{ "best_metric": null, "best_model_checkpoint": null, "epoch": 7.0, "eval_steps": 500, "global_step": 93730, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.04, "learning_rate": 4.813293502613891e-05, "loss": 3.4729, "step": 500 }, { "epoch": 0.07, "learning_rate": 4.626587005227782e-05, "loss": 3.0263, "step": 1000 }, { "epoch": 0.11, "learning_rate": 4.439880507841673e-05, "loss": 2.8813, "step": 1500 }, { "epoch": 0.15, "learning_rate": 4.253174010455564e-05, "loss": 2.7034, "step": 2000 }, { "epoch": 0.19, "learning_rate": 4.066467513069455e-05, "loss": 2.7308, "step": 2500 }, { "epoch": 0.22, "learning_rate": 3.879761015683346e-05, "loss": 2.6246, "step": 3000 }, { "epoch": 0.26, "learning_rate": 3.693054518297237e-05, "loss": 2.5404, "step": 3500 }, { "epoch": 0.3, "learning_rate": 3.506348020911128e-05, "loss": 2.543, "step": 4000 }, { "epoch": 0.34, "learning_rate": 3.319641523525019e-05, "loss": 2.4337, "step": 4500 }, { "epoch": 0.37, "learning_rate": 3.13293502613891e-05, "loss": 2.5003, "step": 5000 }, { "epoch": 0.41, "learning_rate": 2.9462285287528007e-05, "loss": 2.4064, "step": 5500 }, { "epoch": 0.45, "learning_rate": 2.759522031366692e-05, "loss": 2.3045, "step": 6000 }, { "epoch": 0.49, "learning_rate": 2.5728155339805826e-05, "loss": 2.3047, "step": 6500 }, { "epoch": 0.52, "learning_rate": 2.3861090365944735e-05, "loss": 2.2855, "step": 7000 }, { "epoch": 0.56, "learning_rate": 2.1994025392083645e-05, "loss": 2.3088, "step": 7500 }, { "epoch": 0.6, "learning_rate": 2.0126960418222554e-05, "loss": 2.2434, "step": 8000 }, { "epoch": 0.63, "learning_rate": 1.8259895444361464e-05, "loss": 2.215, "step": 8500 }, { "epoch": 0.67, "learning_rate": 1.6392830470500377e-05, "loss": 2.2731, "step": 9000 }, { "epoch": 0.71, "learning_rate": 1.4525765496639285e-05, "loss": 2.1623, "step": 9500 }, { "epoch": 0.75, "learning_rate": 1.2658700522778194e-05, "loss": 2.1395, "step": 10000 }, { "epoch": 0.78, "learning_rate": 1.0791635548917104e-05, "loss": 2.2036, "step": 10500 }, { "epoch": 0.82, "learning_rate": 8.924570575056013e-06, "loss": 2.1406, "step": 11000 }, { "epoch": 0.86, "learning_rate": 7.057505601194922e-06, "loss": 2.0425, "step": 11500 }, { "epoch": 0.9, "learning_rate": 5.190440627333831e-06, "loss": 2.0832, "step": 12000 }, { "epoch": 0.93, "learning_rate": 3.3233756534727413e-06, "loss": 2.0409, "step": 12500 }, { "epoch": 0.97, "learning_rate": 1.4563106796116506e-06, "loss": 2.0602, "step": 13000 }, { "epoch": 1.01, "learning_rate": 3.991784914115011e-05, "loss": 2.1169, "step": 13500 }, { "epoch": 1.05, "learning_rate": 3.95444361463779e-05, "loss": 1.9384, "step": 14000 }, { "epoch": 1.08, "learning_rate": 3.917102315160568e-05, "loss": 1.9064, "step": 14500 }, { "epoch": 1.12, "learning_rate": 3.879761015683346e-05, "loss": 1.9263, "step": 15000 }, { "epoch": 1.16, "learning_rate": 3.8424197162061244e-05, "loss": 1.984, "step": 15500 }, { "epoch": 1.19, "learning_rate": 3.805078416728902e-05, "loss": 1.9026, "step": 16000 }, { "epoch": 1.23, "learning_rate": 3.7677371172516806e-05, "loss": 1.9688, "step": 16500 }, { "epoch": 1.27, "learning_rate": 3.7303958177744584e-05, "loss": 1.9084, "step": 17000 }, { "epoch": 1.31, "learning_rate": 3.693054518297237e-05, "loss": 1.9521, "step": 17500 }, { "epoch": 1.34, "learning_rate": 3.655713218820015e-05, "loss": 1.9084, "step": 18000 }, { "epoch": 1.38, "learning_rate": 3.618371919342793e-05, "loss": 1.957, "step": 18500 }, { "epoch": 1.42, "learning_rate": 3.5810306198655716e-05, "loss": 1.9216, "step": 19000 }, { "epoch": 1.46, "learning_rate": 3.54368932038835e-05, "loss": 1.9009, "step": 19500 }, { "epoch": 1.49, "learning_rate": 3.506348020911128e-05, "loss": 1.9264, "step": 20000 }, { "epoch": 1.53, "learning_rate": 3.469006721433906e-05, "loss": 1.9279, "step": 20500 }, { "epoch": 1.57, "learning_rate": 3.431665421956684e-05, "loss": 1.8553, "step": 21000 }, { "epoch": 1.61, "learning_rate": 3.3943241224794625e-05, "loss": 1.9544, "step": 21500 }, { "epoch": 1.64, "learning_rate": 3.35698282300224e-05, "loss": 1.8713, "step": 22000 }, { "epoch": 1.68, "learning_rate": 3.319641523525019e-05, "loss": 1.877, "step": 22500 }, { "epoch": 1.72, "learning_rate": 3.2823002240477966e-05, "loss": 1.8483, "step": 23000 }, { "epoch": 1.76, "learning_rate": 3.244958924570575e-05, "loss": 1.8154, "step": 23500 }, { "epoch": 1.79, "learning_rate": 3.2076176250933535e-05, "loss": 1.897, "step": 24000 }, { "epoch": 1.83, "learning_rate": 3.170276325616132e-05, "loss": 1.8807, "step": 24500 }, { "epoch": 1.87, "learning_rate": 3.13293502613891e-05, "loss": 1.8696, "step": 25000 }, { "epoch": 1.9, "learning_rate": 3.095593726661688e-05, "loss": 1.8949, "step": 25500 }, { "epoch": 1.94, "learning_rate": 3.058252427184466e-05, "loss": 1.8808, "step": 26000 }, { "epoch": 1.98, "learning_rate": 3.0209111277072444e-05, "loss": 1.8257, "step": 26500 }, { "epoch": 2.02, "learning_rate": 2.9835698282300222e-05, "loss": 1.6733, "step": 27000 }, { "epoch": 2.05, "learning_rate": 2.9462285287528007e-05, "loss": 1.3824, "step": 27500 }, { "epoch": 2.09, "learning_rate": 2.9088872292755788e-05, "loss": 1.4268, "step": 28000 }, { "epoch": 2.13, "learning_rate": 2.8715459297983573e-05, "loss": 1.3661, "step": 28500 }, { "epoch": 2.17, "learning_rate": 2.834204630321135e-05, "loss": 1.3921, "step": 29000 }, { "epoch": 2.2, "learning_rate": 2.7968633308439135e-05, "loss": 1.381, "step": 29500 }, { "epoch": 2.24, "learning_rate": 2.759522031366692e-05, "loss": 1.4431, "step": 30000 }, { "epoch": 2.28, "learning_rate": 2.7221807318894698e-05, "loss": 1.4092, "step": 30500 }, { "epoch": 2.32, "learning_rate": 2.6848394324122482e-05, "loss": 1.3834, "step": 31000 }, { "epoch": 2.35, "learning_rate": 2.6474981329350264e-05, "loss": 1.3976, "step": 31500 }, { "epoch": 2.39, "learning_rate": 2.6101568334578048e-05, "loss": 1.4163, "step": 32000 }, { "epoch": 2.43, "learning_rate": 2.5728155339805826e-05, "loss": 1.463, "step": 32500 }, { "epoch": 2.46, "learning_rate": 2.535474234503361e-05, "loss": 1.3829, "step": 33000 }, { "epoch": 2.5, "learning_rate": 2.498132935026139e-05, "loss": 1.4267, "step": 33500 }, { "epoch": 2.54, "learning_rate": 2.4607916355489173e-05, "loss": 1.4251, "step": 34000 }, { "epoch": 2.58, "learning_rate": 2.4234503360716954e-05, "loss": 1.3971, "step": 34500 }, { "epoch": 2.61, "learning_rate": 2.3861090365944735e-05, "loss": 1.4307, "step": 35000 }, { "epoch": 2.65, "learning_rate": 2.3487677371172517e-05, "loss": 1.3797, "step": 35500 }, { "epoch": 2.69, "learning_rate": 2.3114264376400298e-05, "loss": 1.4136, "step": 36000 }, { "epoch": 2.73, "learning_rate": 2.2740851381628083e-05, "loss": 1.3923, "step": 36500 }, { "epoch": 2.76, "learning_rate": 2.2367438386855864e-05, "loss": 1.4402, "step": 37000 }, { "epoch": 2.8, "learning_rate": 2.1994025392083645e-05, "loss": 1.4485, "step": 37500 }, { "epoch": 2.84, "learning_rate": 2.1620612397311426e-05, "loss": 1.3766, "step": 38000 }, { "epoch": 2.88, "learning_rate": 2.1247199402539207e-05, "loss": 1.3877, "step": 38500 }, { "epoch": 2.91, "learning_rate": 2.0873786407766992e-05, "loss": 1.4218, "step": 39000 }, { "epoch": 2.95, "learning_rate": 2.0500373412994773e-05, "loss": 1.4167, "step": 39500 }, { "epoch": 2.99, "learning_rate": 2.0126960418222554e-05, "loss": 1.3621, "step": 40000 }, { "epoch": 3.02, "learning_rate": 1.9753547423450336e-05, "loss": 1.0586, "step": 40500 }, { "epoch": 3.06, "learning_rate": 1.9380134428678117e-05, "loss": 0.9882, "step": 41000 }, { "epoch": 3.1, "learning_rate": 1.9006721433905898e-05, "loss": 0.9795, "step": 41500 }, { "epoch": 3.14, "learning_rate": 1.8633308439133683e-05, "loss": 0.9992, "step": 42000 }, { "epoch": 3.17, "learning_rate": 1.8259895444361464e-05, "loss": 0.9474, "step": 42500 }, { "epoch": 3.21, "learning_rate": 1.7886482449589245e-05, "loss": 0.9975, "step": 43000 }, { "epoch": 3.25, "learning_rate": 1.7513069454817026e-05, "loss": 0.9606, "step": 43500 }, { "epoch": 3.29, "learning_rate": 1.713965646004481e-05, "loss": 0.9781, "step": 44000 }, { "epoch": 3.32, "learning_rate": 1.6766243465272592e-05, "loss": 1.0437, "step": 44500 }, { "epoch": 3.36, "learning_rate": 1.6392830470500377e-05, "loss": 0.963, "step": 45000 }, { "epoch": 3.4, "learning_rate": 1.6019417475728158e-05, "loss": 1.0407, "step": 45500 }, { "epoch": 3.44, "learning_rate": 1.564600448095594e-05, "loss": 1.0054, "step": 46000 }, { "epoch": 3.47, "learning_rate": 1.527259148618372e-05, "loss": 1.0396, "step": 46500 }, { "epoch": 3.51, "learning_rate": 1.4899178491411503e-05, "loss": 1.0051, "step": 47000 }, { "epoch": 3.55, "learning_rate": 1.4525765496639285e-05, "loss": 0.9828, "step": 47500 }, { "epoch": 3.58, "learning_rate": 1.4152352501867066e-05, "loss": 0.9976, "step": 48000 }, { "epoch": 3.62, "learning_rate": 1.3778939507094849e-05, "loss": 1.0251, "step": 48500 }, { "epoch": 3.66, "learning_rate": 1.340552651232263e-05, "loss": 1.0432, "step": 49000 }, { "epoch": 3.7, "learning_rate": 1.3032113517550413e-05, "loss": 0.9727, "step": 49500 }, { "epoch": 3.73, "learning_rate": 1.2658700522778194e-05, "loss": 0.9822, "step": 50000 }, { "epoch": 3.77, "learning_rate": 1.2285287528005974e-05, "loss": 1.0392, "step": 50500 }, { "epoch": 3.81, "learning_rate": 1.1911874533233758e-05, "loss": 0.9943, "step": 51000 }, { "epoch": 3.85, "learning_rate": 1.153846153846154e-05, "loss": 1.0256, "step": 51500 }, { "epoch": 3.88, "learning_rate": 1.116504854368932e-05, "loss": 1.0166, "step": 52000 }, { "epoch": 3.92, "learning_rate": 1.0791635548917104e-05, "loss": 0.9493, "step": 52500 }, { "epoch": 3.96, "learning_rate": 1.0418222554144885e-05, "loss": 1.0561, "step": 53000 }, { "epoch": 4.0, "learning_rate": 1.0044809559372668e-05, "loss": 0.976, "step": 53500 }, { "epoch": 4.03, "learning_rate": 9.671396564600449e-06, "loss": 0.7222, "step": 54000 }, { "epoch": 4.07, "learning_rate": 9.29798356982823e-06, "loss": 0.6606, "step": 54500 }, { "epoch": 4.11, "learning_rate": 8.924570575056013e-06, "loss": 0.6666, "step": 55000 }, { "epoch": 4.14, "learning_rate": 8.551157580283794e-06, "loss": 0.6524, "step": 55500 }, { "epoch": 4.18, "learning_rate": 8.177744585511576e-06, "loss": 0.6646, "step": 56000 }, { "epoch": 4.22, "learning_rate": 7.804331590739359e-06, "loss": 0.7429, "step": 56500 }, { "epoch": 4.26, "learning_rate": 7.43091859596714e-06, "loss": 0.711, "step": 57000 }, { "epoch": 4.29, "learning_rate": 7.057505601194922e-06, "loss": 0.6983, "step": 57500 }, { "epoch": 4.33, "learning_rate": 6.684092606422704e-06, "loss": 0.6639, "step": 58000 }, { "epoch": 4.37, "learning_rate": 6.310679611650486e-06, "loss": 0.7254, "step": 58500 }, { "epoch": 4.41, "learning_rate": 5.937266616878267e-06, "loss": 0.701, "step": 59000 }, { "epoch": 4.44, "learning_rate": 5.563853622106049e-06, "loss": 0.7424, "step": 59500 }, { "epoch": 4.48, "learning_rate": 5.190440627333831e-06, "loss": 0.6679, "step": 60000 }, { "epoch": 4.52, "learning_rate": 4.817027632561613e-06, "loss": 0.6413, "step": 60500 }, { "epoch": 4.56, "learning_rate": 4.443614637789395e-06, "loss": 0.7014, "step": 61000 }, { "epoch": 4.59, "learning_rate": 4.070201643017177e-06, "loss": 0.6786, "step": 61500 }, { "epoch": 4.63, "learning_rate": 3.6967886482449596e-06, "loss": 0.6792, "step": 62000 }, { "epoch": 4.67, "learning_rate": 3.3233756534727413e-06, "loss": 0.7148, "step": 62500 }, { "epoch": 4.71, "learning_rate": 2.949962658700523e-06, "loss": 0.6512, "step": 63000 }, { "epoch": 4.74, "learning_rate": 2.5765496639283046e-06, "loss": 0.681, "step": 63500 }, { "epoch": 4.78, "learning_rate": 2.203136669156087e-06, "loss": 0.7019, "step": 64000 }, { "epoch": 4.82, "learning_rate": 1.8297236743838687e-06, "loss": 0.5755, "step": 64500 }, { "epoch": 4.85, "learning_rate": 1.4563106796116506e-06, "loss": 0.5506, "step": 65000 }, { "epoch": 4.89, "learning_rate": 1.0828976848394324e-06, "loss": 0.5695, "step": 65500 }, { "epoch": 4.93, "learning_rate": 7.094846900672144e-07, "loss": 0.5534, "step": 66000 }, { "epoch": 4.97, "learning_rate": 3.360716952949963e-07, "loss": 0.5473, "step": 66500 }, { "epoch": 5.0, "learning_rate": 1.425904192894484e-05, "loss": 0.5644, "step": 67000 }, { "epoch": 5.04, "learning_rate": 1.39923183612504e-05, "loss": 0.574, "step": 67500 }, { "epoch": 5.08, "learning_rate": 1.3725594793555959e-05, "loss": 0.5951, "step": 68000 }, { "epoch": 5.12, "learning_rate": 1.345887122586152e-05, "loss": 0.6167, "step": 68500 }, { "epoch": 5.15, "learning_rate": 1.3192147658167076e-05, "loss": 0.6116, "step": 69000 }, { "epoch": 5.19, "learning_rate": 1.2925424090472635e-05, "loss": 0.6739, "step": 69500 }, { "epoch": 5.23, "learning_rate": 1.2658700522778194e-05, "loss": 0.6638, "step": 70000 }, { "epoch": 5.27, "learning_rate": 1.2391976955083752e-05, "loss": 0.6339, "step": 70500 }, { "epoch": 5.3, "learning_rate": 1.212525338738931e-05, "loss": 0.6601, "step": 71000 }, { "epoch": 5.34, "learning_rate": 1.1858529819694868e-05, "loss": 0.6563, "step": 71500 }, { "epoch": 5.38, "learning_rate": 1.1591806252000428e-05, "loss": 0.6491, "step": 72000 }, { "epoch": 5.41, "learning_rate": 1.1325082684305987e-05, "loss": 0.6026, "step": 72500 }, { "epoch": 5.45, "learning_rate": 1.1058359116611544e-05, "loss": 0.6851, "step": 73000 }, { "epoch": 5.49, "learning_rate": 1.0791635548917104e-05, "loss": 0.6337, "step": 73500 }, { "epoch": 5.53, "learning_rate": 1.0524911981222661e-05, "loss": 0.6659, "step": 74000 }, { "epoch": 5.56, "learning_rate": 1.025818841352822e-05, "loss": 0.6691, "step": 74500 }, { "epoch": 5.6, "learning_rate": 9.991464845833778e-06, "loss": 0.6611, "step": 75000 }, { "epoch": 5.64, "learning_rate": 9.724741278139336e-06, "loss": 0.6542, "step": 75500 }, { "epoch": 5.68, "learning_rate": 9.458017710444895e-06, "loss": 0.6849, "step": 76000 }, { "epoch": 5.71, "learning_rate": 9.191294142750454e-06, "loss": 0.6552, "step": 76500 }, { "epoch": 5.75, "learning_rate": 8.924570575056013e-06, "loss": 0.6636, "step": 77000 }, { "epoch": 5.79, "learning_rate": 8.65784700736157e-06, "loss": 0.6323, "step": 77500 }, { "epoch": 5.83, "learning_rate": 8.39112343966713e-06, "loss": 0.6581, "step": 78000 }, { "epoch": 5.86, "learning_rate": 8.124399871972687e-06, "loss": 0.6613, "step": 78500 }, { "epoch": 5.9, "learning_rate": 7.857676304278247e-06, "loss": 0.6816, "step": 79000 }, { "epoch": 5.94, "learning_rate": 7.590952736583805e-06, "loss": 0.6908, "step": 79500 }, { "epoch": 5.97, "learning_rate": 7.324229168889363e-06, "loss": 0.6854, "step": 80000 }, { "epoch": 6.01, "learning_rate": 7.057505601194922e-06, "loss": 0.5705, "step": 80500 }, { "epoch": 6.05, "learning_rate": 6.79078203350048e-06, "loss": 0.4532, "step": 81000 }, { "epoch": 6.09, "learning_rate": 6.5240584658060394e-06, "loss": 0.4785, "step": 81500 }, { "epoch": 6.12, "learning_rate": 6.257334898111597e-06, "loss": 0.4568, "step": 82000 }, { "epoch": 6.16, "learning_rate": 5.990611330417156e-06, "loss": 0.4415, "step": 82500 }, { "epoch": 6.2, "learning_rate": 5.723887762722714e-06, "loss": 0.4598, "step": 83000 }, { "epoch": 6.24, "learning_rate": 5.457164195028273e-06, "loss": 0.4661, "step": 83500 }, { "epoch": 6.27, "learning_rate": 5.190440627333831e-06, "loss": 0.437, "step": 84000 }, { "epoch": 6.31, "learning_rate": 4.92371705963939e-06, "loss": 0.4782, "step": 84500 }, { "epoch": 6.35, "learning_rate": 4.656993491944948e-06, "loss": 0.465, "step": 85000 }, { "epoch": 6.39, "learning_rate": 4.390269924250507e-06, "loss": 0.4491, "step": 85500 }, { "epoch": 6.42, "learning_rate": 4.123546356556066e-06, "loss": 0.4727, "step": 86000 }, { "epoch": 6.46, "learning_rate": 3.856822788861624e-06, "loss": 0.4576, "step": 86500 }, { "epoch": 6.5, "learning_rate": 3.590099221167183e-06, "loss": 0.4565, "step": 87000 }, { "epoch": 6.53, "learning_rate": 3.3233756534727413e-06, "loss": 0.4646, "step": 87500 }, { "epoch": 6.57, "learning_rate": 3.0566520857782996e-06, "loss": 0.4749, "step": 88000 }, { "epoch": 6.61, "learning_rate": 2.789928518083858e-06, "loss": 0.3821, "step": 88500 }, { "epoch": 6.65, "learning_rate": 2.523204950389417e-06, "loss": 0.4083, "step": 89000 }, { "epoch": 6.68, "learning_rate": 2.2564813826949748e-06, "loss": 0.4041, "step": 89500 }, { "epoch": 6.72, "learning_rate": 1.9897578150005336e-06, "loss": 0.4081, "step": 90000 }, { "epoch": 6.76, "learning_rate": 1.723034247306092e-06, "loss": 0.4027, "step": 90500 }, { "epoch": 6.8, "learning_rate": 1.4563106796116506e-06, "loss": 0.4354, "step": 91000 }, { "epoch": 6.83, "learning_rate": 1.189587111917209e-06, "loss": 0.4193, "step": 91500 }, { "epoch": 6.87, "learning_rate": 9.228635442227675e-07, "loss": 0.3877, "step": 92000 }, { "epoch": 6.91, "learning_rate": 6.56139976528326e-07, "loss": 0.4322, "step": 92500 }, { "epoch": 6.95, "learning_rate": 3.894164088338846e-07, "loss": 0.4068, "step": 93000 }, { "epoch": 6.98, "learning_rate": 1.226928411394431e-07, "loss": 0.3968, "step": 93500 }, { "epoch": 7.0, "step": 93730, "total_flos": 2.490638010831667e+16, "train_loss": 0.024883948061263487, "train_runtime": 1768.5377, "train_samples_per_second": 211.99, "train_steps_per_second": 52.999 } ], "logging_steps": 500, "max_steps": 93730, "num_input_tokens_seen": 0, "num_train_epochs": 7, "save_steps": 500, "total_flos": 2.490638010831667e+16, "train_batch_size": 4, "trial_name": null, "trial_params": null }