{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6694302516147691,
  "eval_steps": 500,
  "global_step": 434,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007712330087727755,
      "grad_norm": 0.84765625,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 1.164,
      "memory/device_reserved (GiB)": 246.45,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 5,
      "tokens_per_second_per_gpu": 500.9,
      "total_tokens": 1287402
    },
    {
      "epoch": 0.01542466017545551,
      "grad_norm": 0.9765625,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 1.1463,
      "memory/device_reserved (GiB)": 246.57,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 10,
      "tokens_per_second_per_gpu": 458.45,
      "total_tokens": 2568739
    },
    {
      "epoch": 0.023136990263183263,
      "grad_norm": 0.91015625,
      "learning_rate": 7.000000000000001e-07,
      "loss": 1.1346,
      "memory/device_reserved (GiB)": 246.57,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 15,
      "tokens_per_second_per_gpu": 487.44,
      "total_tokens": 3789106
    },
    {
      "epoch": 0.03084932035091102,
      "grad_norm": 0.90625,
      "learning_rate": 9.500000000000001e-07,
      "loss": 1.1363,
      "memory/device_reserved (GiB)": 246.57,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 20,
      "tokens_per_second_per_gpu": 514.12,
      "total_tokens": 5093129
    },
    {
      "epoch": 0.038561650438638774,
      "grad_norm": 1.0,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 1.118,
      "memory/device_reserved (GiB)": 246.57,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 25,
      "tokens_per_second_per_gpu": 449.18,
      "total_tokens": 6340204
    },
    {
      "epoch": 0.046273980526366526,
      "grad_norm": 0.859375,
      "learning_rate": 1.45e-06,
      "loss": 1.1324,
      "memory/device_reserved (GiB)": 246.57,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 30,
      "tokens_per_second_per_gpu": 450.16,
      "total_tokens": 7552866
    },
    {
      "epoch": 0.053986310614094285,
      "grad_norm": 0.76171875,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 1.1225,
      "memory/device_reserved (GiB)": 246.57,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 35,
      "tokens_per_second_per_gpu": 517.87,
      "total_tokens": 8796011
    },
    {
      "epoch": 0.06169864070182204,
      "grad_norm": 0.8046875,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 1.0583,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 40,
      "tokens_per_second_per_gpu": 478.29,
      "total_tokens": 10026103
    },
    {
      "epoch": 0.06941097078954979,
      "grad_norm": 0.7578125,
      "learning_rate": 2.2e-06,
      "loss": 1.0965,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 45,
      "tokens_per_second_per_gpu": 470.47,
      "total_tokens": 11340325
    },
    {
      "epoch": 0.07712330087727755,
      "grad_norm": 0.65625,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 1.1055,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 50,
      "tokens_per_second_per_gpu": 447.61,
      "total_tokens": 12618872
    },
    {
      "epoch": 0.08483563096500531,
      "grad_norm": 0.61328125,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 1.0971,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 55,
      "tokens_per_second_per_gpu": 468.97,
      "total_tokens": 13906030
    },
    {
      "epoch": 0.09254796105273305,
      "grad_norm": 0.51171875,
      "learning_rate": 2.95e-06,
      "loss": 1.0837,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 60,
      "tokens_per_second_per_gpu": 443.28,
      "total_tokens": 15157386
    },
    {
      "epoch": 0.10026029114046081,
      "grad_norm": 0.474609375,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 1.081,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 65,
      "tokens_per_second_per_gpu": 434.55,
      "total_tokens": 16483820
    },
    {
      "epoch": 0.10797262122818857,
      "grad_norm": 0.486328125,
      "learning_rate": 3.45e-06,
      "loss": 1.0501,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 70,
      "tokens_per_second_per_gpu": 446.29,
      "total_tokens": 17761687
    },
    {
      "epoch": 0.11568495131591632,
      "grad_norm": 0.37890625,
      "learning_rate": 3.7e-06,
      "loss": 1.0402,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 75,
      "tokens_per_second_per_gpu": 499.21,
      "total_tokens": 19043878
    },
    {
      "epoch": 0.12339728140364407,
      "grad_norm": 0.42578125,
      "learning_rate": 3.95e-06,
      "loss": 1.0604,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 80,
      "tokens_per_second_per_gpu": 453.36,
      "total_tokens": 20314866
    },
    {
      "epoch": 0.13110961149137182,
      "grad_norm": 0.373046875,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 1.0382,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 85,
      "tokens_per_second_per_gpu": 453.89,
      "total_tokens": 21555557
    },
    {
      "epoch": 0.13882194157909958,
      "grad_norm": 0.330078125,
      "learning_rate": 4.450000000000001e-06,
      "loss": 1.0753,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 90,
      "tokens_per_second_per_gpu": 510.81,
      "total_tokens": 22834764
    },
    {
      "epoch": 0.14653427166682734,
      "grad_norm": 0.373046875,
      "learning_rate": 4.7e-06,
      "loss": 1.021,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 95,
      "tokens_per_second_per_gpu": 477.3,
      "total_tokens": 24078719
    },
    {
      "epoch": 0.1542466017545551,
      "grad_norm": 0.271484375,
      "learning_rate": 4.95e-06,
      "loss": 1.0379,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 100,
      "tokens_per_second_per_gpu": 495.26,
      "total_tokens": 25368597
    },
    {
      "epoch": 0.16195893184228285,
      "grad_norm": 0.283203125,
      "learning_rate": 4.999950130642566e-06,
      "loss": 1.0399,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 105,
      "tokens_per_second_per_gpu": 458.71,
      "total_tokens": 26667382
    },
    {
      "epoch": 0.16967126193001061,
      "grad_norm": 0.27734375,
      "learning_rate": 4.999747540342916e-06,
      "loss": 1.0052,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 110,
      "tokens_per_second_per_gpu": 454.85,
      "total_tokens": 27935546
    },
    {
      "epoch": 0.17738359201773837,
      "grad_norm": 0.287109375,
      "learning_rate": 4.999389126939546e-06,
      "loss": 0.9942,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 115,
      "tokens_per_second_per_gpu": 488.42,
      "total_tokens": 29232061
    },
    {
      "epoch": 0.1850959221054661,
      "grad_norm": 0.263671875,
      "learning_rate": 4.998874916411828e-06,
      "loss": 1.0166,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 120,
      "tokens_per_second_per_gpu": 479.94,
      "total_tokens": 30496024
    },
    {
      "epoch": 0.19280825219319386,
      "grad_norm": 0.271484375,
      "learning_rate": 4.99820494603199e-06,
      "loss": 0.9909,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 125,
      "tokens_per_second_per_gpu": 462.61,
      "total_tokens": 31699870
    },
    {
      "epoch": 0.20052058228092162,
      "grad_norm": 0.259765625,
      "learning_rate": 4.997379264362414e-06,
      "loss": 0.9874,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 130,
      "tokens_per_second_per_gpu": 467.47,
      "total_tokens": 32956914
    },
    {
      "epoch": 0.20823291236864938,
      "grad_norm": 0.251953125,
      "learning_rate": 4.996397931252118e-06,
      "loss": 1.0041,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 135,
      "tokens_per_second_per_gpu": 459.93,
      "total_tokens": 34182303
    },
    {
      "epoch": 0.21594524245637714,
      "grad_norm": 0.2314453125,
      "learning_rate": 4.9952610178324206e-06,
      "loss": 1.0227,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 140,
      "tokens_per_second_per_gpu": 510.67,
      "total_tokens": 35453388
    },
    {
      "epoch": 0.2236575725441049,
      "grad_norm": 0.2734375,
      "learning_rate": 4.993968606511777e-06,
      "loss": 1.0007,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 145,
      "tokens_per_second_per_gpu": 448.95,
      "total_tokens": 36723948
    },
    {
      "epoch": 0.23136990263183263,
      "grad_norm": 0.24609375,
      "learning_rate": 4.9925207909698115e-06,
      "loss": 0.9714,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 150,
      "tokens_per_second_per_gpu": 460.62,
      "total_tokens": 37953157
    },
    {
      "epoch": 0.2390822327195604,
      "grad_norm": 0.244140625,
      "learning_rate": 4.990917676150527e-06,
      "loss": 1.0193,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 155,
      "tokens_per_second_per_gpu": 509.45,
      "total_tokens": 39278804
    },
    {
      "epoch": 0.24679456280728815,
      "grad_norm": 0.2451171875,
      "learning_rate": 4.989159378254698e-06,
      "loss": 0.9887,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 160,
      "tokens_per_second_per_gpu": 491.62,
      "total_tokens": 40562259
    },
    {
      "epoch": 0.2545068928950159,
      "grad_norm": 0.2353515625,
      "learning_rate": 4.9872460247314455e-06,
      "loss": 0.9637,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 165,
      "tokens_per_second_per_gpu": 423.64,
      "total_tokens": 41779016
    },
    {
      "epoch": 0.26221922298274364,
      "grad_norm": 0.248046875,
      "learning_rate": 4.9851777542690004e-06,
      "loss": 0.9812,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 170,
      "tokens_per_second_per_gpu": 504.59,
      "total_tokens": 43071677
    },
    {
      "epoch": 0.2699315530704714,
      "grad_norm": 0.2412109375,
      "learning_rate": 4.9829547167846515e-06,
      "loss": 0.9513,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 175,
      "tokens_per_second_per_gpu": 440.97,
      "total_tokens": 44311747
    },
    {
      "epoch": 0.27764388315819916,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.9805770734138785e-06,
      "loss": 0.9504,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 180,
      "tokens_per_second_per_gpu": 463.76,
      "total_tokens": 45513801
    },
    {
      "epoch": 0.2853562132459269,
      "grad_norm": 0.2490234375,
      "learning_rate": 4.978044996498671e-06,
      "loss": 0.9967,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 185,
      "tokens_per_second_per_gpu": 450.72,
      "total_tokens": 46737916
    },
    {
      "epoch": 0.2930685433336547,
      "grad_norm": 0.32421875,
      "learning_rate": 4.975358669575039e-06,
      "loss": 0.9927,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 190,
      "tokens_per_second_per_gpu": 482.76,
      "total_tokens": 48020729
    },
    {
      "epoch": 0.30078087342138243,
      "grad_norm": 0.259765625,
      "learning_rate": 4.972518287359707e-06,
      "loss": 0.9893,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 195,
      "tokens_per_second_per_gpu": 437.14,
      "total_tokens": 49235773
    },
    {
      "epoch": 0.3084932035091102,
      "grad_norm": 0.2314453125,
      "learning_rate": 4.969524055735999e-06,
      "loss": 0.9344,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 200,
      "tokens_per_second_per_gpu": 428.14,
      "total_tokens": 50485073
    },
    {
      "epoch": 0.31620553359683795,
      "grad_norm": 0.2294921875,
      "learning_rate": 4.9663761917389195e-06,
      "loss": 0.9669,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 205,
      "tokens_per_second_per_gpu": 513.34,
      "total_tokens": 51774221
    },
    {
      "epoch": 0.3239178636845657,
      "grad_norm": 0.2373046875,
      "learning_rate": 4.9630749235394174e-06,
      "loss": 1.0206,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 210,
      "tokens_per_second_per_gpu": 470.02,
      "total_tokens": 53055366
    },
    {
      "epoch": 0.33163019377229347,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.959620490427851e-06,
      "loss": 1.0136,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 215,
      "tokens_per_second_per_gpu": 474.09,
      "total_tokens": 54318548
    },
    {
      "epoch": 0.33934252386002123,
      "grad_norm": 0.2353515625,
      "learning_rate": 4.95601314279664e-06,
      "loss": 0.98,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 220,
      "tokens_per_second_per_gpu": 441.47,
      "total_tokens": 55573941
    },
    {
      "epoch": 0.347054853947749,
      "grad_norm": 0.21875,
      "learning_rate": 4.952253142122116e-06,
      "loss": 0.9428,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 225,
      "tokens_per_second_per_gpu": 446.8,
      "total_tokens": 56816147
    },
    {
      "epoch": 0.35476718403547675,
      "grad_norm": 0.236328125,
      "learning_rate": 4.948340760945575e-06,
      "loss": 0.9925,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 230,
      "tokens_per_second_per_gpu": 440.95,
      "total_tokens": 58018929
    },
    {
      "epoch": 0.36247951412320445,
      "grad_norm": 0.216796875,
      "learning_rate": 4.944276282853515e-06,
      "loss": 1.0056,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 235,
      "tokens_per_second_per_gpu": 437.83,
      "total_tokens": 59236706
    },
    {
      "epoch": 0.3701918442109322,
      "grad_norm": 0.216796875,
      "learning_rate": 4.940060002457083e-06,
      "loss": 1.0271,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 240,
      "tokens_per_second_per_gpu": 478.28,
      "total_tokens": 60523938
    },
    {
      "epoch": 0.37790417429865997,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.935692225370723e-06,
      "loss": 0.9817,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 245,
      "tokens_per_second_per_gpu": 461.75,
      "total_tokens": 61782785
    },
    {
      "epoch": 0.3856165043863877,
      "grad_norm": 0.2265625,
      "learning_rate": 4.9311732681900195e-06,
      "loss": 1.0036,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 250,
      "tokens_per_second_per_gpu": 455.35,
      "total_tokens": 62995481
    },
    {
      "epoch": 0.3933288344741155,
      "grad_norm": 0.2109375,
      "learning_rate": 4.926503458468756e-06,
      "loss": 1.0026,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 255,
      "tokens_per_second_per_gpu": 485.03,
      "total_tokens": 64329450
    },
    {
      "epoch": 0.40104116456184324,
      "grad_norm": 0.2294921875,
      "learning_rate": 4.921683134695161e-06,
      "loss": 1.0022,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 260,
      "tokens_per_second_per_gpu": 462.42,
      "total_tokens": 65577746
    },
    {
      "epoch": 0.408753494649571,
      "grad_norm": 0.2158203125,
      "learning_rate": 4.916712646267386e-06,
      "loss": 1.0291,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 265,
      "tokens_per_second_per_gpu": 467.37,
      "total_tokens": 66836212
    },
    {
      "epoch": 0.41646582473729876,
      "grad_norm": 0.2333984375,
      "learning_rate": 4.91159235346817e-06,
      "loss": 0.9912,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 270,
      "tokens_per_second_per_gpu": 515.37,
      "total_tokens": 68143788
    },
    {
      "epoch": 0.4241781548250265,
      "grad_norm": 0.2275390625,
      "learning_rate": 4.906322627438728e-06,
      "loss": 0.965,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 275,
      "tokens_per_second_per_gpu": 482.58,
      "total_tokens": 69419203
    },
    {
      "epoch": 0.4318904849127543,
      "grad_norm": 0.251953125,
      "learning_rate": 4.900903850151851e-06,
      "loss": 0.9919,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 280,
      "tokens_per_second_per_gpu": 476.95,
      "total_tokens": 70708858
    },
    {
      "epoch": 0.43960281500048204,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.895336414384216e-06,
      "loss": 1.015,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 285,
      "tokens_per_second_per_gpu": 488.29,
      "total_tokens": 71965415
    },
    {
      "epoch": 0.4473151450882098,
      "grad_norm": 0.2470703125,
      "learning_rate": 4.889620723687917e-06,
      "loss": 0.998,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 290,
      "tokens_per_second_per_gpu": 501.89,
      "total_tokens": 73325345
    },
    {
      "epoch": 0.45502747517593756,
      "grad_norm": 0.24609375,
      "learning_rate": 4.883757192361212e-06,
      "loss": 0.9773,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 295,
      "tokens_per_second_per_gpu": 469.94,
      "total_tokens": 74612028
    },
    {
      "epoch": 0.46273980526366526,
      "grad_norm": 0.2255859375,
      "learning_rate": 4.8777462454184985e-06,
      "loss": 0.9877,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 300,
      "tokens_per_second_per_gpu": 476.32,
      "total_tokens": 75892822
    },
    {
      "epoch": 0.470452135351393,
      "grad_norm": 0.244140625,
      "learning_rate": 4.871588318559497e-06,
      "loss": 1.0243,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 305,
      "tokens_per_second_per_gpu": 451.66,
      "total_tokens": 77185408
    },
    {
      "epoch": 0.4781644654391208,
      "grad_norm": 0.220703125,
      "learning_rate": 4.86528385813768e-06,
      "loss": 0.9481,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 310,
      "tokens_per_second_per_gpu": 441.19,
      "total_tokens": 78404970
    },
    {
      "epoch": 0.48587679552684854,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.8588333211279105e-06,
      "loss": 0.9578,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 315,
      "tokens_per_second_per_gpu": 466.55,
      "total_tokens": 79635002
    },
    {
      "epoch": 0.4935891256145763,
      "grad_norm": 0.212890625,
      "learning_rate": 4.852237175093324e-06,
      "loss": 1.0071,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 320,
      "tokens_per_second_per_gpu": 452.13,
      "total_tokens": 80933611
    },
    {
      "epoch": 0.5013014557023041,
      "grad_norm": 0.2177734375,
      "learning_rate": 4.845495898151433e-06,
      "loss": 0.944,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 325,
      "tokens_per_second_per_gpu": 433.83,
      "total_tokens": 82145228
    },
    {
      "epoch": 0.5090137857900318,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.838609978939473e-06,
      "loss": 1.0259,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 330,
      "tokens_per_second_per_gpu": 453.14,
      "total_tokens": 83357951
    },
    {
      "epoch": 0.5167261158777595,
      "grad_norm": 0.236328125,
      "learning_rate": 4.831579916578984e-06,
      "loss": 0.9701,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 335,
      "tokens_per_second_per_gpu": 466.64,
      "total_tokens": 84573768
    },
    {
      "epoch": 0.5244384459654873,
      "grad_norm": 0.2060546875,
      "learning_rate": 4.824406220639634e-06,
      "loss": 1.005,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 340,
      "tokens_per_second_per_gpu": 460.44,
      "total_tokens": 85851432
    },
    {
      "epoch": 0.532150776053215,
      "grad_norm": 0.2138671875,
      "learning_rate": 4.817089411102277e-06,
      "loss": 0.9687,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 345,
      "tokens_per_second_per_gpu": 440.77,
      "total_tokens": 87088346
    },
    {
      "epoch": 0.5398631061409428,
      "grad_norm": 0.2294921875,
      "learning_rate": 4.809630018321269e-06,
      "loss": 0.9483,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 350,
      "tokens_per_second_per_gpu": 480.0,
      "total_tokens": 88317739
    },
    {
      "epoch": 0.5475754362286706,
      "grad_norm": 0.271484375,
      "learning_rate": 4.802028582986024e-06,
      "loss": 0.9666,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 355,
      "tokens_per_second_per_gpu": 476.5,
      "total_tokens": 89577574
    },
    {
      "epoch": 0.5552877663163983,
      "grad_norm": 0.240234375,
      "learning_rate": 4.794285656081819e-06,
      "loss": 0.9507,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 360,
      "tokens_per_second_per_gpu": 497.81,
      "total_tokens": 90828267
    },
    {
      "epoch": 0.5630000964041261,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.786401798849861e-06,
      "loss": 0.9405,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 365,
      "tokens_per_second_per_gpu": 481.32,
      "total_tokens": 92116882
    },
    {
      "epoch": 0.5707124264918538,
      "grad_norm": 0.2392578125,
      "learning_rate": 4.778377582746601e-06,
      "loss": 1.0099,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 370,
      "tokens_per_second_per_gpu": 459.13,
      "total_tokens": 93354393
    },
    {
      "epoch": 0.5784247565795816,
      "grad_norm": 0.220703125,
      "learning_rate": 4.770213589402317e-06,
      "loss": 0.9608,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 375,
      "tokens_per_second_per_gpu": 485.77,
      "total_tokens": 94596707
    },
    {
      "epoch": 0.5861370866673093,
      "grad_norm": 0.2109375,
      "learning_rate": 4.7619104105789525e-06,
      "loss": 0.9747,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 380,
      "tokens_per_second_per_gpu": 449.15,
      "total_tokens": 95823690
    },
    {
      "epoch": 0.5938494167550371,
      "grad_norm": 0.234375,
      "learning_rate": 4.753468648127223e-06,
      "loss": 0.9593,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 385,
      "tokens_per_second_per_gpu": 462.81,
      "total_tokens": 97064318
    },
    {
      "epoch": 0.6015617468427649,
      "grad_norm": 0.2041015625,
      "learning_rate": 4.74488891394299e-06,
      "loss": 0.984,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 390,
      "tokens_per_second_per_gpu": 467.19,
      "total_tokens": 98318680
    },
    {
      "epoch": 0.6092740769304926,
      "grad_norm": 0.2216796875,
      "learning_rate": 4.7361718299229125e-06,
      "loss": 0.9446,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 395,
      "tokens_per_second_per_gpu": 439.44,
      "total_tokens": 99522489
    },
    {
      "epoch": 0.6169864070182204,
      "grad_norm": 0.2265625,
      "learning_rate": 4.727318027919364e-06,
      "loss": 0.9796,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 400,
      "tokens_per_second_per_gpu": 441.41,
      "total_tokens": 100785080
    },
    {
      "epoch": 0.6246987371059481,
      "grad_norm": 0.23046875,
      "learning_rate": 4.718328149694636e-06,
      "loss": 1.0054,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 405,
      "tokens_per_second_per_gpu": 497.07,
      "total_tokens": 102084350
    },
    {
      "epoch": 0.6324110671936759,
      "grad_norm": 0.236328125,
      "learning_rate": 4.70920284687442e-06,
      "loss": 0.9891,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 410,
      "tokens_per_second_per_gpu": 482.3,
      "total_tokens": 103357801
    },
    {
      "epoch": 0.6401233972814037,
      "grad_norm": 0.2294921875,
      "learning_rate": 4.699942780900575e-06,
      "loss": 0.9616,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 415,
      "tokens_per_second_per_gpu": 499.49,
      "total_tokens": 104613606
    },
    {
      "epoch": 0.6478357273691314,
      "grad_norm": 0.212890625,
      "learning_rate": 4.6905486229831814e-06,
      "loss": 0.9559,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 420,
      "tokens_per_second_per_gpu": 500.21,
      "total_tokens": 105880634
    },
    {
      "epoch": 0.6555480574568592,
      "grad_norm": 0.220703125,
      "learning_rate": 4.68102105405189e-06,
      "loss": 0.9997,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 425,
      "tokens_per_second_per_gpu": 473.56,
      "total_tokens": 107115671
    },
    {
      "epoch": 0.6632603875445869,
      "grad_norm": 0.212890625,
      "learning_rate": 4.671360764706566e-06,
      "loss": 0.9785,
      "memory/device_reserved (GiB)": 246.75,
      "memory/max_active (GiB)": 211.16,
      "memory/max_allocated (GiB)": 211.16,
      "step": 430,
      "tokens_per_second_per_gpu": 507.32,
      "total_tokens": 108431881
    }
  ],
  "logging_steps": 5,
  "max_steps": 1945,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 217,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.746652599408562e+20,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}