{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.720324810557501, "global_step": 65000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.966916338035625e-05, "loss": 7.5608, "step": 500 }, { "epoch": 0.03, "learning_rate": 4.9338326760712497e-05, "loss": 5.5847, "step": 1000 }, { "epoch": 0.04, "learning_rate": 4.900749014106874e-05, "loss": 4.6592, "step": 1500 }, { "epoch": 0.05, "learning_rate": 4.8676653521424984e-05, "loss": 4.0091, "step": 2000 }, { "epoch": 0.07, "learning_rate": 4.8345816901781224e-05, "loss": 3.567, "step": 2500 }, { "epoch": 0.08, "learning_rate": 4.801498028213747e-05, "loss": 3.2476, "step": 3000 }, { "epoch": 0.09, "learning_rate": 4.768414366249372e-05, "loss": 3.0105, "step": 3500 }, { "epoch": 0.11, "learning_rate": 4.735330704284996e-05, "loss": 2.8386, "step": 4000 }, { "epoch": 0.12, "learning_rate": 4.7022470423206206e-05, "loss": 2.7075, "step": 4500 }, { "epoch": 0.13, "learning_rate": 4.669163380356245e-05, "loss": 2.5952, "step": 5000 }, { "epoch": 0.15, "learning_rate": 4.63607971839187e-05, "loss": 2.5073, "step": 5500 }, { "epoch": 0.16, "learning_rate": 4.602996056427494e-05, "loss": 2.4273, "step": 6000 }, { "epoch": 0.17, "learning_rate": 4.569912394463118e-05, "loss": 2.3631, "step": 6500 }, { "epoch": 0.19, "learning_rate": 4.536828732498743e-05, "loss": 2.3059, "step": 7000 }, { "epoch": 0.2, "learning_rate": 4.5037450705343675e-05, "loss": 2.252, "step": 7500 }, { "epoch": 0.21, "learning_rate": 4.470661408569992e-05, "loss": 2.2036, "step": 8000 }, { "epoch": 0.22, "learning_rate": 4.437577746605617e-05, "loss": 2.1583, "step": 8500 }, { "epoch": 0.24, "learning_rate": 4.404494084641241e-05, "loss": 2.1248, "step": 9000 }, { "epoch": 0.25, "learning_rate": 4.3714104226768656e-05, "loss": 2.0883, "step": 9500 }, { "epoch": 0.26, "learning_rate": 4.33832676071249e-05, "loss": 2.0567, "step": 10000 }, { "epoch": 0.28, "learning_rate": 4.3052430987481144e-05, "loss": 2.0281, "step": 10500 }, { "epoch": 0.29, "learning_rate": 4.272159436783739e-05, "loss": 2.0014, "step": 11000 }, { "epoch": 0.3, "learning_rate": 4.239075774819363e-05, "loss": 1.9757, "step": 11500 }, { "epoch": 0.32, "learning_rate": 4.205992112854988e-05, "loss": 1.9505, "step": 12000 }, { "epoch": 0.33, "learning_rate": 4.1729084508906125e-05, "loss": 1.9295, "step": 12500 }, { "epoch": 0.34, "learning_rate": 4.139824788926237e-05, "loss": 1.91, "step": 13000 }, { "epoch": 0.36, "learning_rate": 4.106741126961861e-05, "loss": 1.8905, "step": 13500 }, { "epoch": 0.37, "learning_rate": 4.073657464997486e-05, "loss": 1.875, "step": 14000 }, { "epoch": 0.38, "learning_rate": 4.04057380303311e-05, "loss": 1.8563, "step": 14500 }, { "epoch": 0.4, "learning_rate": 4.007490141068735e-05, "loss": 1.8398, "step": 15000 }, { "epoch": 0.41, "learning_rate": 3.9744064791043594e-05, "loss": 1.8282, "step": 15500 }, { "epoch": 0.42, "learning_rate": 3.9413228171399835e-05, "loss": 1.8094, "step": 16000 }, { "epoch": 0.44, "learning_rate": 3.908239155175608e-05, "loss": 1.7997, "step": 16500 }, { "epoch": 0.45, "learning_rate": 3.875155493211233e-05, "loss": 1.7898, "step": 17000 }, { "epoch": 0.46, "learning_rate": 3.8420718312468576e-05, "loss": 1.7711, "step": 17500 }, { "epoch": 0.48, "learning_rate": 3.8089881692824816e-05, "loss": 1.7582, "step": 18000 }, { "epoch": 0.49, "learning_rate": 3.775904507318106e-05, "loss": 1.7495, "step": 18500 }, { "epoch": 0.5, "learning_rate": 3.7428208453537304e-05, "loss": 1.7372, "step": 19000 }, { "epoch": 0.52, "learning_rate": 3.709737183389355e-05, "loss": 1.726, "step": 19500 }, { "epoch": 0.53, "learning_rate": 3.67665352142498e-05, "loss": 1.7189, "step": 20000 }, { "epoch": 0.54, "learning_rate": 3.6435698594606045e-05, "loss": 1.7042, "step": 20500 }, { "epoch": 0.56, "learning_rate": 3.6104861974962285e-05, "loss": 1.7008, "step": 21000 }, { "epoch": 0.57, "learning_rate": 3.577402535531853e-05, "loss": 1.6903, "step": 21500 }, { "epoch": 0.58, "learning_rate": 3.544318873567477e-05, "loss": 1.6816, "step": 22000 }, { "epoch": 0.6, "learning_rate": 3.511235211603102e-05, "loss": 1.6652, "step": 22500 }, { "epoch": 0.61, "learning_rate": 3.478151549638727e-05, "loss": 1.6645, "step": 23000 }, { "epoch": 0.62, "learning_rate": 3.445067887674351e-05, "loss": 1.6567, "step": 23500 }, { "epoch": 0.64, "learning_rate": 3.4119842257099754e-05, "loss": 1.6457, "step": 24000 }, { "epoch": 0.65, "learning_rate": 3.3789005637456e-05, "loss": 1.64, "step": 24500 }, { "epoch": 0.66, "learning_rate": 3.345816901781225e-05, "loss": 1.6359, "step": 25000 }, { "epoch": 0.67, "learning_rate": 3.3127332398168496e-05, "loss": 1.6238, "step": 25500 }, { "epoch": 0.69, "learning_rate": 3.2796495778524736e-05, "loss": 1.6227, "step": 26000 }, { "epoch": 0.7, "learning_rate": 3.2465659158880976e-05, "loss": 1.6178, "step": 26500 }, { "epoch": 0.71, "learning_rate": 3.213482253923722e-05, "loss": 1.6051, "step": 27000 }, { "epoch": 0.73, "learning_rate": 3.180398591959347e-05, "loss": 1.6022, "step": 27500 }, { "epoch": 0.74, "learning_rate": 3.147314929994972e-05, "loss": 1.5949, "step": 28000 }, { "epoch": 0.75, "learning_rate": 3.114231268030596e-05, "loss": 1.5883, "step": 28500 }, { "epoch": 0.77, "learning_rate": 3.0811476060662205e-05, "loss": 1.586, "step": 29000 }, { "epoch": 0.78, "learning_rate": 3.048063944101845e-05, "loss": 1.5805, "step": 29500 }, { "epoch": 0.79, "learning_rate": 3.0149802821374696e-05, "loss": 1.577, "step": 30000 }, { "epoch": 0.81, "learning_rate": 2.9818966201730943e-05, "loss": 1.5667, "step": 30500 }, { "epoch": 0.82, "learning_rate": 2.9488129582087183e-05, "loss": 1.5675, "step": 31000 }, { "epoch": 0.83, "learning_rate": 2.9157292962443427e-05, "loss": 1.5637, "step": 31500 }, { "epoch": 0.85, "learning_rate": 2.8826456342799674e-05, "loss": 1.5588, "step": 32000 }, { "epoch": 0.86, "learning_rate": 2.849561972315592e-05, "loss": 1.5533, "step": 32500 }, { "epoch": 0.87, "learning_rate": 2.8164783103512165e-05, "loss": 1.5502, "step": 33000 }, { "epoch": 0.89, "learning_rate": 2.7833946483868405e-05, "loss": 1.5442, "step": 33500 }, { "epoch": 0.9, "learning_rate": 2.7503109864224652e-05, "loss": 1.5403, "step": 34000 }, { "epoch": 0.91, "learning_rate": 2.71722732445809e-05, "loss": 1.5389, "step": 34500 }, { "epoch": 0.93, "learning_rate": 2.6841436624937143e-05, "loss": 1.5341, "step": 35000 }, { "epoch": 0.94, "learning_rate": 2.6510600005293383e-05, "loss": 1.5268, "step": 35500 }, { "epoch": 0.95, "learning_rate": 2.617976338564963e-05, "loss": 1.5216, "step": 36000 }, { "epoch": 0.97, "learning_rate": 2.5848926766005877e-05, "loss": 1.5234, "step": 36500 }, { "epoch": 0.98, "learning_rate": 2.551809014636212e-05, "loss": 1.5196, "step": 37000 }, { "epoch": 0.99, "learning_rate": 2.5187253526718368e-05, "loss": 1.5141, "step": 37500 }, { "epoch": 1.01, "learning_rate": 2.4856416907074612e-05, "loss": 1.5129, "step": 38000 }, { "epoch": 1.02, "learning_rate": 2.452558028743086e-05, "loss": 1.5034, "step": 38500 }, { "epoch": 1.03, "learning_rate": 2.41947436677871e-05, "loss": 1.4969, "step": 39000 }, { "epoch": 1.05, "learning_rate": 2.3863907048143346e-05, "loss": 1.4956, "step": 39500 }, { "epoch": 1.06, "learning_rate": 2.353307042849959e-05, "loss": 1.492, "step": 40000 }, { "epoch": 1.07, "learning_rate": 2.3202233808855837e-05, "loss": 1.4883, "step": 40500 }, { "epoch": 1.09, "learning_rate": 2.287139718921208e-05, "loss": 1.4882, "step": 41000 }, { "epoch": 1.1, "learning_rate": 2.2540560569568325e-05, "loss": 1.4835, "step": 41500 }, { "epoch": 1.11, "learning_rate": 2.2209723949924572e-05, "loss": 1.4839, "step": 42000 }, { "epoch": 1.12, "learning_rate": 2.1878887330280815e-05, "loss": 1.4787, "step": 42500 }, { "epoch": 1.14, "learning_rate": 2.154805071063706e-05, "loss": 1.4794, "step": 43000 }, { "epoch": 1.15, "learning_rate": 2.1217214090993303e-05, "loss": 1.4779, "step": 43500 }, { "epoch": 1.16, "learning_rate": 2.088637747134955e-05, "loss": 1.4732, "step": 44000 }, { "epoch": 1.18, "learning_rate": 2.0555540851705797e-05, "loss": 1.4674, "step": 44500 }, { "epoch": 1.19, "learning_rate": 2.0224704232062037e-05, "loss": 1.4697, "step": 45000 }, { "epoch": 1.2, "learning_rate": 1.9893867612418284e-05, "loss": 1.4665, "step": 45500 }, { "epoch": 1.22, "learning_rate": 1.9563030992774528e-05, "loss": 1.4654, "step": 46000 }, { "epoch": 1.23, "learning_rate": 1.9232194373130775e-05, "loss": 1.4608, "step": 46500 }, { "epoch": 1.24, "learning_rate": 1.890135775348702e-05, "loss": 1.4598, "step": 47000 }, { "epoch": 1.26, "learning_rate": 1.8570521133843263e-05, "loss": 1.4592, "step": 47500 }, { "epoch": 1.27, "learning_rate": 1.823968451419951e-05, "loss": 1.4576, "step": 48000 }, { "epoch": 1.28, "learning_rate": 1.7908847894555753e-05, "loss": 1.4539, "step": 48500 }, { "epoch": 1.3, "learning_rate": 1.7578011274911997e-05, "loss": 1.4533, "step": 49000 }, { "epoch": 1.31, "learning_rate": 1.7247174655268244e-05, "loss": 1.4501, "step": 49500 }, { "epoch": 1.32, "learning_rate": 1.6916338035624488e-05, "loss": 1.4458, "step": 50000 }, { "epoch": 1.34, "learning_rate": 1.6585501415980735e-05, "loss": 1.4474, "step": 50500 }, { "epoch": 1.35, "learning_rate": 1.6254664796336975e-05, "loss": 1.4455, "step": 51000 }, { "epoch": 1.36, "learning_rate": 1.5923828176693222e-05, "loss": 1.4427, "step": 51500 }, { "epoch": 1.38, "learning_rate": 1.5592991557049466e-05, "loss": 1.4426, "step": 52000 }, { "epoch": 1.39, "learning_rate": 1.5262154937405713e-05, "loss": 1.4401, "step": 52500 }, { "epoch": 1.4, "learning_rate": 1.4931318317761959e-05, "loss": 1.4407, "step": 53000 }, { "epoch": 1.42, "learning_rate": 1.46004816981182e-05, "loss": 1.4338, "step": 53500 }, { "epoch": 1.43, "learning_rate": 1.4269645078474448e-05, "loss": 1.4365, "step": 54000 }, { "epoch": 1.44, "learning_rate": 1.393880845883069e-05, "loss": 1.4352, "step": 54500 }, { "epoch": 1.46, "learning_rate": 1.3607971839186937e-05, "loss": 1.4321, "step": 55000 }, { "epoch": 1.47, "learning_rate": 1.3277135219543182e-05, "loss": 1.4323, "step": 55500 }, { "epoch": 1.48, "learning_rate": 1.2946298599899426e-05, "loss": 1.4275, "step": 56000 }, { "epoch": 1.5, "learning_rate": 1.2615461980255671e-05, "loss": 1.4308, "step": 56500 }, { "epoch": 1.51, "learning_rate": 1.2284625360611917e-05, "loss": 1.4284, "step": 57000 }, { "epoch": 1.52, "learning_rate": 1.195378874096816e-05, "loss": 1.425, "step": 57500 }, { "epoch": 1.54, "learning_rate": 1.1622952121324406e-05, "loss": 1.4238, "step": 58000 }, { "epoch": 1.55, "learning_rate": 1.129211550168065e-05, "loss": 1.4235, "step": 58500 }, { "epoch": 1.56, "learning_rate": 1.0961278882036895e-05, "loss": 1.4215, "step": 59000 }, { "epoch": 1.57, "learning_rate": 1.063044226239314e-05, "loss": 1.4217, "step": 59500 }, { "epoch": 1.59, "learning_rate": 1.0299605642749386e-05, "loss": 1.4184, "step": 60000 }, { "epoch": 1.6, "learning_rate": 9.96876902310563e-06, "loss": 1.4165, "step": 60500 }, { "epoch": 1.61, "learning_rate": 9.637932403461875e-06, "loss": 1.4175, "step": 61000 }, { "epoch": 1.63, "learning_rate": 9.307095783818118e-06, "loss": 1.4209, "step": 61500 }, { "epoch": 1.64, "learning_rate": 8.976259164174366e-06, "loss": 1.4129, "step": 62000 }, { "epoch": 1.65, "learning_rate": 8.64542254453061e-06, "loss": 1.4149, "step": 62500 }, { "epoch": 1.67, "learning_rate": 8.314585924886855e-06, "loss": 1.4114, "step": 63000 }, { "epoch": 1.68, "learning_rate": 7.983749305243098e-06, "loss": 1.4108, "step": 63500 }, { "epoch": 1.69, "learning_rate": 7.652912685599344e-06, "loss": 1.4141, "step": 64000 }, { "epoch": 1.71, "learning_rate": 7.322076065955589e-06, "loss": 1.4125, "step": 64500 }, { "epoch": 1.72, "learning_rate": 6.991239446311834e-06, "loss": 1.412, "step": 65000 } ], "max_steps": 75566, "num_train_epochs": 2, "total_flos": 1.8615262647316808e+18, "trial_name": null, "trial_params": null }