| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.97561622051418, | |
| "eval_steps": 0, | |
| "global_step": 90000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.010943206027150154, | |
| "learning_rate": 1.6563604240282684e-06, | |
| "loss": 0.0168, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.010196206159889698, | |
| "learning_rate": 3.312720848056537e-06, | |
| "loss": 0.0139, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.009697528555989265, | |
| "learning_rate": 4.969081272084806e-06, | |
| "loss": 0.0142, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.011516646482050419, | |
| "learning_rate": 6.625441696113074e-06, | |
| "loss": 0.0139, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.01910693757236004, | |
| "learning_rate": 8.281802120141344e-06, | |
| "loss": 0.0137, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.004923399072140455, | |
| "learning_rate": 9.938162544169612e-06, | |
| "loss": 0.0139, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.007313380483537912, | |
| "learning_rate": 1.159452296819788e-05, | |
| "loss": 0.0137, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.003835548646748066, | |
| "learning_rate": 1.3250883392226147e-05, | |
| "loss": 0.0139, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.007223702035844326, | |
| "learning_rate": 1.4907243816254417e-05, | |
| "loss": 0.0136, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.010532427579164505, | |
| "learning_rate": 1.6563604240282687e-05, | |
| "loss": 0.0135, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.0022224171552807093, | |
| "learning_rate": 1.8219964664310956e-05, | |
| "loss": 0.0137, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.0046217963099479675, | |
| "learning_rate": 1.9876325088339224e-05, | |
| "loss": 0.0138, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.007814540527760983, | |
| "learning_rate": 2.1532685512367493e-05, | |
| "loss": 0.0136, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.011144978925585747, | |
| "learning_rate": 2.318904593639576e-05, | |
| "loss": 0.0137, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.0014108135364949703, | |
| "learning_rate": 2.484540636042403e-05, | |
| "loss": 0.0138, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.0039382693357765675, | |
| "learning_rate": 2.6501766784452294e-05, | |
| "loss": 0.0135, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.014407187700271606, | |
| "learning_rate": 2.8158127208480566e-05, | |
| "loss": 0.0138, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.003938812296837568, | |
| "learning_rate": 2.9814487632508834e-05, | |
| "loss": 0.0138, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.0075125317089259624, | |
| "learning_rate": 2.983655639540591e-05, | |
| "loss": 0.0141, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.004186388570815325, | |
| "learning_rate": 2.9652498282124278e-05, | |
| "loss": 0.0139, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.004882327280938625, | |
| "learning_rate": 2.9468440168842645e-05, | |
| "loss": 0.0139, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.002360760699957609, | |
| "learning_rate": 2.9284382055561012e-05, | |
| "loss": 0.0138, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.003932056948542595, | |
| "learning_rate": 2.910032394227938e-05, | |
| "loss": 0.0141, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.015781084075570107, | |
| "learning_rate": 2.8916265828997743e-05, | |
| "loss": 0.0138, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.002987222746014595, | |
| "learning_rate": 2.873220771571611e-05, | |
| "loss": 0.0138, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.002478894544765353, | |
| "learning_rate": 2.8548149602434473e-05, | |
| "loss": 0.0138, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.00281110149808228, | |
| "learning_rate": 2.836409148915284e-05, | |
| "loss": 0.0138, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.0016416048165410757, | |
| "learning_rate": 2.8180033375871207e-05, | |
| "loss": 0.0136, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.003159256186336279, | |
| "learning_rate": 2.7995975262589574e-05, | |
| "loss": 0.0139, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.03868310526013374, | |
| "learning_rate": 2.781191714930794e-05, | |
| "loss": 0.0151, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.026466334238648415, | |
| "learning_rate": 2.7627859036026308e-05, | |
| "loss": 0.019, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.014158655889332294, | |
| "learning_rate": 2.7443800922744675e-05, | |
| "loss": 0.0184, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.0531819723546505, | |
| "learning_rate": 2.7259742809463042e-05, | |
| "loss": 0.018, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.031152892857789993, | |
| "learning_rate": 2.707568469618141e-05, | |
| "loss": 0.0163, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.045258529484272, | |
| "learning_rate": 2.6891626582899776e-05, | |
| "loss": 0.0164, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.08072955161333084, | |
| "learning_rate": 2.6707568469618143e-05, | |
| "loss": 0.0158, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.06747995316982269, | |
| "learning_rate": 2.652351035633651e-05, | |
| "loss": 0.0155, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.029322072863578796, | |
| "learning_rate": 2.6339452243054877e-05, | |
| "loss": 0.0151, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.03149860352277756, | |
| "learning_rate": 2.615539412977324e-05, | |
| "loss": 0.0151, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.03845517709851265, | |
| "learning_rate": 2.5971336016491608e-05, | |
| "loss": 0.0152, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.02114521712064743, | |
| "learning_rate": 2.578727790320997e-05, | |
| "loss": 0.0152, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.04469970241189003, | |
| "learning_rate": 2.560321978992834e-05, | |
| "loss": 0.0151, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.03439483791589737, | |
| "learning_rate": 2.5419161676646705e-05, | |
| "loss": 0.0148, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.016261784359812737, | |
| "learning_rate": 2.5235103563365072e-05, | |
| "loss": 0.015, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.04864068329334259, | |
| "learning_rate": 2.505104545008344e-05, | |
| "loss": 0.0147, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.024570118635892868, | |
| "learning_rate": 2.4866987336801806e-05, | |
| "loss": 0.0149, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.015043354593217373, | |
| "learning_rate": 2.4682929223520173e-05, | |
| "loss": 0.0151, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.038648445159196854, | |
| "learning_rate": 2.449887111023854e-05, | |
| "loss": 0.015, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.2623123824596405, | |
| "learning_rate": 2.4314812996956907e-05, | |
| "loss": 0.0148, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.02235906571149826, | |
| "learning_rate": 2.4130754883675274e-05, | |
| "loss": 0.0147, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 0.005854467861354351, | |
| "learning_rate": 2.394669677039364e-05, | |
| "loss": 0.0149, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.011547247879207134, | |
| "learning_rate": 2.376263865711201e-05, | |
| "loss": 0.0147, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.03119933046400547, | |
| "learning_rate": 2.3578580543830375e-05, | |
| "loss": 0.015, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 0.047728102654218674, | |
| "learning_rate": 2.339452243054874e-05, | |
| "loss": 0.0146, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.04931659996509552, | |
| "learning_rate": 2.3210464317267106e-05, | |
| "loss": 0.0145, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.22793345153331757, | |
| "learning_rate": 2.3026406203985473e-05, | |
| "loss": 0.0147, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.046943288296461105, | |
| "learning_rate": 2.2842348090703837e-05, | |
| "loss": 0.0149, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.07938718795776367, | |
| "learning_rate": 2.2658289977422203e-05, | |
| "loss": 0.0147, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.02574564516544342, | |
| "learning_rate": 2.247423186414057e-05, | |
| "loss": 0.0144, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.011776907369494438, | |
| "learning_rate": 2.2290173750858937e-05, | |
| "loss": 0.0147, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.0066869258880615234, | |
| "learning_rate": 2.2106115637577304e-05, | |
| "loss": 0.0147, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 0.010923570953309536, | |
| "learning_rate": 2.192205752429567e-05, | |
| "loss": 0.0145, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 0.03816843405365944, | |
| "learning_rate": 2.173799941101404e-05, | |
| "loss": 0.0149, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.04863005876541138, | |
| "learning_rate": 2.1553941297732405e-05, | |
| "loss": 0.0147, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.09232094883918762, | |
| "learning_rate": 2.1369883184450772e-05, | |
| "loss": 0.0148, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.006221645046025515, | |
| "learning_rate": 2.118582507116914e-05, | |
| "loss": 0.0148, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.008674496784806252, | |
| "learning_rate": 2.1001766957887506e-05, | |
| "loss": 0.0145, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.013610797002911568, | |
| "learning_rate": 2.081770884460587e-05, | |
| "loss": 0.0149, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.020487351343035698, | |
| "learning_rate": 2.0633650731324237e-05, | |
| "loss": 0.0147, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.004388992674648762, | |
| "learning_rate": 2.0449592618042604e-05, | |
| "loss": 0.0146, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.029407095164060593, | |
| "learning_rate": 2.026553450476097e-05, | |
| "loss": 0.0144, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 0.04079248011112213, | |
| "learning_rate": 2.0081476391479335e-05, | |
| "loss": 0.0146, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 0.033315982669591904, | |
| "learning_rate": 1.98974182781977e-05, | |
| "loss": 0.0143, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 0.00864444486796856, | |
| "learning_rate": 1.971336016491607e-05, | |
| "loss": 0.0145, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.03435393422842026, | |
| "learning_rate": 1.9529302051634436e-05, | |
| "loss": 0.0145, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.008053929544985294, | |
| "learning_rate": 1.9345243938352803e-05, | |
| "loss": 0.0146, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.004771470092236996, | |
| "learning_rate": 1.916118582507117e-05, | |
| "loss": 0.0143, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.016594666987657547, | |
| "learning_rate": 1.8977127711789537e-05, | |
| "loss": 0.0149, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.01181800477206707, | |
| "learning_rate": 1.8793069598507904e-05, | |
| "loss": 0.0143, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 0.03508065640926361, | |
| "learning_rate": 1.860901148522627e-05, | |
| "loss": 0.0146, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 0.028093870729207993, | |
| "learning_rate": 1.8424953371944638e-05, | |
| "loss": 0.0146, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 0.05029403790831566, | |
| "learning_rate": 1.8240895258663005e-05, | |
| "loss": 0.0146, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.006528925616294146, | |
| "learning_rate": 1.8056837145381368e-05, | |
| "loss": 0.0142, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 0.02557162009179592, | |
| "learning_rate": 1.7872779032099735e-05, | |
| "loss": 0.0144, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.025360217317938805, | |
| "learning_rate": 1.7688720918818102e-05, | |
| "loss": 0.0146, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.04788580909371376, | |
| "learning_rate": 1.750466280553647e-05, | |
| "loss": 0.0147, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.02906920574605465, | |
| "learning_rate": 1.7320604692254836e-05, | |
| "loss": 0.0144, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.012823808006942272, | |
| "learning_rate": 1.71365465789732e-05, | |
| "loss": 0.0145, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 0.008996455930173397, | |
| "learning_rate": 1.6952488465691567e-05, | |
| "loss": 0.0143, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 0.010119748301804066, | |
| "learning_rate": 1.6768430352409934e-05, | |
| "loss": 0.0146, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.02591855265200138, | |
| "learning_rate": 1.65843722391283e-05, | |
| "loss": 0.0143, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 0.013729671947658062, | |
| "learning_rate": 1.6400314125846668e-05, | |
| "loss": 0.0147, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.0771203562617302, | |
| "learning_rate": 1.6216256012565035e-05, | |
| "loss": 0.0146, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.04501279070973396, | |
| "learning_rate": 1.60321978992834e-05, | |
| "loss": 0.0144, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.03493111953139305, | |
| "learning_rate": 1.584813978600177e-05, | |
| "loss": 0.0144, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 0.01472916454076767, | |
| "learning_rate": 1.5664081672720136e-05, | |
| "loss": 0.0144, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.04763146862387657, | |
| "learning_rate": 1.54800235594385e-05, | |
| "loss": 0.0145, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.024467509239912033, | |
| "learning_rate": 1.5295965446156866e-05, | |
| "loss": 0.0144, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 0.01768341101706028, | |
| "learning_rate": 1.5111907332875235e-05, | |
| "loss": 0.0144, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 0.06102894991636276, | |
| "learning_rate": 1.49278492195936e-05, | |
| "loss": 0.0141, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 0.01851697266101837, | |
| "learning_rate": 1.4743791106311966e-05, | |
| "loss": 0.0142, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 0.015444310382008553, | |
| "learning_rate": 1.4559732993030333e-05, | |
| "loss": 0.0145, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.013120009563863277, | |
| "learning_rate": 1.43756748797487e-05, | |
| "loss": 0.0143, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.01589464209973812, | |
| "learning_rate": 1.4191616766467067e-05, | |
| "loss": 0.0141, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.040490709245204926, | |
| "learning_rate": 1.4007558653185433e-05, | |
| "loss": 0.0144, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 0.025874989107251167, | |
| "learning_rate": 1.38235005399038e-05, | |
| "loss": 0.0143, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.022394156083464622, | |
| "learning_rate": 1.3639442426622166e-05, | |
| "loss": 0.0144, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "grad_norm": 0.010273805819451809, | |
| "learning_rate": 1.3455384313340531e-05, | |
| "loss": 0.0144, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 0.028374383226037025, | |
| "learning_rate": 1.3271326200058898e-05, | |
| "loss": 0.0144, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 0.018441613763570786, | |
| "learning_rate": 1.3087268086777265e-05, | |
| "loss": 0.0145, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 0.006460436619818211, | |
| "learning_rate": 1.2903209973495632e-05, | |
| "loss": 0.0145, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 0.00770485308021307, | |
| "learning_rate": 1.2719151860213999e-05, | |
| "loss": 0.0144, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.00849447026848793, | |
| "learning_rate": 1.2535093746932366e-05, | |
| "loss": 0.0147, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.035621609538793564, | |
| "learning_rate": 1.2351035633650733e-05, | |
| "loss": 0.0145, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.008923010900616646, | |
| "learning_rate": 1.2166977520369098e-05, | |
| "loss": 0.0144, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.0056849876418709755, | |
| "learning_rate": 1.1982919407087464e-05, | |
| "loss": 0.0143, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.0071659935638308525, | |
| "learning_rate": 1.179886129380583e-05, | |
| "loss": 0.0144, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 0.021617043763399124, | |
| "learning_rate": 1.1614803180524198e-05, | |
| "loss": 0.0143, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 0.011144719086587429, | |
| "learning_rate": 1.1430745067242565e-05, | |
| "loss": 0.0142, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.010943782515823841, | |
| "learning_rate": 1.1246686953960932e-05, | |
| "loss": 0.0143, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.010286700911819935, | |
| "learning_rate": 1.1062628840679299e-05, | |
| "loss": 0.0143, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "grad_norm": 0.010169615969061852, | |
| "learning_rate": 1.0878570727397666e-05, | |
| "loss": 0.0143, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "grad_norm": 0.032067082822322845, | |
| "learning_rate": 1.069451261411603e-05, | |
| "loss": 0.0144, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.008680183440446854, | |
| "learning_rate": 1.0510454500834396e-05, | |
| "loss": 0.0143, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.01648719422519207, | |
| "learning_rate": 1.0326396387552763e-05, | |
| "loss": 0.0143, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.0210120789706707, | |
| "learning_rate": 1.014233827427113e-05, | |
| "loss": 0.0146, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 0.034336596727371216, | |
| "learning_rate": 9.958280160989497e-06, | |
| "loss": 0.0144, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 0.03138417750597, | |
| "learning_rate": 9.774222047707864e-06, | |
| "loss": 0.0141, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 0.01799875684082508, | |
| "learning_rate": 9.590163934426231e-06, | |
| "loss": 0.0142, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 0.02960127592086792, | |
| "learning_rate": 9.406105821144595e-06, | |
| "loss": 0.0143, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 0.012712860479950905, | |
| "learning_rate": 9.222047707862962e-06, | |
| "loss": 0.0146, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.009180006571114063, | |
| "learning_rate": 9.037989594581329e-06, | |
| "loss": 0.0143, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.0106426402926445, | |
| "learning_rate": 8.853931481299696e-06, | |
| "loss": 0.0143, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.03638075664639473, | |
| "learning_rate": 8.669873368018063e-06, | |
| "loss": 0.0141, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 0.02028089202940464, | |
| "learning_rate": 8.48581525473643e-06, | |
| "loss": 0.0144, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.004987742286175489, | |
| "learning_rate": 8.301757141454797e-06, | |
| "loss": 0.0143, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 0.012421207502484322, | |
| "learning_rate": 8.117699028173162e-06, | |
| "loss": 0.0145, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.05489884316921234, | |
| "learning_rate": 7.933640914891527e-06, | |
| "loss": 0.0142, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.007833471521735191, | |
| "learning_rate": 7.749582801609894e-06, | |
| "loss": 0.0145, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "grad_norm": 0.014776123687624931, | |
| "learning_rate": 7.565524688328261e-06, | |
| "loss": 0.0142, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 3.11, | |
| "grad_norm": 0.015590249560773373, | |
| "learning_rate": 7.381466575046628e-06, | |
| "loss": 0.0143, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "grad_norm": 0.018214261159300804, | |
| "learning_rate": 7.197408461764995e-06, | |
| "loss": 0.0142, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 0.029773008078336716, | |
| "learning_rate": 7.013350348483361e-06, | |
| "loss": 0.0143, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 3.18, | |
| "grad_norm": 0.028139958158135414, | |
| "learning_rate": 6.8292922352017276e-06, | |
| "loss": 0.0143, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.024558302015066147, | |
| "learning_rate": 6.6452341219200945e-06, | |
| "loss": 0.014, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 0.05188705772161484, | |
| "learning_rate": 6.461176008638461e-06, | |
| "loss": 0.0141, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.02240253984928131, | |
| "learning_rate": 6.277117895356828e-06, | |
| "loss": 0.0142, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 0.015994379296898842, | |
| "learning_rate": 6.093059782075194e-06, | |
| "loss": 0.0143, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 0.014095323160290718, | |
| "learning_rate": 5.909001668793561e-06, | |
| "loss": 0.0141, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 0.0076615894213318825, | |
| "learning_rate": 5.724943555511927e-06, | |
| "loss": 0.0141, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 0.023330098018050194, | |
| "learning_rate": 5.540885442230294e-06, | |
| "loss": 0.0143, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 0.022397508844733238, | |
| "learning_rate": 5.35682732894866e-06, | |
| "loss": 0.0141, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 3.38, | |
| "grad_norm": 0.01998765394091606, | |
| "learning_rate": 5.172769215667027e-06, | |
| "loss": 0.0143, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.07193479686975479, | |
| "learning_rate": 4.988711102385393e-06, | |
| "loss": 0.0143, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 3.42, | |
| "grad_norm": 0.030124777927994728, | |
| "learning_rate": 4.80465298910376e-06, | |
| "loss": 0.0146, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.0762249082326889, | |
| "learning_rate": 4.620594875822126e-06, | |
| "loss": 0.0144, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 0.030013220384716988, | |
| "learning_rate": 4.4365367625404925e-06, | |
| "loss": 0.0143, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 3.49, | |
| "grad_norm": 0.013210024684667587, | |
| "learning_rate": 4.2524786492588595e-06, | |
| "loss": 0.0144, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "grad_norm": 0.021476522088050842, | |
| "learning_rate": 4.068420535977226e-06, | |
| "loss": 0.0145, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 0.005120801739394665, | |
| "learning_rate": 3.884362422695593e-06, | |
| "loss": 0.0142, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 0.03378542512655258, | |
| "learning_rate": 3.7003043094139592e-06, | |
| "loss": 0.0144, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 3.58, | |
| "grad_norm": 0.004559422377496958, | |
| "learning_rate": 3.5162461961323254e-06, | |
| "loss": 0.0143, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.022087154909968376, | |
| "learning_rate": 3.3321880828506924e-06, | |
| "loss": 0.0142, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 0.027302134782075882, | |
| "learning_rate": 3.1481299695690585e-06, | |
| "loss": 0.0142, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 0.0070667564868927, | |
| "learning_rate": 2.964071856287425e-06, | |
| "loss": 0.0142, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 0.004392644390463829, | |
| "learning_rate": 2.7800137430057916e-06, | |
| "loss": 0.014, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 0.004756265785545111, | |
| "learning_rate": 2.595955629724158e-06, | |
| "loss": 0.0144, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 0.028167065232992172, | |
| "learning_rate": 2.4118975164425248e-06, | |
| "loss": 0.0141, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 3.73, | |
| "grad_norm": 0.04241223633289337, | |
| "learning_rate": 2.2278394031608913e-06, | |
| "loss": 0.0143, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.0073333000764250755, | |
| "learning_rate": 2.043781289879258e-06, | |
| "loss": 0.014, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 0.022960776463150978, | |
| "learning_rate": 1.8597231765976245e-06, | |
| "loss": 0.0146, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.009491208009421825, | |
| "learning_rate": 1.675665063315991e-06, | |
| "loss": 0.0143, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 3.82, | |
| "grad_norm": 0.04244249686598778, | |
| "learning_rate": 1.4916069500343576e-06, | |
| "loss": 0.0142, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.0845978856086731, | |
| "learning_rate": 1.307548836752724e-06, | |
| "loss": 0.0139, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 0.012684383429586887, | |
| "learning_rate": 1.1234907234710905e-06, | |
| "loss": 0.0143, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 3.89, | |
| "grad_norm": 0.010290221311151981, | |
| "learning_rate": 9.394326101894571e-07, | |
| "loss": 0.0144, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 0.009122644551098347, | |
| "learning_rate": 7.553744969078238e-07, | |
| "loss": 0.0143, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "grad_norm": 0.02026693895459175, | |
| "learning_rate": 5.713163836261902e-07, | |
| "loss": 0.0142, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 0.03241865336894989, | |
| "learning_rate": 3.8725827034455676e-07, | |
| "loss": 0.0142, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 3.98, | |
| "grad_norm": 0.04753628000617027, | |
| "learning_rate": 2.0320015706292333e-07, | |
| "loss": 0.0142, | |
| "step": 90000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 90552, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 10000, | |
| "total_flos": 0.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |