9.972331897833766e-06, | |
| "loss": 0.0553, | |
| "num_tokens": 15825845.0, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3431151241534989, | |
| "grad_norm": 0.10522339551652994, | |
| "learning_rate": 9.96934622106262e-06, | |
| "loss": 0.048, | |
| "num_tokens": 16263335.0, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.35214446952595935, | |
| "grad_norm": 0.09154677645576041, | |
| "learning_rate": 9.966208146750697e-06, | |
| "loss": 0.0497, | |
| "num_tokens": 16685043.0, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.3611738148984199, | |
| "grad_norm": 0.09239203897782809, | |
| "learning_rate": 9.962917781901308e-06, | |
| "loss": 0.0447, | |
| "num_tokens": 17112536.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.37020316027088035, | |
| "grad_norm": 0.09443543473738454, | |
| "learning_rate": 9.959475238710617e-06, | |
| "loss": 0.038, | |
| "num_tokens": 17536800.0, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3792325056433409, | |
| "grad_norm": 0.08866051485745102, | |
| "learning_rate": 9.955880634563825e-06, | |
| "loss": 0.038, | |
| "num_tokens": 17972150.0, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.38826185101580135, | |
| "grad_norm": 0.0937980936876227, | |
| "learning_rate": 9.952134092031174e-06, | |
| "loss": 0.0454, | |
| "num_tokens": 18396216.0, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3972911963882618, | |
| "grad_norm": 0.08953079302859607, | |
| "learning_rate": 9.948235738863755e-06, | |
| "loss": 0.0395, | |
| "num_tokens": 18821149.0, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.40632054176072235, | |
| "grad_norm": 0.09638013349344338, | |
| "learning_rate": 9.944185707989163e-06, | |
| "loss": 0.0389, | |
| "num_tokens": 19245794.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4153498871331828, | |
| "grad_norm": 0.0963287325583386, | |
| "learning_rate": 9.939984137506954e-06, | |
| "loss": 0.0423, | |
| "num_tokens": 19678417.0, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.42437923250564336, | |
| "grad_norm": 0.09082871490975147, | |
| "learning_rate": 9.93563117068395e-06, | |
| "loss": 0.0377, | |
| "num_tokens": 20092889.0, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.43340857787810383, | |
| "grad_norm": 0.09446657995542025, | |
| "learning_rate": 9.93112695594934e-06, | |
| "loss": 0.0537, | |
| "num_tokens": 20547069.0, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.44243792325056436, | |
| "grad_norm": 0.08479659819661463, | |
| "learning_rate": 9.926471646889624e-06, | |
| "loss": 0.0355, | |
| "num_tokens": 20966538.0, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.45146726862302483, | |
| "grad_norm": 0.10103007225011962, | |
| "learning_rate": 9.921665402243376e-06, | |
| "loss": 0.0447, | |
| "num_tokens": 21400755.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4604966139954853, | |
| "grad_norm": 0.08624391161599489, | |
| "learning_rate": 9.916708385895837e-06, | |
| "loss": 0.0389, | |
| "num_tokens": 21840775.0, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.46952595936794583, | |
| "grad_norm": 0.08394108614320675, | |
| "learning_rate": 9.911600766873312e-06, | |
| "loss": 0.04, | |
| "num_tokens": 22278786.0, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.4785553047404063, | |
| "grad_norm": 0.08578722894055574, | |
| "learning_rate": 9.906342719337427e-06, | |
| "loss": 0.0368, | |
| "num_tokens": 22693466.0, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.48758465011286684, | |
| "grad_norm": 0.08306630403146477, | |
| "learning_rate": 9.900934422579167e-06, | |
| "loss": 0.0313, | |
| "num_tokens": 23112540.0, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4966139954853273, | |
| "grad_norm": 0.08907008819061979, | |
| "learning_rate": 9.895376061012786e-06, | |
| "loss": 0.0368, | |
| "num_tokens": 23564619.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.5056433408577878, | |
| "grad_norm": 0.09449233534316975, | |
| "learning_rate": 9.889667824169498e-06, | |
| "loss": 0.0347, | |
| "num_tokens": 23967845.0, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5146726862302483, | |
| "grad_norm": 0.13169270387732046, | |
| "learning_rate": 9.883809906691031e-06, | |
| "loss": 0.1075, | |
| "num_tokens": 24391020.0, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5237020316027088, | |
| "grad_norm": 0.08626372320969251, | |
| "learning_rate": 9.877802508322977e-06, | |
| "loss": 0.0397, | |
| "num_tokens": 24825527.0, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5327313769751693, | |
| "grad_norm": 0.6199046568163001, | |
| "learning_rate": 9.871645833907992e-06, | |
| "loss": 0.085, | |
| "num_tokens": 25263167.0, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5417607223476298, | |
| "grad_norm": 0.08854553441636301, | |
| "learning_rate": 9.865340093378799e-06, | |
| "loss": 0.0347, | |
| "num_tokens": 25693750.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5507900677200903, | |
| "grad_norm": 0.08190394530809236, | |
| "learning_rate": 9.858885501751044e-06, | |
| "loss": 0.0301, | |
| "num_tokens": 26127066.0, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.5598194130925508, | |
| "grad_norm": 0.07633990311461816, | |
| "learning_rate": 9.852282279115951e-06, | |
| "loss": 0.0331, | |
| "num_tokens": 26553930.0, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5688487584650113, | |
| "grad_norm": 0.10134813831731004, | |
| "learning_rate": 9.84553065063283e-06, | |
| "loss": 0.0482, | |
| "num_tokens": 26997155.0, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5778781038374717, | |
| "grad_norm": 0.07702444941063769, | |
| "learning_rate": 9.838630846521381e-06, | |
| "loss": 0.0289, | |
| "num_tokens": 27416930.0, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5869074492099323, | |
| "grad_norm": 0.08139774767813368, | |
| "learning_rate": 9.831583102053868e-06, | |
| "loss": 0.0342, | |
| "num_tokens": 27849198.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5959367945823928, | |
| "grad_norm": 0.08475766420893233, | |
| "learning_rate": 9.824387657547074e-06, | |
| "loss": 0.0326, | |
| "num_tokens": 28281608.0, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6049661399548533, | |
| "grad_norm": 0.08880561553586032, | |
| "learning_rate": 9.817044758354123e-06, | |
| "loss": 0.0275, | |
| "num_tokens": 28703014.0, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.6139954853273137, | |
| "grad_norm": 0.16108769242249113, | |
| "learning_rate": 9.809554654856106e-06, | |
| "loss": 0.1026, | |
| "num_tokens": 29154694.0, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6230248306997742, | |
| "grad_norm": 0.08758330536346154, | |
| "learning_rate": 9.80191760245354e-06, | |
| "loss": 0.0309, | |
| "num_tokens": 29581312.0, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.6320541760722348, | |
| "grad_norm": 0.08405143622629944, | |
| "learning_rate": 9.794133861557674e-06, | |
| "loss": 0.0311, | |
| "num_tokens": 30010189.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6410835214446953, | |
| "grad_norm": 0.09242952293604854, | |
| "learning_rate": 9.78620369758159e-06, | |
| "loss": 0.0308, | |
| "num_tokens": 30423268.0, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.6501128668171557, | |
| "grad_norm": 0.08547346542642349, | |
| "learning_rate": 9.778127380931165e-06, | |
| "loss": 0.0309, | |
| "num_tokens": 30854686.0, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6591422121896162, | |
| "grad_norm": 0.07877474464629973, | |
| "learning_rate": 9.76990518699585e-06, | |
| "loss": 0.0264, | |
| "num_tokens": 31260959.0, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6681715575620768, | |
| "grad_norm": 0.0952585998223962, | |
| "learning_rate": 9.761537396139277e-06, | |
| "loss": 0.0268, | |
| "num_tokens": 31676000.0, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6772009029345373, | |
| "grad_norm": 0.08841152027185054, | |
| "learning_rate": 9.753024293689696e-06, | |
| "loss": 0.0324, | |
| "num_tokens": 32086486.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6862302483069977, | |
| "grad_norm": 0.08632580074796341, | |
| "learning_rate": 9.744366169930254e-06, | |
| "loss": 0.0372, | |
| "num_tokens": 32498590.0, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6952595936794582, | |
| "grad_norm": 0.08545023498371142, | |
| "learning_rate": 9.735563320089088e-06, | |
| "loss": 0.03, | |
| "num_tokens": 32917173.0, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.7042889390519187, | |
| "grad_norm": 0.10023458177821297, | |
| "learning_rate": 9.72661604432927e-06, | |
| "loss": 0.0316, | |
| "num_tokens": 33349789.0, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7133182844243793, | |
| "grad_norm": 0.09633881794164335, | |
| "learning_rate": 9.717524647738553e-06, | |
| "loss": 0.031, | |
| "num_tokens": 33785863.0, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7223476297968398, | |
| "grad_norm": 0.08583028492237739, | |
| "learning_rate": 9.70828944031899e-06, | |
| "loss": 0.0299, | |
| "num_tokens": 34216063.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7313769751693002, | |
| "grad_norm": 0.08655284410423547, | |
| "learning_rate": 9.698910736976344e-06, | |
| "loss": 0.0261, | |
| "num_tokens": 34655546.0, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7404063205417607, | |
| "grad_norm": 0.08819526204263398, | |
| "learning_rate": 9.689388857509365e-06, | |
| "loss": 0.0335, | |
| "num_tokens": 35085479.0, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7494356659142212, | |
| "grad_norm": 0.0775793985272994, | |
| "learning_rate": 9.679724126598878e-06, | |
| "loss": 0.0254, | |
| "num_tokens": 35510555.0, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7584650112866818, | |
| "grad_norm": 0.09506327636871045, | |
| "learning_rate": 9.669916873796709e-06, | |
| "loss": 0.034, | |
| "num_tokens": 35921862.0, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7674943566591422, | |
| "grad_norm": 0.08310818380762558, | |
| "learning_rate": 9.659967433514458e-06, | |
| "loss": 0.0252, | |
| "num_tokens": 36328530.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7765237020316027, | |
| "grad_norm": 0.07740207993257811, | |
| "learning_rate": 9.649876145012085e-06, | |
| "loss": 0.0278, | |
| "num_tokens": 36772328.0, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7855530474040632, | |
| "grad_norm": 0.088775953244118, | |
| "learning_rate": 9.639643352386353e-06, | |
| "loss": 0.0316, | |
| "num_tokens": 37186946.0, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7945823927765236, | |
| "grad_norm": 0.08696148874584747, | |
| "learning_rate": 9.629269404559081e-06, | |
| "loss": 0.0249, | |
| "num_tokens": 37591509.0, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8036117381489842, | |
| "grad_norm": 0.1487676873622619, | |
| "learning_rate": 9.618754655265262e-06, | |
| "loss": 0.0498, | |
| "num_tokens": 38045842.0, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.8126410835214447, | |
| "grad_norm": 0.08230381913162506, | |
| "learning_rate": 9.608099463040989e-06, | |
| "loss": 0.0258, | |
| "num_tokens": 38465378.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8216704288939052, | |
| "grad_norm": 0.07869393467806507, | |
| "learning_rate": 9.597304191211228e-06, | |
| "loss": 0.0255, | |
| "num_tokens": 38900287.0, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8306997742663657, | |
| "grad_norm": 0.2570682466701682, | |
| "learning_rate": 9.586369207877449e-06, | |
| "loss": 0.0945, | |
| "num_tokens": 39323143.0, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8397291196388262, | |
| "grad_norm": 0.0846434867175951, | |
| "learning_rate": 9.575294885905051e-06, | |
| "loss": 0.0302, | |
| "num_tokens": 39745833.0, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8487584650112867, | |
| "grad_norm": 0.07921974056879383, | |
| "learning_rate": 9.564081602910654e-06, | |
| "loss": 0.0248, | |
| "num_tokens": 40155224.0, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.8577878103837472, | |
| "grad_norm": 0.08337454729573074, | |
| "learning_rate": 9.552729741249235e-06, | |
| "loss": 0.03, | |
| "num_tokens": 40600001.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8668171557562077, | |
| "grad_norm": 0.09668085159830789, | |
| "learning_rate": 9.541239688001076e-06, | |
| "loss": 0.0773, | |
| "num_tokens": 41034208.0, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8758465011286681, | |
| "grad_norm": 0.07704679084970735, | |
| "learning_rate": 9.52961183495857e-06, | |
| "loss": 0.0228, | |
| "num_tokens": 41463499.0, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8848758465011287, | |
| "grad_norm": 0.42293339948498004, | |
| "learning_rate": 9.517846578612866e-06, | |
| "loss": 0.0722, | |
| "num_tokens": 41905599.0, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.8939051918735892, | |
| "grad_norm": 0.09304706758598777, | |
| "learning_rate": 9.505944320140343e-06, | |
| "loss": 0.0286, | |
| "num_tokens": 42324749.0, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.9029345372460497, | |
| "grad_norm": 0.0756645296148333, | |
| "learning_rate": 9.49390546538893e-06, | |
| "loss": 0.0249, | |
| "num_tokens": 42745369.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9119638826185101, | |
| "grad_norm": 0.09351722784422487, | |
| "learning_rate": 9.481730424864276e-06, | |
| "loss": 0.0351, | |
| "num_tokens": 43186876.0, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.9209932279909706, | |
| "grad_norm": 0.07742636072299149, | |
| "learning_rate": 9.469419613715743e-06, | |
| "loss": 0.0231, | |
| "num_tokens": 43605408.0, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9300225733634312, | |
| "grad_norm": 0.08249628831703604, | |
| "learning_rate": 9.456973451722255e-06, | |
| "loss": 0.0252, | |
| "num_tokens": 44020619.0, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9390519187358917, | |
| "grad_norm": 0.08239928085619691, | |
| "learning_rate": 9.44439236327798e-06, | |
| "loss": 0.0357, | |
| "num_tokens": 44482117.0, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.9480812641083521, | |
| "grad_norm": 0.07649857341695085, | |
| "learning_rate": 9.431676777377865e-06, | |
| "loss": 0.0255, | |
| "num_tokens": 44899252.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9571106094808126, | |
| "grad_norm": 0.07792520551168008, | |
| "learning_rate": 9.418827127603e-06, | |
| "loss": 0.029, | |
| "num_tokens": 45332722.0, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.9661399548532731, | |
| "grad_norm": 0.07612403843116701, | |
| "learning_rate": 9.405843852105846e-06, | |
| "loss": 0.0223, | |
| "num_tokens": 45764262.0, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.9751693002257337, | |
| "grad_norm": 0.0785710592412512, | |
| "learning_rate": 9.392727393595278e-06, | |
| "loss": 0.0232, | |
| "num_tokens": 46186006.0, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.9841986455981941, | |
| "grad_norm": 0.07886878116003551, | |
| "learning_rate": 9.379478199321508e-06, | |
| "loss": 0.038, | |
| "num_tokens": 46639566.0, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9932279909706546, | |
| "grad_norm": 0.07681915336448678, | |
| "learning_rate": 9.366096721060817e-06, | |
| "loss": 0.0264, | |
| "num_tokens": 47064244.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.07681915336448678, | |
| "learning_rate": 9.352583415100157e-06, | |
| "loss": 0.1022, | |
| "num_tokens": 47393855.0, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.08535497635602951, | |
| "eval_num_tokens": 47393855.0, | |
| "eval_runtime": 56.3284, | |
| "eval_samples_per_second": 44.471, | |
| "eval_steps_per_second": 5.574, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.0090293453724606, | |
| "grad_norm": 0.17356249840222848, | |
| "learning_rate": 9.3389387422216e-06, | |
| "loss": 0.0204, | |
| "num_tokens": 47806383.0, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.018058690744921, | |
| "grad_norm": 0.07958262229456115, | |
| "learning_rate": 9.325163167686615e-06, | |
| "loss": 0.0202, | |
| "num_tokens": 48241862.0, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0270880361173815, | |
| "grad_norm": 0.07642681533916272, | |
| "learning_rate": 9.311257161220207e-06, | |
| "loss": 0.0241, | |
| "num_tokens": 48651747.0, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.036117381489842, | |
| "grad_norm": 0.07480863020983759, | |
| "learning_rate": 9.297221196994904e-06, | |
| "loss": 0.0566, | |
| "num_tokens": 49088950.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0451467268623025, | |
| "grad_norm": 0.22260772897039935, | |
| "learning_rate": 9.283055753614581e-06, | |
| "loss": 0.0326, | |
| "num_tokens": 49528452.0, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.054176072234763, | |
| "grad_norm": 0.08495744331654775, | |
| "learning_rate": 9.268761314098148e-06, | |
| "loss": 0.0251, | |
| "num_tokens": 49951741.0, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.0632054176072234, | |
| "grad_norm": 0.08591356397624175, | |
| "learning_rate": 9.254338365863079e-06, | |
| "loss": 0.0408, | |
| "num_tokens": 50388295.0, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.072234762979684, | |
| "grad_norm": 0.08378209224429746, | |
| "learning_rate": 9.239787400708779e-06, | |
| "loss": 0.0372, | |
| "num_tokens": 50829191.0, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.0812641083521444, | |
| "grad_norm": 0.08856082886989099, | |
| "learning_rate": 9.225108914799833e-06, | |
| "loss": 0.0215, | |
| "num_tokens": 51266209.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.090293453724605, | |
| "grad_norm": 0.08191730251281082, | |
| "learning_rate": 9.21030340864908e-06, | |
| "loss": 0.0241, | |
| "num_tokens": 51687096.0, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0993227990970655, | |
| "grad_norm": 0.08193039384365579, | |
| "learning_rate": 9.195371387100544e-06, | |
| "loss": 0.0185, | |
| "num_tokens": 52114071.0, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.108352144469526, | |
| "grad_norm": 0.07999371830102796, | |
| "learning_rate": 9.180313359312218e-06, | |
| "loss": 0.0217, | |
| "num_tokens": 52541213.0, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.1173814898419865, | |
| "grad_norm": 0.07935868903198422, | |
| "learning_rate": 9.165129838738706e-06, | |
| "loss": 0.0223, | |
| "num_tokens": 52966113.0, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.1264108352144468, | |
| "grad_norm": 0.08121184818306772, | |
| "learning_rate": 9.14982134311372e-06, | |
| "loss": 0.0269, | |
| "num_tokens": 53418877.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1354401805869074, | |
| "grad_norm": 0.10411396152761351, | |
| "learning_rate": 9.13438839443242e-06, | |
| "loss": 0.0212, | |
| "num_tokens": 53839935.0, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.144469525959368, | |
| "grad_norm": 0.07346431042302812, | |
| "learning_rate": 9.11883151893361e-06, | |
| "loss": 0.0179, | |
| "num_tokens": 54258590.0, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.1534988713318284, | |
| "grad_norm": 0.0740206025776819, | |
| "learning_rate": 9.103151247081803e-06, | |
| "loss": 0.0212, | |
| "num_tokens": 54666548.0, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.162528216704289, | |
| "grad_norm": 0.08228008762099617, | |
| "learning_rate": 9.087348113549134e-06, | |
| "loss": 0.0201, | |
| "num_tokens": 55091316.0, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.1715575620767495, | |
| "grad_norm": 0.0779148259917351, | |
| "learning_rate": 9.071422657197117e-06, | |
| "loss": 0.0202, | |
| "num_tokens": 55507416.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.18058690744921, | |
| "grad_norm": 0.08002903746993835, | |
| "learning_rate": 9.05537542105828e-06, | |
| "loss": 0.0238, | |
| "num_tokens": 55947804.0, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.1896162528216705, | |
| "grad_norm": 0.07906565123280986, | |
| "learning_rate": 9.039206952317655e-06, | |
| "loss": 0.0186, | |
| "num_tokens": 56371231.0, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.1986455981941309, | |
| "grad_norm": 0.07489036883955975, | |
| "learning_rate": 9.022917802294098e-06, | |
| "loss": 0.02, | |
| "num_tokens": 56783891.0, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.2076749435665914, | |
| "grad_norm": 0.07820510364094234, | |
| "learning_rate": 9.006508526421511e-06, | |
| "loss": 0.0216, | |
| "num_tokens": 57202386.0, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.2167042889390518, | |
| "grad_norm": 0.09106232664538468, | |
| "learning_rate": 8.989979684229894e-06, | |
| "loss": 0.0188, | |
| "num_tokens": 57631769.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.2257336343115124, | |
| "grad_norm": 0.08349103688477277, | |
| "learning_rate": 8.973331839326266e-06, | |
| "loss": 0.022, | |
| "num_tokens": 58062896.0, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.234762979683973, | |
| "grad_norm": 0.07090655801699461, | |
| "learning_rate": 8.956565559375452e-06, | |
| "loss": 0.0215, | |
| "num_tokens": 58487448.0, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.2437923250564333, | |
| "grad_norm": 0.08633882584217765, | |
| "learning_rate": 8.93968141608071e-06, | |
| "loss": 0.0207, | |
| "num_tokens": 58922017.0, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.252821670428894, | |
| "grad_norm": 0.07728014844526158, | |
| "learning_rate": 8.922679985164262e-06, | |
| "loss": 0.022, | |
| "num_tokens": 59367436.0, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.2618510158013545, | |
| "grad_norm": 0.09084101538743762, | |
| "learning_rate": 8.905561846347648e-06, | |
| "loss": 0.0232, | |
| "num_tokens": 59795521.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2708803611738149, | |
| "grad_norm": 0.08500823309230846, | |
| "learning_rate": 8.888327583331953e-06, | |
| "loss": 0.0185, | |
| "num_tokens": 60200448.0, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.2799097065462754, | |
| "grad_norm": 0.08835861679468875, | |
| "learning_rate": 8.870977783777917e-06, | |
| "loss": 0.0207, | |
| "num_tokens": 60630275.0, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.2889390519187358, | |
| "grad_norm": 0.08076126658072942, | |
| "learning_rate": 8.853513039285888e-06, | |
| "loss": 0.0328, | |
| "num_tokens": 61053633.0, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.2979683972911964, | |
| "grad_norm": 0.1384536355652561, | |
| "learning_rate": 8.835933945375654e-06, | |
| "loss": 0.0234, | |
| "num_tokens": 61498301.0, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.3069977426636568, | |
| "grad_norm": 0.0829871254442218, | |
| "learning_rate": 8.818241101466135e-06, | |
| "loss": 0.0205, | |
| "num_tokens": 61937666.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.3160270880361173, | |
| "grad_norm": 0.082910689437852, | |
| "learning_rate": 8.800435110854943e-06, | |
| "loss": 0.0201, | |
| "num_tokens": 62349356.0, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.325056433408578, | |
| "grad_norm": 0.07645576769781585, | |
| "learning_rate": 8.78251658069781e-06, | |
| "loss": 0.0222, | |
| "num_tokens": 62773669.0, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.3340857787810383, | |
| "grad_norm": 0.08787635923368299, | |
| "learning_rate": 8.764486121987885e-06, | |
| "loss": 0.0185, | |
| "num_tokens": 63184808.0, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.3431151241534989, | |
| "grad_norm": 0.07676125126637708, | |
| "learning_rate": 8.746344349534905e-06, | |
| "loss": 0.0181, | |
| "num_tokens": 63626245.0, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.3521444695259595, | |
| "grad_norm": 0.0694959330953695, | |
| "learning_rate": 8.728091881944226e-06, | |
| "loss": 0.0213, | |
| "num_tokens": 64068177.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3611738148984198, | |
| "grad_norm": 0.08269786083124728, | |
| "learning_rate": 8.70972934159573e-06, | |
| "loss": 0.0222, | |
| "num_tokens": 64496374.0, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.3702031602708804, | |
| "grad_norm": 0.09481830509518524, | |
| "learning_rate": 8.691257354622602e-06, | |
| "loss": 0.0184, | |
| "num_tokens": 64924919.0, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.379232505643341, | |
| "grad_norm": 0.07325200785757595, | |
| "learning_rate": 8.672676550889985e-06, | |
| "loss": 0.0228, | |
| "num_tokens": 65360268.0, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.3882618510158014, | |
| "grad_norm": 0.08206745377927094, | |
| "learning_rate": 8.653987563973494e-06, | |
| "loss": 0.0182, | |
| "num_tokens": 65781587.0, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.3972911963882617, | |
| "grad_norm": 0.07582039793883952, | |
| "learning_rate": 8.635191031137624e-06, | |
| "loss": 0.0183, | |
| "num_tokens": 66191116.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.4063205417607223, | |
| "grad_norm": 0.0746601188535464, | |
| "learning_rate": 8.616287593314006e-06, | |
| "loss": 0.0205, | |
| "num_tokens": 66617225.0, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.4153498871331829, | |
| "grad_norm": 0.0802654573159223, | |
| "learning_rate": 8.597277895079568e-06, | |
| "loss": 0.0409, | |
| "num_tokens": 67046148.0, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.4243792325056432, | |
| "grad_norm": 0.11469234518973448, | |
| "learning_rate": 8.578162584634537e-06, | |
| "loss": 0.02, | |
| "num_tokens": 67476904.0, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.4334085778781038, | |
| "grad_norm": 0.07976570567339254, | |
| "learning_rate": 8.558942313780357e-06, | |
| "loss": 0.078, | |
| "num_tokens": 67946595.0, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.4424379232505644, | |
| "grad_norm": 0.20695421562611824, | |
| "learning_rate": 8.539617737897452e-06, | |
| "loss": 0.0194, | |
| "num_tokens": 68359880.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.4514672686230248, | |
| "grad_norm": 0.08200978654094307, | |
| "learning_rate": 8.520189515922872e-06, | |
| "loss": 0.0253, | |
| "num_tokens": 68795114.0, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.4604966139954854, | |
| "grad_norm": 0.08580590592640575, | |
| "learning_rate": 8.500658310327842e-06, | |
| "loss": 0.0215, | |
| "num_tokens": 69225525.0, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.469525959367946, | |
| "grad_norm": 0.0789594468693246, | |
| "learning_rate": 8.48102478709516e-06, | |
| "loss": 0.0205, | |
| "num_tokens": 69645836.0, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.4785553047404063, | |
| "grad_norm": 0.0735719259083779, | |
| "learning_rate": 8.461289615696489e-06, | |
| "loss": 0.0188, | |
| "num_tokens": 70067001.0, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.487584650112867, | |
| "grad_norm": 0.07613369106659294, | |
| "learning_rate": 8.441453469069536e-06, | |
| "loss": 0.0296, | |
| "num_tokens": 70492907.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.4966139954853273, | |
| "grad_norm": 0.07543177380385463, | |
| "learning_rate": 8.4215170235951e-06, | |
| "loss": 0.0316, | |
| "num_tokens": 70942223.0, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.5056433408577878, | |
| "grad_norm": 0.09008361072430175, | |
| "learning_rate": 8.401480959074006e-06, | |
| "loss": 0.0176, | |
| "num_tokens": 71371574.0, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.5146726862302482, | |
| "grad_norm": 0.08346708149449512, | |
| "learning_rate": 8.381345958703933e-06, | |
| "loss": 0.0502, | |
| "num_tokens": 71822208.0, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.5237020316027088, | |
| "grad_norm": 0.12181670655808274, | |
| "learning_rate": 8.361112709056115e-06, | |
| "loss": 0.0168, | |
| "num_tokens": 72234675.0, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.5327313769751694, | |
| "grad_norm": 0.07503087447812291, | |
| "learning_rate": 8.340781900051924e-06, | |
| "loss": 0.0185, | |
| "num_tokens": 72652286.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.5417607223476297, | |
| "grad_norm": 0.08398432499044427, | |
| "learning_rate": 8.32035422493935e-06, | |
| "loss": 0.0253, | |
| "num_tokens": 73068061.0, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.5507900677200903, | |
| "grad_norm": 0.07872465338491443, | |
| "learning_rate": 8.299830380269372e-06, | |
| "loss": 0.0213, | |
| "num_tokens": 73496804.0, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.559819413092551, | |
| "grad_norm": 0.0839265636209512, | |
| "learning_rate": 8.27921106587218e-06, | |
| "loss": 0.0182, | |
| "num_tokens": 73913610.0, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.5688487584650113, | |
| "grad_norm": 0.07717964434903289, | |
| "learning_rate": 8.258496984833344e-06, | |
| "loss": 0.0191, | |
| "num_tokens": 74340086.0, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.5778781038374716, | |
| "grad_norm": 0.09088946896779272, | |
| "learning_rate": 8.237688843469815e-06, | |
| "loss": 0.018, | |
| "num_tokens": 74761072.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.5869074492099324, | |
| "grad_norm": 0.08587530602264679, | |
| "learning_rate": 8.216787351305854e-06, | |
| "loss": 0.0341, | |
| "num_tokens": 75200788.0, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.5959367945823928, | |
| "grad_norm": 0.0895332050746481, | |
| "learning_rate": 8.195793221048834e-06, | |
| "loss": 0.0172, | |
| "num_tokens": 75626000.0, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.6049661399548532, | |
| "grad_norm": 0.07005785017925037, | |
| "learning_rate": 8.17470716856494e-06, | |
| "loss": 0.0203, | |
| "num_tokens": 76049835.0, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.6139954853273137, | |
| "grad_norm": 0.06991469620387697, | |
| "learning_rate": 8.153529912854751e-06, | |
| "loss": 0.0161, | |
| "num_tokens": 76485166.0, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.6230248306997743, | |
| "grad_norm": 0.07435841338293188, | |
| "learning_rate": 8.13226217602874e-06, | |
| "loss": 0.0181, | |
| "num_tokens": 76889061.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.6320541760722347, | |
| "grad_norm": 0.07690872751853474, | |
| "learning_rate": 8.110904683282635e-06, | |
| "loss": 0.0924, | |
| "num_tokens": 77325361.0, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.6410835214446953, | |
| "grad_norm": 0.17850649612473476, | |
| "learning_rate": 8.089458162872697e-06, | |
| "loss": 0.0168, | |
| "num_tokens": 77759756.0, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.6501128668171559, | |
| "grad_norm": 0.06746756510504061, | |
| "learning_rate": 8.067923346090888e-06, | |
| "loss": 0.0179, | |
| "num_tokens": 78187190.0, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.6591422121896162, | |
| "grad_norm": 0.07885448482479361, | |
| "learning_rate": 8.046300967239934e-06, | |
| "loss": 0.0166, | |
| "num_tokens": 78598237.0, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.6681715575620768, | |
| "grad_norm": 0.06890297475547125, | |
| "learning_rate": 8.024591763608291e-06, | |
| "loss": 0.0171, | |
| "num_tokens": 79028466.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.6772009029345374, | |
| "grad_norm": 0.0751884704197381, | |
| "learning_rate": 8.002796475444995e-06, | |
| "loss": 0.0179, | |
| "num_tokens": 79444122.0, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.6862302483069977, | |
| "grad_norm": 0.07290866223712479, | |
| "learning_rate": 7.980915845934433e-06, | |
| "loss": 0.0194, | |
| "num_tokens": 79857405.0, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.695259593679458, | |
| "grad_norm": 0.0792398036644149, | |
| "learning_rate": 7.95895062117099e-06, | |
| "loss": 0.0171, | |
| "num_tokens": 80293603.0, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.7042889390519187, | |
| "grad_norm": 0.07077113041595946, | |
| "learning_rate": 7.936901550133616e-06, | |
| "loss": 0.0536, | |
| "num_tokens": 80728108.0, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.7133182844243793, | |
| "grad_norm": 0.47740128118493913, | |
| "learning_rate": 7.914769384660283e-06, | |
| "loss": 0.0178, | |
| "num_tokens": 81173331.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.7223476297968396, | |
| "grad_norm": 0.08235180008975489, | |
| "learning_rate": 7.892554879422351e-06, | |
| "loss": 0.024, | |
| "num_tokens": 81579387.0, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.7313769751693002, | |
| "grad_norm": 0.07101875141394902, | |
| "learning_rate": 7.870258791898832e-06, | |
| "loss": 0.0189, | |
| "num_tokens": 82013021.0, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.7404063205417608, | |
| "grad_norm": 0.07249721287890425, | |
| "learning_rate": 7.847881882350568e-06, | |
| "loss": 0.0169, | |
| "num_tokens": 82440490.0, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.7494356659142212, | |
| "grad_norm": 0.07095993054906387, | |
| "learning_rate": 7.825424913794299e-06, | |
| "loss": 0.0193, | |
| "num_tokens": 82879243.0, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.7584650112866818, | |
| "grad_norm": 0.06608741875722214, | |
| "learning_rate": 7.802888651976647e-06, | |
| "loss": 0.017, | |
| "num_tokens": 83320595.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.7674943566591423, | |
| "grad_norm": 0.07043212785004925, | |
| "learning_rate": 7.78027386534801e-06, | |
| "loss": 0.0167, | |
| "num_tokens": 83736643.0, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.7765237020316027, | |
| "grad_norm": 0.07424356993552451, | |
| "learning_rate": 7.757581325036357e-06, | |
| "loss": 0.027, | |
| "num_tokens": 84180990.0, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.785553047404063, | |
| "grad_norm": 0.07082819077709723, | |
| "learning_rate": 7.73481180482093e-06, | |
| "loss": 0.0184, | |
| "num_tokens": 84610497.0, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.7945823927765236, | |
| "grad_norm": 0.0765803358413884, | |
| "learning_rate": 7.711966081105863e-06, | |
| "loss": 0.0299, | |
| "num_tokens": 85060786.0, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.8036117381489842, | |
| "grad_norm": 0.09743570524376206, | |
| "learning_rate": 7.68904493289371e-06, | |
| "loss": 0.0159, | |
| "num_tokens": 85477616.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.8126410835214446, | |
| "grad_norm": 0.06763165044353958, | |
| "learning_rate": 7.666049141758878e-06, | |
| "loss": 0.0668, | |
| "num_tokens": 85930357.0, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.8216704288939052, | |
| "grad_norm": 0.13642266006295956, | |
| "learning_rate": 7.642979491820974e-06, | |
| "loss": 0.0169, | |
| "num_tokens": 86344995.0, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.8306997742663658, | |
| "grad_norm": 0.07068370731607548, | |
| "learning_rate": 7.619836769718075e-06, | |
| "loss": 0.0158, | |
| "num_tokens": 86757489.0, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.8397291196388261, | |
| "grad_norm": 0.07345491867594905, | |
| "learning_rate": 7.596621764579904e-06, | |
| "loss": 0.0175, | |
| "num_tokens": 87187973.0, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.8487584650112867, | |
| "grad_norm": 0.0842192458018367, | |
| "learning_rate": 7.573335268000918e-06, | |
| "loss": 0.0253, | |
| "num_tokens": 87614553.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.8577878103837473, | |
| "grad_norm": 0.06522749322895638, | |
| "learning_rate": 7.549978074013314e-06, | |
| "loss": 0.0168, | |
| "num_tokens": 88048570.0, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.8668171557562077, | |
| "grad_norm": 0.07495632464623402, | |
| "learning_rate": 7.5265509790599625e-06, | |
| "loss": 0.0166, | |
| "num_tokens": 88466238.0, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.875846501128668, | |
| "grad_norm": 0.07359735329146656, | |
| "learning_rate": 7.503054781967241e-06, | |
| "loss": 0.0185, | |
| "num_tokens": 88895453.0, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.8848758465011288, | |
| "grad_norm": 0.09710562224494881, | |
| "learning_rate": 7.479490283917802e-06, | |
| "loss": 0.0659, | |
| "num_tokens": 89349875.0, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.8939051918735892, | |
| "grad_norm": 0.0702463525638772, | |
| "learning_rate": 7.455858288423249e-06, | |
| "loss": 0.0166, | |
| "num_tokens": 89792563.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.9029345372460496, | |
| "grad_norm": 0.06912331304375105, | |
| "learning_rate": 7.43215960129674e-06, | |
| "loss": 0.0453, | |
| "num_tokens": 90234898.0, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.9119638826185101, | |
| "grad_norm": 0.25433782045873404, | |
| "learning_rate": 7.408395030625513e-06, | |
| "loss": 0.0176, | |
| "num_tokens": 90648682.0, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.9209932279909707, | |
| "grad_norm": 0.07360863662739528, | |
| "learning_rate": 7.384565386743327e-06, | |
| "loss": 0.0181, | |
| "num_tokens": 91063468.0, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.930022573363431, | |
| "grad_norm": 0.07509698446261776, | |
| "learning_rate": 7.360671482202838e-06, | |
| "loss": 0.0252, | |
| "num_tokens": 91494817.0, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.9390519187358917, | |
| "grad_norm": 0.07276762866373235, | |
| "learning_rate": 7.336714131747878e-06, | |
| "loss": 0.0173, | |
| "num_tokens": 91925573.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.9480812641083523, | |
| "grad_norm": 0.0674231394947175, | |
| "learning_rate": 7.312694152285691e-06, | |
| "loss": 0.0657, | |
| "num_tokens": 92350433.0, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.9571106094808126, | |
| "grad_norm": 0.07775923659337187, | |
| "learning_rate": 7.288612362859066e-06, | |
| "loss": 0.0163, | |
| "num_tokens": 92765584.0, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.966139954853273, | |
| "grad_norm": 0.075222930525113, | |
| "learning_rate": 7.2644695846184165e-06, | |
| "loss": 0.0178, | |
| "num_tokens": 93229192.0, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.9751693002257338, | |
| "grad_norm": 0.06923809192473053, | |
| "learning_rate": 7.240266640793774e-06, | |
| "loss": 0.0148, | |
| "num_tokens": 93636026.0, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.9841986455981941, | |
| "grad_norm": 0.06904441894343259, | |
| "learning_rate": 7.216004356666717e-06, | |
| "loss": 0.0173, | |
| "num_tokens": 94060859.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.9932279909706545, | |
| "grad_norm": 0.07164737021846543, | |
| "learning_rate": 7.191683559542238e-06, | |
| "loss": 0.0154, | |
| "num_tokens": 94478604.0, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.07299986656925679, | |
| "learning_rate": 7.167305078720527e-06, | |
| "loss": 0.0152, | |
| "num_tokens": 94789780.0, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.09336748719215393, | |
| "eval_num_tokens": 94789780.0, | |
| "eval_runtime": 55.4758, | |
| "eval_samples_per_second": 45.155, | |
| "eval_steps_per_second": 5.66, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.0090293453724604, | |
| "grad_norm": 0.08097562146839203, | |
| "learning_rate": 7.142869745468697e-06, | |
| "loss": 0.0141, | |
| "num_tokens": 95206304.0, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.018058690744921, | |
| "grad_norm": 0.0657521707564096, | |
| "learning_rate": 7.118378392992436e-06, | |
| "loss": 0.0128, | |
| "num_tokens": 95629591.0, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.0270880361173815, | |
| "grad_norm": 0.06260152099136218, | |
| "learning_rate": 7.093831856407599e-06, | |
| "loss": 0.0247, | |
| "num_tokens": 96043349.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.036117381489842, | |
| "grad_norm": 0.0787462901317735, | |
| "learning_rate": 7.069230972711727e-06, | |
| "loss": 0.0136, | |
| "num_tokens": 96461415.0, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.0451467268623027, | |
| "grad_norm": 0.06859199368166498, | |
| "learning_rate": 7.044576580755517e-06, | |
| "loss": 0.0125, | |
| "num_tokens": 96920134.0, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.054176072234763, | |
| "grad_norm": 0.07142924118779401, | |
| "learning_rate": 7.019869521214206e-06, | |
| "loss": 0.0149, | |
| "num_tokens": 97365121.0, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.0632054176072234, | |
| "grad_norm": 0.06946075051983241, | |
| "learning_rate": 6.995110636558916e-06, | |
| "loss": 0.0132, | |
| "num_tokens": 97777134.0, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.072234762979684, | |
| "grad_norm": 0.07132704808421908, | |
| "learning_rate": 6.970300771027914e-06, | |
| "loss": 0.0135, | |
| "num_tokens": 98209859.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.0812641083521446, | |
| "grad_norm": 0.0787937957547939, | |
| "learning_rate": 6.945440770597845e-06, | |
| "loss": 0.0166, | |
| "num_tokens": 98622007.0, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.090293453724605, | |
| "grad_norm": 0.06889250982502683, | |
| "learning_rate": 6.920531482954863e-06, | |
| "loss": 0.0139, | |
| "num_tokens": 99052847.0, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.0993227990970653, | |
| "grad_norm": 0.06756115890886172, | |
| "learning_rate": 6.895573757465745e-06, | |
| "loss": 0.0128, | |
| "num_tokens": 99471292.0, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.108352144469526, | |
| "grad_norm": 0.06696980846993505, | |
| "learning_rate": 6.870568445148915e-06, | |
| "loss": 0.0133, | |
| "num_tokens": 99901083.0, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.1173814898419865, | |
| "grad_norm": 0.07298831986544797, | |
| "learning_rate": 6.845516398645434e-06, | |
| "loss": 0.0222, | |
| "num_tokens": 100323717.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.126410835214447, | |
| "grad_norm": 0.07277884406791328, | |
| "learning_rate": 6.820418472189926e-06, | |
| "loss": 0.0137, | |
| "num_tokens": 100746774.0, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.1354401805869077, | |
| "grad_norm": 0.07572502256810043, | |
| "learning_rate": 6.795275521581443e-06, | |
| "loss": 0.0592, | |
| "num_tokens": 101186553.0, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 2.144469525959368, | |
| "grad_norm": 0.1286025399754837, | |
| "learning_rate": 6.770088404154293e-06, | |
| "loss": 0.0146, | |
| "num_tokens": 101595403.0, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.1534988713318284, | |
| "grad_norm": 0.07448253421632307, | |
| "learning_rate": 6.744857978748795e-06, | |
| "loss": 0.0153, | |
| "num_tokens": 102015527.0, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 2.1625282167042887, | |
| "grad_norm": 0.07103847764519645, | |
| "learning_rate": 6.719585105682012e-06, | |
| "loss": 0.0166, | |
| "num_tokens": 102435412.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.1715575620767495, | |
| "grad_norm": 0.07961645748931483, | |
| "learning_rate": 6.6942706467183916e-06, | |
| "loss": 0.0157, | |
| "num_tokens": 102851570.0, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 2.18058690744921, | |
| "grad_norm": 0.07140779969396704, | |
| "learning_rate": 6.668915465040403e-06, | |
| "loss": 0.0133, | |
| "num_tokens": 103270322.0, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.1896162528216703, | |
| "grad_norm": 0.06501016705122462, | |
| "learning_rate": 6.643520425219093e-06, | |
| "loss": 0.0126, | |
| "num_tokens": 103685477.0, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 2.198645598194131, | |
| "grad_norm": 0.07536879705533626, | |
| "learning_rate": 6.618086393184601e-06, | |
| "loss": 0.0123, | |
| "num_tokens": 104120983.0, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.2076749435665914, | |
| "grad_norm": 0.06681362355147086, | |
| "learning_rate": 6.592614236196646e-06, | |
| "loss": 0.0194, | |
| "num_tokens": 104536337.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.216704288939052, | |
| "grad_norm": 0.0693251653826736, | |
| "learning_rate": 6.567104822814942e-06, | |
| "loss": 0.013, | |
| "num_tokens": 104955016.0, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.2257336343115126, | |
| "grad_norm": 0.06358470526760071, | |
| "learning_rate": 6.541559022869589e-06, | |
| "loss": 0.0136, | |
| "num_tokens": 105388070.0, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 2.234762979683973, | |
| "grad_norm": 0.2564353264908074, | |
| "learning_rate": 6.515977707431411e-06, | |
| "loss": 0.0434, | |
| "num_tokens": 105818809.0, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.2437923250564333, | |
| "grad_norm": 0.07335440119765696, | |
| "learning_rate": 6.490361748782248e-06, | |
| "loss": 0.0136, | |
| "num_tokens": 106245854.0, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 2.2528216704288937, | |
| "grad_norm": 0.06532922181091284, | |
| "learning_rate": 6.464712020385223e-06, | |
| "loss": 0.013, | |
| "num_tokens": 106670473.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.2618510158013545, | |
| "grad_norm": 0.7266023360680285, | |
| "learning_rate": 6.439029396854955e-06, | |
| "loss": 0.048, | |
| "num_tokens": 107095545.0, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 2.270880361173815, | |
| "grad_norm": 0.07370448837002078, | |
| "learning_rate": 6.4133147539277295e-06, | |
| "loss": 0.0131, | |
| "num_tokens": 107520274.0, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.2799097065462752, | |
| "grad_norm": 0.08390458225564246, | |
| "learning_rate": 6.3875689684316435e-06, | |
| "loss": 0.0339, | |
| "num_tokens": 107953982.0, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.288939051918736, | |
| "grad_norm": 0.06777882375271017, | |
| "learning_rate": 6.361792918256705e-06, | |
| "loss": 0.0142, | |
| "num_tokens": 108388280.0, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.2979683972911964, | |
| "grad_norm": 0.07626936803178642, | |
| "learning_rate": 6.335987482324904e-06, | |
| "loss": 0.0142, | |
| "num_tokens": 108824002.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.3069977426636568, | |
| "grad_norm": 0.06449258340863874, | |
| "learning_rate": 6.310153540560229e-06, | |
| "loss": 0.0219, | |
| "num_tokens": 109244245.0, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.3160270880361176, | |
| "grad_norm": 0.06691990520571968, | |
| "learning_rate": 6.284291973858682e-06, | |
| "loss": 0.0113, | |
| "num_tokens": 109673712.0, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 2.325056433408578, | |
| "grad_norm": 0.06859871365700514, | |
| "learning_rate": 6.25840366405822e-06, | |
| "loss": 0.0134, | |
| "num_tokens": 110102425.0, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.3340857787810383, | |
| "grad_norm": 0.07117195067564289, | |
| "learning_rate": 6.232489493908706e-06, | |
| "loss": 0.0143, | |
| "num_tokens": 110528736.0, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 2.343115124153499, | |
| "grad_norm": 0.07013656339118007, | |
| "learning_rate": 6.2065503470417956e-06, | |
| "loss": 0.0122, | |
| "num_tokens": 110947222.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.3521444695259595, | |
| "grad_norm": 0.06746840266813196, | |
| "learning_rate": 6.180587107940809e-06, | |
| "loss": 0.0405, | |
| "num_tokens": 111396704.0, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 2.36117381489842, | |
| "grad_norm": 0.20371704046600575, | |
| "learning_rate": 6.154600661910577e-06, | |
| "loss": 0.0568, | |
| "num_tokens": 111831800.0, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.37020316027088, | |
| "grad_norm": 0.1429920710461871, | |
| "learning_rate": 6.128591895047243e-06, | |
| "loss": 0.0127, | |
| "num_tokens": 112263654.0, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 2.379232505643341, | |
| "grad_norm": 0.06796938770368956, | |
| "learning_rate": 6.102561694208064e-06, | |
| "loss": 0.0121, | |
| "num_tokens": 112691946.0, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.3882618510158014, | |
| "grad_norm": 0.0666953830758214, | |
| "learning_rate": 6.076510946981155e-06, | |
| "loss": 0.0235, | |
| "num_tokens": 113132551.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.3972911963882617, | |
| "grad_norm": 0.13625856338482667, | |
| "learning_rate": 6.05044054165523e-06, | |
| "loss": 0.0328, | |
| "num_tokens": 113569796.0, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.4063205417607225, | |
| "grad_norm": 0.15413072191676425, | |
| "learning_rate": 6.024351367189314e-06, | |
| "loss": 0.0129, | |
| "num_tokens": 114001801.0, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 2.415349887133183, | |
| "grad_norm": 0.06757492167260408, | |
| "learning_rate": 5.998244313182431e-06, | |
| "loss": 0.0125, | |
| "num_tokens": 114434376.0, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.4243792325056432, | |
| "grad_norm": 0.07107258708384646, | |
| "learning_rate": 5.972120269843263e-06, | |
| "loss": 0.0145, | |
| "num_tokens": 114881898.0, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 2.4334085778781036, | |
| "grad_norm": 0.061338195451882724, | |
| "learning_rate": 5.945980127959812e-06, | |
| "loss": 0.0121, | |
| "num_tokens": 115303494.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.4424379232505644, | |
| "grad_norm": 0.06762066726047494, | |
| "learning_rate": 5.919824778869002e-06, | |
| "loss": 0.0142, | |
| "num_tokens": 115722287.0, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 2.4514672686230248, | |
| "grad_norm": 0.0692200341605998, | |
| "learning_rate": 5.893655114426306e-06, | |
| "loss": 0.012, | |
| "num_tokens": 116161502.0, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.460496613995485, | |
| "grad_norm": 0.06754090447134138, | |
| "learning_rate": 5.867472026975326e-06, | |
| "loss": 0.0127, | |
| "num_tokens": 116581091.0, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 2.469525959367946, | |
| "grad_norm": 0.06225621229357672, | |
| "learning_rate": 5.841276409317366e-06, | |
| "loss": 0.0125, | |
| "num_tokens": 117012555.0, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.4785553047404063, | |
| "grad_norm": 0.06196719021324254, | |
| "learning_rate": 5.815069154680991e-06, | |
| "loss": 0.0122, | |
| "num_tokens": 117435175.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.4875846501128667, | |
| "grad_norm": 0.06697708798601215, | |
| "learning_rate": 5.788851156691569e-06, | |
| "loss": 0.0658, | |
| "num_tokens": 117873663.0, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.4966139954853275, | |
| "grad_norm": 0.34075070167121463, | |
| "learning_rate": 5.7626233093407955e-06, | |
| "loss": 0.0128, | |
| "num_tokens": 118286469.0, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 2.505643340857788, | |
| "grad_norm": 0.07629756246894763, | |
| "learning_rate": 5.7363865069562195e-06, | |
| "loss": 0.0115, | |
| "num_tokens": 118713374.0, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.514672686230248, | |
| "grad_norm": 0.07034796548205811, | |
| "learning_rate": 5.710141644170734e-06, | |
| "loss": 0.0139, | |
| "num_tokens": 119126272.0, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 2.523702031602709, | |
| "grad_norm": 0.08603521531576876, | |
| "learning_rate": 5.683889615892091e-06, | |
| "loss": 0.0212, | |
| "num_tokens": 119554963.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.5327313769751694, | |
| "grad_norm": 0.07088091524565285, | |
| "learning_rate": 5.65763131727236e-06, | |
| "loss": 0.0123, | |
| "num_tokens": 119977012.0, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 2.5417607223476297, | |
| "grad_norm": 0.06614360737927173, | |
| "learning_rate": 5.631367643677428e-06, | |
| "loss": 0.0128, | |
| "num_tokens": 120396254.0, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.55079006772009, | |
| "grad_norm": 0.07042295702919123, | |
| "learning_rate": 5.605099490656459e-06, | |
| "loss": 0.0121, | |
| "num_tokens": 120811188.0, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 2.559819413092551, | |
| "grad_norm": 0.06916246634023543, | |
| "learning_rate": 5.578827753911357e-06, | |
| "loss": 0.0129, | |
| "num_tokens": 121249730.0, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.5688487584650113, | |
| "grad_norm": 0.07209978464901166, | |
| "learning_rate": 5.5525533292662246e-06, | |
| "loss": 0.0136, | |
| "num_tokens": 121653968.0, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.5778781038374716, | |
| "grad_norm": 0.0787892316770162, | |
| "learning_rate": 5.52627711263682e-06, | |
| "loss": 0.0127, | |
| "num_tokens": 122090338.0, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.5869074492099324, | |
| "grad_norm": 0.07387994099802492, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.0126, | |
| "num_tokens": 122507656.0, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 2.595936794582393, | |
| "grad_norm": 0.06821405288648451, | |
| "learning_rate": 5.4737228873631835e-06, | |
| "loss": 0.012, | |
| "num_tokens": 122940825.0, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.604966139954853, | |
| "grad_norm": 0.06756484850678789, | |
| "learning_rate": 5.447446670733777e-06, | |
| "loss": 0.0127, | |
| "num_tokens": 123349895.0, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 2.6139954853273135, | |
| "grad_norm": 0.06308494712154893, | |
| "learning_rate": 5.421172246088645e-06, | |
| "loss": 0.0122, | |
| "num_tokens": 123771878.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.6230248306997743, | |
| "grad_norm": 0.08082656510725997, | |
| "learning_rate": 5.394900509343543e-06, | |
| "loss": 0.0129, | |
| "num_tokens": 124201272.0, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 2.6320541760722347, | |
| "grad_norm": 0.06299806239016838, | |
| "learning_rate": 5.368632356322574e-06, | |
| "loss": 0.0197, | |
| "num_tokens": 124649372.0, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.6410835214446955, | |
| "grad_norm": 0.09342033450420674, | |
| "learning_rate": 5.342368682727641e-06, | |
| "loss": 0.0122, | |
| "num_tokens": 125066840.0, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 2.650112866817156, | |
| "grad_norm": 0.07602506163920753, | |
| "learning_rate": 5.3161103841079105e-06, | |
| "loss": 0.0141, | |
| "num_tokens": 125494250.0, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.659142212189616, | |
| "grad_norm": 0.06866876545387215, | |
| "learning_rate": 5.2898583558292645e-06, | |
| "loss": 0.012, | |
| "num_tokens": 125918388.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.6681715575620766, | |
| "grad_norm": 0.0686873223442973, | |
| "learning_rate": 5.2636134930437836e-06, | |
| "loss": 0.013, | |
| "num_tokens": 126360809.0, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.6772009029345374, | |
| "grad_norm": 0.06418767945659551, | |
| "learning_rate": 5.237376690659206e-06, | |
| "loss": 0.0118, | |
| "num_tokens": 126789778.0, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 2.6862302483069977, | |
| "grad_norm": 0.06467304405936973, | |
| "learning_rate": 5.211148843308432e-06, | |
| "loss": 0.0188, | |
| "num_tokens": 127228131.0, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.695259593679458, | |
| "grad_norm": 0.327408568479494, | |
| "learning_rate": 5.1849308453190105e-06, | |
| "loss": 0.045, | |
| "num_tokens": 127662535.0, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 2.704288939051919, | |
| "grad_norm": 0.07065815550062918, | |
| "learning_rate": 5.158723590682636e-06, | |
| "loss": 0.0153, | |
| "num_tokens": 128099563.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.7133182844243793, | |
| "grad_norm": 0.0854618622284164, | |
| "learning_rate": 5.132527973024677e-06, | |
| "loss": 0.0217, | |
| "num_tokens": 128526843.0, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 2.7223476297968396, | |
| "grad_norm": 0.06637294397685695, | |
| "learning_rate": 5.106344885573695e-06, | |
| "loss": 0.0145, | |
| "num_tokens": 128943620.0, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.7313769751693, | |
| "grad_norm": 0.06892427094488447, | |
| "learning_rate": 5.0801752211309995e-06, | |
| "loss": 0.0144, | |
| "num_tokens": 129365819.0, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 2.740406320541761, | |
| "grad_norm": 0.07451856933575882, | |
| "learning_rate": 5.05401987204019e-06, | |
| "loss": 0.0127, | |
| "num_tokens": 129798803.0, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 2.749435665914221, | |
| "grad_norm": 0.08191501668420725, | |
| "learning_rate": 5.027879730156738e-06, | |
| "loss": 0.013, | |
| "num_tokens": 130218483.0, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.758465011286682, | |
| "grad_norm": 0.07021794114794579, | |
| "learning_rate": 5.001755686817573e-06, | |
| "loss": 0.0121, | |
| "num_tokens": 130649485.0, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 2.7674943566591423, | |
| "grad_norm": 0.08004562057562054, | |
| "learning_rate": 4.975648632810686e-06, | |
| "loss": 0.017, | |
| "num_tokens": 131078439.0, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 2.7765237020316027, | |
| "grad_norm": 0.0684239650400617, | |
| "learning_rate": 4.949559458344771e-06, | |
| "loss": 0.0128, | |
| "num_tokens": 131532361.0, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 2.785553047404063, | |
| "grad_norm": 0.0660413280467444, | |
| "learning_rate": 4.923489053018846e-06, | |
| "loss": 0.012, | |
| "num_tokens": 131952787.0, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 2.7945823927765234, | |
| "grad_norm": 0.06861103363424428, | |
| "learning_rate": 4.897438305791937e-06, | |
| "loss": 0.013, | |
| "num_tokens": 132386508.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.8036117381489842, | |
| "grad_norm": 0.07134277327598654, | |
| "learning_rate": 4.8714081049527565e-06, | |
| "loss": 0.0126, | |
| "num_tokens": 132820484.0, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 2.8126410835214446, | |
| "grad_norm": 0.06294849212931829, | |
| "learning_rate": 4.845399338089425e-06, | |
| "loss": 0.0118, | |
| "num_tokens": 133254008.0, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 2.8216704288939054, | |
| "grad_norm": 0.07024694450157741, | |
| "learning_rate": 4.819412892059192e-06, | |
| "loss": 0.0678, | |
| "num_tokens": 133686531.0, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 2.8306997742663658, | |
| "grad_norm": 0.2582020873633315, | |
| "learning_rate": 4.793449652958207e-06, | |
| "loss": 0.0131, | |
| "num_tokens": 134117853.0, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 2.839729119638826, | |
| "grad_norm": 0.07077002016575418, | |
| "learning_rate": 4.767510506091296e-06, | |
| "loss": 0.0138, | |
| "num_tokens": 134542656.0, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.8487584650112865, | |
| "grad_norm": 0.06199513656975843, | |
| "learning_rate": 4.741596335941782e-06, | |
| "loss": 0.012, | |
| "num_tokens": 134983147.0, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 2.8577878103837473, | |
| "grad_norm": 0.06740110933280159, | |
| "learning_rate": 4.715708026141321e-06, | |
| "loss": 0.0123, | |
| "num_tokens": 135415172.0, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 2.8668171557562077, | |
| "grad_norm": 0.06838469318720133, | |
| "learning_rate": 4.6898464594397715e-06, | |
| "loss": 0.0112, | |
| "num_tokens": 135853491.0, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 2.875846501128668, | |
| "grad_norm": 0.07024949444733218, | |
| "learning_rate": 4.664012517675098e-06, | |
| "loss": 0.0311, | |
| "num_tokens": 136297933.0, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 2.884875846501129, | |
| "grad_norm": 0.1250997981580887, | |
| "learning_rate": 4.638207081743295e-06, | |
| "loss": 0.0556, | |
| "num_tokens": 136746456.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.893905191873589, | |
| "grad_norm": 0.10424208944448676, | |
| "learning_rate": 4.612431031568359e-06, | |
| "loss": 0.0169, | |
| "num_tokens": 137185823.0, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.9029345372460496, | |
| "grad_norm": 0.08211764206486848, | |
| "learning_rate": 4.586685246072272e-06, | |
| "loss": 0.0143, | |
| "num_tokens": 137610632.0, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 2.91196388261851, | |
| "grad_norm": 0.0752587009855602, | |
| "learning_rate": 4.560970603145046e-06, | |
| "loss": 0.0124, | |
| "num_tokens": 138036129.0, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 2.9209932279909707, | |
| "grad_norm": 0.06185391843833257, | |
| "learning_rate": 4.535287979614777e-06, | |
| "loss": 0.0112, | |
| "num_tokens": 138460480.0, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 2.930022573363431, | |
| "grad_norm": 0.06723587071308086, | |
| "learning_rate": 4.5096382512177535e-06, | |
| "loss": 0.0115, | |
| "num_tokens": 138890848.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.939051918735892, | |
| "grad_norm": 0.06238988762166912, | |
| "learning_rate": 4.484022292568593e-06, | |
| "loss": 0.0115, | |
| "num_tokens": 139300121.0, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 2.9480812641083523, | |
| "grad_norm": 0.06938937112343414, | |
| "learning_rate": 4.458440977130413e-06, | |
| "loss": 0.0206, | |
| "num_tokens": 139730061.0, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 2.9571106094808126, | |
| "grad_norm": 0.062009976633795545, | |
| "learning_rate": 4.432895177185061e-06, | |
| "loss": 0.0119, | |
| "num_tokens": 140161307.0, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 2.966139954853273, | |
| "grad_norm": 0.06784066321825818, | |
| "learning_rate": 4.407385763803355e-06, | |
| "loss": 0.0119, | |
| "num_tokens": 140580963.0, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 2.975169300225734, | |
| "grad_norm": 0.06292128977616586, | |
| "learning_rate": 4.381913606815401e-06, | |
| "loss": 0.0127, | |
| "num_tokens": 140998977.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.984198645598194, | |
| "grad_norm": 0.07569293102716057, | |
| "learning_rate": 4.356479574780909e-06, | |
| "loss": 0.0126, | |
| "num_tokens": 141427751.0, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 2.9932279909706545, | |
| "grad_norm": 0.05892529849159098, | |
| "learning_rate": 4.331084534959598e-06, | |
| "loss": 0.0125, | |
| "num_tokens": 141871576.0, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.07768349752650722, | |
| "learning_rate": 4.305729353281608e-06, | |
| "loss": 0.0112, | |
| "num_tokens": 142185499.0, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.10246679186820984, | |
| "eval_num_tokens": 142185499.0, | |
| "eval_runtime": 176.189, | |
| "eval_samples_per_second": 14.218, | |
| "eval_steps_per_second": 1.782, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 3.0090293453724604, | |
| "grad_norm": 0.07152061537608126, | |
| "learning_rate": 4.28041489431799e-06, | |
| "loss": 0.0113, | |
| "num_tokens": 142609048.0, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.018058690744921, | |
| "grad_norm": 0.05547975846200711, | |
| "learning_rate": 4.255142021251206e-06, | |
| "loss": 0.0097, | |
| "num_tokens": 143034905.0, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 3.0270880361173815, | |
| "grad_norm": 0.06145285775610661, | |
| "learning_rate": 4.22991159584571e-06, | |
| "loss": 0.0099, | |
| "num_tokens": 143441637.0, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.036117381489842, | |
| "grad_norm": 0.058702561810255335, | |
| "learning_rate": 4.204724478418558e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 143864737.0, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 3.0451467268623027, | |
| "grad_norm": 0.06716554695100217, | |
| "learning_rate": 4.1795815278100746e-06, | |
| "loss": 0.0102, | |
| "num_tokens": 144287169.0, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.054176072234763, | |
| "grad_norm": 0.10156337114344059, | |
| "learning_rate": 4.154483601354566e-06, | |
| "loss": 0.0161, | |
| "num_tokens": 144729276.0, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 3.0632054176072234, | |
| "grad_norm": 0.06503277327720089, | |
| "learning_rate": 4.129431554851086e-06, | |
| "loss": 0.0105, | |
| "num_tokens": 145152649.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.072234762979684, | |
| "grad_norm": 0.06496822163807517, | |
| "learning_rate": 4.104426242534256e-06, | |
| "loss": 0.0102, | |
| "num_tokens": 145603813.0, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 3.0812641083521446, | |
| "grad_norm": 0.056389637609083165, | |
| "learning_rate": 4.079468517045136e-06, | |
| "loss": 0.0091, | |
| "num_tokens": 146042430.0, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.090293453724605, | |
| "grad_norm": 0.08215256814716321, | |
| "learning_rate": 4.054559229402157e-06, | |
| "loss": 0.0182, | |
| "num_tokens": 146486450.0, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 3.0993227990970653, | |
| "grad_norm": 0.0658823665069719, | |
| "learning_rate": 4.029699228972087e-06, | |
| "loss": 0.0099, | |
| "num_tokens": 146918914.0, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.108352144469526, | |
| "grad_norm": 0.06379382035081951, | |
| "learning_rate": 4.0048893634410865e-06, | |
| "loss": 0.0098, | |
| "num_tokens": 147357955.0, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 3.1173814898419865, | |
| "grad_norm": 0.08656012715805109, | |
| "learning_rate": 3.980130478785794e-06, | |
| "loss": 0.0092, | |
| "num_tokens": 147800926.0, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.126410835214447, | |
| "grad_norm": 0.06257455234456158, | |
| "learning_rate": 3.955423419244484e-06, | |
| "loss": 0.0092, | |
| "num_tokens": 148218943.0, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 3.1354401805869077, | |
| "grad_norm": 0.0627738180223448, | |
| "learning_rate": 3.930769027288273e-06, | |
| "loss": 0.0107, | |
| "num_tokens": 148653844.0, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.144469525959368, | |
| "grad_norm": 0.06243500821032468, | |
| "learning_rate": 3.9061681435924014e-06, | |
| "loss": 0.0092, | |
| "num_tokens": 149091233.0, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 3.1534988713318284, | |
| "grad_norm": 0.060114887497339085, | |
| "learning_rate": 3.881621607007565e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 149519351.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.1625282167042887, | |
| "grad_norm": 0.06684460684294827, | |
| "learning_rate": 3.857130254531303e-06, | |
| "loss": 0.0102, | |
| "num_tokens": 149963954.0, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 3.1715575620767495, | |
| "grad_norm": 0.0581834905948131, | |
| "learning_rate": 3.832694921279474e-06, | |
| "loss": 0.0117, | |
| "num_tokens": 150388491.0, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.18058690744921, | |
| "grad_norm": 0.07664972688212673, | |
| "learning_rate": 3.8083164404577654e-06, | |
| "loss": 0.047, | |
| "num_tokens": 150847232.0, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 3.1896162528216703, | |
| "grad_norm": 0.05840055533816688, | |
| "learning_rate": 3.7839956433332847e-06, | |
| "loss": 0.0297, | |
| "num_tokens": 151284392.0, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.198645598194131, | |
| "grad_norm": 0.09943187660436438, | |
| "learning_rate": 3.759733359206229e-06, | |
| "loss": 0.0095, | |
| "num_tokens": 151731845.0, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 3.2076749435665914, | |
| "grad_norm": 0.058677854163930766, | |
| "learning_rate": 3.735530415381584e-06, | |
| "loss": 0.0095, | |
| "num_tokens": 152170900.0, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.216704288939052, | |
| "grad_norm": 0.060264490178896846, | |
| "learning_rate": 3.7113876371409354e-06, | |
| "loss": 0.0088, | |
| "num_tokens": 152607162.0, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 3.2257336343115126, | |
| "grad_norm": 0.05927279793724808, | |
| "learning_rate": 3.687305847714311e-06, | |
| "loss": 0.009, | |
| "num_tokens": 153028418.0, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.234762979683973, | |
| "grad_norm": 0.06139721772997134, | |
| "learning_rate": 3.6632858682521233e-06, | |
| "loss": 0.0094, | |
| "num_tokens": 153425933.0, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 3.2437923250564333, | |
| "grad_norm": 0.0829022135578992, | |
| "learning_rate": 3.639328517797164e-06, | |
| "loss": 0.0473, | |
| "num_tokens": 153862042.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.2528216704288937, | |
| "grad_norm": 0.06461822963842173, | |
| "learning_rate": 3.6154346132566732e-06, | |
| "loss": 0.012, | |
| "num_tokens": 154292718.0, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 3.2618510158013545, | |
| "grad_norm": 0.06483450196299163, | |
| "learning_rate": 3.5916049693744883e-06, | |
| "loss": 0.0098, | |
| "num_tokens": 154735278.0, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.270880361173815, | |
| "grad_norm": 0.09379714259866676, | |
| "learning_rate": 3.5678403987032616e-06, | |
| "loss": 0.0124, | |
| "num_tokens": 155179493.0, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 3.2799097065462752, | |
| "grad_norm": 0.06565894474694481, | |
| "learning_rate": 3.544141711576754e-06, | |
| "loss": 0.0097, | |
| "num_tokens": 155601091.0, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.288939051918736, | |
| "grad_norm": 0.06116750216624111, | |
| "learning_rate": 3.5205097160821987e-06, | |
| "loss": 0.0094, | |
| "num_tokens": 156037078.0, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 3.2979683972911964, | |
| "grad_norm": 0.06928926670826958, | |
| "learning_rate": 3.4969452180327614e-06, | |
| "loss": 0.0101, | |
| "num_tokens": 156459746.0, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.3069977426636568, | |
| "grad_norm": 0.08409965398766228, | |
| "learning_rate": 3.4734490209400397e-06, | |
| "loss": 0.0274, | |
| "num_tokens": 156901429.0, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 3.3160270880361176, | |
| "grad_norm": 0.061886724574349766, | |
| "learning_rate": 3.450021925986687e-06, | |
| "loss": 0.0084, | |
| "num_tokens": 157331010.0, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.325056433408578, | |
| "grad_norm": 0.05625118622282191, | |
| "learning_rate": 3.4266647319990832e-06, | |
| "loss": 0.009, | |
| "num_tokens": 157772602.0, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 3.3340857787810383, | |
| "grad_norm": 0.059888736611577655, | |
| "learning_rate": 3.403378235420096e-06, | |
| "loss": 0.0092, | |
| "num_tokens": 158189457.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.343115124153499, | |
| "grad_norm": 0.055637198763357, | |
| "learning_rate": 3.380163230281928e-06, | |
| "loss": 0.0089, | |
| "num_tokens": 158626349.0, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 3.3521444695259595, | |
| "grad_norm": 0.058711357359014525, | |
| "learning_rate": 3.3570205081790285e-06, | |
| "loss": 0.01, | |
| "num_tokens": 159042993.0, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.36117381489842, | |
| "grad_norm": 0.15325418912567096, | |
| "learning_rate": 3.3339508582411245e-06, | |
| "loss": 0.03, | |
| "num_tokens": 159461779.0, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 3.37020316027088, | |
| "grad_norm": 0.06129832967810523, | |
| "learning_rate": 3.3109550671062907e-06, | |
| "loss": 0.0106, | |
| "num_tokens": 159889249.0, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.379232505643341, | |
| "grad_norm": 0.06713273236259697, | |
| "learning_rate": 3.288033918894137e-06, | |
| "loss": 0.0095, | |
| "num_tokens": 160318538.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 3.3882618510158014, | |
| "grad_norm": 0.057780993532694884, | |
| "learning_rate": 3.265188195179071e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 160741586.0, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.3972911963882617, | |
| "grad_norm": 0.06396483661772823, | |
| "learning_rate": 3.2424186749636455e-06, | |
| "loss": 0.0099, | |
| "num_tokens": 161157068.0, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 3.4063205417607225, | |
| "grad_norm": 0.11037209671799429, | |
| "learning_rate": 3.2197261346519905e-06, | |
| "loss": 0.0359, | |
| "num_tokens": 161591338.0, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.415349887133183, | |
| "grad_norm": 0.06141290629847806, | |
| "learning_rate": 3.1971113480233556e-06, | |
| "loss": 0.0266, | |
| "num_tokens": 162030602.0, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 3.4243792325056432, | |
| "grad_norm": 0.06327066709618506, | |
| "learning_rate": 3.1745750862057033e-06, | |
| "loss": 0.0096, | |
| "num_tokens": 162466006.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.4334085778781036, | |
| "grad_norm": 0.058909300821178116, | |
| "learning_rate": 3.152118117649433e-06, | |
| "loss": 0.0094, | |
| "num_tokens": 162897575.0, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 3.4424379232505644, | |
| "grad_norm": 0.05997809135270737, | |
| "learning_rate": 3.1297412081011686e-06, | |
| "loss": 0.0096, | |
| "num_tokens": 163303958.0, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.4514672686230248, | |
| "grad_norm": 0.06445221190776994, | |
| "learning_rate": 3.1074451205776505e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 163713904.0, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 3.460496613995485, | |
| "grad_norm": 0.05870409054047719, | |
| "learning_rate": 3.0852306153397194e-06, | |
| "loss": 0.0089, | |
| "num_tokens": 164137525.0, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.469525959367946, | |
| "grad_norm": 0.07569408829366625, | |
| "learning_rate": 3.063098449866384e-06, | |
| "loss": 0.009, | |
| "num_tokens": 164577195.0, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 3.4785553047404063, | |
| "grad_norm": 0.0588097227923788, | |
| "learning_rate": 3.0410493788290114e-06, | |
| "loss": 0.0086, | |
| "num_tokens": 165007790.0, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.4875846501128667, | |
| "grad_norm": 0.07652438346762197, | |
| "learning_rate": 3.019084154065568e-06, | |
| "loss": 0.0236, | |
| "num_tokens": 165448590.0, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 3.4966139954853275, | |
| "grad_norm": 0.06437216729466062, | |
| "learning_rate": 2.997203524555005e-06, | |
| "loss": 0.0094, | |
| "num_tokens": 165878673.0, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.505643340857788, | |
| "grad_norm": 0.06358680380643286, | |
| "learning_rate": 2.97540823639171e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 166305211.0, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 3.514672686230248, | |
| "grad_norm": 0.07750794447714442, | |
| "learning_rate": 2.953699032760067e-06, | |
| "loss": 0.0167, | |
| "num_tokens": 166738863.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.523702031602709, | |
| "grad_norm": 0.07534509702038769, | |
| "learning_rate": 2.932076653909115e-06, | |
| "loss": 0.009, | |
| "num_tokens": 167150386.0, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 3.5327313769751694, | |
| "grad_norm": 0.06216853830987747, | |
| "learning_rate": 2.910541837127305e-06, | |
| "loss": 0.009, | |
| "num_tokens": 167571071.0, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.5417607223476297, | |
| "grad_norm": 0.06388977277596672, | |
| "learning_rate": 2.889095316717366e-06, | |
| "loss": 0.0096, | |
| "num_tokens": 167985607.0, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 3.55079006772009, | |
| "grad_norm": 0.05749415988844605, | |
| "learning_rate": 2.8677378239712607e-06, | |
| "loss": 0.009, | |
| "num_tokens": 168420127.0, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.559819413092551, | |
| "grad_norm": 0.06339615887091614, | |
| "learning_rate": 2.846470087145249e-06, | |
| "loss": 0.0091, | |
| "num_tokens": 168853493.0, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 3.5688487584650113, | |
| "grad_norm": 0.06137274100197093, | |
| "learning_rate": 2.8252928314350626e-06, | |
| "loss": 0.0087, | |
| "num_tokens": 169267253.0, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.5778781038374716, | |
| "grad_norm": 0.059678184298786935, | |
| "learning_rate": 2.804206778951168e-06, | |
| "loss": 0.0086, | |
| "num_tokens": 169692290.0, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 3.5869074492099324, | |
| "grad_norm": 0.09399964545190312, | |
| "learning_rate": 2.7832126486941456e-06, | |
| "loss": 0.0387, | |
| "num_tokens": 170118921.0, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.595936794582393, | |
| "grad_norm": 0.05833904735584078, | |
| "learning_rate": 2.7623111565301863e-06, | |
| "loss": 0.0097, | |
| "num_tokens": 170541906.0, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 3.604966139954853, | |
| "grad_norm": 0.08684244774836425, | |
| "learning_rate": 2.7415030151666567e-06, | |
| "loss": 0.0435, | |
| "num_tokens": 170969534.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.6139954853273135, | |
| "grad_norm": 0.06469737843216128, | |
| "learning_rate": 2.720788934127819e-06, | |
| "loss": 0.0099, | |
| "num_tokens": 171393139.0, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 3.6230248306997743, | |
| "grad_norm": 0.07008366163987513, | |
| "learning_rate": 2.700169619730631e-06, | |
| "loss": 0.0098, | |
| "num_tokens": 171822275.0, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.6320541760722347, | |
| "grad_norm": 0.06074005907395048, | |
| "learning_rate": 2.6796457750606487e-06, | |
| "loss": 0.0097, | |
| "num_tokens": 172269333.0, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 3.6410835214446955, | |
| "grad_norm": 0.06078788588742423, | |
| "learning_rate": 2.659218099948079e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 172666432.0, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.650112866817156, | |
| "grad_norm": 0.06624525044121411, | |
| "learning_rate": 2.6388872909438875e-06, | |
| "loss": 0.0094, | |
| "num_tokens": 173085250.0, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 3.659142212189616, | |
| "grad_norm": 0.05869490549109586, | |
| "learning_rate": 2.618654041296068e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 173503425.0, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 3.6681715575620766, | |
| "grad_norm": 0.06550847475102616, | |
| "learning_rate": 2.5985190409259957e-06, | |
| "loss": 0.0094, | |
| "num_tokens": 173915204.0, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 3.6772009029345374, | |
| "grad_norm": 0.11285521045144888, | |
| "learning_rate": 2.5784829764049013e-06, | |
| "loss": 0.0109, | |
| "num_tokens": 174374964.0, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 3.6862302483069977, | |
| "grad_norm": 0.05975449002422044, | |
| "learning_rate": 2.558546530930466e-06, | |
| "loss": 0.009, | |
| "num_tokens": 174779332.0, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 3.695259593679458, | |
| "grad_norm": 0.08349376485094143, | |
| "learning_rate": 2.5387103843035126e-06, | |
| "loss": 0.0179, | |
| "num_tokens": 175200386.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.704288939051919, | |
| "grad_norm": 0.06215689171823831, | |
| "learning_rate": 2.5189752129048428e-06, | |
| "loss": 0.0088, | |
| "num_tokens": 175605971.0, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 3.7133182844243793, | |
| "grad_norm": 0.05902745197195644, | |
| "learning_rate": 2.49934168967216e-06, | |
| "loss": 0.0365, | |
| "num_tokens": 176029017.0, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 3.7223476297968396, | |
| "grad_norm": 0.13669531400358878, | |
| "learning_rate": 2.4798104840771294e-06, | |
| "loss": 0.0091, | |
| "num_tokens": 176454891.0, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 3.7313769751693, | |
| "grad_norm": 0.06004009215428138, | |
| "learning_rate": 2.46038226210255e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 176885789.0, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 3.740406320541761, | |
| "grad_norm": 0.07913705101953389, | |
| "learning_rate": 2.4410576862196435e-06, | |
| "loss": 0.0142, | |
| "num_tokens": 177301433.0, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 3.749435665914221, | |
| "grad_norm": 0.06212903945062577, | |
| "learning_rate": 2.4218374153654627e-06, | |
| "loss": 0.0095, | |
| "num_tokens": 177722579.0, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 3.758465011286682, | |
| "grad_norm": 0.06270055687189864, | |
| "learning_rate": 2.4027221049204347e-06, | |
| "loss": 0.0094, | |
| "num_tokens": 178144731.0, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 3.7674943566591423, | |
| "grad_norm": 0.0668935291153066, | |
| "learning_rate": 2.383712406685995e-06, | |
| "loss": 0.0101, | |
| "num_tokens": 178574868.0, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 3.7765237020316027, | |
| "grad_norm": 0.0640944822370456, | |
| "learning_rate": 2.364808968862378e-06, | |
| "loss": 0.0096, | |
| "num_tokens": 179005190.0, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 3.785553047404063, | |
| "grad_norm": 0.06369669742321124, | |
| "learning_rate": 2.346012436026508e-06, | |
| "loss": 0.0096, | |
| "num_tokens": 179421217.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.7945823927765234, | |
| "grad_norm": 0.06441713353622074, | |
| "learning_rate": 2.327323449110017e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 179838546.0, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 3.8036117381489842, | |
| "grad_norm": 0.08338546996857724, | |
| "learning_rate": 2.3087426453774002e-06, | |
| "loss": 0.0107, | |
| "num_tokens": 180250539.0, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 3.8126410835214446, | |
| "grad_norm": 0.06738272796438924, | |
| "learning_rate": 2.290270658404271e-06, | |
| "loss": 0.011, | |
| "num_tokens": 180672562.0, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 3.8216704288939054, | |
| "grad_norm": 0.0625991588242634, | |
| "learning_rate": 2.2719081180557757e-06, | |
| "loss": 0.0096, | |
| "num_tokens": 181091537.0, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 3.8306997742663658, | |
| "grad_norm": 0.0556196250975855, | |
| "learning_rate": 2.253655650465096e-06, | |
| "loss": 0.0089, | |
| "num_tokens": 181506969.0, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 3.839729119638826, | |
| "grad_norm": 0.06481115036912619, | |
| "learning_rate": 2.2355138780121166e-06, | |
| "loss": 0.0564, | |
| "num_tokens": 181969132.0, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 3.8487584650112865, | |
| "grad_norm": 0.08347930182037203, | |
| "learning_rate": 2.2174834193021934e-06, | |
| "loss": 0.0091, | |
| "num_tokens": 182397825.0, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 3.8577878103837473, | |
| "grad_norm": 0.059621287329162206, | |
| "learning_rate": 2.199564889145058e-06, | |
| "loss": 0.0172, | |
| "num_tokens": 182846790.0, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 3.8668171557562077, | |
| "grad_norm": 0.07303880000577928, | |
| "learning_rate": 2.181758898533866e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 183262540.0, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 3.875846501128668, | |
| "grad_norm": 0.0673061690918613, | |
| "learning_rate": 2.164066054624347e-06, | |
| "loss": 0.0095, | |
| "num_tokens": 183697485.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.884875846501129, | |
| "grad_norm": 0.0661976596717091, | |
| "learning_rate": 2.146486960714114e-06, | |
| "loss": 0.0089, | |
| "num_tokens": 184125479.0, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 3.893905191873589, | |
| "grad_norm": 0.06398778686313247, | |
| "learning_rate": 2.129022216222085e-06, | |
| "loss": 0.0088, | |
| "num_tokens": 184544539.0, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 3.9029345372460496, | |
| "grad_norm": 0.06215599177607287, | |
| "learning_rate": 2.111672416668048e-06, | |
| "loss": 0.0091, | |
| "num_tokens": 184949325.0, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 3.91196388261851, | |
| "grad_norm": 0.062320619384830064, | |
| "learning_rate": 2.0944381536523526e-06, | |
| "loss": 0.0091, | |
| "num_tokens": 185353902.0, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 3.9209932279909707, | |
| "grad_norm": 0.06084848690325647, | |
| "learning_rate": 2.077320014835738e-06, | |
| "loss": 0.0083, | |
| "num_tokens": 185775888.0, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 3.930022573363431, | |
| "grad_norm": 0.06353024998165271, | |
| "learning_rate": 2.0603185839192914e-06, | |
| "loss": 0.009, | |
| "num_tokens": 186203975.0, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 3.939051918735892, | |
| "grad_norm": 0.058439153731873986, | |
| "learning_rate": 2.043434440624551e-06, | |
| "loss": 0.0086, | |
| "num_tokens": 186639222.0, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 3.9480812641083523, | |
| "grad_norm": 0.05830094939756939, | |
| "learning_rate": 2.0266681606737335e-06, | |
| "loss": 0.0093, | |
| "num_tokens": 187080093.0, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 3.9571106094808126, | |
| "grad_norm": 0.059003884205823975, | |
| "learning_rate": 2.0100203157701066e-06, | |
| "loss": 0.0087, | |
| "num_tokens": 187510794.0, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 3.966139954853273, | |
| "grad_norm": 0.06298185573366215, | |
| "learning_rate": 1.993491473578491e-06, | |
| "loss": 0.01, | |
| "num_tokens": 187954041.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.975169300225734, | |
| "grad_norm": 0.059253636545730855, | |
| "learning_rate": 1.9770821977059026e-06, | |
| "loss": 0.0086, | |
| "num_tokens": 188383605.0, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 3.984198645598194, | |
| "grad_norm": 0.060746167690560224, | |
| "learning_rate": 1.9607930476823467e-06, | |
| "loss": 0.009, | |
| "num_tokens": 188821681.0, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 3.9932279909706545, | |
| "grad_norm": 0.059588516531004394, | |
| "learning_rate": 1.9446245789417194e-06, | |
| "loss": 0.0082, | |
| "num_tokens": 189274421.0, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.0817118222108563, | |
| "learning_rate": 1.928577342802885e-06, | |
| "loss": 0.0086, | |
| "num_tokens": 189586064.0, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.11091578751802444, | |
| "eval_num_tokens": 189586064.0, | |
| "eval_runtime": 55.4572, | |
| "eval_samples_per_second": 45.17, | |
| "eval_steps_per_second": 5.662, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 4.00902934537246, | |
| "grad_norm": 0.052009943856600654, | |
| "learning_rate": 1.9126518864508685e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 190012482.0, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 4.018058690744921, | |
| "grad_norm": 0.0558999480030762, | |
| "learning_rate": 1.8968487529181967e-06, | |
| "loss": 0.0077, | |
| "num_tokens": 190450990.0, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 4.027088036117381, | |
| "grad_norm": 0.052664936025694234, | |
| "learning_rate": 1.8811684810663915e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 190867207.0, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 4.036117381489842, | |
| "grad_norm": 0.052591105688649366, | |
| "learning_rate": 1.8656116055675816e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 191302488.0, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 4.045146726862303, | |
| "grad_norm": 0.05229870599777959, | |
| "learning_rate": 1.85017865688628e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 191733936.0, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 4.054176072234763, | |
| "grad_norm": 0.05419506215141233, | |
| "learning_rate": 1.8348701612612951e-06, | |
| "loss": 0.0071, | |
| "num_tokens": 192167735.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.063205417607223, | |
| "grad_norm": 0.05154741924536706, | |
| "learning_rate": 1.819686640687785e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 192575157.0, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 4.072234762979684, | |
| "grad_norm": 0.1279411133607819, | |
| "learning_rate": 1.8046286128994578e-06, | |
| "loss": 0.0185, | |
| "num_tokens": 193021330.0, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 4.081264108352144, | |
| "grad_norm": 0.05804375874747993, | |
| "learning_rate": 1.7896965913509213e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 193464296.0, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 4.090293453724605, | |
| "grad_norm": 0.05478585010295768, | |
| "learning_rate": 1.7748910852001684e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 193896598.0, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 4.099322799097066, | |
| "grad_norm": 0.09510833668893832, | |
| "learning_rate": 1.7602125992912239e-06, | |
| "loss": 0.0116, | |
| "num_tokens": 194322822.0, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 4.108352144469526, | |
| "grad_norm": 0.09148589236158025, | |
| "learning_rate": 1.7456616341369237e-06, | |
| "loss": 0.0266, | |
| "num_tokens": 194756207.0, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 4.1173814898419865, | |
| "grad_norm": 0.0936052964125868, | |
| "learning_rate": 1.7312386859018517e-06, | |
| "loss": 0.0306, | |
| "num_tokens": 195185054.0, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 4.126410835214447, | |
| "grad_norm": 0.0630492348104541, | |
| "learning_rate": 1.7169442463854208e-06, | |
| "loss": 0.0081, | |
| "num_tokens": 195599914.0, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 4.135440180586907, | |
| "grad_norm": 0.058377931570083946, | |
| "learning_rate": 1.7027788030050967e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 196011879.0, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 4.144469525959368, | |
| "grad_norm": 0.06609945281438827, | |
| "learning_rate": 1.6887428387797942e-06, | |
| "loss": 0.0087, | |
| "num_tokens": 196428514.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.153498871331829, | |
| "grad_norm": 0.058518152958306584, | |
| "learning_rate": 1.6748368323133868e-06, | |
| "loss": 0.0078, | |
| "num_tokens": 196847106.0, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 4.162528216704289, | |
| "grad_norm": 0.06538274508219379, | |
| "learning_rate": 1.6610612577784009e-06, | |
| "loss": 0.015, | |
| "num_tokens": 197290686.0, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 4.1715575620767495, | |
| "grad_norm": 0.05765426974466294, | |
| "learning_rate": 1.6474165848998439e-06, | |
| "loss": 0.0077, | |
| "num_tokens": 197727840.0, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 4.18058690744921, | |
| "grad_norm": 0.05678297815227037, | |
| "learning_rate": 1.633903278939185e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 198145792.0, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 4.18961625282167, | |
| "grad_norm": 0.05345688980658021, | |
| "learning_rate": 1.6205218006784934e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 198577289.0, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 4.198645598194131, | |
| "grad_norm": 0.05771798802728127, | |
| "learning_rate": 1.6072726064047212e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 199021897.0, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 4.207674943566591, | |
| "grad_norm": 0.05564057777959995, | |
| "learning_rate": 1.5941561478941563e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 199437131.0, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 4.216704288939052, | |
| "grad_norm": 0.05403994059952407, | |
| "learning_rate": 1.5811728723970019e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 199863916.0, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.225733634311513, | |
| "grad_norm": 0.05449636042777582, | |
| "learning_rate": 1.568323222622138e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 200298569.0, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 4.234762979683973, | |
| "grad_norm": 0.056568063508186425, | |
| "learning_rate": 1.5556076367220218e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 200746877.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.243792325056433, | |
| "grad_norm": 0.04916400724026921, | |
| "learning_rate": 1.543026548277746e-06, | |
| "loss": 0.0068, | |
| "num_tokens": 201152308.0, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 4.252821670428894, | |
| "grad_norm": 0.05590743077242364, | |
| "learning_rate": 1.5305803862842569e-06, | |
| "loss": 0.0076, | |
| "num_tokens": 201575609.0, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.261851015801354, | |
| "grad_norm": 0.06612932945772447, | |
| "learning_rate": 1.5182695751357245e-06, | |
| "loss": 0.0082, | |
| "num_tokens": 201998154.0, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 4.270880361173815, | |
| "grad_norm": 0.16852448868178485, | |
| "learning_rate": 1.5060945346110707e-06, | |
| "loss": 0.0314, | |
| "num_tokens": 202418768.0, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 4.279909706546276, | |
| "grad_norm": 0.12451963953682407, | |
| "learning_rate": 1.4940556798596585e-06, | |
| "loss": 0.0109, | |
| "num_tokens": 202847509.0, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 4.288939051918736, | |
| "grad_norm": 0.05636694817589496, | |
| "learning_rate": 1.4821534213871344e-06, | |
| "loss": 0.0077, | |
| "num_tokens": 203293633.0, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 4.297968397291196, | |
| "grad_norm": 0.05468619950050741, | |
| "learning_rate": 1.4703881650414304e-06, | |
| "loss": 0.0077, | |
| "num_tokens": 203726576.0, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 4.306997742663657, | |
| "grad_norm": 0.05296646383507029, | |
| "learning_rate": 1.4587603119989263e-06, | |
| "loss": 0.0069, | |
| "num_tokens": 204143709.0, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 4.316027088036117, | |
| "grad_norm": 0.058621647518717554, | |
| "learning_rate": 1.4472702587507655e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 204573732.0, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 4.3250564334085775, | |
| "grad_norm": 0.06658969833141296, | |
| "learning_rate": 1.435918397089347e-06, | |
| "loss": 0.0247, | |
| "num_tokens": 205005583.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.334085778781039, | |
| "grad_norm": 0.05509992027896923, | |
| "learning_rate": 1.4247051140949513e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 205437157.0, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 4.343115124153499, | |
| "grad_norm": 0.056995599613870046, | |
| "learning_rate": 1.4136307921225513e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 205850496.0, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 4.3521444695259595, | |
| "grad_norm": 0.05649954152945347, | |
| "learning_rate": 1.4026958087887723e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 206281600.0, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 4.36117381489842, | |
| "grad_norm": 0.08462415808186972, | |
| "learning_rate": 1.3919005369590132e-06, | |
| "loss": 0.0158, | |
| "num_tokens": 206692358.0, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 4.37020316027088, | |
| "grad_norm": 0.05861019810430756, | |
| "learning_rate": 1.381245344734739e-06, | |
| "loss": 0.0076, | |
| "num_tokens": 207116740.0, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 4.3792325056433405, | |
| "grad_norm": 0.05590320171665174, | |
| "learning_rate": 1.3707305954409194e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 207550858.0, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 4.388261851015802, | |
| "grad_norm": 0.1440931849337204, | |
| "learning_rate": 1.3603566476136488e-06, | |
| "loss": 0.0304, | |
| "num_tokens": 207978354.0, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 4.397291196388262, | |
| "grad_norm": 0.0575755327672531, | |
| "learning_rate": 1.3501238549879156e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 208410302.0, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.4063205417607225, | |
| "grad_norm": 0.05585899759462395, | |
| "learning_rate": 1.3400325664855437e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 208812870.0, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 4.415349887133183, | |
| "grad_norm": 0.05686170026770136, | |
| "learning_rate": 1.3300831262032925e-06, | |
| "loss": 0.0069, | |
| "num_tokens": 209254543.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.424379232505643, | |
| "grad_norm": 0.05803957066651316, | |
| "learning_rate": 1.3202758734011244e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 209669464.0, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 4.433408577878104, | |
| "grad_norm": 0.05687101353160547, | |
| "learning_rate": 1.3106111424906355e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 210092796.0, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 4.442437923250564, | |
| "grad_norm": 0.06650218713012311, | |
| "learning_rate": 1.3010892630236568e-06, | |
| "loss": 0.0545, | |
| "num_tokens": 210539090.0, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 4.451467268623025, | |
| "grad_norm": 0.059652537040254104, | |
| "learning_rate": 1.2917105596810112e-06, | |
| "loss": 0.0081, | |
| "num_tokens": 210962860.0, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 4.460496613995486, | |
| "grad_norm": 0.05691116296595083, | |
| "learning_rate": 1.2824753522614473e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 211389239.0, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 4.469525959367946, | |
| "grad_norm": 0.05271407561778469, | |
| "learning_rate": 1.273383955670732e-06, | |
| "loss": 0.0067, | |
| "num_tokens": 211848433.0, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 4.478555304740406, | |
| "grad_norm": 0.05811717147816959, | |
| "learning_rate": 1.2644366799109118e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 212286944.0, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 4.487584650112867, | |
| "grad_norm": 0.05983073556707705, | |
| "learning_rate": 1.2556338300697485e-06, | |
| "loss": 0.0076, | |
| "num_tokens": 212714521.0, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 4.496613995485327, | |
| "grad_norm": 0.05684584946279294, | |
| "learning_rate": 1.2469757063103061e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 213148368.0, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 4.505643340857787, | |
| "grad_norm": 0.05067708037080972, | |
| "learning_rate": 1.2384626038607255e-06, | |
| "loss": 0.0069, | |
| "num_tokens": 213576096.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.514672686230249, | |
| "grad_norm": 0.060108535921653126, | |
| "learning_rate": 1.2300948130041515e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 213992716.0, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 4.523702031602709, | |
| "grad_norm": 0.0552555230197704, | |
| "learning_rate": 1.2218726190688356e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 214418244.0, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 4.532731376975169, | |
| "grad_norm": 0.08684486719951047, | |
| "learning_rate": 1.2137963024184115e-06, | |
| "loss": 0.0151, | |
| "num_tokens": 214857117.0, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 4.54176072234763, | |
| "grad_norm": 0.055314422617567714, | |
| "learning_rate": 1.2058661384423267e-06, | |
| "loss": 0.0071, | |
| "num_tokens": 215275416.0, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 4.55079006772009, | |
| "grad_norm": 0.05865873272038585, | |
| "learning_rate": 1.1980823975464593e-06, | |
| "loss": 0.0079, | |
| "num_tokens": 215695066.0, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 4.5598194130925505, | |
| "grad_norm": 0.05975989236654091, | |
| "learning_rate": 1.1904453451438951e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 216118838.0, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 4.568848758465011, | |
| "grad_norm": 0.05892246505853784, | |
| "learning_rate": 1.1829552416458775e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 216542622.0, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 4.577878103837472, | |
| "grad_norm": 0.053580754460193454, | |
| "learning_rate": 1.1756123424529266e-06, | |
| "loss": 0.0067, | |
| "num_tokens": 216996879.0, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 4.586907449209932, | |
| "grad_norm": 0.05682211652915156, | |
| "learning_rate": 1.1684168979461336e-06, | |
| "loss": 0.007, | |
| "num_tokens": 217423398.0, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 4.595936794582393, | |
| "grad_norm": 0.0552314425067665, | |
| "learning_rate": 1.1613691534786196e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 217840389.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.604966139954853, | |
| "grad_norm": 0.11448108264486681, | |
| "learning_rate": 1.1544693493671712e-06, | |
| "loss": 0.0117, | |
| "num_tokens": 218270640.0, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 4.6139954853273135, | |
| "grad_norm": 0.16768787318826778, | |
| "learning_rate": 1.1477177208840482e-06, | |
| "loss": 0.0166, | |
| "num_tokens": 218713265.0, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 4.623024830699774, | |
| "grad_norm": 0.055188358952264455, | |
| "learning_rate": 1.1411144982489562e-06, | |
| "loss": 0.0069, | |
| "num_tokens": 219137609.0, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 4.632054176072235, | |
| "grad_norm": 0.17922405875168793, | |
| "learning_rate": 1.1346599066212008e-06, | |
| "loss": 0.0266, | |
| "num_tokens": 219578547.0, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 4.6410835214446955, | |
| "grad_norm": 0.07132874505614062, | |
| "learning_rate": 1.128354166092009e-06, | |
| "loss": 0.0136, | |
| "num_tokens": 220026371.0, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 4.650112866817156, | |
| "grad_norm": 0.06058796922864307, | |
| "learning_rate": 1.1221974916770236e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 220433942.0, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 4.659142212189616, | |
| "grad_norm": 0.05833795875496949, | |
| "learning_rate": 1.11619009330897e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 220867424.0, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 4.668171557562077, | |
| "grad_norm": 0.05870831578481961, | |
| "learning_rate": 1.1103321758305028e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 221290561.0, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 4.677200902934537, | |
| "grad_norm": 0.054615134252345106, | |
| "learning_rate": 1.104623938987216e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 221699415.0, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 4.686230248306998, | |
| "grad_norm": 0.05316946880846092, | |
| "learning_rate": 1.0990655774208339e-06, | |
| "loss": 0.0071, | |
| "num_tokens": 222150407.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.6952595936794586, | |
| "grad_norm": 0.07788416053210236, | |
| "learning_rate": 1.0936572806625755e-06, | |
| "loss": 0.0233, | |
| "num_tokens": 222572283.0, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 4.704288939051919, | |
| "grad_norm": 0.05880412240959133, | |
| "learning_rate": 1.0883992331266883e-06, | |
| "loss": 0.0076, | |
| "num_tokens": 222986703.0, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 4.713318284424379, | |
| "grad_norm": 0.05632076628688802, | |
| "learning_rate": 1.0832916141041655e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 223408969.0, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 4.72234762979684, | |
| "grad_norm": 0.058100528164223286, | |
| "learning_rate": 1.078334597756625e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 223835474.0, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 4.7313769751693, | |
| "grad_norm": 0.052063222481905536, | |
| "learning_rate": 1.0735283531103781e-06, | |
| "loss": 0.0066, | |
| "num_tokens": 224259986.0, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 4.74040632054176, | |
| "grad_norm": 0.055167878798155864, | |
| "learning_rate": 1.0688730440506611e-06, | |
| "loss": 0.008, | |
| "num_tokens": 224699745.0, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 4.749435665914222, | |
| "grad_norm": 0.05267890945523617, | |
| "learning_rate": 1.0643688293160503e-06, | |
| "loss": 0.0067, | |
| "num_tokens": 225145143.0, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 4.758465011286682, | |
| "grad_norm": 0.05312403103681725, | |
| "learning_rate": 1.0600158624930462e-06, | |
| "loss": 0.0071, | |
| "num_tokens": 225578145.0, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 4.767494356659142, | |
| "grad_norm": 0.06796402119311354, | |
| "learning_rate": 1.0558142920108394e-06, | |
| "loss": 0.0067, | |
| "num_tokens": 226003984.0, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 4.776523702031603, | |
| "grad_norm": 0.05356999529102921, | |
| "learning_rate": 1.0517642611362464e-06, | |
| "loss": 0.0068, | |
| "num_tokens": 226426088.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.785553047404063, | |
| "grad_norm": 0.059876875495227244, | |
| "learning_rate": 1.047865907968827e-06, | |
| "loss": 0.0075, | |
| "num_tokens": 226858806.0, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 4.794582392776523, | |
| "grad_norm": 0.0626550898148628, | |
| "learning_rate": 1.0441193654361755e-06, | |
| "loss": 0.0089, | |
| "num_tokens": 227291465.0, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 4.803611738148984, | |
| "grad_norm": 0.05255502973330509, | |
| "learning_rate": 1.0405247612893841e-06, | |
| "loss": 0.007, | |
| "num_tokens": 227721309.0, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 4.812641083521445, | |
| "grad_norm": 0.05438779778903025, | |
| "learning_rate": 1.037082218098692e-06, | |
| "loss": 0.0069, | |
| "num_tokens": 228142508.0, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 4.821670428893905, | |
| "grad_norm": 0.0593963271616768, | |
| "learning_rate": 1.0337918532493027e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 228560446.0, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 4.830699774266366, | |
| "grad_norm": 0.05474712434489935, | |
| "learning_rate": 1.0306537789373832e-06, | |
| "loss": 0.0065, | |
| "num_tokens": 228983150.0, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 4.839729119638826, | |
| "grad_norm": 0.058896302816248663, | |
| "learning_rate": 1.027668102166235e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 229412211.0, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 4.8487584650112865, | |
| "grad_norm": 0.05470480123832545, | |
| "learning_rate": 1.02483492474265e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 229854229.0, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 4.857787810383747, | |
| "grad_norm": 0.07890252865945219, | |
| "learning_rate": 1.0221543432734369e-06, | |
| "loss": 0.008, | |
| "num_tokens": 230272495.0, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 4.866817155756207, | |
| "grad_norm": 0.056142774309556866, | |
| "learning_rate": 1.0196264491621247e-06, | |
| "loss": 0.0071, | |
| "num_tokens": 230670995.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.8758465011286685, | |
| "grad_norm": 0.05995160644721174, | |
| "learning_rate": 1.0172513286058505e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 231097273.0, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 4.884875846501129, | |
| "grad_norm": 0.057626877086779195, | |
| "learning_rate": 1.015029062592418e-06, | |
| "loss": 0.0069, | |
| "num_tokens": 231529230.0, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 4.893905191873589, | |
| "grad_norm": 0.058716050916906605, | |
| "learning_rate": 1.012959726897535e-06, | |
| "loss": 0.0074, | |
| "num_tokens": 231936065.0, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 4.9029345372460496, | |
| "grad_norm": 0.05639352371278753, | |
| "learning_rate": 1.0110433920822306e-06, | |
| "loss": 0.0069, | |
| "num_tokens": 232357557.0, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 4.91196388261851, | |
| "grad_norm": 0.06136250042932289, | |
| "learning_rate": 1.009280123490451e-06, | |
| "loss": 0.0081, | |
| "num_tokens": 232776441.0, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 4.92099322799097, | |
| "grad_norm": 0.058152440154199, | |
| "learning_rate": 1.0076699812468264e-06, | |
| "loss": 0.0072, | |
| "num_tokens": 233194997.0, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 4.9300225733634315, | |
| "grad_norm": 0.054470578247469356, | |
| "learning_rate": 1.0062130202546278e-06, | |
| "loss": 0.007, | |
| "num_tokens": 233614735.0, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 4.939051918735892, | |
| "grad_norm": 0.05476103959928163, | |
| "learning_rate": 1.0049092901938875e-06, | |
| "loss": 0.0067, | |
| "num_tokens": 234054550.0, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 4.948081264108352, | |
| "grad_norm": 0.05951797614235394, | |
| "learning_rate": 1.0037588355197116e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 234481341.0, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 4.957110609480813, | |
| "grad_norm": 0.06242793615084626, | |
| "learning_rate": 1.00276169546076e-06, | |
| "loss": 0.0078, | |
| "num_tokens": 234907817.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.966139954853273, | |
| "grad_norm": 0.0636923634854018, | |
| "learning_rate": 1.0019179040179093e-06, | |
| "loss": 0.0125, | |
| "num_tokens": 235357272.0, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 4.975169300225733, | |
| "grad_norm": 0.055465368286989225, | |
| "learning_rate": 1.0012274899630954e-06, | |
| "loss": 0.0231, | |
| "num_tokens": 235805693.0, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 4.984198645598195, | |
| "grad_norm": 0.05982694743764175, | |
| "learning_rate": 1.0006904768383305e-06, | |
| "loss": 0.007, | |
| "num_tokens": 236221003.0, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 4.993227990970655, | |
| "grad_norm": 0.06260234129879096, | |
| "learning_rate": 1.0003068829549017e-06, | |
| "loss": 0.0073, | |
| "num_tokens": 236660222.0, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.06260234129879096, | |
| "learning_rate": 1.0000767213927445e-06, | |
| "loss": 0.0071, | |
| "num_tokens": 236984937.0, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.12006451934576035, | |
| "eval_num_tokens": 236984937.0, | |
| "eval_runtime": 55.4823, | |
| "eval_samples_per_second": 45.15, | |
| "eval_steps_per_second": 5.659, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 555, | |
| "total_flos": 7.804790123203133e+17, | |
| "train_loss": 0.04724993688717216, | |
| "train_runtime": 7963.8173, | |
| "train_samples_per_second": 8.896, | |
| "train_steps_per_second": 0.07 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 555, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "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": 7.804790123203133e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |