{ "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "global_step": 185568, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01, "learning_rate": 4.9865278496292465e-05, "loss": 2.8053, "step": 500 }, { "epoch": 0.02, "learning_rate": 4.973055699258493e-05, "loss": 2.4623, "step": 1000 }, { "epoch": 0.02, "learning_rate": 4.95958354888774e-05, "loss": 2.33, "step": 1500 }, { "epoch": 0.03, "learning_rate": 4.946111398516986e-05, "loss": 2.2541, "step": 2000 }, { "epoch": 0.04, "learning_rate": 4.9326392481462324e-05, "loss": 2.1654, "step": 2500 }, { "epoch": 0.05, "learning_rate": 4.9191670977754786e-05, "loss": 2.1165, "step": 3000 }, { "epoch": 0.06, "learning_rate": 4.9056949474047256e-05, "loss": 2.0528, "step": 3500 }, { "epoch": 0.06, "learning_rate": 4.892222797033972e-05, "loss": 2.056, "step": 4000 }, { "epoch": 0.07, "learning_rate": 4.878750646663218e-05, "loss": 2.0101, "step": 4500 }, { "epoch": 0.08, "learning_rate": 4.8652784962924645e-05, "loss": 1.9914, "step": 5000 }, { "epoch": 0.09, "learning_rate": 4.8518063459217114e-05, "loss": 1.9736, "step": 5500 }, { "epoch": 0.1, "learning_rate": 4.838334195550957e-05, "loss": 1.951, "step": 6000 }, { "epoch": 0.11, "learning_rate": 4.824862045180203e-05, "loss": 1.9505, "step": 6500 }, { "epoch": 0.11, "learning_rate": 4.8113898948094496e-05, "loss": 1.9406, "step": 7000 }, { "epoch": 0.12, "learning_rate": 4.7979177444386966e-05, "loss": 1.8956, "step": 7500 }, { "epoch": 0.13, "learning_rate": 4.784445594067943e-05, "loss": 1.8892, "step": 8000 }, { "epoch": 0.14, "learning_rate": 4.770973443697189e-05, "loss": 1.8653, "step": 8500 }, { "epoch": 0.15, "learning_rate": 4.7575012933264354e-05, "loss": 1.8717, "step": 9000 }, { "epoch": 0.15, "learning_rate": 4.7440291429556824e-05, "loss": 1.8696, "step": 9500 }, { "epoch": 0.16, "learning_rate": 4.730556992584929e-05, "loss": 1.8364, "step": 10000 }, { "epoch": 0.17, "learning_rate": 4.717084842214175e-05, "loss": 1.8237, "step": 10500 }, { "epoch": 0.18, "learning_rate": 4.703612691843421e-05, "loss": 1.802, "step": 11000 }, { "epoch": 0.19, "learning_rate": 4.690140541472668e-05, "loss": 1.822, "step": 11500 }, { "epoch": 0.19, "learning_rate": 4.6766683911019145e-05, "loss": 1.7857, "step": 12000 }, { "epoch": 0.2, "learning_rate": 4.663196240731161e-05, "loss": 1.8083, "step": 12500 }, { "epoch": 0.21, "learning_rate": 4.649724090360407e-05, "loss": 1.7915, "step": 13000 }, { "epoch": 0.22, "learning_rate": 4.6362519399896534e-05, "loss": 1.793, "step": 13500 }, { "epoch": 0.23, "learning_rate": 4.6227797896189e-05, "loss": 1.767, "step": 14000 }, { "epoch": 0.23, "learning_rate": 4.6093076392481466e-05, "loss": 1.7549, "step": 14500 }, { "epoch": 0.24, "learning_rate": 4.595835488877393e-05, "loss": 1.771, "step": 15000 }, { "epoch": 0.25, "learning_rate": 4.582363338506639e-05, "loss": 1.7506, "step": 15500 }, { "epoch": 0.26, "learning_rate": 4.568891188135886e-05, "loss": 1.7332, "step": 16000 }, { "epoch": 0.27, "learning_rate": 4.5554190377651324e-05, "loss": 1.7562, "step": 16500 }, { "epoch": 0.27, "learning_rate": 4.541946887394379e-05, "loss": 1.7213, "step": 17000 }, { "epoch": 0.28, "learning_rate": 4.528474737023625e-05, "loss": 1.7303, "step": 17500 }, { "epoch": 0.29, "learning_rate": 4.515002586652871e-05, "loss": 1.7064, "step": 18000 }, { "epoch": 0.3, "learning_rate": 4.5015304362821176e-05, "loss": 1.7069, "step": 18500 }, { "epoch": 0.31, "learning_rate": 4.488058285911364e-05, "loss": 1.707, "step": 19000 }, { "epoch": 0.32, "learning_rate": 4.47458613554061e-05, "loss": 1.7, "step": 19500 }, { "epoch": 0.32, "learning_rate": 4.461113985169857e-05, "loss": 1.7024, "step": 20000 }, { "epoch": 0.33, "learning_rate": 4.4476418347991034e-05, "loss": 1.7105, "step": 20500 }, { "epoch": 0.34, "learning_rate": 4.43416968442835e-05, "loss": 1.6947, "step": 21000 }, { "epoch": 0.35, "learning_rate": 4.420697534057596e-05, "loss": 1.6888, "step": 21500 }, { "epoch": 0.36, "learning_rate": 4.407225383686843e-05, "loss": 1.6849, "step": 22000 }, { "epoch": 0.36, "learning_rate": 4.393753233316089e-05, "loss": 1.6862, "step": 22500 }, { "epoch": 0.37, "learning_rate": 4.3802810829453355e-05, "loss": 1.675, "step": 23000 }, { "epoch": 0.38, "learning_rate": 4.366808932574582e-05, "loss": 1.673, "step": 23500 }, { "epoch": 0.39, "learning_rate": 4.353336782203829e-05, "loss": 1.6613, "step": 24000 }, { "epoch": 0.4, "learning_rate": 4.339864631833075e-05, "loss": 1.6772, "step": 24500 }, { "epoch": 0.4, "learning_rate": 4.3263924814623213e-05, "loss": 1.642, "step": 25000 }, { "epoch": 0.41, "learning_rate": 4.3129203310915676e-05, "loss": 1.6658, "step": 25500 }, { "epoch": 0.42, "learning_rate": 4.299448180720814e-05, "loss": 1.6414, "step": 26000 }, { "epoch": 0.43, "learning_rate": 4.285976030350061e-05, "loss": 1.6388, "step": 26500 }, { "epoch": 0.44, "learning_rate": 4.272503879979307e-05, "loss": 1.6249, "step": 27000 }, { "epoch": 0.44, "learning_rate": 4.2590317296085535e-05, "loss": 1.6347, "step": 27500 }, { "epoch": 0.45, "learning_rate": 4.2455595792378e-05, "loss": 1.6197, "step": 28000 }, { "epoch": 0.46, "learning_rate": 4.232087428867047e-05, "loss": 1.6148, "step": 28500 }, { "epoch": 0.47, "learning_rate": 4.218615278496293e-05, "loss": 1.629, "step": 29000 }, { "epoch": 0.48, "learning_rate": 4.205143128125539e-05, "loss": 1.6219, "step": 29500 }, { "epoch": 0.48, "learning_rate": 4.1916709777547856e-05, "loss": 1.619, "step": 30000 }, { "epoch": 0.49, "learning_rate": 4.178198827384032e-05, "loss": 1.6116, "step": 30500 }, { "epoch": 0.5, "learning_rate": 4.164726677013278e-05, "loss": 1.6139, "step": 31000 }, { "epoch": 0.51, "learning_rate": 4.1512545266425244e-05, "loss": 1.6062, "step": 31500 }, { "epoch": 0.52, "learning_rate": 4.137782376271771e-05, "loss": 1.6009, "step": 32000 }, { "epoch": 0.53, "learning_rate": 4.124310225901018e-05, "loss": 1.6162, "step": 32500 }, { "epoch": 0.53, "learning_rate": 4.110838075530264e-05, "loss": 1.5884, "step": 33000 }, { "epoch": 0.54, "learning_rate": 4.09736592515951e-05, "loss": 1.5862, "step": 33500 }, { "epoch": 0.55, "learning_rate": 4.0838937747887565e-05, "loss": 1.5778, "step": 34000 }, { "epoch": 0.56, "learning_rate": 4.0704216244180035e-05, "loss": 1.5676, "step": 34500 }, { "epoch": 0.57, "learning_rate": 4.05694947404725e-05, "loss": 1.5962, "step": 35000 }, { "epoch": 0.57, "learning_rate": 4.043477323676496e-05, "loss": 1.57, "step": 35500 }, { "epoch": 0.58, "learning_rate": 4.0300051733057424e-05, "loss": 1.5756, "step": 36000 }, { "epoch": 0.59, "learning_rate": 4.016533022934989e-05, "loss": 1.5408, "step": 36500 }, { "epoch": 0.6, "learning_rate": 4.0030608725642356e-05, "loss": 1.5559, "step": 37000 }, { "epoch": 0.61, "learning_rate": 3.989588722193482e-05, "loss": 1.5598, "step": 37500 }, { "epoch": 0.61, "learning_rate": 3.976116571822728e-05, "loss": 1.5708, "step": 38000 }, { "epoch": 0.62, "learning_rate": 3.9626444214519745e-05, "loss": 1.5726, "step": 38500 }, { "epoch": 0.63, "learning_rate": 3.9491722710812214e-05, "loss": 1.5552, "step": 39000 }, { "epoch": 0.64, "learning_rate": 3.935700120710468e-05, "loss": 1.5704, "step": 39500 }, { "epoch": 0.65, "learning_rate": 3.922227970339714e-05, "loss": 1.5493, "step": 40000 }, { "epoch": 0.65, "learning_rate": 3.90875581996896e-05, "loss": 1.5585, "step": 40500 }, { "epoch": 0.66, "learning_rate": 3.895283669598207e-05, "loss": 1.5543, "step": 41000 }, { "epoch": 0.67, "learning_rate": 3.8818115192274536e-05, "loss": 1.5492, "step": 41500 }, { "epoch": 0.68, "learning_rate": 3.868339368856699e-05, "loss": 1.5638, "step": 42000 }, { "epoch": 0.69, "learning_rate": 3.8548672184859454e-05, "loss": 1.5583, "step": 42500 }, { "epoch": 0.7, "learning_rate": 3.8413950681151924e-05, "loss": 1.5424, "step": 43000 }, { "epoch": 0.7, "learning_rate": 3.827922917744439e-05, "loss": 1.5306, "step": 43500 }, { "epoch": 0.71, "learning_rate": 3.814450767373685e-05, "loss": 1.54, "step": 44000 }, { "epoch": 0.72, "learning_rate": 3.800978617002931e-05, "loss": 1.5459, "step": 44500 }, { "epoch": 0.73, "learning_rate": 3.787506466632178e-05, "loss": 1.5286, "step": 45000 }, { "epoch": 0.74, "learning_rate": 3.7740343162614245e-05, "loss": 1.5224, "step": 45500 }, { "epoch": 0.74, "learning_rate": 3.760562165890671e-05, "loss": 1.5254, "step": 46000 }, { "epoch": 0.75, "learning_rate": 3.747090015519917e-05, "loss": 1.5088, "step": 46500 }, { "epoch": 0.76, "learning_rate": 3.733617865149164e-05, "loss": 1.5258, "step": 47000 }, { "epoch": 0.77, "learning_rate": 3.7201457147784103e-05, "loss": 1.5163, "step": 47500 }, { "epoch": 0.78, "learning_rate": 3.7066735644076566e-05, "loss": 1.5449, "step": 48000 }, { "epoch": 0.78, "learning_rate": 3.693201414036903e-05, "loss": 1.5275, "step": 48500 }, { "epoch": 0.79, "learning_rate": 3.67972926366615e-05, "loss": 1.5012, "step": 49000 }, { "epoch": 0.8, "learning_rate": 3.666257113295396e-05, "loss": 1.4904, "step": 49500 }, { "epoch": 0.81, "learning_rate": 3.6527849629246425e-05, "loss": 1.5122, "step": 50000 }, { "epoch": 0.82, "learning_rate": 3.639312812553889e-05, "loss": 1.4791, "step": 50500 }, { "epoch": 0.82, "learning_rate": 3.625840662183135e-05, "loss": 1.5132, "step": 51000 }, { "epoch": 0.83, "learning_rate": 3.612368511812382e-05, "loss": 1.5025, "step": 51500 }, { "epoch": 0.84, "learning_rate": 3.598896361441628e-05, "loss": 1.5006, "step": 52000 }, { "epoch": 0.85, "learning_rate": 3.5854242110708746e-05, "loss": 1.4988, "step": 52500 }, { "epoch": 0.86, "learning_rate": 3.571952060700121e-05, "loss": 1.4811, "step": 53000 }, { "epoch": 0.86, "learning_rate": 3.558479910329368e-05, "loss": 1.4799, "step": 53500 }, { "epoch": 0.87, "learning_rate": 3.5450077599586134e-05, "loss": 1.4885, "step": 54000 }, { "epoch": 0.88, "learning_rate": 3.53153560958786e-05, "loss": 1.4879, "step": 54500 }, { "epoch": 0.89, "learning_rate": 3.518063459217106e-05, "loss": 1.4895, "step": 55000 }, { "epoch": 0.9, "learning_rate": 3.504591308846353e-05, "loss": 1.4904, "step": 55500 }, { "epoch": 0.91, "learning_rate": 3.491119158475599e-05, "loss": 1.4779, "step": 56000 }, { "epoch": 0.91, "learning_rate": 3.4776470081048455e-05, "loss": 1.4698, "step": 56500 }, { "epoch": 0.92, "learning_rate": 3.464174857734092e-05, "loss": 1.4857, "step": 57000 }, { "epoch": 0.93, "learning_rate": 3.450702707363339e-05, "loss": 1.4953, "step": 57500 }, { "epoch": 0.94, "learning_rate": 3.437230556992585e-05, "loss": 1.4564, "step": 58000 }, { "epoch": 0.95, "learning_rate": 3.4237584066218314e-05, "loss": 1.4701, "step": 58500 }, { "epoch": 0.95, "learning_rate": 3.4102862562510777e-05, "loss": 1.4557, "step": 59000 }, { "epoch": 0.96, "learning_rate": 3.3968141058803246e-05, "loss": 1.4636, "step": 59500 }, { "epoch": 0.97, "learning_rate": 3.383341955509571e-05, "loss": 1.4657, "step": 60000 }, { "epoch": 0.98, "learning_rate": 3.369869805138817e-05, "loss": 1.4824, "step": 60500 }, { "epoch": 0.99, "learning_rate": 3.3563976547680635e-05, "loss": 1.4775, "step": 61000 }, { "epoch": 0.99, "learning_rate": 3.3429255043973104e-05, "loss": 1.4787, "step": 61500 }, { "epoch": 1.0, "learning_rate": 3.329453354026557e-05, "loss": 1.44, "step": 62000 }, { "epoch": 1.01, "learning_rate": 3.315981203655803e-05, "loss": 1.3586, "step": 62500 }, { "epoch": 1.02, "learning_rate": 3.302509053285049e-05, "loss": 1.3462, "step": 63000 }, { "epoch": 1.03, "learning_rate": 3.2890369029142956e-05, "loss": 1.3497, "step": 63500 }, { "epoch": 1.03, "learning_rate": 3.2755647525435426e-05, "loss": 1.3538, "step": 64000 }, { "epoch": 1.04, "learning_rate": 3.262092602172789e-05, "loss": 1.3349, "step": 64500 }, { "epoch": 1.05, "learning_rate": 3.248620451802035e-05, "loss": 1.3385, "step": 65000 }, { "epoch": 1.06, "learning_rate": 3.2351483014312814e-05, "loss": 1.3367, "step": 65500 }, { "epoch": 1.07, "learning_rate": 3.221676151060528e-05, "loss": 1.3427, "step": 66000 }, { "epoch": 1.08, "learning_rate": 3.208204000689774e-05, "loss": 1.365, "step": 66500 }, { "epoch": 1.08, "learning_rate": 3.19473185031902e-05, "loss": 1.3405, "step": 67000 }, { "epoch": 1.09, "learning_rate": 3.1812596999482666e-05, "loss": 1.3579, "step": 67500 }, { "epoch": 1.1, "learning_rate": 3.1677875495775135e-05, "loss": 1.3267, "step": 68000 }, { "epoch": 1.11, "learning_rate": 3.15431539920676e-05, "loss": 1.3524, "step": 68500 }, { "epoch": 1.12, "learning_rate": 3.140843248836006e-05, "loss": 1.3525, "step": 69000 }, { "epoch": 1.12, "learning_rate": 3.1273710984652524e-05, "loss": 1.3681, "step": 69500 }, { "epoch": 1.13, "learning_rate": 3.1138989480944993e-05, "loss": 1.346, "step": 70000 }, { "epoch": 1.14, "learning_rate": 3.1004267977237456e-05, "loss": 1.3289, "step": 70500 }, { "epoch": 1.15, "learning_rate": 3.086954647352992e-05, "loss": 1.3253, "step": 71000 }, { "epoch": 1.16, "learning_rate": 3.073482496982238e-05, "loss": 1.3319, "step": 71500 }, { "epoch": 1.16, "learning_rate": 3.060010346611485e-05, "loss": 1.3534, "step": 72000 }, { "epoch": 1.17, "learning_rate": 3.0465381962407315e-05, "loss": 1.3415, "step": 72500 }, { "epoch": 1.18, "learning_rate": 3.0330660458699777e-05, "loss": 1.3304, "step": 73000 }, { "epoch": 1.19, "learning_rate": 3.0195938954992244e-05, "loss": 1.3193, "step": 73500 }, { "epoch": 1.2, "learning_rate": 3.0061217451284707e-05, "loss": 1.3279, "step": 74000 }, { "epoch": 1.2, "learning_rate": 2.992649594757717e-05, "loss": 1.3123, "step": 74500 }, { "epoch": 1.21, "learning_rate": 2.9791774443869636e-05, "loss": 1.3458, "step": 75000 }, { "epoch": 1.22, "learning_rate": 2.96570529401621e-05, "loss": 1.3192, "step": 75500 }, { "epoch": 1.23, "learning_rate": 2.9522331436454565e-05, "loss": 1.3382, "step": 76000 }, { "epoch": 1.24, "learning_rate": 2.9387609932747028e-05, "loss": 1.339, "step": 76500 }, { "epoch": 1.24, "learning_rate": 2.9252888429039494e-05, "loss": 1.3377, "step": 77000 }, { "epoch": 1.25, "learning_rate": 2.9118166925331957e-05, "loss": 1.3288, "step": 77500 }, { "epoch": 1.26, "learning_rate": 2.8983445421624416e-05, "loss": 1.3381, "step": 78000 }, { "epoch": 1.27, "learning_rate": 2.8848723917916883e-05, "loss": 1.3094, "step": 78500 }, { "epoch": 1.28, "learning_rate": 2.8714002414209345e-05, "loss": 1.3461, "step": 79000 }, { "epoch": 1.29, "learning_rate": 2.857928091050181e-05, "loss": 1.3408, "step": 79500 }, { "epoch": 1.29, "learning_rate": 2.8444559406794275e-05, "loss": 1.3144, "step": 80000 }, { "epoch": 1.3, "learning_rate": 2.8309837903086737e-05, "loss": 1.3159, "step": 80500 }, { "epoch": 1.31, "learning_rate": 2.8175116399379204e-05, "loss": 1.3294, "step": 81000 }, { "epoch": 1.32, "learning_rate": 2.8040394895671667e-05, "loss": 1.339, "step": 81500 }, { "epoch": 1.33, "learning_rate": 2.7905673391964133e-05, "loss": 1.3309, "step": 82000 }, { "epoch": 1.33, "learning_rate": 2.7770951888256596e-05, "loss": 1.3128, "step": 82500 }, { "epoch": 1.34, "learning_rate": 2.7636230384549062e-05, "loss": 1.3212, "step": 83000 }, { "epoch": 1.35, "learning_rate": 2.7501508880841525e-05, "loss": 1.3325, "step": 83500 }, { "epoch": 1.36, "learning_rate": 2.736678737713399e-05, "loss": 1.3166, "step": 84000 }, { "epoch": 1.37, "learning_rate": 2.7232065873426454e-05, "loss": 1.3058, "step": 84500 }, { "epoch": 1.37, "learning_rate": 2.709734436971892e-05, "loss": 1.3133, "step": 85000 }, { "epoch": 1.38, "learning_rate": 2.6962622866011383e-05, "loss": 1.3134, "step": 85500 }, { "epoch": 1.39, "learning_rate": 2.682790136230385e-05, "loss": 1.321, "step": 86000 }, { "epoch": 1.4, "learning_rate": 2.6693179858596312e-05, "loss": 1.3194, "step": 86500 }, { "epoch": 1.41, "learning_rate": 2.6558458354888775e-05, "loss": 1.2939, "step": 87000 }, { "epoch": 1.41, "learning_rate": 2.642373685118124e-05, "loss": 1.3027, "step": 87500 }, { "epoch": 1.42, "learning_rate": 2.6289015347473704e-05, "loss": 1.3181, "step": 88000 }, { "epoch": 1.43, "learning_rate": 2.615429384376617e-05, "loss": 1.3341, "step": 88500 }, { "epoch": 1.44, "learning_rate": 2.6019572340058633e-05, "loss": 1.3339, "step": 89000 }, { "epoch": 1.45, "learning_rate": 2.58848508363511e-05, "loss": 1.3159, "step": 89500 }, { "epoch": 1.45, "learning_rate": 2.575012933264356e-05, "loss": 1.3121, "step": 90000 }, { "epoch": 1.46, "learning_rate": 2.5615407828936022e-05, "loss": 1.3297, "step": 90500 }, { "epoch": 1.47, "learning_rate": 2.5480686325228488e-05, "loss": 1.3174, "step": 91000 }, { "epoch": 1.48, "learning_rate": 2.534596482152095e-05, "loss": 1.3159, "step": 91500 }, { "epoch": 1.49, "learning_rate": 2.5211243317813417e-05, "loss": 1.3131, "step": 92000 }, { "epoch": 1.5, "learning_rate": 2.507652181410588e-05, "loss": 1.3147, "step": 92500 }, { "epoch": 1.5, "learning_rate": 2.4941800310398343e-05, "loss": 1.3002, "step": 93000 }, { "epoch": 1.51, "learning_rate": 2.480707880669081e-05, "loss": 1.3273, "step": 93500 }, { "epoch": 1.52, "learning_rate": 2.4672357302983272e-05, "loss": 1.3059, "step": 94000 }, { "epoch": 1.53, "learning_rate": 2.453763579927574e-05, "loss": 1.3093, "step": 94500 }, { "epoch": 1.54, "learning_rate": 2.44029142955682e-05, "loss": 1.3131, "step": 95000 }, { "epoch": 1.54, "learning_rate": 2.4268192791860667e-05, "loss": 1.3025, "step": 95500 }, { "epoch": 1.55, "learning_rate": 2.413347128815313e-05, "loss": 1.3011, "step": 96000 }, { "epoch": 1.56, "learning_rate": 2.3998749784445597e-05, "loss": 1.2979, "step": 96500 }, { "epoch": 1.57, "learning_rate": 2.386402828073806e-05, "loss": 1.3008, "step": 97000 }, { "epoch": 1.58, "learning_rate": 2.3729306777030526e-05, "loss": 1.3008, "step": 97500 }, { "epoch": 1.58, "learning_rate": 2.359458527332299e-05, "loss": 1.2968, "step": 98000 }, { "epoch": 1.59, "learning_rate": 2.3459863769615455e-05, "loss": 1.2901, "step": 98500 }, { "epoch": 1.6, "learning_rate": 2.3325142265907914e-05, "loss": 1.2822, "step": 99000 }, { "epoch": 1.61, "learning_rate": 2.319042076220038e-05, "loss": 1.2963, "step": 99500 }, { "epoch": 1.62, "learning_rate": 2.3055699258492843e-05, "loss": 1.2996, "step": 100000 }, { "epoch": 1.62, "learning_rate": 2.292097775478531e-05, "loss": 1.2809, "step": 100500 }, { "epoch": 1.63, "learning_rate": 2.2786256251077773e-05, "loss": 1.2998, "step": 101000 }, { "epoch": 1.64, "learning_rate": 2.265153474737024e-05, "loss": 1.2923, "step": 101500 }, { "epoch": 1.65, "learning_rate": 2.25168132436627e-05, "loss": 1.2957, "step": 102000 }, { "epoch": 1.66, "learning_rate": 2.2382091739955165e-05, "loss": 1.3018, "step": 102500 }, { "epoch": 1.67, "learning_rate": 2.224737023624763e-05, "loss": 1.2862, "step": 103000 }, { "epoch": 1.67, "learning_rate": 2.2112648732540094e-05, "loss": 1.286, "step": 103500 }, { "epoch": 1.68, "learning_rate": 2.197792722883256e-05, "loss": 1.2845, "step": 104000 }, { "epoch": 1.69, "learning_rate": 2.1843205725125023e-05, "loss": 1.2779, "step": 104500 }, { "epoch": 1.7, "learning_rate": 2.1708484221417486e-05, "loss": 1.2727, "step": 105000 }, { "epoch": 1.71, "learning_rate": 2.157376271770995e-05, "loss": 1.2727, "step": 105500 }, { "epoch": 1.71, "learning_rate": 2.1439041214002415e-05, "loss": 1.2889, "step": 106000 }, { "epoch": 1.72, "learning_rate": 2.1304319710294878e-05, "loss": 1.2797, "step": 106500 }, { "epoch": 1.73, "learning_rate": 2.1169598206587344e-05, "loss": 1.2674, "step": 107000 }, { "epoch": 1.74, "learning_rate": 2.1034876702879807e-05, "loss": 1.2805, "step": 107500 }, { "epoch": 1.75, "learning_rate": 2.0900155199172273e-05, "loss": 1.279, "step": 108000 }, { "epoch": 1.75, "learning_rate": 2.0765433695464736e-05, "loss": 1.297, "step": 108500 }, { "epoch": 1.76, "learning_rate": 2.0630712191757202e-05, "loss": 1.2728, "step": 109000 }, { "epoch": 1.77, "learning_rate": 2.0495990688049665e-05, "loss": 1.2926, "step": 109500 }, { "epoch": 1.78, "learning_rate": 2.036126918434213e-05, "loss": 1.2604, "step": 110000 }, { "epoch": 1.79, "learning_rate": 2.0226547680634594e-05, "loss": 1.2553, "step": 110500 }, { "epoch": 1.79, "learning_rate": 2.0091826176927057e-05, "loss": 1.2629, "step": 111000 }, { "epoch": 1.8, "learning_rate": 1.995710467321952e-05, "loss": 1.2653, "step": 111500 }, { "epoch": 1.81, "learning_rate": 1.9822383169511986e-05, "loss": 1.2933, "step": 112000 }, { "epoch": 1.82, "learning_rate": 1.968766166580445e-05, "loss": 1.2675, "step": 112500 }, { "epoch": 1.83, "learning_rate": 1.9552940162096915e-05, "loss": 1.277, "step": 113000 }, { "epoch": 1.83, "learning_rate": 1.9418218658389378e-05, "loss": 1.2757, "step": 113500 }, { "epoch": 1.84, "learning_rate": 1.9283497154681844e-05, "loss": 1.2602, "step": 114000 }, { "epoch": 1.85, "learning_rate": 1.9148775650974307e-05, "loss": 1.2672, "step": 114500 }, { "epoch": 1.86, "learning_rate": 1.901405414726677e-05, "loss": 1.2656, "step": 115000 }, { "epoch": 1.87, "learning_rate": 1.8879332643559236e-05, "loss": 1.2583, "step": 115500 }, { "epoch": 1.88, "learning_rate": 1.87446111398517e-05, "loss": 1.2644, "step": 116000 }, { "epoch": 1.88, "learning_rate": 1.8609889636144165e-05, "loss": 1.267, "step": 116500 }, { "epoch": 1.89, "learning_rate": 1.8475168132436625e-05, "loss": 1.2745, "step": 117000 }, { "epoch": 1.9, "learning_rate": 1.834044662872909e-05, "loss": 1.2545, "step": 117500 }, { "epoch": 1.91, "learning_rate": 1.8205725125021554e-05, "loss": 1.2854, "step": 118000 }, { "epoch": 1.92, "learning_rate": 1.807100362131402e-05, "loss": 1.2675, "step": 118500 }, { "epoch": 1.92, "learning_rate": 1.7936282117606483e-05, "loss": 1.2607, "step": 119000 }, { "epoch": 1.93, "learning_rate": 1.780156061389895e-05, "loss": 1.2594, "step": 119500 }, { "epoch": 1.94, "learning_rate": 1.7666839110191412e-05, "loss": 1.257, "step": 120000 }, { "epoch": 1.95, "learning_rate": 1.753211760648388e-05, "loss": 1.2561, "step": 120500 }, { "epoch": 1.96, "learning_rate": 1.739739610277634e-05, "loss": 1.2429, "step": 121000 }, { "epoch": 1.96, "learning_rate": 1.7262674599068808e-05, "loss": 1.2684, "step": 121500 }, { "epoch": 1.97, "learning_rate": 1.712795309536127e-05, "loss": 1.2395, "step": 122000 }, { "epoch": 1.98, "learning_rate": 1.6993231591653737e-05, "loss": 1.2422, "step": 122500 }, { "epoch": 1.99, "learning_rate": 1.6858510087946196e-05, "loss": 1.2573, "step": 123000 }, { "epoch": 2.0, "learning_rate": 1.6723788584238663e-05, "loss": 1.2511, "step": 123500 }, { "epoch": 2.0, "learning_rate": 1.6589067080531125e-05, "loss": 1.2102, "step": 124000 }, { "epoch": 2.01, "learning_rate": 1.645434557682359e-05, "loss": 1.1356, "step": 124500 }, { "epoch": 2.02, "learning_rate": 1.6319624073116054e-05, "loss": 1.13, "step": 125000 }, { "epoch": 2.03, "learning_rate": 1.618490256940852e-05, "loss": 1.1385, "step": 125500 }, { "epoch": 2.04, "learning_rate": 1.6050181065700984e-05, "loss": 1.1282, "step": 126000 }, { "epoch": 2.05, "learning_rate": 1.591545956199345e-05, "loss": 1.1088, "step": 126500 }, { "epoch": 2.05, "learning_rate": 1.5780738058285913e-05, "loss": 1.1284, "step": 127000 }, { "epoch": 2.06, "learning_rate": 1.5646016554578376e-05, "loss": 1.1302, "step": 127500 }, { "epoch": 2.07, "learning_rate": 1.5511295050870842e-05, "loss": 1.1217, "step": 128000 }, { "epoch": 2.08, "learning_rate": 1.5376573547163305e-05, "loss": 1.139, "step": 128500 }, { "epoch": 2.09, "learning_rate": 1.5241852043455768e-05, "loss": 1.1412, "step": 129000 }, { "epoch": 2.09, "learning_rate": 1.5107130539748232e-05, "loss": 1.124, "step": 129500 }, { "epoch": 2.1, "learning_rate": 1.4972409036040697e-05, "loss": 1.152, "step": 130000 }, { "epoch": 2.11, "learning_rate": 1.4837687532333161e-05, "loss": 1.1365, "step": 130500 }, { "epoch": 2.12, "learning_rate": 1.4702966028625626e-05, "loss": 1.1471, "step": 131000 }, { "epoch": 2.13, "learning_rate": 1.456824452491809e-05, "loss": 1.1427, "step": 131500 }, { "epoch": 2.13, "learning_rate": 1.4433523021210555e-05, "loss": 1.1344, "step": 132000 }, { "epoch": 2.14, "learning_rate": 1.429880151750302e-05, "loss": 1.1466, "step": 132500 }, { "epoch": 2.15, "learning_rate": 1.4164080013795482e-05, "loss": 1.1384, "step": 133000 }, { "epoch": 2.16, "learning_rate": 1.4029358510087947e-05, "loss": 1.1471, "step": 133500 }, { "epoch": 2.17, "learning_rate": 1.3894637006380412e-05, "loss": 1.1282, "step": 134000 }, { "epoch": 2.17, "learning_rate": 1.3759915502672876e-05, "loss": 1.1445, "step": 134500 }, { "epoch": 2.18, "learning_rate": 1.3625193998965339e-05, "loss": 1.1296, "step": 135000 }, { "epoch": 2.19, "learning_rate": 1.3490472495257803e-05, "loss": 1.1235, "step": 135500 }, { "epoch": 2.2, "learning_rate": 1.3355750991550266e-05, "loss": 1.1414, "step": 136000 }, { "epoch": 2.21, "learning_rate": 1.3221029487842731e-05, "loss": 1.135, "step": 136500 }, { "epoch": 2.21, "learning_rate": 1.3086307984135195e-05, "loss": 1.1399, "step": 137000 }, { "epoch": 2.22, "learning_rate": 1.295158648042766e-05, "loss": 1.1459, "step": 137500 }, { "epoch": 2.23, "learning_rate": 1.2816864976720125e-05, "loss": 1.1153, "step": 138000 }, { "epoch": 2.24, "learning_rate": 1.268214347301259e-05, "loss": 1.1347, "step": 138500 }, { "epoch": 2.25, "learning_rate": 1.2547421969305054e-05, "loss": 1.1399, "step": 139000 }, { "epoch": 2.26, "learning_rate": 1.2412700465597517e-05, "loss": 1.1322, "step": 139500 }, { "epoch": 2.26, "learning_rate": 1.2277978961889981e-05, "loss": 1.1387, "step": 140000 }, { "epoch": 2.27, "learning_rate": 1.2143257458182446e-05, "loss": 1.1281, "step": 140500 }, { "epoch": 2.28, "learning_rate": 1.200853595447491e-05, "loss": 1.1289, "step": 141000 }, { "epoch": 2.29, "learning_rate": 1.1873814450767375e-05, "loss": 1.1302, "step": 141500 }, { "epoch": 2.3, "learning_rate": 1.173909294705984e-05, "loss": 1.1282, "step": 142000 }, { "epoch": 2.3, "learning_rate": 1.1604371443352302e-05, "loss": 1.1274, "step": 142500 }, { "epoch": 2.31, "learning_rate": 1.1469649939644767e-05, "loss": 1.1256, "step": 143000 }, { "epoch": 2.32, "learning_rate": 1.1334928435937231e-05, "loss": 1.1255, "step": 143500 }, { "epoch": 2.33, "learning_rate": 1.1200206932229696e-05, "loss": 1.1301, "step": 144000 }, { "epoch": 2.34, "learning_rate": 1.106548542852216e-05, "loss": 1.1407, "step": 144500 }, { "epoch": 2.34, "learning_rate": 1.0930763924814625e-05, "loss": 1.1118, "step": 145000 }, { "epoch": 2.35, "learning_rate": 1.0796042421107088e-05, "loss": 1.1112, "step": 145500 }, { "epoch": 2.36, "learning_rate": 1.0661320917399552e-05, "loss": 1.1235, "step": 146000 }, { "epoch": 2.37, "learning_rate": 1.0526599413692017e-05, "loss": 1.1208, "step": 146500 }, { "epoch": 2.38, "learning_rate": 1.039187790998448e-05, "loss": 1.1206, "step": 147000 }, { "epoch": 2.38, "learning_rate": 1.0257156406276944e-05, "loss": 1.1557, "step": 147500 }, { "epoch": 2.39, "learning_rate": 1.0122434902569409e-05, "loss": 1.1201, "step": 148000 }, { "epoch": 2.4, "learning_rate": 9.987713398861872e-06, "loss": 1.1178, "step": 148500 }, { "epoch": 2.41, "learning_rate": 9.852991895154336e-06, "loss": 1.1284, "step": 149000 }, { "epoch": 2.42, "learning_rate": 9.718270391446801e-06, "loss": 1.1253, "step": 149500 }, { "epoch": 2.42, "learning_rate": 9.583548887739266e-06, "loss": 1.1089, "step": 150000 }, { "epoch": 2.43, "learning_rate": 9.44882738403173e-06, "loss": 1.1147, "step": 150500 }, { "epoch": 2.44, "learning_rate": 9.314105880324195e-06, "loss": 1.1206, "step": 151000 }, { "epoch": 2.45, "learning_rate": 9.179384376616658e-06, "loss": 1.1185, "step": 151500 }, { "epoch": 2.46, "learning_rate": 9.044662872909122e-06, "loss": 1.1251, "step": 152000 }, { "epoch": 2.47, "learning_rate": 8.909941369201587e-06, "loss": 1.1377, "step": 152500 }, { "epoch": 2.47, "learning_rate": 8.775219865494051e-06, "loss": 1.1089, "step": 153000 }, { "epoch": 2.48, "learning_rate": 8.640498361786516e-06, "loss": 1.1066, "step": 153500 }, { "epoch": 2.49, "learning_rate": 8.50577685807898e-06, "loss": 1.0977, "step": 154000 }, { "epoch": 2.5, "learning_rate": 8.371055354371443e-06, "loss": 1.1085, "step": 154500 }, { "epoch": 2.51, "learning_rate": 8.236333850663908e-06, "loss": 1.089, "step": 155000 }, { "epoch": 2.51, "learning_rate": 8.101612346956372e-06, "loss": 1.1056, "step": 155500 }, { "epoch": 2.52, "learning_rate": 7.966890843248837e-06, "loss": 1.1179, "step": 156000 }, { "epoch": 2.53, "learning_rate": 7.832169339541301e-06, "loss": 1.1087, "step": 156500 }, { "epoch": 2.54, "learning_rate": 7.697447835833766e-06, "loss": 1.1138, "step": 157000 }, { "epoch": 2.55, "learning_rate": 7.562726332126228e-06, "loss": 1.1195, "step": 157500 }, { "epoch": 2.55, "learning_rate": 7.428004828418693e-06, "loss": 1.1038, "step": 158000 }, { "epoch": 2.56, "learning_rate": 7.293283324711157e-06, "loss": 1.1086, "step": 158500 }, { "epoch": 2.57, "learning_rate": 7.158561821003622e-06, "loss": 1.109, "step": 159000 }, { "epoch": 2.58, "learning_rate": 7.023840317296086e-06, "loss": 1.1032, "step": 159500 }, { "epoch": 2.59, "learning_rate": 6.889118813588551e-06, "loss": 1.0946, "step": 160000 }, { "epoch": 2.59, "learning_rate": 6.754397309881014e-06, "loss": 1.1043, "step": 160500 }, { "epoch": 2.6, "learning_rate": 6.619675806173478e-06, "loss": 1.1108, "step": 161000 }, { "epoch": 2.61, "learning_rate": 6.484954302465943e-06, "loss": 1.1108, "step": 161500 }, { "epoch": 2.62, "learning_rate": 6.3502327987584074e-06, "loss": 1.1178, "step": 162000 }, { "epoch": 2.63, "learning_rate": 6.215511295050871e-06, "loss": 1.0908, "step": 162500 }, { "epoch": 2.64, "learning_rate": 6.080789791343335e-06, "loss": 1.0962, "step": 163000 }, { "epoch": 2.64, "learning_rate": 5.946068287635799e-06, "loss": 1.131, "step": 163500 }, { "epoch": 2.65, "learning_rate": 5.811346783928263e-06, "loss": 1.1251, "step": 164000 }, { "epoch": 2.66, "learning_rate": 5.676625280220728e-06, "loss": 1.1053, "step": 164500 }, { "epoch": 2.67, "learning_rate": 5.541903776513192e-06, "loss": 1.1148, "step": 165000 }, { "epoch": 2.68, "learning_rate": 5.407182272805656e-06, "loss": 1.1035, "step": 165500 }, { "epoch": 2.68, "learning_rate": 5.2724607690981205e-06, "loss": 1.097, "step": 166000 }, { "epoch": 2.69, "learning_rate": 5.137739265390585e-06, "loss": 1.1098, "step": 166500 }, { "epoch": 2.7, "learning_rate": 5.003017761683049e-06, "loss": 1.0914, "step": 167000 }, { "epoch": 2.71, "learning_rate": 4.868296257975513e-06, "loss": 1.095, "step": 167500 }, { "epoch": 2.72, "learning_rate": 4.733574754267978e-06, "loss": 1.1036, "step": 168000 }, { "epoch": 2.72, "learning_rate": 4.598853250560442e-06, "loss": 1.1007, "step": 168500 }, { "epoch": 2.73, "learning_rate": 4.464131746852906e-06, "loss": 1.1141, "step": 169000 }, { "epoch": 2.74, "learning_rate": 4.32941024314537e-06, "loss": 1.1052, "step": 169500 }, { "epoch": 2.75, "learning_rate": 4.194688739437834e-06, "loss": 1.1109, "step": 170000 }, { "epoch": 2.76, "learning_rate": 4.059967235730298e-06, "loss": 1.0832, "step": 170500 }, { "epoch": 2.76, "learning_rate": 3.925245732022763e-06, "loss": 1.1108, "step": 171000 }, { "epoch": 2.77, "learning_rate": 3.7905242283152265e-06, "loss": 1.0982, "step": 171500 }, { "epoch": 2.78, "learning_rate": 3.655802724607691e-06, "loss": 1.1108, "step": 172000 }, { "epoch": 2.79, "learning_rate": 3.5210812209001556e-06, "loss": 1.1056, "step": 172500 }, { "epoch": 2.8, "learning_rate": 3.3863597171926193e-06, "loss": 1.102, "step": 173000 }, { "epoch": 2.8, "learning_rate": 3.251638213485084e-06, "loss": 1.0955, "step": 173500 }, { "epoch": 2.81, "learning_rate": 3.116916709777548e-06, "loss": 1.0922, "step": 174000 }, { "epoch": 2.82, "learning_rate": 2.982195206070012e-06, "loss": 1.1075, "step": 174500 }, { "epoch": 2.83, "learning_rate": 2.8474737023624763e-06, "loss": 1.0985, "step": 175000 }, { "epoch": 2.84, "learning_rate": 2.712752198654941e-06, "loss": 1.0959, "step": 175500 }, { "epoch": 2.85, "learning_rate": 2.578030694947405e-06, "loss": 1.1047, "step": 176000 }, { "epoch": 2.85, "learning_rate": 2.443309191239869e-06, "loss": 1.101, "step": 176500 }, { "epoch": 2.86, "learning_rate": 2.3085876875323332e-06, "loss": 1.0976, "step": 177000 }, { "epoch": 2.87, "learning_rate": 2.1738661838247974e-06, "loss": 1.0841, "step": 177500 }, { "epoch": 2.88, "learning_rate": 2.0391446801172615e-06, "loss": 1.0921, "step": 178000 }, { "epoch": 2.89, "learning_rate": 1.904423176409726e-06, "loss": 1.0912, "step": 178500 }, { "epoch": 2.89, "learning_rate": 1.7697016727021902e-06, "loss": 1.0973, "step": 179000 }, { "epoch": 2.9, "learning_rate": 1.6349801689946541e-06, "loss": 1.0728, "step": 179500 }, { "epoch": 2.91, "learning_rate": 1.5002586652871185e-06, "loss": 1.0795, "step": 180000 }, { "epoch": 2.92, "learning_rate": 1.3655371615795828e-06, "loss": 1.1133, "step": 180500 }, { "epoch": 2.93, "learning_rate": 1.230815657872047e-06, "loss": 1.0782, "step": 181000 }, { "epoch": 2.93, "learning_rate": 1.0960941541645111e-06, "loss": 1.0806, "step": 181500 }, { "epoch": 2.94, "learning_rate": 9.613726504569755e-07, "loss": 1.0963, "step": 182000 }, { "epoch": 2.95, "learning_rate": 8.266511467494397e-07, "loss": 1.0789, "step": 182500 }, { "epoch": 2.96, "learning_rate": 6.919296430419037e-07, "loss": 1.1024, "step": 183000 }, { "epoch": 2.97, "learning_rate": 5.572081393343681e-07, "loss": 1.0869, "step": 183500 }, { "epoch": 2.97, "learning_rate": 4.224866356268322e-07, "loss": 1.1068, "step": 184000 }, { "epoch": 2.98, "learning_rate": 2.8776513191929646e-07, "loss": 1.0869, "step": 184500 }, { "epoch": 2.99, "learning_rate": 1.5304362821176065e-07, "loss": 1.0886, "step": 185000 }, { "epoch": 3.0, "learning_rate": 1.8322124504224867e-08, "loss": 1.0814, "step": 185500 }, { "epoch": 3.0, "step": 185568, "total_flos": 4.0364234274584986e+17, "train_loss": 1.363925259613666, "train_runtime": 60480.0561, "train_samples_per_second": 30.682, "train_steps_per_second": 3.068 } ], "max_steps": 185568, "num_train_epochs": 3, "total_flos": 4.0364234274584986e+17, "trial_name": null, "trial_params": null }