{
  "best_metric": 0.031337134540081024,
  "best_model_checkpoint": "saves/psy-course/MentaLLaMA-chat-7B/train/fold8/checkpoint-1250",
  "epoch": 4.996682778883518,
  "eval_steps": 50,
  "global_step": 3295,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015164439389631315,
      "grad_norm": 1.9727195501327515,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 1.6899,
      "step": 10
    },
    {
      "epoch": 0.03032887877926263,
      "grad_norm": 2.0087709426879883,
      "learning_rate": 6.060606060606061e-06,
      "loss": 1.7303,
      "step": 20
    },
    {
      "epoch": 0.045493318168893945,
      "grad_norm": 2.2647852897644043,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.5576,
      "step": 30
    },
    {
      "epoch": 0.06065775755852526,
      "grad_norm": 2.9553816318511963,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 1.3227,
      "step": 40
    },
    {
      "epoch": 0.07582219694815658,
      "grad_norm": 0.9407235980033875,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.8816,
      "step": 50
    },
    {
      "epoch": 0.07582219694815658,
      "eval_loss": 0.6520895957946777,
      "eval_runtime": 201.0285,
      "eval_samples_per_second": 5.835,
      "eval_steps_per_second": 5.835,
      "step": 50
    },
    {
      "epoch": 0.09098663633778789,
      "grad_norm": 0.8535279035568237,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.6344,
      "step": 60
    },
    {
      "epoch": 0.1061510757274192,
      "grad_norm": 0.9502021670341492,
      "learning_rate": 2.1212121212121215e-05,
      "loss": 0.4792,
      "step": 70
    },
    {
      "epoch": 0.12131551511705052,
      "grad_norm": 0.742535412311554,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 0.356,
      "step": 80
    },
    {
      "epoch": 0.13647995450668182,
      "grad_norm": 0.6811979413032532,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.1739,
      "step": 90
    },
    {
      "epoch": 0.15164439389631315,
      "grad_norm": 0.5330702662467957,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 0.1876,
      "step": 100
    },
    {
      "epoch": 0.15164439389631315,
      "eval_loss": 0.12008175253868103,
      "eval_runtime": 200.0525,
      "eval_samples_per_second": 5.863,
      "eval_steps_per_second": 5.863,
      "step": 100
    },
    {
      "epoch": 0.16680883328594445,
      "grad_norm": 0.5429323315620422,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.1265,
      "step": 110
    },
    {
      "epoch": 0.18197327267557578,
      "grad_norm": 0.7030391097068787,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.1058,
      "step": 120
    },
    {
      "epoch": 0.19713771206520708,
      "grad_norm": 0.6562190055847168,
      "learning_rate": 3.939393939393939e-05,
      "loss": 0.0875,
      "step": 130
    },
    {
      "epoch": 0.2123021514548384,
      "grad_norm": 0.4261679947376251,
      "learning_rate": 4.242424242424243e-05,
      "loss": 0.0752,
      "step": 140
    },
    {
      "epoch": 0.2274665908444697,
      "grad_norm": 0.49861419200897217,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.0847,
      "step": 150
    },
    {
      "epoch": 0.2274665908444697,
      "eval_loss": 0.0709112212061882,
      "eval_runtime": 199.9166,
      "eval_samples_per_second": 5.867,
      "eval_steps_per_second": 5.867,
      "step": 150
    },
    {
      "epoch": 0.24263103023410104,
      "grad_norm": 0.6327816247940063,
      "learning_rate": 4.848484848484849e-05,
      "loss": 0.0758,
      "step": 160
    },
    {
      "epoch": 0.25779546962373234,
      "grad_norm": 0.739878237247467,
      "learning_rate": 5.151515151515152e-05,
      "loss": 0.07,
      "step": 170
    },
    {
      "epoch": 0.27295990901336364,
      "grad_norm": 0.5013818144798279,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.0762,
      "step": 180
    },
    {
      "epoch": 0.288124348402995,
      "grad_norm": 0.5184940099716187,
      "learning_rate": 5.757575757575758e-05,
      "loss": 0.0737,
      "step": 190
    },
    {
      "epoch": 0.3032887877926263,
      "grad_norm": 0.5929743051528931,
      "learning_rate": 6.060606060606061e-05,
      "loss": 0.0547,
      "step": 200
    },
    {
      "epoch": 0.3032887877926263,
      "eval_loss": 0.059455741196870804,
      "eval_runtime": 200.0736,
      "eval_samples_per_second": 5.863,
      "eval_steps_per_second": 5.863,
      "step": 200
    },
    {
      "epoch": 0.3184532271822576,
      "grad_norm": 0.4454347789287567,
      "learning_rate": 6.363636363636364e-05,
      "loss": 0.0504,
      "step": 210
    },
    {
      "epoch": 0.3336176665718889,
      "grad_norm": 0.5627999901771545,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0545,
      "step": 220
    },
    {
      "epoch": 0.34878210596152026,
      "grad_norm": 0.4958933889865875,
      "learning_rate": 6.96969696969697e-05,
      "loss": 0.0623,
      "step": 230
    },
    {
      "epoch": 0.36394654535115156,
      "grad_norm": 0.3319215178489685,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.0462,
      "step": 240
    },
    {
      "epoch": 0.37911098474078286,
      "grad_norm": 0.5538325309753418,
      "learning_rate": 7.575757575757576e-05,
      "loss": 0.05,
      "step": 250
    },
    {
      "epoch": 0.37911098474078286,
      "eval_loss": 0.050985436886548996,
      "eval_runtime": 199.8857,
      "eval_samples_per_second": 5.868,
      "eval_steps_per_second": 5.868,
      "step": 250
    },
    {
      "epoch": 0.39427542413041416,
      "grad_norm": 0.5419332385063171,
      "learning_rate": 7.878787878787879e-05,
      "loss": 0.0441,
      "step": 260
    },
    {
      "epoch": 0.40943986352004547,
      "grad_norm": 0.2975507378578186,
      "learning_rate": 8.181818181818183e-05,
      "loss": 0.0763,
      "step": 270
    },
    {
      "epoch": 0.4246043029096768,
      "grad_norm": 0.42280369997024536,
      "learning_rate": 8.484848484848486e-05,
      "loss": 0.0493,
      "step": 280
    },
    {
      "epoch": 0.4397687422993081,
      "grad_norm": 0.49099108576774597,
      "learning_rate": 8.787878787878789e-05,
      "loss": 0.0421,
      "step": 290
    },
    {
      "epoch": 0.4549331816889394,
      "grad_norm": 0.8173151612281799,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.0566,
      "step": 300
    },
    {
      "epoch": 0.4549331816889394,
      "eval_loss": 0.04939843714237213,
      "eval_runtime": 199.6146,
      "eval_samples_per_second": 5.876,
      "eval_steps_per_second": 5.876,
      "step": 300
    },
    {
      "epoch": 0.4700976210785707,
      "grad_norm": 0.45777782797813416,
      "learning_rate": 9.393939393939395e-05,
      "loss": 0.0454,
      "step": 310
    },
    {
      "epoch": 0.4852620604682021,
      "grad_norm": 0.37667033076286316,
      "learning_rate": 9.696969696969698e-05,
      "loss": 0.0476,
      "step": 320
    },
    {
      "epoch": 0.5004264998578334,
      "grad_norm": 0.5223247408866882,
      "learning_rate": 0.0001,
      "loss": 0.0506,
      "step": 330
    },
    {
      "epoch": 0.5155909392474647,
      "grad_norm": 0.45672836899757385,
      "learning_rate": 9.999719336268101e-05,
      "loss": 0.0458,
      "step": 340
    },
    {
      "epoch": 0.530755378637096,
      "grad_norm": 0.3252620995044708,
      "learning_rate": 9.998877376581251e-05,
      "loss": 0.057,
      "step": 350
    },
    {
      "epoch": 0.530755378637096,
      "eval_loss": 0.04613954573869705,
      "eval_runtime": 199.5605,
      "eval_samples_per_second": 5.878,
      "eval_steps_per_second": 5.878,
      "step": 350
    },
    {
      "epoch": 0.5459198180267273,
      "grad_norm": 0.39096570014953613,
      "learning_rate": 9.997474215462472e-05,
      "loss": 0.0478,
      "step": 360
    },
    {
      "epoch": 0.5610842574163586,
      "grad_norm": 0.5067103505134583,
      "learning_rate": 9.995510010438337e-05,
      "loss": 0.0344,
      "step": 370
    },
    {
      "epoch": 0.57624869680599,
      "grad_norm": 0.429315984249115,
      "learning_rate": 9.992984982021295e-05,
      "loss": 0.0543,
      "step": 380
    },
    {
      "epoch": 0.5914131361956213,
      "grad_norm": 0.41895803809165955,
      "learning_rate": 9.9898994136849e-05,
      "loss": 0.0408,
      "step": 390
    },
    {
      "epoch": 0.6065775755852526,
      "grad_norm": 0.3504761755466461,
      "learning_rate": 9.986253651832005e-05,
      "loss": 0.0323,
      "step": 400
    },
    {
      "epoch": 0.6065775755852526,
      "eval_loss": 0.04223111271858215,
      "eval_runtime": 199.1425,
      "eval_samples_per_second": 5.89,
      "eval_steps_per_second": 5.89,
      "step": 400
    },
    {
      "epoch": 0.6217420149748839,
      "grad_norm": 0.22722405195236206,
      "learning_rate": 9.982048105755859e-05,
      "loss": 0.0389,
      "step": 410
    },
    {
      "epoch": 0.6369064543645152,
      "grad_norm": 0.37375208735466003,
      "learning_rate": 9.977283247594166e-05,
      "loss": 0.0476,
      "step": 420
    },
    {
      "epoch": 0.6520708937541465,
      "grad_norm": 0.3475249111652374,
      "learning_rate": 9.971959612276076e-05,
      "loss": 0.045,
      "step": 430
    },
    {
      "epoch": 0.6672353331437778,
      "grad_norm": 0.25098511576652527,
      "learning_rate": 9.966077797462129e-05,
      "loss": 0.0382,
      "step": 440
    },
    {
      "epoch": 0.6823997725334091,
      "grad_norm": 0.19550517201423645,
      "learning_rate": 9.959638463477165e-05,
      "loss": 0.0331,
      "step": 450
    },
    {
      "epoch": 0.6823997725334091,
      "eval_loss": 0.03925846517086029,
      "eval_runtime": 198.7747,
      "eval_samples_per_second": 5.901,
      "eval_steps_per_second": 5.901,
      "step": 450
    },
    {
      "epoch": 0.6975642119230405,
      "grad_norm": 0.5500554442405701,
      "learning_rate": 9.952642333236186e-05,
      "loss": 0.0415,
      "step": 460
    },
    {
      "epoch": 0.7127286513126718,
      "grad_norm": 0.47503772377967834,
      "learning_rate": 9.945090192163202e-05,
      "loss": 0.0453,
      "step": 470
    },
    {
      "epoch": 0.7278930907023031,
      "grad_norm": 0.2633504569530487,
      "learning_rate": 9.936982888103051e-05,
      "loss": 0.034,
      "step": 480
    },
    {
      "epoch": 0.7430575300919344,
      "grad_norm": 0.41773566603660583,
      "learning_rate": 9.928321331226219e-05,
      "loss": 0.0417,
      "step": 490
    },
    {
      "epoch": 0.7582219694815657,
      "grad_norm": 0.16469401121139526,
      "learning_rate": 9.919106493926655e-05,
      "loss": 0.0339,
      "step": 500
    },
    {
      "epoch": 0.7582219694815657,
      "eval_loss": 0.04052235931158066,
      "eval_runtime": 198.7973,
      "eval_samples_per_second": 5.9,
      "eval_steps_per_second": 5.9,
      "step": 500
    },
    {
      "epoch": 0.773386408871197,
      "grad_norm": 0.21951507031917572,
      "learning_rate": 9.909339410712612e-05,
      "loss": 0.0414,
      "step": 510
    },
    {
      "epoch": 0.7885508482608283,
      "grad_norm": 0.44605669379234314,
      "learning_rate": 9.8990211780905e-05,
      "loss": 0.0364,
      "step": 520
    },
    {
      "epoch": 0.8037152876504596,
      "grad_norm": 0.3531959652900696,
      "learning_rate": 9.888152954441785e-05,
      "loss": 0.0384,
      "step": 530
    },
    {
      "epoch": 0.8188797270400909,
      "grad_norm": 0.24108946323394775,
      "learning_rate": 9.876735959892953e-05,
      "loss": 0.0354,
      "step": 540
    },
    {
      "epoch": 0.8340441664297223,
      "grad_norm": 0.22522172331809998,
      "learning_rate": 9.864771476178522e-05,
      "loss": 0.0432,
      "step": 550
    },
    {
      "epoch": 0.8340441664297223,
      "eval_loss": 0.03828958794474602,
      "eval_runtime": 198.8582,
      "eval_samples_per_second": 5.899,
      "eval_steps_per_second": 5.899,
      "step": 550
    },
    {
      "epoch": 0.8492086058193536,
      "grad_norm": 0.3373767137527466,
      "learning_rate": 9.852260846497153e-05,
      "loss": 0.0466,
      "step": 560
    },
    {
      "epoch": 0.864373045208985,
      "grad_norm": 0.32619407773017883,
      "learning_rate": 9.839205475360851e-05,
      "loss": 0.0448,
      "step": 570
    },
    {
      "epoch": 0.8795374845986162,
      "grad_norm": 0.3622454106807709,
      "learning_rate": 9.825606828437291e-05,
      "loss": 0.0482,
      "step": 580
    },
    {
      "epoch": 0.8947019239882475,
      "grad_norm": 0.27383357286453247,
      "learning_rate": 9.811466432385267e-05,
      "loss": 0.0341,
      "step": 590
    },
    {
      "epoch": 0.9098663633778789,
      "grad_norm": 0.31864476203918457,
      "learning_rate": 9.796785874683314e-05,
      "loss": 0.0332,
      "step": 600
    },
    {
      "epoch": 0.9098663633778789,
      "eval_loss": 0.03612194210290909,
      "eval_runtime": 198.8666,
      "eval_samples_per_second": 5.898,
      "eval_steps_per_second": 5.898,
      "step": 600
    },
    {
      "epoch": 0.9250308027675102,
      "grad_norm": 0.24628444015979767,
      "learning_rate": 9.781566803451475e-05,
      "loss": 0.0443,
      "step": 610
    },
    {
      "epoch": 0.9401952421571415,
      "grad_norm": 0.20924372971057892,
      "learning_rate": 9.765810927266281e-05,
      "loss": 0.0334,
      "step": 620
    },
    {
      "epoch": 0.9553596815467729,
      "grad_norm": 0.24674858152866364,
      "learning_rate": 9.749520014968934e-05,
      "loss": 0.0341,
      "step": 630
    },
    {
      "epoch": 0.9705241209364042,
      "grad_norm": 0.3071349859237671,
      "learning_rate": 9.732695895466735e-05,
      "loss": 0.0478,
      "step": 640
    },
    {
      "epoch": 0.9856885603260355,
      "grad_norm": 0.2225157916545868,
      "learning_rate": 9.715340457527746e-05,
      "loss": 0.0458,
      "step": 650
    },
    {
      "epoch": 0.9856885603260355,
      "eval_loss": 0.03808917477726936,
      "eval_runtime": 199.4015,
      "eval_samples_per_second": 5.883,
      "eval_steps_per_second": 5.883,
      "step": 650
    },
    {
      "epoch": 1.0008529997156668,
      "grad_norm": 0.17803935706615448,
      "learning_rate": 9.697455649568761e-05,
      "loss": 0.0345,
      "step": 660
    },
    {
      "epoch": 1.016017439105298,
      "grad_norm": 0.37008166313171387,
      "learning_rate": 9.679043479436556e-05,
      "loss": 0.0253,
      "step": 670
    },
    {
      "epoch": 1.0311818784949294,
      "grad_norm": 0.2343502789735794,
      "learning_rate": 9.660106014182489e-05,
      "loss": 0.0242,
      "step": 680
    },
    {
      "epoch": 1.0463463178845607,
      "grad_norm": 0.2453029602766037,
      "learning_rate": 9.640645379830424e-05,
      "loss": 0.0437,
      "step": 690
    },
    {
      "epoch": 1.061510757274192,
      "grad_norm": 0.265876829624176,
      "learning_rate": 9.620663761138067e-05,
      "loss": 0.0281,
      "step": 700
    },
    {
      "epoch": 1.061510757274192,
      "eval_loss": 0.03682916238903999,
      "eval_runtime": 200.4699,
      "eval_samples_per_second": 5.851,
      "eval_steps_per_second": 5.851,
      "step": 700
    },
    {
      "epoch": 1.0766751966638233,
      "grad_norm": 0.10551781207323074,
      "learning_rate": 9.600163401351688e-05,
      "loss": 0.0201,
      "step": 710
    },
    {
      "epoch": 1.0918396360534546,
      "grad_norm": 0.2587718665599823,
      "learning_rate": 9.579146601954276e-05,
      "loss": 0.04,
      "step": 720
    },
    {
      "epoch": 1.1070040754430859,
      "grad_norm": 0.1365862488746643,
      "learning_rate": 9.557615722407177e-05,
      "loss": 0.0334,
      "step": 730
    },
    {
      "epoch": 1.1221685148327172,
      "grad_norm": 0.08451937884092331,
      "learning_rate": 9.535573179885191e-05,
      "loss": 0.0254,
      "step": 740
    },
    {
      "epoch": 1.1373329542223485,
      "grad_norm": 0.12551841139793396,
      "learning_rate": 9.513021449005214e-05,
      "loss": 0.0222,
      "step": 750
    },
    {
      "epoch": 1.1373329542223485,
      "eval_loss": 0.03816114366054535,
      "eval_runtime": 201.2543,
      "eval_samples_per_second": 5.828,
      "eval_steps_per_second": 5.828,
      "step": 750
    },
    {
      "epoch": 1.15249739361198,
      "grad_norm": 0.11542536318302155,
      "learning_rate": 9.489963061548428e-05,
      "loss": 0.0355,
      "step": 760
    },
    {
      "epoch": 1.1676618330016113,
      "grad_norm": 0.09316863119602203,
      "learning_rate": 9.466400606176062e-05,
      "loss": 0.0178,
      "step": 770
    },
    {
      "epoch": 1.1828262723912426,
      "grad_norm": 0.17316240072250366,
      "learning_rate": 9.442336728138779e-05,
      "loss": 0.0219,
      "step": 780
    },
    {
      "epoch": 1.197990711780874,
      "grad_norm": 0.11998700350522995,
      "learning_rate": 9.417774128979706e-05,
      "loss": 0.0329,
      "step": 790
    },
    {
      "epoch": 1.2131551511705052,
      "grad_norm": 0.17929060757160187,
      "learning_rate": 9.39271556623114e-05,
      "loss": 0.0272,
      "step": 800
    },
    {
      "epoch": 1.2131551511705052,
      "eval_loss": 0.03464984521269798,
      "eval_runtime": 202.272,
      "eval_samples_per_second": 5.799,
      "eval_steps_per_second": 5.799,
      "step": 800
    },
    {
      "epoch": 1.2283195905601365,
      "grad_norm": 0.2804087698459625,
      "learning_rate": 9.367163853104975e-05,
      "loss": 0.032,
      "step": 810
    },
    {
      "epoch": 1.2434840299497678,
      "grad_norm": 0.14590689539909363,
      "learning_rate": 9.341121858176876e-05,
      "loss": 0.0353,
      "step": 820
    },
    {
      "epoch": 1.2586484693393991,
      "grad_norm": 0.18488483130931854,
      "learning_rate": 9.314592505064239e-05,
      "loss": 0.0286,
      "step": 830
    },
    {
      "epoch": 1.2738129087290304,
      "grad_norm": 0.1490219384431839,
      "learning_rate": 9.28757877209796e-05,
      "loss": 0.0212,
      "step": 840
    },
    {
      "epoch": 1.2889773481186617,
      "grad_norm": 0.1876908391714096,
      "learning_rate": 9.260083691988084e-05,
      "loss": 0.0303,
      "step": 850
    },
    {
      "epoch": 1.2889773481186617,
      "eval_loss": 0.03515174984931946,
      "eval_runtime": 203.3766,
      "eval_samples_per_second": 5.768,
      "eval_steps_per_second": 5.768,
      "step": 850
    },
    {
      "epoch": 1.304141787508293,
      "grad_norm": 0.251598596572876,
      "learning_rate": 9.232110351483327e-05,
      "loss": 0.0323,
      "step": 860
    },
    {
      "epoch": 1.3193062268979243,
      "grad_norm": 0.3371416926383972,
      "learning_rate": 9.203661891024547e-05,
      "loss": 0.0321,
      "step": 870
    },
    {
      "epoch": 1.3344706662875556,
      "grad_norm": 0.21001723408699036,
      "learning_rate": 9.174741504392173e-05,
      "loss": 0.0237,
      "step": 880
    },
    {
      "epoch": 1.349635105677187,
      "grad_norm": 0.30405065417289734,
      "learning_rate": 9.145352438347662e-05,
      "loss": 0.0319,
      "step": 890
    },
    {
      "epoch": 1.3647995450668184,
      "grad_norm": 0.3439757823944092,
      "learning_rate": 9.115497992268995e-05,
      "loss": 0.0318,
      "step": 900
    },
    {
      "epoch": 1.3647995450668184,
      "eval_loss": 0.035844411700963974,
      "eval_runtime": 203.8426,
      "eval_samples_per_second": 5.754,
      "eval_steps_per_second": 5.754,
      "step": 900
    },
    {
      "epoch": 1.3799639844564497,
      "grad_norm": 0.05537993460893631,
      "learning_rate": 9.085181517780273e-05,
      "loss": 0.0316,
      "step": 910
    },
    {
      "epoch": 1.395128423846081,
      "grad_norm": 0.16362939774990082,
      "learning_rate": 9.054406418375443e-05,
      "loss": 0.0318,
      "step": 920
    },
    {
      "epoch": 1.4102928632357123,
      "grad_norm": 0.18182925879955292,
      "learning_rate": 9.023176149036203e-05,
      "loss": 0.0266,
      "step": 930
    },
    {
      "epoch": 1.4254573026253436,
      "grad_norm": 0.20367887616157532,
      "learning_rate": 8.991494215844132e-05,
      "loss": 0.0358,
      "step": 940
    },
    {
      "epoch": 1.440621742014975,
      "grad_norm": 0.2871052920818329,
      "learning_rate": 8.959364175587069e-05,
      "loss": 0.0233,
      "step": 950
    },
    {
      "epoch": 1.440621742014975,
      "eval_loss": 0.035272736102342606,
      "eval_runtime": 204.45,
      "eval_samples_per_second": 5.737,
      "eval_steps_per_second": 5.737,
      "step": 950
    },
    {
      "epoch": 1.4557861814046062,
      "grad_norm": 0.24572664499282837,
      "learning_rate": 8.926789635359817e-05,
      "loss": 0.0361,
      "step": 960
    },
    {
      "epoch": 1.4709506207942376,
      "grad_norm": 0.08848091959953308,
      "learning_rate": 8.893774252159187e-05,
      "loss": 0.025,
      "step": 970
    },
    {
      "epoch": 1.4861150601838689,
      "grad_norm": 0.18202854692935944,
      "learning_rate": 8.860321732473439e-05,
      "loss": 0.0328,
      "step": 980
    },
    {
      "epoch": 1.5012794995735002,
      "grad_norm": 0.41244789958000183,
      "learning_rate": 8.826435831866184e-05,
      "loss": 0.0338,
      "step": 990
    },
    {
      "epoch": 1.5164439389631315,
      "grad_norm": 0.08242332190275192,
      "learning_rate": 8.79212035455475e-05,
      "loss": 0.0263,
      "step": 1000
    },
    {
      "epoch": 1.5164439389631315,
      "eval_loss": 0.03487205505371094,
      "eval_runtime": 205.0947,
      "eval_samples_per_second": 5.719,
      "eval_steps_per_second": 5.719,
      "step": 1000
    },
    {
      "epoch": 1.5316083783527628,
      "grad_norm": 0.08239326626062393,
      "learning_rate": 8.757379152983103e-05,
      "loss": 0.0245,
      "step": 1010
    },
    {
      "epoch": 1.546772817742394,
      "grad_norm": 0.23835065960884094,
      "learning_rate": 8.722216127389363e-05,
      "loss": 0.028,
      "step": 1020
    },
    {
      "epoch": 1.5619372571320254,
      "grad_norm": 0.2806464433670044,
      "learning_rate": 8.686635225367919e-05,
      "loss": 0.0317,
      "step": 1030
    },
    {
      "epoch": 1.5771016965216567,
      "grad_norm": 0.13419637084007263,
      "learning_rate": 8.650640441426274e-05,
      "loss": 0.0338,
      "step": 1040
    },
    {
      "epoch": 1.592266135911288,
      "grad_norm": 0.37877002358436584,
      "learning_rate": 8.614235816536582e-05,
      "loss": 0.0381,
      "step": 1050
    },
    {
      "epoch": 1.592266135911288,
      "eval_loss": 0.035373929888010025,
      "eval_runtime": 205.8756,
      "eval_samples_per_second": 5.698,
      "eval_steps_per_second": 5.698,
      "step": 1050
    },
    {
      "epoch": 1.6074305753009193,
      "grad_norm": 0.17082390189170837,
      "learning_rate": 8.577425437681994e-05,
      "loss": 0.039,
      "step": 1060
    },
    {
      "epoch": 1.6225950146905506,
      "grad_norm": 0.09400735050439835,
      "learning_rate": 8.540213437397833e-05,
      "loss": 0.03,
      "step": 1070
    },
    {
      "epoch": 1.6377594540801819,
      "grad_norm": 0.11701495945453644,
      "learning_rate": 8.502603993307647e-05,
      "loss": 0.027,
      "step": 1080
    },
    {
      "epoch": 1.6529238934698132,
      "grad_norm": 0.16814526915550232,
      "learning_rate": 8.464601327654207e-05,
      "loss": 0.023,
      "step": 1090
    },
    {
      "epoch": 1.6680883328594445,
      "grad_norm": 0.15352267026901245,
      "learning_rate": 8.4262097068255e-05,
      "loss": 0.0267,
      "step": 1100
    },
    {
      "epoch": 1.6680883328594445,
      "eval_loss": 0.0319027379155159,
      "eval_runtime": 206.408,
      "eval_samples_per_second": 5.683,
      "eval_steps_per_second": 5.683,
      "step": 1100
    },
    {
      "epoch": 1.6832527722490758,
      "grad_norm": 0.535393238067627,
      "learning_rate": 8.387433440875758e-05,
      "loss": 0.0296,
      "step": 1110
    },
    {
      "epoch": 1.698417211638707,
      "grad_norm": 0.4503651261329651,
      "learning_rate": 8.348276883041583e-05,
      "loss": 0.0311,
      "step": 1120
    },
    {
      "epoch": 1.7135816510283386,
      "grad_norm": 0.21617506444454193,
      "learning_rate": 8.308744429253238e-05,
      "loss": 0.0314,
      "step": 1130
    },
    {
      "epoch": 1.72874609041797,
      "grad_norm": 0.32634076476097107,
      "learning_rate": 8.268840517641123e-05,
      "loss": 0.0292,
      "step": 1140
    },
    {
      "epoch": 1.7439105298076012,
      "grad_norm": 0.3362925946712494,
      "learning_rate": 8.228569628037543e-05,
      "loss": 0.0335,
      "step": 1150
    },
    {
      "epoch": 1.7439105298076012,
      "eval_loss": 0.032032061368227005,
      "eval_runtime": 206.8911,
      "eval_samples_per_second": 5.67,
      "eval_steps_per_second": 5.67,
      "step": 1150
    },
    {
      "epoch": 1.7590749691972325,
      "grad_norm": 0.2585095465183258,
      "learning_rate": 8.187936281473758e-05,
      "loss": 0.0285,
      "step": 1160
    },
    {
      "epoch": 1.7742394085868638,
      "grad_norm": 0.14993241429328918,
      "learning_rate": 8.146945039672446e-05,
      "loss": 0.024,
      "step": 1170
    },
    {
      "epoch": 1.789403847976495,
      "grad_norm": 0.41795483231544495,
      "learning_rate": 8.105600504535567e-05,
      "loss": 0.0292,
      "step": 1180
    },
    {
      "epoch": 1.8045682873661264,
      "grad_norm": 0.34870293736457825,
      "learning_rate": 8.06390731762773e-05,
      "loss": 0.0379,
      "step": 1190
    },
    {
      "epoch": 1.8197327267557577,
      "grad_norm": 0.3084681034088135,
      "learning_rate": 8.021870159655109e-05,
      "loss": 0.0228,
      "step": 1200
    },
    {
      "epoch": 1.8197327267557577,
      "eval_loss": 0.03198724240064621,
      "eval_runtime": 206.9606,
      "eval_samples_per_second": 5.668,
      "eval_steps_per_second": 5.668,
      "step": 1200
    },
    {
      "epoch": 1.8348971661453892,
      "grad_norm": 0.3046850860118866,
      "learning_rate": 7.979493749939955e-05,
      "loss": 0.0262,
      "step": 1210
    },
    {
      "epoch": 1.8500616055350205,
      "grad_norm": 0.22466641664505005,
      "learning_rate": 7.93678284589079e-05,
      "loss": 0.0348,
      "step": 1220
    },
    {
      "epoch": 1.8652260449246518,
      "grad_norm": 0.160406231880188,
      "learning_rate": 7.893742242468301e-05,
      "loss": 0.0278,
      "step": 1230
    },
    {
      "epoch": 1.8803904843142831,
      "grad_norm": 0.38902711868286133,
      "learning_rate": 7.850376771647038e-05,
      "loss": 0.0361,
      "step": 1240
    },
    {
      "epoch": 1.8955549237039144,
      "grad_norm": 0.2108290046453476,
      "learning_rate": 7.806691301872958e-05,
      "loss": 0.0311,
      "step": 1250
    },
    {
      "epoch": 1.8955549237039144,
      "eval_loss": 0.031337134540081024,
      "eval_runtime": 207.0505,
      "eval_samples_per_second": 5.665,
      "eval_steps_per_second": 5.665,
      "step": 1250
    },
    {
      "epoch": 1.9107193630935457,
      "grad_norm": 0.07268174737691879,
      "learning_rate": 7.762690737516846e-05,
      "loss": 0.0306,
      "step": 1260
    },
    {
      "epoch": 1.925883802483177,
      "grad_norm": 0.17554602026939392,
      "learning_rate": 7.718380018323742e-05,
      "loss": 0.0233,
      "step": 1270
    },
    {
      "epoch": 1.9410482418728083,
      "grad_norm": 0.18121547996997833,
      "learning_rate": 7.673764118858371e-05,
      "loss": 0.0242,
      "step": 1280
    },
    {
      "epoch": 1.9562126812624396,
      "grad_norm": 0.1335168182849884,
      "learning_rate": 7.628848047946675e-05,
      "loss": 0.0332,
      "step": 1290
    },
    {
      "epoch": 1.971377120652071,
      "grad_norm": 0.18943274021148682,
      "learning_rate": 7.583636848113483e-05,
      "loss": 0.0228,
      "step": 1300
    },
    {
      "epoch": 1.971377120652071,
      "eval_loss": 0.03141847252845764,
      "eval_runtime": 207.6226,
      "eval_samples_per_second": 5.65,
      "eval_steps_per_second": 5.65,
      "step": 1300
    },
    {
      "epoch": 1.9865415600417022,
      "grad_norm": 0.28002068400382996,
      "learning_rate": 7.538135595016423e-05,
      "loss": 0.0309,
      "step": 1310
    },
    {
      "epoch": 2.0017059994313335,
      "grad_norm": 0.16608303785324097,
      "learning_rate": 7.492349396876096e-05,
      "loss": 0.0279,
      "step": 1320
    },
    {
      "epoch": 2.016870438820965,
      "grad_norm": 0.12324703484773636,
      "learning_rate": 7.4462833939026e-05,
      "loss": 0.0216,
      "step": 1330
    },
    {
      "epoch": 2.032034878210596,
      "grad_norm": 0.1612786501646042,
      "learning_rate": 7.399942757718455e-05,
      "loss": 0.0179,
      "step": 1340
    },
    {
      "epoch": 2.0471993176002274,
      "grad_norm": 0.2244199812412262,
      "learning_rate": 7.35333269077802e-05,
      "loss": 0.0157,
      "step": 1350
    },
    {
      "epoch": 2.0471993176002274,
      "eval_loss": 0.03206400200724602,
      "eval_runtime": 208.1756,
      "eval_samples_per_second": 5.635,
      "eval_steps_per_second": 5.635,
      "step": 1350
    },
    {
      "epoch": 2.0623637569898587,
      "grad_norm": 0.3085191547870636,
      "learning_rate": 7.306458425783426e-05,
      "loss": 0.0242,
      "step": 1360
    },
    {
      "epoch": 2.07752819637949,
      "grad_norm": 0.18783684074878693,
      "learning_rate": 7.25932522509713e-05,
      "loss": 0.0169,
      "step": 1370
    },
    {
      "epoch": 2.0926926357691213,
      "grad_norm": 0.2335079312324524,
      "learning_rate": 7.211938380151133e-05,
      "loss": 0.0215,
      "step": 1380
    },
    {
      "epoch": 2.1078570751587526,
      "grad_norm": 0.2962063252925873,
      "learning_rate": 7.164303210852934e-05,
      "loss": 0.0241,
      "step": 1390
    },
    {
      "epoch": 2.123021514548384,
      "grad_norm": 0.15745271742343903,
      "learning_rate": 7.116425064988286e-05,
      "loss": 0.0129,
      "step": 1400
    },
    {
      "epoch": 2.123021514548384,
      "eval_loss": 0.03161005303263664,
      "eval_runtime": 208.076,
      "eval_samples_per_second": 5.637,
      "eval_steps_per_second": 5.637,
      "step": 1400
    },
    {
      "epoch": 2.1381859539380152,
      "grad_norm": 0.34667152166366577,
      "learning_rate": 7.068309317620827e-05,
      "loss": 0.0204,
      "step": 1410
    },
    {
      "epoch": 2.1533503933276466,
      "grad_norm": 0.19276046752929688,
      "learning_rate": 7.019961370488645e-05,
      "loss": 0.0167,
      "step": 1420
    },
    {
      "epoch": 2.168514832717278,
      "grad_norm": 0.23463784158229828,
      "learning_rate": 6.971386651397849e-05,
      "loss": 0.0165,
      "step": 1430
    },
    {
      "epoch": 2.183679272106909,
      "grad_norm": 0.2765744924545288,
      "learning_rate": 6.922590613613211e-05,
      "loss": 0.0148,
      "step": 1440
    },
    {
      "epoch": 2.1988437114965405,
      "grad_norm": 0.15293200314044952,
      "learning_rate": 6.873578735245961e-05,
      "loss": 0.0208,
      "step": 1450
    },
    {
      "epoch": 2.1988437114965405,
      "eval_loss": 0.03302780166268349,
      "eval_runtime": 208.992,
      "eval_samples_per_second": 5.613,
      "eval_steps_per_second": 5.613,
      "step": 1450
    },
    {
      "epoch": 2.2140081508861718,
      "grad_norm": 0.1213361918926239,
      "learning_rate": 6.824356518638775e-05,
      "loss": 0.0218,
      "step": 1460
    },
    {
      "epoch": 2.229172590275803,
      "grad_norm": 0.20920248329639435,
      "learning_rate": 6.774929489748052e-05,
      "loss": 0.0175,
      "step": 1470
    },
    {
      "epoch": 2.2443370296654344,
      "grad_norm": 0.41807928681373596,
      "learning_rate": 6.725303197523548e-05,
      "loss": 0.0269,
      "step": 1480
    },
    {
      "epoch": 2.259501469055066,
      "grad_norm": 0.10990037024021149,
      "learning_rate": 6.675483213285412e-05,
      "loss": 0.017,
      "step": 1490
    },
    {
      "epoch": 2.274665908444697,
      "grad_norm": 0.2917281687259674,
      "learning_rate": 6.625475130098728e-05,
      "loss": 0.0191,
      "step": 1500
    },
    {
      "epoch": 2.274665908444697,
      "eval_loss": 0.03248446062207222,
      "eval_runtime": 208.5219,
      "eval_samples_per_second": 5.625,
      "eval_steps_per_second": 5.625,
      "step": 1500
    },
    {
      "epoch": 2.2898303478343287,
      "grad_norm": 0.18032291531562805,
      "learning_rate": 6.575284562145593e-05,
      "loss": 0.0174,
      "step": 1510
    },
    {
      "epoch": 2.30499478722396,
      "grad_norm": 0.39623764157295227,
      "learning_rate": 6.524917144094851e-05,
      "loss": 0.0242,
      "step": 1520
    },
    {
      "epoch": 2.3201592266135913,
      "grad_norm": 0.04941236600279808,
      "learning_rate": 6.474378530469509e-05,
      "loss": 0.0237,
      "step": 1530
    },
    {
      "epoch": 2.3353236660032226,
      "grad_norm": 0.08923232555389404,
      "learning_rate": 6.42367439501193e-05,
      "loss": 0.0148,
      "step": 1540
    },
    {
      "epoch": 2.350488105392854,
      "grad_norm": 0.25008293986320496,
      "learning_rate": 6.372810430046862e-05,
      "loss": 0.0189,
      "step": 1550
    },
    {
      "epoch": 2.350488105392854,
      "eval_loss": 0.03253987804055214,
      "eval_runtime": 208.5816,
      "eval_samples_per_second": 5.624,
      "eval_steps_per_second": 5.624,
      "step": 1550
    },
    {
      "epoch": 2.365652544782485,
      "grad_norm": 0.17195628583431244,
      "learning_rate": 6.321792345842402e-05,
      "loss": 0.0219,
      "step": 1560
    },
    {
      "epoch": 2.3808169841721165,
      "grad_norm": 0.14535215497016907,
      "learning_rate": 6.270625869968906e-05,
      "loss": 0.0294,
      "step": 1570
    },
    {
      "epoch": 2.395981423561748,
      "grad_norm": 0.19247370958328247,
      "learning_rate": 6.219316746656007e-05,
      "loss": 0.0216,
      "step": 1580
    },
    {
      "epoch": 2.411145862951379,
      "grad_norm": 0.1686420887708664,
      "learning_rate": 6.167870736147713e-05,
      "loss": 0.0217,
      "step": 1590
    },
    {
      "epoch": 2.4263103023410104,
      "grad_norm": 0.20523522794246674,
      "learning_rate": 6.116293614055744e-05,
      "loss": 0.0161,
      "step": 1600
    },
    {
      "epoch": 2.4263103023410104,
      "eval_loss": 0.03252248466014862,
      "eval_runtime": 208.7585,
      "eval_samples_per_second": 5.619,
      "eval_steps_per_second": 5.619,
      "step": 1600
    },
    {
      "epoch": 2.4414747417306417,
      "grad_norm": 0.15517178177833557,
      "learning_rate": 6.06459117071113e-05,
      "loss": 0.0163,
      "step": 1610
    },
    {
      "epoch": 2.456639181120273,
      "grad_norm": 0.30436283349990845,
      "learning_rate": 6.012769210514146e-05,
      "loss": 0.0174,
      "step": 1620
    },
    {
      "epoch": 2.4718036205099043,
      "grad_norm": 0.08914119005203247,
      "learning_rate": 5.9608335512826915e-05,
      "loss": 0.0178,
      "step": 1630
    },
    {
      "epoch": 2.4869680598995356,
      "grad_norm": 0.08494864404201508,
      "learning_rate": 5.908790023599144e-05,
      "loss": 0.0117,
      "step": 1640
    },
    {
      "epoch": 2.502132499289167,
      "grad_norm": 0.23792272806167603,
      "learning_rate": 5.856644470155781e-05,
      "loss": 0.0164,
      "step": 1650
    },
    {
      "epoch": 2.502132499289167,
      "eval_loss": 0.03559296950697899,
      "eval_runtime": 209.1857,
      "eval_samples_per_second": 5.607,
      "eval_steps_per_second": 5.607,
      "step": 1650
    },
    {
      "epoch": 2.5172969386787982,
      "grad_norm": 0.15145932137966156,
      "learning_rate": 5.8044027450988546e-05,
      "loss": 0.021,
      "step": 1660
    },
    {
      "epoch": 2.5324613780684295,
      "grad_norm": 0.15041588246822357,
      "learning_rate": 5.752070713371371e-05,
      "loss": 0.0147,
      "step": 1670
    },
    {
      "epoch": 2.547625817458061,
      "grad_norm": 0.23582014441490173,
      "learning_rate": 5.699654250054662e-05,
      "loss": 0.0143,
      "step": 1680
    },
    {
      "epoch": 2.562790256847692,
      "grad_norm": 0.15433664619922638,
      "learning_rate": 5.647159239708809e-05,
      "loss": 0.0153,
      "step": 1690
    },
    {
      "epoch": 2.5779546962373234,
      "grad_norm": 0.14156116545200348,
      "learning_rate": 5.5945915757120146e-05,
      "loss": 0.0144,
      "step": 1700
    },
    {
      "epoch": 2.5779546962373234,
      "eval_loss": 0.0337643064558506,
      "eval_runtime": 209.5605,
      "eval_samples_per_second": 5.597,
      "eval_steps_per_second": 5.597,
      "step": 1700
    },
    {
      "epoch": 2.5931191356269547,
      "grad_norm": 0.3334906995296478,
      "learning_rate": 5.5419571595989825e-05,
      "loss": 0.0176,
      "step": 1710
    },
    {
      "epoch": 2.608283575016586,
      "grad_norm": 0.37115344405174255,
      "learning_rate": 5.4892619003983734e-05,
      "loss": 0.02,
      "step": 1720
    },
    {
      "epoch": 2.6234480144062173,
      "grad_norm": 0.04717765748500824,
      "learning_rate": 5.436511713969428e-05,
      "loss": 0.0134,
      "step": 1730
    },
    {
      "epoch": 2.6386124537958486,
      "grad_norm": 0.0989762544631958,
      "learning_rate": 5.383712522337817e-05,
      "loss": 0.022,
      "step": 1740
    },
    {
      "epoch": 2.65377689318548,
      "grad_norm": 0.19949494302272797,
      "learning_rate": 5.3308702530308076e-05,
      "loss": 0.0211,
      "step": 1750
    },
    {
      "epoch": 2.65377689318548,
      "eval_loss": 0.032832950353622437,
      "eval_runtime": 209.7443,
      "eval_samples_per_second": 5.593,
      "eval_steps_per_second": 5.593,
      "step": 1750
    },
    {
      "epoch": 2.6689413325751112,
      "grad_norm": 0.18161270022392273,
      "learning_rate": 5.2779908384118025e-05,
      "loss": 0.0233,
      "step": 1760
    },
    {
      "epoch": 2.6841057719647425,
      "grad_norm": 0.25410404801368713,
      "learning_rate": 5.22508021501434e-05,
      "loss": 0.0233,
      "step": 1770
    },
    {
      "epoch": 2.699270211354374,
      "grad_norm": 0.14723050594329834,
      "learning_rate": 5.1721443228756284e-05,
      "loss": 0.015,
      "step": 1780
    },
    {
      "epoch": 2.714434650744005,
      "grad_norm": 0.2623627483844757,
      "learning_rate": 5.119189104869683e-05,
      "loss": 0.0175,
      "step": 1790
    },
    {
      "epoch": 2.729599090133637,
      "grad_norm": 0.14635932445526123,
      "learning_rate": 5.066220506040148e-05,
      "loss": 0.0195,
      "step": 1800
    },
    {
      "epoch": 2.729599090133637,
      "eval_loss": 0.03220284357666969,
      "eval_runtime": 205.6598,
      "eval_samples_per_second": 5.704,
      "eval_steps_per_second": 5.704,
      "step": 1800
    },
    {
      "epoch": 2.7447635295232677,
      "grad_norm": 0.12976735830307007,
      "learning_rate": 5.013244472932872e-05,
      "loss": 0.0202,
      "step": 1810
    },
    {
      "epoch": 2.7599279689128995,
      "grad_norm": 0.216698557138443,
      "learning_rate": 4.960266952928316e-05,
      "loss": 0.0179,
      "step": 1820
    },
    {
      "epoch": 2.7750924083025303,
      "grad_norm": 0.26297900080680847,
      "learning_rate": 4.907293893573867e-05,
      "loss": 0.0163,
      "step": 1830
    },
    {
      "epoch": 2.790256847692162,
      "grad_norm": 0.16788074374198914,
      "learning_rate": 4.8543312419161396e-05,
      "loss": 0.0182,
      "step": 1840
    },
    {
      "epoch": 2.805421287081793,
      "grad_norm": 0.22970642149448395,
      "learning_rate": 4.8013849438333165e-05,
      "loss": 0.0148,
      "step": 1850
    },
    {
      "epoch": 2.805421287081793,
      "eval_loss": 0.033808089792728424,
      "eval_runtime": 209.7324,
      "eval_samples_per_second": 5.593,
      "eval_steps_per_second": 5.593,
      "step": 1850
    },
    {
      "epoch": 2.8205857264714247,
      "grad_norm": 0.36859261989593506,
      "learning_rate": 4.748460943367643e-05,
      "loss": 0.0189,
      "step": 1860
    },
    {
      "epoch": 2.8357501658610555,
      "grad_norm": 0.28842833638191223,
      "learning_rate": 4.695565182058113e-05,
      "loss": 0.0256,
      "step": 1870
    },
    {
      "epoch": 2.8509146052506873,
      "grad_norm": 0.2374098002910614,
      "learning_rate": 4.642703598273431e-05,
      "loss": 0.017,
      "step": 1880
    },
    {
      "epoch": 2.8660790446403186,
      "grad_norm": 0.18494807183742523,
      "learning_rate": 4.589882126545352e-05,
      "loss": 0.0189,
      "step": 1890
    },
    {
      "epoch": 2.88124348402995,
      "grad_norm": 0.16361607611179352,
      "learning_rate": 4.537106696902425e-05,
      "loss": 0.0249,
      "step": 1900
    },
    {
      "epoch": 2.88124348402995,
      "eval_loss": 0.03270639479160309,
      "eval_runtime": 210.1362,
      "eval_samples_per_second": 5.582,
      "eval_steps_per_second": 5.582,
      "step": 1900
    },
    {
      "epoch": 2.896407923419581,
      "grad_norm": 0.6155688166618347,
      "learning_rate": 4.484383234204266e-05,
      "loss": 0.022,
      "step": 1910
    },
    {
      "epoch": 2.9115723628092125,
      "grad_norm": 0.1977698802947998,
      "learning_rate": 4.4317176574763935e-05,
      "loss": 0.0184,
      "step": 1920
    },
    {
      "epoch": 2.926736802198844,
      "grad_norm": 0.14884449541568756,
      "learning_rate": 4.3791158792457334e-05,
      "loss": 0.0211,
      "step": 1930
    },
    {
      "epoch": 2.941901241588475,
      "grad_norm": 0.1196577176451683,
      "learning_rate": 4.3265838048768334e-05,
      "loss": 0.0168,
      "step": 1940
    },
    {
      "epoch": 2.9570656809781064,
      "grad_norm": 0.23351362347602844,
      "learning_rate": 4.274127331908915e-05,
      "loss": 0.0152,
      "step": 1950
    },
    {
      "epoch": 2.9570656809781064,
      "eval_loss": 0.03197691962122917,
      "eval_runtime": 210.4483,
      "eval_samples_per_second": 5.574,
      "eval_steps_per_second": 5.574,
      "step": 1950
    },
    {
      "epoch": 2.9722301203677377,
      "grad_norm": 0.04323707893490791,
      "learning_rate": 4.221752349393758e-05,
      "loss": 0.0152,
      "step": 1960
    },
    {
      "epoch": 2.987394559757369,
      "grad_norm": 0.0944715216755867,
      "learning_rate": 4.1694647372345886e-05,
      "loss": 0.0172,
      "step": 1970
    },
    {
      "epoch": 3.0025589991470003,
      "grad_norm": 0.1532895863056183,
      "learning_rate": 4.117270365525946e-05,
      "loss": 0.0181,
      "step": 1980
    },
    {
      "epoch": 3.0177234385366316,
      "grad_norm": 0.09352372586727142,
      "learning_rate": 4.065175093894694e-05,
      "loss": 0.0104,
      "step": 1990
    },
    {
      "epoch": 3.032887877926263,
      "grad_norm": 0.12653310596942902,
      "learning_rate": 4.013184770842167e-05,
      "loss": 0.0136,
      "step": 2000
    },
    {
      "epoch": 3.032887877926263,
      "eval_loss": 0.03291192650794983,
      "eval_runtime": 210.5061,
      "eval_samples_per_second": 5.572,
      "eval_steps_per_second": 5.572,
      "step": 2000
    },
    {
      "epoch": 3.048052317315894,
      "grad_norm": 0.10540532320737839,
      "learning_rate": 3.9613052330876064e-05,
      "loss": 0.0118,
      "step": 2010
    },
    {
      "epoch": 3.0632167567055255,
      "grad_norm": 0.5384581685066223,
      "learning_rate": 3.909542304912881e-05,
      "loss": 0.0137,
      "step": 2020
    },
    {
      "epoch": 3.078381196095157,
      "grad_norm": 0.18060655891895294,
      "learning_rate": 3.857901797508628e-05,
      "loss": 0.0156,
      "step": 2030
    },
    {
      "epoch": 3.093545635484788,
      "grad_norm": 0.10394889861345291,
      "learning_rate": 3.80638950832186e-05,
      "loss": 0.0182,
      "step": 2040
    },
    {
      "epoch": 3.1087100748744194,
      "grad_norm": 0.18449059128761292,
      "learning_rate": 3.7550112204051014e-05,
      "loss": 0.009,
      "step": 2050
    },
    {
      "epoch": 3.1087100748744194,
      "eval_loss": 0.03408639878034592,
      "eval_runtime": 210.2004,
      "eval_samples_per_second": 5.58,
      "eval_steps_per_second": 5.58,
      "step": 2050
    },
    {
      "epoch": 3.1238745142640507,
      "grad_norm": 0.05877893418073654,
      "learning_rate": 3.703772701767167e-05,
      "loss": 0.0073,
      "step": 2060
    },
    {
      "epoch": 3.139038953653682,
      "grad_norm": 0.1925165206193924,
      "learning_rate": 3.652679704725596e-05,
      "loss": 0.007,
      "step": 2070
    },
    {
      "epoch": 3.1542033930433133,
      "grad_norm": 0.18786293268203735,
      "learning_rate": 3.601737965260882e-05,
      "loss": 0.0079,
      "step": 2080
    },
    {
      "epoch": 3.1693678324329446,
      "grad_norm": 0.229954332113266,
      "learning_rate": 3.550953202372503e-05,
      "loss": 0.0061,
      "step": 2090
    },
    {
      "epoch": 3.184532271822576,
      "grad_norm": 0.3331630229949951,
      "learning_rate": 3.500331117436895e-05,
      "loss": 0.0089,
      "step": 2100
    },
    {
      "epoch": 3.184532271822576,
      "eval_loss": 0.03669964522123337,
      "eval_runtime": 210.13,
      "eval_samples_per_second": 5.582,
      "eval_steps_per_second": 5.582,
      "step": 2100
    },
    {
      "epoch": 3.1996967112122072,
      "grad_norm": 0.14467471837997437,
      "learning_rate": 3.4498773935673626e-05,
      "loss": 0.0101,
      "step": 2110
    },
    {
      "epoch": 3.2148611506018385,
      "grad_norm": 0.18001681566238403,
      "learning_rate": 3.399597694976081e-05,
      "loss": 0.0154,
      "step": 2120
    },
    {
      "epoch": 3.23002558999147,
      "grad_norm": 0.14212116599082947,
      "learning_rate": 3.349497666338187e-05,
      "loss": 0.0109,
      "step": 2130
    },
    {
      "epoch": 3.245190029381101,
      "grad_norm": 0.24891246855258942,
      "learning_rate": 3.299582932158085e-05,
      "loss": 0.0107,
      "step": 2140
    },
    {
      "epoch": 3.260354468770733,
      "grad_norm": 0.1589607298374176,
      "learning_rate": 3.2498590961379996e-05,
      "loss": 0.0127,
      "step": 2150
    },
    {
      "epoch": 3.260354468770733,
      "eval_loss": 0.0364246740937233,
      "eval_runtime": 209.961,
      "eval_samples_per_second": 5.587,
      "eval_steps_per_second": 5.587,
      "step": 2150
    },
    {
      "epoch": 3.2755189081603637,
      "grad_norm": 0.15065188705921173,
      "learning_rate": 3.200331740548887e-05,
      "loss": 0.0121,
      "step": 2160
    },
    {
      "epoch": 3.2906833475499955,
      "grad_norm": 0.17789319157600403,
      "learning_rate": 3.1510064256037274e-05,
      "loss": 0.0125,
      "step": 2170
    },
    {
      "epoch": 3.3058477869396263,
      "grad_norm": 0.1671750396490097,
      "learning_rate": 3.1018886888333065e-05,
      "loss": 0.0088,
      "step": 2180
    },
    {
      "epoch": 3.321012226329258,
      "grad_norm": 0.1114286333322525,
      "learning_rate": 3.052984044464548e-05,
      "loss": 0.0109,
      "step": 2190
    },
    {
      "epoch": 3.3361766657188894,
      "grad_norm": 0.1630328744649887,
      "learning_rate": 3.0042979828014496e-05,
      "loss": 0.0119,
      "step": 2200
    },
    {
      "epoch": 3.3361766657188894,
      "eval_loss": 0.03581894189119339,
      "eval_runtime": 210.4663,
      "eval_samples_per_second": 5.573,
      "eval_steps_per_second": 5.573,
      "step": 2200
    },
    {
      "epoch": 3.3513411051085207,
      "grad_norm": 0.1710463911294937,
      "learning_rate": 2.9558359696087152e-05,
      "loss": 0.0134,
      "step": 2210
    },
    {
      "epoch": 3.366505544498152,
      "grad_norm": 0.1446899175643921,
      "learning_rate": 2.9076034454981332e-05,
      "loss": 0.0116,
      "step": 2220
    },
    {
      "epoch": 3.3816699838877833,
      "grad_norm": 0.23707441985607147,
      "learning_rate": 2.8596058253177932e-05,
      "loss": 0.0105,
      "step": 2230
    },
    {
      "epoch": 3.3968344232774146,
      "grad_norm": 0.057392384856939316,
      "learning_rate": 2.811848497544175e-05,
      "loss": 0.0119,
      "step": 2240
    },
    {
      "epoch": 3.411998862667046,
      "grad_norm": 0.16835249960422516,
      "learning_rate": 2.764336823677216e-05,
      "loss": 0.0118,
      "step": 2250
    },
    {
      "epoch": 3.411998862667046,
      "eval_loss": 0.0358290821313858,
      "eval_runtime": 210.8136,
      "eval_samples_per_second": 5.564,
      "eval_steps_per_second": 5.564,
      "step": 2250
    },
    {
      "epoch": 3.427163302056677,
      "grad_norm": 0.10215135663747787,
      "learning_rate": 2.717076137638388e-05,
      "loss": 0.0116,
      "step": 2260
    },
    {
      "epoch": 3.4423277414463085,
      "grad_norm": 0.2105620950460434,
      "learning_rate": 2.6700717451719066e-05,
      "loss": 0.011,
      "step": 2270
    },
    {
      "epoch": 3.45749218083594,
      "grad_norm": 0.31182989478111267,
      "learning_rate": 2.6233289232490447e-05,
      "loss": 0.0121,
      "step": 2280
    },
    {
      "epoch": 3.472656620225571,
      "grad_norm": 0.17491719126701355,
      "learning_rate": 2.5768529194757474e-05,
      "loss": 0.0077,
      "step": 2290
    },
    {
      "epoch": 3.4878210596152024,
      "grad_norm": 0.2520913779735565,
      "learning_rate": 2.5306489515034713e-05,
      "loss": 0.0084,
      "step": 2300
    },
    {
      "epoch": 3.4878210596152024,
      "eval_loss": 0.0370686799287796,
      "eval_runtime": 210.7211,
      "eval_samples_per_second": 5.567,
      "eval_steps_per_second": 5.567,
      "step": 2300
    },
    {
      "epoch": 3.5029854990048337,
      "grad_norm": 0.14362971484661102,
      "learning_rate": 2.484722206443455e-05,
      "loss": 0.016,
      "step": 2310
    },
    {
      "epoch": 3.518149938394465,
      "grad_norm": 0.09810702502727509,
      "learning_rate": 2.4390778402843605e-05,
      "loss": 0.0107,
      "step": 2320
    },
    {
      "epoch": 3.5333143777840963,
      "grad_norm": 0.2903960049152374,
      "learning_rate": 2.3937209773134472e-05,
      "loss": 0.0092,
      "step": 2330
    },
    {
      "epoch": 3.5484788171737276,
      "grad_norm": 0.3453470766544342,
      "learning_rate": 2.3486567095412864e-05,
      "loss": 0.0096,
      "step": 2340
    },
    {
      "epoch": 3.563643256563359,
      "grad_norm": 0.36739131808280945,
      "learning_rate": 2.3038900961301053e-05,
      "loss": 0.0137,
      "step": 2350
    },
    {
      "epoch": 3.563643256563359,
      "eval_loss": 0.03648369759321213,
      "eval_runtime": 210.7175,
      "eval_samples_per_second": 5.567,
      "eval_steps_per_second": 5.567,
      "step": 2350
    },
    {
      "epoch": 3.57880769595299,
      "grad_norm": 0.13596086204051971,
      "learning_rate": 2.259426162825819e-05,
      "loss": 0.0082,
      "step": 2360
    },
    {
      "epoch": 3.5939721353426215,
      "grad_norm": 0.29473790526390076,
      "learning_rate": 2.215269901393805e-05,
      "loss": 0.0099,
      "step": 2370
    },
    {
      "epoch": 3.609136574732253,
      "grad_norm": 0.1077868714928627,
      "learning_rate": 2.1714262690585192e-05,
      "loss": 0.0079,
      "step": 2380
    },
    {
      "epoch": 3.624301014121884,
      "grad_norm": 0.10145460814237595,
      "learning_rate": 2.1279001879469424e-05,
      "loss": 0.0082,
      "step": 2390
    },
    {
      "epoch": 3.6394654535115154,
      "grad_norm": 0.09785987436771393,
      "learning_rate": 2.084696544536025e-05,
      "loss": 0.0093,
      "step": 2400
    },
    {
      "epoch": 3.6394654535115154,
      "eval_loss": 0.036392148584127426,
      "eval_runtime": 210.5922,
      "eval_samples_per_second": 5.57,
      "eval_steps_per_second": 5.57,
      "step": 2400
    },
    {
      "epoch": 3.6546298929011467,
      "grad_norm": 0.023465588688850403,
      "learning_rate": 2.0418201891040778e-05,
      "loss": 0.0105,
      "step": 2410
    },
    {
      "epoch": 3.669794332290778,
      "grad_norm": 0.1549590677022934,
      "learning_rate": 1.9992759351862772e-05,
      "loss": 0.0164,
      "step": 2420
    },
    {
      "epoch": 3.6849587716804093,
      "grad_norm": 0.24472041428089142,
      "learning_rate": 1.9570685590342463e-05,
      "loss": 0.0073,
      "step": 2430
    },
    {
      "epoch": 3.7001232110700406,
      "grad_norm": 0.27343687415122986,
      "learning_rate": 1.9152027990798748e-05,
      "loss": 0.0109,
      "step": 2440
    },
    {
      "epoch": 3.715287650459672,
      "grad_norm": 0.2274019867181778,
      "learning_rate": 1.8736833554033263e-05,
      "loss": 0.0095,
      "step": 2450
    },
    {
      "epoch": 3.715287650459672,
      "eval_loss": 0.0359375923871994,
      "eval_runtime": 210.6833,
      "eval_samples_per_second": 5.568,
      "eval_steps_per_second": 5.568,
      "step": 2450
    },
    {
      "epoch": 3.7304520898493037,
      "grad_norm": 0.16789063811302185,
      "learning_rate": 1.8325148892054093e-05,
      "loss": 0.0097,
      "step": 2460
    },
    {
      "epoch": 3.7456165292389345,
      "grad_norm": 0.17372868955135345,
      "learning_rate": 1.7917020222842662e-05,
      "loss": 0.0109,
      "step": 2470
    },
    {
      "epoch": 3.7607809686285663,
      "grad_norm": 0.10489284247159958,
      "learning_rate": 1.751249336516513e-05,
      "loss": 0.01,
      "step": 2480
    },
    {
      "epoch": 3.775945408018197,
      "grad_norm": 0.20585931837558746,
      "learning_rate": 1.7111613733428522e-05,
      "loss": 0.009,
      "step": 2490
    },
    {
      "epoch": 3.791109847407829,
      "grad_norm": 0.1862187534570694,
      "learning_rate": 1.671442633258218e-05,
      "loss": 0.0102,
      "step": 2500
    },
    {
      "epoch": 3.791109847407829,
      "eval_loss": 0.03812110051512718,
      "eval_runtime": 210.6984,
      "eval_samples_per_second": 5.567,
      "eval_steps_per_second": 5.567,
      "step": 2500
    },
    {
      "epoch": 3.8062742867974597,
      "grad_norm": 0.4300471246242523,
      "learning_rate": 1.6320975753065453e-05,
      "loss": 0.0085,
      "step": 2510
    },
    {
      "epoch": 3.8214387261870915,
      "grad_norm": 0.3116505742073059,
      "learning_rate": 1.5931306165801452e-05,
      "loss": 0.0063,
      "step": 2520
    },
    {
      "epoch": 3.8366031655767223,
      "grad_norm": 0.14136524498462677,
      "learning_rate": 1.554546131723848e-05,
      "loss": 0.0103,
      "step": 2530
    },
    {
      "epoch": 3.851767604966354,
      "grad_norm": 0.35863766074180603,
      "learning_rate": 1.5163484524438516e-05,
      "loss": 0.009,
      "step": 2540
    },
    {
      "epoch": 3.8669320443559854,
      "grad_norm": 0.24358558654785156,
      "learning_rate": 1.4785418670214496e-05,
      "loss": 0.0137,
      "step": 2550
    },
    {
      "epoch": 3.8669320443559854,
      "eval_loss": 0.03836597874760628,
      "eval_runtime": 210.9446,
      "eval_samples_per_second": 5.561,
      "eval_steps_per_second": 5.561,
      "step": 2550
    },
    {
      "epoch": 3.8820964837456167,
      "grad_norm": 0.08940009027719498,
      "learning_rate": 1.4411306198315777e-05,
      "loss": 0.0075,
      "step": 2560
    },
    {
      "epoch": 3.897260923135248,
      "grad_norm": 0.161200612783432,
      "learning_rate": 1.4041189108663421e-05,
      "loss": 0.0106,
      "step": 2570
    },
    {
      "epoch": 3.9124253625248793,
      "grad_norm": 0.06406702101230621,
      "learning_rate": 1.3675108952634824e-05,
      "loss": 0.0082,
      "step": 2580
    },
    {
      "epoch": 3.9275898019145106,
      "grad_norm": 0.24487881362438202,
      "learning_rate": 1.3313106828399147e-05,
      "loss": 0.0091,
      "step": 2590
    },
    {
      "epoch": 3.942754241304142,
      "grad_norm": 0.3004579544067383,
      "learning_rate": 1.2955223376303205e-05,
      "loss": 0.01,
      "step": 2600
    },
    {
      "epoch": 3.942754241304142,
      "eval_loss": 0.03762030228972435,
      "eval_runtime": 211.0716,
      "eval_samples_per_second": 5.557,
      "eval_steps_per_second": 5.557,
      "step": 2600
    },
    {
      "epoch": 3.957918680693773,
      "grad_norm": 0.1428319662809372,
      "learning_rate": 1.2601498774309112e-05,
      "loss": 0.0098,
      "step": 2610
    },
    {
      "epoch": 3.9730831200834045,
      "grad_norm": 0.314528226852417,
      "learning_rate": 1.2251972733483612e-05,
      "loss": 0.011,
      "step": 2620
    },
    {
      "epoch": 3.9882475594730358,
      "grad_norm": 0.549127459526062,
      "learning_rate": 1.19066844935399e-05,
      "loss": 0.0112,
      "step": 2630
    },
    {
      "epoch": 4.003411998862667,
      "grad_norm": 0.09976401925086975,
      "learning_rate": 1.156567281843241e-05,
      "loss": 0.01,
      "step": 2640
    },
    {
      "epoch": 4.018576438252298,
      "grad_norm": 0.2533574104309082,
      "learning_rate": 1.1228975992004842e-05,
      "loss": 0.0058,
      "step": 2650
    },
    {
      "epoch": 4.018576438252298,
      "eval_loss": 0.0389280766248703,
      "eval_runtime": 210.8888,
      "eval_samples_per_second": 5.562,
      "eval_steps_per_second": 5.562,
      "step": 2650
    },
    {
      "epoch": 4.03374087764193,
      "grad_norm": 0.19056229293346405,
      "learning_rate": 1.0896631813692377e-05,
      "loss": 0.0074,
      "step": 2660
    },
    {
      "epoch": 4.048905317031561,
      "grad_norm": 0.08204106986522675,
      "learning_rate": 1.0568677594277899e-05,
      "loss": 0.0068,
      "step": 2670
    },
    {
      "epoch": 4.064069756421192,
      "grad_norm": 0.07987801730632782,
      "learning_rate": 1.0245150151703509e-05,
      "loss": 0.004,
      "step": 2680
    },
    {
      "epoch": 4.079234195810824,
      "grad_norm": 0.12268516421318054,
      "learning_rate": 9.926085806936918e-06,
      "loss": 0.0054,
      "step": 2690
    },
    {
      "epoch": 4.094398635200455,
      "grad_norm": 0.2826228141784668,
      "learning_rate": 9.611520379894068e-06,
      "loss": 0.0071,
      "step": 2700
    },
    {
      "epoch": 4.094398635200455,
      "eval_loss": 0.040721941739320755,
      "eval_runtime": 210.6065,
      "eval_samples_per_second": 5.57,
      "eval_steps_per_second": 5.57,
      "step": 2700
    },
    {
      "epoch": 4.109563074590087,
      "grad_norm": 0.18833203613758087,
      "learning_rate": 9.301489185417578e-06,
      "loss": 0.0051,
      "step": 2710
    },
    {
      "epoch": 4.1247275139797175,
      "grad_norm": 0.07186464220285416,
      "learning_rate": 8.996027029312314e-06,
      "loss": 0.0037,
      "step": 2720
    },
    {
      "epoch": 4.139891953369349,
      "grad_norm": 0.26854589581489563,
      "learning_rate": 8.695168204437721e-06,
      "loss": 0.0096,
      "step": 2730
    },
    {
      "epoch": 4.15505639275898,
      "grad_norm": 0.2191058099269867,
      "learning_rate": 8.398946486858029e-06,
      "loss": 0.005,
      "step": 2740
    },
    {
      "epoch": 4.170220832148612,
      "grad_norm": 0.04452141746878624,
      "learning_rate": 8.107395132050333e-06,
      "loss": 0.0051,
      "step": 2750
    },
    {
      "epoch": 4.170220832148612,
      "eval_loss": 0.04139946773648262,
      "eval_runtime": 210.785,
      "eval_samples_per_second": 5.565,
      "eval_steps_per_second": 5.565,
      "step": 2750
    },
    {
      "epoch": 4.185385271538243,
      "grad_norm": 0.11700636148452759,
      "learning_rate": 7.820546871171142e-06,
      "loss": 0.0047,
      "step": 2760
    },
    {
      "epoch": 4.200549710927874,
      "grad_norm": 0.09535209834575653,
      "learning_rate": 7.5384339073818145e-06,
      "loss": 0.0039,
      "step": 2770
    },
    {
      "epoch": 4.215714150317505,
      "grad_norm": 0.22407791018486023,
      "learning_rate": 7.261087912233228e-06,
      "loss": 0.0084,
      "step": 2780
    },
    {
      "epoch": 4.230878589707137,
      "grad_norm": 0.1446915864944458,
      "learning_rate": 6.988540022110235e-06,
      "loss": 0.0048,
      "step": 2790
    },
    {
      "epoch": 4.246043029096768,
      "grad_norm": 0.12982718646526337,
      "learning_rate": 6.720820834735969e-06,
      "loss": 0.0077,
      "step": 2800
    },
    {
      "epoch": 4.246043029096768,
      "eval_loss": 0.041931215673685074,
      "eval_runtime": 211.1736,
      "eval_samples_per_second": 5.555,
      "eval_steps_per_second": 5.555,
      "step": 2800
    },
    {
      "epoch": 4.2612074684864,
      "grad_norm": 0.3041308522224426,
      "learning_rate": 6.457960405736968e-06,
      "loss": 0.0073,
      "step": 2810
    },
    {
      "epoch": 4.2763719078760305,
      "grad_norm": 0.047563496977090836,
      "learning_rate": 6.199988245268778e-06,
      "loss": 0.0031,
      "step": 2820
    },
    {
      "epoch": 4.291536347265662,
      "grad_norm": 0.08547382056713104,
      "learning_rate": 5.946933314703157e-06,
      "loss": 0.0046,
      "step": 2830
    },
    {
      "epoch": 4.306700786655293,
      "grad_norm": 0.04036795720458031,
      "learning_rate": 5.698824023376531e-06,
      "loss": 0.0047,
      "step": 2840
    },
    {
      "epoch": 4.321865226044925,
      "grad_norm": 0.07473888993263245,
      "learning_rate": 5.455688225400802e-06,
      "loss": 0.0039,
      "step": 2850
    },
    {
      "epoch": 4.321865226044925,
      "eval_loss": 0.042382657527923584,
      "eval_runtime": 211.489,
      "eval_samples_per_second": 5.546,
      "eval_steps_per_second": 5.546,
      "step": 2850
    },
    {
      "epoch": 4.337029665434556,
      "grad_norm": 0.15295420587062836,
      "learning_rate": 5.217553216536098e-06,
      "loss": 0.008,
      "step": 2860
    },
    {
      "epoch": 4.3521941048241874,
      "grad_norm": 0.2737196385860443,
      "learning_rate": 4.984445731126542e-06,
      "loss": 0.0085,
      "step": 2870
    },
    {
      "epoch": 4.367358544213818,
      "grad_norm": 0.05751258507370949,
      "learning_rate": 4.756391939098853e-06,
      "loss": 0.0038,
      "step": 2880
    },
    {
      "epoch": 4.38252298360345,
      "grad_norm": 0.016851551830768585,
      "learning_rate": 4.533417443024374e-06,
      "loss": 0.0069,
      "step": 2890
    },
    {
      "epoch": 4.397687422993081,
      "grad_norm": 0.03521433472633362,
      "learning_rate": 4.315547275244769e-06,
      "loss": 0.002,
      "step": 2900
    },
    {
      "epoch": 4.397687422993081,
      "eval_loss": 0.04292920231819153,
      "eval_runtime": 211.4803,
      "eval_samples_per_second": 5.547,
      "eval_steps_per_second": 5.547,
      "step": 2900
    },
    {
      "epoch": 4.412851862382713,
      "grad_norm": 0.37077462673187256,
      "learning_rate": 4.10280589506179e-06,
      "loss": 0.006,
      "step": 2910
    },
    {
      "epoch": 4.4280163017723435,
      "grad_norm": 0.21733137965202332,
      "learning_rate": 3.895217185991312e-06,
      "loss": 0.0046,
      "step": 2920
    },
    {
      "epoch": 4.443180741161975,
      "grad_norm": 0.20421354472637177,
      "learning_rate": 3.692804453082038e-06,
      "loss": 0.0048,
      "step": 2930
    },
    {
      "epoch": 4.458345180551606,
      "grad_norm": 0.028284486383199692,
      "learning_rate": 3.495590420299194e-06,
      "loss": 0.0046,
      "step": 2940
    },
    {
      "epoch": 4.473509619941238,
      "grad_norm": 0.2422058880329132,
      "learning_rate": 3.3035972279733207e-06,
      "loss": 0.0036,
      "step": 2950
    },
    {
      "epoch": 4.473509619941238,
      "eval_loss": 0.0432889424264431,
      "eval_runtime": 211.2005,
      "eval_samples_per_second": 5.554,
      "eval_steps_per_second": 5.554,
      "step": 2950
    },
    {
      "epoch": 4.488674059330869,
      "grad_norm": 0.11122108995914459,
      "learning_rate": 3.116846430314796e-06,
      "loss": 0.0062,
      "step": 2960
    },
    {
      "epoch": 4.5038384987205005,
      "grad_norm": 0.05903605371713638,
      "learning_rate": 2.9353589929939087e-06,
      "loss": 0.0037,
      "step": 2970
    },
    {
      "epoch": 4.519002938110132,
      "grad_norm": 0.15656092762947083,
      "learning_rate": 2.7591552907872574e-06,
      "loss": 0.0036,
      "step": 2980
    },
    {
      "epoch": 4.534167377499763,
      "grad_norm": 0.07800609618425369,
      "learning_rate": 2.5882551052902883e-06,
      "loss": 0.0081,
      "step": 2990
    },
    {
      "epoch": 4.549331816889394,
      "grad_norm": 0.46738412976264954,
      "learning_rate": 2.4226776226965453e-06,
      "loss": 0.0083,
      "step": 3000
    },
    {
      "epoch": 4.549331816889394,
      "eval_loss": 0.043931957334280014,
      "eval_runtime": 210.8491,
      "eval_samples_per_second": 5.563,
      "eval_steps_per_second": 5.563,
      "step": 3000
    },
    {
      "epoch": 4.564496256279026,
      "grad_norm": 0.08746257424354553,
      "learning_rate": 2.262441431643697e-06,
      "loss": 0.0064,
      "step": 3010
    },
    {
      "epoch": 4.579660695668657,
      "grad_norm": 0.02448512427508831,
      "learning_rate": 2.1075645211266927e-06,
      "loss": 0.0025,
      "step": 3020
    },
    {
      "epoch": 4.594825135058288,
      "grad_norm": 0.3164784610271454,
      "learning_rate": 1.9580642784782053e-06,
      "loss": 0.0056,
      "step": 3030
    },
    {
      "epoch": 4.60998957444792,
      "grad_norm": 0.17482583224773407,
      "learning_rate": 1.813957487416651e-06,
      "loss": 0.003,
      "step": 3040
    },
    {
      "epoch": 4.625154013837551,
      "grad_norm": 0.1300291121006012,
      "learning_rate": 1.6752603261619315e-06,
      "loss": 0.0049,
      "step": 3050
    },
    {
      "epoch": 4.625154013837551,
      "eval_loss": 0.043927934020757675,
      "eval_runtime": 211.115,
      "eval_samples_per_second": 5.556,
      "eval_steps_per_second": 5.556,
      "step": 3050
    },
    {
      "epoch": 4.640318453227183,
      "grad_norm": 0.22304365038871765,
      "learning_rate": 1.541988365619207e-06,
      "loss": 0.0065,
      "step": 3060
    },
    {
      "epoch": 4.6554828926168135,
      "grad_norm": 0.28815048933029175,
      "learning_rate": 1.4141565676307865e-06,
      "loss": 0.0047,
      "step": 3070
    },
    {
      "epoch": 4.670647332006445,
      "grad_norm": 0.03885040432214737,
      "learning_rate": 1.291779283296468e-06,
      "loss": 0.0026,
      "step": 3080
    },
    {
      "epoch": 4.685811771396076,
      "grad_norm": 0.07859209179878235,
      "learning_rate": 1.1748702513623922e-06,
      "loss": 0.0042,
      "step": 3090
    },
    {
      "epoch": 4.700976210785708,
      "grad_norm": 0.11053480952978134,
      "learning_rate": 1.0634425966786155e-06,
      "loss": 0.0033,
      "step": 3100
    },
    {
      "epoch": 4.700976210785708,
      "eval_loss": 0.04425197094678879,
      "eval_runtime": 211.2975,
      "eval_samples_per_second": 5.551,
      "eval_steps_per_second": 5.551,
      "step": 3100
    },
    {
      "epoch": 4.716140650175339,
      "grad_norm": 0.192493736743927,
      "learning_rate": 9.575088287257118e-07,
      "loss": 0.0045,
      "step": 3110
    },
    {
      "epoch": 4.73130508956497,
      "grad_norm": 0.24704289436340332,
      "learning_rate": 8.570808402103392e-07,
      "loss": 0.0057,
      "step": 3120
    },
    {
      "epoch": 4.746469528954601,
      "grad_norm": 0.20354284346103668,
      "learning_rate": 7.62169905730109e-07,
      "loss": 0.0032,
      "step": 3130
    },
    {
      "epoch": 4.761633968344233,
      "grad_norm": 0.3002487123012543,
      "learning_rate": 6.727866805078531e-07,
      "loss": 0.0035,
      "step": 3140
    },
    {
      "epoch": 4.776798407733864,
      "grad_norm": 0.03897647187113762,
      "learning_rate": 5.889411991953975e-07,
      "loss": 0.0034,
      "step": 3150
    },
    {
      "epoch": 4.776798407733864,
      "eval_loss": 0.044278133660554886,
      "eval_runtime": 211.6727,
      "eval_samples_per_second": 5.542,
      "eval_steps_per_second": 5.542,
      "step": 3150
    },
    {
      "epoch": 4.791962847123496,
      "grad_norm": 0.1218685507774353,
      "learning_rate": 5.106428747470137e-07,
      "loss": 0.0043,
      "step": 3160
    },
    {
      "epoch": 4.8071272865131265,
      "grad_norm": 0.12456027418375015,
      "learning_rate": 4.37900497362681e-07,
      "loss": 0.0049,
      "step": 3170
    },
    {
      "epoch": 4.822291725902758,
      "grad_norm": 0.16677603125572205,
      "learning_rate": 3.7072223350124167e-07,
      "loss": 0.0036,
      "step": 3180
    },
    {
      "epoch": 4.837456165292389,
      "grad_norm": 0.20543785393238068,
      "learning_rate": 3.0911562496358517e-07,
      "loss": 0.0055,
      "step": 3190
    },
    {
      "epoch": 4.852620604682021,
      "grad_norm": 0.16222603619098663,
      "learning_rate": 2.53087588045986e-07,
      "loss": 0.0038,
      "step": 3200
    },
    {
      "epoch": 4.852620604682021,
      "eval_loss": 0.04416884854435921,
      "eval_runtime": 211.7555,
      "eval_samples_per_second": 5.539,
      "eval_steps_per_second": 5.539,
      "step": 3200
    },
    {
      "epoch": 4.867785044071652,
      "grad_norm": 0.23848523199558258,
      "learning_rate": 2.026444127636029e-07,
      "loss": 0.0034,
      "step": 3210
    },
    {
      "epoch": 4.882949483461283,
      "grad_norm": 0.03002069890499115,
      "learning_rate": 1.577917621443825e-07,
      "loss": 0.004,
      "step": 3220
    },
    {
      "epoch": 4.898113922850914,
      "grad_norm": 0.4608694314956665,
      "learning_rate": 1.185346715932345e-07,
      "loss": 0.0064,
      "step": 3230
    },
    {
      "epoch": 4.913278362240546,
      "grad_norm": 0.3726924657821655,
      "learning_rate": 8.48775483267783e-08,
      "loss": 0.0052,
      "step": 3240
    },
    {
      "epoch": 4.928442801630177,
      "grad_norm": 0.046221598982810974,
      "learning_rate": 5.6824170878544414e-08,
      "loss": 0.0029,
      "step": 3250
    },
    {
      "epoch": 4.928442801630177,
      "eval_loss": 0.04429050534963608,
      "eval_runtime": 211.597,
      "eval_samples_per_second": 5.544,
      "eval_steps_per_second": 5.544,
      "step": 3250
    },
    {
      "epoch": 4.943607241019809,
      "grad_norm": 0.1485227346420288,
      "learning_rate": 3.4377688674774866e-08,
      "loss": 0.0034,
      "step": 3260
    },
    {
      "epoch": 4.9587716804094395,
      "grad_norm": 0.1220594197511673,
      "learning_rate": 1.7540621680850466e-08,
      "loss": 0.0059,
      "step": 3270
    },
    {
      "epoch": 4.973936119799071,
      "grad_norm": 0.1424579620361328,
      "learning_rate": 6.314860118400434e-09,
      "loss": 0.0044,
      "step": 3280
    },
    {
      "epoch": 4.989100559188703,
      "grad_norm": 0.03991411253809929,
      "learning_rate": 7.016642530777162e-10,
      "loss": 0.0055,
      "step": 3290
    },
    {
      "epoch": 4.996682778883518,
      "step": 3295,
      "total_flos": 8.907277487397274e+17,
      "train_loss": 0.048833281226591835,
      "train_runtime": 43274.7734,
      "train_samples_per_second": 1.219,
      "train_steps_per_second": 0.076
    }
  ],
  "logging_steps": 10,
  "max_steps": 3295,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.907277487397274e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}