| { | |
| "best_metric": 0.6208789944648743, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.07923145488759037, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00039615727443795186, | |
| "grad_norm": 14.697911262512207, | |
| "learning_rate": 5e-06, | |
| "loss": 2.5987, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00039615727443795186, | |
| "eval_loss": 1.2088950872421265, | |
| "eval_runtime": 341.1554, | |
| "eval_samples_per_second": 12.464, | |
| "eval_steps_per_second": 6.232, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0007923145488759037, | |
| "grad_norm": 16.757488250732422, | |
| "learning_rate": 1e-05, | |
| "loss": 3.394, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0011884718233138556, | |
| "grad_norm": 19.087627410888672, | |
| "learning_rate": 1.5e-05, | |
| "loss": 3.0496, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0015846290977518075, | |
| "grad_norm": 18.1752986907959, | |
| "learning_rate": 2e-05, | |
| "loss": 3.2584, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0019807863721897595, | |
| "grad_norm": 14.575785636901855, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.986, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0023769436466277113, | |
| "grad_norm": 13.09287166595459, | |
| "learning_rate": 3e-05, | |
| "loss": 2.952, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.002773100921065663, | |
| "grad_norm": 11.316363334655762, | |
| "learning_rate": 3.5e-05, | |
| "loss": 2.6524, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.003169258195503615, | |
| "grad_norm": 11.73412036895752, | |
| "learning_rate": 4e-05, | |
| "loss": 2.669, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0035654154699415667, | |
| "grad_norm": 12.15320110321045, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.9467, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.003961572744379519, | |
| "grad_norm": 10.565752983093262, | |
| "learning_rate": 5e-05, | |
| "loss": 2.5971, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004357730018817471, | |
| "grad_norm": 10.899881362915039, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.5928, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.004753887293255423, | |
| "grad_norm": 10.57030200958252, | |
| "learning_rate": 6e-05, | |
| "loss": 3.0548, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.005150044567693374, | |
| "grad_norm": 9.254692077636719, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.5015, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.005546201842131326, | |
| "grad_norm": 9.80490779876709, | |
| "learning_rate": 7e-05, | |
| "loss": 2.2598, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.005942359116569278, | |
| "grad_norm": 10.694642066955566, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.7476, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.00633851639100723, | |
| "grad_norm": 10.605643272399902, | |
| "learning_rate": 8e-05, | |
| "loss": 2.5001, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.006734673665445182, | |
| "grad_norm": 13.457276344299316, | |
| "learning_rate": 8.5e-05, | |
| "loss": 2.797, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0071308309398831335, | |
| "grad_norm": 10.881452560424805, | |
| "learning_rate": 9e-05, | |
| "loss": 2.497, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.007526988214321085, | |
| "grad_norm": 10.051718711853027, | |
| "learning_rate": 9.5e-05, | |
| "loss": 2.4237, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.007923145488759038, | |
| "grad_norm": 9.118144989013672, | |
| "learning_rate": 0.0001, | |
| "loss": 2.2963, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.008319302763196989, | |
| "grad_norm": 10.652849197387695, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 2.4412, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.008715460037634942, | |
| "grad_norm": 12.306429862976074, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 2.689, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.009111617312072893, | |
| "grad_norm": 10.86596393585205, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 2.6179, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.009507774586510845, | |
| "grad_norm": 10.322020530700684, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 2.3587, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.009903931860948796, | |
| "grad_norm": 10.52370834350586, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 2.3825, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.010300089135386749, | |
| "grad_norm": 12.126945495605469, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 2.8174, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0106962464098247, | |
| "grad_norm": 10.495783805847168, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 2.5925, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.011092403684262652, | |
| "grad_norm": 11.440340995788574, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 2.5071, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.011488560958700603, | |
| "grad_norm": 11.15341854095459, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 2.431, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.011884718233138556, | |
| "grad_norm": 12.230130195617676, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 2.8904, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.012280875507576509, | |
| "grad_norm": 11.259504318237305, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 2.7206, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.01267703278201446, | |
| "grad_norm": 12.922420501708984, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 3.0273, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.013073190056452412, | |
| "grad_norm": 9.90955924987793, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 2.0496, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.013469347330890363, | |
| "grad_norm": 13.736802101135254, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 2.8411, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.013865504605328316, | |
| "grad_norm": 10.806241035461426, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 2.2993, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.014261661879766267, | |
| "grad_norm": 10.51795768737793, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 2.3481, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01465781915420422, | |
| "grad_norm": 11.385435104370117, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 2.7561, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01505397642864217, | |
| "grad_norm": 13.5595703125, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 3.1686, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.015450133703080123, | |
| "grad_norm": 12.45494556427002, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 2.5412, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.015846290977518076, | |
| "grad_norm": 11.393204689025879, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 2.3917, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.016242448251956025, | |
| "grad_norm": 14.68457317352295, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 3.2398, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.016638605526393978, | |
| "grad_norm": 12.070296287536621, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 2.431, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.01703476280083193, | |
| "grad_norm": 21.221527099609375, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 3.3495, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.017430920075269883, | |
| "grad_norm": 13.779664039611816, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 2.8153, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.017827077349707832, | |
| "grad_norm": 13.476646423339844, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 3.2993, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.018223234624145785, | |
| "grad_norm": 16.80905532836914, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 3.6182, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.018619391898583738, | |
| "grad_norm": 13.21243953704834, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 3.0093, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.01901554917302169, | |
| "grad_norm": 14.631463050842285, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 2.7913, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.019411706447459643, | |
| "grad_norm": 12.769938468933105, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 2.2544, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.019807863721897592, | |
| "grad_norm": 17.29874038696289, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 3.2499, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.019807863721897592, | |
| "eval_loss": 0.7548462748527527, | |
| "eval_runtime": 343.3265, | |
| "eval_samples_per_second": 12.385, | |
| "eval_steps_per_second": 6.192, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.020204020996335545, | |
| "grad_norm": 13.948724746704102, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 3.4015, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.020600178270773498, | |
| "grad_norm": 8.921982765197754, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 2.2722, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.02099633554521145, | |
| "grad_norm": 7.509298324584961, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 2.9086, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0213924928196494, | |
| "grad_norm": 6.69866943359375, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 2.4154, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.021788650094087352, | |
| "grad_norm": 6.785119533538818, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 2.2951, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.022184807368525305, | |
| "grad_norm": 8.127997398376465, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.9319, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.022580964642963258, | |
| "grad_norm": 7.68143367767334, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 2.6769, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.022977121917401207, | |
| "grad_norm": 7.978808403015137, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 2.5758, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.02337327919183916, | |
| "grad_norm": 7.170241832733154, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 2.3414, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.023769436466277112, | |
| "grad_norm": 6.732669353485107, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 2.1366, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.024165593740715065, | |
| "grad_norm": 7.11915922164917, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 2.3087, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.024561751015153017, | |
| "grad_norm": 7.628853797912598, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 2.4743, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.024957908289590967, | |
| "grad_norm": 8.127632141113281, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 2.3436, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.02535406556402892, | |
| "grad_norm": 7.773597717285156, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 2.0663, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.025750222838466872, | |
| "grad_norm": 9.912444114685059, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 3.0763, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.026146380112904825, | |
| "grad_norm": 8.724308013916016, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 2.5321, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.026542537387342774, | |
| "grad_norm": 8.80415153503418, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 2.7565, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.026938694661780727, | |
| "grad_norm": 9.181212425231934, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.5467, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.02733485193621868, | |
| "grad_norm": 8.560654640197754, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 2.6073, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.027731009210656632, | |
| "grad_norm": 8.651196479797363, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 2.809, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.02812716648509458, | |
| "grad_norm": 7.863115310668945, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 2.3025, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.028523323759532534, | |
| "grad_norm": 8.734823226928711, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 2.7863, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.028919481033970486, | |
| "grad_norm": 8.805511474609375, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 2.483, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.02931563830840844, | |
| "grad_norm": 9.213960647583008, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 2.6902, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.02971179558284639, | |
| "grad_norm": 8.166645050048828, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 2.2442, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.03010795285728434, | |
| "grad_norm": 7.891556262969971, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 2.3533, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.030504110131722294, | |
| "grad_norm": 9.308650970458984, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 2.8334, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.030900267406160246, | |
| "grad_norm": 9.532136917114258, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 2.9882, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0312964246805982, | |
| "grad_norm": 9.510985374450684, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 3.2583, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.03169258195503615, | |
| "grad_norm": 9.957487106323242, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.6734, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.032088739229474104, | |
| "grad_norm": 10.348840713500977, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 3.1382, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.03248489650391205, | |
| "grad_norm": 8.748047828674316, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 2.3051, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.03288105377835, | |
| "grad_norm": 10.452115058898926, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 2.5516, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.033277211052787956, | |
| "grad_norm": 10.838149070739746, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 3.3848, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.03367336832722591, | |
| "grad_norm": 12.059874534606934, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 2.4924, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.03406952560166386, | |
| "grad_norm": 9.75168228149414, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 2.4424, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.034465682876101814, | |
| "grad_norm": 8.269457817077637, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 2.1441, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.034861840150539766, | |
| "grad_norm": 11.684209823608398, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 2.4579, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.03525799742497772, | |
| "grad_norm": 10.999527931213379, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 2.8323, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.035654154699415665, | |
| "grad_norm": 11.552958488464355, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 2.6295, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03605031197385362, | |
| "grad_norm": 12.94347858428955, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 2.2909, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.03644646924829157, | |
| "grad_norm": 9.04222583770752, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 2.0357, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.03684262652272952, | |
| "grad_norm": 11.101726531982422, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 2.8028, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.037238783797167475, | |
| "grad_norm": 11.385910034179688, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 2.4366, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.03763494107160543, | |
| "grad_norm": 10.282135963439941, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 2.6235, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.03803109834604338, | |
| "grad_norm": 13.128955841064453, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 2.8328, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.03842725562048133, | |
| "grad_norm": 11.919032096862793, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 2.9821, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.038823412894919286, | |
| "grad_norm": 13.125587463378906, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 2.9282, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.03921957016935723, | |
| "grad_norm": 11.837300300598145, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 2.3627, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.039615727443795185, | |
| "grad_norm": 13.225713729858398, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 2.665, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.039615727443795185, | |
| "eval_loss": 0.6710785031318665, | |
| "eval_runtime": 343.2884, | |
| "eval_samples_per_second": 12.386, | |
| "eval_steps_per_second": 6.193, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04001188471823314, | |
| "grad_norm": 6.430683135986328, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 2.5807, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.04040804199267109, | |
| "grad_norm": 6.545248508453369, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 2.5388, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.04080419926710904, | |
| "grad_norm": 6.790993690490723, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 2.5627, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.041200356541546995, | |
| "grad_norm": 5.768946170806885, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.2582, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.04159651381598495, | |
| "grad_norm": 7.711808681488037, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 2.291, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0419926710904229, | |
| "grad_norm": 6.631197452545166, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 2.7122, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.042388828364860846, | |
| "grad_norm": 6.862368583679199, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 2.6714, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.0427849856392988, | |
| "grad_norm": 6.4192214012146, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 2.5312, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.04318114291373675, | |
| "grad_norm": 6.572208404541016, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 2.5457, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.043577300188174704, | |
| "grad_norm": 7.45621919631958, | |
| "learning_rate": 5e-05, | |
| "loss": 2.5436, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04397345746261266, | |
| "grad_norm": 6.586299419403076, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 2.5885, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.04436961473705061, | |
| "grad_norm": 7.276597499847412, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 2.1434, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.04476577201148856, | |
| "grad_norm": 7.623659133911133, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 2.8945, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.045161929285926515, | |
| "grad_norm": 7.042168617248535, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 2.3028, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.04555808656036447, | |
| "grad_norm": 7.464590549468994, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 2.3807, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.045954243834802413, | |
| "grad_norm": 6.643574237823486, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.9407, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.046350401109240366, | |
| "grad_norm": 6.645323276519775, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 2.3134, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.04674655838367832, | |
| "grad_norm": 6.613039970397949, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 1.9895, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.04714271565811627, | |
| "grad_norm": 7.568057060241699, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 2.5604, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.047538872932554224, | |
| "grad_norm": 9.300191879272461, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 2.8018, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.04793503020699218, | |
| "grad_norm": 7.3803791999816895, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 2.5437, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.04833118748143013, | |
| "grad_norm": 7.900534152984619, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 2.9423, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.04872734475586808, | |
| "grad_norm": 8.615334510803223, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 2.8762, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.049123502030306035, | |
| "grad_norm": 8.52196216583252, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 2.9788, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.04951965930474398, | |
| "grad_norm": 7.072556495666504, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 2.1147, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.04991581657918193, | |
| "grad_norm": 8.038581848144531, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 2.6251, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.050311973853619886, | |
| "grad_norm": 8.452302932739258, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 2.7774, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.05070813112805784, | |
| "grad_norm": 7.472427845001221, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.3101, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.05110428840249579, | |
| "grad_norm": 7.390994548797607, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 2.2508, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.051500445676933744, | |
| "grad_norm": 6.412713527679443, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 2.2383, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0518966029513717, | |
| "grad_norm": 9.124238014221191, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 2.7566, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.05229276022580965, | |
| "grad_norm": 8.466002464294434, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 2.6618, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.052688917500247595, | |
| "grad_norm": 7.873275279998779, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 2.2592, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.05308507477468555, | |
| "grad_norm": 7.096005439758301, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 2.0873, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0534812320491235, | |
| "grad_norm": 10.305582046508789, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 3.0193, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.05387738932356145, | |
| "grad_norm": 10.208508491516113, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 2.7085, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.054273546597999406, | |
| "grad_norm": 12.142950057983398, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 2.64, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.05466970387243736, | |
| "grad_norm": 10.012589454650879, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 2.8601, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.05506586114687531, | |
| "grad_norm": 10.026057243347168, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 2.3891, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.055462018421313264, | |
| "grad_norm": 9.359071731567383, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 2.8986, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.05585817569575122, | |
| "grad_norm": 9.041458129882812, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 2.4639, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.05625433297018916, | |
| "grad_norm": 9.14760684967041, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 2.4462, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.056650490244627115, | |
| "grad_norm": 9.521930694580078, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 2.3937, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.05704664751906507, | |
| "grad_norm": 8.896791458129883, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 2.4117, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.05744280479350302, | |
| "grad_norm": 12.4694242477417, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 2.8917, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.05783896206794097, | |
| "grad_norm": 9.43685245513916, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.1703, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.058235119342378926, | |
| "grad_norm": 11.866545677185059, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 2.7334, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.05863127661681688, | |
| "grad_norm": 9.459863662719727, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 2.2491, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.05902743389125483, | |
| "grad_norm": 13.348437309265137, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 3.0507, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.05942359116569278, | |
| "grad_norm": 14.311579704284668, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 3.0234, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05942359116569278, | |
| "eval_loss": 0.63045334815979, | |
| "eval_runtime": 343.3222, | |
| "eval_samples_per_second": 12.385, | |
| "eval_steps_per_second": 6.192, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05981974844013073, | |
| "grad_norm": 4.873414516448975, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 2.1627, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.06021590571456868, | |
| "grad_norm": 5.070279121398926, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 2.0904, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.060612062989006635, | |
| "grad_norm": 6.505746364593506, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 2.6791, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.06100822026344459, | |
| "grad_norm": 6.289201259613037, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 2.6375, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.06140437753788254, | |
| "grad_norm": 6.557570457458496, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 2.9108, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.06180053481232049, | |
| "grad_norm": 6.779500484466553, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 2.5379, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.062196692086758446, | |
| "grad_norm": 5.956170082092285, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 2.2845, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.0625928493611964, | |
| "grad_norm": 6.594735145568848, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 2.326, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.06298900663563435, | |
| "grad_norm": 6.56148099899292, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 2.6377, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.0633851639100723, | |
| "grad_norm": 6.894052982330322, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 2.449, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06378132118451026, | |
| "grad_norm": 7.5433831214904785, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 2.5443, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.06417747845894821, | |
| "grad_norm": 6.663662433624268, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 2.3789, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.06457363573338615, | |
| "grad_norm": 6.053978443145752, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 1.9774, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.0649697930078241, | |
| "grad_norm": 6.855613708496094, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 2.3337, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.06536595028226205, | |
| "grad_norm": 7.299771308898926, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 2.7787, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.0657621075567, | |
| "grad_norm": 8.048273086547852, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 2.7755, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.06615826483113796, | |
| "grad_norm": 7.242124080657959, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 2.204, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.06655442210557591, | |
| "grad_norm": 7.778584957122803, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 2.2592, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.06695057938001386, | |
| "grad_norm": 7.791292190551758, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 2.4431, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.06734673665445182, | |
| "grad_norm": 7.882373332977295, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.7833, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.06774289392888977, | |
| "grad_norm": 8.636392593383789, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 2.8624, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.06813905120332772, | |
| "grad_norm": 7.124153137207031, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 2.0618, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.06853520847776567, | |
| "grad_norm": 7.282960414886475, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 2.5303, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.06893136575220363, | |
| "grad_norm": 7.873406887054443, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 2.5966, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.06932752302664158, | |
| "grad_norm": 6.703880310058594, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 2.068, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.06972368030107953, | |
| "grad_norm": 7.491240501403809, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.5188, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.07011983757551749, | |
| "grad_norm": 10.332925796508789, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 3.4564, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.07051599484995544, | |
| "grad_norm": 8.192689895629883, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 2.4159, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.07091215212439339, | |
| "grad_norm": 7.419778823852539, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 2.2505, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.07130830939883133, | |
| "grad_norm": 8.588955879211426, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 2.3399, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.07170446667326928, | |
| "grad_norm": 7.702329158782959, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 1.9926, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.07210062394770723, | |
| "grad_norm": 7.002358913421631, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.2113, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.07249678122214519, | |
| "grad_norm": 8.050068855285645, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 2.264, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.07289293849658314, | |
| "grad_norm": 8.686763763427734, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 2.6803, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.07328909577102109, | |
| "grad_norm": 7.135583400726318, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 2.1053, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.07368525304545905, | |
| "grad_norm": 9.32211685180664, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 2.8225, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.074081410319897, | |
| "grad_norm": 10.14301586151123, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 2.2446, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.07447756759433495, | |
| "grad_norm": 8.889291763305664, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 2.1738, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.0748737248687729, | |
| "grad_norm": 8.034686088562012, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 2.0352, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.07526988214321086, | |
| "grad_norm": 9.324980735778809, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 2.1299, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07566603941764881, | |
| "grad_norm": 9.698161125183105, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 2.6616, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.07606219669208676, | |
| "grad_norm": 11.161945343017578, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 3.0046, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.07645835396652471, | |
| "grad_norm": 9.398616790771484, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 2.4149, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.07685451124096267, | |
| "grad_norm": 10.527079582214355, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 2.583, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.07725066851540062, | |
| "grad_norm": 11.699482917785645, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 2.7391, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.07764682578983857, | |
| "grad_norm": 14.22352123260498, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 3.0816, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.07804298306427651, | |
| "grad_norm": 11.533976554870605, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 3.2266, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.07843914033871446, | |
| "grad_norm": 11.76283073425293, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 3.1872, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.07883529761315242, | |
| "grad_norm": 11.078360557556152, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 2.615, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.07923145488759037, | |
| "grad_norm": 12.138440132141113, | |
| "learning_rate": 0.0, | |
| "loss": 3.0815, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07923145488759037, | |
| "eval_loss": 0.6208789944648743, | |
| "eval_runtime": 343.2192, | |
| "eval_samples_per_second": 12.389, | |
| "eval_steps_per_second": 6.194, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 200, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.861996017975296e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |