{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.985976939856653,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "learning_rate": 4.2e-05,
      "loss": 0.8909,
      "step": 50
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.1e-05,
      "loss": 0.8023,
      "step": 100
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.7678152918815613,
      "eval_runtime": 2324.1166,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 100
    },
    {
      "epoch": 0.19,
      "learning_rate": 9.895140664961637e-05,
      "loss": 0.7815,
      "step": 150
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.767263427109975e-05,
      "loss": 0.8184,
      "step": 200
    },
    {
      "epoch": 0.25,
      "eval_loss": 0.7796664834022522,
      "eval_runtime": 2323.4202,
      "eval_samples_per_second": 0.633,
      "eval_steps_per_second": 0.633,
      "step": 200
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.639386189258313e-05,
      "loss": 0.7501,
      "step": 250
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.51150895140665e-05,
      "loss": 0.7786,
      "step": 300
    },
    {
      "epoch": 0.37,
      "eval_loss": 0.7461986541748047,
      "eval_runtime": 2323.156,
      "eval_samples_per_second": 0.633,
      "eval_steps_per_second": 0.633,
      "step": 300
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.383631713554988e-05,
      "loss": 0.7247,
      "step": 350
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.255754475703325e-05,
      "loss": 0.699,
      "step": 400
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.7154579162597656,
      "eval_runtime": 2326.716,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 400
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.127877237851663e-05,
      "loss": 0.7498,
      "step": 450
    },
    {
      "epoch": 0.62,
      "learning_rate": 9e-05,
      "loss": 0.6729,
      "step": 500
    },
    {
      "epoch": 0.62,
      "eval_loss": 0.6918142437934875,
      "eval_runtime": 2326.9582,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 500
    },
    {
      "epoch": 0.69,
      "learning_rate": 8.872122762148338e-05,
      "loss": 0.6821,
      "step": 550
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.744245524296676e-05,
      "loss": 0.7392,
      "step": 600
    },
    {
      "epoch": 0.75,
      "eval_loss": 0.6949509978294373,
      "eval_runtime": 2328.8704,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 600
    },
    {
      "epoch": 0.81,
      "learning_rate": 8.616368286445013e-05,
      "loss": 0.6848,
      "step": 650
    },
    {
      "epoch": 0.87,
      "learning_rate": 8.488491048593351e-05,
      "loss": 0.7601,
      "step": 700
    },
    {
      "epoch": 0.87,
      "eval_loss": 0.6541839241981506,
      "eval_runtime": 2322.3854,
      "eval_samples_per_second": 0.633,
      "eval_steps_per_second": 0.633,
      "step": 700
    },
    {
      "epoch": 0.93,
      "learning_rate": 8.360613810741689e-05,
      "loss": 0.7163,
      "step": 750
    },
    {
      "epoch": 1.0,
      "learning_rate": 8.232736572890026e-05,
      "loss": 0.7121,
      "step": 800
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.7020866274833679,
      "eval_runtime": 2329.0246,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 800
    },
    {
      "epoch": 1.06,
      "learning_rate": 8.104859335038364e-05,
      "loss": 0.6303,
      "step": 850
    },
    {
      "epoch": 1.12,
      "learning_rate": 7.992327365728901e-05,
      "loss": 0.834,
      "step": 900
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.8280826210975647,
      "eval_runtime": 2322.7118,
      "eval_samples_per_second": 0.633,
      "eval_steps_per_second": 0.633,
      "step": 900
    },
    {
      "epoch": 1.18,
      "learning_rate": 7.867007672634272e-05,
      "loss": 0.7102,
      "step": 950
    },
    {
      "epoch": 1.25,
      "learning_rate": 7.73913043478261e-05,
      "loss": 0.6845,
      "step": 1000
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.684939444065094,
      "eval_runtime": 2322.9147,
      "eval_samples_per_second": 0.633,
      "eval_steps_per_second": 0.633,
      "step": 1000
    },
    {
      "epoch": 1.31,
      "learning_rate": 7.611253196930947e-05,
      "loss": 0.6918,
      "step": 1050
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.483375959079285e-05,
      "loss": 0.7139,
      "step": 1100
    },
    {
      "epoch": 1.37,
      "eval_loss": 0.682366132736206,
      "eval_runtime": 2323.8642,
      "eval_samples_per_second": 0.633,
      "eval_steps_per_second": 0.633,
      "step": 1100
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.355498721227622e-05,
      "loss": 0.6513,
      "step": 1150
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.22762148337596e-05,
      "loss": 0.6515,
      "step": 1200
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.6614900827407837,
      "eval_runtime": 2324.2619,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 1200
    },
    {
      "epoch": 1.56,
      "learning_rate": 7.099744245524297e-05,
      "loss": 0.6325,
      "step": 1250
    },
    {
      "epoch": 1.62,
      "learning_rate": 6.971867007672635e-05,
      "loss": 0.6658,
      "step": 1300
    },
    {
      "epoch": 1.62,
      "eval_loss": 0.6691812872886658,
      "eval_runtime": 2322.8992,
      "eval_samples_per_second": 0.633,
      "eval_steps_per_second": 0.633,
      "step": 1300
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.843989769820973e-05,
      "loss": 0.6709,
      "step": 1350
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.71611253196931e-05,
      "loss": 0.6571,
      "step": 1400
    },
    {
      "epoch": 1.75,
      "eval_loss": 0.7406654357910156,
      "eval_runtime": 2323.316,
      "eval_samples_per_second": 0.633,
      "eval_steps_per_second": 0.633,
      "step": 1400
    },
    {
      "epoch": 1.81,
      "learning_rate": 6.588235294117648e-05,
      "loss": 0.6527,
      "step": 1450
    },
    {
      "epoch": 1.87,
      "learning_rate": 6.460358056265985e-05,
      "loss": 0.6432,
      "step": 1500
    },
    {
      "epoch": 1.87,
      "eval_loss": 0.6645843982696533,
      "eval_runtime": 2325.3057,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 1500
    },
    {
      "epoch": 1.93,
      "learning_rate": 6.332480818414323e-05,
      "loss": 0.6998,
      "step": 1550
    },
    {
      "epoch": 1.99,
      "learning_rate": 6.20460358056266e-05,
      "loss": 0.6777,
      "step": 1600
    },
    {
      "epoch": 1.99,
      "eval_loss": 0.7001596093177795,
      "eval_runtime": 2324.698,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 1600
    },
    {
      "epoch": 2.06,
      "learning_rate": 6.0767263427109976e-05,
      "loss": 0.63,
      "step": 1650
    },
    {
      "epoch": 2.12,
      "learning_rate": 5.948849104859335e-05,
      "loss": 0.6027,
      "step": 1700
    },
    {
      "epoch": 2.12,
      "eval_loss": 0.6740280389785767,
      "eval_runtime": 2325.0824,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 1700
    },
    {
      "epoch": 2.18,
      "learning_rate": 5.820971867007673e-05,
      "loss": 0.6171,
      "step": 1750
    },
    {
      "epoch": 2.24,
      "learning_rate": 5.6930946291560104e-05,
      "loss": 0.5711,
      "step": 1800
    },
    {
      "epoch": 2.24,
      "eval_loss": 0.6770710349082947,
      "eval_runtime": 2324.7006,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 1800
    },
    {
      "epoch": 2.31,
      "learning_rate": 5.565217391304348e-05,
      "loss": 0.6193,
      "step": 1850
    },
    {
      "epoch": 2.37,
      "learning_rate": 5.4373401534526856e-05,
      "loss": 0.6544,
      "step": 1900
    },
    {
      "epoch": 2.37,
      "eval_loss": 0.7468971014022827,
      "eval_runtime": 2324.7323,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 1900
    },
    {
      "epoch": 2.43,
      "learning_rate": 5.309462915601023e-05,
      "loss": 0.6559,
      "step": 1950
    },
    {
      "epoch": 2.49,
      "learning_rate": 5.181585677749361e-05,
      "loss": 0.6981,
      "step": 2000
    },
    {
      "epoch": 2.49,
      "eval_loss": 0.6552309393882751,
      "eval_runtime": 2325.8405,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2000
    },
    {
      "epoch": 2.56,
      "learning_rate": 5.0537084398976984e-05,
      "loss": 0.6161,
      "step": 2050
    },
    {
      "epoch": 2.62,
      "learning_rate": 4.925831202046036e-05,
      "loss": 0.6425,
      "step": 2100
    },
    {
      "epoch": 2.62,
      "eval_loss": 0.6608405113220215,
      "eval_runtime": 2324.8602,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2100
    },
    {
      "epoch": 2.68,
      "learning_rate": 4.7979539641943737e-05,
      "loss": 0.5966,
      "step": 2150
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.670076726342711e-05,
      "loss": 0.6435,
      "step": 2200
    },
    {
      "epoch": 2.74,
      "eval_loss": 0.6722382307052612,
      "eval_runtime": 2325.4748,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2200
    },
    {
      "epoch": 2.8,
      "learning_rate": 4.542199488491049e-05,
      "loss": 0.6493,
      "step": 2250
    },
    {
      "epoch": 2.87,
      "learning_rate": 4.4143222506393865e-05,
      "loss": 0.6281,
      "step": 2300
    },
    {
      "epoch": 2.87,
      "eval_loss": 0.6553158760070801,
      "eval_runtime": 2324.9442,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2300
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.286445012787724e-05,
      "loss": 0.5495,
      "step": 2350
    },
    {
      "epoch": 2.99,
      "learning_rate": 4.158567774936061e-05,
      "loss": 0.6129,
      "step": 2400
    },
    {
      "epoch": 2.99,
      "eval_loss": 0.6536511182785034,
      "eval_runtime": 2325.6669,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2400
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.0306905370843986e-05,
      "loss": 0.5085,
      "step": 2450
    },
    {
      "epoch": 3.12,
      "learning_rate": 3.902813299232736e-05,
      "loss": 0.5443,
      "step": 2500
    },
    {
      "epoch": 3.12,
      "eval_loss": 0.659809947013855,
      "eval_runtime": 2326.2914,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2500
    },
    {
      "epoch": 3.18,
      "learning_rate": 3.774936061381074e-05,
      "loss": 0.5447,
      "step": 2550
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.6470588235294114e-05,
      "loss": 0.535,
      "step": 2600
    },
    {
      "epoch": 3.24,
      "eval_loss": 0.6724178194999695,
      "eval_runtime": 2327.4156,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2600
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.519181585677749e-05,
      "loss": 0.6094,
      "step": 2650
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.3913043478260867e-05,
      "loss": 0.5347,
      "step": 2700
    },
    {
      "epoch": 3.37,
      "eval_loss": 0.6599487662315369,
      "eval_runtime": 2327.143,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2700
    },
    {
      "epoch": 3.43,
      "learning_rate": 3.263427109974424e-05,
      "loss": 0.572,
      "step": 2750
    },
    {
      "epoch": 3.49,
      "learning_rate": 3.135549872122762e-05,
      "loss": 0.5493,
      "step": 2800
    },
    {
      "epoch": 3.49,
      "eval_loss": 0.6231494545936584,
      "eval_runtime": 2326.2678,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2800
    },
    {
      "epoch": 3.55,
      "learning_rate": 3.0076726342710998e-05,
      "loss": 0.5566,
      "step": 2850
    },
    {
      "epoch": 3.61,
      "learning_rate": 2.8797953964194374e-05,
      "loss": 0.5564,
      "step": 2900
    },
    {
      "epoch": 3.61,
      "eval_loss": 0.6593871116638184,
      "eval_runtime": 2327.2395,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 2900
    },
    {
      "epoch": 3.68,
      "learning_rate": 2.751918158567775e-05,
      "loss": 0.5787,
      "step": 2950
    },
    {
      "epoch": 3.74,
      "learning_rate": 2.6240409207161126e-05,
      "loss": 0.509,
      "step": 3000
    },
    {
      "epoch": 3.74,
      "eval_loss": 0.6348086595535278,
      "eval_runtime": 2326.4619,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 3000
    },
    {
      "epoch": 3.8,
      "learning_rate": 2.4961636828644502e-05,
      "loss": 0.5356,
      "step": 3050
    },
    {
      "epoch": 3.86,
      "learning_rate": 2.368286445012788e-05,
      "loss": 0.52,
      "step": 3100
    },
    {
      "epoch": 3.86,
      "eval_loss": 0.6384520530700684,
      "eval_runtime": 2327.6186,
      "eval_samples_per_second": 0.632,
      "eval_steps_per_second": 0.632,
      "step": 3100
    },
    {
      "epoch": 3.93,
      "learning_rate": 2.2404092071611255e-05,
      "loss": 0.5418,
      "step": 3150
    },
    {
      "epoch": 3.99,
      "learning_rate": 2.112531969309463e-05,
      "loss": 0.5569,
      "step": 3200
    },
    {
      "epoch": 3.99,
      "eval_loss": 0.6191104650497437,
      "eval_runtime": 2329.026,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 3200
    },
    {
      "epoch": 4.05,
      "learning_rate": 1.9846547314578007e-05,
      "loss": 0.4716,
      "step": 3250
    },
    {
      "epoch": 4.11,
      "learning_rate": 1.8567774936061383e-05,
      "loss": 0.501,
      "step": 3300
    },
    {
      "epoch": 4.11,
      "eval_loss": 0.6244591474533081,
      "eval_runtime": 2328.4154,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 3300
    },
    {
      "epoch": 4.18,
      "learning_rate": 1.728900255754476e-05,
      "loss": 0.4853,
      "step": 3350
    },
    {
      "epoch": 4.24,
      "learning_rate": 1.6010230179028135e-05,
      "loss": 0.5105,
      "step": 3400
    },
    {
      "epoch": 4.24,
      "eval_loss": 0.6227459907531738,
      "eval_runtime": 2328.6203,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 3400
    },
    {
      "epoch": 4.3,
      "learning_rate": 1.4731457800511508e-05,
      "loss": 0.4757,
      "step": 3450
    },
    {
      "epoch": 4.36,
      "learning_rate": 1.3452685421994884e-05,
      "loss": 0.5064,
      "step": 3500
    },
    {
      "epoch": 4.36,
      "eval_loss": 0.628924548625946,
      "eval_runtime": 2328.7385,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 3500
    },
    {
      "epoch": 4.43,
      "learning_rate": 1.2173913043478261e-05,
      "loss": 0.5107,
      "step": 3550
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.0895140664961638e-05,
      "loss": 0.5088,
      "step": 3600
    },
    {
      "epoch": 4.49,
      "eval_loss": 0.620337724685669,
      "eval_runtime": 2329.0799,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 3600
    },
    {
      "epoch": 4.55,
      "learning_rate": 9.616368286445012e-06,
      "loss": 0.5184,
      "step": 3650
    },
    {
      "epoch": 4.61,
      "learning_rate": 8.337595907928388e-06,
      "loss": 0.5246,
      "step": 3700
    },
    {
      "epoch": 4.61,
      "eval_loss": 0.618162989616394,
      "eval_runtime": 2327.9739,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 3700
    },
    {
      "epoch": 4.67,
      "learning_rate": 7.058823529411765e-06,
      "loss": 0.5027,
      "step": 3750
    },
    {
      "epoch": 4.74,
      "learning_rate": 5.780051150895141e-06,
      "loss": 0.4927,
      "step": 3800
    },
    {
      "epoch": 4.74,
      "eval_loss": 0.6198447942733765,
      "eval_runtime": 2328.8672,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 3800
    },
    {
      "epoch": 4.8,
      "learning_rate": 4.501278772378517e-06,
      "loss": 0.5568,
      "step": 3850
    },
    {
      "epoch": 4.86,
      "learning_rate": 3.2225063938618927e-06,
      "loss": 0.5038,
      "step": 3900
    },
    {
      "epoch": 4.86,
      "eval_loss": 0.6127236485481262,
      "eval_runtime": 2329.1619,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 3900
    },
    {
      "epoch": 4.92,
      "learning_rate": 1.943734015345269e-06,
      "loss": 0.5114,
      "step": 3950
    },
    {
      "epoch": 4.99,
      "learning_rate": 6.649616368286445e-07,
      "loss": 0.5494,
      "step": 4000
    },
    {
      "epoch": 4.99,
      "eval_loss": 0.614413321018219,
      "eval_runtime": 2328.9991,
      "eval_samples_per_second": 0.631,
      "eval_steps_per_second": 0.631,
      "step": 4000
    }
  ],
  "max_steps": 4010,
  "num_train_epochs": 5,
  "total_flos": 3.2509882552308204e+18,
  "trial_name": null,
  "trial_params": null
}