| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 671, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0014903129657228018, | |
| "grad_norm": 0.7207538217956673, | |
| "learning_rate": 4.411764705882353e-08, | |
| "loss": 1.6077, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007451564828614009, | |
| "grad_norm": 0.65849854443806, | |
| "learning_rate": 2.2058823529411765e-07, | |
| "loss": 1.5945, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014903129657228018, | |
| "grad_norm": 0.6800875831730971, | |
| "learning_rate": 4.411764705882353e-07, | |
| "loss": 1.6315, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.022354694485842028, | |
| "grad_norm": 0.8644630471455123, | |
| "learning_rate": 6.61764705882353e-07, | |
| "loss": 1.6271, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.029806259314456036, | |
| "grad_norm": 0.7771872952949634, | |
| "learning_rate": 8.823529411764706e-07, | |
| "loss": 1.5795, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.037257824143070044, | |
| "grad_norm": 0.7397975175907138, | |
| "learning_rate": 1.1029411764705884e-06, | |
| "loss": 1.6225, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.044709388971684055, | |
| "grad_norm": 0.7702121293753343, | |
| "learning_rate": 1.323529411764706e-06, | |
| "loss": 1.6526, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05216095380029806, | |
| "grad_norm": 0.6644067274265962, | |
| "learning_rate": 1.5441176470588234e-06, | |
| "loss": 1.6006, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05961251862891207, | |
| "grad_norm": 0.6017633448350802, | |
| "learning_rate": 1.7647058823529412e-06, | |
| "loss": 1.6434, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06706408345752608, | |
| "grad_norm": 0.4509127973744261, | |
| "learning_rate": 1.9852941176470586e-06, | |
| "loss": 1.5392, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.07451564828614009, | |
| "grad_norm": 0.4610064447084038, | |
| "learning_rate": 2.2058823529411767e-06, | |
| "loss": 1.589, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08196721311475409, | |
| "grad_norm": 0.35955958459549675, | |
| "learning_rate": 2.4264705882352943e-06, | |
| "loss": 1.5594, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08941877794336811, | |
| "grad_norm": 0.3478335776850152, | |
| "learning_rate": 2.647058823529412e-06, | |
| "loss": 1.5798, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09687034277198212, | |
| "grad_norm": 0.3527200373964567, | |
| "learning_rate": 2.8676470588235296e-06, | |
| "loss": 1.5693, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.10432190760059612, | |
| "grad_norm": 0.29600438876672497, | |
| "learning_rate": 2.999918570372821e-06, | |
| "loss": 1.5045, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11177347242921014, | |
| "grad_norm": 0.2900544302757877, | |
| "learning_rate": 2.9990025885979037e-06, | |
| "loss": 1.483, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11922503725782414, | |
| "grad_norm": 0.2408233475597495, | |
| "learning_rate": 2.997069461623824e-06, | |
| "loss": 1.5672, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12667660208643816, | |
| "grad_norm": 0.28176854976259386, | |
| "learning_rate": 2.9941205011700118e-06, | |
| "loss": 1.5082, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.13412816691505217, | |
| "grad_norm": 0.19913384731707087, | |
| "learning_rate": 2.990157708247667e-06, | |
| "loss": 1.4819, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14157973174366617, | |
| "grad_norm": 0.21083590352306433, | |
| "learning_rate": 2.9851837718019762e-06, | |
| "loss": 1.5531, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.14903129657228018, | |
| "grad_norm": 0.20730699550159845, | |
| "learning_rate": 2.9792020668875367e-06, | |
| "loss": 1.481, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15648286140089418, | |
| "grad_norm": 0.18165412005960324, | |
| "learning_rate": 2.9722166523782167e-06, | |
| "loss": 1.4844, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.16393442622950818, | |
| "grad_norm": 0.15838156755040073, | |
| "learning_rate": 2.964232268213018e-06, | |
| "loss": 1.4455, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17138599105812222, | |
| "grad_norm": 0.16340737391300794, | |
| "learning_rate": 2.955254332179797e-06, | |
| "loss": 1.469, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.17883755588673622, | |
| "grad_norm": 0.14759052394480532, | |
| "learning_rate": 2.9452889362390366e-06, | |
| "loss": 1.5597, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18628912071535023, | |
| "grad_norm": 0.12914782899327282, | |
| "learning_rate": 2.9343428423901614e-06, | |
| "loss": 1.4765, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.19374068554396423, | |
| "grad_norm": 0.12460225982681429, | |
| "learning_rate": 2.9224234780831905e-06, | |
| "loss": 1.4656, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.20119225037257824, | |
| "grad_norm": 0.14217631629657101, | |
| "learning_rate": 2.9095389311788626e-06, | |
| "loss": 1.4872, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.20864381520119224, | |
| "grad_norm": 0.1272863323148308, | |
| "learning_rate": 2.8956979444606303e-06, | |
| "loss": 1.4314, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21609538002980627, | |
| "grad_norm": 0.11983511963861297, | |
| "learning_rate": 2.8809099097022624e-06, | |
| "loss": 1.4734, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.22354694485842028, | |
| "grad_norm": 0.12007606131316517, | |
| "learning_rate": 2.8651848612950768e-06, | |
| "loss": 1.4997, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23099850968703428, | |
| "grad_norm": 0.11593831929818996, | |
| "learning_rate": 2.848533469439122e-06, | |
| "loss": 1.3879, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.23845007451564829, | |
| "grad_norm": 0.11598915986268504, | |
| "learning_rate": 2.8309670329029358e-06, | |
| "loss": 1.4192, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2459016393442623, | |
| "grad_norm": 0.11190528057252458, | |
| "learning_rate": 2.8124974713567872e-06, | |
| "loss": 1.452, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2533532041728763, | |
| "grad_norm": 0.1203691003615664, | |
| "learning_rate": 2.79313731728461e-06, | |
| "loss": 1.4819, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2608047690014903, | |
| "grad_norm": 0.10952238143478435, | |
| "learning_rate": 2.772899707480108e-06, | |
| "loss": 1.4173, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.26825633383010433, | |
| "grad_norm": 0.10685505615273189, | |
| "learning_rate": 2.7517983741328146e-06, | |
| "loss": 1.4078, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2757078986587183, | |
| "grad_norm": 0.11464827031091669, | |
| "learning_rate": 2.729847635510137e-06, | |
| "loss": 1.5007, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.28315946348733234, | |
| "grad_norm": 0.11122715995599067, | |
| "learning_rate": 2.70706238624173e-06, | |
| "loss": 1.4541, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2906110283159464, | |
| "grad_norm": 0.10752842550767219, | |
| "learning_rate": 2.6834580872127733e-06, | |
| "loss": 1.4129, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.29806259314456035, | |
| "grad_norm": 0.11490358820200401, | |
| "learning_rate": 2.6590507550730175e-06, | |
| "loss": 1.3986, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3055141579731744, | |
| "grad_norm": 0.10706266255477938, | |
| "learning_rate": 2.6338569513687182e-06, | |
| "loss": 1.4404, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.31296572280178836, | |
| "grad_norm": 0.1069097244298094, | |
| "learning_rate": 2.6078937713048357e-06, | |
| "loss": 1.3969, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3204172876304024, | |
| "grad_norm": 0.1160964307611302, | |
| "learning_rate": 2.581178832145114e-06, | |
| "loss": 1.4672, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.32786885245901637, | |
| "grad_norm": 0.1097725420424586, | |
| "learning_rate": 2.553730261257924e-06, | |
| "loss": 1.387, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3353204172876304, | |
| "grad_norm": 0.10407180190964474, | |
| "learning_rate": 2.525566683815973e-06, | |
| "loss": 1.3973, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.34277198211624443, | |
| "grad_norm": 0.11536030542488797, | |
| "learning_rate": 2.496707210158233e-06, | |
| "loss": 1.4624, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3502235469448584, | |
| "grad_norm": 0.10982670956660034, | |
| "learning_rate": 2.4671714228226542e-06, | |
| "loss": 1.4078, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.35767511177347244, | |
| "grad_norm": 0.10587771136409477, | |
| "learning_rate": 2.4369793632584796e-06, | |
| "loss": 1.4285, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3651266766020864, | |
| "grad_norm": 0.1208719398977082, | |
| "learning_rate": 2.4061515182271535e-06, | |
| "loss": 1.4726, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.37257824143070045, | |
| "grad_norm": 0.10511587252373167, | |
| "learning_rate": 2.3747088059010745e-06, | |
| "loss": 1.3548, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38002980625931443, | |
| "grad_norm": 0.10957225921751226, | |
| "learning_rate": 2.342672561669611e-06, | |
| "loss": 1.4678, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.38748137108792846, | |
| "grad_norm": 0.10693917993857549, | |
| "learning_rate": 2.3100645236620133e-06, | |
| "loss": 1.3974, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3949329359165425, | |
| "grad_norm": 0.11581766322136706, | |
| "learning_rate": 2.276906817997054e-06, | |
| "loss": 1.4098, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.40238450074515647, | |
| "grad_norm": 0.10904876454822796, | |
| "learning_rate": 2.2432219437693897e-06, | |
| "loss": 1.4056, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4098360655737705, | |
| "grad_norm": 0.10831229242143947, | |
| "learning_rate": 2.209032757782848e-06, | |
| "loss": 1.3776, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4172876304023845, | |
| "grad_norm": 0.11150103499322561, | |
| "learning_rate": 2.174362459040989e-06, | |
| "loss": 1.4279, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4247391952309985, | |
| "grad_norm": 0.11628953499023266, | |
| "learning_rate": 2.139234573005468e-06, | |
| "loss": 1.4328, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.43219076005961254, | |
| "grad_norm": 0.1154264307480958, | |
| "learning_rate": 2.1036729356328806e-06, | |
| "loss": 1.4109, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4396423248882265, | |
| "grad_norm": 0.1060574220828385, | |
| "learning_rate": 2.06770167720092e-06, | |
| "loss": 1.3781, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.44709388971684055, | |
| "grad_norm": 0.1053110442104737, | |
| "learning_rate": 2.0313452059348308e-06, | |
| "loss": 1.4444, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.10693306802856366, | |
| "learning_rate": 1.99462819144525e-06, | |
| "loss": 1.4359, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.46199701937406856, | |
| "grad_norm": 0.11652642471380918, | |
| "learning_rate": 1.957575547988697e-06, | |
| "loss": 1.4028, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.46944858420268254, | |
| "grad_norm": 0.11274466501498542, | |
| "learning_rate": 1.9202124175620545e-06, | |
| "loss": 1.3804, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.47690014903129657, | |
| "grad_norm": 0.10861122941192258, | |
| "learning_rate": 1.8825641528425148e-06, | |
| "loss": 1.3951, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4843517138599106, | |
| "grad_norm": 0.11658514112563267, | |
| "learning_rate": 1.8446562999845715e-06, | |
| "loss": 1.4377, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4918032786885246, | |
| "grad_norm": 0.11414477697164481, | |
| "learning_rate": 1.8065145812857305e-06, | |
| "loss": 1.3746, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4992548435171386, | |
| "grad_norm": 0.10815121580373825, | |
| "learning_rate": 1.7681648777326943e-06, | |
| "loss": 1.4422, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.5067064083457526, | |
| "grad_norm": 0.11037495106798235, | |
| "learning_rate": 1.7296332114398704e-06, | |
| "loss": 1.3666, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5141579731743666, | |
| "grad_norm": 0.11676408141846001, | |
| "learning_rate": 1.6909457279921186e-06, | |
| "loss": 1.3734, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5216095380029806, | |
| "grad_norm": 0.1113201753442444, | |
| "learning_rate": 1.6521286787037178e-06, | |
| "loss": 1.4135, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5290611028315947, | |
| "grad_norm": 0.10247556885521109, | |
| "learning_rate": 1.613208402805586e-06, | |
| "loss": 1.4152, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5365126676602087, | |
| "grad_norm": 0.10445203605746987, | |
| "learning_rate": 1.5742113095728515e-06, | |
| "loss": 1.4376, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5439642324888226, | |
| "grad_norm": 0.10879333034477139, | |
| "learning_rate": 1.535163860404891e-06, | |
| "loss": 1.455, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5514157973174366, | |
| "grad_norm": 0.12869015642921014, | |
| "learning_rate": 1.4960925508699984e-06, | |
| "loss": 1.3821, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5588673621460507, | |
| "grad_norm": 0.10954384701245683, | |
| "learning_rate": 1.4570238927268746e-06, | |
| "loss": 1.4397, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5663189269746647, | |
| "grad_norm": 0.10796984723976862, | |
| "learning_rate": 1.4179843959351213e-06, | |
| "loss": 1.4644, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5737704918032787, | |
| "grad_norm": 0.11517040756519294, | |
| "learning_rate": 1.3790005506669643e-06, | |
| "loss": 1.4344, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5812220566318927, | |
| "grad_norm": 0.10926990342446409, | |
| "learning_rate": 1.340098809332401e-06, | |
| "loss": 1.4211, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5886736214605067, | |
| "grad_norm": 0.1115330774443057, | |
| "learning_rate": 1.3013055686299683e-06, | |
| "loss": 1.3726, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5961251862891207, | |
| "grad_norm": 0.10589693766514106, | |
| "learning_rate": 1.2626471516353158e-06, | |
| "loss": 1.4593, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6035767511177347, | |
| "grad_norm": 0.11370583262546306, | |
| "learning_rate": 1.22414978993974e-06, | |
| "loss": 1.3516, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6110283159463488, | |
| "grad_norm": 0.12504328760615357, | |
| "learning_rate": 1.1858396058507837e-06, | |
| "loss": 1.5039, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6184798807749627, | |
| "grad_norm": 0.11743701858841901, | |
| "learning_rate": 1.1477425946670016e-06, | |
| "loss": 1.413, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6259314456035767, | |
| "grad_norm": 0.11045498569848493, | |
| "learning_rate": 1.1098846070389027e-06, | |
| "loss": 1.3941, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6333830104321908, | |
| "grad_norm": 0.12144356998222744, | |
| "learning_rate": 1.0722913314280395e-06, | |
| "loss": 1.4172, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6408345752608048, | |
| "grad_norm": 0.10365373765687404, | |
| "learning_rate": 1.0349882766761573e-06, | |
| "loss": 1.3337, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6482861400894188, | |
| "grad_norm": 0.10647853031859235, | |
| "learning_rate": 9.980007546962206e-07, | |
| "loss": 1.3959, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6557377049180327, | |
| "grad_norm": 0.10731308955650487, | |
| "learning_rate": 9.613538632970634e-07, | |
| "loss": 1.4023, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6631892697466468, | |
| "grad_norm": 0.10450909600455241, | |
| "learning_rate": 9.250724691533223e-07, | |
| "loss": 1.3872, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6706408345752608, | |
| "grad_norm": 0.1041016914874247, | |
| "learning_rate": 8.891811909322058e-07, | |
| "loss": 1.4465, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6780923994038748, | |
| "grad_norm": 0.120460469459905, | |
| "learning_rate": 8.537043825885445e-07, | |
| "loss": 1.4501, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6855439642324889, | |
| "grad_norm": 0.1136665297817651, | |
| "learning_rate": 8.186661168394658e-07, | |
| "loss": 1.3963, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6929955290611028, | |
| "grad_norm": 0.11135508784322877, | |
| "learning_rate": 7.840901688299e-07, | |
| "loss": 1.4571, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7004470938897168, | |
| "grad_norm": 0.09679813596282592, | |
| "learning_rate": 7.500000000000003e-07, | |
| "loss": 1.3349, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7078986587183308, | |
| "grad_norm": 0.11033066007905798, | |
| "learning_rate": 7.16418742165435e-07, | |
| "loss": 1.4129, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7153502235469449, | |
| "grad_norm": 0.10642022161621655, | |
| "learning_rate": 6.83369181821336e-07, | |
| "loss": 1.4064, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7228017883755589, | |
| "grad_norm": 0.11256587052450513, | |
| "learning_rate": 6.508737446805704e-07, | |
| "loss": 1.4562, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7302533532041728, | |
| "grad_norm": 0.11453839206549224, | |
| "learning_rate": 6.189544804568165e-07, | |
| "loss": 1.375, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7377049180327869, | |
| "grad_norm": 0.10923398824173401, | |
| "learning_rate": 5.876330479027766e-07, | |
| "loss": 1.3996, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7451564828614009, | |
| "grad_norm": 0.10567442549564693, | |
| "learning_rate": 5.56930700113673e-07, | |
| "loss": 1.3151, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7526080476900149, | |
| "grad_norm": 0.11241400299999167, | |
| "learning_rate": 5.268682701060012e-07, | |
| "loss": 1.3493, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7600596125186289, | |
| "grad_norm": 0.12270530511888589, | |
| "learning_rate": 4.974661566813315e-07, | |
| "loss": 1.4556, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.767511177347243, | |
| "grad_norm": 0.11388307104576878, | |
| "learning_rate": 4.6874431058474127e-07, | |
| "loss": 1.3664, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7749627421758569, | |
| "grad_norm": 0.11683819283505842, | |
| "learning_rate": 4.4072222096727663e-07, | |
| "loss": 1.3655, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7824143070044709, | |
| "grad_norm": 0.11164956386870885, | |
| "learning_rate": 4.1341890216162934e-07, | |
| "loss": 1.4173, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.789865871833085, | |
| "grad_norm": 0.115617873949176, | |
| "learning_rate": 3.868528807799988e-07, | |
| "loss": 1.4142, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.797317436661699, | |
| "grad_norm": 0.11850575487319875, | |
| "learning_rate": 3.610421831428953e-07, | |
| "loss": 1.4074, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8047690014903129, | |
| "grad_norm": 0.10945804497786706, | |
| "learning_rate": 3.36004323047419e-07, | |
| "loss": 1.4065, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.812220566318927, | |
| "grad_norm": 0.1207296244907688, | |
| "learning_rate": 3.11756289883306e-07, | |
| "loss": 1.4036, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.819672131147541, | |
| "grad_norm": 0.11721523752905044, | |
| "learning_rate": 2.883145371048133e-07, | |
| "loss": 1.4412, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.827123695976155, | |
| "grad_norm": 0.1082841170490658, | |
| "learning_rate": 2.656949710662591e-07, | |
| "loss": 1.4002, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.834575260804769, | |
| "grad_norm": 0.10851598999478153, | |
| "learning_rate": 2.4391294022879947e-07, | |
| "loss": 1.3989, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.842026825633383, | |
| "grad_norm": 0.11994637882887171, | |
| "learning_rate": 2.2298322474575838e-07, | |
| "loss": 1.4249, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.849478390461997, | |
| "grad_norm": 0.11657770238088389, | |
| "learning_rate": 2.0292002643358892e-07, | |
| "loss": 1.3606, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.856929955290611, | |
| "grad_norm": 0.12222665777231102, | |
| "learning_rate": 1.8373695913525317e-07, | |
| "loss": 1.4042, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8643815201192251, | |
| "grad_norm": 0.11933591684408021, | |
| "learning_rate": 1.6544703948258172e-07, | |
| "loss": 1.3801, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8718330849478391, | |
| "grad_norm": 0.11512513535214042, | |
| "learning_rate": 1.4806267806386093e-07, | |
| "loss": 1.4339, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.879284649776453, | |
| "grad_norm": 0.10887999147831859, | |
| "learning_rate": 1.3159567100265506e-07, | |
| "loss": 1.3578, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.886736214605067, | |
| "grad_norm": 0.10936406330567362, | |
| "learning_rate": 1.1605719195356806e-07, | |
| "loss": 1.4209, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.8941877794336811, | |
| "grad_norm": 0.11262116794609218, | |
| "learning_rate": 1.0145778452038629e-07, | |
| "loss": 1.3144, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9016393442622951, | |
| "grad_norm": 0.10899372973775676, | |
| "learning_rate": 8.780735510173316e-08, | |
| "loss": 1.3914, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.10967422102767481, | |
| "learning_rate": 7.51151661691048e-08, | |
| "loss": 1.4328, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9165424739195231, | |
| "grad_norm": 0.1195412132912122, | |
| "learning_rate": 6.338982998183856e-08, | |
| "loss": 1.4603, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9239940387481371, | |
| "grad_norm": 0.11487548750982872, | |
| "learning_rate": 5.263930274328044e-08, | |
| "loss": 1.3718, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9314456035767511, | |
| "grad_norm": 0.10425903788261691, | |
| "learning_rate": 4.287087920212035e-08, | |
| "loss": 1.3577, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9388971684053651, | |
| "grad_norm": 0.10782320142147576, | |
| "learning_rate": 3.4091187702554485e-08, | |
| "loss": 1.4204, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9463487332339792, | |
| "grad_norm": 0.11831408481668713, | |
| "learning_rate": 2.630618568663584e-08, | |
| "loss": 1.3797, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9538002980625931, | |
| "grad_norm": 0.10973676611006619, | |
| "learning_rate": 1.9521155651863854e-08, | |
| "loss": 1.4409, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.9612518628912071, | |
| "grad_norm": 0.1141554461719489, | |
| "learning_rate": 1.3740701566756276e-08, | |
| "loss": 1.3842, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.9687034277198212, | |
| "grad_norm": 0.11403561646592845, | |
| "learning_rate": 8.968745746835983e-09, | |
| "loss": 1.3295, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9761549925484352, | |
| "grad_norm": 0.11455606102253561, | |
| "learning_rate": 5.208526193150764e-09, | |
| "loss": 1.434, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9836065573770492, | |
| "grad_norm": 0.10758938639100596, | |
| "learning_rate": 2.462594395134854e-09, | |
| "loss": 1.433, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9910581222056631, | |
| "grad_norm": 0.11199281621555614, | |
| "learning_rate": 7.328135993011631e-10, | |
| "loss": 1.4072, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.9985096870342772, | |
| "grad_norm": 0.10839803620089897, | |
| "learning_rate": 2.035754493812103e-11, | |
| "loss": 1.4364, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.4019134044647217, | |
| "eval_runtime": 563.36, | |
| "eval_samples_per_second": 4.235, | |
| "eval_steps_per_second": 0.133, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 671, | |
| "total_flos": 2236028102377472.0, | |
| "train_loss": 1.4386844755817987, | |
| "train_runtime": 17062.4084, | |
| "train_samples_per_second": 1.258, | |
| "train_steps_per_second": 0.039 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 671, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2236028102377472.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |