| { | |
| "best_metric": 1.0096042156219482, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-200", | |
| "epoch": 0.0614486519701974, | |
| "eval_steps": 50, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00030724325985098704, | |
| "grad_norm": 23.011367797851562, | |
| "learning_rate": 5e-06, | |
| "loss": 6.3985, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00030724325985098704, | |
| "eval_loss": 2.409876585006714, | |
| "eval_runtime": 441.0598, | |
| "eval_samples_per_second": 12.429, | |
| "eval_steps_per_second": 6.215, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0006144865197019741, | |
| "grad_norm": 32.28715515136719, | |
| "learning_rate": 1e-05, | |
| "loss": 7.734, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0009217297795529611, | |
| "grad_norm": 28.548389434814453, | |
| "learning_rate": 1.5e-05, | |
| "loss": 7.4653, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0012289730394039482, | |
| "grad_norm": 24.919015884399414, | |
| "learning_rate": 2e-05, | |
| "loss": 7.3706, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.001536216299254935, | |
| "grad_norm": 22.995384216308594, | |
| "learning_rate": 2.5e-05, | |
| "loss": 6.938, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0018434595591059221, | |
| "grad_norm": 19.593135833740234, | |
| "learning_rate": 3e-05, | |
| "loss": 7.122, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0021507028189569092, | |
| "grad_norm": 18.867374420166016, | |
| "learning_rate": 3.5e-05, | |
| "loss": 5.8915, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0024579460788078963, | |
| "grad_norm": 15.144004821777344, | |
| "learning_rate": 4e-05, | |
| "loss": 5.7937, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.002765189338658883, | |
| "grad_norm": 13.516443252563477, | |
| "learning_rate": 4.5e-05, | |
| "loss": 5.6253, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.00307243259850987, | |
| "grad_norm": 14.180933952331543, | |
| "learning_rate": 5e-05, | |
| "loss": 5.3894, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.003379675858360857, | |
| "grad_norm": 16.558490753173828, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 4.628, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0036869191182118443, | |
| "grad_norm": 14.397795677185059, | |
| "learning_rate": 6e-05, | |
| "loss": 5.3816, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.003994162378062831, | |
| "grad_norm": 15.309121131896973, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 4.9166, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0043014056379138185, | |
| "grad_norm": 15.53266716003418, | |
| "learning_rate": 7e-05, | |
| "loss": 4.5761, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.004608648897764805, | |
| "grad_norm": 13.632649421691895, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 3.6609, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.004915892157615793, | |
| "grad_norm": 13.333073616027832, | |
| "learning_rate": 8e-05, | |
| "loss": 4.0839, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.005223135417466779, | |
| "grad_norm": 14.622278213500977, | |
| "learning_rate": 8.5e-05, | |
| "loss": 4.4781, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.005530378677317766, | |
| "grad_norm": 12.793045997619629, | |
| "learning_rate": 9e-05, | |
| "loss": 4.9027, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0058376219371687535, | |
| "grad_norm": 11.85013484954834, | |
| "learning_rate": 9.5e-05, | |
| "loss": 3.8893, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.00614486519701974, | |
| "grad_norm": 13.697391510009766, | |
| "learning_rate": 0.0001, | |
| "loss": 4.6765, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.006452108456870728, | |
| "grad_norm": 10.900583267211914, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 3.559, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.006759351716721714, | |
| "grad_norm": 15.428689002990723, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 4.3751, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.007066594976572701, | |
| "grad_norm": 11.113511085510254, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 4.1964, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0073738382364236885, | |
| "grad_norm": 12.102518081665039, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 4.4795, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.007681081496274675, | |
| "grad_norm": 11.722882270812988, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 4.2897, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.007988324756125662, | |
| "grad_norm": 11.406682968139648, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 5.0885, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.00829556801597665, | |
| "grad_norm": 11.243815422058105, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 4.4903, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.008602811275827637, | |
| "grad_norm": 11.628358840942383, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 4.4378, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.008910054535678624, | |
| "grad_norm": 10.681234359741211, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 4.691, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.00921729779552961, | |
| "grad_norm": 10.783159255981445, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 4.9312, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.009524541055380597, | |
| "grad_norm": 10.892800331115723, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 5.2761, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.009831784315231585, | |
| "grad_norm": 11.34912109375, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 5.4007, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.010139027575082572, | |
| "grad_norm": 13.050414085388184, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 5.5961, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.010446270834933559, | |
| "grad_norm": 11.140335083007812, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 4.5804, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.010753514094784545, | |
| "grad_norm": 11.838504791259766, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 5.553, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.011060757354635532, | |
| "grad_norm": 12.498956680297852, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 4.8211, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.01136800061448652, | |
| "grad_norm": 12.233826637268066, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 5.3758, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.011675243874337507, | |
| "grad_norm": 12.875245094299316, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 5.2887, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.011982487134188494, | |
| "grad_norm": 14.482268333435059, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 5.3555, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.01228973039403948, | |
| "grad_norm": 12.469817161560059, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 5.1884, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.012596973653890467, | |
| "grad_norm": 14.905447959899902, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 5.7557, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.012904216913741455, | |
| "grad_norm": 12.736101150512695, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 4.3636, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.013211460173592442, | |
| "grad_norm": 13.541585922241211, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 5.2159, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.013518703433443429, | |
| "grad_norm": 14.56790542602539, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 6.1784, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.013825946693294415, | |
| "grad_norm": 14.048850059509277, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 5.7569, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.014133189953145402, | |
| "grad_norm": 14.115535736083984, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 5.7633, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.01444043321299639, | |
| "grad_norm": 13.446468353271484, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 5.7742, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.014747676472847377, | |
| "grad_norm": 15.913680076599121, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 5.6512, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.015054919732698364, | |
| "grad_norm": 13.415315628051758, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 4.9549, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.01536216299254935, | |
| "grad_norm": 22.350265502929688, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 6.6309, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01536216299254935, | |
| "eval_loss": 1.475726842880249, | |
| "eval_runtime": 443.646, | |
| "eval_samples_per_second": 12.357, | |
| "eval_steps_per_second": 6.178, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015669406252400337, | |
| "grad_norm": 17.104642868041992, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 6.866, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.015976649512251324, | |
| "grad_norm": 12.813888549804688, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 6.4915, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.01628389277210231, | |
| "grad_norm": 8.647307395935059, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 4.6191, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.0165911360319533, | |
| "grad_norm": 8.637353897094727, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 5.2535, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.016898379291804287, | |
| "grad_norm": 8.05450439453125, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 4.8999, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.017205622551655274, | |
| "grad_norm": 8.113150596618652, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 4.5401, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.01751286581150626, | |
| "grad_norm": 9.034808158874512, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 5.2058, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.017820109071357247, | |
| "grad_norm": 9.577351570129395, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 4.8838, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.018127352331208234, | |
| "grad_norm": 8.906129837036133, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 4.4069, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.01843459559105922, | |
| "grad_norm": 8.663689613342285, | |
| "learning_rate": 8.83022221559489e-05, | |
| "loss": 4.3744, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.018741838850910207, | |
| "grad_norm": 8.294697761535645, | |
| "learning_rate": 8.773547901113862e-05, | |
| "loss": 4.1245, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.019049082110761194, | |
| "grad_norm": 8.371333122253418, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 4.5234, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.019356325370612184, | |
| "grad_norm": 8.438100814819336, | |
| "learning_rate": 8.656768508095853e-05, | |
| "loss": 4.2958, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.01966356863046317, | |
| "grad_norm": 8.364363670349121, | |
| "learning_rate": 8.596699001693255e-05, | |
| "loss": 3.934, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.019970811890314157, | |
| "grad_norm": 7.366670608520508, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 3.5163, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.020278055150165144, | |
| "grad_norm": 7.6314921379089355, | |
| "learning_rate": 8.473291852294987e-05, | |
| "loss": 3.7853, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.02058529841001613, | |
| "grad_norm": 7.76265811920166, | |
| "learning_rate": 8.409991800312493e-05, | |
| "loss": 3.8884, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.020892541669867117, | |
| "grad_norm": 8.086345672607422, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 3.5251, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.021199784929718104, | |
| "grad_norm": 6.946031093597412, | |
| "learning_rate": 8.280295144952536e-05, | |
| "loss": 3.593, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.02150702818956909, | |
| "grad_norm": 7.788767337799072, | |
| "learning_rate": 8.213938048432697e-05, | |
| "loss": 4.1621, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.021814271449420077, | |
| "grad_norm": 8.629899978637695, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 3.7884, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.022121514709271064, | |
| "grad_norm": 6.985069274902344, | |
| "learning_rate": 8.07830737662829e-05, | |
| "loss": 3.9133, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.022428757969122054, | |
| "grad_norm": 6.972623825073242, | |
| "learning_rate": 8.009075115760243e-05, | |
| "loss": 3.205, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.02273600122897304, | |
| "grad_norm": 7.396304130554199, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 3.4196, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.023043244488824027, | |
| "grad_norm": 7.159494876861572, | |
| "learning_rate": 7.86788218175523e-05, | |
| "loss": 3.5241, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.023350487748675014, | |
| "grad_norm": 7.689444541931152, | |
| "learning_rate": 7.795964517353735e-05, | |
| "loss": 4.3798, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.023657731008526, | |
| "grad_norm": 8.844693183898926, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 4.4668, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.023964974268376987, | |
| "grad_norm": 8.752437591552734, | |
| "learning_rate": 7.649596321166024e-05, | |
| "loss": 4.4342, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.024272217528227974, | |
| "grad_norm": 8.479850769042969, | |
| "learning_rate": 7.575190374550272e-05, | |
| "loss": 4.2627, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.02457946078807896, | |
| "grad_norm": 8.894695281982422, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 4.7375, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.024886704047929947, | |
| "grad_norm": 9.680545806884766, | |
| "learning_rate": 7.424048101231686e-05, | |
| "loss": 5.2633, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.025193947307780934, | |
| "grad_norm": 9.162818908691406, | |
| "learning_rate": 7.347357813929454e-05, | |
| "loss": 4.553, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.025501190567631924, | |
| "grad_norm": 9.10145378112793, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 4.7129, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.02580843382748291, | |
| "grad_norm": 9.074843406677246, | |
| "learning_rate": 7.191855733945387e-05, | |
| "loss": 4.0724, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.026115677087333897, | |
| "grad_norm": 9.07311725616455, | |
| "learning_rate": 7.113091308703498e-05, | |
| "loss": 4.7301, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.026422920347184884, | |
| "grad_norm": 11.810867309570312, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 5.2496, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.02673016360703587, | |
| "grad_norm": 10.740492820739746, | |
| "learning_rate": 6.953655642446368e-05, | |
| "loss": 5.3248, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.027037406866886857, | |
| "grad_norm": 10.580676078796387, | |
| "learning_rate": 6.873032967079561e-05, | |
| "loss": 4.8344, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.027344650126737844, | |
| "grad_norm": 11.064846992492676, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 4.8108, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.02765189338658883, | |
| "grad_norm": 11.720850944519043, | |
| "learning_rate": 6.710100716628344e-05, | |
| "loss": 4.8933, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.027959136646439817, | |
| "grad_norm": 9.551570892333984, | |
| "learning_rate": 6.627840772285784e-05, | |
| "loss": 4.714, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.028266379906290804, | |
| "grad_norm": 11.187040328979492, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 4.073, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.028573623166141794, | |
| "grad_norm": 9.788082122802734, | |
| "learning_rate": 6.461858523613684e-05, | |
| "loss": 4.7062, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.02888086642599278, | |
| "grad_norm": 10.513997077941895, | |
| "learning_rate": 6.378186779084995e-05, | |
| "loss": 4.612, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.029188109685843767, | |
| "grad_norm": 12.733663558959961, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 5.4784, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.029495352945694754, | |
| "grad_norm": 13.620742797851562, | |
| "learning_rate": 6.209609477998338e-05, | |
| "loss": 5.3613, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.02980259620554574, | |
| "grad_norm": 13.19336986541748, | |
| "learning_rate": 6.124755271719325e-05, | |
| "loss": 5.6018, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.030109839465396727, | |
| "grad_norm": 13.040425300598145, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 5.054, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.030417082725247714, | |
| "grad_norm": 14.068294525146484, | |
| "learning_rate": 5.9540449768827246e-05, | |
| "loss": 5.6559, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.0307243259850987, | |
| "grad_norm": 14.678174018859863, | |
| "learning_rate": 5.868240888334653e-05, | |
| "loss": 5.2266, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0307243259850987, | |
| "eval_loss": 1.2480071783065796, | |
| "eval_runtime": 443.7094, | |
| "eval_samples_per_second": 12.355, | |
| "eval_steps_per_second": 6.177, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.031031569244949687, | |
| "grad_norm": 10.580477714538574, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 5.5392, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.031338812504800674, | |
| "grad_norm": 9.869646072387695, | |
| "learning_rate": 5.695865504800327e-05, | |
| "loss": 5.5595, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.03164605576465166, | |
| "grad_norm": 9.99475383758545, | |
| "learning_rate": 5.6093467170257374e-05, | |
| "loss": 4.8398, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.03195329902450265, | |
| "grad_norm": 6.947811126708984, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 4.2438, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.032260542284353634, | |
| "grad_norm": 6.170292377471924, | |
| "learning_rate": 5.435778713738292e-05, | |
| "loss": 4.6814, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.03256778554420462, | |
| "grad_norm": 6.3110246658325195, | |
| "learning_rate": 5.348782368720626e-05, | |
| "loss": 4.4389, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.032875028804055614, | |
| "grad_norm": 6.087858200073242, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 4.1039, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.0331822720639066, | |
| "grad_norm": 6.2835211753845215, | |
| "learning_rate": 5.174497483512506e-05, | |
| "loss": 4.0166, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.03348951532375759, | |
| "grad_norm": 6.776898384094238, | |
| "learning_rate": 5.0872620321864185e-05, | |
| "loss": 4.2686, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.033796758583608574, | |
| "grad_norm": 6.588527202606201, | |
| "learning_rate": 5e-05, | |
| "loss": 3.7269, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.03410400184345956, | |
| "grad_norm": 7.68696403503418, | |
| "learning_rate": 4.912737967813583e-05, | |
| "loss": 3.9392, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.03441124510331055, | |
| "grad_norm": 5.877866744995117, | |
| "learning_rate": 4.825502516487497e-05, | |
| "loss": 3.5263, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.034718488363161534, | |
| "grad_norm": 7.236352920532227, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 3.7052, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.03502573162301252, | |
| "grad_norm": 6.501185894012451, | |
| "learning_rate": 4.6512176312793736e-05, | |
| "loss": 2.9949, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.03533297488286351, | |
| "grad_norm": 6.9282941818237305, | |
| "learning_rate": 4.564221286261709e-05, | |
| "loss": 3.5694, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.035640218142714494, | |
| "grad_norm": 7.355767250061035, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 3.9165, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.03594746140256548, | |
| "grad_norm": 7.192873954772949, | |
| "learning_rate": 4.390653282974264e-05, | |
| "loss": 3.1828, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.03625470466241647, | |
| "grad_norm": 6.482280254364014, | |
| "learning_rate": 4.3041344951996746e-05, | |
| "loss": 3.2275, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.036561947922267454, | |
| "grad_norm": 6.704769611358643, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 3.1593, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.03686919118211844, | |
| "grad_norm": 6.919817924499512, | |
| "learning_rate": 4.131759111665349e-05, | |
| "loss": 3.0843, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03717643444196943, | |
| "grad_norm": 7.4176506996154785, | |
| "learning_rate": 4.045955023117276e-05, | |
| "loss": 3.4191, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.037483677701820414, | |
| "grad_norm": 7.300436496734619, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 3.7824, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.0377909209616714, | |
| "grad_norm": 6.638614177703857, | |
| "learning_rate": 3.875244728280676e-05, | |
| "loss": 3.3756, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.03809816422152239, | |
| "grad_norm": 6.030972957611084, | |
| "learning_rate": 3.790390522001662e-05, | |
| "loss": 3.3504, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.038405407481373374, | |
| "grad_norm": 7.475324630737305, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 3.5805, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03871265074122437, | |
| "grad_norm": 7.8704609870910645, | |
| "learning_rate": 3.6218132209150045e-05, | |
| "loss": 3.6184, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.039019894001075354, | |
| "grad_norm": 7.465869426727295, | |
| "learning_rate": 3.5381414763863166e-05, | |
| "loss": 4.1467, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.03932713726092634, | |
| "grad_norm": 7.257153511047363, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 4.2342, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.03963438052077733, | |
| "grad_norm": 7.452193260192871, | |
| "learning_rate": 3.372159227714218e-05, | |
| "loss": 4.3429, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.039941623780628314, | |
| "grad_norm": 7.880541801452637, | |
| "learning_rate": 3.289899283371657e-05, | |
| "loss": 4.3863, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0402488670404793, | |
| "grad_norm": 8.45700740814209, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 4.8662, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.04055611030033029, | |
| "grad_norm": 8.435225486755371, | |
| "learning_rate": 3.12696703292044e-05, | |
| "loss": 4.2266, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.040863353560181274, | |
| "grad_norm": 9.154268264770508, | |
| "learning_rate": 3.046344357553632e-05, | |
| "loss": 4.2599, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.04117059682003226, | |
| "grad_norm": 7.695627212524414, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 4.0229, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.04147784007988325, | |
| "grad_norm": 9.314230918884277, | |
| "learning_rate": 2.886908691296504e-05, | |
| "loss": 4.8999, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.041785083339734234, | |
| "grad_norm": 9.278669357299805, | |
| "learning_rate": 2.8081442660546125e-05, | |
| "loss": 4.6078, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.04209232659958522, | |
| "grad_norm": 10.638617515563965, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 4.6089, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.04239956985943621, | |
| "grad_norm": 8.870823860168457, | |
| "learning_rate": 2.6526421860705473e-05, | |
| "loss": 4.3292, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.042706813119287194, | |
| "grad_norm": 9.005095481872559, | |
| "learning_rate": 2.575951898768315e-05, | |
| "loss": 4.5548, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.04301405637913818, | |
| "grad_norm": 10.735925674438477, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 4.8442, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.04332129963898917, | |
| "grad_norm": 9.419788360595703, | |
| "learning_rate": 2.4248096254497288e-05, | |
| "loss": 4.926, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.043628542898840154, | |
| "grad_norm": 9.60610580444336, | |
| "learning_rate": 2.350403678833976e-05, | |
| "loss": 4.2713, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.04393578615869114, | |
| "grad_norm": 8.531740188598633, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 4.0917, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.04424302941854213, | |
| "grad_norm": 9.785737037658691, | |
| "learning_rate": 2.2040354826462668e-05, | |
| "loss": 4.6779, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.044550272678393114, | |
| "grad_norm": 10.093374252319336, | |
| "learning_rate": 2.132117818244771e-05, | |
| "loss": 4.713, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.04485751593824411, | |
| "grad_norm": 11.576554298400879, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 5.0362, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.045164759198095095, | |
| "grad_norm": 12.363187789916992, | |
| "learning_rate": 1.9909248842397584e-05, | |
| "loss": 5.3881, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.04547200245794608, | |
| "grad_norm": 11.872509956359863, | |
| "learning_rate": 1.9216926233717085e-05, | |
| "loss": 4.8039, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.04577924571779707, | |
| "grad_norm": 13.223235130310059, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 5.4591, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.046086488977648055, | |
| "grad_norm": 13.729710578918457, | |
| "learning_rate": 1.7860619515673033e-05, | |
| "loss": 5.6227, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.046086488977648055, | |
| "eval_loss": 1.0585381984710693, | |
| "eval_runtime": 443.8526, | |
| "eval_samples_per_second": 12.351, | |
| "eval_steps_per_second": 6.175, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.04639373223749904, | |
| "grad_norm": 6.492424964904785, | |
| "learning_rate": 1.7197048550474643e-05, | |
| "loss": 4.6689, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.04670097549735003, | |
| "grad_norm": 6.938760757446289, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 4.6549, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.047008218757201015, | |
| "grad_norm": 6.7988762855529785, | |
| "learning_rate": 1.5900081996875083e-05, | |
| "loss": 4.5316, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.047315462017052, | |
| "grad_norm": 6.588143825531006, | |
| "learning_rate": 1.526708147705013e-05, | |
| "loss": 3.9889, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.04762270527690299, | |
| "grad_norm": 7.157522678375244, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 4.0667, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.047929948536753975, | |
| "grad_norm": 6.799620628356934, | |
| "learning_rate": 1.4033009983067452e-05, | |
| "loss": 4.1346, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.04823719179660496, | |
| "grad_norm": 6.208632469177246, | |
| "learning_rate": 1.3432314919041478e-05, | |
| "loss": 4.036, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.04854443505645595, | |
| "grad_norm": 7.102668762207031, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 4.2118, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.048851678316306935, | |
| "grad_norm": 5.840615749359131, | |
| "learning_rate": 1.22645209888614e-05, | |
| "loss": 3.6561, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.04915892157615792, | |
| "grad_norm": 6.531946182250977, | |
| "learning_rate": 1.1697777844051105e-05, | |
| "loss": 3.5827, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.04946616483600891, | |
| "grad_norm": 6.030324459075928, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 3.4239, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.049773408095859895, | |
| "grad_norm": 6.251836776733398, | |
| "learning_rate": 1.0599462319663905e-05, | |
| "loss": 3.4889, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.05008065135571088, | |
| "grad_norm": 6.034832000732422, | |
| "learning_rate": 1.006822449763537e-05, | |
| "loss": 3.4204, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.05038789461556187, | |
| "grad_norm": 6.695008277893066, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 3.3347, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.05069513787541286, | |
| "grad_norm": 6.495909214019775, | |
| "learning_rate": 9.042397785550405e-06, | |
| "loss": 3.6318, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.05100238113526385, | |
| "grad_norm": 6.50161600112915, | |
| "learning_rate": 8.548121372247918e-06, | |
| "loss": 3.5788, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.051309624395114835, | |
| "grad_norm": 6.825322151184082, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 3.5707, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.05161686765496582, | |
| "grad_norm": 6.349621772766113, | |
| "learning_rate": 7.597595192178702e-06, | |
| "loss": 3.1345, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.05192411091481681, | |
| "grad_norm": 5.96699333190918, | |
| "learning_rate": 7.1416349648943894e-06, | |
| "loss": 3.3035, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.052231354174667795, | |
| "grad_norm": 5.0452046394348145, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.964, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.05253859743451878, | |
| "grad_norm": 5.954025745391846, | |
| "learning_rate": 6.269014643030213e-06, | |
| "loss": 3.4229, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.05284584069436977, | |
| "grad_norm": 5.680531024932861, | |
| "learning_rate": 5.852620357053651e-06, | |
| "loss": 3.0665, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.053153083954220755, | |
| "grad_norm": 6.695633888244629, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 3.4811, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.05346032721407174, | |
| "grad_norm": 6.21097469329834, | |
| "learning_rate": 5.060297685041659e-06, | |
| "loss": 3.3157, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.05376757047392273, | |
| "grad_norm": 6.927323341369629, | |
| "learning_rate": 4.684610648167503e-06, | |
| "loss": 3.661, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.054074813733773715, | |
| "grad_norm": 7.5244669914245605, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 4.2119, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.0543820569936247, | |
| "grad_norm": 6.910071849822998, | |
| "learning_rate": 3.974757327377981e-06, | |
| "loss": 4.0973, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.05468930025347569, | |
| "grad_norm": 8.820836067199707, | |
| "learning_rate": 3.6408072716606346e-06, | |
| "loss": 4.4176, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.054996543513326675, | |
| "grad_norm": 6.843106269836426, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 4.0182, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.05530378677317766, | |
| "grad_norm": 6.976109027862549, | |
| "learning_rate": 3.0153689607045845e-06, | |
| "loss": 3.6852, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.05561103003302865, | |
| "grad_norm": 7.7527689933776855, | |
| "learning_rate": 2.724071220034158e-06, | |
| "loss": 4.1718, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.055918273292879635, | |
| "grad_norm": 8.710504531860352, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 3.9291, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.05622551655273062, | |
| "grad_norm": 8.126774787902832, | |
| "learning_rate": 2.1847622018482283e-06, | |
| "loss": 4.0263, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.05653275981258161, | |
| "grad_norm": 8.905202865600586, | |
| "learning_rate": 1.9369152030840556e-06, | |
| "loss": 4.5138, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.0568400030724326, | |
| "grad_norm": 8.544697761535645, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 4.6821, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.05714724633228359, | |
| "grad_norm": 8.854402542114258, | |
| "learning_rate": 1.4852136862001764e-06, | |
| "loss": 4.187, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.057454489592134575, | |
| "grad_norm": 9.037402153015137, | |
| "learning_rate": 1.2814967607382432e-06, | |
| "loss": 4.2726, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.05776173285198556, | |
| "grad_norm": 9.918100357055664, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 5.0083, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.05806897611183655, | |
| "grad_norm": 9.055500030517578, | |
| "learning_rate": 9.186408276168013e-07, | |
| "loss": 4.6175, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.058376219371687535, | |
| "grad_norm": 9.965828895568848, | |
| "learning_rate": 7.596123493895991e-07, | |
| "loss": 4.9183, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.05868346263153852, | |
| "grad_norm": 8.950596809387207, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 4.2883, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.05899070589138951, | |
| "grad_norm": 9.814715385437012, | |
| "learning_rate": 4.865965629214819e-07, | |
| "loss": 4.4442, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.059297949151240495, | |
| "grad_norm": 8.11826229095459, | |
| "learning_rate": 3.7269241793390085e-07, | |
| "loss": 3.7846, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.05960519241109148, | |
| "grad_norm": 11.356589317321777, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 4.5596, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.05991243567094247, | |
| "grad_norm": 10.671046257019043, | |
| "learning_rate": 1.9026509541272275e-07, | |
| "loss": 4.8289, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.060219678930793455, | |
| "grad_norm": 11.361414909362793, | |
| "learning_rate": 1.2179748700879012e-07, | |
| "loss": 5.3315, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.06052692219064444, | |
| "grad_norm": 10.357075691223145, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 4.7601, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.06083416545049543, | |
| "grad_norm": 12.07322883605957, | |
| "learning_rate": 3.04586490452119e-08, | |
| "loss": 5.5341, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.061141408710346415, | |
| "grad_norm": 13.117568016052246, | |
| "learning_rate": 7.615242180436522e-09, | |
| "loss": 5.3338, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.0614486519701974, | |
| "grad_norm": 17.20490074157715, | |
| "learning_rate": 0.0, | |
| "loss": 5.5981, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.0614486519701974, | |
| "eval_loss": 1.0096042156219482, | |
| "eval_runtime": 443.3236, | |
| "eval_samples_per_second": 12.366, | |
| "eval_steps_per_second": 6.183, | |
| "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": 2.865576738249769e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |