{
  "best_metric": 0.7940782904624939,
  "best_model_checkpoint": "models_v3/albertina-100m-pt-41-128-t0-v0/checkpoint-65000",
  "epoch": 9.40529590507886,
  "eval_steps": 5000,
  "global_step": 65000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 3.617421501953408e-06,
      "loss": 1.794,
      "step": 500
    },
    {
      "epoch": 0.14,
      "learning_rate": 7.234843003906816e-06,
      "loss": 1.6062,
      "step": 1000
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.0852264505860224e-05,
      "loss": 1.5053,
      "step": 1500
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.4469686007813631e-05,
      "loss": 1.4352,
      "step": 2000
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.808710750976704e-05,
      "loss": 1.3906,
      "step": 2500
    },
    {
      "epoch": 0.43,
      "learning_rate": 2.1704529011720447e-05,
      "loss": 1.3456,
      "step": 3000
    },
    {
      "epoch": 0.51,
      "learning_rate": 2.5321950513673853e-05,
      "loss": 1.3123,
      "step": 3500
    },
    {
      "epoch": 0.58,
      "learning_rate": 2.8939372015627263e-05,
      "loss": 1.2811,
      "step": 4000
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.255679351758067e-05,
      "loss": 1.2554,
      "step": 4500
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.617421501953408e-05,
      "loss": 1.2324,
      "step": 5000
    },
    {
      "epoch": 0.72,
      "eval_loss": 1.2159796953201294,
      "eval_runtime": 335.7202,
      "eval_samples_per_second": 384.555,
      "eval_steps_per_second": 1.504,
      "step": 5000
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.979163652148749e-05,
      "loss": 1.2154,
      "step": 5500
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.3409058023440894e-05,
      "loss": 1.1969,
      "step": 6000
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.70264795253943e-05,
      "loss": 1.1784,
      "step": 6500
    },
    {
      "epoch": 1.01,
      "learning_rate": 4.9928455441405815e-05,
      "loss": 1.165,
      "step": 7000
    },
    {
      "epoch": 1.09,
      "learning_rate": 4.9526519718966546e-05,
      "loss": 1.1452,
      "step": 7500
    },
    {
      "epoch": 1.16,
      "learning_rate": 4.912458399652728e-05,
      "loss": 1.133,
      "step": 8000
    },
    {
      "epoch": 1.23,
      "learning_rate": 4.8722648274088015e-05,
      "loss": 1.115,
      "step": 8500
    },
    {
      "epoch": 1.3,
      "learning_rate": 4.8320712551648746e-05,
      "loss": 1.1026,
      "step": 9000
    },
    {
      "epoch": 1.37,
      "learning_rate": 4.791877682920948e-05,
      "loss": 1.0872,
      "step": 9500
    },
    {
      "epoch": 1.45,
      "learning_rate": 4.751684110677021e-05,
      "loss": 1.0802,
      "step": 10000
    },
    {
      "epoch": 1.45,
      "eval_loss": 1.0682921409606934,
      "eval_runtime": 335.7245,
      "eval_samples_per_second": 384.55,
      "eval_steps_per_second": 1.504,
      "step": 10000
    },
    {
      "epoch": 1.52,
      "learning_rate": 4.711490538433094e-05,
      "loss": 1.0682,
      "step": 10500
    },
    {
      "epoch": 1.59,
      "learning_rate": 4.671296966189167e-05,
      "loss": 1.0589,
      "step": 11000
    },
    {
      "epoch": 1.66,
      "learning_rate": 4.63110339394524e-05,
      "loss": 1.0491,
      "step": 11500
    },
    {
      "epoch": 1.74,
      "learning_rate": 4.590909821701313e-05,
      "loss": 1.0425,
      "step": 12000
    },
    {
      "epoch": 1.81,
      "learning_rate": 4.550716249457387e-05,
      "loss": 1.0313,
      "step": 12500
    },
    {
      "epoch": 1.88,
      "learning_rate": 4.51052267721346e-05,
      "loss": 1.0224,
      "step": 13000
    },
    {
      "epoch": 1.95,
      "learning_rate": 4.470329104969533e-05,
      "loss": 1.0129,
      "step": 13500
    },
    {
      "epoch": 2.03,
      "learning_rate": 4.4301355327256064e-05,
      "loss": 1.0066,
      "step": 14000
    },
    {
      "epoch": 2.1,
      "learning_rate": 4.38994196048168e-05,
      "loss": 0.998,
      "step": 14500
    },
    {
      "epoch": 2.17,
      "learning_rate": 4.349748388237753e-05,
      "loss": 0.9909,
      "step": 15000
    },
    {
      "epoch": 2.17,
      "eval_loss": 0.9950871467590332,
      "eval_runtime": 335.5649,
      "eval_samples_per_second": 384.733,
      "eval_steps_per_second": 1.505,
      "step": 15000
    },
    {
      "epoch": 2.24,
      "learning_rate": 4.3095548159938264e-05,
      "loss": 0.9861,
      "step": 15500
    },
    {
      "epoch": 2.32,
      "learning_rate": 4.2693612437498996e-05,
      "loss": 0.9806,
      "step": 16000
    },
    {
      "epoch": 2.39,
      "learning_rate": 4.2291676715059733e-05,
      "loss": 0.9738,
      "step": 16500
    },
    {
      "epoch": 2.46,
      "learning_rate": 4.1889740992620465e-05,
      "loss": 0.9713,
      "step": 17000
    },
    {
      "epoch": 2.53,
      "learning_rate": 4.1487805270181196e-05,
      "loss": 0.9684,
      "step": 17500
    },
    {
      "epoch": 2.6,
      "learning_rate": 4.108586954774193e-05,
      "loss": 0.9595,
      "step": 18000
    },
    {
      "epoch": 2.68,
      "learning_rate": 4.0683933825302665e-05,
      "loss": 0.9576,
      "step": 18500
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.0281998102863396e-05,
      "loss": 0.9507,
      "step": 19000
    },
    {
      "epoch": 2.82,
      "learning_rate": 3.988006238042413e-05,
      "loss": 0.946,
      "step": 19500
    },
    {
      "epoch": 2.89,
      "learning_rate": 3.947812665798486e-05,
      "loss": 0.9471,
      "step": 20000
    },
    {
      "epoch": 2.89,
      "eval_loss": 0.947361171245575,
      "eval_runtime": 335.631,
      "eval_samples_per_second": 384.658,
      "eval_steps_per_second": 1.505,
      "step": 20000
    },
    {
      "epoch": 2.97,
      "learning_rate": 3.907619093554559e-05,
      "loss": 0.939,
      "step": 20500
    },
    {
      "epoch": 3.04,
      "learning_rate": 3.867425521310632e-05,
      "loss": 0.9318,
      "step": 21000
    },
    {
      "epoch": 3.11,
      "learning_rate": 3.827231949066705e-05,
      "loss": 0.9259,
      "step": 21500
    },
    {
      "epoch": 3.18,
      "learning_rate": 3.787038376822778e-05,
      "loss": 0.9219,
      "step": 22000
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.746844804578852e-05,
      "loss": 0.9174,
      "step": 22500
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.706651232334925e-05,
      "loss": 0.9179,
      "step": 23000
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.666457660090998e-05,
      "loss": 0.9122,
      "step": 23500
    },
    {
      "epoch": 3.47,
      "learning_rate": 3.6262640878470714e-05,
      "loss": 0.9121,
      "step": 24000
    },
    {
      "epoch": 3.55,
      "learning_rate": 3.586070515603145e-05,
      "loss": 0.9079,
      "step": 24500
    },
    {
      "epoch": 3.62,
      "learning_rate": 3.545876943359218e-05,
      "loss": 0.9019,
      "step": 25000
    },
    {
      "epoch": 3.62,
      "eval_loss": 0.9143123030662537,
      "eval_runtime": 336.064,
      "eval_samples_per_second": 384.162,
      "eval_steps_per_second": 1.503,
      "step": 25000
    },
    {
      "epoch": 3.69,
      "learning_rate": 3.5056833711152914e-05,
      "loss": 0.8999,
      "step": 25500
    },
    {
      "epoch": 3.76,
      "learning_rate": 3.4654897988713646e-05,
      "loss": 0.8967,
      "step": 26000
    },
    {
      "epoch": 3.83,
      "learning_rate": 3.425296226627438e-05,
      "loss": 0.8928,
      "step": 26500
    },
    {
      "epoch": 3.91,
      "learning_rate": 3.3851026543835115e-05,
      "loss": 0.8937,
      "step": 27000
    },
    {
      "epoch": 3.98,
      "learning_rate": 3.3449090821395846e-05,
      "loss": 0.8882,
      "step": 27500
    },
    {
      "epoch": 4.05,
      "learning_rate": 3.304715509895658e-05,
      "loss": 0.8859,
      "step": 28000
    },
    {
      "epoch": 4.12,
      "learning_rate": 3.264521937651731e-05,
      "loss": 0.8792,
      "step": 28500
    },
    {
      "epoch": 4.2,
      "learning_rate": 3.2243283654078046e-05,
      "loss": 0.8769,
      "step": 29000
    },
    {
      "epoch": 4.27,
      "learning_rate": 3.184134793163878e-05,
      "loss": 0.8752,
      "step": 29500
    },
    {
      "epoch": 4.34,
      "learning_rate": 3.14394122091995e-05,
      "loss": 0.8715,
      "step": 30000
    },
    {
      "epoch": 4.34,
      "eval_loss": 0.8864349722862244,
      "eval_runtime": 335.5739,
      "eval_samples_per_second": 384.723,
      "eval_steps_per_second": 1.505,
      "step": 30000
    },
    {
      "epoch": 4.41,
      "learning_rate": 3.103747648676024e-05,
      "loss": 0.8749,
      "step": 30500
    },
    {
      "epoch": 4.49,
      "learning_rate": 3.063554076432097e-05,
      "loss": 0.8685,
      "step": 31000
    },
    {
      "epoch": 4.56,
      "learning_rate": 3.02336050418817e-05,
      "loss": 0.8663,
      "step": 31500
    },
    {
      "epoch": 4.63,
      "learning_rate": 2.9831669319442436e-05,
      "loss": 0.8624,
      "step": 32000
    },
    {
      "epoch": 4.7,
      "learning_rate": 2.9429733597003167e-05,
      "loss": 0.8613,
      "step": 32500
    },
    {
      "epoch": 4.77,
      "learning_rate": 2.90277978745639e-05,
      "loss": 0.858,
      "step": 33000
    },
    {
      "epoch": 4.85,
      "learning_rate": 2.8625862152124633e-05,
      "loss": 0.8589,
      "step": 33500
    },
    {
      "epoch": 4.92,
      "learning_rate": 2.8223926429685364e-05,
      "loss": 0.8572,
      "step": 34000
    },
    {
      "epoch": 4.99,
      "learning_rate": 2.78219907072461e-05,
      "loss": 0.8504,
      "step": 34500
    },
    {
      "epoch": 5.06,
      "learning_rate": 2.742005498480683e-05,
      "loss": 0.8471,
      "step": 35000
    },
    {
      "epoch": 5.06,
      "eval_loss": 0.8663927316665649,
      "eval_runtime": 335.8915,
      "eval_samples_per_second": 384.359,
      "eval_steps_per_second": 1.503,
      "step": 35000
    },
    {
      "epoch": 5.14,
      "learning_rate": 2.7018119262367564e-05,
      "loss": 0.8465,
      "step": 35500
    },
    {
      "epoch": 5.21,
      "learning_rate": 2.6616183539928296e-05,
      "loss": 0.8425,
      "step": 36000
    },
    {
      "epoch": 5.28,
      "learning_rate": 2.621424781748903e-05,
      "loss": 0.8417,
      "step": 36500
    },
    {
      "epoch": 5.35,
      "learning_rate": 2.581231209504976e-05,
      "loss": 0.8385,
      "step": 37000
    },
    {
      "epoch": 5.43,
      "learning_rate": 2.5410376372610496e-05,
      "loss": 0.8394,
      "step": 37500
    },
    {
      "epoch": 5.5,
      "learning_rate": 2.5008440650171227e-05,
      "loss": 0.8373,
      "step": 38000
    },
    {
      "epoch": 5.57,
      "learning_rate": 2.4606504927731958e-05,
      "loss": 0.8344,
      "step": 38500
    },
    {
      "epoch": 5.64,
      "learning_rate": 2.420456920529269e-05,
      "loss": 0.8347,
      "step": 39000
    },
    {
      "epoch": 5.72,
      "learning_rate": 2.3802633482853424e-05,
      "loss": 0.8324,
      "step": 39500
    },
    {
      "epoch": 5.79,
      "learning_rate": 2.3400697760414155e-05,
      "loss": 0.8315,
      "step": 40000
    },
    {
      "epoch": 5.79,
      "eval_loss": 0.8464680910110474,
      "eval_runtime": 335.9122,
      "eval_samples_per_second": 384.336,
      "eval_steps_per_second": 1.503,
      "step": 40000
    },
    {
      "epoch": 5.86,
      "learning_rate": 2.299876203797489e-05,
      "loss": 0.8305,
      "step": 40500
    },
    {
      "epoch": 5.93,
      "learning_rate": 2.259682631553562e-05,
      "loss": 0.8274,
      "step": 41000
    },
    {
      "epoch": 6.0,
      "learning_rate": 2.2194890593096355e-05,
      "loss": 0.8229,
      "step": 41500
    },
    {
      "epoch": 6.08,
      "learning_rate": 2.1792954870657086e-05,
      "loss": 0.8202,
      "step": 42000
    },
    {
      "epoch": 6.15,
      "learning_rate": 2.1391019148217817e-05,
      "loss": 0.8183,
      "step": 42500
    },
    {
      "epoch": 6.22,
      "learning_rate": 2.098908342577855e-05,
      "loss": 0.8182,
      "step": 43000
    },
    {
      "epoch": 6.29,
      "learning_rate": 2.0587147703339283e-05,
      "loss": 0.8172,
      "step": 43500
    },
    {
      "epoch": 6.37,
      "learning_rate": 2.0185211980900014e-05,
      "loss": 0.815,
      "step": 44000
    },
    {
      "epoch": 6.44,
      "learning_rate": 1.978327625846075e-05,
      "loss": 0.8136,
      "step": 44500
    },
    {
      "epoch": 6.51,
      "learning_rate": 1.938134053602148e-05,
      "loss": 0.8123,
      "step": 45000
    },
    {
      "epoch": 6.51,
      "eval_loss": 0.8314308524131775,
      "eval_runtime": 335.9134,
      "eval_samples_per_second": 384.334,
      "eval_steps_per_second": 1.503,
      "step": 45000
    },
    {
      "epoch": 6.58,
      "learning_rate": 1.8979404813582214e-05,
      "loss": 0.813,
      "step": 45500
    },
    {
      "epoch": 6.66,
      "learning_rate": 1.8577469091142946e-05,
      "loss": 0.8101,
      "step": 46000
    },
    {
      "epoch": 6.73,
      "learning_rate": 1.817553336870368e-05,
      "loss": 0.812,
      "step": 46500
    },
    {
      "epoch": 6.8,
      "learning_rate": 1.777359764626441e-05,
      "loss": 0.8069,
      "step": 47000
    },
    {
      "epoch": 6.87,
      "learning_rate": 1.7371661923825142e-05,
      "loss": 0.8048,
      "step": 47500
    },
    {
      "epoch": 6.95,
      "learning_rate": 1.6969726201385873e-05,
      "loss": 0.8064,
      "step": 48000
    },
    {
      "epoch": 7.02,
      "learning_rate": 1.6567790478946608e-05,
      "loss": 0.8019,
      "step": 48500
    },
    {
      "epoch": 7.09,
      "learning_rate": 1.616585475650734e-05,
      "loss": 0.7993,
      "step": 49000
    },
    {
      "epoch": 7.16,
      "learning_rate": 1.5763919034068074e-05,
      "loss": 0.8011,
      "step": 49500
    },
    {
      "epoch": 7.23,
      "learning_rate": 1.5361983311628805e-05,
      "loss": 0.7977,
      "step": 50000
    },
    {
      "epoch": 7.23,
      "eval_loss": 0.8229328989982605,
      "eval_runtime": 335.9175,
      "eval_samples_per_second": 384.33,
      "eval_steps_per_second": 1.503,
      "step": 50000
    },
    {
      "epoch": 7.31,
      "learning_rate": 1.4960047589189538e-05,
      "loss": 0.797,
      "step": 50500
    },
    {
      "epoch": 7.38,
      "learning_rate": 1.455811186675027e-05,
      "loss": 0.7943,
      "step": 51000
    },
    {
      "epoch": 7.45,
      "learning_rate": 1.4156176144311003e-05,
      "loss": 0.7934,
      "step": 51500
    },
    {
      "epoch": 7.52,
      "learning_rate": 1.3754240421871733e-05,
      "loss": 0.7926,
      "step": 52000
    },
    {
      "epoch": 7.6,
      "learning_rate": 1.3352304699432466e-05,
      "loss": 0.7939,
      "step": 52500
    },
    {
      "epoch": 7.67,
      "learning_rate": 1.2950368976993198e-05,
      "loss": 0.7922,
      "step": 53000
    },
    {
      "epoch": 7.74,
      "learning_rate": 1.2548433254553931e-05,
      "loss": 0.7919,
      "step": 53500
    },
    {
      "epoch": 7.81,
      "learning_rate": 1.2146497532114664e-05,
      "loss": 0.7886,
      "step": 54000
    },
    {
      "epoch": 7.89,
      "learning_rate": 1.1744561809675397e-05,
      "loss": 0.7866,
      "step": 54500
    },
    {
      "epoch": 7.96,
      "learning_rate": 1.134262608723613e-05,
      "loss": 0.7856,
      "step": 55000
    },
    {
      "epoch": 7.96,
      "eval_loss": 0.8111724257469177,
      "eval_runtime": 335.2542,
      "eval_samples_per_second": 385.09,
      "eval_steps_per_second": 1.506,
      "step": 55000
    },
    {
      "epoch": 8.03,
      "learning_rate": 1.0940690364796863e-05,
      "loss": 0.787,
      "step": 55500
    },
    {
      "epoch": 8.1,
      "learning_rate": 1.0538754642357594e-05,
      "loss": 0.7837,
      "step": 56000
    },
    {
      "epoch": 8.18,
      "learning_rate": 1.0136818919918327e-05,
      "loss": 0.7834,
      "step": 56500
    },
    {
      "epoch": 8.25,
      "learning_rate": 9.73488319747906e-06,
      "loss": 0.7817,
      "step": 57000
    },
    {
      "epoch": 8.32,
      "learning_rate": 9.332947475039792e-06,
      "loss": 0.7803,
      "step": 57500
    },
    {
      "epoch": 8.39,
      "learning_rate": 8.931011752600525e-06,
      "loss": 0.7793,
      "step": 58000
    },
    {
      "epoch": 8.46,
      "learning_rate": 8.529076030161256e-06,
      "loss": 0.7802,
      "step": 58500
    },
    {
      "epoch": 8.54,
      "learning_rate": 8.127140307721989e-06,
      "loss": 0.7788,
      "step": 59000
    },
    {
      "epoch": 8.61,
      "learning_rate": 7.725204585282722e-06,
      "loss": 0.7754,
      "step": 59500
    },
    {
      "epoch": 8.68,
      "learning_rate": 7.323268862843455e-06,
      "loss": 0.776,
      "step": 60000
    },
    {
      "epoch": 8.68,
      "eval_loss": 0.8020525574684143,
      "eval_runtime": 335.4375,
      "eval_samples_per_second": 384.879,
      "eval_steps_per_second": 1.505,
      "step": 60000
    },
    {
      "epoch": 8.75,
      "learning_rate": 6.921333140404188e-06,
      "loss": 0.7767,
      "step": 60500
    },
    {
      "epoch": 8.83,
      "learning_rate": 6.519397417964919e-06,
      "loss": 0.7766,
      "step": 61000
    },
    {
      "epoch": 8.9,
      "learning_rate": 6.117461695525652e-06,
      "loss": 0.7727,
      "step": 61500
    },
    {
      "epoch": 8.97,
      "learning_rate": 5.7155259730863845e-06,
      "loss": 0.7739,
      "step": 62000
    },
    {
      "epoch": 9.04,
      "learning_rate": 5.3135902506471165e-06,
      "loss": 0.7723,
      "step": 62500
    },
    {
      "epoch": 9.12,
      "learning_rate": 4.911654528207849e-06,
      "loss": 0.771,
      "step": 63000
    },
    {
      "epoch": 9.19,
      "learning_rate": 4.509718805768582e-06,
      "loss": 0.7674,
      "step": 63500
    },
    {
      "epoch": 9.26,
      "learning_rate": 4.107783083329314e-06,
      "loss": 0.772,
      "step": 64000
    },
    {
      "epoch": 9.33,
      "learning_rate": 3.705847360890047e-06,
      "loss": 0.7683,
      "step": 64500
    },
    {
      "epoch": 9.41,
      "learning_rate": 3.303911638450779e-06,
      "loss": 0.7705,
      "step": 65000
    },
    {
      "epoch": 9.41,
      "eval_loss": 0.7940782904624939,
      "eval_runtime": 335.4617,
      "eval_samples_per_second": 384.852,
      "eval_steps_per_second": 1.505,
      "step": 65000
    }
  ],
  "logging_steps": 500,
  "max_steps": 69110,
  "num_train_epochs": 10,
  "save_steps": 5000,
  "total_flos": 1.2759454800715415e+18,
  "trial_name": null,
  "trial_params": null
}