{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 250.0,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "grad_norm": 13.149889945983887,
      "learning_rate": 1.8e-06,
      "loss": 1.4142,
      "step": 10
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.3455119132995605,
      "learning_rate": 3.8e-06,
      "loss": 1.2033,
      "step": 20
    },
    {
      "epoch": 0.75,
      "grad_norm": 3.1435389518737793,
      "learning_rate": 5.8e-06,
      "loss": 0.6036,
      "step": 30
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.23374342918396,
      "learning_rate": 7.8e-06,
      "loss": 0.3176,
      "step": 40
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.6511033773422241,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.2071,
      "step": 50
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.4293673038482666,
      "learning_rate": 1.18e-05,
      "loss": 0.1331,
      "step": 60
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.4380099773406982,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.09,
      "step": 70
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8167829513549805,
      "learning_rate": 1.58e-05,
      "loss": 0.0758,
      "step": 80
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.9906498789787292,
      "learning_rate": 1.78e-05,
      "loss": 0.0559,
      "step": 90
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.940280556678772,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 0.0599,
      "step": 100
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.5989202260971069,
      "learning_rate": 2.18e-05,
      "loss": 0.0496,
      "step": 110
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.060807228088379,
      "learning_rate": 2.38e-05,
      "loss": 0.0467,
      "step": 120
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.7101704478263855,
      "learning_rate": 2.58e-05,
      "loss": 0.0433,
      "step": 130
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.6841318011283875,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 0.0421,
      "step": 140
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.9395173788070679,
      "learning_rate": 2.98e-05,
      "loss": 0.0414,
      "step": 150
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6530765295028687,
      "learning_rate": 3.18e-05,
      "loss": 0.0364,
      "step": 160
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.5677581429481506,
      "learning_rate": 3.38e-05,
      "loss": 0.0316,
      "step": 170
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.6642166376113892,
      "learning_rate": 3.58e-05,
      "loss": 0.0374,
      "step": 180
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.5084357857704163,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 0.033,
      "step": 190
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.40896496176719666,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 0.036,
      "step": 200
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.8260584473609924,
      "learning_rate": 4.18e-05,
      "loss": 0.0413,
      "step": 210
    },
    {
      "epoch": 5.5,
      "grad_norm": 1.0218405723571777,
      "learning_rate": 4.38e-05,
      "loss": 0.0395,
      "step": 220
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.7389956712722778,
      "learning_rate": 4.58e-05,
      "loss": 0.0303,
      "step": 230
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.8274843096733093,
      "learning_rate": 4.78e-05,
      "loss": 0.0384,
      "step": 240
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.9301137328147888,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 0.0361,
      "step": 250
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.8189093470573425,
      "learning_rate": 5.1800000000000005e-05,
      "loss": 0.0362,
      "step": 260
    },
    {
      "epoch": 6.75,
      "grad_norm": 1.1223821640014648,
      "learning_rate": 5.380000000000001e-05,
      "loss": 0.037,
      "step": 270
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.6035957336425781,
      "learning_rate": 5.580000000000001e-05,
      "loss": 0.0298,
      "step": 280
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.35879412293434143,
      "learning_rate": 5.7799999999999995e-05,
      "loss": 0.038,
      "step": 290
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.6832134127616882,
      "learning_rate": 5.9800000000000003e-05,
      "loss": 0.0342,
      "step": 300
    },
    {
      "epoch": 7.75,
      "grad_norm": 1.0018675327301025,
      "learning_rate": 6.18e-05,
      "loss": 0.0449,
      "step": 310
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.648226261138916,
      "learning_rate": 6.38e-05,
      "loss": 0.033,
      "step": 320
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.8478636145591736,
      "learning_rate": 6.58e-05,
      "loss": 0.0326,
      "step": 330
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.6271067261695862,
      "learning_rate": 6.780000000000001e-05,
      "loss": 0.0361,
      "step": 340
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.8157103061676025,
      "learning_rate": 6.98e-05,
      "loss": 0.0332,
      "step": 350
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.5046970248222351,
      "learning_rate": 7.18e-05,
      "loss": 0.0333,
      "step": 360
    },
    {
      "epoch": 9.25,
      "grad_norm": 0.4200487434864044,
      "learning_rate": 7.38e-05,
      "loss": 0.0378,
      "step": 370
    },
    {
      "epoch": 9.5,
      "grad_norm": 0.4756053686141968,
      "learning_rate": 7.58e-05,
      "loss": 0.0353,
      "step": 380
    },
    {
      "epoch": 9.75,
      "grad_norm": 0.5499545335769653,
      "learning_rate": 7.780000000000001e-05,
      "loss": 0.0388,
      "step": 390
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.5970879197120667,
      "learning_rate": 7.98e-05,
      "loss": 0.0292,
      "step": 400
    },
    {
      "epoch": 10.25,
      "grad_norm": 0.4654370844364166,
      "learning_rate": 8.18e-05,
      "loss": 0.0346,
      "step": 410
    },
    {
      "epoch": 10.5,
      "grad_norm": 0.4801628589630127,
      "learning_rate": 8.38e-05,
      "loss": 0.034,
      "step": 420
    },
    {
      "epoch": 10.75,
      "grad_norm": 0.5524302124977112,
      "learning_rate": 8.58e-05,
      "loss": 0.034,
      "step": 430
    },
    {
      "epoch": 11.0,
      "grad_norm": 0.6977431774139404,
      "learning_rate": 8.78e-05,
      "loss": 0.0303,
      "step": 440
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.9158258438110352,
      "learning_rate": 8.98e-05,
      "loss": 0.0313,
      "step": 450
    },
    {
      "epoch": 11.5,
      "grad_norm": 0.9180648326873779,
      "learning_rate": 9.180000000000001e-05,
      "loss": 0.0361,
      "step": 460
    },
    {
      "epoch": 11.75,
      "grad_norm": 0.5505082011222839,
      "learning_rate": 9.38e-05,
      "loss": 0.0375,
      "step": 470
    },
    {
      "epoch": 12.0,
      "grad_norm": 0.5206027626991272,
      "learning_rate": 9.58e-05,
      "loss": 0.0361,
      "step": 480
    },
    {
      "epoch": 12.25,
      "grad_norm": 0.6917493939399719,
      "learning_rate": 9.78e-05,
      "loss": 0.0282,
      "step": 490
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.9478809237480164,
      "learning_rate": 9.98e-05,
      "loss": 0.0337,
      "step": 500
    },
    {
      "epoch": 12.75,
      "grad_norm": 0.5493677258491516,
      "learning_rate": 9.9999778549206e-05,
      "loss": 0.0317,
      "step": 510
    },
    {
      "epoch": 13.0,
      "grad_norm": 0.561867892742157,
      "learning_rate": 9.999901304280685e-05,
      "loss": 0.0346,
      "step": 520
    },
    {
      "epoch": 13.25,
      "grad_norm": 0.3665112853050232,
      "learning_rate": 9.999770075521164e-05,
      "loss": 0.0376,
      "step": 530
    },
    {
      "epoch": 13.5,
      "grad_norm": 0.5569449663162231,
      "learning_rate": 9.99958417007713e-05,
      "loss": 0.0327,
      "step": 540
    },
    {
      "epoch": 13.75,
      "grad_norm": 0.5099448561668396,
      "learning_rate": 9.999343589981615e-05,
      "loss": 0.0303,
      "step": 550
    },
    {
      "epoch": 14.0,
      "grad_norm": 0.6323690414428711,
      "learning_rate": 9.999048337865568e-05,
      "loss": 0.0262,
      "step": 560
    },
    {
      "epoch": 14.25,
      "grad_norm": 0.28097397089004517,
      "learning_rate": 9.998698416957815e-05,
      "loss": 0.0264,
      "step": 570
    },
    {
      "epoch": 14.5,
      "grad_norm": 0.8999001383781433,
      "learning_rate": 9.998293831085037e-05,
      "loss": 0.0264,
      "step": 580
    },
    {
      "epoch": 14.75,
      "grad_norm": 0.4309835731983185,
      "learning_rate": 9.997834584671719e-05,
      "loss": 0.0285,
      "step": 590
    },
    {
      "epoch": 15.0,
      "grad_norm": 0.4197908639907837,
      "learning_rate": 9.997320682740107e-05,
      "loss": 0.029,
      "step": 600
    },
    {
      "epoch": 15.25,
      "grad_norm": 0.4405575394630432,
      "learning_rate": 9.996752130910149e-05,
      "loss": 0.032,
      "step": 610
    },
    {
      "epoch": 15.5,
      "grad_norm": 0.3575686812400818,
      "learning_rate": 9.99612893539944e-05,
      "loss": 0.0232,
      "step": 620
    },
    {
      "epoch": 15.75,
      "grad_norm": 0.615228533744812,
      "learning_rate": 9.995451103023144e-05,
      "loss": 0.0233,
      "step": 630
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.45813262462615967,
      "learning_rate": 9.994718641193928e-05,
      "loss": 0.025,
      "step": 640
    },
    {
      "epoch": 16.25,
      "grad_norm": 0.4395560026168823,
      "learning_rate": 9.993931557921874e-05,
      "loss": 0.0264,
      "step": 650
    },
    {
      "epoch": 16.5,
      "grad_norm": 0.5942047238349915,
      "learning_rate": 9.993089861814402e-05,
      "loss": 0.0268,
      "step": 660
    },
    {
      "epoch": 16.75,
      "grad_norm": 0.6569145917892456,
      "learning_rate": 9.992193562076166e-05,
      "loss": 0.033,
      "step": 670
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.6226444840431213,
      "learning_rate": 9.991242668508954e-05,
      "loss": 0.0266,
      "step": 680
    },
    {
      "epoch": 17.25,
      "grad_norm": 0.47699910402297974,
      "learning_rate": 9.990237191511587e-05,
      "loss": 0.0297,
      "step": 690
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.6073171496391296,
      "learning_rate": 9.989177142079802e-05,
      "loss": 0.0296,
      "step": 700
    },
    {
      "epoch": 17.75,
      "grad_norm": 0.35824280977249146,
      "learning_rate": 9.988062531806126e-05,
      "loss": 0.0327,
      "step": 710
    },
    {
      "epoch": 18.0,
      "grad_norm": 0.5538836121559143,
      "learning_rate": 9.986893372879762e-05,
      "loss": 0.0292,
      "step": 720
    },
    {
      "epoch": 18.25,
      "grad_norm": 0.4948605000972748,
      "learning_rate": 9.985669678086443e-05,
      "loss": 0.0292,
      "step": 730
    },
    {
      "epoch": 18.5,
      "grad_norm": 0.38511842489242554,
      "learning_rate": 9.984391460808298e-05,
      "loss": 0.0221,
      "step": 740
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.4333004653453827,
      "learning_rate": 9.983058735023709e-05,
      "loss": 0.0226,
      "step": 750
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.6658897399902344,
      "learning_rate": 9.98167151530715e-05,
      "loss": 0.0205,
      "step": 760
    },
    {
      "epoch": 19.25,
      "grad_norm": 0.5551711916923523,
      "learning_rate": 9.980229816829034e-05,
      "loss": 0.0218,
      "step": 770
    },
    {
      "epoch": 19.5,
      "grad_norm": 0.39259231090545654,
      "learning_rate": 9.978733655355544e-05,
      "loss": 0.0258,
      "step": 780
    },
    {
      "epoch": 19.75,
      "grad_norm": 0.45398399233818054,
      "learning_rate": 9.977183047248464e-05,
      "loss": 0.0196,
      "step": 790
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.37839898467063904,
      "learning_rate": 9.975578009464992e-05,
      "loss": 0.0243,
      "step": 800
    },
    {
      "epoch": 20.25,
      "grad_norm": 0.8339304327964783,
      "learning_rate": 9.97391855955757e-05,
      "loss": 0.0263,
      "step": 810
    },
    {
      "epoch": 20.5,
      "grad_norm": 0.6552790403366089,
      "learning_rate": 9.972204715673669e-05,
      "loss": 0.0296,
      "step": 820
    },
    {
      "epoch": 20.75,
      "grad_norm": 0.36044272780418396,
      "learning_rate": 9.970436496555617e-05,
      "loss": 0.0216,
      "step": 830
    },
    {
      "epoch": 21.0,
      "grad_norm": 0.45362478494644165,
      "learning_rate": 9.968613921540373e-05,
      "loss": 0.024,
      "step": 840
    },
    {
      "epoch": 21.25,
      "grad_norm": 0.37655627727508545,
      "learning_rate": 9.966737010559326e-05,
      "loss": 0.0265,
      "step": 850
    },
    {
      "epoch": 21.5,
      "grad_norm": 0.6144092082977295,
      "learning_rate": 9.964805784138072e-05,
      "loss": 0.0231,
      "step": 860
    },
    {
      "epoch": 21.75,
      "grad_norm": 0.463652640581131,
      "learning_rate": 9.962820263396195e-05,
      "loss": 0.0218,
      "step": 870
    },
    {
      "epoch": 22.0,
      "grad_norm": 0.27026480436325073,
      "learning_rate": 9.960780470047033e-05,
      "loss": 0.0196,
      "step": 880
    },
    {
      "epoch": 22.25,
      "grad_norm": 0.41693419218063354,
      "learning_rate": 9.958686426397437e-05,
      "loss": 0.0169,
      "step": 890
    },
    {
      "epoch": 22.5,
      "grad_norm": 0.4871846139431,
      "learning_rate": 9.956538155347534e-05,
      "loss": 0.0226,
      "step": 900
    },
    {
      "epoch": 22.75,
      "grad_norm": 0.4026087522506714,
      "learning_rate": 9.95433568039047e-05,
      "loss": 0.0161,
      "step": 910
    },
    {
      "epoch": 23.0,
      "grad_norm": 0.26234668493270874,
      "learning_rate": 9.952079025612162e-05,
      "loss": 0.0189,
      "step": 920
    },
    {
      "epoch": 23.25,
      "grad_norm": 0.38857871294021606,
      "learning_rate": 9.949768215691022e-05,
      "loss": 0.0199,
      "step": 930
    },
    {
      "epoch": 23.5,
      "grad_norm": 0.3500604033470154,
      "learning_rate": 9.9474032758977e-05,
      "loss": 0.016,
      "step": 940
    },
    {
      "epoch": 23.75,
      "grad_norm": 0.3929100036621094,
      "learning_rate": 9.944984232094794e-05,
      "loss": 0.0181,
      "step": 950
    },
    {
      "epoch": 24.0,
      "grad_norm": 0.2801556885242462,
      "learning_rate": 9.942511110736584e-05,
      "loss": 0.0187,
      "step": 960
    },
    {
      "epoch": 24.25,
      "grad_norm": 0.3869239091873169,
      "learning_rate": 9.939983938868726e-05,
      "loss": 0.017,
      "step": 970
    },
    {
      "epoch": 24.5,
      "grad_norm": 0.3851812481880188,
      "learning_rate": 9.93740274412797e-05,
      "loss": 0.0192,
      "step": 980
    },
    {
      "epoch": 24.75,
      "grad_norm": 0.623075544834137,
      "learning_rate": 9.934767554741846e-05,
      "loss": 0.0246,
      "step": 990
    },
    {
      "epoch": 25.0,
      "grad_norm": 0.35598379373550415,
      "learning_rate": 9.932078399528361e-05,
      "loss": 0.0206,
      "step": 1000
    },
    {
      "epoch": 25.25,
      "grad_norm": 0.6533331871032715,
      "learning_rate": 9.929335307895689e-05,
      "loss": 0.0209,
      "step": 1010
    },
    {
      "epoch": 25.5,
      "grad_norm": 0.32766997814178467,
      "learning_rate": 9.926538309841839e-05,
      "loss": 0.0165,
      "step": 1020
    },
    {
      "epoch": 25.75,
      "grad_norm": 0.4293682277202606,
      "learning_rate": 9.923687435954334e-05,
      "loss": 0.0196,
      "step": 1030
    },
    {
      "epoch": 26.0,
      "grad_norm": 0.4344434142112732,
      "learning_rate": 9.920782717409873e-05,
      "loss": 0.0229,
      "step": 1040
    },
    {
      "epoch": 26.25,
      "grad_norm": 0.4949270188808441,
      "learning_rate": 9.917824185973994e-05,
      "loss": 0.0186,
      "step": 1050
    },
    {
      "epoch": 26.5,
      "grad_norm": 0.6668685674667358,
      "learning_rate": 9.914811874000723e-05,
      "loss": 0.0245,
      "step": 1060
    },
    {
      "epoch": 26.75,
      "grad_norm": 0.7803040742874146,
      "learning_rate": 9.911745814432218e-05,
      "loss": 0.0243,
      "step": 1070
    },
    {
      "epoch": 27.0,
      "grad_norm": 0.3841216564178467,
      "learning_rate": 9.90862604079842e-05,
      "loss": 0.0207,
      "step": 1080
    },
    {
      "epoch": 27.25,
      "grad_norm": 0.5297211408615112,
      "learning_rate": 9.90545258721667e-05,
      "loss": 0.0246,
      "step": 1090
    },
    {
      "epoch": 27.5,
      "grad_norm": 0.23917119204998016,
      "learning_rate": 9.90222548839135e-05,
      "loss": 0.0182,
      "step": 1100
    },
    {
      "epoch": 27.75,
      "grad_norm": 0.3145104646682739,
      "learning_rate": 9.898944779613495e-05,
      "loss": 0.0171,
      "step": 1110
    },
    {
      "epoch": 28.0,
      "grad_norm": 0.4466751217842102,
      "learning_rate": 9.89561049676041e-05,
      "loss": 0.0183,
      "step": 1120
    },
    {
      "epoch": 28.25,
      "grad_norm": 0.6861326098442078,
      "learning_rate": 9.89222267629528e-05,
      "loss": 0.0221,
      "step": 1130
    },
    {
      "epoch": 28.5,
      "grad_norm": 0.4263695478439331,
      "learning_rate": 9.888781355266763e-05,
      "loss": 0.0233,
      "step": 1140
    },
    {
      "epoch": 28.75,
      "grad_norm": 0.6137635111808777,
      "learning_rate": 9.885286571308598e-05,
      "loss": 0.0224,
      "step": 1150
    },
    {
      "epoch": 29.0,
      "grad_norm": 0.527621865272522,
      "learning_rate": 9.881738362639182e-05,
      "loss": 0.0177,
      "step": 1160
    },
    {
      "epoch": 29.25,
      "grad_norm": 0.39209073781967163,
      "learning_rate": 9.878136768061154e-05,
      "loss": 0.0198,
      "step": 1170
    },
    {
      "epoch": 29.5,
      "grad_norm": 0.6302863359451294,
      "learning_rate": 9.874481826960979e-05,
      "loss": 0.0226,
      "step": 1180
    },
    {
      "epoch": 29.75,
      "grad_norm": 0.34944644570350647,
      "learning_rate": 9.870773579308503e-05,
      "loss": 0.0159,
      "step": 1190
    },
    {
      "epoch": 30.0,
      "grad_norm": 0.4721897840499878,
      "learning_rate": 9.867012065656533e-05,
      "loss": 0.0206,
      "step": 1200
    },
    {
      "epoch": 30.25,
      "grad_norm": 0.3942127227783203,
      "learning_rate": 9.863197327140376e-05,
      "loss": 0.0192,
      "step": 1210
    },
    {
      "epoch": 30.5,
      "grad_norm": 0.2768412232398987,
      "learning_rate": 9.859329405477403e-05,
      "loss": 0.0138,
      "step": 1220
    },
    {
      "epoch": 30.75,
      "grad_norm": 0.33834919333457947,
      "learning_rate": 9.855408342966585e-05,
      "loss": 0.018,
      "step": 1230
    },
    {
      "epoch": 31.0,
      "grad_norm": 0.2718036472797394,
      "learning_rate": 9.851434182488033e-05,
      "loss": 0.0176,
      "step": 1240
    },
    {
      "epoch": 31.25,
      "grad_norm": 0.3397132158279419,
      "learning_rate": 9.84740696750253e-05,
      "loss": 0.0196,
      "step": 1250
    },
    {
      "epoch": 31.5,
      "grad_norm": 0.3387533128261566,
      "learning_rate": 9.843326742051055e-05,
      "loss": 0.0207,
      "step": 1260
    },
    {
      "epoch": 31.75,
      "grad_norm": 0.4069482088088989,
      "learning_rate": 9.839193550754297e-05,
      "loss": 0.0205,
      "step": 1270
    },
    {
      "epoch": 32.0,
      "grad_norm": 0.3819906413555145,
      "learning_rate": 9.835007438812177e-05,
      "loss": 0.0244,
      "step": 1280
    },
    {
      "epoch": 32.25,
      "grad_norm": 0.5072693824768066,
      "learning_rate": 9.830768452003341e-05,
      "loss": 0.0255,
      "step": 1290
    },
    {
      "epoch": 32.5,
      "grad_norm": 0.5479236841201782,
      "learning_rate": 9.826476636684671e-05,
      "loss": 0.0255,
      "step": 1300
    },
    {
      "epoch": 32.75,
      "grad_norm": 0.6064186692237854,
      "learning_rate": 9.822132039790773e-05,
      "loss": 0.0166,
      "step": 1310
    },
    {
      "epoch": 33.0,
      "grad_norm": 0.5842517614364624,
      "learning_rate": 9.817734708833461e-05,
      "loss": 0.0209,
      "step": 1320
    },
    {
      "epoch": 33.25,
      "grad_norm": 0.26301082968711853,
      "learning_rate": 9.813284691901243e-05,
      "loss": 0.0174,
      "step": 1330
    },
    {
      "epoch": 33.5,
      "grad_norm": 0.4220535159111023,
      "learning_rate": 9.808782037658792e-05,
      "loss": 0.0201,
      "step": 1340
    },
    {
      "epoch": 33.75,
      "grad_norm": 0.42425140738487244,
      "learning_rate": 9.804226795346411e-05,
      "loss": 0.0146,
      "step": 1350
    },
    {
      "epoch": 34.0,
      "grad_norm": 0.27794766426086426,
      "learning_rate": 9.799619014779503e-05,
      "loss": 0.0163,
      "step": 1360
    },
    {
      "epoch": 34.25,
      "grad_norm": 0.4631710350513458,
      "learning_rate": 9.794958746348013e-05,
      "loss": 0.0175,
      "step": 1370
    },
    {
      "epoch": 34.5,
      "grad_norm": 0.4688563644886017,
      "learning_rate": 9.790246041015896e-05,
      "loss": 0.0196,
      "step": 1380
    },
    {
      "epoch": 34.75,
      "grad_norm": 0.5662594437599182,
      "learning_rate": 9.785480950320538e-05,
      "loss": 0.017,
      "step": 1390
    },
    {
      "epoch": 35.0,
      "grad_norm": 0.35163137316703796,
      "learning_rate": 9.78066352637221e-05,
      "loss": 0.0183,
      "step": 1400
    },
    {
      "epoch": 35.25,
      "grad_norm": 0.3633674681186676,
      "learning_rate": 9.775793821853488e-05,
      "loss": 0.0159,
      "step": 1410
    },
    {
      "epoch": 35.5,
      "grad_norm": 0.31508517265319824,
      "learning_rate": 9.77087189001868e-05,
      "loss": 0.0176,
      "step": 1420
    },
    {
      "epoch": 35.75,
      "grad_norm": 0.37547263503074646,
      "learning_rate": 9.765897784693243e-05,
      "loss": 0.0175,
      "step": 1430
    },
    {
      "epoch": 36.0,
      "grad_norm": 0.3714698255062103,
      "learning_rate": 9.760871560273197e-05,
      "loss": 0.0209,
      "step": 1440
    },
    {
      "epoch": 36.25,
      "grad_norm": 0.6465221047401428,
      "learning_rate": 9.755793271724526e-05,
      "loss": 0.0204,
      "step": 1450
    },
    {
      "epoch": 36.5,
      "grad_norm": 0.6143114566802979,
      "learning_rate": 9.750662974582584e-05,
      "loss": 0.0196,
      "step": 1460
    },
    {
      "epoch": 36.75,
      "grad_norm": 0.48664823174476624,
      "learning_rate": 9.745480724951473e-05,
      "loss": 0.0173,
      "step": 1470
    },
    {
      "epoch": 37.0,
      "grad_norm": 0.47144779562950134,
      "learning_rate": 9.740246579503447e-05,
      "loss": 0.0243,
      "step": 1480
    },
    {
      "epoch": 37.25,
      "grad_norm": 0.35993170738220215,
      "learning_rate": 9.734960595478284e-05,
      "loss": 0.0152,
      "step": 1490
    },
    {
      "epoch": 37.5,
      "grad_norm": 0.3285447657108307,
      "learning_rate": 9.729622830682657e-05,
      "loss": 0.0189,
      "step": 1500
    },
    {
      "epoch": 37.75,
      "grad_norm": 0.4205745458602905,
      "learning_rate": 9.724233343489504e-05,
      "loss": 0.0222,
      "step": 1510
    },
    {
      "epoch": 38.0,
      "grad_norm": 0.2668244242668152,
      "learning_rate": 9.718792192837396e-05,
      "loss": 0.0163,
      "step": 1520
    },
    {
      "epoch": 38.25,
      "grad_norm": 0.4396367371082306,
      "learning_rate": 9.713299438229886e-05,
      "loss": 0.0193,
      "step": 1530
    },
    {
      "epoch": 38.5,
      "grad_norm": 0.40211695432662964,
      "learning_rate": 9.707755139734855e-05,
      "loss": 0.0164,
      "step": 1540
    },
    {
      "epoch": 38.75,
      "grad_norm": 0.4458588659763336,
      "learning_rate": 9.702159357983866e-05,
      "loss": 0.0208,
      "step": 1550
    },
    {
      "epoch": 39.0,
      "grad_norm": 0.22913917899131775,
      "learning_rate": 9.696512154171492e-05,
      "loss": 0.0163,
      "step": 1560
    },
    {
      "epoch": 39.25,
      "grad_norm": 0.4234263002872467,
      "learning_rate": 9.690813590054645e-05,
      "loss": 0.0158,
      "step": 1570
    },
    {
      "epoch": 39.5,
      "grad_norm": 0.41260766983032227,
      "learning_rate": 9.685063727951914e-05,
      "loss": 0.0183,
      "step": 1580
    },
    {
      "epoch": 39.75,
      "grad_norm": 0.4093533754348755,
      "learning_rate": 9.679262630742865e-05,
      "loss": 0.0215,
      "step": 1590
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.5214777588844299,
      "learning_rate": 9.673410361867373e-05,
      "loss": 0.0183,
      "step": 1600
    },
    {
      "epoch": 40.25,
      "grad_norm": 0.3588806986808777,
      "learning_rate": 9.667506985324909e-05,
      "loss": 0.0181,
      "step": 1610
    },
    {
      "epoch": 40.5,
      "grad_norm": 0.3200751543045044,
      "learning_rate": 9.661552565673855e-05,
      "loss": 0.0147,
      "step": 1620
    },
    {
      "epoch": 40.75,
      "grad_norm": 0.2634463608264923,
      "learning_rate": 9.655547168030789e-05,
      "loss": 0.0139,
      "step": 1630
    },
    {
      "epoch": 41.0,
      "grad_norm": 0.362644225358963,
      "learning_rate": 9.649490858069777e-05,
      "loss": 0.0142,
      "step": 1640
    },
    {
      "epoch": 41.25,
      "grad_norm": 0.32056477665901184,
      "learning_rate": 9.643383702021658e-05,
      "loss": 0.0156,
      "step": 1650
    },
    {
      "epoch": 41.5,
      "grad_norm": 0.4088948369026184,
      "learning_rate": 9.637225766673307e-05,
      "loss": 0.0188,
      "step": 1660
    },
    {
      "epoch": 41.75,
      "grad_norm": 0.37446433305740356,
      "learning_rate": 9.631017119366922e-05,
      "loss": 0.015,
      "step": 1670
    },
    {
      "epoch": 42.0,
      "grad_norm": 0.4627027213573456,
      "learning_rate": 9.624757827999273e-05,
      "loss": 0.0163,
      "step": 1680
    },
    {
      "epoch": 42.25,
      "grad_norm": 0.22955283522605896,
      "learning_rate": 9.618447961020971e-05,
      "loss": 0.019,
      "step": 1690
    },
    {
      "epoch": 42.5,
      "grad_norm": 0.27593836188316345,
      "learning_rate": 9.612087587435707e-05,
      "loss": 0.0137,
      "step": 1700
    },
    {
      "epoch": 42.75,
      "grad_norm": 0.3275546431541443,
      "learning_rate": 9.605676776799508e-05,
      "loss": 0.0152,
      "step": 1710
    },
    {
      "epoch": 43.0,
      "grad_norm": 0.38114628195762634,
      "learning_rate": 9.599215599219973e-05,
      "loss": 0.0181,
      "step": 1720
    },
    {
      "epoch": 43.25,
      "grad_norm": 0.3989734649658203,
      "learning_rate": 9.592704125355505e-05,
      "loss": 0.019,
      "step": 1730
    },
    {
      "epoch": 43.5,
      "grad_norm": 0.31094446778297424,
      "learning_rate": 9.586142426414538e-05,
      "loss": 0.0179,
      "step": 1740
    },
    {
      "epoch": 43.75,
      "grad_norm": 0.17701998353004456,
      "learning_rate": 9.57953057415476e-05,
      "loss": 0.0199,
      "step": 1750
    },
    {
      "epoch": 44.0,
      "grad_norm": 0.29390594363212585,
      "learning_rate": 9.572868640882328e-05,
      "loss": 0.0163,
      "step": 1760
    },
    {
      "epoch": 44.25,
      "grad_norm": 0.3040190041065216,
      "learning_rate": 9.56615669945108e-05,
      "loss": 0.0155,
      "step": 1770
    },
    {
      "epoch": 44.5,
      "grad_norm": 0.44003409147262573,
      "learning_rate": 9.55939482326173e-05,
      "loss": 0.0185,
      "step": 1780
    },
    {
      "epoch": 44.75,
      "grad_norm": 0.3427186608314514,
      "learning_rate": 9.552583086261069e-05,
      "loss": 0.0191,
      "step": 1790
    },
    {
      "epoch": 45.0,
      "grad_norm": 0.3714499771595001,
      "learning_rate": 9.545721562941168e-05,
      "loss": 0.0175,
      "step": 1800
    },
    {
      "epoch": 45.25,
      "grad_norm": 0.30852749943733215,
      "learning_rate": 9.538810328338543e-05,
      "loss": 0.0164,
      "step": 1810
    },
    {
      "epoch": 45.5,
      "grad_norm": 0.2884804904460907,
      "learning_rate": 9.531849458033349e-05,
      "loss": 0.0161,
      "step": 1820
    },
    {
      "epoch": 45.75,
      "grad_norm": 0.3582738935947418,
      "learning_rate": 9.524839028148547e-05,
      "loss": 0.0146,
      "step": 1830
    },
    {
      "epoch": 46.0,
      "grad_norm": 0.32050254940986633,
      "learning_rate": 9.517779115349077e-05,
      "loss": 0.0136,
      "step": 1840
    },
    {
      "epoch": 46.25,
      "grad_norm": 0.40590667724609375,
      "learning_rate": 9.510669796841014e-05,
      "loss": 0.0165,
      "step": 1850
    },
    {
      "epoch": 46.5,
      "grad_norm": 0.2890171408653259,
      "learning_rate": 9.503511150370727e-05,
      "loss": 0.016,
      "step": 1860
    },
    {
      "epoch": 46.75,
      "grad_norm": 0.29964378476142883,
      "learning_rate": 9.496303254224024e-05,
      "loss": 0.0152,
      "step": 1870
    },
    {
      "epoch": 47.0,
      "grad_norm": 0.27994900941848755,
      "learning_rate": 9.489046187225306e-05,
      "loss": 0.0186,
      "step": 1880
    },
    {
      "epoch": 47.25,
      "grad_norm": 0.31143495440483093,
      "learning_rate": 9.481740028736692e-05,
      "loss": 0.0133,
      "step": 1890
    },
    {
      "epoch": 47.5,
      "grad_norm": 0.5050742626190186,
      "learning_rate": 9.474384858657164e-05,
      "loss": 0.019,
      "step": 1900
    },
    {
      "epoch": 47.75,
      "grad_norm": 0.45632266998291016,
      "learning_rate": 9.466980757421679e-05,
      "loss": 0.0203,
      "step": 1910
    },
    {
      "epoch": 48.0,
      "grad_norm": 0.362568199634552,
      "learning_rate": 9.459527806000305e-05,
      "loss": 0.0152,
      "step": 1920
    },
    {
      "epoch": 48.25,
      "grad_norm": 0.2526082694530487,
      "learning_rate": 9.452026085897325e-05,
      "loss": 0.0154,
      "step": 1930
    },
    {
      "epoch": 48.5,
      "grad_norm": 0.29366403818130493,
      "learning_rate": 9.444475679150348e-05,
      "loss": 0.0184,
      "step": 1940
    },
    {
      "epoch": 48.75,
      "grad_norm": 0.311564177274704,
      "learning_rate": 9.436876668329411e-05,
      "loss": 0.0155,
      "step": 1950
    },
    {
      "epoch": 49.0,
      "grad_norm": 0.42421355843544006,
      "learning_rate": 9.429229136536079e-05,
      "loss": 0.0126,
      "step": 1960
    },
    {
      "epoch": 49.25,
      "grad_norm": 0.608485221862793,
      "learning_rate": 9.421533167402534e-05,
      "loss": 0.0198,
      "step": 1970
    },
    {
      "epoch": 49.5,
      "grad_norm": 0.5886660814285278,
      "learning_rate": 9.413788845090666e-05,
      "loss": 0.0234,
      "step": 1980
    },
    {
      "epoch": 49.75,
      "grad_norm": 0.5250411033630371,
      "learning_rate": 9.405996254291136e-05,
      "loss": 0.0218,
      "step": 1990
    },
    {
      "epoch": 50.0,
      "grad_norm": 0.2709248661994934,
      "learning_rate": 9.398155480222474e-05,
      "loss": 0.0161,
      "step": 2000
    },
    {
      "epoch": 50.25,
      "grad_norm": 0.1994844675064087,
      "learning_rate": 9.390266608630128e-05,
      "loss": 0.0136,
      "step": 2010
    },
    {
      "epoch": 50.5,
      "grad_norm": 0.17653261125087738,
      "learning_rate": 9.38232972578553e-05,
      "loss": 0.0163,
      "step": 2020
    },
    {
      "epoch": 50.75,
      "grad_norm": 0.22169753909111023,
      "learning_rate": 9.374344918485164e-05,
      "loss": 0.0199,
      "step": 2030
    },
    {
      "epoch": 51.0,
      "grad_norm": 0.28427475690841675,
      "learning_rate": 9.366312274049602e-05,
      "loss": 0.018,
      "step": 2040
    },
    {
      "epoch": 51.25,
      "grad_norm": 0.20987115800380707,
      "learning_rate": 9.358231880322554e-05,
      "loss": 0.0135,
      "step": 2050
    },
    {
      "epoch": 51.5,
      "grad_norm": 0.18959645926952362,
      "learning_rate": 9.350103825669916e-05,
      "loss": 0.0138,
      "step": 2060
    },
    {
      "epoch": 51.75,
      "grad_norm": 0.3204769790172577,
      "learning_rate": 9.341928198978787e-05,
      "loss": 0.0135,
      "step": 2070
    },
    {
      "epoch": 52.0,
      "grad_norm": 0.4980137348175049,
      "learning_rate": 9.333705089656512e-05,
      "loss": 0.0188,
      "step": 2080
    },
    {
      "epoch": 52.25,
      "grad_norm": 0.4129096567630768,
      "learning_rate": 9.325434587629698e-05,
      "loss": 0.012,
      "step": 2090
    },
    {
      "epoch": 52.5,
      "grad_norm": 0.4252679944038391,
      "learning_rate": 9.31711678334323e-05,
      "loss": 0.0151,
      "step": 2100
    },
    {
      "epoch": 52.75,
      "grad_norm": 0.23144613206386566,
      "learning_rate": 9.308751767759282e-05,
      "loss": 0.014,
      "step": 2110
    },
    {
      "epoch": 53.0,
      "grad_norm": 0.3891558349132538,
      "learning_rate": 9.300339632356325e-05,
      "loss": 0.0148,
      "step": 2120
    },
    {
      "epoch": 53.25,
      "grad_norm": 0.49045929312705994,
      "learning_rate": 9.291880469128124e-05,
      "loss": 0.0152,
      "step": 2130
    },
    {
      "epoch": 53.5,
      "grad_norm": 0.40814849734306335,
      "learning_rate": 9.283374370582732e-05,
      "loss": 0.0114,
      "step": 2140
    },
    {
      "epoch": 53.75,
      "grad_norm": 0.43267515301704407,
      "learning_rate": 9.274821429741482e-05,
      "loss": 0.0242,
      "step": 2150
    },
    {
      "epoch": 54.0,
      "grad_norm": 0.2985181510448456,
      "learning_rate": 9.266221740137961e-05,
      "loss": 0.0134,
      "step": 2160
    },
    {
      "epoch": 54.25,
      "grad_norm": 0.3468068540096283,
      "learning_rate": 9.257575395817001e-05,
      "loss": 0.0142,
      "step": 2170
    },
    {
      "epoch": 54.5,
      "grad_norm": 0.22969213128089905,
      "learning_rate": 9.248882491333637e-05,
      "loss": 0.0092,
      "step": 2180
    },
    {
      "epoch": 54.75,
      "grad_norm": 0.22361359000205994,
      "learning_rate": 9.240143121752076e-05,
      "loss": 0.0137,
      "step": 2190
    },
    {
      "epoch": 55.0,
      "grad_norm": 0.24732734262943268,
      "learning_rate": 9.23135738264467e-05,
      "loss": 0.0141,
      "step": 2200
    },
    {
      "epoch": 55.25,
      "grad_norm": 0.20617972314357758,
      "learning_rate": 9.222525370090849e-05,
      "loss": 0.0093,
      "step": 2210
    },
    {
      "epoch": 55.5,
      "grad_norm": 0.2772747576236725,
      "learning_rate": 9.213647180676088e-05,
      "loss": 0.0112,
      "step": 2220
    },
    {
      "epoch": 55.75,
      "grad_norm": 0.21540410816669464,
      "learning_rate": 9.204722911490846e-05,
      "loss": 0.0134,
      "step": 2230
    },
    {
      "epoch": 56.0,
      "grad_norm": 0.20707572996616364,
      "learning_rate": 9.1957526601295e-05,
      "loss": 0.0135,
      "step": 2240
    },
    {
      "epoch": 56.25,
      "grad_norm": 0.24764341115951538,
      "learning_rate": 9.186736524689281e-05,
      "loss": 0.0103,
      "step": 2250
    },
    {
      "epoch": 56.5,
      "grad_norm": 0.31629952788352966,
      "learning_rate": 9.177674603769204e-05,
      "loss": 0.017,
      "step": 2260
    },
    {
      "epoch": 56.75,
      "grad_norm": 0.29424598813056946,
      "learning_rate": 9.168566996468983e-05,
      "loss": 0.019,
      "step": 2270
    },
    {
      "epoch": 57.0,
      "grad_norm": 0.24435587227344513,
      "learning_rate": 9.159413802387951e-05,
      "loss": 0.0133,
      "step": 2280
    },
    {
      "epoch": 57.25,
      "grad_norm": 0.2817462086677551,
      "learning_rate": 9.150215121623974e-05,
      "loss": 0.0153,
      "step": 2290
    },
    {
      "epoch": 57.5,
      "grad_norm": 0.17769835889339447,
      "learning_rate": 9.140971054772349e-05,
      "loss": 0.0143,
      "step": 2300
    },
    {
      "epoch": 57.75,
      "grad_norm": 0.2751213014125824,
      "learning_rate": 9.131681702924713e-05,
      "loss": 0.0154,
      "step": 2310
    },
    {
      "epoch": 58.0,
      "grad_norm": 0.20847024023532867,
      "learning_rate": 9.122347167667926e-05,
      "loss": 0.0138,
      "step": 2320
    },
    {
      "epoch": 58.25,
      "grad_norm": 0.22063089907169342,
      "learning_rate": 9.112967551082973e-05,
      "loss": 0.0158,
      "step": 2330
    },
    {
      "epoch": 58.5,
      "grad_norm": 0.39056122303009033,
      "learning_rate": 9.103542955743835e-05,
      "loss": 0.0202,
      "step": 2340
    },
    {
      "epoch": 58.75,
      "grad_norm": 0.3934856355190277,
      "learning_rate": 9.094073484716381e-05,
      "loss": 0.0159,
      "step": 2350
    },
    {
      "epoch": 59.0,
      "grad_norm": 0.41992074251174927,
      "learning_rate": 9.084559241557226e-05,
      "loss": 0.0152,
      "step": 2360
    },
    {
      "epoch": 59.25,
      "grad_norm": 0.30681025981903076,
      "learning_rate": 9.075000330312608e-05,
      "loss": 0.012,
      "step": 2370
    },
    {
      "epoch": 59.5,
      "grad_norm": 0.21425344049930573,
      "learning_rate": 9.065396855517253e-05,
      "loss": 0.0121,
      "step": 2380
    },
    {
      "epoch": 59.75,
      "grad_norm": 0.24510438740253448,
      "learning_rate": 9.055748922193219e-05,
      "loss": 0.0144,
      "step": 2390
    },
    {
      "epoch": 60.0,
      "grad_norm": 0.2346717268228531,
      "learning_rate": 9.046056635848761e-05,
      "loss": 0.0197,
      "step": 2400
    },
    {
      "epoch": 60.25,
      "grad_norm": 0.40343689918518066,
      "learning_rate": 9.036320102477169e-05,
      "loss": 0.012,
      "step": 2410
    },
    {
      "epoch": 60.5,
      "grad_norm": 0.447936087846756,
      "learning_rate": 9.02653942855561e-05,
      "loss": 0.0119,
      "step": 2420
    },
    {
      "epoch": 60.75,
      "grad_norm": 0.43714478611946106,
      "learning_rate": 9.016714721043971e-05,
      "loss": 0.0141,
      "step": 2430
    },
    {
      "epoch": 61.0,
      "grad_norm": 0.2527328133583069,
      "learning_rate": 9.006846087383675e-05,
      "loss": 0.0134,
      "step": 2440
    },
    {
      "epoch": 61.25,
      "grad_norm": 0.26533153653144836,
      "learning_rate": 8.996933635496523e-05,
      "loss": 0.0146,
      "step": 2450
    },
    {
      "epoch": 61.5,
      "grad_norm": 0.18538479506969452,
      "learning_rate": 8.986977473783498e-05,
      "loss": 0.014,
      "step": 2460
    },
    {
      "epoch": 61.75,
      "grad_norm": 0.4042245149612427,
      "learning_rate": 8.97697771112359e-05,
      "loss": 0.0147,
      "step": 2470
    },
    {
      "epoch": 62.0,
      "grad_norm": 0.4519740045070648,
      "learning_rate": 8.966934456872602e-05,
      "loss": 0.0126,
      "step": 2480
    },
    {
      "epoch": 62.25,
      "grad_norm": 0.5037399530410767,
      "learning_rate": 8.95684782086195e-05,
      "loss": 0.0153,
      "step": 2490
    },
    {
      "epoch": 62.5,
      "grad_norm": 0.36741331219673157,
      "learning_rate": 8.946717913397476e-05,
      "loss": 0.0154,
      "step": 2500
    },
    {
      "epoch": 62.75,
      "grad_norm": 0.4969617426395416,
      "learning_rate": 8.93654484525822e-05,
      "loss": 0.016,
      "step": 2510
    },
    {
      "epoch": 63.0,
      "grad_norm": 0.3934231698513031,
      "learning_rate": 8.926328727695226e-05,
      "loss": 0.0155,
      "step": 2520
    },
    {
      "epoch": 63.25,
      "grad_norm": 0.3299417793750763,
      "learning_rate": 8.916069672430319e-05,
      "loss": 0.0123,
      "step": 2530
    },
    {
      "epoch": 63.5,
      "grad_norm": 0.4059097170829773,
      "learning_rate": 8.905767791654884e-05,
      "loss": 0.015,
      "step": 2540
    },
    {
      "epoch": 63.75,
      "grad_norm": 0.2909546494483948,
      "learning_rate": 8.895423198028638e-05,
      "loss": 0.0158,
      "step": 2550
    },
    {
      "epoch": 64.0,
      "grad_norm": 0.26669272780418396,
      "learning_rate": 8.885036004678402e-05,
      "loss": 0.0184,
      "step": 2560
    },
    {
      "epoch": 64.25,
      "grad_norm": 0.3337773382663727,
      "learning_rate": 8.874606325196857e-05,
      "loss": 0.0133,
      "step": 2570
    },
    {
      "epoch": 64.5,
      "grad_norm": 0.2659240961074829,
      "learning_rate": 8.864134273641304e-05,
      "loss": 0.0174,
      "step": 2580
    },
    {
      "epoch": 64.75,
      "grad_norm": 0.20509116351604462,
      "learning_rate": 8.853619964532427e-05,
      "loss": 0.0156,
      "step": 2590
    },
    {
      "epoch": 65.0,
      "grad_norm": 0.3653193414211273,
      "learning_rate": 8.843063512853019e-05,
      "loss": 0.0151,
      "step": 2600
    },
    {
      "epoch": 65.25,
      "grad_norm": 0.29937827587127686,
      "learning_rate": 8.832465034046749e-05,
      "loss": 0.0145,
      "step": 2610
    },
    {
      "epoch": 65.5,
      "grad_norm": 0.2896886467933655,
      "learning_rate": 8.821824644016882e-05,
      "loss": 0.0128,
      "step": 2620
    },
    {
      "epoch": 65.75,
      "grad_norm": 0.3686462342739105,
      "learning_rate": 8.811142459125019e-05,
      "loss": 0.0106,
      "step": 2630
    },
    {
      "epoch": 66.0,
      "grad_norm": 0.23774658143520355,
      "learning_rate": 8.800418596189822e-05,
      "loss": 0.0116,
      "step": 2640
    },
    {
      "epoch": 66.25,
      "grad_norm": 0.2722029387950897,
      "learning_rate": 8.789653172485737e-05,
      "loss": 0.0083,
      "step": 2650
    },
    {
      "epoch": 66.5,
      "grad_norm": 0.22109919786453247,
      "learning_rate": 8.778846305741715e-05,
      "loss": 0.0179,
      "step": 2660
    },
    {
      "epoch": 66.75,
      "grad_norm": 0.3148277699947357,
      "learning_rate": 8.767998114139918e-05,
      "loss": 0.0166,
      "step": 2670
    },
    {
      "epoch": 67.0,
      "grad_norm": 0.3373554050922394,
      "learning_rate": 8.757108716314429e-05,
      "loss": 0.0117,
      "step": 2680
    },
    {
      "epoch": 67.25,
      "grad_norm": 0.29900553822517395,
      "learning_rate": 8.746178231349962e-05,
      "loss": 0.0163,
      "step": 2690
    },
    {
      "epoch": 67.5,
      "grad_norm": 0.2480359971523285,
      "learning_rate": 8.735206778780549e-05,
      "loss": 0.017,
      "step": 2700
    },
    {
      "epoch": 67.75,
      "grad_norm": 0.3013180196285248,
      "learning_rate": 8.724194478588234e-05,
      "loss": 0.0162,
      "step": 2710
    },
    {
      "epoch": 68.0,
      "grad_norm": 0.23710669577121735,
      "learning_rate": 8.713141451201772e-05,
      "loss": 0.0149,
      "step": 2720
    },
    {
      "epoch": 68.25,
      "grad_norm": 0.3515462875366211,
      "learning_rate": 8.702047817495295e-05,
      "loss": 0.0134,
      "step": 2730
    },
    {
      "epoch": 68.5,
      "grad_norm": 0.3761940896511078,
      "learning_rate": 8.69091369878701e-05,
      "loss": 0.0159,
      "step": 2740
    },
    {
      "epoch": 68.75,
      "grad_norm": 0.3726300299167633,
      "learning_rate": 8.679739216837849e-05,
      "loss": 0.0122,
      "step": 2750
    },
    {
      "epoch": 69.0,
      "grad_norm": 0.26920396089553833,
      "learning_rate": 8.66852449385016e-05,
      "loss": 0.0129,
      "step": 2760
    },
    {
      "epoch": 69.25,
      "grad_norm": 0.378433495759964,
      "learning_rate": 8.657269652466356e-05,
      "loss": 0.0114,
      "step": 2770
    },
    {
      "epoch": 69.5,
      "grad_norm": 0.3709133267402649,
      "learning_rate": 8.645974815767577e-05,
      "loss": 0.0131,
      "step": 2780
    },
    {
      "epoch": 69.75,
      "grad_norm": 0.22590842843055725,
      "learning_rate": 8.634640107272351e-05,
      "loss": 0.0144,
      "step": 2790
    },
    {
      "epoch": 70.0,
      "grad_norm": 0.3659012019634247,
      "learning_rate": 8.623265650935234e-05,
      "loss": 0.0116,
      "step": 2800
    },
    {
      "epoch": 70.25,
      "grad_norm": 0.28762730956077576,
      "learning_rate": 8.611851571145456e-05,
      "loss": 0.0133,
      "step": 2810
    },
    {
      "epoch": 70.5,
      "grad_norm": 0.44853711128234863,
      "learning_rate": 8.600397992725566e-05,
      "loss": 0.0126,
      "step": 2820
    },
    {
      "epoch": 70.75,
      "grad_norm": 0.3202303349971771,
      "learning_rate": 8.588905040930061e-05,
      "loss": 0.0135,
      "step": 2830
    },
    {
      "epoch": 71.0,
      "grad_norm": 0.27304911613464355,
      "learning_rate": 8.577372841444022e-05,
      "loss": 0.0103,
      "step": 2840
    },
    {
      "epoch": 71.25,
      "grad_norm": 0.3241058588027954,
      "learning_rate": 8.565801520381736e-05,
      "loss": 0.0144,
      "step": 2850
    },
    {
      "epoch": 71.5,
      "grad_norm": 0.3490132689476013,
      "learning_rate": 8.554191204285313e-05,
      "loss": 0.0166,
      "step": 2860
    },
    {
      "epoch": 71.75,
      "grad_norm": 0.22821162641048431,
      "learning_rate": 8.542542020123315e-05,
      "loss": 0.0118,
      "step": 2870
    },
    {
      "epoch": 72.0,
      "grad_norm": 0.21379388868808746,
      "learning_rate": 8.530854095289347e-05,
      "loss": 0.0122,
      "step": 2880
    },
    {
      "epoch": 72.25,
      "grad_norm": 0.2390654981136322,
      "learning_rate": 8.519127557600688e-05,
      "loss": 0.0117,
      "step": 2890
    },
    {
      "epoch": 72.5,
      "grad_norm": 0.2505422532558441,
      "learning_rate": 8.507362535296871e-05,
      "loss": 0.015,
      "step": 2900
    },
    {
      "epoch": 72.75,
      "grad_norm": 0.31641829013824463,
      "learning_rate": 8.495559157038299e-05,
      "loss": 0.015,
      "step": 2910
    },
    {
      "epoch": 73.0,
      "grad_norm": 0.5015877485275269,
      "learning_rate": 8.483717551904823e-05,
      "loss": 0.0152,
      "step": 2920
    },
    {
      "epoch": 73.25,
      "grad_norm": 0.373805433511734,
      "learning_rate": 8.47183784939434e-05,
      "loss": 0.0098,
      "step": 2930
    },
    {
      "epoch": 73.5,
      "grad_norm": 0.31168583035469055,
      "learning_rate": 8.459920179421374e-05,
      "loss": 0.0102,
      "step": 2940
    },
    {
      "epoch": 73.75,
      "grad_norm": 0.25392434000968933,
      "learning_rate": 8.447964672315656e-05,
      "loss": 0.0121,
      "step": 2950
    },
    {
      "epoch": 74.0,
      "grad_norm": 0.2583267390727997,
      "learning_rate": 8.435971458820692e-05,
      "loss": 0.014,
      "step": 2960
    },
    {
      "epoch": 74.25,
      "grad_norm": 0.2887478470802307,
      "learning_rate": 8.423940670092345e-05,
      "loss": 0.0143,
      "step": 2970
    },
    {
      "epoch": 74.5,
      "grad_norm": 0.20327523350715637,
      "learning_rate": 8.411872437697394e-05,
      "loss": 0.014,
      "step": 2980
    },
    {
      "epoch": 74.75,
      "grad_norm": 0.20265914499759674,
      "learning_rate": 8.399766893612096e-05,
      "loss": 0.0102,
      "step": 2990
    },
    {
      "epoch": 75.0,
      "grad_norm": 0.32352325320243835,
      "learning_rate": 8.38762417022074e-05,
      "loss": 0.0109,
      "step": 3000
    },
    {
      "epoch": 75.25,
      "grad_norm": 0.23712804913520813,
      "learning_rate": 8.375444400314204e-05,
      "loss": 0.0145,
      "step": 3010
    },
    {
      "epoch": 75.5,
      "grad_norm": 0.2874774634838104,
      "learning_rate": 8.3632277170885e-05,
      "loss": 0.0125,
      "step": 3020
    },
    {
      "epoch": 75.75,
      "grad_norm": 0.23495231568813324,
      "learning_rate": 8.350974254143318e-05,
      "loss": 0.0153,
      "step": 3030
    },
    {
      "epoch": 76.0,
      "grad_norm": 0.25911620259284973,
      "learning_rate": 8.338684145480566e-05,
      "loss": 0.0136,
      "step": 3040
    },
    {
      "epoch": 76.25,
      "grad_norm": 0.28189530968666077,
      "learning_rate": 8.326357525502904e-05,
      "loss": 0.0113,
      "step": 3050
    },
    {
      "epoch": 76.5,
      "grad_norm": 0.3920529782772064,
      "learning_rate": 8.313994529012273e-05,
      "loss": 0.0194,
      "step": 3060
    },
    {
      "epoch": 76.75,
      "grad_norm": 0.30341243743896484,
      "learning_rate": 8.301595291208422e-05,
      "loss": 0.0174,
      "step": 3070
    },
    {
      "epoch": 77.0,
      "grad_norm": 0.417825847864151,
      "learning_rate": 8.289159947687427e-05,
      "loss": 0.0172,
      "step": 3080
    },
    {
      "epoch": 77.25,
      "grad_norm": 0.3963508903980255,
      "learning_rate": 8.276688634440216e-05,
      "loss": 0.014,
      "step": 3090
    },
    {
      "epoch": 77.5,
      "grad_norm": 0.3563499450683594,
      "learning_rate": 8.26418148785107e-05,
      "loss": 0.0161,
      "step": 3100
    },
    {
      "epoch": 77.75,
      "grad_norm": 0.3115830719470978,
      "learning_rate": 8.251638644696141e-05,
      "loss": 0.0149,
      "step": 3110
    },
    {
      "epoch": 78.0,
      "grad_norm": 0.3986978530883789,
      "learning_rate": 8.23906024214195e-05,
      "loss": 0.0153,
      "step": 3120
    },
    {
      "epoch": 78.25,
      "grad_norm": 0.27240195870399475,
      "learning_rate": 8.226446417743897e-05,
      "loss": 0.0126,
      "step": 3130
    },
    {
      "epoch": 78.5,
      "grad_norm": 0.2304350584745407,
      "learning_rate": 8.213797309444742e-05,
      "loss": 0.0142,
      "step": 3140
    },
    {
      "epoch": 78.75,
      "grad_norm": 0.1990521103143692,
      "learning_rate": 8.201113055573105e-05,
      "loss": 0.0126,
      "step": 3150
    },
    {
      "epoch": 79.0,
      "grad_norm": 0.1642603874206543,
      "learning_rate": 8.188393794841958e-05,
      "loss": 0.0121,
      "step": 3160
    },
    {
      "epoch": 79.25,
      "grad_norm": 0.23555266857147217,
      "learning_rate": 8.175639666347094e-05,
      "loss": 0.0111,
      "step": 3170
    },
    {
      "epoch": 79.5,
      "grad_norm": 0.28729990124702454,
      "learning_rate": 8.162850809565623e-05,
      "loss": 0.0152,
      "step": 3180
    },
    {
      "epoch": 79.75,
      "grad_norm": 0.33808666467666626,
      "learning_rate": 8.150027364354431e-05,
      "loss": 0.0127,
      "step": 3190
    },
    {
      "epoch": 80.0,
      "grad_norm": 0.4436752200126648,
      "learning_rate": 8.137169470948662e-05,
      "loss": 0.0149,
      "step": 3200
    },
    {
      "epoch": 80.25,
      "grad_norm": 0.4043782949447632,
      "learning_rate": 8.124277269960179e-05,
      "loss": 0.0146,
      "step": 3210
    },
    {
      "epoch": 80.5,
      "grad_norm": 0.23915347456932068,
      "learning_rate": 8.111350902376023e-05,
      "loss": 0.014,
      "step": 3220
    },
    {
      "epoch": 80.75,
      "grad_norm": 0.175531268119812,
      "learning_rate": 8.098390509556883e-05,
      "loss": 0.0098,
      "step": 3230
    },
    {
      "epoch": 81.0,
      "grad_norm": 0.3109751045703888,
      "learning_rate": 8.085396233235536e-05,
      "loss": 0.0161,
      "step": 3240
    },
    {
      "epoch": 81.25,
      "grad_norm": 0.19472594559192657,
      "learning_rate": 8.072368215515306e-05,
      "loss": 0.0113,
      "step": 3250
    },
    {
      "epoch": 81.5,
      "grad_norm": 0.2841004431247711,
      "learning_rate": 8.059306598868506e-05,
      "loss": 0.0128,
      "step": 3260
    },
    {
      "epoch": 81.75,
      "grad_norm": 0.23794560134410858,
      "learning_rate": 8.046211526134888e-05,
      "loss": 0.0137,
      "step": 3270
    },
    {
      "epoch": 82.0,
      "grad_norm": 0.3024290204048157,
      "learning_rate": 8.033083140520065e-05,
      "loss": 0.0135,
      "step": 3280
    },
    {
      "epoch": 82.25,
      "grad_norm": 0.3859151303768158,
      "learning_rate": 8.019921585593962e-05,
      "loss": 0.0108,
      "step": 3290
    },
    {
      "epoch": 82.5,
      "grad_norm": 0.30000364780426025,
      "learning_rate": 8.006727005289232e-05,
      "loss": 0.0111,
      "step": 3300
    },
    {
      "epoch": 82.75,
      "grad_norm": 0.23926366865634918,
      "learning_rate": 7.993499543899692e-05,
      "loss": 0.01,
      "step": 3310
    },
    {
      "epoch": 83.0,
      "grad_norm": 0.36803534626960754,
      "learning_rate": 7.980239346078742e-05,
      "loss": 0.0116,
      "step": 3320
    },
    {
      "epoch": 83.25,
      "grad_norm": 0.5533159971237183,
      "learning_rate": 7.966946556837778e-05,
      "loss": 0.0115,
      "step": 3330
    },
    {
      "epoch": 83.5,
      "grad_norm": 0.2806551158428192,
      "learning_rate": 7.953621321544616e-05,
      "loss": 0.0121,
      "step": 3340
    },
    {
      "epoch": 83.75,
      "grad_norm": 0.32250022888183594,
      "learning_rate": 7.940263785921896e-05,
      "loss": 0.0111,
      "step": 3350
    },
    {
      "epoch": 84.0,
      "grad_norm": 0.21136438846588135,
      "learning_rate": 7.926874096045482e-05,
      "loss": 0.0124,
      "step": 3360
    },
    {
      "epoch": 84.25,
      "grad_norm": 0.29519808292388916,
      "learning_rate": 7.913452398342881e-05,
      "loss": 0.0119,
      "step": 3370
    },
    {
      "epoch": 84.5,
      "grad_norm": 0.20215508341789246,
      "learning_rate": 7.89999883959163e-05,
      "loss": 0.0188,
      "step": 3380
    },
    {
      "epoch": 84.75,
      "grad_norm": 0.20980972051620483,
      "learning_rate": 7.886513566917687e-05,
      "loss": 0.0141,
      "step": 3390
    },
    {
      "epoch": 85.0,
      "grad_norm": 0.15065793693065643,
      "learning_rate": 7.872996727793838e-05,
      "loss": 0.0119,
      "step": 3400
    },
    {
      "epoch": 85.25,
      "grad_norm": 0.30695223808288574,
      "learning_rate": 7.859448470038069e-05,
      "loss": 0.0151,
      "step": 3410
    },
    {
      "epoch": 85.5,
      "grad_norm": 0.17749333381652832,
      "learning_rate": 7.845868941811956e-05,
      "loss": 0.0125,
      "step": 3420
    },
    {
      "epoch": 85.75,
      "grad_norm": 0.16342045366764069,
      "learning_rate": 7.832258291619043e-05,
      "loss": 0.0111,
      "step": 3430
    },
    {
      "epoch": 86.0,
      "grad_norm": 0.29444870352745056,
      "learning_rate": 7.81861666830322e-05,
      "loss": 0.0112,
      "step": 3440
    },
    {
      "epoch": 86.25,
      "grad_norm": 0.4270837903022766,
      "learning_rate": 7.804944221047097e-05,
      "loss": 0.0137,
      "step": 3450
    },
    {
      "epoch": 86.5,
      "grad_norm": 0.3473292887210846,
      "learning_rate": 7.791241099370364e-05,
      "loss": 0.0136,
      "step": 3460
    },
    {
      "epoch": 86.75,
      "grad_norm": 0.38873690366744995,
      "learning_rate": 7.777507453128163e-05,
      "loss": 0.0107,
      "step": 3470
    },
    {
      "epoch": 87.0,
      "grad_norm": 0.2898962199687958,
      "learning_rate": 7.763743432509451e-05,
      "loss": 0.0097,
      "step": 3480
    },
    {
      "epoch": 87.25,
      "grad_norm": 0.2728941738605499,
      "learning_rate": 7.749949188035353e-05,
      "loss": 0.013,
      "step": 3490
    },
    {
      "epoch": 87.5,
      "grad_norm": 0.203449085354805,
      "learning_rate": 7.736124870557516e-05,
      "loss": 0.016,
      "step": 3500
    },
    {
      "epoch": 87.75,
      "grad_norm": 0.28565430641174316,
      "learning_rate": 7.722270631256459e-05,
      "loss": 0.0123,
      "step": 3510
    },
    {
      "epoch": 88.0,
      "grad_norm": 0.32342642545700073,
      "learning_rate": 7.708386621639925e-05,
      "loss": 0.0134,
      "step": 3520
    },
    {
      "epoch": 88.25,
      "grad_norm": 0.26711976528167725,
      "learning_rate": 7.694472993541219e-05,
      "loss": 0.0113,
      "step": 3530
    },
    {
      "epoch": 88.5,
      "grad_norm": 0.26559197902679443,
      "learning_rate": 7.680529899117547e-05,
      "loss": 0.0137,
      "step": 3540
    },
    {
      "epoch": 88.75,
      "grad_norm": 0.32358163595199585,
      "learning_rate": 7.666557490848358e-05,
      "loss": 0.0159,
      "step": 3550
    },
    {
      "epoch": 89.0,
      "grad_norm": 0.26062920689582825,
      "learning_rate": 7.65255592153367e-05,
      "loss": 0.0126,
      "step": 3560
    },
    {
      "epoch": 89.25,
      "grad_norm": 0.22964859008789062,
      "learning_rate": 7.638525344292402e-05,
      "loss": 0.0123,
      "step": 3570
    },
    {
      "epoch": 89.5,
      "grad_norm": 0.17719513177871704,
      "learning_rate": 7.624465912560697e-05,
      "loss": 0.0113,
      "step": 3580
    },
    {
      "epoch": 89.75,
      "grad_norm": 0.23025670647621155,
      "learning_rate": 7.610377780090249e-05,
      "loss": 0.0156,
      "step": 3590
    },
    {
      "epoch": 90.0,
      "grad_norm": 0.23307792842388153,
      "learning_rate": 7.596261100946618e-05,
      "loss": 0.0123,
      "step": 3600
    },
    {
      "epoch": 90.25,
      "grad_norm": 0.17993809282779694,
      "learning_rate": 7.582116029507542e-05,
      "loss": 0.0107,
      "step": 3610
    },
    {
      "epoch": 90.5,
      "grad_norm": 0.17383725941181183,
      "learning_rate": 7.56794272046126e-05,
      "loss": 0.0099,
      "step": 3620
    },
    {
      "epoch": 90.75,
      "grad_norm": 0.24793113768100739,
      "learning_rate": 7.55374132880481e-05,
      "loss": 0.0117,
      "step": 3630
    },
    {
      "epoch": 91.0,
      "grad_norm": 0.3454414904117584,
      "learning_rate": 7.539512009842333e-05,
      "loss": 0.0143,
      "step": 3640
    },
    {
      "epoch": 91.25,
      "grad_norm": 0.22444981336593628,
      "learning_rate": 7.525254919183382e-05,
      "loss": 0.0112,
      "step": 3650
    },
    {
      "epoch": 91.5,
      "grad_norm": 0.3071042001247406,
      "learning_rate": 7.510970212741215e-05,
      "loss": 0.0141,
      "step": 3660
    },
    {
      "epoch": 91.75,
      "grad_norm": 0.13223372399806976,
      "learning_rate": 7.496658046731096e-05,
      "loss": 0.0094,
      "step": 3670
    },
    {
      "epoch": 92.0,
      "grad_norm": 0.1920013278722763,
      "learning_rate": 7.482318577668578e-05,
      "loss": 0.011,
      "step": 3680
    },
    {
      "epoch": 92.25,
      "grad_norm": 0.1618950366973877,
      "learning_rate": 7.467951962367796e-05,
      "loss": 0.0119,
      "step": 3690
    },
    {
      "epoch": 92.5,
      "grad_norm": 0.2321500927209854,
      "learning_rate": 7.453558357939755e-05,
      "loss": 0.0133,
      "step": 3700
    },
    {
      "epoch": 92.75,
      "grad_norm": 0.310839980840683,
      "learning_rate": 7.439137921790606e-05,
      "loss": 0.0112,
      "step": 3710
    },
    {
      "epoch": 93.0,
      "grad_norm": 0.1530730277299881,
      "learning_rate": 7.42469081161993e-05,
      "loss": 0.0149,
      "step": 3720
    },
    {
      "epoch": 93.25,
      "grad_norm": 0.3117826282978058,
      "learning_rate": 7.410217185419006e-05,
      "loss": 0.0147,
      "step": 3730
    },
    {
      "epoch": 93.5,
      "grad_norm": 0.20995257794857025,
      "learning_rate": 7.395717201469095e-05,
      "loss": 0.0105,
      "step": 3740
    },
    {
      "epoch": 93.75,
      "grad_norm": 0.20923268795013428,
      "learning_rate": 7.381191018339696e-05,
      "loss": 0.0141,
      "step": 3750
    },
    {
      "epoch": 94.0,
      "grad_norm": 0.16217997670173645,
      "learning_rate": 7.36663879488682e-05,
      "loss": 0.0109,
      "step": 3760
    },
    {
      "epoch": 94.25,
      "grad_norm": 0.1856212615966797,
      "learning_rate": 7.352060690251254e-05,
      "loss": 0.011,
      "step": 3770
    },
    {
      "epoch": 94.5,
      "grad_norm": 0.274571031332016,
      "learning_rate": 7.337456863856811e-05,
      "loss": 0.0148,
      "step": 3780
    },
    {
      "epoch": 94.75,
      "grad_norm": 0.2003510743379593,
      "learning_rate": 7.3228274754086e-05,
      "loss": 0.0103,
      "step": 3790
    },
    {
      "epoch": 95.0,
      "grad_norm": 0.25044745206832886,
      "learning_rate": 7.308172684891267e-05,
      "loss": 0.0116,
      "step": 3800
    },
    {
      "epoch": 95.25,
      "grad_norm": 0.26451465487480164,
      "learning_rate": 7.293492652567255e-05,
      "loss": 0.0136,
      "step": 3810
    },
    {
      "epoch": 95.5,
      "grad_norm": 0.21916158497333527,
      "learning_rate": 7.278787538975043e-05,
      "loss": 0.0094,
      "step": 3820
    },
    {
      "epoch": 95.75,
      "grad_norm": 0.19029930233955383,
      "learning_rate": 7.2640575049274e-05,
      "loss": 0.0143,
      "step": 3830
    },
    {
      "epoch": 96.0,
      "grad_norm": 0.20158341526985168,
      "learning_rate": 7.249302711509616e-05,
      "loss": 0.0126,
      "step": 3840
    },
    {
      "epoch": 96.25,
      "grad_norm": 0.2865701913833618,
      "learning_rate": 7.23452332007775e-05,
      "loss": 0.0108,
      "step": 3850
    },
    {
      "epoch": 96.5,
      "grad_norm": 0.3106520175933838,
      "learning_rate": 7.219719492256858e-05,
      "loss": 0.0094,
      "step": 3860
    },
    {
      "epoch": 96.75,
      "grad_norm": 0.2534642815589905,
      "learning_rate": 7.20489138993923e-05,
      "loss": 0.0142,
      "step": 3870
    },
    {
      "epoch": 97.0,
      "grad_norm": 0.267780065536499,
      "learning_rate": 7.190039175282614e-05,
      "loss": 0.0104,
      "step": 3880
    },
    {
      "epoch": 97.25,
      "grad_norm": 0.23589850962162018,
      "learning_rate": 7.175163010708455e-05,
      "loss": 0.0105,
      "step": 3890
    },
    {
      "epoch": 97.5,
      "grad_norm": 0.274258017539978,
      "learning_rate": 7.1602630589001e-05,
      "loss": 0.0114,
      "step": 3900
    },
    {
      "epoch": 97.75,
      "grad_norm": 0.19494503736495972,
      "learning_rate": 7.14533948280104e-05,
      "loss": 0.0106,
      "step": 3910
    },
    {
      "epoch": 98.0,
      "grad_norm": 0.1633615493774414,
      "learning_rate": 7.130392445613109e-05,
      "loss": 0.0093,
      "step": 3920
    },
    {
      "epoch": 98.25,
      "grad_norm": 0.3438349664211273,
      "learning_rate": 7.115422110794711e-05,
      "loss": 0.0097,
      "step": 3930
    },
    {
      "epoch": 98.5,
      "grad_norm": 0.2983504831790924,
      "learning_rate": 7.100428642059033e-05,
      "loss": 0.0096,
      "step": 3940
    },
    {
      "epoch": 98.75,
      "grad_norm": 0.2533060610294342,
      "learning_rate": 7.08541220337224e-05,
      "loss": 0.0109,
      "step": 3950
    },
    {
      "epoch": 99.0,
      "grad_norm": 0.2611391842365265,
      "learning_rate": 7.070372958951706e-05,
      "loss": 0.0145,
      "step": 3960
    },
    {
      "epoch": 99.25,
      "grad_norm": 0.12219741195440292,
      "learning_rate": 7.055311073264194e-05,
      "loss": 0.0101,
      "step": 3970
    },
    {
      "epoch": 99.5,
      "grad_norm": 0.23046687245368958,
      "learning_rate": 7.040226711024077e-05,
      "loss": 0.0114,
      "step": 3980
    },
    {
      "epoch": 99.75,
      "grad_norm": 0.28500500321388245,
      "learning_rate": 7.02512003719152e-05,
      "loss": 0.0102,
      "step": 3990
    },
    {
      "epoch": 100.0,
      "grad_norm": 0.28569942712783813,
      "learning_rate": 7.00999121697069e-05,
      "loss": 0.0075,
      "step": 4000
    },
    {
      "epoch": 100.25,
      "grad_norm": 0.2506348788738251,
      "learning_rate": 6.99484041580794e-05,
      "loss": 0.0094,
      "step": 4010
    },
    {
      "epoch": 100.5,
      "grad_norm": 0.19107772409915924,
      "learning_rate": 6.979667799390004e-05,
      "loss": 0.0073,
      "step": 4020
    },
    {
      "epoch": 100.75,
      "grad_norm": 0.24592654407024384,
      "learning_rate": 6.964473533642185e-05,
      "loss": 0.0087,
      "step": 4030
    },
    {
      "epoch": 101.0,
      "grad_norm": 0.13424347341060638,
      "learning_rate": 6.949257784726539e-05,
      "loss": 0.0112,
      "step": 4040
    },
    {
      "epoch": 101.25,
      "grad_norm": 0.18724428117275238,
      "learning_rate": 6.934020719040056e-05,
      "loss": 0.0114,
      "step": 4050
    },
    {
      "epoch": 101.5,
      "grad_norm": 0.3430480360984802,
      "learning_rate": 6.918762503212848e-05,
      "loss": 0.0112,
      "step": 4060
    },
    {
      "epoch": 101.75,
      "grad_norm": 0.41409286856651306,
      "learning_rate": 6.903483304106319e-05,
      "loss": 0.0121,
      "step": 4070
    },
    {
      "epoch": 102.0,
      "grad_norm": 0.209912970662117,
      "learning_rate": 6.888183288811341e-05,
      "loss": 0.0119,
      "step": 4080
    },
    {
      "epoch": 102.25,
      "grad_norm": 0.438534140586853,
      "learning_rate": 6.87286262464643e-05,
      "loss": 0.0119,
      "step": 4090
    },
    {
      "epoch": 102.5,
      "grad_norm": 0.3194650113582611,
      "learning_rate": 6.857521479155915e-05,
      "loss": 0.0099,
      "step": 4100
    },
    {
      "epoch": 102.75,
      "grad_norm": 0.2407376617193222,
      "learning_rate": 6.842160020108104e-05,
      "loss": 0.0122,
      "step": 4110
    },
    {
      "epoch": 103.0,
      "grad_norm": 0.16524438560009003,
      "learning_rate": 6.826778415493455e-05,
      "loss": 0.0107,
      "step": 4120
    },
    {
      "epoch": 103.25,
      "grad_norm": 0.2268092781305313,
      "learning_rate": 6.811376833522729e-05,
      "loss": 0.0122,
      "step": 4130
    },
    {
      "epoch": 103.5,
      "grad_norm": 0.2470845878124237,
      "learning_rate": 6.795955442625159e-05,
      "loss": 0.0117,
      "step": 4140
    },
    {
      "epoch": 103.75,
      "grad_norm": 0.4060851037502289,
      "learning_rate": 6.780514411446608e-05,
      "loss": 0.014,
      "step": 4150
    },
    {
      "epoch": 104.0,
      "grad_norm": 0.25651276111602783,
      "learning_rate": 6.765053908847716e-05,
      "loss": 0.0099,
      "step": 4160
    },
    {
      "epoch": 104.25,
      "grad_norm": 0.3012460768222809,
      "learning_rate": 6.749574103902064e-05,
      "loss": 0.0118,
      "step": 4170
    },
    {
      "epoch": 104.5,
      "grad_norm": 0.3353002965450287,
      "learning_rate": 6.734075165894317e-05,
      "loss": 0.014,
      "step": 4180
    },
    {
      "epoch": 104.75,
      "grad_norm": 0.41446223855018616,
      "learning_rate": 6.71855726431838e-05,
      "loss": 0.0131,
      "step": 4190
    },
    {
      "epoch": 105.0,
      "grad_norm": 0.24888281524181366,
      "learning_rate": 6.703020568875538e-05,
      "loss": 0.0164,
      "step": 4200
    },
    {
      "epoch": 105.25,
      "grad_norm": 0.3254045844078064,
      "learning_rate": 6.687465249472603e-05,
      "loss": 0.0115,
      "step": 4210
    },
    {
      "epoch": 105.5,
      "grad_norm": 0.3065283000469208,
      "learning_rate": 6.671891476220055e-05,
      "loss": 0.0098,
      "step": 4220
    },
    {
      "epoch": 105.75,
      "grad_norm": 0.26779666543006897,
      "learning_rate": 6.656299419430183e-05,
      "loss": 0.009,
      "step": 4230
    },
    {
      "epoch": 106.0,
      "grad_norm": 0.2584543526172638,
      "learning_rate": 6.640689249615223e-05,
      "loss": 0.0089,
      "step": 4240
    },
    {
      "epoch": 106.25,
      "grad_norm": 0.39003482460975647,
      "learning_rate": 6.625061137485491e-05,
      "loss": 0.0107,
      "step": 4250
    },
    {
      "epoch": 106.5,
      "grad_norm": 0.19243167340755463,
      "learning_rate": 6.609415253947517e-05,
      "loss": 0.0111,
      "step": 4260
    },
    {
      "epoch": 106.75,
      "grad_norm": 0.2182435542345047,
      "learning_rate": 6.593751770102178e-05,
      "loss": 0.0124,
      "step": 4270
    },
    {
      "epoch": 107.0,
      "grad_norm": 0.1765671819448471,
      "learning_rate": 6.578070857242823e-05,
      "loss": 0.0112,
      "step": 4280
    },
    {
      "epoch": 107.25,
      "grad_norm": 0.20157065987586975,
      "learning_rate": 6.562372686853402e-05,
      "loss": 0.0092,
      "step": 4290
    },
    {
      "epoch": 107.5,
      "grad_norm": 0.28021109104156494,
      "learning_rate": 6.546657430606593e-05,
      "loss": 0.0096,
      "step": 4300
    },
    {
      "epoch": 107.75,
      "grad_norm": 0.27157220244407654,
      "learning_rate": 6.530925260361918e-05,
      "loss": 0.0112,
      "step": 4310
    },
    {
      "epoch": 108.0,
      "grad_norm": 0.265423446893692,
      "learning_rate": 6.515176348163871e-05,
      "loss": 0.0094,
      "step": 4320
    },
    {
      "epoch": 108.25,
      "grad_norm": 0.23950523138046265,
      "learning_rate": 6.499410866240032e-05,
      "loss": 0.0109,
      "step": 4330
    },
    {
      "epoch": 108.5,
      "grad_norm": 0.17336910963058472,
      "learning_rate": 6.48362898699919e-05,
      "loss": 0.0093,
      "step": 4340
    },
    {
      "epoch": 108.75,
      "grad_norm": 0.19538640975952148,
      "learning_rate": 6.467830883029443e-05,
      "loss": 0.0088,
      "step": 4350
    },
    {
      "epoch": 109.0,
      "grad_norm": 0.26300865411758423,
      "learning_rate": 6.452016727096326e-05,
      "loss": 0.008,
      "step": 4360
    },
    {
      "epoch": 109.25,
      "grad_norm": 0.4020210802555084,
      "learning_rate": 6.436186692140916e-05,
      "loss": 0.0086,
      "step": 4370
    },
    {
      "epoch": 109.5,
      "grad_norm": 0.15748588740825653,
      "learning_rate": 6.420340951277938e-05,
      "loss": 0.0075,
      "step": 4380
    },
    {
      "epoch": 109.75,
      "grad_norm": 0.2636542320251465,
      "learning_rate": 6.404479677793874e-05,
      "loss": 0.0091,
      "step": 4390
    },
    {
      "epoch": 110.0,
      "grad_norm": 0.2198638617992401,
      "learning_rate": 6.388603045145075e-05,
      "loss": 0.0123,
      "step": 4400
    },
    {
      "epoch": 110.25,
      "grad_norm": 0.2724881172180176,
      "learning_rate": 6.372711226955843e-05,
      "loss": 0.0147,
      "step": 4410
    },
    {
      "epoch": 110.5,
      "grad_norm": 0.3189849853515625,
      "learning_rate": 6.356804397016564e-05,
      "loss": 0.0118,
      "step": 4420
    },
    {
      "epoch": 110.75,
      "grad_norm": 0.24298754334449768,
      "learning_rate": 6.340882729281779e-05,
      "loss": 0.0131,
      "step": 4430
    },
    {
      "epoch": 111.0,
      "grad_norm": 0.3537801206111908,
      "learning_rate": 6.324946397868294e-05,
      "loss": 0.011,
      "step": 4440
    },
    {
      "epoch": 111.25,
      "grad_norm": 0.430812269449234,
      "learning_rate": 6.308995577053276e-05,
      "loss": 0.0162,
      "step": 4450
    },
    {
      "epoch": 111.5,
      "grad_norm": 0.18427328765392303,
      "learning_rate": 6.293030441272347e-05,
      "loss": 0.0104,
      "step": 4460
    },
    {
      "epoch": 111.75,
      "grad_norm": 0.3262835144996643,
      "learning_rate": 6.277051165117677e-05,
      "loss": 0.012,
      "step": 4470
    },
    {
      "epoch": 112.0,
      "grad_norm": 0.3320060968399048,
      "learning_rate": 6.261057923336064e-05,
      "loss": 0.0102,
      "step": 4480
    },
    {
      "epoch": 112.25,
      "grad_norm": 0.3201114535331726,
      "learning_rate": 6.245050890827042e-05,
      "loss": 0.0114,
      "step": 4490
    },
    {
      "epoch": 112.5,
      "grad_norm": 0.19090622663497925,
      "learning_rate": 6.229030242640952e-05,
      "loss": 0.0119,
      "step": 4500
    },
    {
      "epoch": 112.75,
      "grad_norm": 0.3037988841533661,
      "learning_rate": 6.212996153977037e-05,
      "loss": 0.0113,
      "step": 4510
    },
    {
      "epoch": 113.0,
      "grad_norm": 0.27645960450172424,
      "learning_rate": 6.196948800181523e-05,
      "loss": 0.0133,
      "step": 4520
    },
    {
      "epoch": 113.25,
      "grad_norm": 0.22976121306419373,
      "learning_rate": 6.180888356745695e-05,
      "loss": 0.011,
      "step": 4530
    },
    {
      "epoch": 113.5,
      "grad_norm": 0.2242169827222824,
      "learning_rate": 6.164814999303995e-05,
      "loss": 0.0108,
      "step": 4540
    },
    {
      "epoch": 113.75,
      "grad_norm": 0.26727139949798584,
      "learning_rate": 6.148728903632081e-05,
      "loss": 0.0093,
      "step": 4550
    },
    {
      "epoch": 114.0,
      "grad_norm": 0.30186429619789124,
      "learning_rate": 6.132630245644921e-05,
      "loss": 0.0096,
      "step": 4560
    },
    {
      "epoch": 114.25,
      "grad_norm": 0.2465476542711258,
      "learning_rate": 6.116519201394857e-05,
      "loss": 0.0079,
      "step": 4570
    },
    {
      "epoch": 114.5,
      "grad_norm": 0.22743864357471466,
      "learning_rate": 6.10039594706969e-05,
      "loss": 0.0101,
      "step": 4580
    },
    {
      "epoch": 114.75,
      "grad_norm": 0.3261238932609558,
      "learning_rate": 6.084260658990744e-05,
      "loss": 0.0088,
      "step": 4590
    },
    {
      "epoch": 115.0,
      "grad_norm": 0.17115327715873718,
      "learning_rate": 6.068113513610943e-05,
      "loss": 0.0088,
      "step": 4600
    },
    {
      "epoch": 115.25,
      "grad_norm": 0.2672770023345947,
      "learning_rate": 6.0519546875128876e-05,
      "loss": 0.0088,
      "step": 4610
    },
    {
      "epoch": 115.5,
      "grad_norm": 0.2757582664489746,
      "learning_rate": 6.035784357406906e-05,
      "loss": 0.012,
      "step": 4620
    },
    {
      "epoch": 115.75,
      "grad_norm": 0.2405472844839096,
      "learning_rate": 6.01960270012914e-05,
      "loss": 0.0136,
      "step": 4630
    },
    {
      "epoch": 116.0,
      "grad_norm": 0.2878851890563965,
      "learning_rate": 6.003409892639599e-05,
      "loss": 0.0104,
      "step": 4640
    },
    {
      "epoch": 116.25,
      "grad_norm": 0.3704386353492737,
      "learning_rate": 5.9872061120202336e-05,
      "loss": 0.012,
      "step": 4650
    },
    {
      "epoch": 116.5,
      "grad_norm": 0.19459301233291626,
      "learning_rate": 5.9709915354729914e-05,
      "loss": 0.0108,
      "step": 4660
    },
    {
      "epoch": 116.75,
      "grad_norm": 0.2355605810880661,
      "learning_rate": 5.9547663403178824e-05,
      "loss": 0.0184,
      "step": 4670
    },
    {
      "epoch": 117.0,
      "grad_norm": 0.3160833716392517,
      "learning_rate": 5.9385307039910445e-05,
      "loss": 0.0103,
      "step": 4680
    },
    {
      "epoch": 117.25,
      "grad_norm": 0.2000928819179535,
      "learning_rate": 5.922284804042792e-05,
      "loss": 0.0086,
      "step": 4690
    },
    {
      "epoch": 117.5,
      "grad_norm": 0.14318124949932098,
      "learning_rate": 5.906028818135687e-05,
      "loss": 0.0201,
      "step": 4700
    },
    {
      "epoch": 117.75,
      "grad_norm": 0.21847835183143616,
      "learning_rate": 5.889762924042585e-05,
      "loss": 0.0167,
      "step": 4710
    },
    {
      "epoch": 118.0,
      "grad_norm": 0.3395015597343445,
      "learning_rate": 5.873487299644699e-05,
      "loss": 0.0103,
      "step": 4720
    },
    {
      "epoch": 118.25,
      "grad_norm": 0.21214354038238525,
      "learning_rate": 5.857202122929649e-05,
      "loss": 0.0107,
      "step": 4730
    },
    {
      "epoch": 118.5,
      "grad_norm": 0.2229463905096054,
      "learning_rate": 5.840907571989518e-05,
      "loss": 0.011,
      "step": 4740
    },
    {
      "epoch": 118.75,
      "grad_norm": 0.22180570662021637,
      "learning_rate": 5.824603825018904e-05,
      "loss": 0.0091,
      "step": 4750
    },
    {
      "epoch": 119.0,
      "grad_norm": 0.2962054908275604,
      "learning_rate": 5.808291060312975e-05,
      "loss": 0.0124,
      "step": 4760
    },
    {
      "epoch": 119.25,
      "grad_norm": 0.23632937669754028,
      "learning_rate": 5.7919694562655083e-05,
      "loss": 0.0116,
      "step": 4770
    },
    {
      "epoch": 119.5,
      "grad_norm": 0.2798871695995331,
      "learning_rate": 5.775639191366954e-05,
      "loss": 0.0117,
      "step": 4780
    },
    {
      "epoch": 119.75,
      "grad_norm": 0.2946552634239197,
      "learning_rate": 5.75930044420247e-05,
      "loss": 0.0126,
      "step": 4790
    },
    {
      "epoch": 120.0,
      "grad_norm": 0.31767404079437256,
      "learning_rate": 5.74295339344998e-05,
      "loss": 0.0095,
      "step": 4800
    },
    {
      "epoch": 120.25,
      "grad_norm": 0.20685528218746185,
      "learning_rate": 5.726598217878211e-05,
      "loss": 0.0117,
      "step": 4810
    },
    {
      "epoch": 120.5,
      "grad_norm": 0.23895318806171417,
      "learning_rate": 5.71023509634474e-05,
      "loss": 0.0104,
      "step": 4820
    },
    {
      "epoch": 120.75,
      "grad_norm": 0.28225383162498474,
      "learning_rate": 5.693864207794049e-05,
      "loss": 0.008,
      "step": 4830
    },
    {
      "epoch": 121.0,
      "grad_norm": 0.2577228248119354,
      "learning_rate": 5.677485731255545e-05,
      "loss": 0.0118,
      "step": 4840
    },
    {
      "epoch": 121.25,
      "grad_norm": 0.22346603870391846,
      "learning_rate": 5.6610998458416296e-05,
      "loss": 0.0182,
      "step": 4850
    },
    {
      "epoch": 121.5,
      "grad_norm": 0.17444445192813873,
      "learning_rate": 5.644706730745716e-05,
      "loss": 0.0113,
      "step": 4860
    },
    {
      "epoch": 121.75,
      "grad_norm": 0.1593342125415802,
      "learning_rate": 5.628306565240287e-05,
      "loss": 0.011,
      "step": 4870
    },
    {
      "epoch": 122.0,
      "grad_norm": 0.22958680987358093,
      "learning_rate": 5.611899528674923e-05,
      "loss": 0.0097,
      "step": 4880
    },
    {
      "epoch": 122.25,
      "grad_norm": 0.26986101269721985,
      "learning_rate": 5.595485800474349e-05,
      "loss": 0.0117,
      "step": 4890
    },
    {
      "epoch": 122.5,
      "grad_norm": 0.2303800880908966,
      "learning_rate": 5.579065560136467e-05,
      "loss": 0.0096,
      "step": 4900
    },
    {
      "epoch": 122.75,
      "grad_norm": 0.28559139370918274,
      "learning_rate": 5.562638987230392e-05,
      "loss": 0.0105,
      "step": 4910
    },
    {
      "epoch": 123.0,
      "grad_norm": 0.20101706683635712,
      "learning_rate": 5.546206261394498e-05,
      "loss": 0.0083,
      "step": 4920
    },
    {
      "epoch": 123.25,
      "grad_norm": 0.2006942480802536,
      "learning_rate": 5.529767562334437e-05,
      "loss": 0.0082,
      "step": 4930
    },
    {
      "epoch": 123.5,
      "grad_norm": 0.2533215582370758,
      "learning_rate": 5.5133230698211926e-05,
      "loss": 0.0086,
      "step": 4940
    },
    {
      "epoch": 123.75,
      "grad_norm": 0.2552017867565155,
      "learning_rate": 5.496872963689096e-05,
      "loss": 0.0107,
      "step": 4950
    },
    {
      "epoch": 124.0,
      "grad_norm": 0.4189196825027466,
      "learning_rate": 5.4804174238338756e-05,
      "loss": 0.0129,
      "step": 4960
    },
    {
      "epoch": 124.25,
      "grad_norm": 0.23606009781360626,
      "learning_rate": 5.463956630210678e-05,
      "loss": 0.0074,
      "step": 4970
    },
    {
      "epoch": 124.5,
      "grad_norm": 0.25102007389068604,
      "learning_rate": 5.4474907628321046e-05,
      "loss": 0.0115,
      "step": 4980
    },
    {
      "epoch": 124.75,
      "grad_norm": 0.19480489194393158,
      "learning_rate": 5.431020001766244e-05,
      "loss": 0.0123,
      "step": 4990
    },
    {
      "epoch": 125.0,
      "grad_norm": 0.28561678528785706,
      "learning_rate": 5.4145445271346986e-05,
      "loss": 0.0111,
      "step": 5000
    },
    {
      "epoch": 125.25,
      "grad_norm": 0.20620761811733246,
      "learning_rate": 5.398064519110622e-05,
      "loss": 0.0102,
      "step": 5010
    },
    {
      "epoch": 125.5,
      "grad_norm": 0.33209967613220215,
      "learning_rate": 5.3815801579167394e-05,
      "loss": 0.0131,
      "step": 5020
    },
    {
      "epoch": 125.75,
      "grad_norm": 0.23769836127758026,
      "learning_rate": 5.365091623823382e-05,
      "loss": 0.0093,
      "step": 5030
    },
    {
      "epoch": 126.0,
      "grad_norm": 0.23213163018226624,
      "learning_rate": 5.348599097146521e-05,
      "loss": 0.009,
      "step": 5040
    },
    {
      "epoch": 126.25,
      "grad_norm": 0.1278630793094635,
      "learning_rate": 5.3321027582457836e-05,
      "loss": 0.0086,
      "step": 5050
    },
    {
      "epoch": 126.5,
      "grad_norm": 0.2442750185728073,
      "learning_rate": 5.315602787522491e-05,
      "loss": 0.0117,
      "step": 5060
    },
    {
      "epoch": 126.75,
      "grad_norm": 0.19907784461975098,
      "learning_rate": 5.299099365417678e-05,
      "loss": 0.0085,
      "step": 5070
    },
    {
      "epoch": 127.0,
      "grad_norm": 0.21254177391529083,
      "learning_rate": 5.2825926724101236e-05,
      "loss": 0.007,
      "step": 5080
    },
    {
      "epoch": 127.25,
      "grad_norm": 0.12308558821678162,
      "learning_rate": 5.26608288901438e-05,
      "loss": 0.0079,
      "step": 5090
    },
    {
      "epoch": 127.5,
      "grad_norm": 0.12892907857894897,
      "learning_rate": 5.24957019577879e-05,
      "loss": 0.0091,
      "step": 5100
    },
    {
      "epoch": 127.75,
      "grad_norm": 0.177867129445076,
      "learning_rate": 5.2330547732835266e-05,
      "loss": 0.0129,
      "step": 5110
    },
    {
      "epoch": 128.0,
      "grad_norm": 0.19430725276470184,
      "learning_rate": 5.2165368021385996e-05,
      "loss": 0.0088,
      "step": 5120
    },
    {
      "epoch": 128.25,
      "grad_norm": 0.21545594930648804,
      "learning_rate": 5.200016462981897e-05,
      "loss": 0.0146,
      "step": 5130
    },
    {
      "epoch": 128.5,
      "grad_norm": 0.148879274725914,
      "learning_rate": 5.1834939364772015e-05,
      "loss": 0.0075,
      "step": 5140
    },
    {
      "epoch": 128.75,
      "grad_norm": 0.2902442514896393,
      "learning_rate": 5.166969403312214e-05,
      "loss": 0.0097,
      "step": 5150
    },
    {
      "epoch": 129.0,
      "grad_norm": 0.23031367361545563,
      "learning_rate": 5.1504430441965844e-05,
      "loss": 0.0075,
      "step": 5160
    },
    {
      "epoch": 129.25,
      "grad_norm": 0.13191789388656616,
      "learning_rate": 5.133915039859923e-05,
      "loss": 0.0091,
      "step": 5170
    },
    {
      "epoch": 129.5,
      "grad_norm": 0.15518267452716827,
      "learning_rate": 5.1173855710498444e-05,
      "loss": 0.0064,
      "step": 5180
    },
    {
      "epoch": 129.75,
      "grad_norm": 0.21780794858932495,
      "learning_rate": 5.100854818529967e-05,
      "loss": 0.0081,
      "step": 5190
    },
    {
      "epoch": 130.0,
      "grad_norm": 0.17131520807743073,
      "learning_rate": 5.084322963077951e-05,
      "loss": 0.0118,
      "step": 5200
    },
    {
      "epoch": 130.25,
      "grad_norm": 0.27276504039764404,
      "learning_rate": 5.067790185483522e-05,
      "loss": 0.012,
      "step": 5210
    },
    {
      "epoch": 130.5,
      "grad_norm": 0.22213535010814667,
      "learning_rate": 5.0512566665464844e-05,
      "loss": 0.0095,
      "step": 5220
    },
    {
      "epoch": 130.75,
      "grad_norm": 0.3615487217903137,
      "learning_rate": 5.034722587074755e-05,
      "loss": 0.0115,
      "step": 5230
    },
    {
      "epoch": 131.0,
      "grad_norm": 0.1875467300415039,
      "learning_rate": 5.018188127882375e-05,
      "loss": 0.0096,
      "step": 5240
    },
    {
      "epoch": 131.25,
      "grad_norm": 0.3611842393875122,
      "learning_rate": 5.0016534697875417e-05,
      "loss": 0.0071,
      "step": 5250
    },
    {
      "epoch": 131.5,
      "grad_norm": 0.1838546097278595,
      "learning_rate": 4.9851187936106294e-05,
      "loss": 0.0079,
      "step": 5260
    },
    {
      "epoch": 131.75,
      "grad_norm": 0.3148566782474518,
      "learning_rate": 4.968584280172206e-05,
      "loss": 0.0099,
      "step": 5270
    },
    {
      "epoch": 132.0,
      "grad_norm": 0.14675936102867126,
      "learning_rate": 4.95205011029106e-05,
      "loss": 0.0089,
      "step": 5280
    },
    {
      "epoch": 132.25,
      "grad_norm": 0.4316335916519165,
      "learning_rate": 4.935516464782227e-05,
      "loss": 0.0089,
      "step": 5290
    },
    {
      "epoch": 132.5,
      "grad_norm": 0.19409996271133423,
      "learning_rate": 4.918983524455003e-05,
      "loss": 0.0087,
      "step": 5300
    },
    {
      "epoch": 132.75,
      "grad_norm": 0.1797177940607071,
      "learning_rate": 4.9024514701109766e-05,
      "loss": 0.0078,
      "step": 5310
    },
    {
      "epoch": 133.0,
      "grad_norm": 0.16304469108581543,
      "learning_rate": 4.885920482542043e-05,
      "loss": 0.0098,
      "step": 5320
    },
    {
      "epoch": 133.25,
      "grad_norm": 0.18975849449634552,
      "learning_rate": 4.869390742528438e-05,
      "loss": 0.0086,
      "step": 5330
    },
    {
      "epoch": 133.5,
      "grad_norm": 0.1938469260931015,
      "learning_rate": 4.852862430836744e-05,
      "loss": 0.0073,
      "step": 5340
    },
    {
      "epoch": 133.75,
      "grad_norm": 0.18156132102012634,
      "learning_rate": 4.836335728217933e-05,
      "loss": 0.0087,
      "step": 5350
    },
    {
      "epoch": 134.0,
      "grad_norm": 0.19061091542243958,
      "learning_rate": 4.819810815405379e-05,
      "loss": 0.0087,
      "step": 5360
    },
    {
      "epoch": 134.25,
      "grad_norm": 0.18980200588703156,
      "learning_rate": 4.803287873112877e-05,
      "loss": 0.0105,
      "step": 5370
    },
    {
      "epoch": 134.5,
      "grad_norm": 0.18439440429210663,
      "learning_rate": 4.786767082032681e-05,
      "loss": 0.0105,
      "step": 5380
    },
    {
      "epoch": 134.75,
      "grad_norm": 0.22925539314746857,
      "learning_rate": 4.77024862283351e-05,
      "loss": 0.0099,
      "step": 5390
    },
    {
      "epoch": 135.0,
      "grad_norm": 0.23364628851413727,
      "learning_rate": 4.753732676158593e-05,
      "loss": 0.0073,
      "step": 5400
    },
    {
      "epoch": 135.25,
      "grad_norm": 0.3074020743370056,
      "learning_rate": 4.737219422623672e-05,
      "loss": 0.0078,
      "step": 5410
    },
    {
      "epoch": 135.5,
      "grad_norm": 0.24079379439353943,
      "learning_rate": 4.720709042815044e-05,
      "loss": 0.0084,
      "step": 5420
    },
    {
      "epoch": 135.75,
      "grad_norm": 0.28063663840293884,
      "learning_rate": 4.704201717287578e-05,
      "loss": 0.0064,
      "step": 5430
    },
    {
      "epoch": 136.0,
      "grad_norm": 0.40149784088134766,
      "learning_rate": 4.6876976265627404e-05,
      "loss": 0.0073,
      "step": 5440
    },
    {
      "epoch": 136.25,
      "grad_norm": 0.17972993850708008,
      "learning_rate": 4.671196951126626e-05,
      "loss": 0.0073,
      "step": 5450
    },
    {
      "epoch": 136.5,
      "grad_norm": 0.32252639532089233,
      "learning_rate": 4.654699871427971e-05,
      "loss": 0.0127,
      "step": 5460
    },
    {
      "epoch": 136.75,
      "grad_norm": 0.17494434118270874,
      "learning_rate": 4.6382065678762034e-05,
      "loss": 0.009,
      "step": 5470
    },
    {
      "epoch": 137.0,
      "grad_norm": 0.144280344247818,
      "learning_rate": 4.6217172208394424e-05,
      "loss": 0.0103,
      "step": 5480
    },
    {
      "epoch": 137.25,
      "grad_norm": 0.20007732510566711,
      "learning_rate": 4.605232010642549e-05,
      "loss": 0.0056,
      "step": 5490
    },
    {
      "epoch": 137.5,
      "grad_norm": 0.21483971178531647,
      "learning_rate": 4.588751117565142e-05,
      "loss": 0.008,
      "step": 5500
    },
    {
      "epoch": 137.75,
      "grad_norm": 0.18367771804332733,
      "learning_rate": 4.5722747218396214e-05,
      "loss": 0.008,
      "step": 5510
    },
    {
      "epoch": 138.0,
      "grad_norm": 0.16514046490192413,
      "learning_rate": 4.5558030036492194e-05,
      "loss": 0.0081,
      "step": 5520
    },
    {
      "epoch": 138.25,
      "grad_norm": 0.14982137084007263,
      "learning_rate": 4.539336143125999e-05,
      "loss": 0.0058,
      "step": 5530
    },
    {
      "epoch": 138.5,
      "grad_norm": 0.10444972664117813,
      "learning_rate": 4.522874320348916e-05,
      "loss": 0.007,
      "step": 5540
    },
    {
      "epoch": 138.75,
      "grad_norm": 0.1786758452653885,
      "learning_rate": 4.506417715341821e-05,
      "loss": 0.0138,
      "step": 5550
    },
    {
      "epoch": 139.0,
      "grad_norm": 0.1584421694278717,
      "learning_rate": 4.489966508071511e-05,
      "loss": 0.0065,
      "step": 5560
    },
    {
      "epoch": 139.25,
      "grad_norm": 0.20712852478027344,
      "learning_rate": 4.4735208784457575e-05,
      "loss": 0.0083,
      "step": 5570
    },
    {
      "epoch": 139.5,
      "grad_norm": 0.1956453025341034,
      "learning_rate": 4.457081006311325e-05,
      "loss": 0.0073,
      "step": 5580
    },
    {
      "epoch": 139.75,
      "grad_norm": 0.15756206214427948,
      "learning_rate": 4.440647071452027e-05,
      "loss": 0.0073,
      "step": 5590
    },
    {
      "epoch": 140.0,
      "grad_norm": 0.2822488844394684,
      "learning_rate": 4.424219253586737e-05,
      "loss": 0.0101,
      "step": 5600
    },
    {
      "epoch": 140.25,
      "grad_norm": 0.2170170098543167,
      "learning_rate": 4.407797732367443e-05,
      "loss": 0.0085,
      "step": 5610
    },
    {
      "epoch": 140.5,
      "grad_norm": 0.14802826941013336,
      "learning_rate": 4.391382687377268e-05,
      "loss": 0.0094,
      "step": 5620
    },
    {
      "epoch": 140.75,
      "grad_norm": 0.19422343373298645,
      "learning_rate": 4.374974298128512e-05,
      "loss": 0.0084,
      "step": 5630
    },
    {
      "epoch": 141.0,
      "grad_norm": 0.1759997010231018,
      "learning_rate": 4.358572744060699e-05,
      "loss": 0.008,
      "step": 5640
    },
    {
      "epoch": 141.25,
      "grad_norm": 0.20682178437709808,
      "learning_rate": 4.342178204538588e-05,
      "loss": 0.0101,
      "step": 5650
    },
    {
      "epoch": 141.5,
      "grad_norm": 0.14558851718902588,
      "learning_rate": 4.325790858850241e-05,
      "loss": 0.0106,
      "step": 5660
    },
    {
      "epoch": 141.75,
      "grad_norm": 0.11343575268983841,
      "learning_rate": 4.309410886205043e-05,
      "loss": 0.0075,
      "step": 5670
    },
    {
      "epoch": 142.0,
      "grad_norm": 0.21630960702896118,
      "learning_rate": 4.293038465731752e-05,
      "loss": 0.0088,
      "step": 5680
    },
    {
      "epoch": 142.25,
      "grad_norm": 0.1380923092365265,
      "learning_rate": 4.276673776476533e-05,
      "loss": 0.0064,
      "step": 5690
    },
    {
      "epoch": 142.5,
      "grad_norm": 0.2421133816242218,
      "learning_rate": 4.260316997401007e-05,
      "loss": 0.0097,
      "step": 5700
    },
    {
      "epoch": 142.75,
      "grad_norm": 0.21211817860603333,
      "learning_rate": 4.243968307380293e-05,
      "loss": 0.0093,
      "step": 5710
    },
    {
      "epoch": 143.0,
      "grad_norm": 0.1438850462436676,
      "learning_rate": 4.22762788520104e-05,
      "loss": 0.0092,
      "step": 5720
    },
    {
      "epoch": 143.25,
      "grad_norm": 0.10566523671150208,
      "learning_rate": 4.211295909559491e-05,
      "loss": 0.0095,
      "step": 5730
    },
    {
      "epoch": 143.5,
      "grad_norm": 0.13774773478507996,
      "learning_rate": 4.194972559059511e-05,
      "loss": 0.0103,
      "step": 5740
    },
    {
      "epoch": 143.75,
      "grad_norm": 0.2047201693058014,
      "learning_rate": 4.178658012210651e-05,
      "loss": 0.0085,
      "step": 5750
    },
    {
      "epoch": 144.0,
      "grad_norm": 0.09229153394699097,
      "learning_rate": 4.162352447426177e-05,
      "loss": 0.0069,
      "step": 5760
    },
    {
      "epoch": 144.25,
      "grad_norm": 0.2979641556739807,
      "learning_rate": 4.146056043021135e-05,
      "loss": 0.0075,
      "step": 5770
    },
    {
      "epoch": 144.5,
      "grad_norm": 0.20068320631980896,
      "learning_rate": 4.1297689772103944e-05,
      "loss": 0.0082,
      "step": 5780
    },
    {
      "epoch": 144.75,
      "grad_norm": 0.1853872686624527,
      "learning_rate": 4.113491428106694e-05,
      "loss": 0.0085,
      "step": 5790
    },
    {
      "epoch": 145.0,
      "grad_norm": 0.30404505133628845,
      "learning_rate": 4.0972235737187055e-05,
      "loss": 0.0102,
      "step": 5800
    },
    {
      "epoch": 145.25,
      "grad_norm": 0.20279182493686676,
      "learning_rate": 4.080965591949076e-05,
      "loss": 0.0075,
      "step": 5810
    },
    {
      "epoch": 145.5,
      "grad_norm": 0.2873016595840454,
      "learning_rate": 4.0647176605924924e-05,
      "loss": 0.0055,
      "step": 5820
    },
    {
      "epoch": 145.75,
      "grad_norm": 0.2151174396276474,
      "learning_rate": 4.0484799573337255e-05,
      "loss": 0.0089,
      "step": 5830
    },
    {
      "epoch": 146.0,
      "grad_norm": 0.13774517178535461,
      "learning_rate": 4.032252659745699e-05,
      "loss": 0.0104,
      "step": 5840
    },
    {
      "epoch": 146.25,
      "grad_norm": 0.31033048033714294,
      "learning_rate": 4.016035945287539e-05,
      "loss": 0.0081,
      "step": 5850
    },
    {
      "epoch": 146.5,
      "grad_norm": 0.18268409371376038,
      "learning_rate": 3.999829991302635e-05,
      "loss": 0.0102,
      "step": 5860
    },
    {
      "epoch": 146.75,
      "grad_norm": 0.1759636551141739,
      "learning_rate": 3.983634975016707e-05,
      "loss": 0.0131,
      "step": 5870
    },
    {
      "epoch": 147.0,
      "grad_norm": 0.12363249808549881,
      "learning_rate": 3.967451073535854e-05,
      "loss": 0.0073,
      "step": 5880
    },
    {
      "epoch": 147.25,
      "grad_norm": 0.13208454847335815,
      "learning_rate": 3.951278463844633e-05,
      "loss": 0.0116,
      "step": 5890
    },
    {
      "epoch": 147.5,
      "grad_norm": 0.25533878803253174,
      "learning_rate": 3.935117322804111e-05,
      "loss": 0.0109,
      "step": 5900
    },
    {
      "epoch": 147.75,
      "grad_norm": 0.1883186548948288,
      "learning_rate": 3.918967827149938e-05,
      "loss": 0.007,
      "step": 5910
    },
    {
      "epoch": 148.0,
      "grad_norm": 0.23646943271160126,
      "learning_rate": 3.9028301534904094e-05,
      "loss": 0.008,
      "step": 5920
    },
    {
      "epoch": 148.25,
      "grad_norm": 0.19803385436534882,
      "learning_rate": 3.88670447830454e-05,
      "loss": 0.0076,
      "step": 5930
    },
    {
      "epoch": 148.5,
      "grad_norm": 0.09101427346467972,
      "learning_rate": 3.870590977940132e-05,
      "loss": 0.0073,
      "step": 5940
    },
    {
      "epoch": 148.75,
      "grad_norm": 0.19984644651412964,
      "learning_rate": 3.8544898286118404e-05,
      "loss": 0.0068,
      "step": 5950
    },
    {
      "epoch": 149.0,
      "grad_norm": 0.1640167087316513,
      "learning_rate": 3.838401206399257e-05,
      "loss": 0.006,
      "step": 5960
    },
    {
      "epoch": 149.25,
      "grad_norm": 0.18579047918319702,
      "learning_rate": 3.822325287244975e-05,
      "loss": 0.0053,
      "step": 5970
    },
    {
      "epoch": 149.5,
      "grad_norm": 0.152107372879982,
      "learning_rate": 3.8062622469526725e-05,
      "loss": 0.0064,
      "step": 5980
    },
    {
      "epoch": 149.75,
      "grad_norm": 0.16540779173374176,
      "learning_rate": 3.790212261185183e-05,
      "loss": 0.007,
      "step": 5990
    },
    {
      "epoch": 150.0,
      "grad_norm": 0.15848080813884735,
      "learning_rate": 3.7741755054625794e-05,
      "loss": 0.0062,
      "step": 6000
    },
    {
      "epoch": 150.25,
      "grad_norm": 0.21890725195407867,
      "learning_rate": 3.758152155160255e-05,
      "loss": 0.009,
      "step": 6010
    },
    {
      "epoch": 150.5,
      "grad_norm": 0.14560624957084656,
      "learning_rate": 3.742142385506999e-05,
      "loss": 0.0059,
      "step": 6020
    },
    {
      "epoch": 150.75,
      "grad_norm": 0.16762469708919525,
      "learning_rate": 3.72614637158309e-05,
      "loss": 0.0082,
      "step": 6030
    },
    {
      "epoch": 151.0,
      "grad_norm": 0.17747336626052856,
      "learning_rate": 3.710164288318371e-05,
      "loss": 0.0076,
      "step": 6040
    },
    {
      "epoch": 151.25,
      "grad_norm": 0.11070247739553452,
      "learning_rate": 3.694196310490345e-05,
      "loss": 0.008,
      "step": 6050
    },
    {
      "epoch": 151.5,
      "grad_norm": 0.17882029712200165,
      "learning_rate": 3.678242612722259e-05,
      "loss": 0.0068,
      "step": 6060
    },
    {
      "epoch": 151.75,
      "grad_norm": 0.235889732837677,
      "learning_rate": 3.6623033694811953e-05,
      "loss": 0.0064,
      "step": 6070
    },
    {
      "epoch": 152.0,
      "grad_norm": 0.31911328434944153,
      "learning_rate": 3.6463787550761665e-05,
      "loss": 0.0078,
      "step": 6080
    },
    {
      "epoch": 152.25,
      "grad_norm": 0.1889229267835617,
      "learning_rate": 3.630468943656202e-05,
      "loss": 0.0089,
      "step": 6090
    },
    {
      "epoch": 152.5,
      "grad_norm": 0.27575376629829407,
      "learning_rate": 3.6145741092084523e-05,
      "loss": 0.0087,
      "step": 6100
    },
    {
      "epoch": 152.75,
      "grad_norm": 0.1767558604478836,
      "learning_rate": 3.598694425556278e-05,
      "loss": 0.0063,
      "step": 6110
    },
    {
      "epoch": 153.0,
      "grad_norm": 0.15844060480594635,
      "learning_rate": 3.58283006635736e-05,
      "loss": 0.0069,
      "step": 6120
    },
    {
      "epoch": 153.25,
      "grad_norm": 0.13357259333133698,
      "learning_rate": 3.566981205101781e-05,
      "loss": 0.0069,
      "step": 6130
    },
    {
      "epoch": 153.5,
      "grad_norm": 0.1877880096435547,
      "learning_rate": 3.5511480151101556e-05,
      "loss": 0.0073,
      "step": 6140
    },
    {
      "epoch": 153.75,
      "grad_norm": 0.2021484524011612,
      "learning_rate": 3.5353306695317104e-05,
      "loss": 0.0071,
      "step": 6150
    },
    {
      "epoch": 154.0,
      "grad_norm": 0.1393587738275528,
      "learning_rate": 3.519529341342402e-05,
      "loss": 0.0067,
      "step": 6160
    },
    {
      "epoch": 154.25,
      "grad_norm": 0.19375017285346985,
      "learning_rate": 3.503744203343026e-05,
      "loss": 0.0063,
      "step": 6170
    },
    {
      "epoch": 154.5,
      "grad_norm": 0.12254150956869125,
      "learning_rate": 3.487975428157318e-05,
      "loss": 0.0066,
      "step": 6180
    },
    {
      "epoch": 154.75,
      "grad_norm": 0.1326659768819809,
      "learning_rate": 3.472223188230083e-05,
      "loss": 0.0077,
      "step": 6190
    },
    {
      "epoch": 155.0,
      "grad_norm": 0.1555151343345642,
      "learning_rate": 3.4564876558252866e-05,
      "loss": 0.0066,
      "step": 6200
    },
    {
      "epoch": 155.25,
      "grad_norm": 0.29415765404701233,
      "learning_rate": 3.440769003024195e-05,
      "loss": 0.0087,
      "step": 6210
    },
    {
      "epoch": 155.5,
      "grad_norm": 0.20861037075519562,
      "learning_rate": 3.425067401723477e-05,
      "loss": 0.0065,
      "step": 6220
    },
    {
      "epoch": 155.75,
      "grad_norm": 0.22378014028072357,
      "learning_rate": 3.409383023633325e-05,
      "loss": 0.0061,
      "step": 6230
    },
    {
      "epoch": 156.0,
      "grad_norm": 0.1030755341053009,
      "learning_rate": 3.3937160402755894e-05,
      "loss": 0.0066,
      "step": 6240
    },
    {
      "epoch": 156.25,
      "grad_norm": 0.15028028190135956,
      "learning_rate": 3.378066622981885e-05,
      "loss": 0.0058,
      "step": 6250
    },
    {
      "epoch": 156.5,
      "grad_norm": 0.21049033105373383,
      "learning_rate": 3.362434942891738e-05,
      "loss": 0.0074,
      "step": 6260
    },
    {
      "epoch": 156.75,
      "grad_norm": 0.22261491417884827,
      "learning_rate": 3.346821170950693e-05,
      "loss": 0.0087,
      "step": 6270
    },
    {
      "epoch": 157.0,
      "grad_norm": 0.15786141157150269,
      "learning_rate": 3.3312254779084585e-05,
      "loss": 0.0084,
      "step": 6280
    },
    {
      "epoch": 157.25,
      "grad_norm": 0.17248444259166718,
      "learning_rate": 3.315648034317039e-05,
      "loss": 0.0068,
      "step": 6290
    },
    {
      "epoch": 157.5,
      "grad_norm": 0.22754375636577606,
      "learning_rate": 3.3000890105288564e-05,
      "loss": 0.0088,
      "step": 6300
    },
    {
      "epoch": 157.75,
      "grad_norm": 0.22142088413238525,
      "learning_rate": 3.284548576694908e-05,
      "loss": 0.0062,
      "step": 6310
    },
    {
      "epoch": 158.0,
      "grad_norm": 0.17295658588409424,
      "learning_rate": 3.2690269027628815e-05,
      "loss": 0.0067,
      "step": 6320
    },
    {
      "epoch": 158.25,
      "grad_norm": 0.17775364220142365,
      "learning_rate": 3.253524158475324e-05,
      "loss": 0.0053,
      "step": 6330
    },
    {
      "epoch": 158.5,
      "grad_norm": 0.1789032220840454,
      "learning_rate": 3.238040513367757e-05,
      "loss": 0.0062,
      "step": 6340
    },
    {
      "epoch": 158.75,
      "grad_norm": 0.21051982045173645,
      "learning_rate": 3.222576136766843e-05,
      "loss": 0.008,
      "step": 6350
    },
    {
      "epoch": 159.0,
      "grad_norm": 0.2447560429573059,
      "learning_rate": 3.2071311977885324e-05,
      "loss": 0.0069,
      "step": 6360
    },
    {
      "epoch": 159.25,
      "grad_norm": 0.19626660645008087,
      "learning_rate": 3.191705865336197e-05,
      "loss": 0.0061,
      "step": 6370
    },
    {
      "epoch": 159.5,
      "grad_norm": 0.1637764275074005,
      "learning_rate": 3.1763003080988075e-05,
      "loss": 0.0083,
      "step": 6380
    },
    {
      "epoch": 159.75,
      "grad_norm": 0.15946759283542633,
      "learning_rate": 3.160914694549063e-05,
      "loss": 0.0065,
      "step": 6390
    },
    {
      "epoch": 160.0,
      "grad_norm": 0.17754492163658142,
      "learning_rate": 3.145549192941573e-05,
      "loss": 0.0071,
      "step": 6400
    },
    {
      "epoch": 160.25,
      "grad_norm": 0.1859520822763443,
      "learning_rate": 3.130203971310999e-05,
      "loss": 0.0055,
      "step": 6410
    },
    {
      "epoch": 160.5,
      "grad_norm": 0.15614630281925201,
      "learning_rate": 3.114879197470225e-05,
      "loss": 0.0064,
      "step": 6420
    },
    {
      "epoch": 160.75,
      "grad_norm": 0.17567554116249084,
      "learning_rate": 3.0995750390085285e-05,
      "loss": 0.0051,
      "step": 6430
    },
    {
      "epoch": 161.0,
      "grad_norm": 0.15353727340698242,
      "learning_rate": 3.084291663289728e-05,
      "loss": 0.007,
      "step": 6440
    },
    {
      "epoch": 161.25,
      "grad_norm": 0.16327092051506042,
      "learning_rate": 3.069029237450375e-05,
      "loss": 0.0055,
      "step": 6450
    },
    {
      "epoch": 161.5,
      "grad_norm": 0.13596777617931366,
      "learning_rate": 3.053787928397911e-05,
      "loss": 0.0044,
      "step": 6460
    },
    {
      "epoch": 161.75,
      "grad_norm": 0.12858083844184875,
      "learning_rate": 3.0385679028088526e-05,
      "loss": 0.005,
      "step": 6470
    },
    {
      "epoch": 162.0,
      "grad_norm": 0.15615972876548767,
      "learning_rate": 3.023369327126959e-05,
      "loss": 0.0062,
      "step": 6480
    },
    {
      "epoch": 162.25,
      "grad_norm": 0.3536645472049713,
      "learning_rate": 3.0081923675614198e-05,
      "loss": 0.0068,
      "step": 6490
    },
    {
      "epoch": 162.5,
      "grad_norm": 0.12195973843336105,
      "learning_rate": 2.993037190085034e-05,
      "loss": 0.0049,
      "step": 6500
    },
    {
      "epoch": 162.75,
      "grad_norm": 0.20225834846496582,
      "learning_rate": 2.977903960432392e-05,
      "loss": 0.0066,
      "step": 6510
    },
    {
      "epoch": 163.0,
      "grad_norm": 0.17703953385353088,
      "learning_rate": 2.9627928440980722e-05,
      "loss": 0.0057,
      "step": 6520
    },
    {
      "epoch": 163.25,
      "grad_norm": 0.13128487765789032,
      "learning_rate": 2.9477040063348183e-05,
      "loss": 0.0051,
      "step": 6530
    },
    {
      "epoch": 163.5,
      "grad_norm": 0.23100434243679047,
      "learning_rate": 2.9326376121517456e-05,
      "loss": 0.0084,
      "step": 6540
    },
    {
      "epoch": 163.75,
      "grad_norm": 0.1476689726114273,
      "learning_rate": 2.9175938263125236e-05,
      "loss": 0.0062,
      "step": 6550
    },
    {
      "epoch": 164.0,
      "grad_norm": 0.16052044928073883,
      "learning_rate": 2.9025728133335873e-05,
      "loss": 0.0057,
      "step": 6560
    },
    {
      "epoch": 164.25,
      "grad_norm": 0.1788512021303177,
      "learning_rate": 2.8875747374823288e-05,
      "loss": 0.0073,
      "step": 6570
    },
    {
      "epoch": 164.5,
      "grad_norm": 0.1804950088262558,
      "learning_rate": 2.872599762775298e-05,
      "loss": 0.0066,
      "step": 6580
    },
    {
      "epoch": 164.75,
      "grad_norm": 0.22115811705589294,
      "learning_rate": 2.857648052976425e-05,
      "loss": 0.0105,
      "step": 6590
    },
    {
      "epoch": 165.0,
      "grad_norm": 0.19970916211605072,
      "learning_rate": 2.8427197715952047e-05,
      "loss": 0.0064,
      "step": 6600
    },
    {
      "epoch": 165.25,
      "grad_norm": 0.16419823467731476,
      "learning_rate": 2.8278150818849393e-05,
      "loss": 0.0059,
      "step": 6610
    },
    {
      "epoch": 165.5,
      "grad_norm": 0.2086798995733261,
      "learning_rate": 2.812934146840922e-05,
      "loss": 0.0058,
      "step": 6620
    },
    {
      "epoch": 165.75,
      "grad_norm": 0.18059085309505463,
      "learning_rate": 2.7980771291986764e-05,
      "loss": 0.0064,
      "step": 6630
    },
    {
      "epoch": 166.0,
      "grad_norm": 0.2674952745437622,
      "learning_rate": 2.783244191432167e-05,
      "loss": 0.0062,
      "step": 6640
    },
    {
      "epoch": 166.25,
      "grad_norm": 0.1242055743932724,
      "learning_rate": 2.768435495752022e-05,
      "loss": 0.0071,
      "step": 6650
    },
    {
      "epoch": 166.5,
      "grad_norm": 0.10322605073451996,
      "learning_rate": 2.753651204103771e-05,
      "loss": 0.005,
      "step": 6660
    },
    {
      "epoch": 166.75,
      "grad_norm": 0.13184568285942078,
      "learning_rate": 2.7388914781660523e-05,
      "loss": 0.006,
      "step": 6670
    },
    {
      "epoch": 167.0,
      "grad_norm": 0.08305729925632477,
      "learning_rate": 2.7241564793488693e-05,
      "loss": 0.0043,
      "step": 6680
    },
    {
      "epoch": 167.25,
      "grad_norm": 0.193691223859787,
      "learning_rate": 2.7094463687918037e-05,
      "loss": 0.007,
      "step": 6690
    },
    {
      "epoch": 167.5,
      "grad_norm": 0.11765813082456589,
      "learning_rate": 2.694761307362268e-05,
      "loss": 0.0049,
      "step": 6700
    },
    {
      "epoch": 167.75,
      "grad_norm": 0.12011758238077164,
      "learning_rate": 2.6801014556537467e-05,
      "loss": 0.0061,
      "step": 6710
    },
    {
      "epoch": 168.0,
      "grad_norm": 0.14065556228160858,
      "learning_rate": 2.6654669739840243e-05,
      "loss": 0.0049,
      "step": 6720
    },
    {
      "epoch": 168.25,
      "grad_norm": 0.15679895877838135,
      "learning_rate": 2.650858022393451e-05,
      "loss": 0.0053,
      "step": 6730
    },
    {
      "epoch": 168.5,
      "grad_norm": 0.1384173333644867,
      "learning_rate": 2.6362747606431747e-05,
      "loss": 0.0051,
      "step": 6740
    },
    {
      "epoch": 168.75,
      "grad_norm": 0.12798947095870972,
      "learning_rate": 2.6217173482134172e-05,
      "loss": 0.0064,
      "step": 6750
    },
    {
      "epoch": 169.0,
      "grad_norm": 0.1402834951877594,
      "learning_rate": 2.6071859443017044e-05,
      "loss": 0.0074,
      "step": 6760
    },
    {
      "epoch": 169.25,
      "grad_norm": 0.1125163584947586,
      "learning_rate": 2.5926807078211414e-05,
      "loss": 0.0045,
      "step": 6770
    },
    {
      "epoch": 169.5,
      "grad_norm": 0.12739096581935883,
      "learning_rate": 2.5782017973986728e-05,
      "loss": 0.0056,
      "step": 6780
    },
    {
      "epoch": 169.75,
      "grad_norm": 0.11030346900224686,
      "learning_rate": 2.5637493713733374e-05,
      "loss": 0.0039,
      "step": 6790
    },
    {
      "epoch": 170.0,
      "grad_norm": 0.1563323736190796,
      "learning_rate": 2.549323587794559e-05,
      "loss": 0.0056,
      "step": 6800
    },
    {
      "epoch": 170.25,
      "grad_norm": 0.1213504895567894,
      "learning_rate": 2.5349246044203895e-05,
      "loss": 0.0046,
      "step": 6810
    },
    {
      "epoch": 170.5,
      "grad_norm": 0.17844437062740326,
      "learning_rate": 2.520552578715808e-05,
      "loss": 0.0078,
      "step": 6820
    },
    {
      "epoch": 170.75,
      "grad_norm": 0.16773608326911926,
      "learning_rate": 2.506207667850981e-05,
      "loss": 0.007,
      "step": 6830
    },
    {
      "epoch": 171.0,
      "grad_norm": 0.09557987004518509,
      "learning_rate": 2.4918900286995555e-05,
      "loss": 0.0047,
      "step": 6840
    },
    {
      "epoch": 171.25,
      "grad_norm": 0.19044405221939087,
      "learning_rate": 2.4775998178369458e-05,
      "loss": 0.0042,
      "step": 6850
    },
    {
      "epoch": 171.5,
      "grad_norm": 0.16859029233455658,
      "learning_rate": 2.4633371915386017e-05,
      "loss": 0.0069,
      "step": 6860
    },
    {
      "epoch": 171.75,
      "grad_norm": 0.20419123768806458,
      "learning_rate": 2.4491023057783235e-05,
      "loss": 0.0063,
      "step": 6870
    },
    {
      "epoch": 172.0,
      "grad_norm": 0.1309284120798111,
      "learning_rate": 2.4348953162265375e-05,
      "loss": 0.0065,
      "step": 6880
    },
    {
      "epoch": 172.25,
      "grad_norm": 0.20550362765789032,
      "learning_rate": 2.420716378248607e-05,
      "loss": 0.0079,
      "step": 6890
    },
    {
      "epoch": 172.5,
      "grad_norm": 0.2013462632894516,
      "learning_rate": 2.4065656469031266e-05,
      "loss": 0.0075,
      "step": 6900
    },
    {
      "epoch": 172.75,
      "grad_norm": 0.14772872626781464,
      "learning_rate": 2.3924432769402268e-05,
      "loss": 0.0055,
      "step": 6910
    },
    {
      "epoch": 173.0,
      "grad_norm": 0.1719931662082672,
      "learning_rate": 2.3783494227998844e-05,
      "loss": 0.0061,
      "step": 6920
    },
    {
      "epoch": 173.25,
      "grad_norm": 0.11645688116550446,
      "learning_rate": 2.3642842386102264e-05,
      "loss": 0.0056,
      "step": 6930
    },
    {
      "epoch": 173.5,
      "grad_norm": 0.1909249871969223,
      "learning_rate": 2.3502478781858567e-05,
      "loss": 0.0085,
      "step": 6940
    },
    {
      "epoch": 173.75,
      "grad_norm": 0.23779629170894623,
      "learning_rate": 2.3362404950261628e-05,
      "loss": 0.0052,
      "step": 6950
    },
    {
      "epoch": 174.0,
      "grad_norm": 0.161439910531044,
      "learning_rate": 2.3222622423136458e-05,
      "loss": 0.0055,
      "step": 6960
    },
    {
      "epoch": 174.25,
      "grad_norm": 0.15141646564006805,
      "learning_rate": 2.3083132729122332e-05,
      "loss": 0.0061,
      "step": 6970
    },
    {
      "epoch": 174.5,
      "grad_norm": 0.09013067185878754,
      "learning_rate": 2.294393739365621e-05,
      "loss": 0.0046,
      "step": 6980
    },
    {
      "epoch": 174.75,
      "grad_norm": 0.10369148850440979,
      "learning_rate": 2.2805037938956e-05,
      "loss": 0.0077,
      "step": 6990
    },
    {
      "epoch": 175.0,
      "grad_norm": 0.13797912001609802,
      "learning_rate": 2.266643588400386e-05,
      "loss": 0.0064,
      "step": 7000
    },
    {
      "epoch": 175.25,
      "grad_norm": 0.09309276193380356,
      "learning_rate": 2.252813274452969e-05,
      "loss": 0.007,
      "step": 7010
    },
    {
      "epoch": 175.5,
      "grad_norm": 0.15403063595294952,
      "learning_rate": 2.2390130032994427e-05,
      "loss": 0.0048,
      "step": 7020
    },
    {
      "epoch": 175.75,
      "grad_norm": 0.1723039150238037,
      "learning_rate": 2.2252429258573633e-05,
      "loss": 0.0049,
      "step": 7030
    },
    {
      "epoch": 176.0,
      "grad_norm": 0.22523076832294464,
      "learning_rate": 2.2115031927140904e-05,
      "loss": 0.0039,
      "step": 7040
    },
    {
      "epoch": 176.25,
      "grad_norm": 0.1994743049144745,
      "learning_rate": 2.1977939541251463e-05,
      "loss": 0.0051,
      "step": 7050
    },
    {
      "epoch": 176.5,
      "grad_norm": 0.21922245621681213,
      "learning_rate": 2.1841153600125684e-05,
      "loss": 0.0095,
      "step": 7060
    },
    {
      "epoch": 176.75,
      "grad_norm": 0.14716701209545135,
      "learning_rate": 2.170467559963267e-05,
      "loss": 0.0076,
      "step": 7070
    },
    {
      "epoch": 177.0,
      "grad_norm": 0.1428631842136383,
      "learning_rate": 2.1568507032273982e-05,
      "loss": 0.0048,
      "step": 7080
    },
    {
      "epoch": 177.25,
      "grad_norm": 0.12290722876787186,
      "learning_rate": 2.1432649387167264e-05,
      "loss": 0.0044,
      "step": 7090
    },
    {
      "epoch": 177.5,
      "grad_norm": 0.10502045601606369,
      "learning_rate": 2.1297104150029973e-05,
      "loss": 0.0045,
      "step": 7100
    },
    {
      "epoch": 177.75,
      "grad_norm": 0.14660410583019257,
      "learning_rate": 2.116187280316307e-05,
      "loss": 0.0054,
      "step": 7110
    },
    {
      "epoch": 178.0,
      "grad_norm": 0.1619156002998352,
      "learning_rate": 2.1026956825434908e-05,
      "loss": 0.0065,
      "step": 7120
    },
    {
      "epoch": 178.25,
      "grad_norm": 0.22219659388065338,
      "learning_rate": 2.0892357692265017e-05,
      "loss": 0.0045,
      "step": 7130
    },
    {
      "epoch": 178.5,
      "grad_norm": 0.13960318267345428,
      "learning_rate": 2.0758076875607947e-05,
      "loss": 0.0054,
      "step": 7140
    },
    {
      "epoch": 178.75,
      "grad_norm": 0.17951151728630066,
      "learning_rate": 2.0624115843937207e-05,
      "loss": 0.0056,
      "step": 7150
    },
    {
      "epoch": 179.0,
      "grad_norm": 0.18173860013484955,
      "learning_rate": 2.0490476062229157e-05,
      "loss": 0.0074,
      "step": 7160
    },
    {
      "epoch": 179.25,
      "grad_norm": 0.1216687336564064,
      "learning_rate": 2.035715899194704e-05,
      "loss": 0.0047,
      "step": 7170
    },
    {
      "epoch": 179.5,
      "grad_norm": 0.1189367026090622,
      "learning_rate": 2.022416609102499e-05,
      "loss": 0.0047,
      "step": 7180
    },
    {
      "epoch": 179.75,
      "grad_norm": 0.14440025389194489,
      "learning_rate": 2.009149881385205e-05,
      "loss": 0.0047,
      "step": 7190
    },
    {
      "epoch": 180.0,
      "grad_norm": 0.11506537348031998,
      "learning_rate": 1.995915861125634e-05,
      "loss": 0.0077,
      "step": 7200
    },
    {
      "epoch": 180.25,
      "grad_norm": 0.19950439035892487,
      "learning_rate": 1.9827146930489065e-05,
      "loss": 0.0051,
      "step": 7210
    },
    {
      "epoch": 180.5,
      "grad_norm": 0.10524751245975494,
      "learning_rate": 1.9695465215208848e-05,
      "loss": 0.0041,
      "step": 7220
    },
    {
      "epoch": 180.75,
      "grad_norm": 0.12314356863498688,
      "learning_rate": 1.9564114905465813e-05,
      "loss": 0.0046,
      "step": 7230
    },
    {
      "epoch": 181.0,
      "grad_norm": 0.0873304083943367,
      "learning_rate": 1.9433097437685936e-05,
      "loss": 0.006,
      "step": 7240
    },
    {
      "epoch": 181.25,
      "grad_norm": 0.09135473519563675,
      "learning_rate": 1.930241424465521e-05,
      "loss": 0.0041,
      "step": 7250
    },
    {
      "epoch": 181.5,
      "grad_norm": 0.13335098326206207,
      "learning_rate": 1.9172066755504115e-05,
      "loss": 0.0054,
      "step": 7260
    },
    {
      "epoch": 181.75,
      "grad_norm": 0.15308566391468048,
      "learning_rate": 1.9042056395691914e-05,
      "loss": 0.0042,
      "step": 7270
    },
    {
      "epoch": 182.0,
      "grad_norm": 0.12902922928333282,
      "learning_rate": 1.8912384586991066e-05,
      "loss": 0.0058,
      "step": 7280
    },
    {
      "epoch": 182.25,
      "grad_norm": 0.13669724762439728,
      "learning_rate": 1.8783052747471717e-05,
      "loss": 0.0044,
      "step": 7290
    },
    {
      "epoch": 182.5,
      "grad_norm": 0.13429778814315796,
      "learning_rate": 1.865406229148611e-05,
      "loss": 0.0066,
      "step": 7300
    },
    {
      "epoch": 182.75,
      "grad_norm": 0.15294934809207916,
      "learning_rate": 1.8525414629653233e-05,
      "loss": 0.0066,
      "step": 7310
    },
    {
      "epoch": 183.0,
      "grad_norm": 0.336831659078598,
      "learning_rate": 1.8397111168843255e-05,
      "loss": 0.0075,
      "step": 7320
    },
    {
      "epoch": 183.25,
      "grad_norm": 0.14708802103996277,
      "learning_rate": 1.8269153312162323e-05,
      "loss": 0.0074,
      "step": 7330
    },
    {
      "epoch": 183.5,
      "grad_norm": 0.1348143070936203,
      "learning_rate": 1.8141542458937054e-05,
      "loss": 0.0048,
      "step": 7340
    },
    {
      "epoch": 183.75,
      "grad_norm": 0.1689818948507309,
      "learning_rate": 1.8014280004699268e-05,
      "loss": 0.0069,
      "step": 7350
    },
    {
      "epoch": 184.0,
      "grad_norm": 0.09998207539319992,
      "learning_rate": 1.788736734117078e-05,
      "loss": 0.0054,
      "step": 7360
    },
    {
      "epoch": 184.25,
      "grad_norm": 0.12446059286594391,
      "learning_rate": 1.7760805856248152e-05,
      "loss": 0.0043,
      "step": 7370
    },
    {
      "epoch": 184.5,
      "grad_norm": 0.07528968900442123,
      "learning_rate": 1.7634596933987518e-05,
      "loss": 0.0055,
      "step": 7380
    },
    {
      "epoch": 184.75,
      "grad_norm": 0.12255606800317764,
      "learning_rate": 1.7508741954589404e-05,
      "loss": 0.0044,
      "step": 7390
    },
    {
      "epoch": 185.0,
      "grad_norm": 0.2018868625164032,
      "learning_rate": 1.7383242294383717e-05,
      "loss": 0.0054,
      "step": 7400
    },
    {
      "epoch": 185.25,
      "grad_norm": 0.11658778786659241,
      "learning_rate": 1.7258099325814632e-05,
      "loss": 0.0063,
      "step": 7410
    },
    {
      "epoch": 185.5,
      "grad_norm": 0.11764887720346451,
      "learning_rate": 1.7133314417425594e-05,
      "loss": 0.0044,
      "step": 7420
    },
    {
      "epoch": 185.75,
      "grad_norm": 0.14045648276805878,
      "learning_rate": 1.7008888933844408e-05,
      "loss": 0.005,
      "step": 7430
    },
    {
      "epoch": 186.0,
      "grad_norm": 0.20185740292072296,
      "learning_rate": 1.6884824235768172e-05,
      "loss": 0.0053,
      "step": 7440
    },
    {
      "epoch": 186.25,
      "grad_norm": 0.19346244633197784,
      "learning_rate": 1.6761121679948592e-05,
      "loss": 0.0081,
      "step": 7450
    },
    {
      "epoch": 186.5,
      "grad_norm": 0.13126075267791748,
      "learning_rate": 1.663778261917695e-05,
      "loss": 0.0049,
      "step": 7460
    },
    {
      "epoch": 186.75,
      "grad_norm": 0.17565523087978363,
      "learning_rate": 1.651480840226952e-05,
      "loss": 0.0037,
      "step": 7470
    },
    {
      "epoch": 187.0,
      "grad_norm": 0.17475895583629608,
      "learning_rate": 1.639220037405258e-05,
      "loss": 0.0071,
      "step": 7480
    },
    {
      "epoch": 187.25,
      "grad_norm": 0.14766661822795868,
      "learning_rate": 1.6269959875347906e-05,
      "loss": 0.0043,
      "step": 7490
    },
    {
      "epoch": 187.5,
      "grad_norm": 0.20571599900722504,
      "learning_rate": 1.614808824295802e-05,
      "loss": 0.0086,
      "step": 7500
    },
    {
      "epoch": 187.75,
      "grad_norm": 0.1388307809829712,
      "learning_rate": 1.602658680965152e-05,
      "loss": 0.0049,
      "step": 7510
    },
    {
      "epoch": 188.0,
      "grad_norm": 0.21784968674182892,
      "learning_rate": 1.5905456904148686e-05,
      "loss": 0.0049,
      "step": 7520
    },
    {
      "epoch": 188.25,
      "grad_norm": 0.07575284689664841,
      "learning_rate": 1.57846998511067e-05,
      "loss": 0.0045,
      "step": 7530
    },
    {
      "epoch": 188.5,
      "grad_norm": 0.15673498809337616,
      "learning_rate": 1.566431697110538e-05,
      "loss": 0.0047,
      "step": 7540
    },
    {
      "epoch": 188.75,
      "grad_norm": 0.09218557924032211,
      "learning_rate": 1.554430958063259e-05,
      "loss": 0.0057,
      "step": 7550
    },
    {
      "epoch": 189.0,
      "grad_norm": 0.15179941058158875,
      "learning_rate": 1.5424678992069912e-05,
      "loss": 0.0041,
      "step": 7560
    },
    {
      "epoch": 189.25,
      "grad_norm": 0.12532265484333038,
      "learning_rate": 1.5305426513678362e-05,
      "loss": 0.0035,
      "step": 7570
    },
    {
      "epoch": 189.5,
      "grad_norm": 0.14931413531303406,
      "learning_rate": 1.518655344958388e-05,
      "loss": 0.0041,
      "step": 7580
    },
    {
      "epoch": 189.75,
      "grad_norm": 0.10902183502912521,
      "learning_rate": 1.5068061099763275e-05,
      "loss": 0.0077,
      "step": 7590
    },
    {
      "epoch": 190.0,
      "grad_norm": 0.19096030294895172,
      "learning_rate": 1.494995076002988e-05,
      "loss": 0.0059,
      "step": 7600
    },
    {
      "epoch": 190.25,
      "grad_norm": 0.155643031001091,
      "learning_rate": 1.4832223722019456e-05,
      "loss": 0.0043,
      "step": 7610
    },
    {
      "epoch": 190.5,
      "grad_norm": 0.15102514624595642,
      "learning_rate": 1.4714881273176035e-05,
      "loss": 0.0043,
      "step": 7620
    },
    {
      "epoch": 190.75,
      "grad_norm": 0.07781857997179031,
      "learning_rate": 1.4597924696737835e-05,
      "loss": 0.0038,
      "step": 7630
    },
    {
      "epoch": 191.0,
      "grad_norm": 0.12486126273870468,
      "learning_rate": 1.4481355271723252e-05,
      "loss": 0.0041,
      "step": 7640
    },
    {
      "epoch": 191.25,
      "grad_norm": 0.09264986962080002,
      "learning_rate": 1.4365174272916809e-05,
      "loss": 0.0053,
      "step": 7650
    },
    {
      "epoch": 191.5,
      "grad_norm": 0.18692605197429657,
      "learning_rate": 1.4249382970855319e-05,
      "loss": 0.0043,
      "step": 7660
    },
    {
      "epoch": 191.75,
      "grad_norm": 0.06757988780736923,
      "learning_rate": 1.4133982631813903e-05,
      "loss": 0.0054,
      "step": 7670
    },
    {
      "epoch": 192.0,
      "grad_norm": 0.14858606457710266,
      "learning_rate": 1.4018974517792194e-05,
      "loss": 0.0053,
      "step": 7680
    },
    {
      "epoch": 192.25,
      "grad_norm": 0.10279243439435959,
      "learning_rate": 1.390435988650048e-05,
      "loss": 0.0039,
      "step": 7690
    },
    {
      "epoch": 192.5,
      "grad_norm": 0.19160929322242737,
      "learning_rate": 1.3790139991346006e-05,
      "loss": 0.0052,
      "step": 7700
    },
    {
      "epoch": 192.75,
      "grad_norm": 0.16689233481884003,
      "learning_rate": 1.367631608141926e-05,
      "loss": 0.0039,
      "step": 7710
    },
    {
      "epoch": 193.0,
      "grad_norm": 0.07786817848682404,
      "learning_rate": 1.3562889401480278e-05,
      "loss": 0.0042,
      "step": 7720
    },
    {
      "epoch": 193.25,
      "grad_norm": 0.07822061330080032,
      "learning_rate": 1.3449861191945074e-05,
      "loss": 0.0039,
      "step": 7730
    },
    {
      "epoch": 193.5,
      "grad_norm": 0.11227503418922424,
      "learning_rate": 1.3337232688872009e-05,
      "loss": 0.0036,
      "step": 7740
    },
    {
      "epoch": 193.75,
      "grad_norm": 0.09792827814817429,
      "learning_rate": 1.3225005123948364e-05,
      "loss": 0.005,
      "step": 7750
    },
    {
      "epoch": 194.0,
      "grad_norm": 0.10192038118839264,
      "learning_rate": 1.311317972447681e-05,
      "loss": 0.0047,
      "step": 7760
    },
    {
      "epoch": 194.25,
      "grad_norm": 0.18177363276481628,
      "learning_rate": 1.3001757713361996e-05,
      "loss": 0.0043,
      "step": 7770
    },
    {
      "epoch": 194.5,
      "grad_norm": 0.08618608862161636,
      "learning_rate": 1.2890740309097204e-05,
      "loss": 0.0036,
      "step": 7780
    },
    {
      "epoch": 194.75,
      "grad_norm": 0.1522602140903473,
      "learning_rate": 1.2780128725750944e-05,
      "loss": 0.0047,
      "step": 7790
    },
    {
      "epoch": 195.0,
      "grad_norm": 0.06630995124578476,
      "learning_rate": 1.266992417295379e-05,
      "loss": 0.0045,
      "step": 7800
    },
    {
      "epoch": 195.25,
      "grad_norm": 0.11913304775953293,
      "learning_rate": 1.2560127855885073e-05,
      "loss": 0.0037,
      "step": 7810
    },
    {
      "epoch": 195.5,
      "grad_norm": 0.07550306618213654,
      "learning_rate": 1.2450740975259745e-05,
      "loss": 0.0044,
      "step": 7820
    },
    {
      "epoch": 195.75,
      "grad_norm": 0.12006478756666183,
      "learning_rate": 1.234176472731517e-05,
      "loss": 0.0053,
      "step": 7830
    },
    {
      "epoch": 196.0,
      "grad_norm": 0.10951386392116547,
      "learning_rate": 1.2233200303798158e-05,
      "loss": 0.006,
      "step": 7840
    },
    {
      "epoch": 196.25,
      "grad_norm": 0.087790846824646,
      "learning_rate": 1.2125048891951846e-05,
      "loss": 0.007,
      "step": 7850
    },
    {
      "epoch": 196.5,
      "grad_norm": 0.23221564292907715,
      "learning_rate": 1.2017311674502745e-05,
      "loss": 0.0051,
      "step": 7860
    },
    {
      "epoch": 196.75,
      "grad_norm": 0.1510801911354065,
      "learning_rate": 1.1909989829647822e-05,
      "loss": 0.0032,
      "step": 7870
    },
    {
      "epoch": 197.0,
      "grad_norm": 0.1127174124121666,
      "learning_rate": 1.1803084531041553e-05,
      "loss": 0.0041,
      "step": 7880
    },
    {
      "epoch": 197.25,
      "grad_norm": 0.1897778958082199,
      "learning_rate": 1.1696596947783162e-05,
      "loss": 0.0048,
      "step": 7890
    },
    {
      "epoch": 197.5,
      "grad_norm": 0.1312101036310196,
      "learning_rate": 1.1590528244403803e-05,
      "loss": 0.0056,
      "step": 7900
    },
    {
      "epoch": 197.75,
      "grad_norm": 0.16427397727966309,
      "learning_rate": 1.148487958085382e-05,
      "loss": 0.0039,
      "step": 7910
    },
    {
      "epoch": 198.0,
      "grad_norm": 0.14600983262062073,
      "learning_rate": 1.1379652112490086e-05,
      "loss": 0.0036,
      "step": 7920
    },
    {
      "epoch": 198.25,
      "grad_norm": 0.13961434364318848,
      "learning_rate": 1.1274846990063315e-05,
      "loss": 0.0045,
      "step": 7930
    },
    {
      "epoch": 198.5,
      "grad_norm": 0.13402146100997925,
      "learning_rate": 1.117046535970554e-05,
      "loss": 0.004,
      "step": 7940
    },
    {
      "epoch": 198.75,
      "grad_norm": 0.07271775603294373,
      "learning_rate": 1.106650836291755e-05,
      "loss": 0.0047,
      "step": 7950
    },
    {
      "epoch": 199.0,
      "grad_norm": 0.09896207600831985,
      "learning_rate": 1.0962977136556418e-05,
      "loss": 0.0058,
      "step": 7960
    },
    {
      "epoch": 199.25,
      "grad_norm": 0.21567489206790924,
      "learning_rate": 1.0859872812823024e-05,
      "loss": 0.0101,
      "step": 7970
    },
    {
      "epoch": 199.5,
      "grad_norm": 0.08785581588745117,
      "learning_rate": 1.0757196519249747e-05,
      "loss": 0.0061,
      "step": 7980
    },
    {
      "epoch": 199.75,
      "grad_norm": 0.1338099092245102,
      "learning_rate": 1.0654949378688077e-05,
      "loss": 0.0053,
      "step": 7990
    },
    {
      "epoch": 200.0,
      "grad_norm": 0.07626625150442123,
      "learning_rate": 1.0553132509296376e-05,
      "loss": 0.0099,
      "step": 8000
    },
    {
      "epoch": 200.25,
      "grad_norm": 0.10850115120410919,
      "learning_rate": 1.0451747024527613e-05,
      "loss": 0.0041,
      "step": 8010
    },
    {
      "epoch": 200.5,
      "grad_norm": 0.09877875447273254,
      "learning_rate": 1.0350794033117189e-05,
      "loss": 0.0045,
      "step": 8020
    },
    {
      "epoch": 200.75,
      "grad_norm": 0.1204601377248764,
      "learning_rate": 1.0250274639070856e-05,
      "loss": 0.0042,
      "step": 8030
    },
    {
      "epoch": 201.0,
      "grad_norm": 0.1761455088853836,
      "learning_rate": 1.0150189941652599e-05,
      "loss": 0.0046,
      "step": 8040
    },
    {
      "epoch": 201.25,
      "grad_norm": 0.06259100139141083,
      "learning_rate": 1.0050541035372635e-05,
      "loss": 0.004,
      "step": 8050
    },
    {
      "epoch": 201.5,
      "grad_norm": 0.1302749216556549,
      "learning_rate": 9.951329009975458e-06,
      "loss": 0.0042,
      "step": 8060
    },
    {
      "epoch": 201.75,
      "grad_norm": 0.1296391487121582,
      "learning_rate": 9.852554950427845e-06,
      "loss": 0.0036,
      "step": 8070
    },
    {
      "epoch": 202.0,
      "grad_norm": 0.0929638147354126,
      "learning_rate": 9.754219936907105e-06,
      "loss": 0.0036,
      "step": 8080
    },
    {
      "epoch": 202.25,
      "grad_norm": 0.09810598194599152,
      "learning_rate": 9.656325044789194e-06,
      "loss": 0.003,
      "step": 8090
    },
    {
      "epoch": 202.5,
      "grad_norm": 0.1328025460243225,
      "learning_rate": 9.55887134463697e-06,
      "loss": 0.0049,
      "step": 8100
    },
    {
      "epoch": 202.75,
      "grad_norm": 0.19226421415805817,
      "learning_rate": 9.461859902188475e-06,
      "loss": 0.0039,
      "step": 8110
    },
    {
      "epoch": 203.0,
      "grad_norm": 0.07218576222658157,
      "learning_rate": 9.365291778345303e-06,
      "loss": 0.0037,
      "step": 8120
    },
    {
      "epoch": 203.25,
      "grad_norm": 0.10829392820596695,
      "learning_rate": 9.269168029160991e-06,
      "loss": 0.0049,
      "step": 8130
    },
    {
      "epoch": 203.5,
      "grad_norm": 0.11217450350522995,
      "learning_rate": 9.173489705829447e-06,
      "loss": 0.0041,
      "step": 8140
    },
    {
      "epoch": 203.75,
      "grad_norm": 0.10603217035531998,
      "learning_rate": 9.078257854673516e-06,
      "loss": 0.0027,
      "step": 8150
    },
    {
      "epoch": 204.0,
      "grad_norm": 0.13703373074531555,
      "learning_rate": 8.983473517133429e-06,
      "loss": 0.005,
      "step": 8160
    },
    {
      "epoch": 204.25,
      "grad_norm": 0.1856386959552765,
      "learning_rate": 8.889137729755537e-06,
      "loss": 0.0043,
      "step": 8170
    },
    {
      "epoch": 204.5,
      "grad_norm": 0.12523987889289856,
      "learning_rate": 8.79525152418087e-06,
      "loss": 0.0037,
      "step": 8180
    },
    {
      "epoch": 204.75,
      "grad_norm": 0.17454007267951965,
      "learning_rate": 8.701815927133961e-06,
      "loss": 0.0034,
      "step": 8190
    },
    {
      "epoch": 205.0,
      "grad_norm": 0.22606174647808075,
      "learning_rate": 8.608831960411534e-06,
      "loss": 0.0053,
      "step": 8200
    },
    {
      "epoch": 205.25,
      "grad_norm": 0.12180226296186447,
      "learning_rate": 8.516300640871321e-06,
      "loss": 0.0031,
      "step": 8210
    },
    {
      "epoch": 205.5,
      "grad_norm": 0.08695121854543686,
      "learning_rate": 8.424222980421038e-06,
      "loss": 0.0038,
      "step": 8220
    },
    {
      "epoch": 205.75,
      "grad_norm": 0.13644380867481232,
      "learning_rate": 8.332599986007184e-06,
      "loss": 0.0058,
      "step": 8230
    },
    {
      "epoch": 206.0,
      "grad_norm": 0.1191595196723938,
      "learning_rate": 8.241432659604203e-06,
      "loss": 0.0041,
      "step": 8240
    },
    {
      "epoch": 206.25,
      "grad_norm": 0.08874445408582687,
      "learning_rate": 8.150721998203331e-06,
      "loss": 0.0048,
      "step": 8250
    },
    {
      "epoch": 206.5,
      "grad_norm": 0.07907148450613022,
      "learning_rate": 8.06046899380184e-06,
      "loss": 0.0032,
      "step": 8260
    },
    {
      "epoch": 206.75,
      "grad_norm": 0.10713402926921844,
      "learning_rate": 7.970674633392133e-06,
      "loss": 0.0035,
      "step": 8270
    },
    {
      "epoch": 207.0,
      "grad_norm": 0.11685764789581299,
      "learning_rate": 7.881339898950924e-06,
      "loss": 0.0037,
      "step": 8280
    },
    {
      "epoch": 207.25,
      "grad_norm": 0.13961216807365417,
      "learning_rate": 7.792465767428597e-06,
      "loss": 0.0075,
      "step": 8290
    },
    {
      "epoch": 207.5,
      "grad_norm": 0.1829424947500229,
      "learning_rate": 7.704053210738376e-06,
      "loss": 0.0045,
      "step": 8300
    },
    {
      "epoch": 207.75,
      "grad_norm": 0.1581076681613922,
      "learning_rate": 7.6161031957458494e-06,
      "loss": 0.0032,
      "step": 8310
    },
    {
      "epoch": 208.0,
      "grad_norm": 0.08495005965232849,
      "learning_rate": 7.5286166842582605e-06,
      "loss": 0.0033,
      "step": 8320
    },
    {
      "epoch": 208.25,
      "grad_norm": 0.1170332059264183,
      "learning_rate": 7.4415946330140814e-06,
      "loss": 0.0062,
      "step": 8330
    },
    {
      "epoch": 208.5,
      "grad_norm": 0.05432112142443657,
      "learning_rate": 7.3550379936725644e-06,
      "loss": 0.0038,
      "step": 8340
    },
    {
      "epoch": 208.75,
      "grad_norm": 0.07181187719106674,
      "learning_rate": 7.2689477128032035e-06,
      "loss": 0.0032,
      "step": 8350
    },
    {
      "epoch": 209.0,
      "grad_norm": 0.08231719583272934,
      "learning_rate": 7.183324731875551e-06,
      "loss": 0.0044,
      "step": 8360
    },
    {
      "epoch": 209.25,
      "grad_norm": 0.0662977546453476,
      "learning_rate": 7.098169987248782e-06,
      "loss": 0.0037,
      "step": 8370
    },
    {
      "epoch": 209.5,
      "grad_norm": 0.09705691784620285,
      "learning_rate": 7.013484410161553e-06,
      "loss": 0.0038,
      "step": 8380
    },
    {
      "epoch": 209.75,
      "grad_norm": 0.08059106767177582,
      "learning_rate": 6.92926892672176e-06,
      "loss": 0.0025,
      "step": 8390
    },
    {
      "epoch": 210.0,
      "grad_norm": 0.07411599904298782,
      "learning_rate": 6.845524457896446e-06,
      "loss": 0.004,
      "step": 8400
    },
    {
      "epoch": 210.25,
      "grad_norm": 0.12973953783512115,
      "learning_rate": 6.7622519195017165e-06,
      "loss": 0.0042,
      "step": 8410
    },
    {
      "epoch": 210.5,
      "grad_norm": 0.07746973633766174,
      "learning_rate": 6.679452222192684e-06,
      "loss": 0.0034,
      "step": 8420
    },
    {
      "epoch": 210.75,
      "grad_norm": 0.08138252794742584,
      "learning_rate": 6.597126271453579e-06,
      "loss": 0.0035,
      "step": 8430
    },
    {
      "epoch": 211.0,
      "grad_norm": 0.10311781615018845,
      "learning_rate": 6.51527496758782e-06,
      "loss": 0.0041,
      "step": 8440
    },
    {
      "epoch": 211.25,
      "grad_norm": 0.10894886404275894,
      "learning_rate": 6.433899205708155e-06,
      "loss": 0.0047,
      "step": 8450
    },
    {
      "epoch": 211.5,
      "grad_norm": 0.22221334278583527,
      "learning_rate": 6.352999875726856e-06,
      "loss": 0.0033,
      "step": 8460
    },
    {
      "epoch": 211.75,
      "grad_norm": 0.11968138813972473,
      "learning_rate": 6.272577862346052e-06,
      "loss": 0.0032,
      "step": 8470
    },
    {
      "epoch": 212.0,
      "grad_norm": 0.14150996506214142,
      "learning_rate": 6.192634045047996e-06,
      "loss": 0.004,
      "step": 8480
    },
    {
      "epoch": 212.25,
      "grad_norm": 0.15094904601573944,
      "learning_rate": 6.113169298085458e-06,
      "loss": 0.0038,
      "step": 8490
    },
    {
      "epoch": 212.5,
      "grad_norm": 0.14192979037761688,
      "learning_rate": 6.034184490472195e-06,
      "loss": 0.0075,
      "step": 8500
    },
    {
      "epoch": 212.75,
      "grad_norm": 0.0771539956331253,
      "learning_rate": 5.955680485973386e-06,
      "loss": 0.0046,
      "step": 8510
    },
    {
      "epoch": 213.0,
      "grad_norm": 0.14663337171077728,
      "learning_rate": 5.877658143096265e-06,
      "loss": 0.0034,
      "step": 8520
    },
    {
      "epoch": 213.25,
      "grad_norm": 0.1873416304588318,
      "learning_rate": 5.800118315080661e-06,
      "loss": 0.0045,
      "step": 8530
    },
    {
      "epoch": 213.5,
      "grad_norm": 0.09949974715709686,
      "learning_rate": 5.723061849889716e-06,
      "loss": 0.0033,
      "step": 8540
    },
    {
      "epoch": 213.75,
      "grad_norm": 0.10987547785043716,
      "learning_rate": 5.646489590200604e-06,
      "loss": 0.0033,
      "step": 8550
    },
    {
      "epoch": 214.0,
      "grad_norm": 0.10888543725013733,
      "learning_rate": 5.570402373395256e-06,
      "loss": 0.0028,
      "step": 8560
    },
    {
      "epoch": 214.25,
      "grad_norm": 0.07901471853256226,
      "learning_rate": 5.494801031551305e-06,
      "loss": 0.0033,
      "step": 8570
    },
    {
      "epoch": 214.5,
      "grad_norm": 0.16858859360218048,
      "learning_rate": 5.41968639143291e-06,
      "loss": 0.0025,
      "step": 8580
    },
    {
      "epoch": 214.75,
      "grad_norm": 0.06866457313299179,
      "learning_rate": 5.345059274481751e-06,
      "loss": 0.0041,
      "step": 8590
    },
    {
      "epoch": 215.0,
      "grad_norm": 0.07472264021635056,
      "learning_rate": 5.270920496808002e-06,
      "loss": 0.0027,
      "step": 8600
    },
    {
      "epoch": 215.25,
      "grad_norm": 0.08445484936237335,
      "learning_rate": 5.1972708691814695e-06,
      "loss": 0.0026,
      "step": 8610
    },
    {
      "epoch": 215.5,
      "grad_norm": 0.09009315073490143,
      "learning_rate": 5.124111197022674e-06,
      "loss": 0.003,
      "step": 8620
    },
    {
      "epoch": 215.75,
      "grad_norm": 0.18223707377910614,
      "learning_rate": 5.051442280394081e-06,
      "loss": 0.0041,
      "step": 8630
    },
    {
      "epoch": 216.0,
      "grad_norm": 0.11874091625213623,
      "learning_rate": 4.979264913991322e-06,
      "loss": 0.0043,
      "step": 8640
    },
    {
      "epoch": 216.25,
      "grad_norm": 0.05684478208422661,
      "learning_rate": 4.907579887134489e-06,
      "loss": 0.0025,
      "step": 8650
    },
    {
      "epoch": 216.5,
      "grad_norm": 0.10025288909673691,
      "learning_rate": 4.836387983759572e-06,
      "loss": 0.003,
      "step": 8660
    },
    {
      "epoch": 216.75,
      "grad_norm": 0.06758899241685867,
      "learning_rate": 4.765689982409816e-06,
      "loss": 0.0063,
      "step": 8670
    },
    {
      "epoch": 217.0,
      "grad_norm": 0.10817938297986984,
      "learning_rate": 4.695486656227233e-06,
      "loss": 0.0036,
      "step": 8680
    },
    {
      "epoch": 217.25,
      "grad_norm": 0.10704730451107025,
      "learning_rate": 4.625778772944156e-06,
      "loss": 0.0042,
      "step": 8690
    },
    {
      "epoch": 217.5,
      "grad_norm": 0.07309766113758087,
      "learning_rate": 4.556567094874825e-06,
      "loss": 0.0043,
      "step": 8700
    },
    {
      "epoch": 217.75,
      "grad_norm": 0.11254431307315826,
      "learning_rate": 4.487852378907059e-06,
      "loss": 0.0032,
      "step": 8710
    },
    {
      "epoch": 218.0,
      "grad_norm": 0.08993148058652878,
      "learning_rate": 4.419635376493986e-06,
      "loss": 0.0047,
      "step": 8720
    },
    {
      "epoch": 218.25,
      "grad_norm": 0.06215588003396988,
      "learning_rate": 4.351916833645825e-06,
      "loss": 0.0025,
      "step": 8730
    },
    {
      "epoch": 218.5,
      "grad_norm": 0.05950621888041496,
      "learning_rate": 4.284697490921691e-06,
      "loss": 0.0039,
      "step": 8740
    },
    {
      "epoch": 218.75,
      "grad_norm": 0.09767914563417435,
      "learning_rate": 4.2179780834215585e-06,
      "loss": 0.004,
      "step": 8750
    },
    {
      "epoch": 219.0,
      "grad_norm": 0.17002426087856293,
      "learning_rate": 4.151759340778178e-06,
      "loss": 0.0032,
      "step": 8760
    },
    {
      "epoch": 219.25,
      "grad_norm": 0.09146512299776077,
      "learning_rate": 4.086041987149109e-06,
      "loss": 0.0047,
      "step": 8770
    },
    {
      "epoch": 219.5,
      "grad_norm": 0.0718497633934021,
      "learning_rate": 4.020826741208811e-06,
      "loss": 0.0027,
      "step": 8780
    },
    {
      "epoch": 219.75,
      "grad_norm": 0.11396536231040955,
      "learning_rate": 3.956114316140746e-06,
      "loss": 0.003,
      "step": 8790
    },
    {
      "epoch": 220.0,
      "grad_norm": 0.1519760638475418,
      "learning_rate": 3.891905419629643e-06,
      "loss": 0.0051,
      "step": 8800
    },
    {
      "epoch": 220.25,
      "grad_norm": 0.07514660060405731,
      "learning_rate": 3.8282007538536946e-06,
      "loss": 0.0039,
      "step": 8810
    },
    {
      "epoch": 220.5,
      "grad_norm": 0.20433557033538818,
      "learning_rate": 3.7650010154769265e-06,
      "loss": 0.004,
      "step": 8820
    },
    {
      "epoch": 220.75,
      "grad_norm": 0.07488595694303513,
      "learning_rate": 3.7023068956415608e-06,
      "loss": 0.0055,
      "step": 8830
    },
    {
      "epoch": 221.0,
      "grad_norm": 0.15067219734191895,
      "learning_rate": 3.6401190799604303e-06,
      "loss": 0.0038,
      "step": 8840
    },
    {
      "epoch": 221.25,
      "grad_norm": 0.10581692308187485,
      "learning_rate": 3.578438248509536e-06,
      "loss": 0.0033,
      "step": 8850
    },
    {
      "epoch": 221.5,
      "grad_norm": 0.13871265947818756,
      "learning_rate": 3.5172650758205583e-06,
      "loss": 0.0027,
      "step": 8860
    },
    {
      "epoch": 221.75,
      "grad_norm": 0.11503514647483826,
      "learning_rate": 3.45660023087353e-06,
      "loss": 0.0031,
      "step": 8870
    },
    {
      "epoch": 222.0,
      "grad_norm": 0.14610545337200165,
      "learning_rate": 3.3964443770894528e-06,
      "loss": 0.0032,
      "step": 8880
    },
    {
      "epoch": 222.25,
      "grad_norm": 0.05525064840912819,
      "learning_rate": 3.3367981723231245e-06,
      "loss": 0.0022,
      "step": 8890
    },
    {
      "epoch": 222.5,
      "grad_norm": 0.1545409858226776,
      "learning_rate": 3.2776622688558746e-06,
      "loss": 0.004,
      "step": 8900
    },
    {
      "epoch": 222.75,
      "grad_norm": 0.06882253289222717,
      "learning_rate": 3.2190373133884677e-06,
      "loss": 0.0034,
      "step": 8910
    },
    {
      "epoch": 223.0,
      "grad_norm": 0.07723452895879745,
      "learning_rate": 3.1609239470340446e-06,
      "loss": 0.0033,
      "step": 8920
    },
    {
      "epoch": 223.25,
      "grad_norm": 0.05147286877036095,
      "learning_rate": 3.1033228053110373e-06,
      "loss": 0.0033,
      "step": 8930
    },
    {
      "epoch": 223.5,
      "grad_norm": 0.08452828973531723,
      "learning_rate": 3.0462345181363314e-06,
      "loss": 0.0027,
      "step": 8940
    },
    {
      "epoch": 223.75,
      "grad_norm": 0.07312655448913574,
      "learning_rate": 2.9896597098182654e-06,
      "loss": 0.0028,
      "step": 8950
    },
    {
      "epoch": 224.0,
      "grad_norm": 0.19344282150268555,
      "learning_rate": 2.933598999049891e-06,
      "loss": 0.0028,
      "step": 8960
    },
    {
      "epoch": 224.25,
      "grad_norm": 0.049786657094955444,
      "learning_rate": 2.8780529989021697e-06,
      "loss": 0.0035,
      "step": 8970
    },
    {
      "epoch": 224.5,
      "grad_norm": 0.09912026673555374,
      "learning_rate": 2.823022316817242e-06,
      "loss": 0.004,
      "step": 8980
    },
    {
      "epoch": 224.75,
      "grad_norm": 0.15381453931331635,
      "learning_rate": 2.7685075546018456e-06,
      "loss": 0.0028,
      "step": 8990
    },
    {
      "epoch": 225.0,
      "grad_norm": 0.09527960419654846,
      "learning_rate": 2.7145093084206598e-06,
      "loss": 0.0081,
      "step": 9000
    },
    {
      "epoch": 225.25,
      "grad_norm": 0.046797893941402435,
      "learning_rate": 2.661028168789892e-06,
      "loss": 0.0028,
      "step": 9010
    },
    {
      "epoch": 225.5,
      "grad_norm": 0.12533637881278992,
      "learning_rate": 2.6080647205706855e-06,
      "loss": 0.0047,
      "step": 9020
    },
    {
      "epoch": 225.75,
      "grad_norm": 0.14190489053726196,
      "learning_rate": 2.555619542962834e-06,
      "loss": 0.0026,
      "step": 9030
    },
    {
      "epoch": 226.0,
      "grad_norm": 0.08050890266895294,
      "learning_rate": 2.503693209498409e-06,
      "loss": 0.0023,
      "step": 9040
    },
    {
      "epoch": 226.25,
      "grad_norm": 0.10872689634561539,
      "learning_rate": 2.452286288035449e-06,
      "loss": 0.0038,
      "step": 9050
    },
    {
      "epoch": 226.5,
      "grad_norm": 0.07601405680179596,
      "learning_rate": 2.4013993407518363e-06,
      "loss": 0.0022,
      "step": 9060
    },
    {
      "epoch": 226.75,
      "grad_norm": 0.22459723055362701,
      "learning_rate": 2.351032924139063e-06,
      "loss": 0.0041,
      "step": 9070
    },
    {
      "epoch": 227.0,
      "grad_norm": 0.11310923099517822,
      "learning_rate": 2.30118758899619e-06,
      "loss": 0.0033,
      "step": 9080
    },
    {
      "epoch": 227.25,
      "grad_norm": 0.08301689475774765,
      "learning_rate": 2.2518638804238157e-06,
      "loss": 0.0029,
      "step": 9090
    },
    {
      "epoch": 227.5,
      "grad_norm": 0.06724400073289871,
      "learning_rate": 2.203062337818118e-06,
      "loss": 0.0044,
      "step": 9100
    },
    {
      "epoch": 227.75,
      "grad_norm": 0.13066349923610687,
      "learning_rate": 2.1547834948649483e-06,
      "loss": 0.0032,
      "step": 9110
    },
    {
      "epoch": 228.0,
      "grad_norm": 0.18874166905879974,
      "learning_rate": 2.1070278795340017e-06,
      "loss": 0.0038,
      "step": 9120
    },
    {
      "epoch": 228.25,
      "grad_norm": 0.087242990732193,
      "learning_rate": 2.059796014073029e-06,
      "loss": 0.0034,
      "step": 9130
    },
    {
      "epoch": 228.5,
      "grad_norm": 0.1420496702194214,
      "learning_rate": 2.01308841500214e-06,
      "loss": 0.0034,
      "step": 9140
    },
    {
      "epoch": 228.75,
      "grad_norm": 0.04035526141524315,
      "learning_rate": 1.9669055931081704e-06,
      "loss": 0.0022,
      "step": 9150
    },
    {
      "epoch": 229.0,
      "grad_norm": 0.08264325559139252,
      "learning_rate": 1.9212480534390507e-06,
      "loss": 0.0024,
      "step": 9160
    },
    {
      "epoch": 229.25,
      "grad_norm": 0.06740362197160721,
      "learning_rate": 1.8761162952983246e-06,
      "loss": 0.0045,
      "step": 9170
    },
    {
      "epoch": 229.5,
      "grad_norm": 0.041461482644081116,
      "learning_rate": 1.8315108122396618e-06,
      "loss": 0.0036,
      "step": 9180
    },
    {
      "epoch": 229.75,
      "grad_norm": 0.06524401158094406,
      "learning_rate": 1.787432092061475e-06,
      "loss": 0.0031,
      "step": 9190
    },
    {
      "epoch": 230.0,
      "grad_norm": 0.09731091558933258,
      "learning_rate": 1.743880616801602e-06,
      "loss": 0.0053,
      "step": 9200
    },
    {
      "epoch": 230.25,
      "grad_norm": 0.15023453533649445,
      "learning_rate": 1.7008568627319865e-06,
      "loss": 0.0022,
      "step": 9210
    },
    {
      "epoch": 230.5,
      "grad_norm": 0.08554910868406296,
      "learning_rate": 1.6583613003535226e-06,
      "loss": 0.0055,
      "step": 9220
    },
    {
      "epoch": 230.75,
      "grad_norm": 0.060999348759651184,
      "learning_rate": 1.6163943943908522e-06,
      "loss": 0.0026,
      "step": 9230
    },
    {
      "epoch": 231.0,
      "grad_norm": 0.10573070496320724,
      "learning_rate": 1.5749566037873476e-06,
      "loss": 0.005,
      "step": 9240
    },
    {
      "epoch": 231.25,
      "grad_norm": 0.07315921783447266,
      "learning_rate": 1.5340483817000428e-06,
      "loss": 0.0045,
      "step": 9250
    },
    {
      "epoch": 231.5,
      "grad_norm": 0.08707199990749359,
      "learning_rate": 1.4936701754947101e-06,
      "loss": 0.003,
      "step": 9260
    },
    {
      "epoch": 231.75,
      "grad_norm": 0.03300639986991882,
      "learning_rate": 1.4538224267409361e-06,
      "loss": 0.0034,
      "step": 9270
    },
    {
      "epoch": 232.0,
      "grad_norm": 0.06528275460004807,
      "learning_rate": 1.414505571207314e-06,
      "loss": 0.0027,
      "step": 9280
    },
    {
      "epoch": 232.25,
      "grad_norm": 0.10502535104751587,
      "learning_rate": 1.3757200388566816e-06,
      "loss": 0.0037,
      "step": 9290
    },
    {
      "epoch": 232.5,
      "grad_norm": 0.1114937961101532,
      "learning_rate": 1.3374662538414074e-06,
      "loss": 0.0033,
      "step": 9300
    },
    {
      "epoch": 232.75,
      "grad_norm": 0.1334535777568817,
      "learning_rate": 1.2997446344987617e-06,
      "loss": 0.0032,
      "step": 9310
    },
    {
      "epoch": 233.0,
      "grad_norm": 0.15189684927463531,
      "learning_rate": 1.262555593346315e-06,
      "loss": 0.0045,
      "step": 9320
    },
    {
      "epoch": 233.25,
      "grad_norm": 0.03467800095677376,
      "learning_rate": 1.2258995370774685e-06,
      "loss": 0.0026,
      "step": 9330
    },
    {
      "epoch": 233.5,
      "grad_norm": 0.16183927655220032,
      "learning_rate": 1.1897768665569798e-06,
      "loss": 0.0037,
      "step": 9340
    },
    {
      "epoch": 233.75,
      "grad_norm": 0.06361699104309082,
      "learning_rate": 1.1541879768165954e-06,
      "loss": 0.0055,
      "step": 9350
    },
    {
      "epoch": 234.0,
      "grad_norm": 0.08980465680360794,
      "learning_rate": 1.1191332570507085e-06,
      "loss": 0.0034,
      "step": 9360
    },
    {
      "epoch": 234.25,
      "grad_norm": 0.05992332473397255,
      "learning_rate": 1.0846130906121132e-06,
      "loss": 0.0025,
      "step": 9370
    },
    {
      "epoch": 234.5,
      "grad_norm": 0.17438055574893951,
      "learning_rate": 1.0506278550078131e-06,
      "loss": 0.0051,
      "step": 9380
    },
    {
      "epoch": 234.75,
      "grad_norm": 0.058787088841199875,
      "learning_rate": 1.0171779218949185e-06,
      "loss": 0.0049,
      "step": 9390
    },
    {
      "epoch": 235.0,
      "grad_norm": 0.15100450813770294,
      "learning_rate": 9.842636570765174e-07,
      "loss": 0.0041,
      "step": 9400
    },
    {
      "epoch": 235.25,
      "grad_norm": 0.08583474904298782,
      "learning_rate": 9.518854204977612e-07,
      "loss": 0.0025,
      "step": 9410
    },
    {
      "epoch": 235.5,
      "grad_norm": 0.06979444622993469,
      "learning_rate": 9.200435662418349e-07,
      "loss": 0.0033,
      "step": 9420
    },
    {
      "epoch": 235.75,
      "grad_norm": 0.11030140519142151,
      "learning_rate": 8.887384425261658e-07,
      "loss": 0.0034,
      "step": 9430
    },
    {
      "epoch": 236.0,
      "grad_norm": 0.06448012590408325,
      "learning_rate": 8.579703916985648e-07,
      "loss": 0.0027,
      "step": 9440
    },
    {
      "epoch": 236.25,
      "grad_norm": 0.167857825756073,
      "learning_rate": 8.277397502335194e-07,
      "loss": 0.0026,
      "step": 9450
    },
    {
      "epoch": 236.5,
      "grad_norm": 0.09123705327510834,
      "learning_rate": 7.980468487284675e-07,
      "loss": 0.0036,
      "step": 9460
    },
    {
      "epoch": 236.75,
      "grad_norm": 0.06160133704543114,
      "learning_rate": 7.688920119002297e-07,
      "loss": 0.0039,
      "step": 9470
    },
    {
      "epoch": 237.0,
      "grad_norm": 0.08287883549928665,
      "learning_rate": 7.402755585814269e-07,
      "loss": 0.0029,
      "step": 9480
    },
    {
      "epoch": 237.25,
      "grad_norm": 0.05484398826956749,
      "learning_rate": 7.121978017170073e-07,
      "loss": 0.0023,
      "step": 9490
    },
    {
      "epoch": 237.5,
      "grad_norm": 0.04693023860454559,
      "learning_rate": 6.846590483608306e-07,
      "loss": 0.0041,
      "step": 9500
    },
    {
      "epoch": 237.75,
      "grad_norm": 0.14454308152198792,
      "learning_rate": 6.576595996722834e-07,
      "loss": 0.0035,
      "step": 9510
    },
    {
      "epoch": 238.0,
      "grad_norm": 0.06262636929750443,
      "learning_rate": 6.311997509130141e-07,
      "loss": 0.002,
      "step": 9520
    },
    {
      "epoch": 238.25,
      "grad_norm": 0.03196113556623459,
      "learning_rate": 6.052797914436803e-07,
      "loss": 0.006,
      "step": 9530
    },
    {
      "epoch": 238.5,
      "grad_norm": 0.08593147993087769,
      "learning_rate": 5.799000047208181e-07,
      "loss": 0.0049,
      "step": 9540
    },
    {
      "epoch": 238.75,
      "grad_norm": 0.0887538343667984,
      "learning_rate": 5.550606682937054e-07,
      "loss": 0.0031,
      "step": 9550
    },
    {
      "epoch": 239.0,
      "grad_norm": 0.13892526924610138,
      "learning_rate": 5.307620538013481e-07,
      "loss": 0.0038,
      "step": 9560
    },
    {
      "epoch": 239.25,
      "grad_norm": 0.1304452270269394,
      "learning_rate": 5.070044269694874e-07,
      "loss": 0.0038,
      "step": 9570
    },
    {
      "epoch": 239.5,
      "grad_norm": 0.1446593701839447,
      "learning_rate": 4.837880476077417e-07,
      "loss": 0.0035,
      "step": 9580
    },
    {
      "epoch": 239.75,
      "grad_norm": 0.1164969801902771,
      "learning_rate": 4.6111316960670835e-07,
      "loss": 0.0032,
      "step": 9590
    },
    {
      "epoch": 240.0,
      "grad_norm": 0.15182846784591675,
      "learning_rate": 4.389800409352218e-07,
      "loss": 0.0036,
      "step": 9600
    },
    {
      "epoch": 240.25,
      "grad_norm": 0.02840401977300644,
      "learning_rate": 4.173889036376277e-07,
      "loss": 0.0042,
      "step": 9610
    },
    {
      "epoch": 240.5,
      "grad_norm": 0.09419921785593033,
      "learning_rate": 3.963399938311463e-07,
      "loss": 0.002,
      "step": 9620
    },
    {
      "epoch": 240.75,
      "grad_norm": 0.16660663485527039,
      "learning_rate": 3.7583354170328545e-07,
      "loss": 0.0043,
      "step": 9630
    },
    {
      "epoch": 241.0,
      "grad_norm": 0.13299746811389923,
      "learning_rate": 3.558697715093207e-07,
      "loss": 0.0029,
      "step": 9640
    },
    {
      "epoch": 241.25,
      "grad_norm": 0.12092949450016022,
      "learning_rate": 3.3644890156983576e-07,
      "loss": 0.0061,
      "step": 9650
    },
    {
      "epoch": 241.5,
      "grad_norm": 0.060206662863492966,
      "learning_rate": 3.175711442683638e-07,
      "loss": 0.0025,
      "step": 9660
    },
    {
      "epoch": 241.75,
      "grad_norm": 0.08426082134246826,
      "learning_rate": 2.9923670604902197e-07,
      "loss": 0.0028,
      "step": 9670
    },
    {
      "epoch": 242.0,
      "grad_norm": 0.0851016417145729,
      "learning_rate": 2.814457874143028e-07,
      "loss": 0.0042,
      "step": 9680
    },
    {
      "epoch": 242.25,
      "grad_norm": 0.07830564677715302,
      "learning_rate": 2.641985829228366e-07,
      "loss": 0.0023,
      "step": 9690
    },
    {
      "epoch": 242.5,
      "grad_norm": 0.11447057127952576,
      "learning_rate": 2.474952811872877e-07,
      "loss": 0.0043,
      "step": 9700
    },
    {
      "epoch": 242.75,
      "grad_norm": 0.05448366329073906,
      "learning_rate": 2.3133606487228397e-07,
      "loss": 0.0025,
      "step": 9710
    },
    {
      "epoch": 243.0,
      "grad_norm": 0.03906814008951187,
      "learning_rate": 2.157211106924295e-07,
      "loss": 0.0027,
      "step": 9720
    },
    {
      "epoch": 243.25,
      "grad_norm": 0.039081498980522156,
      "learning_rate": 2.006505894103672e-07,
      "loss": 0.0023,
      "step": 9730
    },
    {
      "epoch": 243.5,
      "grad_norm": 0.027017461135983467,
      "learning_rate": 1.8612466583489696e-07,
      "loss": 0.0034,
      "step": 9740
    },
    {
      "epoch": 243.75,
      "grad_norm": 0.09606248885393143,
      "learning_rate": 1.7214349881918834e-07,
      "loss": 0.0048,
      "step": 9750
    },
    {
      "epoch": 244.0,
      "grad_norm": 0.05149327963590622,
      "learning_rate": 1.5870724125904845e-07,
      "loss": 0.0033,
      "step": 9760
    },
    {
      "epoch": 244.25,
      "grad_norm": 0.14646421372890472,
      "learning_rate": 1.4581604009124006e-07,
      "loss": 0.004,
      "step": 9770
    },
    {
      "epoch": 244.5,
      "grad_norm": 0.06484408676624298,
      "learning_rate": 1.334700362918717e-07,
      "loss": 0.0025,
      "step": 9780
    },
    {
      "epoch": 244.75,
      "grad_norm": 0.037307512015104294,
      "learning_rate": 1.2166936487486015e-07,
      "loss": 0.0022,
      "step": 9790
    },
    {
      "epoch": 245.0,
      "grad_norm": 0.1416601985692978,
      "learning_rate": 1.1041415489045914e-07,
      "loss": 0.0028,
      "step": 9800
    },
    {
      "epoch": 245.25,
      "grad_norm": 0.18991032242774963,
      "learning_rate": 9.970452942384412e-08,
      "loss": 0.0035,
      "step": 9810
    },
    {
      "epoch": 245.5,
      "grad_norm": 0.07187339663505554,
      "learning_rate": 8.954060559375754e-08,
      "loss": 0.0023,
      "step": 9820
    },
    {
      "epoch": 245.75,
      "grad_norm": 0.07545599341392517,
      "learning_rate": 7.99224945512489e-08,
      "loss": 0.0028,
      "step": 9830
    },
    {
      "epoch": 246.0,
      "grad_norm": 0.10132917016744614,
      "learning_rate": 7.085030147843675e-08,
      "loss": 0.0037,
      "step": 9840
    },
    {
      "epoch": 246.25,
      "grad_norm": 0.06773098558187485,
      "learning_rate": 6.232412558736523e-08,
      "loss": 0.0049,
      "step": 9850
    },
    {
      "epoch": 246.5,
      "grad_norm": 0.19954784214496613,
      "learning_rate": 5.434406011893822e-08,
      "loss": 0.0029,
      "step": 9860
    },
    {
      "epoch": 246.75,
      "grad_norm": 0.1477789282798767,
      "learning_rate": 4.6910192341864664e-08,
      "loss": 0.0027,
      "step": 9870
    },
    {
      "epoch": 247.0,
      "grad_norm": 0.063265360891819,
      "learning_rate": 4.0022603551737035e-08,
      "loss": 0.0066,
      "step": 9880
    },
    {
      "epoch": 247.25,
      "grad_norm": 0.054880768060684204,
      "learning_rate": 3.3681369070120985e-08,
      "loss": 0.0033,
      "step": 9890
    },
    {
      "epoch": 247.5,
      "grad_norm": 0.02149050496518612,
      "learning_rate": 2.7886558243744866e-08,
      "loss": 0.0031,
      "step": 9900
    },
    {
      "epoch": 247.75,
      "grad_norm": 0.18347109854221344,
      "learning_rate": 2.2638234443722596e-08,
      "loss": 0.0036,
      "step": 9910
    },
    {
      "epoch": 248.0,
      "grad_norm": 0.0956336110830307,
      "learning_rate": 1.7936455064887504e-08,
      "loss": 0.0044,
      "step": 9920
    },
    {
      "epoch": 248.25,
      "grad_norm": 0.18302513659000397,
      "learning_rate": 1.378127152514841e-08,
      "loss": 0.0031,
      "step": 9930
    },
    {
      "epoch": 248.5,
      "grad_norm": 0.08520723134279251,
      "learning_rate": 1.0172729264917857e-08,
      "loss": 0.0071,
      "step": 9940
    },
    {
      "epoch": 248.75,
      "grad_norm": 0.039946626871824265,
      "learning_rate": 7.1108677466458215e-09,
      "loss": 0.0026,
      "step": 9950
    },
    {
      "epoch": 249.0,
      "grad_norm": 0.17577043175697327,
      "learning_rate": 4.595720454353414e-09,
      "loss": 0.0047,
      "step": 9960
    },
    {
      "epoch": 249.25,
      "grad_norm": 0.1010713055729866,
      "learning_rate": 2.627314893294264e-09,
      "loss": 0.0034,
      "step": 9970
    },
    {
      "epoch": 249.5,
      "grad_norm": 0.04403097182512283,
      "learning_rate": 1.2056725896270048e-09,
      "loss": 0.0041,
      "step": 9980
    },
    {
      "epoch": 249.75,
      "grad_norm": 0.10400509834289551,
      "learning_rate": 3.308090902098826e-10,
      "loss": 0.0024,
      "step": 9990
    },
    {
      "epoch": 250.0,
      "grad_norm": 0.0933779925107956,
      "learning_rate": 2.7339624120159555e-12,
      "loss": 0.0028,
      "step": 10000
    },
    {
      "epoch": 250.0,
      "step": 10000,
      "total_flos": 0.0,
      "train_loss": 0.015269231259077788,
      "train_runtime": 7033.4938,
      "train_samples_per_second": 45.497,
      "train_steps_per_second": 1.422
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 250,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}