| { | |
| "best_metric": 0.6193411350250244, | |
| "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": 14.495563507080078, | |
| "learning_rate": 5e-06, | |
| "loss": 2.5987, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00039615727443795186, | |
| "eval_loss": 1.2088950872421265, | |
| "eval_runtime": 337.0637, | |
| "eval_samples_per_second": 12.615, | |
| "eval_steps_per_second": 6.307, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0007923145488759037, | |
| "grad_norm": 16.614559173583984, | |
| "learning_rate": 1e-05, | |
| "loss": 3.394, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0011884718233138556, | |
| "grad_norm": 18.835763931274414, | |
| "learning_rate": 1.5e-05, | |
| "loss": 3.0506, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0015846290977518075, | |
| "grad_norm": 18.204811096191406, | |
| "learning_rate": 2e-05, | |
| "loss": 3.2621, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0019807863721897595, | |
| "grad_norm": 14.541211128234863, | |
| "learning_rate": 2.5e-05, | |
| "loss": 2.9941, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0023769436466277113, | |
| "grad_norm": 13.049044609069824, | |
| "learning_rate": 3e-05, | |
| "loss": 2.9559, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.002773100921065663, | |
| "grad_norm": 11.005773544311523, | |
| "learning_rate": 3.5e-05, | |
| "loss": 2.655, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.003169258195503615, | |
| "grad_norm": 11.66140365600586, | |
| "learning_rate": 4e-05, | |
| "loss": 2.6704, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0035654154699415667, | |
| "grad_norm": 12.335329055786133, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.9587, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.003961572744379519, | |
| "grad_norm": 10.606803894042969, | |
| "learning_rate": 5e-05, | |
| "loss": 2.5995, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.004357730018817471, | |
| "grad_norm": 11.140325546264648, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.5927, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.004753887293255423, | |
| "grad_norm": 10.353896141052246, | |
| "learning_rate": 6e-05, | |
| "loss": 3.0605, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.005150044567693374, | |
| "grad_norm": 9.193838119506836, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 2.5009, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.005546201842131326, | |
| "grad_norm": 9.763623237609863, | |
| "learning_rate": 7e-05, | |
| "loss": 2.2701, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.005942359116569278, | |
| "grad_norm": 10.643813133239746, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.7406, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.00633851639100723, | |
| "grad_norm": 10.669593811035156, | |
| "learning_rate": 8e-05, | |
| "loss": 2.4921, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.006734673665445182, | |
| "grad_norm": 11.960747718811035, | |
| "learning_rate": 8.5e-05, | |
| "loss": 2.7845, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0071308309398831335, | |
| "grad_norm": 10.707191467285156, | |
| "learning_rate": 9e-05, | |
| "loss": 2.4889, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.007526988214321085, | |
| "grad_norm": 9.775712966918945, | |
| "learning_rate": 9.5e-05, | |
| "loss": 2.4272, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.007923145488759038, | |
| "grad_norm": 9.274944305419922, | |
| "learning_rate": 0.0001, | |
| "loss": 2.3026, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.008319302763196989, | |
| "grad_norm": 10.755288124084473, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 2.449, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.008715460037634942, | |
| "grad_norm": 12.539787292480469, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 2.6936, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.009111617312072893, | |
| "grad_norm": 11.043819427490234, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 2.6237, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.009507774586510845, | |
| "grad_norm": 10.186410903930664, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 2.3617, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.009903931860948796, | |
| "grad_norm": 10.530620574951172, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 2.3778, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.010300089135386749, | |
| "grad_norm": 13.25385856628418, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 2.8179, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0106962464098247, | |
| "grad_norm": 10.420865058898926, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 2.5846, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.011092403684262652, | |
| "grad_norm": 11.93872356414795, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 2.5176, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.011488560958700603, | |
| "grad_norm": 11.279356002807617, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 2.4314, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.011884718233138556, | |
| "grad_norm": 11.756826400756836, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 2.8805, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.012280875507576509, | |
| "grad_norm": 11.12096881866455, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 2.7252, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.01267703278201446, | |
| "grad_norm": 12.88371467590332, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 3.0314, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.013073190056452412, | |
| "grad_norm": 10.017961502075195, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 2.0577, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.013469347330890363, | |
| "grad_norm": 13.735207557678223, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 2.8325, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.013865504605328316, | |
| "grad_norm": 10.726877212524414, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 2.3005, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.014261661879766267, | |
| "grad_norm": 10.604691505432129, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 2.341, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01465781915420422, | |
| "grad_norm": 11.629590034484863, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 2.7549, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.01505397642864217, | |
| "grad_norm": 14.335546493530273, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 3.1741, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.015450133703080123, | |
| "grad_norm": 11.522375106811523, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 2.5413, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.015846290977518076, | |
| "grad_norm": 11.158613204956055, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 2.4075, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.016242448251956025, | |
| "grad_norm": 15.320098876953125, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 3.231, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.016638605526393978, | |
| "grad_norm": 13.03093147277832, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 2.4665, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.01703476280083193, | |
| "grad_norm": 16.613243103027344, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 3.294, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.017430920075269883, | |
| "grad_norm": 14.335494995117188, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 2.815, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.017827077349707832, | |
| "grad_norm": 13.271525382995605, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 3.2896, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.018223234624145785, | |
| "grad_norm": 20.365318298339844, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 3.6564, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.018619391898583738, | |
| "grad_norm": 13.140840530395508, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 2.9866, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.01901554917302169, | |
| "grad_norm": 13.794374465942383, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 2.8268, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.019411706447459643, | |
| "grad_norm": 13.193543434143066, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 2.2512, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.019807863721897592, | |
| "grad_norm": 17.28192138671875, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 3.2413, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.019807863721897592, | |
| "eval_loss": 0.7558191418647766, | |
| "eval_runtime": 338.9865, | |
| "eval_samples_per_second": 12.543, | |
| "eval_steps_per_second": 6.272, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.020204020996335545, | |
| "grad_norm": 13.773219108581543, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 3.3845, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.020600178270773498, | |
| "grad_norm": 9.051798820495605, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 2.2873, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.02099633554521145, | |
| "grad_norm": 7.766826629638672, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 2.9136, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0213924928196494, | |
| "grad_norm": 6.526791572570801, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 2.4049, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.021788650094087352, | |
| "grad_norm": 7.026968479156494, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 2.2845, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.022184807368525305, | |
| "grad_norm": 8.329981803894043, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.9457, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.022580964642963258, | |
| "grad_norm": 7.5735626220703125, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 2.7011, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.022977121917401207, | |
| "grad_norm": 7.662240028381348, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 2.574, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.02337327919183916, | |
| "grad_norm": 7.075723171234131, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 2.3364, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.023769436466277112, | |
| "grad_norm": 6.765683174133301, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 2.1123, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.024165593740715065, | |
| "grad_norm": 7.3614397048950195, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 2.3193, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.024561751015153017, | |
| "grad_norm": 8.243191719055176, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 2.4855, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.024957908289590967, | |
| "grad_norm": 7.8791680335998535, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 2.3414, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.02535406556402892, | |
| "grad_norm": 8.094986915588379, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 2.0803, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.025750222838466872, | |
| "grad_norm": 10.040471076965332, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 3.0746, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.026146380112904825, | |
| "grad_norm": 8.993818283081055, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 2.556, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.026542537387342774, | |
| "grad_norm": 14.81818962097168, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 2.764, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.026938694661780727, | |
| "grad_norm": 9.128986358642578, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.5565, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.02733485193621868, | |
| "grad_norm": 8.368912696838379, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 2.6037, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.027731009210656632, | |
| "grad_norm": 9.154232025146484, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 2.8112, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.02812716648509458, | |
| "grad_norm": 7.73675012588501, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 2.2716, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.028523323759532534, | |
| "grad_norm": 8.645244598388672, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 2.7805, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.028919481033970486, | |
| "grad_norm": 8.736540794372559, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 2.494, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.02931563830840844, | |
| "grad_norm": 9.59007453918457, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 2.7142, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.02971179558284639, | |
| "grad_norm": 8.552069664001465, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 2.2465, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.03010795285728434, | |
| "grad_norm": 7.933364391326904, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 2.3576, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.030504110131722294, | |
| "grad_norm": 9.648405075073242, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 2.8443, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.030900267406160246, | |
| "grad_norm": 10.026429176330566, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 2.994, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0312964246805982, | |
| "grad_norm": 9.404814720153809, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 3.2357, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.03169258195503615, | |
| "grad_norm": 9.66026782989502, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.6634, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.032088739229474104, | |
| "grad_norm": 10.365447998046875, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 3.1117, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.03248489650391205, | |
| "grad_norm": 8.9589262008667, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 2.3134, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.03288105377835, | |
| "grad_norm": 11.305326461791992, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 2.5728, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.033277211052787956, | |
| "grad_norm": 10.456574440002441, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 3.3556, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.03367336832722591, | |
| "grad_norm": 11.175398826599121, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 2.5528, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.03406952560166386, | |
| "grad_norm": 9.353387832641602, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 2.4417, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.034465682876101814, | |
| "grad_norm": 8.356178283691406, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 2.166, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.034861840150539766, | |
| "grad_norm": 9.713857650756836, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 2.4152, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.03525799742497772, | |
| "grad_norm": 10.957006454467773, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 2.8269, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.035654154699415665, | |
| "grad_norm": 10.82498836517334, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 2.5913, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.03605031197385362, | |
| "grad_norm": 13.92136287689209, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 2.3097, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.03644646924829157, | |
| "grad_norm": 8.223499298095703, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 2.0382, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.03684262652272952, | |
| "grad_norm": 13.45029354095459, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 2.8246, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.037238783797167475, | |
| "grad_norm": 12.594712257385254, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 2.4397, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.03763494107160543, | |
| "grad_norm": 9.746724128723145, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 2.6255, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.03803109834604338, | |
| "grad_norm": 12.217853546142578, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 2.7994, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.03842725562048133, | |
| "grad_norm": 11.843147277832031, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 2.9499, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.038823412894919286, | |
| "grad_norm": 12.877901077270508, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 2.9359, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.03921957016935723, | |
| "grad_norm": 11.579832077026367, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 2.3228, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.039615727443795185, | |
| "grad_norm": 12.583760261535645, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 2.6903, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.039615727443795185, | |
| "eval_loss": 0.6697877645492554, | |
| "eval_runtime": 339.4624, | |
| "eval_samples_per_second": 12.526, | |
| "eval_steps_per_second": 6.263, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04001188471823314, | |
| "grad_norm": 6.360177040100098, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 2.5962, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.04040804199267109, | |
| "grad_norm": 6.526253700256348, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 2.5529, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.04080419926710904, | |
| "grad_norm": 6.883899211883545, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 2.5643, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.041200356541546995, | |
| "grad_norm": 5.620289325714111, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.2494, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.04159651381598495, | |
| "grad_norm": 7.3505401611328125, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 2.278, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0419926710904229, | |
| "grad_norm": 6.482987880706787, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 2.6903, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.042388828364860846, | |
| "grad_norm": 6.914910793304443, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 2.7104, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.0427849856392988, | |
| "grad_norm": 6.627212047576904, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 2.5426, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.04318114291373675, | |
| "grad_norm": 6.562288761138916, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 2.5501, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.043577300188174704, | |
| "grad_norm": 7.48563814163208, | |
| "learning_rate": 5e-05, | |
| "loss": 2.5545, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.04397345746261266, | |
| "grad_norm": 6.828021049499512, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 2.6011, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.04436961473705061, | |
| "grad_norm": 6.753318786621094, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 2.1348, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.04476577201148856, | |
| "grad_norm": 8.458637237548828, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 2.8946, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.045161929285926515, | |
| "grad_norm": 6.85118293762207, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 2.3202, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.04555808656036447, | |
| "grad_norm": 7.629461288452148, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 2.3985, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.045954243834802413, | |
| "grad_norm": 6.77186393737793, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.931, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.046350401109240366, | |
| "grad_norm": 6.728777885437012, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 2.3199, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.04674655838367832, | |
| "grad_norm": 6.741183280944824, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 1.9732, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.04714271565811627, | |
| "grad_norm": 7.823344707489014, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 2.5642, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.047538872932554224, | |
| "grad_norm": 9.227714538574219, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 2.8144, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.04793503020699218, | |
| "grad_norm": 7.38051176071167, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 2.5403, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.04833118748143013, | |
| "grad_norm": 8.060673713684082, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 2.9318, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.04872734475586808, | |
| "grad_norm": 8.677541732788086, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 2.8636, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.049123502030306035, | |
| "grad_norm": 8.290096282958984, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 2.9682, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.04951965930474398, | |
| "grad_norm": 7.013187408447266, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 2.0986, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.04991581657918193, | |
| "grad_norm": 7.830003261566162, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 2.6328, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.050311973853619886, | |
| "grad_norm": 8.967572212219238, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 2.7992, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.05070813112805784, | |
| "grad_norm": 7.2473907470703125, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.2991, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.05110428840249579, | |
| "grad_norm": 8.087238311767578, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 2.2725, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.051500445676933744, | |
| "grad_norm": 6.30985164642334, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 2.2416, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0518966029513717, | |
| "grad_norm": 9.02031421661377, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 2.7646, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.05229276022580965, | |
| "grad_norm": 9.004183769226074, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 2.6726, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.052688917500247595, | |
| "grad_norm": 8.092021942138672, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 2.2824, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.05308507477468555, | |
| "grad_norm": 7.0188703536987305, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 2.0659, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0534812320491235, | |
| "grad_norm": 9.969956398010254, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 3.0098, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.05387738932356145, | |
| "grad_norm": 9.049392700195312, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 2.6349, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.054273546597999406, | |
| "grad_norm": 10.317535400390625, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 2.5721, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.05466970387243736, | |
| "grad_norm": 10.092238426208496, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 2.8833, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.05506586114687531, | |
| "grad_norm": 10.784772872924805, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 2.398, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.055462018421313264, | |
| "grad_norm": 9.963539123535156, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 2.9027, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.05585817569575122, | |
| "grad_norm": 8.670623779296875, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 2.4363, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.05625433297018916, | |
| "grad_norm": 8.980132102966309, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 2.4414, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.056650490244627115, | |
| "grad_norm": 10.55542278289795, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 2.432, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.05704664751906507, | |
| "grad_norm": 9.629323959350586, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 2.451, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.05744280479350302, | |
| "grad_norm": 11.688613891601562, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 2.9462, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.05783896206794097, | |
| "grad_norm": 9.729354858398438, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.1626, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.058235119342378926, | |
| "grad_norm": 12.347620964050293, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 2.7457, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.05863127661681688, | |
| "grad_norm": 9.605571746826172, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 2.2582, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.05902743389125483, | |
| "grad_norm": 13.267195701599121, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 3.0407, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.05942359116569278, | |
| "grad_norm": 14.605932235717773, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 2.9647, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05942359116569278, | |
| "eval_loss": 0.6294466853141785, | |
| "eval_runtime": 340.6566, | |
| "eval_samples_per_second": 12.482, | |
| "eval_steps_per_second": 6.241, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.05981974844013073, | |
| "grad_norm": 4.8937788009643555, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 2.1533, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.06021590571456868, | |
| "grad_norm": 5.875320911407471, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 2.0878, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.060612062989006635, | |
| "grad_norm": 6.488078594207764, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 2.6752, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.06100822026344459, | |
| "grad_norm": 6.426373481750488, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 2.639, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.06140437753788254, | |
| "grad_norm": 6.61283540725708, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 2.8925, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.06180053481232049, | |
| "grad_norm": 6.746737957000732, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 2.5429, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.062196692086758446, | |
| "grad_norm": 5.992364883422852, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 2.3126, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.0625928493611964, | |
| "grad_norm": 6.37977933883667, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 2.2988, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.06298900663563435, | |
| "grad_norm": 6.413293838500977, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 2.6317, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.0633851639100723, | |
| "grad_norm": 6.493679046630859, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 2.4086, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.06378132118451026, | |
| "grad_norm": 7.268621921539307, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 2.534, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.06417747845894821, | |
| "grad_norm": 6.6669230461120605, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 2.3878, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.06457363573338615, | |
| "grad_norm": 6.02770471572876, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 1.9933, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.0649697930078241, | |
| "grad_norm": 7.1139326095581055, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 2.3498, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.06536595028226205, | |
| "grad_norm": 7.1752610206604, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 2.7669, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.0657621075567, | |
| "grad_norm": 8.256402015686035, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 2.7502, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.06615826483113796, | |
| "grad_norm": 7.215768814086914, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 2.2185, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.06655442210557591, | |
| "grad_norm": 7.52031946182251, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 2.2662, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.06695057938001386, | |
| "grad_norm": 7.943612575531006, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 2.4433, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.06734673665445182, | |
| "grad_norm": 8.064732551574707, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.7827, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.06774289392888977, | |
| "grad_norm": 8.597424507141113, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 2.8361, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.06813905120332772, | |
| "grad_norm": 7.607199668884277, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 2.032, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.06853520847776567, | |
| "grad_norm": 7.377221584320068, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 2.5357, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.06893136575220363, | |
| "grad_norm": 7.95181941986084, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 2.6041, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.06932752302664158, | |
| "grad_norm": 6.826475620269775, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 2.0748, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.06972368030107953, | |
| "grad_norm": 7.466386795043945, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.5444, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.07011983757551749, | |
| "grad_norm": 9.77598762512207, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 3.4488, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.07051599484995544, | |
| "grad_norm": 8.368354797363281, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 2.3909, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.07091215212439339, | |
| "grad_norm": 7.990671157836914, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 2.2496, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.07130830939883133, | |
| "grad_norm": 8.505362510681152, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 2.3256, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.07170446667326928, | |
| "grad_norm": 7.670304298400879, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 2.0108, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.07210062394770723, | |
| "grad_norm": 7.144657135009766, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.1987, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.07249678122214519, | |
| "grad_norm": 7.975318908691406, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 2.2561, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.07289293849658314, | |
| "grad_norm": 11.282790184020996, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 2.6531, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.07328909577102109, | |
| "grad_norm": 7.126767158508301, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 2.111, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.07368525304545905, | |
| "grad_norm": 9.32459545135498, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 2.8113, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.074081410319897, | |
| "grad_norm": 9.883260726928711, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 2.2595, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.07447756759433495, | |
| "grad_norm": 8.952421188354492, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 2.1888, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.0748737248687729, | |
| "grad_norm": 8.629685401916504, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 2.0509, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.07526988214321086, | |
| "grad_norm": 8.945446968078613, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 2.0993, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.07566603941764881, | |
| "grad_norm": 9.784154891967773, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 2.693, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.07606219669208676, | |
| "grad_norm": 12.141247749328613, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 3.0066, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.07645835396652471, | |
| "grad_norm": 9.189563751220703, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 2.3949, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.07685451124096267, | |
| "grad_norm": 10.456135749816895, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 2.578, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.07725066851540062, | |
| "grad_norm": 11.396419525146484, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 2.6849, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.07764682578983857, | |
| "grad_norm": 12.435629844665527, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 3.1103, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.07804298306427651, | |
| "grad_norm": 11.277499198913574, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 3.2427, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.07843914033871446, | |
| "grad_norm": 11.502141952514648, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 3.1574, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.07883529761315242, | |
| "grad_norm": 11.261180877685547, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 2.6461, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.07923145488759037, | |
| "grad_norm": 11.87040901184082, | |
| "learning_rate": 0.0, | |
| "loss": 3.0392, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07923145488759037, | |
| "eval_loss": 0.6193411350250244, | |
| "eval_runtime": 340.4464, | |
| "eval_samples_per_second": 12.489, | |
| "eval_steps_per_second": 6.245, | |
| "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 | |
| } | |