| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 200, | |
| "global_step": 134, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014925373134328358, | |
| "grad_norm": 2.4583136454633454, | |
| "learning_rate": 9.99862592554908e-06, | |
| "loss": 0.1709, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.029850746268656716, | |
| "grad_norm": 1.2601474874951664, | |
| "learning_rate": 9.994504457428557e-06, | |
| "loss": 0.1095, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.04477611940298507, | |
| "grad_norm": 1.8921245514823541, | |
| "learning_rate": 9.987637860920053e-06, | |
| "loss": 0.1123, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.05970149253731343, | |
| "grad_norm": 1.0381477254848812, | |
| "learning_rate": 9.978029910109491e-06, | |
| "loss": 0.0897, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.07462686567164178, | |
| "grad_norm": 0.8231161139163885, | |
| "learning_rate": 9.965685885812773e-06, | |
| "loss": 0.0804, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08955223880597014, | |
| "grad_norm": 3.3306861693595704, | |
| "learning_rate": 9.950612572673255e-06, | |
| "loss": 0.1197, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.1044776119402985, | |
| "grad_norm": 0.8873942793156309, | |
| "learning_rate": 9.932818255432733e-06, | |
| "loss": 0.1052, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.11940298507462686, | |
| "grad_norm": 0.7726110244054883, | |
| "learning_rate": 9.91231271437788e-06, | |
| "loss": 0.09, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.13432835820895522, | |
| "grad_norm": 0.7713735038013977, | |
| "learning_rate": 9.889107219964726e-06, | |
| "loss": 0.0847, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.14925373134328357, | |
| "grad_norm": 0.8179878874298719, | |
| "learning_rate": 9.863214526624065e-06, | |
| "loss": 0.0899, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16417910447761194, | |
| "grad_norm": 0.793854117532711, | |
| "learning_rate": 9.834648865751254e-06, | |
| "loss": 0.0885, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1791044776119403, | |
| "grad_norm": 0.6649283964833135, | |
| "learning_rate": 9.803425937884202e-06, | |
| "loss": 0.078, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.19402985074626866, | |
| "grad_norm": 0.8490726920421898, | |
| "learning_rate": 9.769562904073896e-06, | |
| "loss": 0.0878, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.208955223880597, | |
| "grad_norm": 0.7411456396869534, | |
| "learning_rate": 9.733078376452172e-06, | |
| "loss": 0.0881, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.22388059701492538, | |
| "grad_norm": 0.7448122603540034, | |
| "learning_rate": 9.693992408001934e-06, | |
| "loss": 0.091, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.23880597014925373, | |
| "grad_norm": 0.672661340651816, | |
| "learning_rate": 9.652326481535434e-06, | |
| "loss": 0.0847, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2537313432835821, | |
| "grad_norm": 0.6295351920280576, | |
| "learning_rate": 9.608103497886687e-06, | |
| "loss": 0.0751, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.26865671641791045, | |
| "grad_norm": 0.5987419993650692, | |
| "learning_rate": 9.561347763324484e-06, | |
| "loss": 0.0757, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2835820895522388, | |
| "grad_norm": 0.6862099670787631, | |
| "learning_rate": 9.512084976192944e-06, | |
| "loss": 0.0832, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.29850746268656714, | |
| "grad_norm": 0.6474317718234073, | |
| "learning_rate": 9.460342212786933e-06, | |
| "loss": 0.0785, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.31343283582089554, | |
| "grad_norm": 0.6520670338068375, | |
| "learning_rate": 9.406147912470142e-06, | |
| "loss": 0.0832, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.3283582089552239, | |
| "grad_norm": 0.6875547799836488, | |
| "learning_rate": 9.349531862043952e-06, | |
| "loss": 0.0909, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.34328358208955223, | |
| "grad_norm": 0.6406155904616685, | |
| "learning_rate": 9.290525179375722e-06, | |
| "loss": 0.0817, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3582089552238806, | |
| "grad_norm": 0.6570196434589318, | |
| "learning_rate": 9.229160296295488e-06, | |
| "loss": 0.0847, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.373134328358209, | |
| "grad_norm": 0.5570171509580777, | |
| "learning_rate": 9.165470940770458e-06, | |
| "loss": 0.0739, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3880597014925373, | |
| "grad_norm": 0.620261743808898, | |
| "learning_rate": 9.099492118367123e-06, | |
| "loss": 0.0861, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.40298507462686567, | |
| "grad_norm": 0.6886602468358877, | |
| "learning_rate": 9.03126009301115e-06, | |
| "loss": 0.0943, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.417910447761194, | |
| "grad_norm": 0.7131708436753863, | |
| "learning_rate": 8.960812367055646e-06, | |
| "loss": 0.0892, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.43283582089552236, | |
| "grad_norm": 0.6391387675910597, | |
| "learning_rate": 8.888187660668762e-06, | |
| "loss": 0.0832, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.44776119402985076, | |
| "grad_norm": 0.6940047679170313, | |
| "learning_rate": 8.81342589055191e-06, | |
| "loss": 0.0906, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.4626865671641791, | |
| "grad_norm": 0.6503008059328946, | |
| "learning_rate": 8.736568148000386e-06, | |
| "loss": 0.0968, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.47761194029850745, | |
| "grad_norm": 0.6090207680984978, | |
| "learning_rate": 8.657656676318346e-06, | |
| "loss": 0.0847, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4925373134328358, | |
| "grad_norm": 0.6434605637551735, | |
| "learning_rate": 8.576734847600639e-06, | |
| "loss": 0.0853, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.5074626865671642, | |
| "grad_norm": 0.6377279956113436, | |
| "learning_rate": 8.49384713889421e-06, | |
| "loss": 0.0899, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.5223880597014925, | |
| "grad_norm": 0.644362530558179, | |
| "learning_rate": 8.40903910775219e-06, | |
| "loss": 0.0827, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.5373134328358209, | |
| "grad_norm": 0.642985561685258, | |
| "learning_rate": 8.32235736719411e-06, | |
| "loss": 0.0888, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.5522388059701493, | |
| "grad_norm": 0.6790092652999941, | |
| "learning_rate": 8.233849560085994e-06, | |
| "loss": 0.0846, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.5671641791044776, | |
| "grad_norm": 0.7854755458868027, | |
| "learning_rate": 8.143564332954426e-06, | |
| "loss": 0.1031, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.582089552238806, | |
| "grad_norm": 0.6030440972462469, | |
| "learning_rate": 8.051551309248961e-06, | |
| "loss": 0.0849, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 0.6418639082548031, | |
| "learning_rate": 7.957861062067614e-06, | |
| "loss": 0.0852, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.6119402985074627, | |
| "grad_norm": 0.6566500879981874, | |
| "learning_rate": 7.86254508636036e-06, | |
| "loss": 0.0887, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.6268656716417911, | |
| "grad_norm": 0.6290631945385817, | |
| "learning_rate": 7.765655770625997e-06, | |
| "loss": 0.0855, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.6417910447761194, | |
| "grad_norm": 0.6140886580472066, | |
| "learning_rate": 7.667246368117852e-06, | |
| "loss": 0.0818, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.6567164179104478, | |
| "grad_norm": 0.6174619223952549, | |
| "learning_rate": 7.56737096757421e-06, | |
| "loss": 0.0856, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.6716417910447762, | |
| "grad_norm": 0.6195862366153396, | |
| "learning_rate": 7.466084463489537e-06, | |
| "loss": 0.0801, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.6865671641791045, | |
| "grad_norm": 0.6331901290551701, | |
| "learning_rate": 7.363442525942827e-06, | |
| "loss": 0.0844, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.7014925373134329, | |
| "grad_norm": 0.7083908752048309, | |
| "learning_rate": 7.25950156999967e-06, | |
| "loss": 0.0894, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.7164179104477612, | |
| "grad_norm": 0.6388671244490378, | |
| "learning_rate": 7.1543187247048525e-06, | |
| "loss": 0.0906, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.7313432835820896, | |
| "grad_norm": 0.5777376817133963, | |
| "learning_rate": 7.047951801682533e-06, | |
| "loss": 0.0775, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.746268656716418, | |
| "grad_norm": 0.5907825361104937, | |
| "learning_rate": 6.9404592633612486e-06, | |
| "loss": 0.08, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.7611940298507462, | |
| "grad_norm": 0.6197333346964937, | |
| "learning_rate": 6.831900190841232e-06, | |
| "loss": 0.0852, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.7761194029850746, | |
| "grad_norm": 0.580913389747538, | |
| "learning_rate": 6.722334251421665e-06, | |
| "loss": 0.0786, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.7910447761194029, | |
| "grad_norm": 0.6769846397314266, | |
| "learning_rate": 6.611821665805769e-06, | |
| "loss": 0.0859, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.8059701492537313, | |
| "grad_norm": 0.6493382927830379, | |
| "learning_rate": 6.500423175001705e-06, | |
| "loss": 0.0918, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.8208955223880597, | |
| "grad_norm": 0.6002157339319569, | |
| "learning_rate": 6.388200006937503e-06, | |
| "loss": 0.089, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.835820895522388, | |
| "grad_norm": 0.5447504002720616, | |
| "learning_rate": 6.275213842808383e-06, | |
| "loss": 0.0731, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.8507462686567164, | |
| "grad_norm": 0.5621404387468115, | |
| "learning_rate": 6.161526783174917e-06, | |
| "loss": 0.0762, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.8656716417910447, | |
| "grad_norm": 0.6279143619008688, | |
| "learning_rate": 6.047201313830724e-06, | |
| "loss": 0.0921, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.8805970149253731, | |
| "grad_norm": 0.5567853485138207, | |
| "learning_rate": 5.932300271458406e-06, | |
| "loss": 0.0692, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.8955223880597015, | |
| "grad_norm": 0.5831390070498211, | |
| "learning_rate": 5.816886809092651e-06, | |
| "loss": 0.0777, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.9104477611940298, | |
| "grad_norm": 0.575437949186205, | |
| "learning_rate": 5.701024361409431e-06, | |
| "loss": 0.0803, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.9253731343283582, | |
| "grad_norm": 0.6339642524977537, | |
| "learning_rate": 5.584776609860414e-06, | |
| "loss": 0.0893, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.9402985074626866, | |
| "grad_norm": 0.5924535700885266, | |
| "learning_rate": 5.468207447671755e-06, | |
| "loss": 0.0844, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.9552238805970149, | |
| "grad_norm": 0.6155460855755083, | |
| "learning_rate": 5.351380944726465e-06, | |
| "loss": 0.0836, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.9701492537313433, | |
| "grad_norm": 0.640329747223743, | |
| "learning_rate": 5.234361312349701e-06, | |
| "loss": 0.0951, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.9850746268656716, | |
| "grad_norm": 0.5300911342515855, | |
| "learning_rate": 5.117212868016303e-06, | |
| "loss": 0.0655, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.4687094870612933, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0424, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0149253731343284, | |
| "grad_norm": 0.4076938808158179, | |
| "learning_rate": 4.882787131983698e-06, | |
| "loss": 0.0319, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.0298507462686568, | |
| "grad_norm": 0.44699344595961127, | |
| "learning_rate": 4.765638687650299e-06, | |
| "loss": 0.0397, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.044776119402985, | |
| "grad_norm": 0.44767644296408526, | |
| "learning_rate": 4.6486190552735375e-06, | |
| "loss": 0.0335, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.0597014925373134, | |
| "grad_norm": 0.37505346262796413, | |
| "learning_rate": 4.531792552328247e-06, | |
| "loss": 0.0285, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.0746268656716418, | |
| "grad_norm": 0.4657077829483448, | |
| "learning_rate": 4.415223390139588e-06, | |
| "loss": 0.0326, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.0895522388059702, | |
| "grad_norm": 0.34306186183329557, | |
| "learning_rate": 4.2989756385905715e-06, | |
| "loss": 0.0244, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.1044776119402986, | |
| "grad_norm": 0.39032497882907513, | |
| "learning_rate": 4.183113190907349e-06, | |
| "loss": 0.027, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.1194029850746268, | |
| "grad_norm": 0.4702211175747881, | |
| "learning_rate": 4.067699728541595e-06, | |
| "loss": 0.0316, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.1343283582089552, | |
| "grad_norm": 0.4455655473637308, | |
| "learning_rate": 3.952798686169279e-06, | |
| "loss": 0.0303, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.1492537313432836, | |
| "grad_norm": 0.4371493860049261, | |
| "learning_rate": 3.838473216825085e-06, | |
| "loss": 0.0282, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.164179104477612, | |
| "grad_norm": 0.4289807711491057, | |
| "learning_rate": 3.7247861571916183e-06, | |
| "loss": 0.0272, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.1791044776119404, | |
| "grad_norm": 0.5368576489602601, | |
| "learning_rate": 3.611799993062497e-06, | |
| "loss": 0.0351, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.1940298507462686, | |
| "grad_norm": 0.6142429871519881, | |
| "learning_rate": 3.4995768249982975e-06, | |
| "loss": 0.0377, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.208955223880597, | |
| "grad_norm": 0.4437616080442814, | |
| "learning_rate": 3.388178334194232e-06, | |
| "loss": 0.0254, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.2238805970149254, | |
| "grad_norm": 0.425005368183552, | |
| "learning_rate": 3.2776657485783357e-06, | |
| "loss": 0.0217, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.2388059701492538, | |
| "grad_norm": 0.45656133078022915, | |
| "learning_rate": 3.168099809158769e-06, | |
| "loss": 0.0257, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.2537313432835822, | |
| "grad_norm": 0.543019359956107, | |
| "learning_rate": 3.059540736638751e-06, | |
| "loss": 0.0291, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.2686567164179103, | |
| "grad_norm": 0.48489281581823257, | |
| "learning_rate": 2.9520481983174675e-06, | |
| "loss": 0.0249, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.2835820895522387, | |
| "grad_norm": 0.5065165276548232, | |
| "learning_rate": 2.8456812752951483e-06, | |
| "loss": 0.0226, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.2985074626865671, | |
| "grad_norm": 0.5835532502940851, | |
| "learning_rate": 2.740498430000332e-06, | |
| "loss": 0.0262, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.3134328358208955, | |
| "grad_norm": 0.41781499689334245, | |
| "learning_rate": 2.636557474057173e-06, | |
| "loss": 0.0171, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.328358208955224, | |
| "grad_norm": 0.60361339503175, | |
| "learning_rate": 2.533915536510464e-06, | |
| "loss": 0.0249, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.3432835820895521, | |
| "grad_norm": 0.47456828088394937, | |
| "learning_rate": 2.4326290324257896e-06, | |
| "loss": 0.0246, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.3582089552238805, | |
| "grad_norm": 0.4465597229891302, | |
| "learning_rate": 2.3327536318821496e-06, | |
| "loss": 0.0183, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.373134328358209, | |
| "grad_norm": 0.49082928585159896, | |
| "learning_rate": 2.234344229374003e-06, | |
| "loss": 0.0272, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.3880597014925373, | |
| "grad_norm": 0.4703289245897022, | |
| "learning_rate": 2.1374549136396417e-06, | |
| "loss": 0.0259, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.4029850746268657, | |
| "grad_norm": 0.4919381501347153, | |
| "learning_rate": 2.042138937932388e-06, | |
| "loss": 0.0244, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.417910447761194, | |
| "grad_norm": 0.48234747726237315, | |
| "learning_rate": 1.9484486907510405e-06, | |
| "loss": 0.0259, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.4328358208955223, | |
| "grad_norm": 0.46679191916730023, | |
| "learning_rate": 1.856435667045577e-06, | |
| "loss": 0.0248, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.4477611940298507, | |
| "grad_norm": 0.4915526917266499, | |
| "learning_rate": 1.7661504399140066e-06, | |
| "loss": 0.0253, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.462686567164179, | |
| "grad_norm": 0.4959939539468189, | |
| "learning_rate": 1.677642632805892e-06, | |
| "loss": 0.029, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.4776119402985075, | |
| "grad_norm": 0.47533058613210927, | |
| "learning_rate": 1.5909608922478108e-06, | |
| "loss": 0.0222, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.4925373134328357, | |
| "grad_norm": 0.41184618371805026, | |
| "learning_rate": 1.5061528611057917e-06, | |
| "loss": 0.0186, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5074626865671643, | |
| "grad_norm": 0.5029260437214431, | |
| "learning_rate": 1.4232651523993635e-06, | |
| "loss": 0.0246, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.5223880597014925, | |
| "grad_norm": 0.4856272674409962, | |
| "learning_rate": 1.3423433236816563e-06, | |
| "loss": 0.0262, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.537313432835821, | |
| "grad_norm": 0.5515452132397586, | |
| "learning_rate": 1.2634318519996148e-06, | |
| "loss": 0.0265, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.5522388059701493, | |
| "grad_norm": 0.48385801987338495, | |
| "learning_rate": 1.186574109448091e-06, | |
| "loss": 0.0253, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.5671641791044775, | |
| "grad_norm": 0.4977052401413162, | |
| "learning_rate": 1.1118123393312397e-06, | |
| "loss": 0.0317, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.582089552238806, | |
| "grad_norm": 0.5013927727792972, | |
| "learning_rate": 1.0391876329443534e-06, | |
| "loss": 0.0248, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.5970149253731343, | |
| "grad_norm": 0.5012798267287072, | |
| "learning_rate": 9.687399069888515e-07, | |
| "loss": 0.0253, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.6119402985074627, | |
| "grad_norm": 0.4798753241925998, | |
| "learning_rate": 9.005078816328772e-07, | |
| "loss": 0.0208, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.626865671641791, | |
| "grad_norm": 0.4892720130785747, | |
| "learning_rate": 8.345290592295429e-07, | |
| "loss": 0.0256, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.6417910447761193, | |
| "grad_norm": 0.44880610381720976, | |
| "learning_rate": 7.708397037045129e-07, | |
| "loss": 0.024, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.6567164179104479, | |
| "grad_norm": 0.5198878749228643, | |
| "learning_rate": 7.094748206242797e-07, | |
| "loss": 0.0282, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.671641791044776, | |
| "grad_norm": 0.4259309199030039, | |
| "learning_rate": 6.50468137956049e-07, | |
| "loss": 0.0203, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.6865671641791045, | |
| "grad_norm": 0.44723898312223315, | |
| "learning_rate": 5.938520875298587e-07, | |
| "loss": 0.0262, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.7014925373134329, | |
| "grad_norm": 0.4649229524932995, | |
| "learning_rate": 5.396577872130676e-07, | |
| "loss": 0.0234, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.716417910447761, | |
| "grad_norm": 0.46689935485099987, | |
| "learning_rate": 4.879150238070585e-07, | |
| "loss": 0.0235, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.7313432835820897, | |
| "grad_norm": 0.4250501253858005, | |
| "learning_rate": 4.386522366755169e-07, | |
| "loss": 0.021, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.7462686567164178, | |
| "grad_norm": 0.4965625077414352, | |
| "learning_rate": 3.918965021133131e-07, | |
| "loss": 0.0285, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.7611940298507462, | |
| "grad_norm": 0.5192751153508381, | |
| "learning_rate": 3.4767351846456744e-07, | |
| "loss": 0.0294, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.7761194029850746, | |
| "grad_norm": 0.4555879578505303, | |
| "learning_rate": 3.0600759199806815e-07, | |
| "loss": 0.026, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.7910447761194028, | |
| "grad_norm": 0.4702463723473376, | |
| "learning_rate": 2.669216235478295e-07, | |
| "loss": 0.0268, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.8059701492537314, | |
| "grad_norm": 0.5014638902986028, | |
| "learning_rate": 2.3043709592610486e-07, | |
| "loss": 0.0282, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.8208955223880596, | |
| "grad_norm": 0.5015871929717038, | |
| "learning_rate": 1.9657406211579966e-07, | |
| "loss": 0.0276, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.835820895522388, | |
| "grad_norm": 0.46199043552912816, | |
| "learning_rate": 1.6535113424874683e-07, | |
| "loss": 0.0253, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.8507462686567164, | |
| "grad_norm": 0.472386855814729, | |
| "learning_rate": 1.3678547337593494e-07, | |
| "loss": 0.0231, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.8656716417910446, | |
| "grad_norm": 0.44628541420234796, | |
| "learning_rate": 1.1089278003527438e-07, | |
| "loss": 0.0229, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.8805970149253732, | |
| "grad_norm": 0.4945658271403232, | |
| "learning_rate": 8.768728562211948e-08, | |
| "loss": 0.0259, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.8955223880597014, | |
| "grad_norm": 0.4631666908930044, | |
| "learning_rate": 6.718174456726789e-08, | |
| "loss": 0.0254, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.9104477611940298, | |
| "grad_norm": 0.47821378359248506, | |
| "learning_rate": 4.9387427326745287e-08, | |
| "loss": 0.0257, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.9253731343283582, | |
| "grad_norm": 0.4463060792083917, | |
| "learning_rate": 3.431411418722941e-08, | |
| "loss": 0.023, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.9402985074626866, | |
| "grad_norm": 0.4458879106581698, | |
| "learning_rate": 2.1970089890509527e-08, | |
| "loss": 0.0237, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.955223880597015, | |
| "grad_norm": 0.39654579447619215, | |
| "learning_rate": 1.2362139079949431e-08, | |
| "loss": 0.0193, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.9701492537313432, | |
| "grad_norm": 0.4384754664189478, | |
| "learning_rate": 5.495542571443135e-09, | |
| "loss": 0.0227, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.9850746268656716, | |
| "grad_norm": 0.40331068244695245, | |
| "learning_rate": 1.3740744509205263e-09, | |
| "loss": 0.0209, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.3037784601924941, | |
| "learning_rate": 0.0, | |
| "loss": 0.013, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "step": 134, | |
| "total_flos": 8781261963264.0, | |
| "train_loss": 0.05639281026574213, | |
| "train_runtime": 528.9145, | |
| "train_samples_per_second": 2.0, | |
| "train_steps_per_second": 0.253 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 134, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 2000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8781261963264.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |