{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 11646,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 4.9570668040528936e-05,
      "loss": 2.5131,
      "step": 100
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.914133608105788e-05,
      "loss": 2.1951,
      "step": 200
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.871200412158681e-05,
      "loss": 1.9936,
      "step": 300
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.828267216211575e-05,
      "loss": 1.8599,
      "step": 400
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.7853340202644685e-05,
      "loss": 1.7738,
      "step": 500
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.7424008243173625e-05,
      "loss": 1.7447,
      "step": 600
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.699467628370256e-05,
      "loss": 1.664,
      "step": 700
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.65653443242315e-05,
      "loss": 1.6198,
      "step": 800
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.6136012364760434e-05,
      "loss": 1.5546,
      "step": 900
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.570668040528937e-05,
      "loss": 1.5249,
      "step": 1000
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.527734844581831e-05,
      "loss": 1.4905,
      "step": 1100
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.484801648634724e-05,
      "loss": 1.4535,
      "step": 1200
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.441868452687618e-05,
      "loss": 1.4336,
      "step": 1300
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.3989352567405116e-05,
      "loss": 1.4008,
      "step": 1400
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.356002060793406e-05,
      "loss": 1.3552,
      "step": 1500
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.313068864846299e-05,
      "loss": 1.3854,
      "step": 1600
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.270135668899193e-05,
      "loss": 1.3722,
      "step": 1700
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.2272024729520865e-05,
      "loss": 1.3153,
      "step": 1800
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.1842692770049805e-05,
      "loss": 1.2735,
      "step": 1900
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.1413360810578746e-05,
      "loss": 1.293,
      "step": 2000
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.098402885110768e-05,
      "loss": 1.2661,
      "step": 2100
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.055469689163662e-05,
      "loss": 1.2543,
      "step": 2200
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.0125364932165554e-05,
      "loss": 1.2635,
      "step": 2300
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.969603297269449e-05,
      "loss": 1.2306,
      "step": 2400
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.926670101322343e-05,
      "loss": 1.2114,
      "step": 2500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.883736905375236e-05,
      "loss": 1.2305,
      "step": 2600
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.84080370942813e-05,
      "loss": 1.1961,
      "step": 2700
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.797870513481024e-05,
      "loss": 1.1866,
      "step": 2800
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.754937317533918e-05,
      "loss": 1.1808,
      "step": 2900
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.712004121586811e-05,
      "loss": 1.1883,
      "step": 3000
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.669070925639705e-05,
      "loss": 1.1482,
      "step": 3100
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.6261377296925985e-05,
      "loss": 1.1377,
      "step": 3200
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.5832045337454926e-05,
      "loss": 1.1463,
      "step": 3300
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.540271337798386e-05,
      "loss": 1.1347,
      "step": 3400
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.4973381418512793e-05,
      "loss": 1.1217,
      "step": 3500
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.4544049459041734e-05,
      "loss": 1.1249,
      "step": 3600
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.411471749957067e-05,
      "loss": 1.0972,
      "step": 3700
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.368538554009961e-05,
      "loss": 1.1206,
      "step": 3800
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.325605358062854e-05,
      "loss": 1.0817,
      "step": 3900
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.282672162115748e-05,
      "loss": 1.0486,
      "step": 4000
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.2397389661686417e-05,
      "loss": 1.0737,
      "step": 4100
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.196805770221536e-05,
      "loss": 1.0448,
      "step": 4200
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.153872574274429e-05,
      "loss": 1.0633,
      "step": 4300
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.1109393783273225e-05,
      "loss": 1.0297,
      "step": 4400
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.0680061823802165e-05,
      "loss": 1.0405,
      "step": 4500
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.0250729864331102e-05,
      "loss": 1.024,
      "step": 4600
    },
    {
      "epoch": 1.21,
      "learning_rate": 2.9821397904860036e-05,
      "loss": 0.9993,
      "step": 4700
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.9392065945388973e-05,
      "loss": 1.0126,
      "step": 4800
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.896273398591791e-05,
      "loss": 1.0055,
      "step": 4900
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.8533402026446848e-05,
      "loss": 1.0017,
      "step": 5000
    },
    {
      "epoch": 1.31,
      "learning_rate": 2.8104070066975785e-05,
      "loss": 1.0052,
      "step": 5100
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.7674738107504722e-05,
      "loss": 0.995,
      "step": 5200
    },
    {
      "epoch": 1.37,
      "learning_rate": 2.724540614803366e-05,
      "loss": 1.0071,
      "step": 5300
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.6816074188562596e-05,
      "loss": 0.9835,
      "step": 5400
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.6386742229091534e-05,
      "loss": 0.9986,
      "step": 5500
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.5957410269620467e-05,
      "loss": 0.9837,
      "step": 5600
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.552807831014941e-05,
      "loss": 0.9897,
      "step": 5700
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.509874635067835e-05,
      "loss": 0.996,
      "step": 5800
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.4669414391207282e-05,
      "loss": 0.9831,
      "step": 5900
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.424008243173622e-05,
      "loss": 0.9518,
      "step": 6000
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3810750472265157e-05,
      "loss": 0.9758,
      "step": 6100
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.3381418512794094e-05,
      "loss": 0.9629,
      "step": 6200
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.295208655332303e-05,
      "loss": 0.9607,
      "step": 6300
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.2522754593851965e-05,
      "loss": 0.9619,
      "step": 6400
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.2093422634380902e-05,
      "loss": 0.9528,
      "step": 6500
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.1664090674909843e-05,
      "loss": 0.9504,
      "step": 6600
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.123475871543878e-05,
      "loss": 0.9569,
      "step": 6700
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.0805426755967717e-05,
      "loss": 0.9721,
      "step": 6800
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.0376094796496654e-05,
      "loss": 0.9521,
      "step": 6900
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.994676283702559e-05,
      "loss": 0.9583,
      "step": 7000
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.9517430877554525e-05,
      "loss": 0.9444,
      "step": 7100
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.9088098918083462e-05,
      "loss": 0.9116,
      "step": 7200
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.86587669586124e-05,
      "loss": 0.9321,
      "step": 7300
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.8229434999141337e-05,
      "loss": 0.9296,
      "step": 7400
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.7800103039670274e-05,
      "loss": 0.9124,
      "step": 7500
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.737077108019921e-05,
      "loss": 0.9234,
      "step": 7600
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.6941439120728148e-05,
      "loss": 0.9581,
      "step": 7700
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.6512107161257085e-05,
      "loss": 0.9333,
      "step": 7800
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.6082775201786023e-05,
      "loss": 0.8964,
      "step": 7900
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.5653443242314956e-05,
      "loss": 0.9115,
      "step": 8000
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.5224111282843895e-05,
      "loss": 0.9046,
      "step": 8100
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.4794779323372832e-05,
      "loss": 0.8765,
      "step": 8200
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.4365447363901768e-05,
      "loss": 0.8779,
      "step": 8300
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.3936115404430705e-05,
      "loss": 0.8801,
      "step": 8400
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3506783444959644e-05,
      "loss": 0.8843,
      "step": 8500
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.3077451485488581e-05,
      "loss": 0.8848,
      "step": 8600
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2648119526017518e-05,
      "loss": 0.8957,
      "step": 8700
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2218787566546454e-05,
      "loss": 0.8997,
      "step": 8800
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.1789455607075391e-05,
      "loss": 0.8977,
      "step": 8900
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.1360123647604328e-05,
      "loss": 0.8699,
      "step": 9000
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.0930791688133265e-05,
      "loss": 0.8895,
      "step": 9100
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.0501459728662203e-05,
      "loss": 0.8848,
      "step": 9200
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.007212776919114e-05,
      "loss": 0.8647,
      "step": 9300
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.642795809720075e-06,
      "loss": 0.8836,
      "step": 9400
    },
    {
      "epoch": 2.45,
      "learning_rate": 9.213463850249012e-06,
      "loss": 0.8824,
      "step": 9500
    },
    {
      "epoch": 2.47,
      "learning_rate": 8.78413189077795e-06,
      "loss": 0.8672,
      "step": 9600
    },
    {
      "epoch": 2.5,
      "learning_rate": 8.354799931306887e-06,
      "loss": 0.881,
      "step": 9700
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.925467971835824e-06,
      "loss": 0.8505,
      "step": 9800
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.496136012364761e-06,
      "loss": 0.8925,
      "step": 9900
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.066804052893698e-06,
      "loss": 0.8876,
      "step": 10000
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.637472093422635e-06,
      "loss": 0.8716,
      "step": 10100
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.208140133951572e-06,
      "loss": 0.875,
      "step": 10200
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.778808174480509e-06,
      "loss": 0.8848,
      "step": 10300
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.349476215009445e-06,
      "loss": 0.8629,
      "step": 10400
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.9201442555383824e-06,
      "loss": 0.8797,
      "step": 10500
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.49081229606732e-06,
      "loss": 0.8478,
      "step": 10600
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.061480336596257e-06,
      "loss": 0.8526,
      "step": 10700
    },
    {
      "epoch": 2.78,
      "learning_rate": 3.6321483771251936e-06,
      "loss": 0.8619,
      "step": 10800
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.2028164176541303e-06,
      "loss": 0.8923,
      "step": 10900
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.7734844581830675e-06,
      "loss": 0.8737,
      "step": 11000
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.3441524987120042e-06,
      "loss": 0.8556,
      "step": 11100
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.9148205392409414e-06,
      "loss": 0.8767,
      "step": 11200
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.4854885797698782e-06,
      "loss": 0.886,
      "step": 11300
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.0561566202988151e-06,
      "loss": 0.8638,
      "step": 11400
    },
    {
      "epoch": 2.96,
      "learning_rate": 6.26824660827752e-07,
      "loss": 0.8885,
      "step": 11500
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.97492701356689e-07,
      "loss": 0.881,
      "step": 11600
    },
    {
      "epoch": 3.0,
      "step": 11646,
      "total_flos": 6.280174860167578e+16,
      "train_loss": 1.0872868923270456,
      "train_runtime": 30832.4334,
      "train_samples_per_second": 6.043,
      "train_steps_per_second": 0.378
    }
  ],
  "max_steps": 11646,
  "num_train_epochs": 3,
  "total_flos": 6.280174860167578e+16,
  "trial_name": null,
  "trial_params": null
}