{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.717472118959108,
  "eval_steps": 50,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 0.0,
      "loss": 1.3449,
      "step": 1
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 1.2712,
      "step": 10
    },
    {
      "epoch": 0.07,
      "learning_rate": 9.600000000000001e-06,
      "loss": 1.1556,
      "step": 20
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.56e-05,
      "loss": 1.0713,
      "step": 30
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.16e-05,
      "loss": 0.9959,
      "step": 40
    },
    {
      "epoch": 0.19,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 1.0538,
      "step": 50
    },
    {
      "epoch": 0.19,
      "eval_loss": 1.1364344358444214,
      "eval_runtime": 14.0151,
      "eval_samples_per_second": 8.562,
      "eval_steps_per_second": 2.141,
      "step": 50
    },
    {
      "epoch": 0.22,
      "learning_rate": 2.9997949574887035e-05,
      "loss": 0.7802,
      "step": 60
    },
    {
      "epoch": 0.26,
      "learning_rate": 2.9981549537224573e-05,
      "loss": 0.7425,
      "step": 70
    },
    {
      "epoch": 0.3,
      "learning_rate": 2.994876739510005e-05,
      "loss": 0.8096,
      "step": 80
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.9899638995304575e-05,
      "loss": 0.71,
      "step": 90
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.9834218059022027e-05,
      "loss": 0.7744,
      "step": 100
    },
    {
      "epoch": 0.37,
      "eval_loss": 0.7776542901992798,
      "eval_runtime": 14.0597,
      "eval_samples_per_second": 8.535,
      "eval_steps_per_second": 2.134,
      "step": 100
    },
    {
      "epoch": 0.41,
      "learning_rate": 2.9752576123085737e-05,
      "loss": 0.6735,
      "step": 110
    },
    {
      "epoch": 0.45,
      "learning_rate": 2.9654802461753992e-05,
      "loss": 0.6107,
      "step": 120
    },
    {
      "epoch": 0.48,
      "learning_rate": 2.9541003989089956e-05,
      "loss": 0.6387,
      "step": 130
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.9411305142052725e-05,
      "loss": 0.5779,
      "step": 140
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.9265847744427305e-05,
      "loss": 0.5936,
      "step": 150
    },
    {
      "epoch": 0.56,
      "eval_loss": 0.6507006287574768,
      "eval_runtime": 14.0635,
      "eval_samples_per_second": 8.533,
      "eval_steps_per_second": 2.133,
      "step": 150
    },
    {
      "epoch": 0.59,
      "learning_rate": 2.910479085174242e-05,
      "loss": 0.5965,
      "step": 160
    },
    {
      "epoch": 0.63,
      "learning_rate": 2.8928310577345608e-05,
      "loss": 0.5969,
      "step": 170
    },
    {
      "epoch": 0.67,
      "learning_rate": 2.873659989982586e-05,
      "loss": 0.5735,
      "step": 180
    },
    {
      "epoch": 0.71,
      "learning_rate": 2.8529868451994387e-05,
      "loss": 0.5535,
      "step": 190
    },
    {
      "epoch": 0.74,
      "learning_rate": 2.830834229165418e-05,
      "loss": 0.5449,
      "step": 200
    },
    {
      "epoch": 0.74,
      "eval_loss": 0.608721137046814,
      "eval_runtime": 14.0592,
      "eval_samples_per_second": 8.535,
      "eval_steps_per_second": 2.134,
      "step": 200
    },
    {
      "epoch": 0.78,
      "learning_rate": 2.807226365440916e-05,
      "loss": 0.5445,
      "step": 210
    },
    {
      "epoch": 0.82,
      "learning_rate": 2.7821890688783088e-05,
      "loss": 0.6112,
      "step": 220
    },
    {
      "epoch": 0.86,
      "learning_rate": 2.7557497173937928e-05,
      "loss": 0.5294,
      "step": 230
    },
    {
      "epoch": 0.89,
      "learning_rate": 2.727937222030039e-05,
      "loss": 0.4957,
      "step": 240
    },
    {
      "epoch": 0.93,
      "learning_rate": 2.698781995342387e-05,
      "loss": 0.501,
      "step": 250
    },
    {
      "epoch": 0.93,
      "eval_loss": 0.5840339660644531,
      "eval_runtime": 14.0673,
      "eval_samples_per_second": 8.53,
      "eval_steps_per_second": 2.133,
      "step": 250
    },
    {
      "epoch": 0.97,
      "learning_rate": 2.668315918143169e-05,
      "loss": 0.5568,
      "step": 260
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.6365723046405022e-05,
      "loss": 0.5425,
      "step": 270
    },
    {
      "epoch": 1.04,
      "learning_rate": 2.6035858660096975e-05,
      "loss": 0.5089,
      "step": 280
    },
    {
      "epoch": 1.08,
      "learning_rate": 2.5693926724370958e-05,
      "loss": 0.5014,
      "step": 290
    },
    {
      "epoch": 1.12,
      "learning_rate": 2.534030113677849e-05,
      "loss": 0.5752,
      "step": 300
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.555169939994812,
      "eval_runtime": 14.0736,
      "eval_samples_per_second": 8.527,
      "eval_steps_per_second": 2.132,
      "step": 300
    },
    {
      "epoch": 1.15,
      "learning_rate": 2.4975368581707724e-05,
      "loss": 0.4937,
      "step": 310
    },
    {
      "epoch": 1.19,
      "learning_rate": 2.459952810754975e-05,
      "loss": 0.446,
      "step": 320
    },
    {
      "epoch": 1.23,
      "learning_rate": 2.4213190690345018e-05,
      "loss": 0.4632,
      "step": 330
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.3816778784387097e-05,
      "loss": 0.4789,
      "step": 340
    },
    {
      "epoch": 1.3,
      "learning_rate": 2.3410725860275092e-05,
      "loss": 0.4542,
      "step": 350
    },
    {
      "epoch": 1.3,
      "eval_loss": 0.5419090390205383,
      "eval_runtime": 14.073,
      "eval_samples_per_second": 8.527,
      "eval_steps_per_second": 2.132,
      "step": 350
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.2995475930919907e-05,
      "loss": 0.4636,
      "step": 360
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.257148306602266e-05,
      "loss": 0.5061,
      "step": 370
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.2139210895556104e-05,
      "loss": 0.4431,
      "step": 380
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.16991321027921e-05,
      "loss": 0.4565,
      "step": 390
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.1251727907429357e-05,
      "loss": 0.5115,
      "step": 400
    },
    {
      "epoch": 1.49,
      "eval_loss": 0.5242533087730408,
      "eval_runtime": 14.071,
      "eval_samples_per_second": 8.528,
      "eval_steps_per_second": 2.132,
      "step": 400
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.0797487539386782e-05,
      "loss": 0.4748,
      "step": 410
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.033690770383775e-05,
      "loss": 0.4308,
      "step": 420
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.9870492038070255e-05,
      "loss": 0.4808,
      "step": 430
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.9398750560766973e-05,
      "loss": 0.5061,
      "step": 440
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.8922199114307297e-05,
      "loss": 0.4224,
      "step": 450
    },
    {
      "epoch": 1.67,
      "eval_loss": 0.518813967704773,
      "eval_runtime": 14.0743,
      "eval_samples_per_second": 8.526,
      "eval_steps_per_second": 2.132,
      "step": 450
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.8441358800701276e-05,
      "loss": 0.4525,
      "step": 460
    },
    {
      "epoch": 1.75,
      "learning_rate": 1.7956755411772203e-05,
      "loss": 0.4854,
      "step": 470
    },
    {
      "epoch": 1.78,
      "learning_rate": 1.746891885421101e-05,
      "loss": 0.4529,
      "step": 480
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.6978382570131037e-05,
      "loss": 0.4786,
      "step": 490
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.6485682953756945e-05,
      "loss": 0.4486,
      "step": 500
    },
    {
      "epoch": 1.86,
      "eval_loss": 0.5055068731307983,
      "eval_runtime": 14.0709,
      "eval_samples_per_second": 8.528,
      "eval_steps_per_second": 2.132,
      "step": 500
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.5991358764885492e-05,
      "loss": 0.4303,
      "step": 510
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.549595053975962e-05,
      "loss": 0.4589,
      "step": 520
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.5e-05,
      "loss": 0.4715,
      "step": 530
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.4504049460240376e-05,
      "loss": 0.4354,
      "step": 540
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.400864123511451e-05,
      "loss": 0.3865,
      "step": 550
    },
    {
      "epoch": 2.04,
      "eval_loss": 0.5038172006607056,
      "eval_runtime": 14.0798,
      "eval_samples_per_second": 8.523,
      "eval_steps_per_second": 2.131,
      "step": 550
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.351431704624306e-05,
      "loss": 0.4119,
      "step": 560
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.3021617429868966e-05,
      "loss": 0.4405,
      "step": 570
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.2531081145788989e-05,
      "loss": 0.4435,
      "step": 580
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.2043244588227797e-05,
      "loss": 0.4083,
      "step": 590
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.1558641199298728e-05,
      "loss": 0.4193,
      "step": 600
    },
    {
      "epoch": 2.23,
      "eval_loss": 0.5048139691352844,
      "eval_runtime": 14.07,
      "eval_samples_per_second": 8.529,
      "eval_steps_per_second": 2.132,
      "step": 600
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.1077800885692704e-05,
      "loss": 0.4138,
      "step": 610
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.0601249439233031e-05,
      "loss": 0.4357,
      "step": 620
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.0129507961929749e-05,
      "loss": 0.475,
      "step": 630
    },
    {
      "epoch": 2.38,
      "learning_rate": 9.663092296162252e-06,
      "loss": 0.4294,
      "step": 640
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.20251246061322e-06,
      "loss": 0.4294,
      "step": 650
    },
    {
      "epoch": 2.42,
      "eval_loss": 0.49953868985176086,
      "eval_runtime": 14.0784,
      "eval_samples_per_second": 8.524,
      "eval_steps_per_second": 2.131,
      "step": 650
    },
    {
      "epoch": 2.45,
      "learning_rate": 8.748272092570648e-06,
      "loss": 0.4213,
      "step": 660
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.300867897207903e-06,
      "loss": 0.4353,
      "step": 670
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.860789104443897e-06,
      "loss": 0.413,
      "step": 680
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.4285169339773486e-06,
      "loss": 0.4153,
      "step": 690
    },
    {
      "epoch": 2.6,
      "learning_rate": 7.0045240690800975e-06,
      "loss": 0.4077,
      "step": 700
    },
    {
      "epoch": 2.6,
      "eval_loss": 0.5014118552207947,
      "eval_runtime": 14.0698,
      "eval_samples_per_second": 8.529,
      "eval_steps_per_second": 2.132,
      "step": 700
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.589274139724911e-06,
      "loss": 0.4025,
      "step": 710
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.1832212156129045e-06,
      "loss": 0.4911,
      "step": 720
    },
    {
      "epoch": 2.71,
      "learning_rate": 5.786809309654983e-06,
      "loss": 0.4088,
      "step": 730
    },
    {
      "epoch": 2.75,
      "learning_rate": 5.400471892450251e-06,
      "loss": 0.4276,
      "step": 740
    },
    {
      "epoch": 2.79,
      "learning_rate": 5.024631418292275e-06,
      "loss": 0.4667,
      "step": 750
    },
    {
      "epoch": 2.79,
      "eval_loss": 0.49848347902297974,
      "eval_runtime": 14.0784,
      "eval_samples_per_second": 8.524,
      "eval_steps_per_second": 2.131,
      "step": 750
    },
    {
      "epoch": 2.83,
      "learning_rate": 4.659698863221513e-06,
      "loss": 0.4215,
      "step": 760
    },
    {
      "epoch": 2.86,
      "learning_rate": 4.306073275629045e-06,
      "loss": 0.4274,
      "step": 770
    },
    {
      "epoch": 2.9,
      "learning_rate": 3.964141339903026e-06,
      "loss": 0.3821,
      "step": 780
    },
    {
      "epoch": 2.94,
      "learning_rate": 3.634276953594982e-06,
      "loss": 0.4385,
      "step": 790
    },
    {
      "epoch": 2.97,
      "learning_rate": 3.3168408185683153e-06,
      "loss": 0.4226,
      "step": 800
    },
    {
      "epoch": 2.97,
      "eval_loss": 0.4937375485897064,
      "eval_runtime": 14.0647,
      "eval_samples_per_second": 8.532,
      "eval_steps_per_second": 2.133,
      "step": 800
    },
    {
      "epoch": 3.01,
      "learning_rate": 3.0121800465761298e-06,
      "loss": 0.4213,
      "step": 810
    },
    {
      "epoch": 3.05,
      "learning_rate": 2.720627779699615e-06,
      "loss": 0.4184,
      "step": 820
    },
    {
      "epoch": 3.09,
      "learning_rate": 2.442502826062072e-06,
      "loss": 0.4083,
      "step": 830
    },
    {
      "epoch": 3.12,
      "learning_rate": 2.1781093112169132e-06,
      "loss": 0.4407,
      "step": 840
    },
    {
      "epoch": 3.16,
      "learning_rate": 1.927736345590839e-06,
      "loss": 0.4195,
      "step": 850
    },
    {
      "epoch": 3.16,
      "eval_loss": 0.49196699261665344,
      "eval_runtime": 14.0663,
      "eval_samples_per_second": 8.531,
      "eval_steps_per_second": 2.133,
      "step": 850
    },
    {
      "epoch": 3.2,
      "learning_rate": 1.691657708345823e-06,
      "loss": 0.3772,
      "step": 860
    },
    {
      "epoch": 3.23,
      "learning_rate": 1.4701315480056165e-06,
      "loss": 0.4214,
      "step": 870
    },
    {
      "epoch": 3.27,
      "learning_rate": 1.2634001001741375e-06,
      "loss": 0.4077,
      "step": 880
    },
    {
      "epoch": 3.31,
      "learning_rate": 1.0716894226543954e-06,
      "loss": 0.373,
      "step": 890
    },
    {
      "epoch": 3.35,
      "learning_rate": 8.952091482575825e-07,
      "loss": 0.338,
      "step": 900
    },
    {
      "epoch": 3.35,
      "eval_loss": 0.4922858476638794,
      "eval_runtime": 14.0689,
      "eval_samples_per_second": 8.529,
      "eval_steps_per_second": 2.132,
      "step": 900
    },
    {
      "epoch": 3.38,
      "learning_rate": 7.341522555726971e-07,
      "loss": 0.4324,
      "step": 910
    },
    {
      "epoch": 3.42,
      "learning_rate": 5.886948579472779e-07,
      "loss": 0.4012,
      "step": 920
    },
    {
      "epoch": 3.46,
      "learning_rate": 4.589960109100444e-07,
      "loss": 0.4066,
      "step": 930
    },
    {
      "epoch": 3.49,
      "learning_rate": 3.45197538246011e-07,
      "loss": 0.3547,
      "step": 940
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.474238769142645e-07,
      "loss": 0.3943,
      "step": 950
    },
    {
      "epoch": 3.53,
      "eval_loss": 0.4926001727581024,
      "eval_runtime": 14.07,
      "eval_samples_per_second": 8.529,
      "eval_steps_per_second": 2.132,
      "step": 950
    },
    {
      "epoch": 3.57,
      "learning_rate": 1.657819409779726e-07,
      "loss": 0.4137,
      "step": 960
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.0036100469542786e-07,
      "loss": 0.4237,
      "step": 970
    },
    {
      "epoch": 3.64,
      "learning_rate": 5.1232604899952296e-08,
      "loss": 0.3914,
      "step": 980
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.8450462775428946e-08,
      "loss": 0.4274,
      "step": 990
    },
    {
      "epoch": 3.72,
      "learning_rate": 2.0504251129649378e-09,
      "loss": 0.3953,
      "step": 1000
    },
    {
      "epoch": 3.72,
      "eval_loss": 0.4926711618900299,
      "eval_runtime": 14.0668,
      "eval_samples_per_second": 8.531,
      "eval_steps_per_second": 2.133,
      "step": 1000
    },
    {
      "epoch": 3.72,
      "step": 1000,
      "total_flos": 1.0108845062671565e+17,
      "train_loss": 0.509855135679245,
      "train_runtime": 2165.2959,
      "train_samples_per_second": 1.847,
      "train_steps_per_second": 0.462
    }
  ],
  "logging_steps": 10,
  "max_steps": 1000,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 1.0108845062671565e+17,
  "trial_name": null,
  "trial_params": null
}