{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 24.868123587038433,
  "eval_steps": 500,
  "global_step": 4125,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06028636021100226,
      "grad_norm": 14.353901908542692,
      "learning_rate": 5e-06,
      "loss": 1.0595,
      "step": 10
    },
    {
      "epoch": 0.12057272042200452,
      "grad_norm": 1.3180865308999647,
      "learning_rate": 5e-06,
      "loss": 0.9403,
      "step": 20
    },
    {
      "epoch": 0.1808590806330068,
      "grad_norm": 0.9782137323160548,
      "learning_rate": 5e-06,
      "loss": 0.89,
      "step": 30
    },
    {
      "epoch": 0.24114544084400905,
      "grad_norm": 1.3518217541449848,
      "learning_rate": 5e-06,
      "loss": 0.8618,
      "step": 40
    },
    {
      "epoch": 0.30143180105501133,
      "grad_norm": 0.8934828143592951,
      "learning_rate": 5e-06,
      "loss": 0.8447,
      "step": 50
    },
    {
      "epoch": 0.3617181612660136,
      "grad_norm": 0.9919016971761021,
      "learning_rate": 5e-06,
      "loss": 0.8318,
      "step": 60
    },
    {
      "epoch": 0.42200452147701584,
      "grad_norm": 0.9227497685306041,
      "learning_rate": 5e-06,
      "loss": 0.8177,
      "step": 70
    },
    {
      "epoch": 0.4822908816880181,
      "grad_norm": 0.9363087183332998,
      "learning_rate": 5e-06,
      "loss": 0.8106,
      "step": 80
    },
    {
      "epoch": 0.5425772418990203,
      "grad_norm": 1.0736624468965914,
      "learning_rate": 5e-06,
      "loss": 0.8052,
      "step": 90
    },
    {
      "epoch": 0.6028636021100227,
      "grad_norm": 1.353564519927803,
      "learning_rate": 5e-06,
      "loss": 0.7907,
      "step": 100
    },
    {
      "epoch": 0.6631499623210249,
      "grad_norm": 1.0455405570196998,
      "learning_rate": 5e-06,
      "loss": 0.7913,
      "step": 110
    },
    {
      "epoch": 0.7234363225320272,
      "grad_norm": 0.9569192147667731,
      "learning_rate": 5e-06,
      "loss": 0.7828,
      "step": 120
    },
    {
      "epoch": 0.7837226827430294,
      "grad_norm": 0.8298672277424939,
      "learning_rate": 5e-06,
      "loss": 0.7841,
      "step": 130
    },
    {
      "epoch": 0.8440090429540317,
      "grad_norm": 1.2437896161299962,
      "learning_rate": 5e-06,
      "loss": 0.7832,
      "step": 140
    },
    {
      "epoch": 0.9042954031650339,
      "grad_norm": 0.6290099294011746,
      "learning_rate": 5e-06,
      "loss": 0.7818,
      "step": 150
    },
    {
      "epoch": 0.9645817633760362,
      "grad_norm": 0.683050547264518,
      "learning_rate": 5e-06,
      "loss": 0.7807,
      "step": 160
    },
    {
      "epoch": 0.9947249434815373,
      "eval_loss": 0.7690147161483765,
      "eval_runtime": 114.9859,
      "eval_samples_per_second": 38.866,
      "eval_steps_per_second": 0.609,
      "step": 165
    },
    {
      "epoch": 1.0248681235870385,
      "grad_norm": 0.8375316338814506,
      "learning_rate": 5e-06,
      "loss": 0.7706,
      "step": 170
    },
    {
      "epoch": 1.0851544837980407,
      "grad_norm": 0.8491856111019711,
      "learning_rate": 5e-06,
      "loss": 0.7207,
      "step": 180
    },
    {
      "epoch": 1.145440844009043,
      "grad_norm": 1.00264106211233,
      "learning_rate": 5e-06,
      "loss": 0.7264,
      "step": 190
    },
    {
      "epoch": 1.2057272042200453,
      "grad_norm": 0.9524591725477187,
      "learning_rate": 5e-06,
      "loss": 0.7223,
      "step": 200
    },
    {
      "epoch": 1.2660135644310475,
      "grad_norm": 0.8200876998100735,
      "learning_rate": 5e-06,
      "loss": 0.7237,
      "step": 210
    },
    {
      "epoch": 1.3262999246420497,
      "grad_norm": 0.6972921329207209,
      "learning_rate": 5e-06,
      "loss": 0.7233,
      "step": 220
    },
    {
      "epoch": 1.3865862848530521,
      "grad_norm": 1.3472894348419586,
      "learning_rate": 5e-06,
      "loss": 0.7175,
      "step": 230
    },
    {
      "epoch": 1.4468726450640543,
      "grad_norm": 0.8080326013334395,
      "learning_rate": 5e-06,
      "loss": 0.7203,
      "step": 240
    },
    {
      "epoch": 1.5071590052750565,
      "grad_norm": 0.6690842132293575,
      "learning_rate": 5e-06,
      "loss": 0.7184,
      "step": 250
    },
    {
      "epoch": 1.5674453654860587,
      "grad_norm": 0.6680726643782731,
      "learning_rate": 5e-06,
      "loss": 0.7193,
      "step": 260
    },
    {
      "epoch": 1.627731725697061,
      "grad_norm": 0.6582407897638277,
      "learning_rate": 5e-06,
      "loss": 0.7171,
      "step": 270
    },
    {
      "epoch": 1.6880180859080633,
      "grad_norm": 0.6058749327234759,
      "learning_rate": 5e-06,
      "loss": 0.718,
      "step": 280
    },
    {
      "epoch": 1.7483044461190655,
      "grad_norm": 0.6771950284954183,
      "learning_rate": 5e-06,
      "loss": 0.717,
      "step": 290
    },
    {
      "epoch": 1.8085908063300677,
      "grad_norm": 0.620278984859575,
      "learning_rate": 5e-06,
      "loss": 0.721,
      "step": 300
    },
    {
      "epoch": 1.8688771665410702,
      "grad_norm": 0.7376417188133093,
      "learning_rate": 5e-06,
      "loss": 0.7187,
      "step": 310
    },
    {
      "epoch": 1.9291635267520724,
      "grad_norm": 0.841815742734191,
      "learning_rate": 5e-06,
      "loss": 0.7168,
      "step": 320
    },
    {
      "epoch": 1.9894498869630746,
      "grad_norm": 0.9694154154724743,
      "learning_rate": 5e-06,
      "loss": 0.7139,
      "step": 330
    },
    {
      "epoch": 1.995478522984175,
      "eval_loss": 0.7520068287849426,
      "eval_runtime": 112.311,
      "eval_samples_per_second": 39.791,
      "eval_steps_per_second": 0.623,
      "step": 331
    },
    {
      "epoch": 2.049736247174077,
      "grad_norm": 1.3857092066503702,
      "learning_rate": 5e-06,
      "loss": 0.6895,
      "step": 340
    },
    {
      "epoch": 2.110022607385079,
      "grad_norm": 0.9308939279408932,
      "learning_rate": 5e-06,
      "loss": 0.661,
      "step": 350
    },
    {
      "epoch": 2.1703089675960814,
      "grad_norm": 1.7049059921843441,
      "learning_rate": 5e-06,
      "loss": 0.6596,
      "step": 360
    },
    {
      "epoch": 2.230595327807084,
      "grad_norm": 1.1664692858604568,
      "learning_rate": 5e-06,
      "loss": 0.6626,
      "step": 370
    },
    {
      "epoch": 2.290881688018086,
      "grad_norm": 0.8099811209522968,
      "learning_rate": 5e-06,
      "loss": 0.6609,
      "step": 380
    },
    {
      "epoch": 2.351168048229088,
      "grad_norm": 0.7769380652338045,
      "learning_rate": 5e-06,
      "loss": 0.6633,
      "step": 390
    },
    {
      "epoch": 2.4114544084400906,
      "grad_norm": 1.0202300923789958,
      "learning_rate": 5e-06,
      "loss": 0.667,
      "step": 400
    },
    {
      "epoch": 2.4717407686510926,
      "grad_norm": 0.9120675026610565,
      "learning_rate": 5e-06,
      "loss": 0.6654,
      "step": 410
    },
    {
      "epoch": 2.532027128862095,
      "grad_norm": 0.8057600298686861,
      "learning_rate": 5e-06,
      "loss": 0.6643,
      "step": 420
    },
    {
      "epoch": 2.5923134890730974,
      "grad_norm": 0.6688016636812687,
      "learning_rate": 5e-06,
      "loss": 0.6657,
      "step": 430
    },
    {
      "epoch": 2.6525998492840994,
      "grad_norm": 0.7419224625762498,
      "learning_rate": 5e-06,
      "loss": 0.6669,
      "step": 440
    },
    {
      "epoch": 2.712886209495102,
      "grad_norm": 0.7294710798953756,
      "learning_rate": 5e-06,
      "loss": 0.664,
      "step": 450
    },
    {
      "epoch": 2.7731725697061043,
      "grad_norm": 0.7239503892331308,
      "learning_rate": 5e-06,
      "loss": 0.6628,
      "step": 460
    },
    {
      "epoch": 2.8334589299171062,
      "grad_norm": 0.6873637028133951,
      "learning_rate": 5e-06,
      "loss": 0.6634,
      "step": 470
    },
    {
      "epoch": 2.8937452901281087,
      "grad_norm": 0.7070671521244221,
      "learning_rate": 5e-06,
      "loss": 0.67,
      "step": 480
    },
    {
      "epoch": 2.9540316503391106,
      "grad_norm": 0.6467492636416056,
      "learning_rate": 5e-06,
      "loss": 0.6642,
      "step": 490
    },
    {
      "epoch": 2.9962321024868124,
      "eval_loss": 0.7524702548980713,
      "eval_runtime": 115.8914,
      "eval_samples_per_second": 38.562,
      "eval_steps_per_second": 0.604,
      "step": 497
    },
    {
      "epoch": 3.014318010550113,
      "grad_norm": 1.2600144566497484,
      "learning_rate": 5e-06,
      "loss": 0.6595,
      "step": 500
    },
    {
      "epoch": 3.0746043707611155,
      "grad_norm": 0.9439871932745402,
      "learning_rate": 5e-06,
      "loss": 0.6121,
      "step": 510
    },
    {
      "epoch": 3.1348907309721175,
      "grad_norm": 0.9269396364377557,
      "learning_rate": 5e-06,
      "loss": 0.6091,
      "step": 520
    },
    {
      "epoch": 3.19517709118312,
      "grad_norm": 1.4073687560739623,
      "learning_rate": 5e-06,
      "loss": 0.6146,
      "step": 530
    },
    {
      "epoch": 3.255463451394122,
      "grad_norm": 1.0909582280984875,
      "learning_rate": 5e-06,
      "loss": 0.6135,
      "step": 540
    },
    {
      "epoch": 3.3157498116051243,
      "grad_norm": 1.6855484277386785,
      "learning_rate": 5e-06,
      "loss": 0.6181,
      "step": 550
    },
    {
      "epoch": 3.3760361718161267,
      "grad_norm": 1.3960879287332164,
      "learning_rate": 5e-06,
      "loss": 0.6204,
      "step": 560
    },
    {
      "epoch": 3.4363225320271287,
      "grad_norm": 1.3111215484703225,
      "learning_rate": 5e-06,
      "loss": 0.6166,
      "step": 570
    },
    {
      "epoch": 3.496608892238131,
      "grad_norm": 1.4456626678129865,
      "learning_rate": 5e-06,
      "loss": 0.6187,
      "step": 580
    },
    {
      "epoch": 3.5568952524491335,
      "grad_norm": 0.9267972418344134,
      "learning_rate": 5e-06,
      "loss": 0.6198,
      "step": 590
    },
    {
      "epoch": 3.6171816126601355,
      "grad_norm": 1.5516401845259948,
      "learning_rate": 5e-06,
      "loss": 0.6216,
      "step": 600
    },
    {
      "epoch": 3.677467972871138,
      "grad_norm": 1.5014637761221297,
      "learning_rate": 5e-06,
      "loss": 0.618,
      "step": 610
    },
    {
      "epoch": 3.7377543330821403,
      "grad_norm": 1.1059687872166368,
      "learning_rate": 5e-06,
      "loss": 0.6185,
      "step": 620
    },
    {
      "epoch": 3.7980406932931423,
      "grad_norm": 1.2950526783062104,
      "learning_rate": 5e-06,
      "loss": 0.6207,
      "step": 630
    },
    {
      "epoch": 3.8583270535041447,
      "grad_norm": 0.9542402867171873,
      "learning_rate": 5e-06,
      "loss": 0.618,
      "step": 640
    },
    {
      "epoch": 3.918613413715147,
      "grad_norm": 0.7924967732032556,
      "learning_rate": 5e-06,
      "loss": 0.6163,
      "step": 650
    },
    {
      "epoch": 3.978899773926149,
      "grad_norm": 0.9450387587222794,
      "learning_rate": 5e-06,
      "loss": 0.6186,
      "step": 660
    },
    {
      "epoch": 3.99698568198945,
      "eval_loss": 0.7615257501602173,
      "eval_runtime": 111.9191,
      "eval_samples_per_second": 39.931,
      "eval_steps_per_second": 0.625,
      "step": 663
    },
    {
      "epoch": 4.039186134137151,
      "grad_norm": 0.9971689199182937,
      "learning_rate": 5e-06,
      "loss": 0.5958,
      "step": 670
    },
    {
      "epoch": 4.099472494348154,
      "grad_norm": 0.7965799505534403,
      "learning_rate": 5e-06,
      "loss": 0.5627,
      "step": 680
    },
    {
      "epoch": 4.159758854559156,
      "grad_norm": 0.8634698321641859,
      "learning_rate": 5e-06,
      "loss": 0.5642,
      "step": 690
    },
    {
      "epoch": 4.220045214770158,
      "grad_norm": 0.8538005391887241,
      "learning_rate": 5e-06,
      "loss": 0.5721,
      "step": 700
    },
    {
      "epoch": 4.280331574981161,
      "grad_norm": 0.8087517000342034,
      "learning_rate": 5e-06,
      "loss": 0.5702,
      "step": 710
    },
    {
      "epoch": 4.340617935192163,
      "grad_norm": 0.7416451232136964,
      "learning_rate": 5e-06,
      "loss": 0.5735,
      "step": 720
    },
    {
      "epoch": 4.400904295403165,
      "grad_norm": 0.683331985209307,
      "learning_rate": 5e-06,
      "loss": 0.576,
      "step": 730
    },
    {
      "epoch": 4.461190655614168,
      "grad_norm": 0.7531824862918711,
      "learning_rate": 5e-06,
      "loss": 0.5747,
      "step": 740
    },
    {
      "epoch": 4.52147701582517,
      "grad_norm": 0.6679489381777426,
      "learning_rate": 5e-06,
      "loss": 0.5788,
      "step": 750
    },
    {
      "epoch": 4.581763376036172,
      "grad_norm": 0.8184328898914945,
      "learning_rate": 5e-06,
      "loss": 0.5773,
      "step": 760
    },
    {
      "epoch": 4.642049736247174,
      "grad_norm": 0.7709343247764259,
      "learning_rate": 5e-06,
      "loss": 0.5747,
      "step": 770
    },
    {
      "epoch": 4.702336096458176,
      "grad_norm": 0.8008305213299817,
      "learning_rate": 5e-06,
      "loss": 0.5789,
      "step": 780
    },
    {
      "epoch": 4.762622456669178,
      "grad_norm": 0.770151242560135,
      "learning_rate": 5e-06,
      "loss": 0.5808,
      "step": 790
    },
    {
      "epoch": 4.822908816880181,
      "grad_norm": 0.8007315199653617,
      "learning_rate": 5e-06,
      "loss": 0.5766,
      "step": 800
    },
    {
      "epoch": 4.883195177091183,
      "grad_norm": 0.8153924743627864,
      "learning_rate": 5e-06,
      "loss": 0.5794,
      "step": 810
    },
    {
      "epoch": 4.943481537302185,
      "grad_norm": 0.7155989123366884,
      "learning_rate": 5e-06,
      "loss": 0.5777,
      "step": 820
    },
    {
      "epoch": 4.997739261492088,
      "eval_loss": 0.7784610390663147,
      "eval_runtime": 113.9486,
      "eval_samples_per_second": 39.219,
      "eval_steps_per_second": 0.614,
      "step": 829
    },
    {
      "epoch": 5.003767897513188,
      "grad_norm": 1.5968784870292558,
      "learning_rate": 5e-06,
      "loss": 0.5864,
      "step": 830
    },
    {
      "epoch": 5.06405425772419,
      "grad_norm": 1.3160101230444552,
      "learning_rate": 5e-06,
      "loss": 0.5136,
      "step": 840
    },
    {
      "epoch": 5.124340617935192,
      "grad_norm": 0.9218240884563489,
      "learning_rate": 5e-06,
      "loss": 0.514,
      "step": 850
    },
    {
      "epoch": 5.184626978146194,
      "grad_norm": 0.943292128668918,
      "learning_rate": 5e-06,
      "loss": 0.5129,
      "step": 860
    },
    {
      "epoch": 5.244913338357197,
      "grad_norm": 0.9210116431359558,
      "learning_rate": 5e-06,
      "loss": 0.5147,
      "step": 870
    },
    {
      "epoch": 5.305199698568199,
      "grad_norm": 0.9123827846726384,
      "learning_rate": 5e-06,
      "loss": 0.5152,
      "step": 880
    },
    {
      "epoch": 5.365486058779201,
      "grad_norm": 0.8967759483580899,
      "learning_rate": 5e-06,
      "loss": 0.5246,
      "step": 890
    },
    {
      "epoch": 5.425772418990204,
      "grad_norm": 1.0833892045281441,
      "learning_rate": 5e-06,
      "loss": 0.5173,
      "step": 900
    },
    {
      "epoch": 5.486058779201206,
      "grad_norm": 0.8452587501323132,
      "learning_rate": 5e-06,
      "loss": 0.5205,
      "step": 910
    },
    {
      "epoch": 5.546345139412208,
      "grad_norm": 0.9036749110113976,
      "learning_rate": 5e-06,
      "loss": 0.5213,
      "step": 920
    },
    {
      "epoch": 5.6066314996232105,
      "grad_norm": 0.8270647016134544,
      "learning_rate": 5e-06,
      "loss": 0.5214,
      "step": 930
    },
    {
      "epoch": 5.6669178598342125,
      "grad_norm": 0.760029827349666,
      "learning_rate": 5e-06,
      "loss": 0.5233,
      "step": 940
    },
    {
      "epoch": 5.7272042200452145,
      "grad_norm": 1.1680570931837515,
      "learning_rate": 5e-06,
      "loss": 0.5284,
      "step": 950
    },
    {
      "epoch": 5.787490580256217,
      "grad_norm": 1.0453134822125045,
      "learning_rate": 5e-06,
      "loss": 0.5273,
      "step": 960
    },
    {
      "epoch": 5.847776940467219,
      "grad_norm": 0.8063546838482688,
      "learning_rate": 5e-06,
      "loss": 0.5265,
      "step": 970
    },
    {
      "epoch": 5.908063300678221,
      "grad_norm": 0.8921967199529918,
      "learning_rate": 5e-06,
      "loss": 0.5316,
      "step": 980
    },
    {
      "epoch": 5.968349660889224,
      "grad_norm": 0.8942899369042694,
      "learning_rate": 5e-06,
      "loss": 0.5287,
      "step": 990
    },
    {
      "epoch": 5.998492840994725,
      "eval_loss": 0.815434992313385,
      "eval_runtime": 116.1895,
      "eval_samples_per_second": 38.463,
      "eval_steps_per_second": 0.602,
      "step": 995
    },
    {
      "epoch": 6.028636021100226,
      "grad_norm": 1.6908391333792119,
      "learning_rate": 5e-06,
      "loss": 0.5023,
      "step": 1000
    },
    {
      "epoch": 6.088922381311228,
      "grad_norm": 1.2916624414612379,
      "learning_rate": 5e-06,
      "loss": 0.4578,
      "step": 1010
    },
    {
      "epoch": 6.149208741522231,
      "grad_norm": 1.1040568786409888,
      "learning_rate": 5e-06,
      "loss": 0.454,
      "step": 1020
    },
    {
      "epoch": 6.209495101733233,
      "grad_norm": 1.0463754179790778,
      "learning_rate": 5e-06,
      "loss": 0.4592,
      "step": 1030
    },
    {
      "epoch": 6.269781461944235,
      "grad_norm": 1.0493678707628407,
      "learning_rate": 5e-06,
      "loss": 0.4582,
      "step": 1040
    },
    {
      "epoch": 6.330067822155238,
      "grad_norm": 1.105648007217362,
      "learning_rate": 5e-06,
      "loss": 0.4593,
      "step": 1050
    },
    {
      "epoch": 6.39035418236624,
      "grad_norm": 1.0291790248523935,
      "learning_rate": 5e-06,
      "loss": 0.4629,
      "step": 1060
    },
    {
      "epoch": 6.450640542577242,
      "grad_norm": 1.2514460402035084,
      "learning_rate": 5e-06,
      "loss": 0.4634,
      "step": 1070
    },
    {
      "epoch": 6.510926902788244,
      "grad_norm": 0.8518373738036685,
      "learning_rate": 5e-06,
      "loss": 0.4654,
      "step": 1080
    },
    {
      "epoch": 6.571213262999247,
      "grad_norm": 1.071998502866085,
      "learning_rate": 5e-06,
      "loss": 0.4663,
      "step": 1090
    },
    {
      "epoch": 6.6314996232102486,
      "grad_norm": 0.9000534182028347,
      "learning_rate": 5e-06,
      "loss": 0.4658,
      "step": 1100
    },
    {
      "epoch": 6.691785983421251,
      "grad_norm": 0.8995635602376101,
      "learning_rate": 5e-06,
      "loss": 0.469,
      "step": 1110
    },
    {
      "epoch": 6.752072343632253,
      "grad_norm": 0.9011803423576781,
      "learning_rate": 5e-06,
      "loss": 0.4715,
      "step": 1120
    },
    {
      "epoch": 6.812358703843255,
      "grad_norm": 0.9140285189147836,
      "learning_rate": 5e-06,
      "loss": 0.4763,
      "step": 1130
    },
    {
      "epoch": 6.872645064054257,
      "grad_norm": 1.0476263187239303,
      "learning_rate": 5e-06,
      "loss": 0.4781,
      "step": 1140
    },
    {
      "epoch": 6.93293142426526,
      "grad_norm": 1.1369535585860062,
      "learning_rate": 5e-06,
      "loss": 0.4702,
      "step": 1150
    },
    {
      "epoch": 6.993217784476262,
      "grad_norm": 1.2720345806831712,
      "learning_rate": 5e-06,
      "loss": 0.473,
      "step": 1160
    },
    {
      "epoch": 6.999246420497363,
      "eval_loss": 0.8709951639175415,
      "eval_runtime": 116.2672,
      "eval_samples_per_second": 38.437,
      "eval_steps_per_second": 0.602,
      "step": 1161
    },
    {
      "epoch": 7.053504144687264,
      "grad_norm": 1.6175687599813067,
      "learning_rate": 5e-06,
      "loss": 0.4154,
      "step": 1170
    },
    {
      "epoch": 7.113790504898267,
      "grad_norm": 1.2621822187556238,
      "learning_rate": 5e-06,
      "loss": 0.3985,
      "step": 1180
    },
    {
      "epoch": 7.174076865109269,
      "grad_norm": 1.3109771889362003,
      "learning_rate": 5e-06,
      "loss": 0.3937,
      "step": 1190
    },
    {
      "epoch": 7.234363225320271,
      "grad_norm": 1.1476426236895934,
      "learning_rate": 5e-06,
      "loss": 0.3986,
      "step": 1200
    },
    {
      "epoch": 7.294649585531274,
      "grad_norm": 1.1513686323462657,
      "learning_rate": 5e-06,
      "loss": 0.4003,
      "step": 1210
    },
    {
      "epoch": 7.354935945742276,
      "grad_norm": 1.2318284334824612,
      "learning_rate": 5e-06,
      "loss": 0.402,
      "step": 1220
    },
    {
      "epoch": 7.415222305953278,
      "grad_norm": 1.3293034219686575,
      "learning_rate": 5e-06,
      "loss": 0.4041,
      "step": 1230
    },
    {
      "epoch": 7.475508666164281,
      "grad_norm": 1.3126587055332364,
      "learning_rate": 5e-06,
      "loss": 0.4028,
      "step": 1240
    },
    {
      "epoch": 7.535795026375283,
      "grad_norm": 1.4541660160673635,
      "learning_rate": 5e-06,
      "loss": 0.4046,
      "step": 1250
    },
    {
      "epoch": 7.596081386586285,
      "grad_norm": 1.1677113248358406,
      "learning_rate": 5e-06,
      "loss": 0.4064,
      "step": 1260
    },
    {
      "epoch": 7.6563677467972875,
      "grad_norm": 1.2339252526811149,
      "learning_rate": 5e-06,
      "loss": 0.4057,
      "step": 1270
    },
    {
      "epoch": 7.7166541070082895,
      "grad_norm": 0.9795667261721808,
      "learning_rate": 5e-06,
      "loss": 0.4105,
      "step": 1280
    },
    {
      "epoch": 7.776940467219291,
      "grad_norm": 1.3122595751880688,
      "learning_rate": 5e-06,
      "loss": 0.4115,
      "step": 1290
    },
    {
      "epoch": 7.837226827430294,
      "grad_norm": 1.3363743108253807,
      "learning_rate": 5e-06,
      "loss": 0.4094,
      "step": 1300
    },
    {
      "epoch": 7.897513187641296,
      "grad_norm": 1.2278606177387434,
      "learning_rate": 5e-06,
      "loss": 0.411,
      "step": 1310
    },
    {
      "epoch": 7.957799547852298,
      "grad_norm": 1.099485555308857,
      "learning_rate": 5e-06,
      "loss": 0.4134,
      "step": 1320
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.9475046396255493,
      "eval_runtime": 118.217,
      "eval_samples_per_second": 37.803,
      "eval_steps_per_second": 0.592,
      "step": 1327
    },
    {
      "epoch": 8.0180859080633,
      "grad_norm": 2.747551612605114,
      "learning_rate": 5e-06,
      "loss": 0.3977,
      "step": 1330
    },
    {
      "epoch": 8.078372268274302,
      "grad_norm": 1.7260940558261295,
      "learning_rate": 5e-06,
      "loss": 0.3441,
      "step": 1340
    },
    {
      "epoch": 8.138658628485306,
      "grad_norm": 1.2842675938679606,
      "learning_rate": 5e-06,
      "loss": 0.3397,
      "step": 1350
    },
    {
      "epoch": 8.198944988696308,
      "grad_norm": 1.1398592198842168,
      "learning_rate": 5e-06,
      "loss": 0.3442,
      "step": 1360
    },
    {
      "epoch": 8.25923134890731,
      "grad_norm": 1.153557070127385,
      "learning_rate": 5e-06,
      "loss": 0.3454,
      "step": 1370
    },
    {
      "epoch": 8.319517709118312,
      "grad_norm": 1.3384373413827295,
      "learning_rate": 5e-06,
      "loss": 0.3438,
      "step": 1380
    },
    {
      "epoch": 8.379804069329314,
      "grad_norm": 1.3554470516864743,
      "learning_rate": 5e-06,
      "loss": 0.3493,
      "step": 1390
    },
    {
      "epoch": 8.440090429540316,
      "grad_norm": 1.325212009488347,
      "learning_rate": 5e-06,
      "loss": 0.3528,
      "step": 1400
    },
    {
      "epoch": 8.500376789751318,
      "grad_norm": 1.74556781346271,
      "learning_rate": 5e-06,
      "loss": 0.3548,
      "step": 1410
    },
    {
      "epoch": 8.560663149962322,
      "grad_norm": 1.3741632805919384,
      "learning_rate": 5e-06,
      "loss": 0.3518,
      "step": 1420
    },
    {
      "epoch": 8.620949510173324,
      "grad_norm": 1.3019952245252948,
      "learning_rate": 5e-06,
      "loss": 0.3521,
      "step": 1430
    },
    {
      "epoch": 8.681235870384326,
      "grad_norm": 1.100365748448614,
      "learning_rate": 5e-06,
      "loss": 0.3573,
      "step": 1440
    },
    {
      "epoch": 8.741522230595328,
      "grad_norm": 1.3911327396968252,
      "learning_rate": 5e-06,
      "loss": 0.3574,
      "step": 1450
    },
    {
      "epoch": 8.80180859080633,
      "grad_norm": 1.3853784407579808,
      "learning_rate": 5e-06,
      "loss": 0.3606,
      "step": 1460
    },
    {
      "epoch": 8.862094951017333,
      "grad_norm": 1.278532404930149,
      "learning_rate": 5e-06,
      "loss": 0.3567,
      "step": 1470
    },
    {
      "epoch": 8.922381311228335,
      "grad_norm": 1.207780040603517,
      "learning_rate": 5e-06,
      "loss": 0.3565,
      "step": 1480
    },
    {
      "epoch": 8.982667671439337,
      "grad_norm": 1.262306381996002,
      "learning_rate": 5e-06,
      "loss": 0.3615,
      "step": 1490
    },
    {
      "epoch": 8.994724943481538,
      "eval_loss": 1.0203213691711426,
      "eval_runtime": 114.5597,
      "eval_samples_per_second": 39.01,
      "eval_steps_per_second": 0.611,
      "step": 1492
    },
    {
      "epoch": 9.04295403165034,
      "grad_norm": 1.9088952161459727,
      "learning_rate": 5e-06,
      "loss": 0.3188,
      "step": 1500
    },
    {
      "epoch": 9.103240391861341,
      "grad_norm": 1.5762821812746035,
      "learning_rate": 5e-06,
      "loss": 0.2921,
      "step": 1510
    },
    {
      "epoch": 9.163526752072343,
      "grad_norm": 1.892762439817433,
      "learning_rate": 5e-06,
      "loss": 0.2866,
      "step": 1520
    },
    {
      "epoch": 9.223813112283345,
      "grad_norm": 1.5993468972103952,
      "learning_rate": 5e-06,
      "loss": 0.2913,
      "step": 1530
    },
    {
      "epoch": 9.284099472494349,
      "grad_norm": 1.7180835806783503,
      "learning_rate": 5e-06,
      "loss": 0.2925,
      "step": 1540
    },
    {
      "epoch": 9.34438583270535,
      "grad_norm": 1.4283006573501846,
      "learning_rate": 5e-06,
      "loss": 0.2936,
      "step": 1550
    },
    {
      "epoch": 9.404672192916353,
      "grad_norm": 1.472590439034633,
      "learning_rate": 5e-06,
      "loss": 0.2957,
      "step": 1560
    },
    {
      "epoch": 9.464958553127355,
      "grad_norm": 1.6627373496351208,
      "learning_rate": 5e-06,
      "loss": 0.2968,
      "step": 1570
    },
    {
      "epoch": 9.525244913338357,
      "grad_norm": 1.6479857681742245,
      "learning_rate": 5e-06,
      "loss": 0.2968,
      "step": 1580
    },
    {
      "epoch": 9.585531273549359,
      "grad_norm": 1.451039229532527,
      "learning_rate": 5e-06,
      "loss": 0.2969,
      "step": 1590
    },
    {
      "epoch": 9.645817633760362,
      "grad_norm": 1.338137019241527,
      "learning_rate": 5e-06,
      "loss": 0.2999,
      "step": 1600
    },
    {
      "epoch": 9.706103993971364,
      "grad_norm": 1.4364529449975805,
      "learning_rate": 5e-06,
      "loss": 0.301,
      "step": 1610
    },
    {
      "epoch": 9.766390354182366,
      "grad_norm": 1.6181472883899526,
      "learning_rate": 5e-06,
      "loss": 0.3021,
      "step": 1620
    },
    {
      "epoch": 9.826676714393368,
      "grad_norm": 1.2929981294290136,
      "learning_rate": 5e-06,
      "loss": 0.3047,
      "step": 1630
    },
    {
      "epoch": 9.88696307460437,
      "grad_norm": 1.5569082204346147,
      "learning_rate": 5e-06,
      "loss": 0.3039,
      "step": 1640
    },
    {
      "epoch": 9.947249434815372,
      "grad_norm": 1.2611067636631264,
      "learning_rate": 5e-06,
      "loss": 0.3057,
      "step": 1650
    },
    {
      "epoch": 9.995478522984175,
      "eval_loss": 1.1177444458007812,
      "eval_runtime": 116.3468,
      "eval_samples_per_second": 38.411,
      "eval_steps_per_second": 0.602,
      "step": 1658
    },
    {
      "epoch": 10.007535795026376,
      "grad_norm": 1.793363117063337,
      "learning_rate": 5e-06,
      "loss": 0.3015,
      "step": 1660
    },
    {
      "epoch": 10.067822155237378,
      "grad_norm": 1.9015543562249504,
      "learning_rate": 5e-06,
      "loss": 0.2446,
      "step": 1670
    },
    {
      "epoch": 10.12810851544838,
      "grad_norm": 1.6166742540525376,
      "learning_rate": 5e-06,
      "loss": 0.2417,
      "step": 1680
    },
    {
      "epoch": 10.188394875659382,
      "grad_norm": 1.5826959730874415,
      "learning_rate": 5e-06,
      "loss": 0.2407,
      "step": 1690
    },
    {
      "epoch": 10.248681235870384,
      "grad_norm": 1.313614154478162,
      "learning_rate": 5e-06,
      "loss": 0.2421,
      "step": 1700
    },
    {
      "epoch": 10.308967596081386,
      "grad_norm": 1.8551338463863283,
      "learning_rate": 5e-06,
      "loss": 0.2432,
      "step": 1710
    },
    {
      "epoch": 10.369253956292388,
      "grad_norm": 1.3937043583045259,
      "learning_rate": 5e-06,
      "loss": 0.2426,
      "step": 1720
    },
    {
      "epoch": 10.429540316503392,
      "grad_norm": 1.4908947612613312,
      "learning_rate": 5e-06,
      "loss": 0.2458,
      "step": 1730
    },
    {
      "epoch": 10.489826676714394,
      "grad_norm": 1.3513472181146375,
      "learning_rate": 5e-06,
      "loss": 0.2468,
      "step": 1740
    },
    {
      "epoch": 10.550113036925396,
      "grad_norm": 1.2442859437459592,
      "learning_rate": 5e-06,
      "loss": 0.2472,
      "step": 1750
    },
    {
      "epoch": 10.610399397136398,
      "grad_norm": 1.61049205004398,
      "learning_rate": 5e-06,
      "loss": 0.2497,
      "step": 1760
    },
    {
      "epoch": 10.6706857573474,
      "grad_norm": 1.4687600663257643,
      "learning_rate": 5e-06,
      "loss": 0.2513,
      "step": 1770
    },
    {
      "epoch": 10.730972117558402,
      "grad_norm": 1.3544154919254643,
      "learning_rate": 5e-06,
      "loss": 0.2534,
      "step": 1780
    },
    {
      "epoch": 10.791258477769405,
      "grad_norm": 1.4310611416705883,
      "learning_rate": 5e-06,
      "loss": 0.2534,
      "step": 1790
    },
    {
      "epoch": 10.851544837980407,
      "grad_norm": 1.5722579222869695,
      "learning_rate": 5e-06,
      "loss": 0.2543,
      "step": 1800
    },
    {
      "epoch": 10.91183119819141,
      "grad_norm": 1.520704900113862,
      "learning_rate": 5e-06,
      "loss": 0.254,
      "step": 1810
    },
    {
      "epoch": 10.972117558402411,
      "grad_norm": 1.4393868378454422,
      "learning_rate": 5e-06,
      "loss": 0.2565,
      "step": 1820
    },
    {
      "epoch": 10.996232102486813,
      "eval_loss": 1.236780047416687,
      "eval_runtime": 115.9111,
      "eval_samples_per_second": 38.555,
      "eval_steps_per_second": 0.604,
      "step": 1824
    },
    {
      "epoch": 11.032403918613413,
      "grad_norm": 2.415959560977028,
      "learning_rate": 5e-06,
      "loss": 0.2285,
      "step": 1830
    },
    {
      "epoch": 11.092690278824415,
      "grad_norm": 1.5442333094800837,
      "learning_rate": 5e-06,
      "loss": 0.1975,
      "step": 1840
    },
    {
      "epoch": 11.152976639035419,
      "grad_norm": 2.1161682250412257,
      "learning_rate": 5e-06,
      "loss": 0.1968,
      "step": 1850
    },
    {
      "epoch": 11.213262999246421,
      "grad_norm": 1.6062184104077026,
      "learning_rate": 5e-06,
      "loss": 0.1967,
      "step": 1860
    },
    {
      "epoch": 11.273549359457423,
      "grad_norm": 1.6876217090206025,
      "learning_rate": 5e-06,
      "loss": 0.1937,
      "step": 1870
    },
    {
      "epoch": 11.333835719668425,
      "grad_norm": 1.756341727895367,
      "learning_rate": 5e-06,
      "loss": 0.1966,
      "step": 1880
    },
    {
      "epoch": 11.394122079879427,
      "grad_norm": 1.5508699967732422,
      "learning_rate": 5e-06,
      "loss": 0.1997,
      "step": 1890
    },
    {
      "epoch": 11.454408440090429,
      "grad_norm": 1.5669766694757912,
      "learning_rate": 5e-06,
      "loss": 0.2023,
      "step": 1900
    },
    {
      "epoch": 11.51469480030143,
      "grad_norm": 1.4502558429874435,
      "learning_rate": 5e-06,
      "loss": 0.202,
      "step": 1910
    },
    {
      "epoch": 11.574981160512435,
      "grad_norm": 1.432868245038754,
      "learning_rate": 5e-06,
      "loss": 0.2027,
      "step": 1920
    },
    {
      "epoch": 11.635267520723437,
      "grad_norm": 1.944951997696116,
      "learning_rate": 5e-06,
      "loss": 0.2051,
      "step": 1930
    },
    {
      "epoch": 11.695553880934439,
      "grad_norm": 1.5628919140219797,
      "learning_rate": 5e-06,
      "loss": 0.2047,
      "step": 1940
    },
    {
      "epoch": 11.75584024114544,
      "grad_norm": 1.8554418586433572,
      "learning_rate": 5e-06,
      "loss": 0.2036,
      "step": 1950
    },
    {
      "epoch": 11.816126601356443,
      "grad_norm": 1.5083234117594833,
      "learning_rate": 5e-06,
      "loss": 0.2057,
      "step": 1960
    },
    {
      "epoch": 11.876412961567446,
      "grad_norm": 1.6850324704140989,
      "learning_rate": 5e-06,
      "loss": 0.2084,
      "step": 1970
    },
    {
      "epoch": 11.936699321778448,
      "grad_norm": 1.518863675506513,
      "learning_rate": 5e-06,
      "loss": 0.2081,
      "step": 1980
    },
    {
      "epoch": 11.99698568198945,
      "grad_norm": 2.025449919349873,
      "learning_rate": 5e-06,
      "loss": 0.2099,
      "step": 1990
    },
    {
      "epoch": 11.99698568198945,
      "eval_loss": 1.3552354574203491,
      "eval_runtime": 116.1042,
      "eval_samples_per_second": 38.491,
      "eval_steps_per_second": 0.603,
      "step": 1990
    },
    {
      "epoch": 12.057272042200452,
      "grad_norm": 1.9707275891443514,
      "learning_rate": 5e-06,
      "loss": 0.161,
      "step": 2000
    },
    {
      "epoch": 12.117558402411454,
      "grad_norm": 1.60811233115537,
      "learning_rate": 5e-06,
      "loss": 0.1527,
      "step": 2010
    },
    {
      "epoch": 12.177844762622456,
      "grad_norm": 1.6751301942391925,
      "learning_rate": 5e-06,
      "loss": 0.1541,
      "step": 2020
    },
    {
      "epoch": 12.238131122833458,
      "grad_norm": 1.4492067853763235,
      "learning_rate": 5e-06,
      "loss": 0.1559,
      "step": 2030
    },
    {
      "epoch": 12.298417483044462,
      "grad_norm": 1.9142452633048677,
      "learning_rate": 5e-06,
      "loss": 0.1576,
      "step": 2040
    },
    {
      "epoch": 12.358703843255464,
      "grad_norm": 1.4385285667791092,
      "learning_rate": 5e-06,
      "loss": 0.1589,
      "step": 2050
    },
    {
      "epoch": 12.418990203466466,
      "grad_norm": 1.4928747098321213,
      "learning_rate": 5e-06,
      "loss": 0.1608,
      "step": 2060
    },
    {
      "epoch": 12.479276563677468,
      "grad_norm": 1.5019521655912937,
      "learning_rate": 5e-06,
      "loss": 0.1605,
      "step": 2070
    },
    {
      "epoch": 12.53956292388847,
      "grad_norm": 1.874845955363987,
      "learning_rate": 5e-06,
      "loss": 0.1612,
      "step": 2080
    },
    {
      "epoch": 12.599849284099472,
      "grad_norm": 1.885802643703012,
      "learning_rate": 5e-06,
      "loss": 0.1614,
      "step": 2090
    },
    {
      "epoch": 12.660135644310476,
      "grad_norm": 1.515486697883317,
      "learning_rate": 5e-06,
      "loss": 0.1622,
      "step": 2100
    },
    {
      "epoch": 12.720422004521478,
      "grad_norm": 1.7537896738178245,
      "learning_rate": 5e-06,
      "loss": 0.1644,
      "step": 2110
    },
    {
      "epoch": 12.78070836473248,
      "grad_norm": 1.6471681441773147,
      "learning_rate": 5e-06,
      "loss": 0.1647,
      "step": 2120
    },
    {
      "epoch": 12.840994724943481,
      "grad_norm": 1.590934554816475,
      "learning_rate": 5e-06,
      "loss": 0.1648,
      "step": 2130
    },
    {
      "epoch": 12.901281085154483,
      "grad_norm": 1.7730249951617123,
      "learning_rate": 5e-06,
      "loss": 0.1656,
      "step": 2140
    },
    {
      "epoch": 12.961567445365485,
      "grad_norm": 1.6645957171669876,
      "learning_rate": 5e-06,
      "loss": 0.1676,
      "step": 2150
    },
    {
      "epoch": 12.997739261492088,
      "eval_loss": 1.5070669651031494,
      "eval_runtime": 116.1666,
      "eval_samples_per_second": 38.471,
      "eval_steps_per_second": 0.603,
      "step": 2156
    },
    {
      "epoch": 13.02185380557649,
      "grad_norm": 2.462408072338029,
      "learning_rate": 5e-06,
      "loss": 0.1517,
      "step": 2160
    },
    {
      "epoch": 13.082140165787491,
      "grad_norm": 2.4318360950844804,
      "learning_rate": 5e-06,
      "loss": 0.1195,
      "step": 2170
    },
    {
      "epoch": 13.142426525998493,
      "grad_norm": 1.7426593578459117,
      "learning_rate": 5e-06,
      "loss": 0.1201,
      "step": 2180
    },
    {
      "epoch": 13.202712886209495,
      "grad_norm": 2.2383839547509696,
      "learning_rate": 5e-06,
      "loss": 0.1191,
      "step": 2190
    },
    {
      "epoch": 13.262999246420497,
      "grad_norm": 1.5553317913023317,
      "learning_rate": 5e-06,
      "loss": 0.118,
      "step": 2200
    },
    {
      "epoch": 13.323285606631499,
      "grad_norm": 1.4669913420252614,
      "learning_rate": 5e-06,
      "loss": 0.119,
      "step": 2210
    },
    {
      "epoch": 13.383571966842501,
      "grad_norm": 1.8080283494554128,
      "learning_rate": 5e-06,
      "loss": 0.1196,
      "step": 2220
    },
    {
      "epoch": 13.443858327053505,
      "grad_norm": 1.646647327329277,
      "learning_rate": 5e-06,
      "loss": 0.1233,
      "step": 2230
    },
    {
      "epoch": 13.504144687264507,
      "grad_norm": 1.8533611593149397,
      "learning_rate": 5e-06,
      "loss": 0.1241,
      "step": 2240
    },
    {
      "epoch": 13.564431047475509,
      "grad_norm": 1.6777607818909321,
      "learning_rate": 5e-06,
      "loss": 0.126,
      "step": 2250
    },
    {
      "epoch": 13.62471740768651,
      "grad_norm": 1.7745416454251923,
      "learning_rate": 5e-06,
      "loss": 0.1266,
      "step": 2260
    },
    {
      "epoch": 13.685003767897513,
      "grad_norm": 1.7096006593564836,
      "learning_rate": 5e-06,
      "loss": 0.1282,
      "step": 2270
    },
    {
      "epoch": 13.745290128108515,
      "grad_norm": 1.8638678064551115,
      "learning_rate": 5e-06,
      "loss": 0.1267,
      "step": 2280
    },
    {
      "epoch": 13.805576488319518,
      "grad_norm": 1.4262510400727582,
      "learning_rate": 5e-06,
      "loss": 0.1273,
      "step": 2290
    },
    {
      "epoch": 13.86586284853052,
      "grad_norm": 1.8828592364046501,
      "learning_rate": 5e-06,
      "loss": 0.1269,
      "step": 2300
    },
    {
      "epoch": 13.926149208741522,
      "grad_norm": 1.6583448018945106,
      "learning_rate": 5e-06,
      "loss": 0.1273,
      "step": 2310
    },
    {
      "epoch": 13.986435568952524,
      "grad_norm": 1.548491818827999,
      "learning_rate": 5e-06,
      "loss": 0.1283,
      "step": 2320
    },
    {
      "epoch": 13.998492840994725,
      "eval_loss": 1.6323552131652832,
      "eval_runtime": 115.7608,
      "eval_samples_per_second": 38.605,
      "eval_steps_per_second": 0.605,
      "step": 2322
    },
    {
      "epoch": 14.046721929163526,
      "grad_norm": 2.0937924453069203,
      "learning_rate": 5e-06,
      "loss": 0.1006,
      "step": 2330
    },
    {
      "epoch": 14.107008289374528,
      "grad_norm": 2.139024110314688,
      "learning_rate": 5e-06,
      "loss": 0.0897,
      "step": 2340
    },
    {
      "epoch": 14.167294649585532,
      "grad_norm": 1.5948191216888472,
      "learning_rate": 5e-06,
      "loss": 0.092,
      "step": 2350
    },
    {
      "epoch": 14.227581009796534,
      "grad_norm": 1.6158036371008326,
      "learning_rate": 5e-06,
      "loss": 0.0916,
      "step": 2360
    },
    {
      "epoch": 14.287867370007536,
      "grad_norm": 1.6380184320632138,
      "learning_rate": 5e-06,
      "loss": 0.0946,
      "step": 2370
    },
    {
      "epoch": 14.348153730218538,
      "grad_norm": 1.5125683894811608,
      "learning_rate": 5e-06,
      "loss": 0.0926,
      "step": 2380
    },
    {
      "epoch": 14.40844009042954,
      "grad_norm": 1.8291017955611948,
      "learning_rate": 5e-06,
      "loss": 0.0967,
      "step": 2390
    },
    {
      "epoch": 14.468726450640542,
      "grad_norm": 1.6743144432735446,
      "learning_rate": 5e-06,
      "loss": 0.097,
      "step": 2400
    },
    {
      "epoch": 14.529012810851544,
      "grad_norm": 1.7276135456842643,
      "learning_rate": 5e-06,
      "loss": 0.096,
      "step": 2410
    },
    {
      "epoch": 14.589299171062548,
      "grad_norm": 1.734533902590376,
      "learning_rate": 5e-06,
      "loss": 0.0977,
      "step": 2420
    },
    {
      "epoch": 14.64958553127355,
      "grad_norm": 1.7106037137785368,
      "learning_rate": 5e-06,
      "loss": 0.0974,
      "step": 2430
    },
    {
      "epoch": 14.709871891484552,
      "grad_norm": 1.848167736152076,
      "learning_rate": 5e-06,
      "loss": 0.098,
      "step": 2440
    },
    {
      "epoch": 14.770158251695554,
      "grad_norm": 2.0729946870310907,
      "learning_rate": 5e-06,
      "loss": 0.0988,
      "step": 2450
    },
    {
      "epoch": 14.830444611906556,
      "grad_norm": 1.6522773246662863,
      "learning_rate": 5e-06,
      "loss": 0.0996,
      "step": 2460
    },
    {
      "epoch": 14.890730972117558,
      "grad_norm": 1.8165186575808512,
      "learning_rate": 5e-06,
      "loss": 0.1018,
      "step": 2470
    },
    {
      "epoch": 14.951017332328561,
      "grad_norm": 1.8324940323277443,
      "learning_rate": 5e-06,
      "loss": 0.1022,
      "step": 2480
    },
    {
      "epoch": 14.999246420497363,
      "eval_loss": 1.7541841268539429,
      "eval_runtime": 116.4567,
      "eval_samples_per_second": 38.375,
      "eval_steps_per_second": 0.601,
      "step": 2488
    },
    {
      "epoch": 15.011303692539563,
      "grad_norm": 1.7122346572238332,
      "learning_rate": 5e-06,
      "loss": 0.0969,
      "step": 2490
    },
    {
      "epoch": 15.071590052750565,
      "grad_norm": 1.7869038836513413,
      "learning_rate": 5e-06,
      "loss": 0.07,
      "step": 2500
    },
    {
      "epoch": 15.131876412961567,
      "grad_norm": 1.5692589038411329,
      "learning_rate": 5e-06,
      "loss": 0.069,
      "step": 2510
    },
    {
      "epoch": 15.19216277317257,
      "grad_norm": 1.5568543176798626,
      "learning_rate": 5e-06,
      "loss": 0.0714,
      "step": 2520
    },
    {
      "epoch": 15.252449133383571,
      "grad_norm": 1.8979966448489778,
      "learning_rate": 5e-06,
      "loss": 0.0721,
      "step": 2530
    },
    {
      "epoch": 15.312735493594575,
      "grad_norm": 1.6079155193057422,
      "learning_rate": 5e-06,
      "loss": 0.0734,
      "step": 2540
    },
    {
      "epoch": 15.373021853805577,
      "grad_norm": 1.5682954069117738,
      "learning_rate": 5e-06,
      "loss": 0.0731,
      "step": 2550
    },
    {
      "epoch": 15.433308214016579,
      "grad_norm": 1.607771288791899,
      "learning_rate": 5e-06,
      "loss": 0.0736,
      "step": 2560
    },
    {
      "epoch": 15.493594574227581,
      "grad_norm": 1.592532966261963,
      "learning_rate": 5e-06,
      "loss": 0.0751,
      "step": 2570
    },
    {
      "epoch": 15.553880934438583,
      "grad_norm": 2.106042369619277,
      "learning_rate": 5e-06,
      "loss": 0.0765,
      "step": 2580
    },
    {
      "epoch": 15.614167294649585,
      "grad_norm": 1.9643387382353286,
      "learning_rate": 5e-06,
      "loss": 0.0764,
      "step": 2590
    },
    {
      "epoch": 15.674453654860589,
      "grad_norm": 2.356472326588705,
      "learning_rate": 5e-06,
      "loss": 0.0766,
      "step": 2600
    },
    {
      "epoch": 15.73474001507159,
      "grad_norm": 1.7285204414699735,
      "learning_rate": 5e-06,
      "loss": 0.0768,
      "step": 2610
    },
    {
      "epoch": 15.795026375282593,
      "grad_norm": 1.656141946179993,
      "learning_rate": 5e-06,
      "loss": 0.0781,
      "step": 2620
    },
    {
      "epoch": 15.855312735493595,
      "grad_norm": 1.751656627220462,
      "learning_rate": 5e-06,
      "loss": 0.0778,
      "step": 2630
    },
    {
      "epoch": 15.915599095704597,
      "grad_norm": 1.9120339283286005,
      "learning_rate": 5e-06,
      "loss": 0.0777,
      "step": 2640
    },
    {
      "epoch": 15.975885455915598,
      "grad_norm": 1.7091309765672063,
      "learning_rate": 5e-06,
      "loss": 0.0779,
      "step": 2650
    },
    {
      "epoch": 16.0,
      "eval_loss": 1.8729431629180908,
      "eval_runtime": 130.3121,
      "eval_samples_per_second": 34.295,
      "eval_steps_per_second": 0.537,
      "step": 2654
    },
    {
      "epoch": 16.0361718161266,
      "grad_norm": 1.6873009916146906,
      "learning_rate": 5e-06,
      "loss": 0.0634,
      "step": 2660
    },
    {
      "epoch": 16.096458176337602,
      "grad_norm": 1.4499459070972593,
      "learning_rate": 5e-06,
      "loss": 0.0517,
      "step": 2670
    },
    {
      "epoch": 16.156744536548604,
      "grad_norm": 1.518251208097316,
      "learning_rate": 5e-06,
      "loss": 0.0532,
      "step": 2680
    },
    {
      "epoch": 16.217030896759606,
      "grad_norm": 1.7901282259982598,
      "learning_rate": 5e-06,
      "loss": 0.0547,
      "step": 2690
    },
    {
      "epoch": 16.277317256970612,
      "grad_norm": 1.6296841948609884,
      "learning_rate": 5e-06,
      "loss": 0.0553,
      "step": 2700
    },
    {
      "epoch": 16.337603617181614,
      "grad_norm": 1.9077348260850149,
      "learning_rate": 5e-06,
      "loss": 0.0551,
      "step": 2710
    },
    {
      "epoch": 16.397889977392616,
      "grad_norm": 1.6275792847492503,
      "learning_rate": 5e-06,
      "loss": 0.0556,
      "step": 2720
    },
    {
      "epoch": 16.458176337603618,
      "grad_norm": 1.8060616409571004,
      "learning_rate": 5e-06,
      "loss": 0.0569,
      "step": 2730
    },
    {
      "epoch": 16.51846269781462,
      "grad_norm": 1.6503072239622354,
      "learning_rate": 5e-06,
      "loss": 0.0568,
      "step": 2740
    },
    {
      "epoch": 16.578749058025622,
      "grad_norm": 1.5506661680891511,
      "learning_rate": 5e-06,
      "loss": 0.0572,
      "step": 2750
    },
    {
      "epoch": 16.639035418236624,
      "grad_norm": 1.6807035773937569,
      "learning_rate": 5e-06,
      "loss": 0.0583,
      "step": 2760
    },
    {
      "epoch": 16.699321778447626,
      "grad_norm": 1.794347099575113,
      "learning_rate": 5e-06,
      "loss": 0.0589,
      "step": 2770
    },
    {
      "epoch": 16.759608138658628,
      "grad_norm": 1.6112070974863837,
      "learning_rate": 5e-06,
      "loss": 0.0591,
      "step": 2780
    },
    {
      "epoch": 16.81989449886963,
      "grad_norm": 1.5309535448790481,
      "learning_rate": 5e-06,
      "loss": 0.0595,
      "step": 2790
    },
    {
      "epoch": 16.88018085908063,
      "grad_norm": 1.6962829999060645,
      "learning_rate": 5e-06,
      "loss": 0.0595,
      "step": 2800
    },
    {
      "epoch": 16.940467219291634,
      "grad_norm": 1.5160770218318527,
      "learning_rate": 5e-06,
      "loss": 0.0607,
      "step": 2810
    },
    {
      "epoch": 16.994724943481536,
      "eval_loss": 1.9862152338027954,
      "eval_runtime": 135.1067,
      "eval_samples_per_second": 33.078,
      "eval_steps_per_second": 0.518,
      "step": 2819
    },
    {
      "epoch": 17.00075357950264,
      "grad_norm": 2.340332137108714,
      "learning_rate": 5e-06,
      "loss": 0.0615,
      "step": 2820
    },
    {
      "epoch": 17.06103993971364,
      "grad_norm": 1.4236109373520875,
      "learning_rate": 5e-06,
      "loss": 0.0397,
      "step": 2830
    },
    {
      "epoch": 17.121326299924643,
      "grad_norm": 1.5339925411690258,
      "learning_rate": 5e-06,
      "loss": 0.0409,
      "step": 2840
    },
    {
      "epoch": 17.181612660135645,
      "grad_norm": 1.5223509157958206,
      "learning_rate": 5e-06,
      "loss": 0.0416,
      "step": 2850
    },
    {
      "epoch": 17.241899020346647,
      "grad_norm": 1.7444832258175693,
      "learning_rate": 5e-06,
      "loss": 0.0425,
      "step": 2860
    },
    {
      "epoch": 17.30218538055765,
      "grad_norm": 1.7829806133150694,
      "learning_rate": 5e-06,
      "loss": 0.0426,
      "step": 2870
    },
    {
      "epoch": 17.36247174076865,
      "grad_norm": 1.7776363684895005,
      "learning_rate": 5e-06,
      "loss": 0.0437,
      "step": 2880
    },
    {
      "epoch": 17.422758100979653,
      "grad_norm": 1.449976018411429,
      "learning_rate": 5e-06,
      "loss": 0.0434,
      "step": 2890
    },
    {
      "epoch": 17.483044461190655,
      "grad_norm": 1.4737183761321249,
      "learning_rate": 5e-06,
      "loss": 0.0443,
      "step": 2900
    },
    {
      "epoch": 17.543330821401657,
      "grad_norm": 1.5570813152529674,
      "learning_rate": 5e-06,
      "loss": 0.0439,
      "step": 2910
    },
    {
      "epoch": 17.60361718161266,
      "grad_norm": 1.6635463270221298,
      "learning_rate": 5e-06,
      "loss": 0.0451,
      "step": 2920
    },
    {
      "epoch": 17.66390354182366,
      "grad_norm": 1.5633000807042863,
      "learning_rate": 5e-06,
      "loss": 0.0452,
      "step": 2930
    },
    {
      "epoch": 17.724189902034666,
      "grad_norm": 1.5756718088358965,
      "learning_rate": 5e-06,
      "loss": 0.0453,
      "step": 2940
    },
    {
      "epoch": 17.78447626224567,
      "grad_norm": 1.5086652524908912,
      "learning_rate": 5e-06,
      "loss": 0.0461,
      "step": 2950
    },
    {
      "epoch": 17.84476262245667,
      "grad_norm": 1.543593492657027,
      "learning_rate": 5e-06,
      "loss": 0.0471,
      "step": 2960
    },
    {
      "epoch": 17.905048982667672,
      "grad_norm": 1.8850354234893694,
      "learning_rate": 5e-06,
      "loss": 0.0473,
      "step": 2970
    },
    {
      "epoch": 17.965335342878674,
      "grad_norm": 1.5531948749022353,
      "learning_rate": 5e-06,
      "loss": 0.0481,
      "step": 2980
    },
    {
      "epoch": 17.995478522984175,
      "eval_loss": 2.0547046661376953,
      "eval_runtime": 131.4004,
      "eval_samples_per_second": 34.011,
      "eval_steps_per_second": 0.533,
      "step": 2985
    },
    {
      "epoch": 18.025621703089676,
      "grad_norm": 1.5738867889136725,
      "learning_rate": 5e-06,
      "loss": 0.041,
      "step": 2990
    },
    {
      "epoch": 18.08590806330068,
      "grad_norm": 1.7413824765739192,
      "learning_rate": 5e-06,
      "loss": 0.0319,
      "step": 3000
    },
    {
      "epoch": 18.14619442351168,
      "grad_norm": 1.349257585093222,
      "learning_rate": 5e-06,
      "loss": 0.0321,
      "step": 3010
    },
    {
      "epoch": 18.206480783722682,
      "grad_norm": 1.5030527061221075,
      "learning_rate": 5e-06,
      "loss": 0.033,
      "step": 3020
    },
    {
      "epoch": 18.266767143933684,
      "grad_norm": 1.3839757017227572,
      "learning_rate": 5e-06,
      "loss": 0.0338,
      "step": 3030
    },
    {
      "epoch": 18.327053504144686,
      "grad_norm": 1.3639549234993298,
      "learning_rate": 5e-06,
      "loss": 0.034,
      "step": 3040
    },
    {
      "epoch": 18.38733986435569,
      "grad_norm": 1.491804900184454,
      "learning_rate": 5e-06,
      "loss": 0.0344,
      "step": 3050
    },
    {
      "epoch": 18.44762622456669,
      "grad_norm": 1.7182953986203733,
      "learning_rate": 5e-06,
      "loss": 0.0345,
      "step": 3060
    },
    {
      "epoch": 18.507912584777696,
      "grad_norm": 1.5927159416873042,
      "learning_rate": 5e-06,
      "loss": 0.035,
      "step": 3070
    },
    {
      "epoch": 18.568198944988698,
      "grad_norm": 1.48048633519207,
      "learning_rate": 5e-06,
      "loss": 0.0351,
      "step": 3080
    },
    {
      "epoch": 18.6284853051997,
      "grad_norm": 1.5724972628180884,
      "learning_rate": 5e-06,
      "loss": 0.0357,
      "step": 3090
    },
    {
      "epoch": 18.6887716654107,
      "grad_norm": 1.460368142482016,
      "learning_rate": 5e-06,
      "loss": 0.0367,
      "step": 3100
    },
    {
      "epoch": 18.749058025621704,
      "grad_norm": 1.5605571994020253,
      "learning_rate": 5e-06,
      "loss": 0.037,
      "step": 3110
    },
    {
      "epoch": 18.809344385832706,
      "grad_norm": 1.809142213762153,
      "learning_rate": 5e-06,
      "loss": 0.0364,
      "step": 3120
    },
    {
      "epoch": 18.869630746043708,
      "grad_norm": 1.6290220037965493,
      "learning_rate": 5e-06,
      "loss": 0.0367,
      "step": 3130
    },
    {
      "epoch": 18.92991710625471,
      "grad_norm": 1.441383521222635,
      "learning_rate": 5e-06,
      "loss": 0.0373,
      "step": 3140
    },
    {
      "epoch": 18.99020346646571,
      "grad_norm": 1.4650505575255777,
      "learning_rate": 5e-06,
      "loss": 0.038,
      "step": 3150
    },
    {
      "epoch": 18.99623210248681,
      "eval_loss": 2.1350691318511963,
      "eval_runtime": 116.4178,
      "eval_samples_per_second": 38.388,
      "eval_steps_per_second": 0.601,
      "step": 3151
    },
    {
      "epoch": 19.050489826676714,
      "grad_norm": 1.2364462769526054,
      "learning_rate": 5e-06,
      "loss": 0.0273,
      "step": 3160
    },
    {
      "epoch": 19.110776186887716,
      "grad_norm": 1.2847338125252712,
      "learning_rate": 5e-06,
      "loss": 0.0263,
      "step": 3170
    },
    {
      "epoch": 19.171062547098717,
      "grad_norm": 1.4364005652974698,
      "learning_rate": 5e-06,
      "loss": 0.027,
      "step": 3180
    },
    {
      "epoch": 19.23134890730972,
      "grad_norm": 1.4324790800038558,
      "learning_rate": 5e-06,
      "loss": 0.0278,
      "step": 3190
    },
    {
      "epoch": 19.291635267520725,
      "grad_norm": 1.3696161004609035,
      "learning_rate": 5e-06,
      "loss": 0.0268,
      "step": 3200
    },
    {
      "epoch": 19.351921627731727,
      "grad_norm": 1.4042401515519942,
      "learning_rate": 5e-06,
      "loss": 0.0276,
      "step": 3210
    },
    {
      "epoch": 19.41220798794273,
      "grad_norm": 1.4201103499411916,
      "learning_rate": 5e-06,
      "loss": 0.0282,
      "step": 3220
    },
    {
      "epoch": 19.47249434815373,
      "grad_norm": 1.4787498619597446,
      "learning_rate": 5e-06,
      "loss": 0.029,
      "step": 3230
    },
    {
      "epoch": 19.532780708364733,
      "grad_norm": 1.3042565121452792,
      "learning_rate": 5e-06,
      "loss": 0.0289,
      "step": 3240
    },
    {
      "epoch": 19.593067068575735,
      "grad_norm": 1.5616956908222805,
      "learning_rate": 5e-06,
      "loss": 0.0291,
      "step": 3250
    },
    {
      "epoch": 19.653353428786737,
      "grad_norm": 1.3916636056123413,
      "learning_rate": 5e-06,
      "loss": 0.0294,
      "step": 3260
    },
    {
      "epoch": 19.71363978899774,
      "grad_norm": 1.5340949993651216,
      "learning_rate": 5e-06,
      "loss": 0.0302,
      "step": 3270
    },
    {
      "epoch": 19.77392614920874,
      "grad_norm": 1.567955217260001,
      "learning_rate": 5e-06,
      "loss": 0.03,
      "step": 3280
    },
    {
      "epoch": 19.834212509419743,
      "grad_norm": 1.4614834017683382,
      "learning_rate": 5e-06,
      "loss": 0.0302,
      "step": 3290
    },
    {
      "epoch": 19.894498869630745,
      "grad_norm": 1.3384969584915662,
      "learning_rate": 5e-06,
      "loss": 0.0299,
      "step": 3300
    },
    {
      "epoch": 19.954785229841747,
      "grad_norm": 1.6563798754782657,
      "learning_rate": 5e-06,
      "loss": 0.0306,
      "step": 3310
    },
    {
      "epoch": 19.99698568198945,
      "eval_loss": 2.2255122661590576,
      "eval_runtime": 131.2328,
      "eval_samples_per_second": 34.054,
      "eval_steps_per_second": 0.533,
      "step": 3317
    },
    {
      "epoch": 20.015071590052752,
      "grad_norm": 1.2478793413376605,
      "learning_rate": 5e-06,
      "loss": 0.0282,
      "step": 3320
    },
    {
      "epoch": 20.075357950263754,
      "grad_norm": 1.5652614584389117,
      "learning_rate": 5e-06,
      "loss": 0.0201,
      "step": 3330
    },
    {
      "epoch": 20.135644310474756,
      "grad_norm": 1.6731787596786802,
      "learning_rate": 5e-06,
      "loss": 0.0215,
      "step": 3340
    },
    {
      "epoch": 20.195930670685758,
      "grad_norm": 1.2852656312690158,
      "learning_rate": 5e-06,
      "loss": 0.0215,
      "step": 3350
    },
    {
      "epoch": 20.25621703089676,
      "grad_norm": 1.189629660977167,
      "learning_rate": 5e-06,
      "loss": 0.0218,
      "step": 3360
    },
    {
      "epoch": 20.316503391107762,
      "grad_norm": 1.280211241755101,
      "learning_rate": 5e-06,
      "loss": 0.0225,
      "step": 3370
    },
    {
      "epoch": 20.376789751318764,
      "grad_norm": 1.3671510362477786,
      "learning_rate": 5e-06,
      "loss": 0.0233,
      "step": 3380
    },
    {
      "epoch": 20.437076111529766,
      "grad_norm": 1.2479455610052839,
      "learning_rate": 5e-06,
      "loss": 0.0235,
      "step": 3390
    },
    {
      "epoch": 20.497362471740768,
      "grad_norm": 1.3487469571257409,
      "learning_rate": 5e-06,
      "loss": 0.024,
      "step": 3400
    },
    {
      "epoch": 20.55764883195177,
      "grad_norm": 1.296539761041376,
      "learning_rate": 5e-06,
      "loss": 0.0239,
      "step": 3410
    },
    {
      "epoch": 20.617935192162772,
      "grad_norm": 1.3688473131734242,
      "learning_rate": 5e-06,
      "loss": 0.0243,
      "step": 3420
    },
    {
      "epoch": 20.678221552373774,
      "grad_norm": 1.3764265527684842,
      "learning_rate": 5e-06,
      "loss": 0.0246,
      "step": 3430
    },
    {
      "epoch": 20.738507912584776,
      "grad_norm": 1.5833602766458634,
      "learning_rate": 5e-06,
      "loss": 0.0249,
      "step": 3440
    },
    {
      "epoch": 20.79879427279578,
      "grad_norm": 1.528464092769155,
      "learning_rate": 5e-06,
      "loss": 0.0258,
      "step": 3450
    },
    {
      "epoch": 20.859080633006784,
      "grad_norm": 1.605313026206358,
      "learning_rate": 5e-06,
      "loss": 0.026,
      "step": 3460
    },
    {
      "epoch": 20.919366993217785,
      "grad_norm": 1.3923344492334147,
      "learning_rate": 5e-06,
      "loss": 0.0259,
      "step": 3470
    },
    {
      "epoch": 20.979653353428787,
      "grad_norm": 1.4204063096393778,
      "learning_rate": 5e-06,
      "loss": 0.0256,
      "step": 3480
    },
    {
      "epoch": 20.997739261492086,
      "eval_loss": 2.269946813583374,
      "eval_runtime": 116.5456,
      "eval_samples_per_second": 38.345,
      "eval_steps_per_second": 0.601,
      "step": 3483
    },
    {
      "epoch": 21.03993971363979,
      "grad_norm": 1.310936775235027,
      "learning_rate": 5e-06,
      "loss": 0.0208,
      "step": 3490
    },
    {
      "epoch": 21.10022607385079,
      "grad_norm": 1.1547005623677518,
      "learning_rate": 5e-06,
      "loss": 0.0186,
      "step": 3500
    },
    {
      "epoch": 21.160512434061793,
      "grad_norm": 1.2938934451769595,
      "learning_rate": 5e-06,
      "loss": 0.0194,
      "step": 3510
    },
    {
      "epoch": 21.220798794272795,
      "grad_norm": 1.484320867371485,
      "learning_rate": 5e-06,
      "loss": 0.0196,
      "step": 3520
    },
    {
      "epoch": 21.281085154483797,
      "grad_norm": 1.4005288463687333,
      "learning_rate": 5e-06,
      "loss": 0.02,
      "step": 3530
    },
    {
      "epoch": 21.3413715146948,
      "grad_norm": 1.6013274204687469,
      "learning_rate": 5e-06,
      "loss": 0.0201,
      "step": 3540
    },
    {
      "epoch": 21.4016578749058,
      "grad_norm": 1.4800630752519932,
      "learning_rate": 5e-06,
      "loss": 0.02,
      "step": 3550
    },
    {
      "epoch": 21.461944235116803,
      "grad_norm": 1.497381379010042,
      "learning_rate": 5e-06,
      "loss": 0.0203,
      "step": 3560
    },
    {
      "epoch": 21.52223059532781,
      "grad_norm": 1.2261010629715117,
      "learning_rate": 5e-06,
      "loss": 0.0208,
      "step": 3570
    },
    {
      "epoch": 21.58251695553881,
      "grad_norm": 1.353310824251755,
      "learning_rate": 5e-06,
      "loss": 0.0211,
      "step": 3580
    },
    {
      "epoch": 21.642803315749813,
      "grad_norm": 1.2550244372657888,
      "learning_rate": 5e-06,
      "loss": 0.0217,
      "step": 3590
    },
    {
      "epoch": 21.703089675960815,
      "grad_norm": 1.3870144386972025,
      "learning_rate": 5e-06,
      "loss": 0.0218,
      "step": 3600
    },
    {
      "epoch": 21.763376036171817,
      "grad_norm": 1.2996006226877992,
      "learning_rate": 5e-06,
      "loss": 0.0217,
      "step": 3610
    },
    {
      "epoch": 21.82366239638282,
      "grad_norm": 1.3533095992501387,
      "learning_rate": 5e-06,
      "loss": 0.0216,
      "step": 3620
    },
    {
      "epoch": 21.88394875659382,
      "grad_norm": 1.3112078284304194,
      "learning_rate": 5e-06,
      "loss": 0.0217,
      "step": 3630
    },
    {
      "epoch": 21.944235116804823,
      "grad_norm": 1.3288506152434172,
      "learning_rate": 5e-06,
      "loss": 0.0221,
      "step": 3640
    },
    {
      "epoch": 21.998492840994725,
      "eval_loss": 2.351545572280884,
      "eval_runtime": 116.2788,
      "eval_samples_per_second": 38.433,
      "eval_steps_per_second": 0.602,
      "step": 3649
    },
    {
      "epoch": 22.004521477015825,
      "grad_norm": 1.0649685278407737,
      "learning_rate": 5e-06,
      "loss": 0.0218,
      "step": 3650
    },
    {
      "epoch": 22.064807837226827,
      "grad_norm": 1.2396804319512318,
      "learning_rate": 5e-06,
      "loss": 0.0158,
      "step": 3660
    },
    {
      "epoch": 22.12509419743783,
      "grad_norm": 1.2203455145168154,
      "learning_rate": 5e-06,
      "loss": 0.0163,
      "step": 3670
    },
    {
      "epoch": 22.18538055764883,
      "grad_norm": 1.1547196338255674,
      "learning_rate": 5e-06,
      "loss": 0.0169,
      "step": 3680
    },
    {
      "epoch": 22.245666917859833,
      "grad_norm": 1.309239398813011,
      "learning_rate": 5e-06,
      "loss": 0.0173,
      "step": 3690
    },
    {
      "epoch": 22.305953278070838,
      "grad_norm": 1.2683908516455489,
      "learning_rate": 5e-06,
      "loss": 0.0173,
      "step": 3700
    },
    {
      "epoch": 22.36623963828184,
      "grad_norm": 1.5982780924133835,
      "learning_rate": 5e-06,
      "loss": 0.0183,
      "step": 3710
    },
    {
      "epoch": 22.426525998492842,
      "grad_norm": 1.4555126747204115,
      "learning_rate": 5e-06,
      "loss": 0.0184,
      "step": 3720
    },
    {
      "epoch": 22.486812358703844,
      "grad_norm": 1.3675517314579124,
      "learning_rate": 5e-06,
      "loss": 0.0186,
      "step": 3730
    },
    {
      "epoch": 22.547098718914846,
      "grad_norm": 1.2088813887554757,
      "learning_rate": 5e-06,
      "loss": 0.0188,
      "step": 3740
    },
    {
      "epoch": 22.607385079125848,
      "grad_norm": 1.157896013440299,
      "learning_rate": 5e-06,
      "loss": 0.0194,
      "step": 3750
    },
    {
      "epoch": 22.66767143933685,
      "grad_norm": 1.4152388376390785,
      "learning_rate": 5e-06,
      "loss": 0.0189,
      "step": 3760
    },
    {
      "epoch": 22.727957799547852,
      "grad_norm": 1.2554294042934306,
      "learning_rate": 5e-06,
      "loss": 0.019,
      "step": 3770
    },
    {
      "epoch": 22.788244159758854,
      "grad_norm": 1.3329931616963409,
      "learning_rate": 5e-06,
      "loss": 0.019,
      "step": 3780
    },
    {
      "epoch": 22.848530519969856,
      "grad_norm": 1.256565715756104,
      "learning_rate": 5e-06,
      "loss": 0.0189,
      "step": 3790
    },
    {
      "epoch": 22.908816880180858,
      "grad_norm": 1.368011051794652,
      "learning_rate": 5e-06,
      "loss": 0.0195,
      "step": 3800
    },
    {
      "epoch": 22.96910324039186,
      "grad_norm": 1.4311223887923525,
      "learning_rate": 5e-06,
      "loss": 0.0197,
      "step": 3810
    },
    {
      "epoch": 22.99924642049736,
      "eval_loss": 2.359902858734131,
      "eval_runtime": 117.0325,
      "eval_samples_per_second": 38.186,
      "eval_steps_per_second": 0.598,
      "step": 3815
    },
    {
      "epoch": 23.029389600602865,
      "grad_norm": 1.114496963570823,
      "learning_rate": 5e-06,
      "loss": 0.0171,
      "step": 3820
    },
    {
      "epoch": 23.089675960813867,
      "grad_norm": 1.1656429989984474,
      "learning_rate": 5e-06,
      "loss": 0.0149,
      "step": 3830
    },
    {
      "epoch": 23.14996232102487,
      "grad_norm": 1.3181375592191023,
      "learning_rate": 5e-06,
      "loss": 0.0152,
      "step": 3840
    },
    {
      "epoch": 23.21024868123587,
      "grad_norm": 1.214891422163979,
      "learning_rate": 5e-06,
      "loss": 0.0157,
      "step": 3850
    },
    {
      "epoch": 23.270535041446873,
      "grad_norm": 1.1329477739782727,
      "learning_rate": 5e-06,
      "loss": 0.0159,
      "step": 3860
    },
    {
      "epoch": 23.330821401657875,
      "grad_norm": 1.178978239439168,
      "learning_rate": 5e-06,
      "loss": 0.0163,
      "step": 3870
    },
    {
      "epoch": 23.391107761868877,
      "grad_norm": 1.1072042006237766,
      "learning_rate": 5e-06,
      "loss": 0.0165,
      "step": 3880
    },
    {
      "epoch": 23.45139412207988,
      "grad_norm": 1.2281267859101168,
      "learning_rate": 5e-06,
      "loss": 0.0168,
      "step": 3890
    },
    {
      "epoch": 23.51168048229088,
      "grad_norm": 1.1558899107856422,
      "learning_rate": 5e-06,
      "loss": 0.0174,
      "step": 3900
    },
    {
      "epoch": 23.571966842501883,
      "grad_norm": 1.217612194947995,
      "learning_rate": 5e-06,
      "loss": 0.0173,
      "step": 3910
    },
    {
      "epoch": 23.632253202712885,
      "grad_norm": 1.2440180440988633,
      "learning_rate": 5e-06,
      "loss": 0.0174,
      "step": 3920
    },
    {
      "epoch": 23.692539562923887,
      "grad_norm": 1.3420865819465875,
      "learning_rate": 5e-06,
      "loss": 0.0174,
      "step": 3930
    },
    {
      "epoch": 23.75282592313489,
      "grad_norm": 1.3593190067484822,
      "learning_rate": 5e-06,
      "loss": 0.018,
      "step": 3940
    },
    {
      "epoch": 23.813112283345895,
      "grad_norm": 1.2367975594859635,
      "learning_rate": 5e-06,
      "loss": 0.0177,
      "step": 3950
    },
    {
      "epoch": 23.873398643556897,
      "grad_norm": 1.2214254875430963,
      "learning_rate": 5e-06,
      "loss": 0.0179,
      "step": 3960
    },
    {
      "epoch": 23.9336850037679,
      "grad_norm": 1.20689456847013,
      "learning_rate": 5e-06,
      "loss": 0.018,
      "step": 3970
    },
    {
      "epoch": 23.9939713639789,
      "grad_norm": 1.2666923485205688,
      "learning_rate": 5e-06,
      "loss": 0.0186,
      "step": 3980
    },
    {
      "epoch": 24.0,
      "eval_loss": 2.3888232707977295,
      "eval_runtime": 116.343,
      "eval_samples_per_second": 38.412,
      "eval_steps_per_second": 0.602,
      "step": 3981
    },
    {
      "epoch": 24.054257724189902,
      "grad_norm": 1.1066074003233175,
      "learning_rate": 5e-06,
      "loss": 0.0141,
      "step": 3990
    },
    {
      "epoch": 24.114544084400904,
      "grad_norm": 1.0389652033545478,
      "learning_rate": 5e-06,
      "loss": 0.0141,
      "step": 4000
    },
    {
      "epoch": 24.174830444611906,
      "grad_norm": 1.1258379344398495,
      "learning_rate": 5e-06,
      "loss": 0.0144,
      "step": 4010
    },
    {
      "epoch": 24.23511680482291,
      "grad_norm": 1.1758249329629795,
      "learning_rate": 5e-06,
      "loss": 0.0148,
      "step": 4020
    },
    {
      "epoch": 24.29540316503391,
      "grad_norm": 1.1741885490501298,
      "learning_rate": 5e-06,
      "loss": 0.0151,
      "step": 4030
    },
    {
      "epoch": 24.355689525244912,
      "grad_norm": 1.1622125971160786,
      "learning_rate": 5e-06,
      "loss": 0.0156,
      "step": 4040
    },
    {
      "epoch": 24.415975885455914,
      "grad_norm": 1.0899820074898727,
      "learning_rate": 5e-06,
      "loss": 0.0154,
      "step": 4050
    },
    {
      "epoch": 24.476262245666916,
      "grad_norm": 1.2599823712922462,
      "learning_rate": 5e-06,
      "loss": 0.0159,
      "step": 4060
    },
    {
      "epoch": 24.53654860587792,
      "grad_norm": 1.140178001121795,
      "learning_rate": 5e-06,
      "loss": 0.0157,
      "step": 4070
    },
    {
      "epoch": 24.596834966088924,
      "grad_norm": 1.1230288891139422,
      "learning_rate": 5e-06,
      "loss": 0.0158,
      "step": 4080
    },
    {
      "epoch": 24.657121326299926,
      "grad_norm": 1.1218784125193524,
      "learning_rate": 5e-06,
      "loss": 0.0163,
      "step": 4090
    },
    {
      "epoch": 24.717407686510928,
      "grad_norm": 1.1524016261830858,
      "learning_rate": 5e-06,
      "loss": 0.0164,
      "step": 4100
    },
    {
      "epoch": 24.77769404672193,
      "grad_norm": 1.15414666614049,
      "learning_rate": 5e-06,
      "loss": 0.0166,
      "step": 4110
    },
    {
      "epoch": 24.83798040693293,
      "grad_norm": 1.3244431296269448,
      "learning_rate": 5e-06,
      "loss": 0.0169,
      "step": 4120
    },
    {
      "epoch": 24.868123587038433,
      "eval_loss": 2.4057514667510986,
      "eval_runtime": 113.5854,
      "eval_samples_per_second": 39.345,
      "eval_steps_per_second": 0.616,
      "step": 4125
    },
    {
      "epoch": 24.868123587038433,
      "step": 4125,
      "total_flos": 6909319257784320.0,
      "train_loss": 0.2661740447047985,
      "train_runtime": 139979.5045,
      "train_samples_per_second": 15.162,
      "train_steps_per_second": 0.029
    }
  ],
  "logging_steps": 10,
  "max_steps": 4125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 25,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6909319257784320.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}