{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 3250,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03076923076923077,
      "grad_norm": 1.4984049797058105,
      "learning_rate": 0.0001999953280342959,
      "loss": 2.6105,
      "step": 10
    },
    {
      "epoch": 0.06153846153846154,
      "grad_norm": 13.759217262268066,
      "learning_rate": 0.00019998131257372876,
      "loss": 2.1182,
      "step": 20
    },
    {
      "epoch": 0.09230769230769231,
      "grad_norm": 1.4669537544250488,
      "learning_rate": 0.0001999579549278937,
      "loss": 1.9057,
      "step": 30
    },
    {
      "epoch": 0.12307692307692308,
      "grad_norm": 1.4035037755966187,
      "learning_rate": 0.00019992525727931303,
      "loss": 1.6194,
      "step": 40
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 1.9111855030059814,
      "learning_rate": 0.00019988322268323268,
      "loss": 1.4734,
      "step": 50
    },
    {
      "epoch": 0.18461538461538463,
      "grad_norm": 1.690224051475525,
      "learning_rate": 0.0001998318550673364,
      "loss": 1.4012,
      "step": 60
    },
    {
      "epoch": 0.2153846153846154,
      "grad_norm": 1.4565370082855225,
      "learning_rate": 0.00019977115923137912,
      "loss": 1.3154,
      "step": 70
    },
    {
      "epoch": 0.24615384615384617,
      "grad_norm": 1.325324535369873,
      "learning_rate": 0.00019970114084673796,
      "loss": 1.2612,
      "step": 80
    },
    {
      "epoch": 0.27692307692307694,
      "grad_norm": 1.8117148876190186,
      "learning_rate": 0.0001996218064558829,
      "loss": 1.1993,
      "step": 90
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 1.4131654500961304,
      "learning_rate": 0.00019953316347176488,
      "loss": 1.1835,
      "step": 100
    },
    {
      "epoch": 0.3384615384615385,
      "grad_norm": 1.9338051080703735,
      "learning_rate": 0.00019943522017712358,
      "loss": 1.2223,
      "step": 110
    },
    {
      "epoch": 0.36923076923076925,
      "grad_norm": 1.7367668151855469,
      "learning_rate": 0.0001993279857237133,
      "loss": 1.1528,
      "step": 120
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5504117012023926,
      "learning_rate": 0.0001992114701314478,
      "loss": 1.1799,
      "step": 130
    },
    {
      "epoch": 0.4307692307692308,
      "grad_norm": 1.7593408823013306,
      "learning_rate": 0.0001990856842874641,
      "loss": 1.1915,
      "step": 140
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 1.5112876892089844,
      "learning_rate": 0.0001989506399451051,
      "loss": 1.1552,
      "step": 150
    },
    {
      "epoch": 0.49230769230769234,
      "grad_norm": 1.5315320491790771,
      "learning_rate": 0.00019880634972282166,
      "loss": 1.1271,
      "step": 160
    },
    {
      "epoch": 0.5230769230769231,
      "grad_norm": 1.7272446155548096,
      "learning_rate": 0.0001986528271029931,
      "loss": 1.1321,
      "step": 170
    },
    {
      "epoch": 0.5538461538461539,
      "grad_norm": 1.8666999340057373,
      "learning_rate": 0.00019849008643066772,
      "loss": 1.1367,
      "step": 180
    },
    {
      "epoch": 0.5846153846153846,
      "grad_norm": 1.8758796453475952,
      "learning_rate": 0.00019831814291222232,
      "loss": 1.1219,
      "step": 190
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 1.7113131284713745,
      "learning_rate": 0.00019813701261394136,
      "loss": 1.1101,
      "step": 200
    },
    {
      "epoch": 0.6461538461538462,
      "grad_norm": 1.506138801574707,
      "learning_rate": 0.0001979467124605156,
      "loss": 1.0801,
      "step": 210
    },
    {
      "epoch": 0.676923076923077,
      "grad_norm": 1.4458012580871582,
      "learning_rate": 0.0001977472602334609,
      "loss": 1.0838,
      "step": 220
    },
    {
      "epoch": 0.7076923076923077,
      "grad_norm": 1.615057349205017,
      "learning_rate": 0.0001975386745694565,
      "loss": 1.094,
      "step": 230
    },
    {
      "epoch": 0.7384615384615385,
      "grad_norm": 1.4620922803878784,
      "learning_rate": 0.00019732097495860386,
      "loss": 1.0689,
      "step": 240
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 1.5983144044876099,
      "learning_rate": 0.0001970941817426052,
      "loss": 1.072,
      "step": 250
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5557613372802734,
      "learning_rate": 0.0001968583161128631,
      "loss": 1.0974,
      "step": 260
    },
    {
      "epoch": 0.8307692307692308,
      "grad_norm": 1.674525499343872,
      "learning_rate": 0.00019661340010850026,
      "loss": 1.0958,
      "step": 270
    },
    {
      "epoch": 0.8615384615384616,
      "grad_norm": 1.9557188749313354,
      "learning_rate": 0.00019635945661430006,
      "loss": 1.0713,
      "step": 280
    },
    {
      "epoch": 0.8923076923076924,
      "grad_norm": 1.408217191696167,
      "learning_rate": 0.00019609650935856844,
      "loss": 1.0681,
      "step": 290
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 1.6660012006759644,
      "learning_rate": 0.00019582458291091663,
      "loss": 1.0639,
      "step": 300
    },
    {
      "epoch": 0.9538461538461539,
      "grad_norm": 1.5910792350769043,
      "learning_rate": 0.00019554370267996538,
      "loss": 1.0844,
      "step": 310
    },
    {
      "epoch": 0.9846153846153847,
      "grad_norm": 1.698133111000061,
      "learning_rate": 0.0001952538949109708,
      "loss": 1.1323,
      "step": 320
    },
    {
      "epoch": 1.0153846153846153,
      "grad_norm": 1.7201091051101685,
      "learning_rate": 0.00019495518668337201,
      "loss": 1.0761,
      "step": 330
    },
    {
      "epoch": 1.0461538461538462,
      "grad_norm": 1.7263455390930176,
      "learning_rate": 0.00019464760590826098,
      "loss": 1.0388,
      "step": 340
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 1.8437412977218628,
      "learning_rate": 0.0001943311813257743,
      "loss": 1.0229,
      "step": 350
    },
    {
      "epoch": 1.1076923076923078,
      "grad_norm": 1.4422318935394287,
      "learning_rate": 0.00019400594250240798,
      "loss": 1.0218,
      "step": 360
    },
    {
      "epoch": 1.1384615384615384,
      "grad_norm": 1.765499472618103,
      "learning_rate": 0.0001936719198282545,
      "loss": 1.0087,
      "step": 370
    },
    {
      "epoch": 1.1692307692307693,
      "grad_norm": 1.500490427017212,
      "learning_rate": 0.00019332914451416347,
      "loss": 1.0239,
      "step": 380
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.5604429244995117,
      "learning_rate": 0.00019297764858882514,
      "loss": 1.0271,
      "step": 390
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 1.636246681213379,
      "learning_rate": 0.00019261746489577765,
      "loss": 1.0537,
      "step": 400
    },
    {
      "epoch": 1.2615384615384615,
      "grad_norm": 1.7009079456329346,
      "learning_rate": 0.00019224862709033824,
      "loss": 1.0267,
      "step": 410
    },
    {
      "epoch": 1.2923076923076924,
      "grad_norm": 1.7331833839416504,
      "learning_rate": 0.00019187116963645842,
      "loss": 0.962,
      "step": 420
    },
    {
      "epoch": 1.323076923076923,
      "grad_norm": 1.5424834489822388,
      "learning_rate": 0.00019148512780350384,
      "loss": 0.9834,
      "step": 430
    },
    {
      "epoch": 1.353846153846154,
      "grad_norm": 1.474749207496643,
      "learning_rate": 0.0001910905376629585,
      "loss": 1.0047,
      "step": 440
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 1.739359974861145,
      "learning_rate": 0.00019068743608505455,
      "loss": 1.0188,
      "step": 450
    },
    {
      "epoch": 1.4153846153846155,
      "grad_norm": 1.52376389503479,
      "learning_rate": 0.0001902758607353269,
      "loss": 1.0226,
      "step": 460
    },
    {
      "epoch": 1.4461538461538461,
      "grad_norm": 1.723073124885559,
      "learning_rate": 0.0001898558500710939,
      "loss": 1.0054,
      "step": 470
    },
    {
      "epoch": 1.476923076923077,
      "grad_norm": 1.7302210330963135,
      "learning_rate": 0.00018942744333786397,
      "loss": 1.0335,
      "step": 480
    },
    {
      "epoch": 1.5076923076923077,
      "grad_norm": 1.5802934169769287,
      "learning_rate": 0.0001889906805656684,
      "loss": 0.9924,
      "step": 490
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 1.383296012878418,
      "learning_rate": 0.000188545602565321,
      "loss": 1.0083,
      "step": 500
    },
    {
      "epoch": 1.5692307692307692,
      "grad_norm": 1.5573105812072754,
      "learning_rate": 0.00018809225092460488,
      "loss": 0.9982,
      "step": 510
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5339369773864746,
      "learning_rate": 0.00018763066800438636,
      "loss": 0.9827,
      "step": 520
    },
    {
      "epoch": 1.6307692307692307,
      "grad_norm": 1.5264880657196045,
      "learning_rate": 0.00018716089693465696,
      "loss": 0.9879,
      "step": 530
    },
    {
      "epoch": 1.6615384615384614,
      "grad_norm": 1.4193902015686035,
      "learning_rate": 0.00018668298161050309,
      "loss": 0.9825,
      "step": 540
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 1.6493202447891235,
      "learning_rate": 0.00018619696668800492,
      "loss": 1.0143,
      "step": 550
    },
    {
      "epoch": 1.7230769230769232,
      "grad_norm": 1.4842519760131836,
      "learning_rate": 0.00018570289758006346,
      "loss": 0.9721,
      "step": 560
    },
    {
      "epoch": 1.7538461538461538,
      "grad_norm": 1.617529034614563,
      "learning_rate": 0.0001852008204521572,
      "loss": 0.9826,
      "step": 570
    },
    {
      "epoch": 1.7846153846153845,
      "grad_norm": 1.6360862255096436,
      "learning_rate": 0.0001846907822180286,
      "loss": 0.9236,
      "step": 580
    },
    {
      "epoch": 1.8153846153846154,
      "grad_norm": 1.664921760559082,
      "learning_rate": 0.00018417283053530044,
      "loss": 1.0072,
      "step": 590
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 1.5803855657577515,
      "learning_rate": 0.00018364701380102266,
      "loss": 1.0154,
      "step": 600
    },
    {
      "epoch": 1.876923076923077,
      "grad_norm": 1.6201574802398682,
      "learning_rate": 0.0001831133811471503,
      "loss": 1.0299,
      "step": 610
    },
    {
      "epoch": 1.9076923076923076,
      "grad_norm": 1.5785369873046875,
      "learning_rate": 0.0001825719824359524,
      "loss": 1.015,
      "step": 620
    },
    {
      "epoch": 1.9384615384615385,
      "grad_norm": 1.6001813411712646,
      "learning_rate": 0.0001820228682553533,
      "loss": 0.967,
      "step": 630
    },
    {
      "epoch": 1.9692307692307693,
      "grad_norm": 1.5755499601364136,
      "learning_rate": 0.00018146608991420534,
      "loss": 0.9779,
      "step": 640
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.593526840209961,
      "learning_rate": 0.00018090169943749476,
      "loss": 1.0293,
      "step": 650
    },
    {
      "epoch": 2.0307692307692307,
      "grad_norm": 2.3367538452148438,
      "learning_rate": 0.00018032974956148063,
      "loss": 0.9708,
      "step": 660
    },
    {
      "epoch": 2.0615384615384613,
      "grad_norm": 1.7174549102783203,
      "learning_rate": 0.00017975029372876706,
      "loss": 0.9657,
      "step": 670
    },
    {
      "epoch": 2.0923076923076924,
      "grad_norm": 1.5391074419021606,
      "learning_rate": 0.0001791633860833096,
      "loss": 0.9353,
      "step": 680
    },
    {
      "epoch": 2.123076923076923,
      "grad_norm": 1.7437571287155151,
      "learning_rate": 0.00017856908146535603,
      "loss": 0.9348,
      "step": 690
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 1.6768618822097778,
      "learning_rate": 0.00017796743540632223,
      "loss": 0.9376,
      "step": 700
    },
    {
      "epoch": 2.184615384615385,
      "grad_norm": 1.6524372100830078,
      "learning_rate": 0.00017735850412360331,
      "loss": 0.9116,
      "step": 710
    },
    {
      "epoch": 2.2153846153846155,
      "grad_norm": 1.7284245491027832,
      "learning_rate": 0.00017674234451532065,
      "loss": 0.9618,
      "step": 720
    },
    {
      "epoch": 2.246153846153846,
      "grad_norm": 1.5894191265106201,
      "learning_rate": 0.00017611901415500535,
      "loss": 0.924,
      "step": 730
    },
    {
      "epoch": 2.276923076923077,
      "grad_norm": 1.6842291355133057,
      "learning_rate": 0.00017548857128621875,
      "loss": 0.9143,
      "step": 740
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 1.7472116947174072,
      "learning_rate": 0.00017485107481711012,
      "loss": 0.9417,
      "step": 750
    },
    {
      "epoch": 2.3384615384615386,
      "grad_norm": 1.4523415565490723,
      "learning_rate": 0.00017420658431491223,
      "loss": 0.9279,
      "step": 760
    },
    {
      "epoch": 2.3692307692307693,
      "grad_norm": 1.6412981748580933,
      "learning_rate": 0.00017355516000037554,
      "loss": 0.9273,
      "step": 770
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.49977445602417,
      "learning_rate": 0.00017289686274214118,
      "loss": 0.9011,
      "step": 780
    },
    {
      "epoch": 2.430769230769231,
      "grad_norm": 2.0299291610717773,
      "learning_rate": 0.0001722317540510534,
      "loss": 0.919,
      "step": 790
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 1.8127689361572266,
      "learning_rate": 0.00017155989607441213,
      "loss": 0.9446,
      "step": 800
    },
    {
      "epoch": 2.4923076923076923,
      "grad_norm": 1.7033181190490723,
      "learning_rate": 0.00017088135159016584,
      "loss": 0.9044,
      "step": 810
    },
    {
      "epoch": 2.523076923076923,
      "grad_norm": 1.5404083728790283,
      "learning_rate": 0.00017019618400104572,
      "loss": 0.8942,
      "step": 820
    },
    {
      "epoch": 2.5538461538461537,
      "grad_norm": 1.6343528032302856,
      "learning_rate": 0.00016950445732864127,
      "loss": 0.98,
      "step": 830
    },
    {
      "epoch": 2.5846153846153848,
      "grad_norm": 1.7268812656402588,
      "learning_rate": 0.00016880623620741842,
      "loss": 0.9107,
      "step": 840
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 1.6956366300582886,
      "learning_rate": 0.00016810158587867973,
      "loss": 0.9351,
      "step": 850
    },
    {
      "epoch": 2.646153846153846,
      "grad_norm": 1.7267690896987915,
      "learning_rate": 0.0001673905721844686,
      "loss": 0.9147,
      "step": 860
    },
    {
      "epoch": 2.676923076923077,
      "grad_norm": 1.5935823917388916,
      "learning_rate": 0.00016667326156141692,
      "loss": 0.9393,
      "step": 870
    },
    {
      "epoch": 2.707692307692308,
      "grad_norm": 1.5465987920761108,
      "learning_rate": 0.00016594972103453726,
      "loss": 0.9158,
      "step": 880
    },
    {
      "epoch": 2.7384615384615385,
      "grad_norm": 1.7121493816375732,
      "learning_rate": 0.0001652200182109602,
      "loss": 0.934,
      "step": 890
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 1.7504358291625977,
      "learning_rate": 0.00016448422127361706,
      "loss": 0.9277,
      "step": 900
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.8299874067306519,
      "learning_rate": 0.000163742398974869,
      "loss": 0.9498,
      "step": 910
    },
    {
      "epoch": 2.830769230769231,
      "grad_norm": 1.7550499439239502,
      "learning_rate": 0.00016299462063008272,
      "loss": 0.9769,
      "step": 920
    },
    {
      "epoch": 2.8615384615384616,
      "grad_norm": 1.6253832578659058,
      "learning_rate": 0.00016224095611115384,
      "loss": 0.8633,
      "step": 930
    },
    {
      "epoch": 2.8923076923076922,
      "grad_norm": 1.7190197706222534,
      "learning_rate": 0.00016148147583997812,
      "loss": 0.9252,
      "step": 940
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 1.6005258560180664,
      "learning_rate": 0.00016071625078187114,
      "loss": 0.9606,
      "step": 950
    },
    {
      "epoch": 2.953846153846154,
      "grad_norm": 1.6922186613082886,
      "learning_rate": 0.0001599453524389374,
      "loss": 0.8922,
      "step": 960
    },
    {
      "epoch": 2.9846153846153847,
      "grad_norm": 1.7878159284591675,
      "learning_rate": 0.00015916885284338937,
      "loss": 0.9142,
      "step": 970
    },
    {
      "epoch": 3.0153846153846153,
      "grad_norm": 1.6686667203903198,
      "learning_rate": 0.00015838682455081657,
      "loss": 0.925,
      "step": 980
    },
    {
      "epoch": 3.046153846153846,
      "grad_norm": 1.6054987907409668,
      "learning_rate": 0.00015759934063340627,
      "loss": 0.9249,
      "step": 990
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 1.7218083143234253,
      "learning_rate": 0.00015680647467311557,
      "loss": 0.8723,
      "step": 1000
    },
    {
      "epoch": 3.1076923076923078,
      "grad_norm": 2.051055431365967,
      "learning_rate": 0.00015600830075479603,
      "loss": 0.8255,
      "step": 1010
    },
    {
      "epoch": 3.1384615384615384,
      "grad_norm": 1.7609128952026367,
      "learning_rate": 0.00015520489345927096,
      "loss": 0.8821,
      "step": 1020
    },
    {
      "epoch": 3.169230769230769,
      "grad_norm": 1.690056562423706,
      "learning_rate": 0.00015439632785636706,
      "loss": 0.9113,
      "step": 1030
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.9333645105361938,
      "learning_rate": 0.00015358267949789966,
      "loss": 0.9241,
      "step": 1040
    },
    {
      "epoch": 3.230769230769231,
      "grad_norm": 1.7492021322250366,
      "learning_rate": 0.0001527640244106133,
      "loss": 0.8993,
      "step": 1050
    },
    {
      "epoch": 3.2615384615384615,
      "grad_norm": 1.8243930339813232,
      "learning_rate": 0.00015194043908907775,
      "loss": 0.8928,
      "step": 1060
    },
    {
      "epoch": 3.292307692307692,
      "grad_norm": 1.7545908689498901,
      "learning_rate": 0.00015111200048854056,
      "loss": 0.8879,
      "step": 1070
    },
    {
      "epoch": 3.3230769230769233,
      "grad_norm": 1.7745007276535034,
      "learning_rate": 0.00015027878601773633,
      "loss": 0.869,
      "step": 1080
    },
    {
      "epoch": 3.353846153846154,
      "grad_norm": 1.8173617124557495,
      "learning_rate": 0.0001494408735316537,
      "loss": 0.9033,
      "step": 1090
    },
    {
      "epoch": 3.3846153846153846,
      "grad_norm": 1.9644584655761719,
      "learning_rate": 0.0001485983413242606,
      "loss": 0.8392,
      "step": 1100
    },
    {
      "epoch": 3.4153846153846152,
      "grad_norm": 1.8126754760742188,
      "learning_rate": 0.00014775126812118864,
      "loss": 0.8928,
      "step": 1110
    },
    {
      "epoch": 3.4461538461538463,
      "grad_norm": 1.8512457609176636,
      "learning_rate": 0.00014689973307237687,
      "loss": 0.887,
      "step": 1120
    },
    {
      "epoch": 3.476923076923077,
      "grad_norm": 1.7760255336761475,
      "learning_rate": 0.00014604381574467615,
      "loss": 0.8951,
      "step": 1130
    },
    {
      "epoch": 3.5076923076923077,
      "grad_norm": 1.7298815250396729,
      "learning_rate": 0.0001451835961144145,
      "loss": 0.8799,
      "step": 1140
    },
    {
      "epoch": 3.5384615384615383,
      "grad_norm": 1.6533232927322388,
      "learning_rate": 0.00014431915455992414,
      "loss": 0.8652,
      "step": 1150
    },
    {
      "epoch": 3.569230769230769,
      "grad_norm": 1.8500922918319702,
      "learning_rate": 0.000143450571854031,
      "loss": 0.8864,
      "step": 1160
    },
    {
      "epoch": 3.6,
      "grad_norm": 1.719185471534729,
      "learning_rate": 0.00014257792915650728,
      "loss": 0.8602,
      "step": 1170
    },
    {
      "epoch": 3.6307692307692307,
      "grad_norm": 2.0936434268951416,
      "learning_rate": 0.00014170130800648814,
      "loss": 0.8538,
      "step": 1180
    },
    {
      "epoch": 3.6615384615384614,
      "grad_norm": 1.9639226198196411,
      "learning_rate": 0.0001408207903148525,
      "loss": 0.8279,
      "step": 1190
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 2.0342531204223633,
      "learning_rate": 0.00013993645835656953,
      "loss": 0.8683,
      "step": 1200
    },
    {
      "epoch": 3.723076923076923,
      "grad_norm": 1.8024394512176514,
      "learning_rate": 0.0001390483947630109,
      "loss": 0.888,
      "step": 1210
    },
    {
      "epoch": 3.753846153846154,
      "grad_norm": 1.8016464710235596,
      "learning_rate": 0.00013815668251422952,
      "loss": 0.9126,
      "step": 1220
    },
    {
      "epoch": 3.7846153846153845,
      "grad_norm": 1.8998101949691772,
      "learning_rate": 0.0001372614049312064,
      "loss": 0.8799,
      "step": 1230
    },
    {
      "epoch": 3.815384615384615,
      "grad_norm": 1.674023151397705,
      "learning_rate": 0.0001363626456680647,
      "loss": 0.8822,
      "step": 1240
    },
    {
      "epoch": 3.8461538461538463,
      "grad_norm": 1.7552895545959473,
      "learning_rate": 0.00013546048870425356,
      "loss": 0.8456,
      "step": 1250
    },
    {
      "epoch": 3.876923076923077,
      "grad_norm": 1.790547490119934,
      "learning_rate": 0.00013455501833670088,
      "loss": 0.8149,
      "step": 1260
    },
    {
      "epoch": 3.9076923076923076,
      "grad_norm": 1.8295682668685913,
      "learning_rate": 0.0001336463191719367,
      "loss": 0.8622,
      "step": 1270
    },
    {
      "epoch": 3.9384615384615387,
      "grad_norm": 1.7753140926361084,
      "learning_rate": 0.00013273447611818767,
      "loss": 0.8678,
      "step": 1280
    },
    {
      "epoch": 3.9692307692307693,
      "grad_norm": 1.808028221130371,
      "learning_rate": 0.00013181957437744332,
      "loss": 0.8578,
      "step": 1290
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.848429799079895,
      "learning_rate": 0.00013090169943749476,
      "loss": 0.8463,
      "step": 1300
    },
    {
      "epoch": 4.030769230769231,
      "grad_norm": 1.9943642616271973,
      "learning_rate": 0.00012998093706394675,
      "loss": 0.8016,
      "step": 1310
    },
    {
      "epoch": 4.061538461538461,
      "grad_norm": 1.9723416566848755,
      "learning_rate": 0.00012905737329220392,
      "loss": 0.8379,
      "step": 1320
    },
    {
      "epoch": 4.092307692307692,
      "grad_norm": 2.0291624069213867,
      "learning_rate": 0.00012813109441943166,
      "loss": 0.8041,
      "step": 1330
    },
    {
      "epoch": 4.123076923076923,
      "grad_norm": 1.9695005416870117,
      "learning_rate": 0.00012720218699649243,
      "loss": 0.8507,
      "step": 1340
    },
    {
      "epoch": 4.153846153846154,
      "grad_norm": 2.0986056327819824,
      "learning_rate": 0.0001262707378198587,
      "loss": 0.8786,
      "step": 1350
    },
    {
      "epoch": 4.184615384615385,
      "grad_norm": 1.849661946296692,
      "learning_rate": 0.00012533683392350263,
      "loss": 0.8659,
      "step": 1360
    },
    {
      "epoch": 4.2153846153846155,
      "grad_norm": 1.8133331537246704,
      "learning_rate": 0.00012440056257076375,
      "loss": 0.833,
      "step": 1370
    },
    {
      "epoch": 4.246153846153846,
      "grad_norm": 2.1380434036254883,
      "learning_rate": 0.00012346201124619502,
      "loss": 0.8067,
      "step": 1380
    },
    {
      "epoch": 4.276923076923077,
      "grad_norm": 1.7417466640472412,
      "learning_rate": 0.00012252126764738844,
      "loss": 0.8342,
      "step": 1390
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 1.8836528062820435,
      "learning_rate": 0.00012157841967678063,
      "loss": 0.8251,
      "step": 1400
    },
    {
      "epoch": 4.338461538461538,
      "grad_norm": 2.2744300365448,
      "learning_rate": 0.00012063355543343924,
      "loss": 0.8138,
      "step": 1410
    },
    {
      "epoch": 4.36923076923077,
      "grad_norm": 2.028259754180908,
      "learning_rate": 0.00011968676320483103,
      "loss": 0.8063,
      "step": 1420
    },
    {
      "epoch": 4.4,
      "grad_norm": 1.8546655178070068,
      "learning_rate": 0.00011873813145857249,
      "loss": 0.8415,
      "step": 1430
    },
    {
      "epoch": 4.430769230769231,
      "grad_norm": 1.8695989847183228,
      "learning_rate": 0.00011778774883416323,
      "loss": 0.8486,
      "step": 1440
    },
    {
      "epoch": 4.461538461538462,
      "grad_norm": 1.8322254419326782,
      "learning_rate": 0.00011683570413470383,
      "loss": 0.8131,
      "step": 1450
    },
    {
      "epoch": 4.492307692307692,
      "grad_norm": 2.126826047897339,
      "learning_rate": 0.00011588208631859807,
      "loss": 0.8307,
      "step": 1460
    },
    {
      "epoch": 4.523076923076923,
      "grad_norm": 1.9609365463256836,
      "learning_rate": 0.00011492698449124042,
      "loss": 0.8343,
      "step": 1470
    },
    {
      "epoch": 4.553846153846154,
      "grad_norm": 1.7528473138809204,
      "learning_rate": 0.0001139704878966906,
      "loss": 0.8788,
      "step": 1480
    },
    {
      "epoch": 4.584615384615384,
      "grad_norm": 1.9300537109375,
      "learning_rate": 0.00011301268590933434,
      "loss": 0.8691,
      "step": 1490
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 2.0728390216827393,
      "learning_rate": 0.0001120536680255323,
      "loss": 0.8018,
      "step": 1500
    },
    {
      "epoch": 4.6461538461538465,
      "grad_norm": 2.0605084896087646,
      "learning_rate": 0.00011109352385525783,
      "loss": 0.8635,
      "step": 1510
    },
    {
      "epoch": 4.676923076923077,
      "grad_norm": 1.8925397396087646,
      "learning_rate": 0.00011013234311372353,
      "loss": 0.7742,
      "step": 1520
    },
    {
      "epoch": 4.707692307692308,
      "grad_norm": 1.926761507987976,
      "learning_rate": 0.00010917021561299863,
      "loss": 0.8611,
      "step": 1530
    },
    {
      "epoch": 4.7384615384615385,
      "grad_norm": 1.9038374423980713,
      "learning_rate": 0.00010820723125361684,
      "loss": 0.8712,
      "step": 1540
    },
    {
      "epoch": 4.769230769230769,
      "grad_norm": 2.0429656505584717,
      "learning_rate": 0.00010724348001617625,
      "loss": 0.8328,
      "step": 1550
    },
    {
      "epoch": 4.8,
      "grad_norm": 2.0092427730560303,
      "learning_rate": 0.00010627905195293135,
      "loss": 0.8373,
      "step": 1560
    },
    {
      "epoch": 4.8307692307692305,
      "grad_norm": 1.9809768199920654,
      "learning_rate": 0.00010531403717937887,
      "loss": 0.8246,
      "step": 1570
    },
    {
      "epoch": 4.861538461538462,
      "grad_norm": 1.7576910257339478,
      "learning_rate": 0.00010434852586583736,
      "loss": 0.8373,
      "step": 1580
    },
    {
      "epoch": 4.892307692307693,
      "grad_norm": 1.7677350044250488,
      "learning_rate": 0.00010338260822902167,
      "loss": 0.7948,
      "step": 1590
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 1.678804636001587,
      "learning_rate": 0.00010241637452361323,
      "loss": 0.8287,
      "step": 1600
    },
    {
      "epoch": 4.953846153846154,
      "grad_norm": 1.6783541440963745,
      "learning_rate": 0.00010144991503382674,
      "loss": 0.7825,
      "step": 1610
    },
    {
      "epoch": 4.984615384615385,
      "grad_norm": 1.9260199069976807,
      "learning_rate": 0.00010048332006497406,
      "loss": 0.8559,
      "step": 1620
    },
    {
      "epoch": 5.015384615384615,
      "grad_norm": 1.8397397994995117,
      "learning_rate": 9.9516679935026e-05,
      "loss": 0.8318,
      "step": 1630
    },
    {
      "epoch": 5.046153846153846,
      "grad_norm": 2.1345362663269043,
      "learning_rate": 9.855008496617327e-05,
      "loss": 0.8066,
      "step": 1640
    },
    {
      "epoch": 5.076923076923077,
      "grad_norm": 1.9365698099136353,
      "learning_rate": 9.75836254763868e-05,
      "loss": 0.7916,
      "step": 1650
    },
    {
      "epoch": 5.107692307692307,
      "grad_norm": 2.0976245403289795,
      "learning_rate": 9.661739177097836e-05,
      "loss": 0.7567,
      "step": 1660
    },
    {
      "epoch": 5.138461538461539,
      "grad_norm": 1.8643560409545898,
      "learning_rate": 9.565147413416266e-05,
      "loss": 0.7796,
      "step": 1670
    },
    {
      "epoch": 5.1692307692307695,
      "grad_norm": 1.9309316873550415,
      "learning_rate": 9.468596282062114e-05,
      "loss": 0.7884,
      "step": 1680
    },
    {
      "epoch": 5.2,
      "grad_norm": 1.8694071769714355,
      "learning_rate": 9.372094804706867e-05,
      "loss": 0.805,
      "step": 1690
    },
    {
      "epoch": 5.230769230769231,
      "grad_norm": 1.7506272792816162,
      "learning_rate": 9.275651998382377e-05,
      "loss": 0.7667,
      "step": 1700
    },
    {
      "epoch": 5.2615384615384615,
      "grad_norm": 2.0087368488311768,
      "learning_rate": 9.179276874638315e-05,
      "loss": 0.8121,
      "step": 1710
    },
    {
      "epoch": 5.292307692307692,
      "grad_norm": 1.9830255508422852,
      "learning_rate": 9.082978438700138e-05,
      "loss": 0.7741,
      "step": 1720
    },
    {
      "epoch": 5.323076923076923,
      "grad_norm": 2.119537591934204,
      "learning_rate": 8.986765688627652e-05,
      "loss": 0.7927,
      "step": 1730
    },
    {
      "epoch": 5.3538461538461535,
      "grad_norm": 1.9642573595046997,
      "learning_rate": 8.890647614474224e-05,
      "loss": 0.7995,
      "step": 1740
    },
    {
      "epoch": 5.384615384615385,
      "grad_norm": 2.063324213027954,
      "learning_rate": 8.79463319744677e-05,
      "loss": 0.8108,
      "step": 1750
    },
    {
      "epoch": 5.415384615384616,
      "grad_norm": 2.1207165718078613,
      "learning_rate": 8.698731409066568e-05,
      "loss": 0.7708,
      "step": 1760
    },
    {
      "epoch": 5.446153846153846,
      "grad_norm": 2.086155891418457,
      "learning_rate": 8.602951210330942e-05,
      "loss": 0.8002,
      "step": 1770
    },
    {
      "epoch": 5.476923076923077,
      "grad_norm": 2.0774879455566406,
      "learning_rate": 8.50730155087596e-05,
      "loss": 0.7897,
      "step": 1780
    },
    {
      "epoch": 5.507692307692308,
      "grad_norm": 2.012019157409668,
      "learning_rate": 8.411791368140196e-05,
      "loss": 0.8106,
      "step": 1790
    },
    {
      "epoch": 5.538461538461538,
      "grad_norm": 2.1534152030944824,
      "learning_rate": 8.316429586529615e-05,
      "loss": 0.8051,
      "step": 1800
    },
    {
      "epoch": 5.569230769230769,
      "grad_norm": 1.9364471435546875,
      "learning_rate": 8.221225116583678e-05,
      "loss": 0.8359,
      "step": 1810
    },
    {
      "epoch": 5.6,
      "grad_norm": 2.1292238235473633,
      "learning_rate": 8.126186854142752e-05,
      "loss": 0.824,
      "step": 1820
    },
    {
      "epoch": 5.63076923076923,
      "grad_norm": 1.9687916040420532,
      "learning_rate": 8.0313236795169e-05,
      "loss": 0.8499,
      "step": 1830
    },
    {
      "epoch": 5.661538461538462,
      "grad_norm": 1.9281556606292725,
      "learning_rate": 7.936644456656081e-05,
      "loss": 0.7799,
      "step": 1840
    },
    {
      "epoch": 5.6923076923076925,
      "grad_norm": 2.123990297317505,
      "learning_rate": 7.84215803232194e-05,
      "loss": 0.8438,
      "step": 1850
    },
    {
      "epoch": 5.723076923076923,
      "grad_norm": 1.9118835926055908,
      "learning_rate": 7.747873235261157e-05,
      "loss": 0.7557,
      "step": 1860
    },
    {
      "epoch": 5.753846153846154,
      "grad_norm": 1.90312922000885,
      "learning_rate": 7.6537988753805e-05,
      "loss": 0.7715,
      "step": 1870
    },
    {
      "epoch": 5.7846153846153845,
      "grad_norm": 1.9930269718170166,
      "learning_rate": 7.559943742923626e-05,
      "loss": 0.8177,
      "step": 1880
    },
    {
      "epoch": 5.815384615384615,
      "grad_norm": 2.398965358734131,
      "learning_rate": 7.466316607649738e-05,
      "loss": 0.8063,
      "step": 1890
    },
    {
      "epoch": 5.846153846153846,
      "grad_norm": 2.19911527633667,
      "learning_rate": 7.372926218014131e-05,
      "loss": 0.7767,
      "step": 1900
    },
    {
      "epoch": 5.876923076923077,
      "grad_norm": 2.059420585632324,
      "learning_rate": 7.279781300350758e-05,
      "loss": 0.7907,
      "step": 1910
    },
    {
      "epoch": 5.907692307692308,
      "grad_norm": 2.0258114337921143,
      "learning_rate": 7.186890558056836e-05,
      "loss": 0.7923,
      "step": 1920
    },
    {
      "epoch": 5.938461538461539,
      "grad_norm": 1.8502027988433838,
      "learning_rate": 7.094262670779612e-05,
      "loss": 0.8067,
      "step": 1930
    },
    {
      "epoch": 5.969230769230769,
      "grad_norm": 2.0791757106781006,
      "learning_rate": 7.00190629360533e-05,
      "loss": 0.8026,
      "step": 1940
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.09051513671875,
      "learning_rate": 6.909830056250527e-05,
      "loss": 0.8339,
      "step": 1950
    },
    {
      "epoch": 6.030769230769231,
      "grad_norm": 1.9199817180633545,
      "learning_rate": 6.81804256225567e-05,
      "loss": 0.752,
      "step": 1960
    },
    {
      "epoch": 6.061538461538461,
      "grad_norm": 2.218315362930298,
      "learning_rate": 6.726552388181233e-05,
      "loss": 0.7989,
      "step": 1970
    },
    {
      "epoch": 6.092307692307692,
      "grad_norm": 1.7573823928833008,
      "learning_rate": 6.63536808280633e-05,
      "loss": 0.735,
      "step": 1980
    },
    {
      "epoch": 6.123076923076923,
      "grad_norm": 1.8894169330596924,
      "learning_rate": 6.544498166329913e-05,
      "loss": 0.7633,
      "step": 1990
    },
    {
      "epoch": 6.153846153846154,
      "grad_norm": 2.069439649581909,
      "learning_rate": 6.453951129574644e-05,
      "loss": 0.7882,
      "step": 2000
    },
    {
      "epoch": 6.184615384615385,
      "grad_norm": 2.2288105487823486,
      "learning_rate": 6.36373543319353e-05,
      "loss": 0.773,
      "step": 2010
    },
    {
      "epoch": 6.2153846153846155,
      "grad_norm": 2.023712396621704,
      "learning_rate": 6.273859506879365e-05,
      "loss": 0.753,
      "step": 2020
    },
    {
      "epoch": 6.246153846153846,
      "grad_norm": 2.0464038848876953,
      "learning_rate": 6.18433174857705e-05,
      "loss": 0.7616,
      "step": 2030
    },
    {
      "epoch": 6.276923076923077,
      "grad_norm": 2.011988878250122,
      "learning_rate": 6.095160523698913e-05,
      "loss": 0.7611,
      "step": 2040
    },
    {
      "epoch": 6.3076923076923075,
      "grad_norm": 2.3154494762420654,
      "learning_rate": 6.006354164343046e-05,
      "loss": 0.7943,
      "step": 2050
    },
    {
      "epoch": 6.338461538461538,
      "grad_norm": 2.325244903564453,
      "learning_rate": 5.917920968514752e-05,
      "loss": 0.7529,
      "step": 2060
    },
    {
      "epoch": 6.36923076923077,
      "grad_norm": 1.9565290212631226,
      "learning_rate": 5.829869199351188e-05,
      "loss": 0.7952,
      "step": 2070
    },
    {
      "epoch": 6.4,
      "grad_norm": 1.8974802494049072,
      "learning_rate": 5.7422070843492734e-05,
      "loss": 0.7621,
      "step": 2080
    },
    {
      "epoch": 6.430769230769231,
      "grad_norm": 2.1027376651763916,
      "learning_rate": 5.654942814596902e-05,
      "loss": 0.758,
      "step": 2090
    },
    {
      "epoch": 6.461538461538462,
      "grad_norm": 1.9659016132354736,
      "learning_rate": 5.568084544007588e-05,
      "loss": 0.8167,
      "step": 2100
    },
    {
      "epoch": 6.492307692307692,
      "grad_norm": 2.048273801803589,
      "learning_rate": 5.481640388558551e-05,
      "loss": 0.7517,
      "step": 2110
    },
    {
      "epoch": 6.523076923076923,
      "grad_norm": 2.1136772632598877,
      "learning_rate": 5.395618425532389e-05,
      "loss": 0.7406,
      "step": 2120
    },
    {
      "epoch": 6.553846153846154,
      "grad_norm": 2.157108783721924,
      "learning_rate": 5.3100266927623156e-05,
      "loss": 0.8148,
      "step": 2130
    },
    {
      "epoch": 6.584615384615384,
      "grad_norm": 2.106790781021118,
      "learning_rate": 5.2248731878811365e-05,
      "loss": 0.7784,
      "step": 2140
    },
    {
      "epoch": 6.615384615384615,
      "grad_norm": 1.9885175228118896,
      "learning_rate": 5.14016586757394e-05,
      "loss": 0.787,
      "step": 2150
    },
    {
      "epoch": 6.6461538461538465,
      "grad_norm": 2.2793591022491455,
      "learning_rate": 5.055912646834635e-05,
      "loss": 0.7643,
      "step": 2160
    },
    {
      "epoch": 6.676923076923077,
      "grad_norm": 1.9444079399108887,
      "learning_rate": 4.972121398226371e-05,
      "loss": 0.7571,
      "step": 2170
    },
    {
      "epoch": 6.707692307692308,
      "grad_norm": 2.114750385284424,
      "learning_rate": 4.888799951145948e-05,
      "loss": 0.7698,
      "step": 2180
    },
    {
      "epoch": 6.7384615384615385,
      "grad_norm": 1.8953299522399902,
      "learning_rate": 4.805956091092227e-05,
      "loss": 0.7318,
      "step": 2190
    },
    {
      "epoch": 6.769230769230769,
      "grad_norm": 2.02268385887146,
      "learning_rate": 4.723597558938672e-05,
      "loss": 0.7826,
      "step": 2200
    },
    {
      "epoch": 6.8,
      "grad_norm": 2.0683560371398926,
      "learning_rate": 4.6417320502100316e-05,
      "loss": 0.7363,
      "step": 2210
    },
    {
      "epoch": 6.8307692307692305,
      "grad_norm": 2.056417465209961,
      "learning_rate": 4.5603672143632944e-05,
      "loss": 0.7779,
      "step": 2220
    },
    {
      "epoch": 6.861538461538462,
      "grad_norm": 2.0059967041015625,
      "learning_rate": 4.479510654072909e-05,
      "loss": 0.8147,
      "step": 2230
    },
    {
      "epoch": 6.892307692307693,
      "grad_norm": 1.9279494285583496,
      "learning_rate": 4.399169924520403e-05,
      "loss": 0.7647,
      "step": 2240
    },
    {
      "epoch": 6.923076923076923,
      "grad_norm": 2.5176281929016113,
      "learning_rate": 4.3193525326884435e-05,
      "loss": 0.8265,
      "step": 2250
    },
    {
      "epoch": 6.953846153846154,
      "grad_norm": 1.9771342277526855,
      "learning_rate": 4.240065936659374e-05,
      "loss": 0.7755,
      "step": 2260
    },
    {
      "epoch": 6.984615384615385,
      "grad_norm": 2.1674654483795166,
      "learning_rate": 4.161317544918345e-05,
      "loss": 0.7682,
      "step": 2270
    },
    {
      "epoch": 7.015384615384615,
      "grad_norm": 1.9835222959518433,
      "learning_rate": 4.0831147156610684e-05,
      "loss": 0.7977,
      "step": 2280
    },
    {
      "epoch": 7.046153846153846,
      "grad_norm": 2.224958658218384,
      "learning_rate": 4.005464756106262e-05,
      "loss": 0.7346,
      "step": 2290
    },
    {
      "epoch": 7.076923076923077,
      "grad_norm": 2.0972986221313477,
      "learning_rate": 3.9283749218128885e-05,
      "loss": 0.7259,
      "step": 2300
    },
    {
      "epoch": 7.107692307692307,
      "grad_norm": 2.0841434001922607,
      "learning_rate": 3.851852416002187e-05,
      "loss": 0.7083,
      "step": 2310
    },
    {
      "epoch": 7.138461538461539,
      "grad_norm": 2.0690999031066895,
      "learning_rate": 3.775904388884618e-05,
      "loss": 0.7639,
      "step": 2320
    },
    {
      "epoch": 7.1692307692307695,
      "grad_norm": 2.232002019882202,
      "learning_rate": 3.7005379369917325e-05,
      "loss": 0.7578,
      "step": 2330
    },
    {
      "epoch": 7.2,
      "grad_norm": 2.051391363143921,
      "learning_rate": 3.6257601025131026e-05,
      "loss": 0.7454,
      "step": 2340
    },
    {
      "epoch": 7.230769230769231,
      "grad_norm": 2.056314468383789,
      "learning_rate": 3.5515778726382966e-05,
      "loss": 0.7374,
      "step": 2350
    },
    {
      "epoch": 7.2615384615384615,
      "grad_norm": 2.0361640453338623,
      "learning_rate": 3.477998178903982e-05,
      "loss": 0.7634,
      "step": 2360
    },
    {
      "epoch": 7.292307692307692,
      "grad_norm": 2.221268653869629,
      "learning_rate": 3.4050278965462764e-05,
      "loss": 0.7884,
      "step": 2370
    },
    {
      "epoch": 7.323076923076923,
      "grad_norm": 2.4283814430236816,
      "learning_rate": 3.3326738438583114e-05,
      "loss": 0.7827,
      "step": 2380
    },
    {
      "epoch": 7.3538461538461535,
      "grad_norm": 2.1344385147094727,
      "learning_rate": 3.2609427815531426e-05,
      "loss": 0.7281,
      "step": 2390
    },
    {
      "epoch": 7.384615384615385,
      "grad_norm": 2.1491506099700928,
      "learning_rate": 3.1898414121320276e-05,
      "loss": 0.7854,
      "step": 2400
    },
    {
      "epoch": 7.415384615384616,
      "grad_norm": 2.2601113319396973,
      "learning_rate": 3.11937637925816e-05,
      "loss": 0.7765,
      "step": 2410
    },
    {
      "epoch": 7.446153846153846,
      "grad_norm": 1.77631676197052,
      "learning_rate": 3.0495542671358746e-05,
      "loss": 0.7188,
      "step": 2420
    },
    {
      "epoch": 7.476923076923077,
      "grad_norm": 1.9240316152572632,
      "learning_rate": 2.9803815998954332e-05,
      "loss": 0.7751,
      "step": 2430
    },
    {
      "epoch": 7.507692307692308,
      "grad_norm": 1.9639230966567993,
      "learning_rate": 2.9118648409834205e-05,
      "loss": 0.7346,
      "step": 2440
    },
    {
      "epoch": 7.538461538461538,
      "grad_norm": 1.9350874423980713,
      "learning_rate": 2.84401039255879e-05,
      "loss": 0.7412,
      "step": 2450
    },
    {
      "epoch": 7.569230769230769,
      "grad_norm": 2.0040721893310547,
      "learning_rate": 2.7768245948946612e-05,
      "loss": 0.7466,
      "step": 2460
    },
    {
      "epoch": 7.6,
      "grad_norm": 1.894864559173584,
      "learning_rate": 2.7103137257858868e-05,
      "loss": 0.7722,
      "step": 2470
    },
    {
      "epoch": 7.63076923076923,
      "grad_norm": 2.5914828777313232,
      "learning_rate": 2.6444839999624494e-05,
      "loss": 0.7724,
      "step": 2480
    },
    {
      "epoch": 7.661538461538462,
      "grad_norm": 2.3115148544311523,
      "learning_rate": 2.5793415685087797e-05,
      "loss": 0.8039,
      "step": 2490
    },
    {
      "epoch": 7.6923076923076925,
      "grad_norm": 1.9655823707580566,
      "learning_rate": 2.514892518288988e-05,
      "loss": 0.7421,
      "step": 2500
    },
    {
      "epoch": 7.723076923076923,
      "grad_norm": 2.0706140995025635,
      "learning_rate": 2.4511428713781238e-05,
      "loss": 0.7631,
      "step": 2510
    },
    {
      "epoch": 7.753846153846154,
      "grad_norm": 1.910049557685852,
      "learning_rate": 2.3880985844994674e-05,
      "loss": 0.7399,
      "step": 2520
    },
    {
      "epoch": 7.7846153846153845,
      "grad_norm": 2.1228690147399902,
      "learning_rate": 2.3257655484679374e-05,
      "loss": 0.7572,
      "step": 2530
    },
    {
      "epoch": 7.815384615384615,
      "grad_norm": 2.168492555618286,
      "learning_rate": 2.2641495876396713e-05,
      "loss": 0.7703,
      "step": 2540
    },
    {
      "epoch": 7.846153846153846,
      "grad_norm": 1.8387569189071655,
      "learning_rate": 2.2032564593677774e-05,
      "loss": 0.7217,
      "step": 2550
    },
    {
      "epoch": 7.876923076923077,
      "grad_norm": 2.0391128063201904,
      "learning_rate": 2.1430918534643996e-05,
      "loss": 0.752,
      "step": 2560
    },
    {
      "epoch": 7.907692307692308,
      "grad_norm": 2.573230743408203,
      "learning_rate": 2.0836613916690428e-05,
      "loss": 0.7175,
      "step": 2570
    },
    {
      "epoch": 7.938461538461539,
      "grad_norm": 2.0587637424468994,
      "learning_rate": 2.024970627123295e-05,
      "loss": 0.8003,
      "step": 2580
    },
    {
      "epoch": 7.969230769230769,
      "grad_norm": 2.126275062561035,
      "learning_rate": 1.967025043851939e-05,
      "loss": 0.7484,
      "step": 2590
    },
    {
      "epoch": 8.0,
      "grad_norm": 2.07289719581604,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 0.7706,
      "step": 2600
    },
    {
      "epoch": 8.03076923076923,
      "grad_norm": 2.0858993530273438,
      "learning_rate": 1.8533910085794713e-05,
      "loss": 0.7456,
      "step": 2610
    },
    {
      "epoch": 8.061538461538461,
      "grad_norm": 2.192556619644165,
      "learning_rate": 1.7977131744646724e-05,
      "loss": 0.7149,
      "step": 2620
    },
    {
      "epoch": 8.092307692307692,
      "grad_norm": 1.8930575847625732,
      "learning_rate": 1.7428017564047594e-05,
      "loss": 0.7205,
      "step": 2630
    },
    {
      "epoch": 8.123076923076923,
      "grad_norm": 2.337496042251587,
      "learning_rate": 1.6886618852849724e-05,
      "loss": 0.7782,
      "step": 2640
    },
    {
      "epoch": 8.153846153846153,
      "grad_norm": 2.018669366836548,
      "learning_rate": 1.6352986198977325e-05,
      "loss": 0.7889,
      "step": 2650
    },
    {
      "epoch": 8.184615384615384,
      "grad_norm": 2.315338134765625,
      "learning_rate": 1.5827169464699576e-05,
      "loss": 0.7248,
      "step": 2660
    },
    {
      "epoch": 8.215384615384615,
      "grad_norm": 2.2808115482330322,
      "learning_rate": 1.530921778197142e-05,
      "loss": 0.7778,
      "step": 2670
    },
    {
      "epoch": 8.246153846153845,
      "grad_norm": 2.2129008769989014,
      "learning_rate": 1.4799179547842822e-05,
      "loss": 0.7097,
      "step": 2680
    },
    {
      "epoch": 8.276923076923078,
      "grad_norm": 2.071539878845215,
      "learning_rate": 1.429710241993656e-05,
      "loss": 0.7472,
      "step": 2690
    },
    {
      "epoch": 8.307692307692308,
      "grad_norm": 2.134145736694336,
      "learning_rate": 1.3803033311995072e-05,
      "loss": 0.7591,
      "step": 2700
    },
    {
      "epoch": 8.338461538461539,
      "grad_norm": 2.066234827041626,
      "learning_rate": 1.3317018389496927e-05,
      "loss": 0.7537,
      "step": 2710
    },
    {
      "epoch": 8.36923076923077,
      "grad_norm": 2.0239179134368896,
      "learning_rate": 1.2839103065343083e-05,
      "loss": 0.8152,
      "step": 2720
    },
    {
      "epoch": 8.4,
      "grad_norm": 2.076111316680908,
      "learning_rate": 1.2369331995613665e-05,
      "loss": 0.7242,
      "step": 2730
    },
    {
      "epoch": 8.430769230769231,
      "grad_norm": 2.0224554538726807,
      "learning_rate": 1.1907749075395147e-05,
      "loss": 0.7407,
      "step": 2740
    },
    {
      "epoch": 8.461538461538462,
      "grad_norm": 2.344095230102539,
      "learning_rate": 1.1454397434679021e-05,
      "loss": 0.8089,
      "step": 2750
    },
    {
      "epoch": 8.492307692307692,
      "grad_norm": 2.1623709201812744,
      "learning_rate": 1.1009319434331622e-05,
      "loss": 0.7209,
      "step": 2760
    },
    {
      "epoch": 8.523076923076923,
      "grad_norm": 2.238278865814209,
      "learning_rate": 1.0572556662136035e-05,
      "loss": 0.6911,
      "step": 2770
    },
    {
      "epoch": 8.553846153846154,
      "grad_norm": 2.4050676822662354,
      "learning_rate": 1.014414992890611e-05,
      "loss": 0.7138,
      "step": 2780
    },
    {
      "epoch": 8.584615384615384,
      "grad_norm": 2.0228419303894043,
      "learning_rate": 9.724139264673116e-06,
      "loss": 0.7452,
      "step": 2790
    },
    {
      "epoch": 8.615384615384615,
      "grad_norm": 2.2917470932006836,
      "learning_rate": 9.31256391494546e-06,
      "loss": 0.6921,
      "step": 2800
    },
    {
      "epoch": 8.646153846153846,
      "grad_norm": 1.8075125217437744,
      "learning_rate": 8.909462337041507e-06,
      "loss": 0.734,
      "step": 2810
    },
    {
      "epoch": 8.676923076923076,
      "grad_norm": 2.01468825340271,
      "learning_rate": 8.514872196496183e-06,
      "loss": 0.7335,
      "step": 2820
    },
    {
      "epoch": 8.707692307692307,
      "grad_norm": 2.298112630844116,
      "learning_rate": 8.128830363541574e-06,
      "loss": 0.7094,
      "step": 2830
    },
    {
      "epoch": 8.73846153846154,
      "grad_norm": 2.1774020195007324,
      "learning_rate": 7.751372909661769e-06,
      "loss": 0.8051,
      "step": 2840
    },
    {
      "epoch": 8.76923076923077,
      "grad_norm": 2.1409170627593994,
      "learning_rate": 7.382535104222366e-06,
      "loss": 0.7679,
      "step": 2850
    },
    {
      "epoch": 8.8,
      "grad_norm": 2.082650899887085,
      "learning_rate": 7.022351411174866e-06,
      "loss": 0.7277,
      "step": 2860
    },
    {
      "epoch": 8.830769230769231,
      "grad_norm": 2.0364997386932373,
      "learning_rate": 6.670855485836525e-06,
      "loss": 0.7121,
      "step": 2870
    },
    {
      "epoch": 8.861538461538462,
      "grad_norm": 2.206631898880005,
      "learning_rate": 6.32808017174551e-06,
      "loss": 0.741,
      "step": 2880
    },
    {
      "epoch": 8.892307692307693,
      "grad_norm": 2.1782455444335938,
      "learning_rate": 5.994057497592031e-06,
      "loss": 0.7468,
      "step": 2890
    },
    {
      "epoch": 8.923076923076923,
      "grad_norm": 2.374114751815796,
      "learning_rate": 5.668818674225685e-06,
      "loss": 0.7532,
      "step": 2900
    },
    {
      "epoch": 8.953846153846154,
      "grad_norm": 2.0386927127838135,
      "learning_rate": 5.3523940917390215e-06,
      "loss": 0.7524,
      "step": 2910
    },
    {
      "epoch": 8.984615384615385,
      "grad_norm": 1.9349490404129028,
      "learning_rate": 5.0448133166279944e-06,
      "loss": 0.7254,
      "step": 2920
    },
    {
      "epoch": 9.015384615384615,
      "grad_norm": 2.218142032623291,
      "learning_rate": 4.746105089029229e-06,
      "loss": 0.744,
      "step": 2930
    },
    {
      "epoch": 9.046153846153846,
      "grad_norm": 2.208949565887451,
      "learning_rate": 4.4562973200346416e-06,
      "loss": 0.7338,
      "step": 2940
    },
    {
      "epoch": 9.076923076923077,
      "grad_norm": 1.8909201622009277,
      "learning_rate": 4.175417089083378e-06,
      "loss": 0.7232,
      "step": 2950
    },
    {
      "epoch": 9.107692307692307,
      "grad_norm": 2.163405179977417,
      "learning_rate": 3.903490641431573e-06,
      "loss": 0.7349,
      "step": 2960
    },
    {
      "epoch": 9.138461538461538,
      "grad_norm": 2.3600995540618896,
      "learning_rate": 3.6405433856999684e-06,
      "loss": 0.7678,
      "step": 2970
    },
    {
      "epoch": 9.169230769230769,
      "grad_norm": 2.188930034637451,
      "learning_rate": 3.3865998914997643e-06,
      "loss": 0.7418,
      "step": 2980
    },
    {
      "epoch": 9.2,
      "grad_norm": 2.2357707023620605,
      "learning_rate": 3.1416838871368924e-06,
      "loss": 0.7426,
      "step": 2990
    },
    {
      "epoch": 9.23076923076923,
      "grad_norm": 1.9671388864517212,
      "learning_rate": 2.905818257394799e-06,
      "loss": 0.7298,
      "step": 3000
    },
    {
      "epoch": 9.261538461538462,
      "grad_norm": 2.478001594543457,
      "learning_rate": 2.679025041396155e-06,
      "loss": 0.7185,
      "step": 3010
    },
    {
      "epoch": 9.292307692307693,
      "grad_norm": 2.290837287902832,
      "learning_rate": 2.461325430543482e-06,
      "loss": 0.7444,
      "step": 3020
    },
    {
      "epoch": 9.323076923076924,
      "grad_norm": 2.3114941120147705,
      "learning_rate": 2.2527397665391027e-06,
      "loss": 0.7101,
      "step": 3030
    },
    {
      "epoch": 9.353846153846154,
      "grad_norm": 1.9803047180175781,
      "learning_rate": 2.053287539484405e-06,
      "loss": 0.6882,
      "step": 3040
    },
    {
      "epoch": 9.384615384615385,
      "grad_norm": 2.2022945880889893,
      "learning_rate": 1.8629873860586566e-06,
      "loss": 0.6975,
      "step": 3050
    },
    {
      "epoch": 9.415384615384616,
      "grad_norm": 1.9524860382080078,
      "learning_rate": 1.6818570877776718e-06,
      "loss": 0.6846,
      "step": 3060
    },
    {
      "epoch": 9.446153846153846,
      "grad_norm": 2.2137529850006104,
      "learning_rate": 1.5099135693322774e-06,
      "loss": 0.7273,
      "step": 3070
    },
    {
      "epoch": 9.476923076923077,
      "grad_norm": 2.1356773376464844,
      "learning_rate": 1.3471728970068987e-06,
      "loss": 0.7458,
      "step": 3080
    },
    {
      "epoch": 9.507692307692308,
      "grad_norm": 2.098583221435547,
      "learning_rate": 1.1936502771783486e-06,
      "loss": 0.7143,
      "step": 3090
    },
    {
      "epoch": 9.538461538461538,
      "grad_norm": 2.1301488876342773,
      "learning_rate": 1.0493600548948878e-06,
      "loss": 0.7496,
      "step": 3100
    },
    {
      "epoch": 9.569230769230769,
      "grad_norm": 2.2320845127105713,
      "learning_rate": 9.143157125359514e-07,
      "loss": 0.7591,
      "step": 3110
    },
    {
      "epoch": 9.6,
      "grad_norm": 2.2147562503814697,
      "learning_rate": 7.885298685522235e-07,
      "loss": 0.7303,
      "step": 3120
    },
    {
      "epoch": 9.63076923076923,
      "grad_norm": 2.2397658824920654,
      "learning_rate": 6.720142762867032e-07,
      "loss": 0.7555,
      "step": 3130
    },
    {
      "epoch": 9.661538461538461,
      "grad_norm": 2.177968740463257,
      "learning_rate": 5.647798228764156e-07,
      "loss": 0.7479,
      "step": 3140
    },
    {
      "epoch": 9.692307692307692,
      "grad_norm": 2.243819236755371,
      "learning_rate": 4.668365282351372e-07,
      "loss": 0.7534,
      "step": 3150
    },
    {
      "epoch": 9.723076923076922,
      "grad_norm": 2.2024483680725098,
      "learning_rate": 3.781935441171336e-07,
      "loss": 0.7508,
      "step": 3160
    },
    {
      "epoch": 9.753846153846155,
      "grad_norm": 2.2095537185668945,
      "learning_rate": 2.988591532620322e-07,
      "loss": 0.7463,
      "step": 3170
    },
    {
      "epoch": 9.784615384615385,
      "grad_norm": 2.1779325008392334,
      "learning_rate": 2.288407686208971e-07,
      "loss": 0.7814,
      "step": 3180
    },
    {
      "epoch": 9.815384615384616,
      "grad_norm": 2.161087989807129,
      "learning_rate": 1.68144932663572e-07,
      "loss": 0.7342,
      "step": 3190
    },
    {
      "epoch": 9.846153846153847,
      "grad_norm": 2.0373594760894775,
      "learning_rate": 1.1677731676733584e-07,
      "loss": 0.7054,
      "step": 3200
    },
    {
      "epoch": 9.876923076923077,
      "grad_norm": 2.077526330947876,
      "learning_rate": 7.474272068698218e-08,
      "loss": 0.7355,
      "step": 3210
    },
    {
      "epoch": 9.907692307692308,
      "grad_norm": 2.0211634635925293,
      "learning_rate": 4.2045072106333684e-08,
      "loss": 0.763,
      "step": 3220
    },
    {
      "epoch": 9.938461538461539,
      "grad_norm": 2.205732822418213,
      "learning_rate": 1.8687426271246645e-08,
      "loss": 0.7732,
      "step": 3230
    },
    {
      "epoch": 9.96923076923077,
      "grad_norm": 2.1264724731445312,
      "learning_rate": 4.6719657041283114e-09,
      "loss": 0.7328,
      "step": 3240
    },
    {
      "epoch": 10.0,
      "grad_norm": 2.179966449737549,
      "learning_rate": 0.0,
      "loss": 0.7768,
      "step": 3250
    }
  ],
  "logging_steps": 10,
  "max_steps": 3250,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2869986411741184e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}