{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 1000,
  "global_step": 4425,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011299435028248588,
      "grad_norm": 0.7552605867385864,
      "learning_rate": 4.5146726862302485e-06,
      "loss": 1.0465,
      "step": 10
    },
    {
      "epoch": 0.022598870056497175,
      "grad_norm": 0.8854597210884094,
      "learning_rate": 9.029345372460497e-06,
      "loss": 1.0939,
      "step": 20
    },
    {
      "epoch": 0.03389830508474576,
      "grad_norm": 0.7813611030578613,
      "learning_rate": 1.3544018058690747e-05,
      "loss": 1.0292,
      "step": 30
    },
    {
      "epoch": 0.04519774011299435,
      "grad_norm": 0.6397517919540405,
      "learning_rate": 1.8058690744920994e-05,
      "loss": 0.995,
      "step": 40
    },
    {
      "epoch": 0.05649717514124294,
      "grad_norm": 0.4727611839771271,
      "learning_rate": 2.2573363431151244e-05,
      "loss": 0.896,
      "step": 50
    },
    {
      "epoch": 0.06779661016949153,
      "grad_norm": 0.37453365325927734,
      "learning_rate": 2.7088036117381494e-05,
      "loss": 0.8037,
      "step": 60
    },
    {
      "epoch": 0.07909604519774012,
      "grad_norm": 0.5303831696510315,
      "learning_rate": 3.1602708803611745e-05,
      "loss": 0.7872,
      "step": 70
    },
    {
      "epoch": 0.0903954802259887,
      "grad_norm": 0.3839394152164459,
      "learning_rate": 3.611738148984199e-05,
      "loss": 0.7697,
      "step": 80
    },
    {
      "epoch": 0.1016949152542373,
      "grad_norm": 0.39426788687705994,
      "learning_rate": 4.063205417607224e-05,
      "loss": 0.7481,
      "step": 90
    },
    {
      "epoch": 0.11299435028248588,
      "grad_norm": 0.29685771465301514,
      "learning_rate": 4.514672686230249e-05,
      "loss": 0.7166,
      "step": 100
    },
    {
      "epoch": 0.12429378531073447,
      "grad_norm": 0.3550776243209839,
      "learning_rate": 4.966139954853273e-05,
      "loss": 0.7455,
      "step": 110
    },
    {
      "epoch": 0.13559322033898305,
      "grad_norm": 0.49571502208709717,
      "learning_rate": 5.417607223476299e-05,
      "loss": 0.7451,
      "step": 120
    },
    {
      "epoch": 0.14689265536723164,
      "grad_norm": 0.3369849622249603,
      "learning_rate": 5.869074492099323e-05,
      "loss": 0.7235,
      "step": 130
    },
    {
      "epoch": 0.15819209039548024,
      "grad_norm": 0.37830305099487305,
      "learning_rate": 6.320541760722349e-05,
      "loss": 0.7201,
      "step": 140
    },
    {
      "epoch": 0.1694915254237288,
      "grad_norm": 0.42834165692329407,
      "learning_rate": 6.772009029345373e-05,
      "loss": 0.7199,
      "step": 150
    },
    {
      "epoch": 0.1807909604519774,
      "grad_norm": 0.3480307459831238,
      "learning_rate": 7.223476297968398e-05,
      "loss": 0.7134,
      "step": 160
    },
    {
      "epoch": 0.192090395480226,
      "grad_norm": 0.47217118740081787,
      "learning_rate": 7.674943566591423e-05,
      "loss": 0.7193,
      "step": 170
    },
    {
      "epoch": 0.2033898305084746,
      "grad_norm": 0.43827682733535767,
      "learning_rate": 8.126410835214448e-05,
      "loss": 0.7049,
      "step": 180
    },
    {
      "epoch": 0.21468926553672316,
      "grad_norm": 0.38405266404151917,
      "learning_rate": 8.577878103837473e-05,
      "loss": 0.719,
      "step": 190
    },
    {
      "epoch": 0.22598870056497175,
      "grad_norm": 0.48938581347465515,
      "learning_rate": 9.029345372460498e-05,
      "loss": 0.7077,
      "step": 200
    },
    {
      "epoch": 0.23728813559322035,
      "grad_norm": 0.4330425560474396,
      "learning_rate": 9.480812641083521e-05,
      "loss": 0.7301,
      "step": 210
    },
    {
      "epoch": 0.24858757062146894,
      "grad_norm": 0.3442066013813019,
      "learning_rate": 9.932279909706546e-05,
      "loss": 0.7174,
      "step": 220
    },
    {
      "epoch": 0.2598870056497175,
      "grad_norm": 0.41479647159576416,
      "learning_rate": 0.00010383747178329573,
      "loss": 0.7163,
      "step": 230
    },
    {
      "epoch": 0.2711864406779661,
      "grad_norm": 0.4269746243953705,
      "learning_rate": 0.00010835214446952598,
      "loss": 0.7206,
      "step": 240
    },
    {
      "epoch": 0.2824858757062147,
      "grad_norm": 0.4595146179199219,
      "learning_rate": 0.00011286681715575623,
      "loss": 0.7198,
      "step": 250
    },
    {
      "epoch": 0.2937853107344633,
      "grad_norm": 0.459258496761322,
      "learning_rate": 0.00011738148984198646,
      "loss": 0.6791,
      "step": 260
    },
    {
      "epoch": 0.3050847457627119,
      "grad_norm": 0.40073147416114807,
      "learning_rate": 0.00012189616252821671,
      "loss": 0.7379,
      "step": 270
    },
    {
      "epoch": 0.3163841807909605,
      "grad_norm": 0.4646053910255432,
      "learning_rate": 0.00012641083521444698,
      "loss": 0.7146,
      "step": 280
    },
    {
      "epoch": 0.327683615819209,
      "grad_norm": 0.5096935629844666,
      "learning_rate": 0.00013092550790067722,
      "loss": 0.6855,
      "step": 290
    },
    {
      "epoch": 0.3389830508474576,
      "grad_norm": 0.37696900963783264,
      "learning_rate": 0.00013544018058690745,
      "loss": 0.7169,
      "step": 300
    },
    {
      "epoch": 0.3502824858757062,
      "grad_norm": 0.344178169965744,
      "learning_rate": 0.00013995485327313772,
      "loss": 0.7079,
      "step": 310
    },
    {
      "epoch": 0.3615819209039548,
      "grad_norm": 0.33671554923057556,
      "learning_rate": 0.00014446952595936795,
      "loss": 0.68,
      "step": 320
    },
    {
      "epoch": 0.3728813559322034,
      "grad_norm": 0.365024596452713,
      "learning_rate": 0.00014898419864559822,
      "loss": 0.7247,
      "step": 330
    },
    {
      "epoch": 0.384180790960452,
      "grad_norm": 0.36559364199638367,
      "learning_rate": 0.00015349887133182845,
      "loss": 0.6753,
      "step": 340
    },
    {
      "epoch": 0.3954802259887006,
      "grad_norm": 0.3757419288158417,
      "learning_rate": 0.0001580135440180587,
      "loss": 0.6616,
      "step": 350
    },
    {
      "epoch": 0.4067796610169492,
      "grad_norm": 0.5549542307853699,
      "learning_rate": 0.00016252821670428895,
      "loss": 0.686,
      "step": 360
    },
    {
      "epoch": 0.4180790960451977,
      "grad_norm": 0.3681774139404297,
      "learning_rate": 0.0001670428893905192,
      "loss": 0.7086,
      "step": 370
    },
    {
      "epoch": 0.4293785310734463,
      "grad_norm": 0.3036465346813202,
      "learning_rate": 0.00017155756207674945,
      "loss": 0.6966,
      "step": 380
    },
    {
      "epoch": 0.4406779661016949,
      "grad_norm": 0.32314741611480713,
      "learning_rate": 0.0001760722347629797,
      "loss": 0.7406,
      "step": 390
    },
    {
      "epoch": 0.4519774011299435,
      "grad_norm": 0.41190987825393677,
      "learning_rate": 0.00018058690744920995,
      "loss": 0.6772,
      "step": 400
    },
    {
      "epoch": 0.4632768361581921,
      "grad_norm": 0.28181225061416626,
      "learning_rate": 0.0001851015801354402,
      "loss": 0.6843,
      "step": 410
    },
    {
      "epoch": 0.4745762711864407,
      "grad_norm": 0.2925048768520355,
      "learning_rate": 0.00018961625282167043,
      "loss": 0.7206,
      "step": 420
    },
    {
      "epoch": 0.4858757062146893,
      "grad_norm": 0.33646613359451294,
      "learning_rate": 0.0001941309255079007,
      "loss": 0.6879,
      "step": 430
    },
    {
      "epoch": 0.4971751412429379,
      "grad_norm": 0.40007829666137695,
      "learning_rate": 0.00019864559819413093,
      "loss": 0.7032,
      "step": 440
    },
    {
      "epoch": 0.5084745762711864,
      "grad_norm": 0.29928356409072876,
      "learning_rate": 0.00019999847502678901,
      "loss": 0.7086,
      "step": 450
    },
    {
      "epoch": 0.519774011299435,
      "grad_norm": 0.285306453704834,
      "learning_rate": 0.00019999100588221217,
      "loss": 0.6709,
      "step": 460
    },
    {
      "epoch": 0.5310734463276836,
      "grad_norm": 0.5055399537086487,
      "learning_rate": 0.00019997731293347473,
      "loss": 0.6901,
      "step": 470
    },
    {
      "epoch": 0.5423728813559322,
      "grad_norm": 0.3232295513153076,
      "learning_rate": 0.00019995739703287558,
      "loss": 0.6748,
      "step": 480
    },
    {
      "epoch": 0.5536723163841808,
      "grad_norm": 0.40595105290412903,
      "learning_rate": 0.00019993125942005297,
      "loss": 0.6908,
      "step": 490
    },
    {
      "epoch": 0.5649717514124294,
      "grad_norm": 0.29886409640312195,
      "learning_rate": 0.00019989890172190697,
      "loss": 0.6961,
      "step": 500
    },
    {
      "epoch": 0.576271186440678,
      "grad_norm": 0.3396940231323242,
      "learning_rate": 0.00019986032595249855,
      "loss": 0.6741,
      "step": 510
    },
    {
      "epoch": 0.5875706214689266,
      "grad_norm": 0.3772926926612854,
      "learning_rate": 0.00019981553451292396,
      "loss": 0.6681,
      "step": 520
    },
    {
      "epoch": 0.5988700564971752,
      "grad_norm": 0.3138965964317322,
      "learning_rate": 0.00019976453019116547,
      "loss": 0.7027,
      "step": 530
    },
    {
      "epoch": 0.6101694915254238,
      "grad_norm": 0.31275707483291626,
      "learning_rate": 0.00019970731616191772,
      "loss": 0.6955,
      "step": 540
    },
    {
      "epoch": 0.6214689265536724,
      "grad_norm": 0.28057295083999634,
      "learning_rate": 0.00019964389598639012,
      "loss": 0.6617,
      "step": 550
    },
    {
      "epoch": 0.632768361581921,
      "grad_norm": 0.5602757930755615,
      "learning_rate": 0.00019957427361208522,
      "loss": 0.6896,
      "step": 560
    },
    {
      "epoch": 0.6440677966101694,
      "grad_norm": 0.43884241580963135,
      "learning_rate": 0.00019949845337255306,
      "loss": 0.6786,
      "step": 570
    },
    {
      "epoch": 0.655367231638418,
      "grad_norm": 0.3063630759716034,
      "learning_rate": 0.00019941643998712128,
      "loss": 0.685,
      "step": 580
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.2912139594554901,
      "learning_rate": 0.00019932823856060153,
      "loss": 0.691,
      "step": 590
    },
    {
      "epoch": 0.6779661016949152,
      "grad_norm": 0.5503478646278381,
      "learning_rate": 0.00019923385458297168,
      "loss": 0.7076,
      "step": 600
    },
    {
      "epoch": 0.6892655367231638,
      "grad_norm": 0.30678126215934753,
      "learning_rate": 0.00019913329392903396,
      "loss": 0.6628,
      "step": 610
    },
    {
      "epoch": 0.7005649717514124,
      "grad_norm": 0.341691792011261,
      "learning_rate": 0.00019902656285804954,
      "loss": 0.6973,
      "step": 620
    },
    {
      "epoch": 0.711864406779661,
      "grad_norm": 0.2795998454093933,
      "learning_rate": 0.0001989136680133488,
      "loss": 0.668,
      "step": 630
    },
    {
      "epoch": 0.7231638418079096,
      "grad_norm": 0.3217742443084717,
      "learning_rate": 0.00019879461642191774,
      "loss": 0.6839,
      "step": 640
    },
    {
      "epoch": 0.7344632768361582,
      "grad_norm": 0.2668191194534302,
      "learning_rate": 0.00019866941549396075,
      "loss": 0.6929,
      "step": 650
    },
    {
      "epoch": 0.7457627118644068,
      "grad_norm": 0.8005201816558838,
      "learning_rate": 0.0001985380730224393,
      "loss": 0.701,
      "step": 660
    },
    {
      "epoch": 0.7570621468926554,
      "grad_norm": 0.29280129075050354,
      "learning_rate": 0.00019840059718258693,
      "loss": 0.7222,
      "step": 670
    },
    {
      "epoch": 0.768361581920904,
      "grad_norm": 0.3449852764606476,
      "learning_rate": 0.00019825699653140032,
      "loss": 0.6919,
      "step": 680
    },
    {
      "epoch": 0.7796610169491526,
      "grad_norm": 0.2684054970741272,
      "learning_rate": 0.0001981072800071066,
      "loss": 0.6859,
      "step": 690
    },
    {
      "epoch": 0.7909604519774012,
      "grad_norm": 0.3716861307621002,
      "learning_rate": 0.00019795145692860726,
      "loss": 0.666,
      "step": 700
    },
    {
      "epoch": 0.8022598870056498,
      "grad_norm": 0.3001028597354889,
      "learning_rate": 0.00019778953699489785,
      "loss": 0.6587,
      "step": 710
    },
    {
      "epoch": 0.8135593220338984,
      "grad_norm": 0.476034551858902,
      "learning_rate": 0.0001976215302844644,
      "loss": 0.6766,
      "step": 720
    },
    {
      "epoch": 0.8248587570621468,
      "grad_norm": 0.3516661524772644,
      "learning_rate": 0.00019744744725465617,
      "loss": 0.6836,
      "step": 730
    },
    {
      "epoch": 0.8361581920903954,
      "grad_norm": 0.275155246257782,
      "learning_rate": 0.00019726729874103448,
      "loss": 0.6729,
      "step": 740
    },
    {
      "epoch": 0.847457627118644,
      "grad_norm": 0.31217560172080994,
      "learning_rate": 0.00019708109595669858,
      "loss": 0.6811,
      "step": 750
    },
    {
      "epoch": 0.8587570621468926,
      "grad_norm": 0.3042563199996948,
      "learning_rate": 0.00019688885049158752,
      "loss": 0.7353,
      "step": 760
    },
    {
      "epoch": 0.8700564971751412,
      "grad_norm": 0.3564906120300293,
      "learning_rate": 0.0001966905743117588,
      "loss": 0.6938,
      "step": 770
    },
    {
      "epoch": 0.8813559322033898,
      "grad_norm": 0.3310820162296295,
      "learning_rate": 0.00019648627975864355,
      "loss": 0.6724,
      "step": 780
    },
    {
      "epoch": 0.8926553672316384,
      "grad_norm": 0.4258658289909363,
      "learning_rate": 0.00019627597954827833,
      "loss": 0.6901,
      "step": 790
    },
    {
      "epoch": 0.903954802259887,
      "grad_norm": 0.3146076202392578,
      "learning_rate": 0.00019605968677051378,
      "loss": 0.7008,
      "step": 800
    },
    {
      "epoch": 0.9152542372881356,
      "grad_norm": 0.3338530957698822,
      "learning_rate": 0.0001958374148881996,
      "loss": 0.6567,
      "step": 810
    },
    {
      "epoch": 0.9265536723163842,
      "grad_norm": 0.3822636008262634,
      "learning_rate": 0.00019560917773634676,
      "loss": 0.6756,
      "step": 820
    },
    {
      "epoch": 0.9378531073446328,
      "grad_norm": 0.3074444830417633,
      "learning_rate": 0.00019537498952126642,
      "loss": 0.6958,
      "step": 830
    },
    {
      "epoch": 0.9491525423728814,
      "grad_norm": 0.31832125782966614,
      "learning_rate": 0.00019513486481968545,
      "loss": 0.6871,
      "step": 840
    },
    {
      "epoch": 0.96045197740113,
      "grad_norm": 0.3670668303966522,
      "learning_rate": 0.00019488881857783935,
      "loss": 0.6956,
      "step": 850
    },
    {
      "epoch": 0.9717514124293786,
      "grad_norm": 0.2784234881401062,
      "learning_rate": 0.00019463686611054172,
      "loss": 0.6865,
      "step": 860
    },
    {
      "epoch": 0.9830508474576272,
      "grad_norm": 0.3210557997226715,
      "learning_rate": 0.00019437902310023126,
      "loss": 0.6845,
      "step": 870
    },
    {
      "epoch": 0.9943502824858758,
      "grad_norm": 0.3313763737678528,
      "learning_rate": 0.0001941153055959954,
      "loss": 0.6589,
      "step": 880
    },
    {
      "epoch": 1.0056497175141244,
      "grad_norm": 0.35079240798950195,
      "learning_rate": 0.00019384573001257155,
      "loss": 0.671,
      "step": 890
    },
    {
      "epoch": 1.0169491525423728,
      "grad_norm": 0.3715152442455292,
      "learning_rate": 0.0001935703131293252,
      "loss": 0.608,
      "step": 900
    },
    {
      "epoch": 1.0282485875706215,
      "grad_norm": 0.3646804988384247,
      "learning_rate": 0.00019328907208920567,
      "loss": 0.6457,
      "step": 910
    },
    {
      "epoch": 1.03954802259887,
      "grad_norm": 0.3106479048728943,
      "learning_rate": 0.00019300202439767898,
      "loss": 0.6105,
      "step": 920
    },
    {
      "epoch": 1.0508474576271187,
      "grad_norm": 0.29659685492515564,
      "learning_rate": 0.00019270918792163827,
      "loss": 0.6221,
      "step": 930
    },
    {
      "epoch": 1.0621468926553672,
      "grad_norm": 0.4973548948764801,
      "learning_rate": 0.00019241058088829166,
      "loss": 0.6465,
      "step": 940
    },
    {
      "epoch": 1.073446327683616,
      "grad_norm": 0.4067288637161255,
      "learning_rate": 0.00019210622188402788,
      "loss": 0.6374,
      "step": 950
    },
    {
      "epoch": 1.0847457627118644,
      "grad_norm": 0.4106499254703522,
      "learning_rate": 0.00019179612985325908,
      "loss": 0.6226,
      "step": 960
    },
    {
      "epoch": 1.0960451977401129,
      "grad_norm": 0.31531280279159546,
      "learning_rate": 0.000191480324097242,
      "loss": 0.6604,
      "step": 970
    },
    {
      "epoch": 1.1073446327683616,
      "grad_norm": 0.3855787515640259,
      "learning_rate": 0.00019115882427287638,
      "loss": 0.6466,
      "step": 980
    },
    {
      "epoch": 1.11864406779661,
      "grad_norm": 0.36246171593666077,
      "learning_rate": 0.00019083165039148153,
      "loss": 0.6315,
      "step": 990
    },
    {
      "epoch": 1.1299435028248588,
      "grad_norm": 0.4511447846889496,
      "learning_rate": 0.00019049882281755066,
      "loss": 0.6341,
      "step": 1000
    },
    {
      "epoch": 1.1299435028248588,
      "eval_loss": 0.7100452184677124,
      "eval_runtime": 16.7354,
      "eval_samples_per_second": 89.093,
      "eval_steps_per_second": 11.174,
      "step": 1000
    },
    {
      "epoch": 1.1412429378531073,
      "grad_norm": 0.39097025990486145,
      "learning_rate": 0.00019016036226748346,
      "loss": 0.6144,
      "step": 1010
    },
    {
      "epoch": 1.152542372881356,
      "grad_norm": 0.40734922885894775,
      "learning_rate": 0.00018981628980829652,
      "loss": 0.6458,
      "step": 1020
    },
    {
      "epoch": 1.1638418079096045,
      "grad_norm": 0.3473665118217468,
      "learning_rate": 0.00018946662685631203,
      "loss": 0.6417,
      "step": 1030
    },
    {
      "epoch": 1.1751412429378532,
      "grad_norm": 0.41758203506469727,
      "learning_rate": 0.0001891113951758249,
      "loss": 0.6355,
      "step": 1040
    },
    {
      "epoch": 1.1864406779661016,
      "grad_norm": 0.6268119215965271,
      "learning_rate": 0.0001887506168777479,
      "loss": 0.6332,
      "step": 1050
    },
    {
      "epoch": 1.1977401129943503,
      "grad_norm": 0.3336067795753479,
      "learning_rate": 0.00018838431441823547,
      "loss": 0.6445,
      "step": 1060
    },
    {
      "epoch": 1.2090395480225988,
      "grad_norm": 0.3316787779331207,
      "learning_rate": 0.00018801251059728604,
      "loss": 0.6314,
      "step": 1070
    },
    {
      "epoch": 1.2203389830508475,
      "grad_norm": 0.304683119058609,
      "learning_rate": 0.00018763522855732266,
      "loss": 0.6232,
      "step": 1080
    },
    {
      "epoch": 1.231638418079096,
      "grad_norm": 0.34174299240112305,
      "learning_rate": 0.00018725249178175286,
      "loss": 0.6561,
      "step": 1090
    },
    {
      "epoch": 1.2429378531073447,
      "grad_norm": 0.3202422559261322,
      "learning_rate": 0.0001868643240935066,
      "loss": 0.6066,
      "step": 1100
    },
    {
      "epoch": 1.2542372881355932,
      "grad_norm": 0.3420744836330414,
      "learning_rate": 0.0001864707496535537,
      "loss": 0.6367,
      "step": 1110
    },
    {
      "epoch": 1.2655367231638417,
      "grad_norm": 0.38407158851623535,
      "learning_rate": 0.00018607179295939983,
      "loss": 0.6278,
      "step": 1120
    },
    {
      "epoch": 1.2768361581920904,
      "grad_norm": 0.37982526421546936,
      "learning_rate": 0.00018566747884356182,
      "loss": 0.6209,
      "step": 1130
    },
    {
      "epoch": 1.288135593220339,
      "grad_norm": 0.33808642625808716,
      "learning_rate": 0.00018525783247202188,
      "loss": 0.6525,
      "step": 1140
    },
    {
      "epoch": 1.2994350282485876,
      "grad_norm": 0.45425844192504883,
      "learning_rate": 0.00018484287934266122,
      "loss": 0.6367,
      "step": 1150
    },
    {
      "epoch": 1.310734463276836,
      "grad_norm": 0.422880083322525,
      "learning_rate": 0.00018442264528367304,
      "loss": 0.6499,
      "step": 1160
    },
    {
      "epoch": 1.3220338983050848,
      "grad_norm": 0.3406778573989868,
      "learning_rate": 0.0001839971564519547,
      "loss": 0.6478,
      "step": 1170
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.37413087487220764,
      "learning_rate": 0.00018356643933147986,
      "loss": 0.6382,
      "step": 1180
    },
    {
      "epoch": 1.344632768361582,
      "grad_norm": 0.3171542286872864,
      "learning_rate": 0.00018313052073164982,
      "loss": 0.6373,
      "step": 1190
    },
    {
      "epoch": 1.3559322033898304,
      "grad_norm": 0.3208696246147156,
      "learning_rate": 0.00018268942778562494,
      "loss": 0.6231,
      "step": 1200
    },
    {
      "epoch": 1.3672316384180792,
      "grad_norm": 0.37951213121414185,
      "learning_rate": 0.0001822431879486357,
      "loss": 0.6659,
      "step": 1210
    },
    {
      "epoch": 1.3785310734463276,
      "grad_norm": 0.3654930591583252,
      "learning_rate": 0.00018179182899627386,
      "loss": 0.6467,
      "step": 1220
    },
    {
      "epoch": 1.3898305084745763,
      "grad_norm": 0.5204548239707947,
      "learning_rate": 0.00018133537902276342,
      "loss": 0.6479,
      "step": 1230
    },
    {
      "epoch": 1.4011299435028248,
      "grad_norm": 0.33258357644081116,
      "learning_rate": 0.0001808738664392122,
      "loss": 0.613,
      "step": 1240
    },
    {
      "epoch": 1.4124293785310735,
      "grad_norm": 0.3374476134777069,
      "learning_rate": 0.0001804073199718432,
      "loss": 0.6308,
      "step": 1250
    },
    {
      "epoch": 1.423728813559322,
      "grad_norm": 0.3677334189414978,
      "learning_rate": 0.0001799357686602068,
      "loss": 0.6412,
      "step": 1260
    },
    {
      "epoch": 1.4350282485875705,
      "grad_norm": 0.503189742565155,
      "learning_rate": 0.00017945924185537295,
      "loss": 0.6242,
      "step": 1270
    },
    {
      "epoch": 1.4463276836158192,
      "grad_norm": 0.44822338223457336,
      "learning_rate": 0.00017897776921810453,
      "loss": 0.6269,
      "step": 1280
    },
    {
      "epoch": 1.457627118644068,
      "grad_norm": 0.38298189640045166,
      "learning_rate": 0.00017849138071701092,
      "loss": 0.6206,
      "step": 1290
    },
    {
      "epoch": 1.4689265536723164,
      "grad_norm": 0.40119263529777527,
      "learning_rate": 0.0001780001066266829,
      "loss": 0.6398,
      "step": 1300
    },
    {
      "epoch": 1.4802259887005649,
      "grad_norm": 0.5280100703239441,
      "learning_rate": 0.00017750397752580796,
      "loss": 0.6299,
      "step": 1310
    },
    {
      "epoch": 1.4915254237288136,
      "grad_norm": 0.35528528690338135,
      "learning_rate": 0.00017700302429526723,
      "loss": 0.6511,
      "step": 1320
    },
    {
      "epoch": 1.5028248587570623,
      "grad_norm": 0.3741821348667145,
      "learning_rate": 0.00017649727811621315,
      "loss": 0.6564,
      "step": 1330
    },
    {
      "epoch": 1.5141242937853108,
      "grad_norm": 0.5371260046958923,
      "learning_rate": 0.0001759867704681288,
      "loss": 0.6186,
      "step": 1340
    },
    {
      "epoch": 1.5254237288135593,
      "grad_norm": 0.3590337336063385,
      "learning_rate": 0.00017547153312686832,
      "loss": 0.6372,
      "step": 1350
    },
    {
      "epoch": 1.536723163841808,
      "grad_norm": 0.34499940276145935,
      "learning_rate": 0.00017495159816267929,
      "loss": 0.6442,
      "step": 1360
    },
    {
      "epoch": 1.5480225988700564,
      "grad_norm": 0.45834365487098694,
      "learning_rate": 0.00017442699793820631,
      "loss": 0.6363,
      "step": 1370
    },
    {
      "epoch": 1.559322033898305,
      "grad_norm": 0.3389630615711212,
      "learning_rate": 0.00017389776510647688,
      "loss": 0.6399,
      "step": 1380
    },
    {
      "epoch": 1.5706214689265536,
      "grad_norm": 1.0764904022216797,
      "learning_rate": 0.0001733639326088688,
      "loss": 0.6073,
      "step": 1390
    },
    {
      "epoch": 1.5819209039548023,
      "grad_norm": 0.32130682468414307,
      "learning_rate": 0.00017282553367305975,
      "loss": 0.65,
      "step": 1400
    },
    {
      "epoch": 1.5932203389830508,
      "grad_norm": 0.5010311603546143,
      "learning_rate": 0.00017228260181095925,
      "loss": 0.626,
      "step": 1410
    },
    {
      "epoch": 1.6045197740112993,
      "grad_norm": 0.3811778128147125,
      "learning_rate": 0.00017173517081662255,
      "loss": 0.6565,
      "step": 1420
    },
    {
      "epoch": 1.615819209039548,
      "grad_norm": 0.793699324131012,
      "learning_rate": 0.00017118327476414735,
      "loss": 0.6555,
      "step": 1430
    },
    {
      "epoch": 1.6271186440677967,
      "grad_norm": 0.33635199069976807,
      "learning_rate": 0.0001706269480055528,
      "loss": 0.6431,
      "step": 1440
    },
    {
      "epoch": 1.6384180790960452,
      "grad_norm": 0.46125972270965576,
      "learning_rate": 0.00017006622516864127,
      "loss": 0.6383,
      "step": 1450
    },
    {
      "epoch": 1.6497175141242937,
      "grad_norm": 0.37351667881011963,
      "learning_rate": 0.00016950114115484315,
      "loss": 0.6304,
      "step": 1460
    },
    {
      "epoch": 1.6610169491525424,
      "grad_norm": 0.35749056935310364,
      "learning_rate": 0.0001689317311370442,
      "loss": 0.6647,
      "step": 1470
    },
    {
      "epoch": 1.672316384180791,
      "grad_norm": 0.3775118887424469,
      "learning_rate": 0.0001683580305573966,
      "loss": 0.6288,
      "step": 1480
    },
    {
      "epoch": 1.6836158192090396,
      "grad_norm": 0.3228208124637604,
      "learning_rate": 0.0001677800751251126,
      "loss": 0.6067,
      "step": 1490
    },
    {
      "epoch": 1.694915254237288,
      "grad_norm": 0.38675859570503235,
      "learning_rate": 0.00016719790081424192,
      "loss": 0.6438,
      "step": 1500
    },
    {
      "epoch": 1.7062146892655368,
      "grad_norm": 0.3832201063632965,
      "learning_rate": 0.0001666115438614328,
      "loss": 0.6269,
      "step": 1510
    },
    {
      "epoch": 1.7175141242937855,
      "grad_norm": 0.5429854989051819,
      "learning_rate": 0.00016602104076367624,
      "loss": 0.6397,
      "step": 1520
    },
    {
      "epoch": 1.7288135593220337,
      "grad_norm": 0.4048645496368408,
      "learning_rate": 0.0001654264282760343,
      "loss": 0.6402,
      "step": 1530
    },
    {
      "epoch": 1.7401129943502824,
      "grad_norm": 0.3748385012149811,
      "learning_rate": 0.00016482774340935267,
      "loss": 0.626,
      "step": 1540
    },
    {
      "epoch": 1.7514124293785311,
      "grad_norm": 0.44016194343566895,
      "learning_rate": 0.00016422502342795648,
      "loss": 0.6542,
      "step": 1550
    },
    {
      "epoch": 1.7627118644067796,
      "grad_norm": 0.34061527252197266,
      "learning_rate": 0.00016361830584733124,
      "loss": 0.6472,
      "step": 1560
    },
    {
      "epoch": 1.774011299435028,
      "grad_norm": 0.3848430812358856,
      "learning_rate": 0.00016300762843178754,
      "loss": 0.6304,
      "step": 1570
    },
    {
      "epoch": 1.7853107344632768,
      "grad_norm": 0.3756044805049896,
      "learning_rate": 0.00016239302919211052,
      "loss": 0.6161,
      "step": 1580
    },
    {
      "epoch": 1.7966101694915255,
      "grad_norm": 0.40179935097694397,
      "learning_rate": 0.0001617745463831939,
      "loss": 0.6428,
      "step": 1590
    },
    {
      "epoch": 1.807909604519774,
      "grad_norm": 0.4611279368400574,
      "learning_rate": 0.00016115221850165892,
      "loss": 0.6661,
      "step": 1600
    },
    {
      "epoch": 1.8192090395480225,
      "grad_norm": 0.5279085636138916,
      "learning_rate": 0.0001605260842834581,
      "loss": 0.6293,
      "step": 1610
    },
    {
      "epoch": 1.8305084745762712,
      "grad_norm": 0.5711485743522644,
      "learning_rate": 0.00015989618270146423,
      "loss": 0.6368,
      "step": 1620
    },
    {
      "epoch": 1.84180790960452,
      "grad_norm": 0.37090200185775757,
      "learning_rate": 0.00015926255296304454,
      "loss": 0.6297,
      "step": 1630
    },
    {
      "epoch": 1.8531073446327684,
      "grad_norm": 0.3836909830570221,
      "learning_rate": 0.00015862523450762032,
      "loss": 0.6817,
      "step": 1640
    },
    {
      "epoch": 1.8644067796610169,
      "grad_norm": 0.4297361969947815,
      "learning_rate": 0.00015798426700421196,
      "loss": 0.6586,
      "step": 1650
    },
    {
      "epoch": 1.8757062146892656,
      "grad_norm": 0.3696504235267639,
      "learning_rate": 0.0001573396903489699,
      "loss": 0.6466,
      "step": 1660
    },
    {
      "epoch": 1.8870056497175143,
      "grad_norm": 0.39131200313568115,
      "learning_rate": 0.00015669154466269137,
      "loss": 0.6448,
      "step": 1670
    },
    {
      "epoch": 1.8983050847457628,
      "grad_norm": 0.3583720624446869,
      "learning_rate": 0.00015603987028832303,
      "loss": 0.6385,
      "step": 1680
    },
    {
      "epoch": 1.9096045197740112,
      "grad_norm": 0.40514451265335083,
      "learning_rate": 0.00015538470778844995,
      "loss": 0.6368,
      "step": 1690
    },
    {
      "epoch": 1.92090395480226,
      "grad_norm": 0.4644288122653961,
      "learning_rate": 0.00015472609794277083,
      "loss": 0.6341,
      "step": 1700
    },
    {
      "epoch": 1.9322033898305084,
      "grad_norm": 0.6364967823028564,
      "learning_rate": 0.00015406408174555976,
      "loss": 0.619,
      "step": 1710
    },
    {
      "epoch": 1.943502824858757,
      "grad_norm": 0.3623166084289551,
      "learning_rate": 0.00015339870040311448,
      "loss": 0.6591,
      "step": 1720
    },
    {
      "epoch": 1.9548022598870056,
      "grad_norm": 0.39671990275382996,
      "learning_rate": 0.00015272999533119162,
      "loss": 0.645,
      "step": 1730
    },
    {
      "epoch": 1.9661016949152543,
      "grad_norm": 0.45107388496398926,
      "learning_rate": 0.0001520580081524289,
      "loss": 0.625,
      "step": 1740
    },
    {
      "epoch": 1.9774011299435028,
      "grad_norm": 0.4473719596862793,
      "learning_rate": 0.00015138278069375423,
      "loss": 0.6481,
      "step": 1750
    },
    {
      "epoch": 1.9887005649717513,
      "grad_norm": 0.390774130821228,
      "learning_rate": 0.00015070435498378243,
      "loss": 0.6482,
      "step": 1760
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.41401803493499756,
      "learning_rate": 0.00015002277325019904,
      "loss": 0.623,
      "step": 1770
    },
    {
      "epoch": 2.0112994350282487,
      "grad_norm": 0.7757445573806763,
      "learning_rate": 0.00014933807791713203,
      "loss": 0.573,
      "step": 1780
    },
    {
      "epoch": 2.022598870056497,
      "grad_norm": 0.44817861914634705,
      "learning_rate": 0.00014865031160251116,
      "loss": 0.6054,
      "step": 1790
    },
    {
      "epoch": 2.0338983050847457,
      "grad_norm": 0.4326503574848175,
      "learning_rate": 0.0001479595171154152,
      "loss": 0.5724,
      "step": 1800
    },
    {
      "epoch": 2.0451977401129944,
      "grad_norm": 0.4020352065563202,
      "learning_rate": 0.00014726573745340747,
      "loss": 0.5536,
      "step": 1810
    },
    {
      "epoch": 2.056497175141243,
      "grad_norm": 0.3883920907974243,
      "learning_rate": 0.00014656901579985934,
      "loss": 0.5591,
      "step": 1820
    },
    {
      "epoch": 2.0677966101694913,
      "grad_norm": 0.4672842025756836,
      "learning_rate": 0.00014586939552126255,
      "loss": 0.5467,
      "step": 1830
    },
    {
      "epoch": 2.07909604519774,
      "grad_norm": 0.5037193894386292,
      "learning_rate": 0.0001451669201645298,
      "loss": 0.5548,
      "step": 1840
    },
    {
      "epoch": 2.0903954802259888,
      "grad_norm": 0.46837809681892395,
      "learning_rate": 0.00014446163345428411,
      "loss": 0.5464,
      "step": 1850
    },
    {
      "epoch": 2.1016949152542375,
      "grad_norm": 0.4075257480144501,
      "learning_rate": 0.00014375357929013763,
      "loss": 0.5658,
      "step": 1860
    },
    {
      "epoch": 2.1129943502824857,
      "grad_norm": 0.5770477652549744,
      "learning_rate": 0.00014304280174395862,
      "loss": 0.5657,
      "step": 1870
    },
    {
      "epoch": 2.1242937853107344,
      "grad_norm": 0.38950198888778687,
      "learning_rate": 0.00014232934505712877,
      "loss": 0.5488,
      "step": 1880
    },
    {
      "epoch": 2.135593220338983,
      "grad_norm": 0.45963385701179504,
      "learning_rate": 0.00014161325363778914,
      "loss": 0.5536,
      "step": 1890
    },
    {
      "epoch": 2.146892655367232,
      "grad_norm": 0.5212920904159546,
      "learning_rate": 0.00014089457205807614,
      "loss": 0.5518,
      "step": 1900
    },
    {
      "epoch": 2.15819209039548,
      "grad_norm": 0.7819460034370422,
      "learning_rate": 0.00014017334505134722,
      "loss": 0.5493,
      "step": 1910
    },
    {
      "epoch": 2.169491525423729,
      "grad_norm": 0.41791173815727234,
      "learning_rate": 0.00013944961750939644,
      "loss": 0.5685,
      "step": 1920
    },
    {
      "epoch": 2.1807909604519775,
      "grad_norm": 0.40155965089797974,
      "learning_rate": 0.00013872343447966033,
      "loss": 0.5385,
      "step": 1930
    },
    {
      "epoch": 2.1920903954802258,
      "grad_norm": 0.3540368974208832,
      "learning_rate": 0.00013799484116241382,
      "loss": 0.5601,
      "step": 1940
    },
    {
      "epoch": 2.2033898305084745,
      "grad_norm": 0.45144760608673096,
      "learning_rate": 0.00013726388290795697,
      "loss": 0.5547,
      "step": 1950
    },
    {
      "epoch": 2.214689265536723,
      "grad_norm": 0.4121219515800476,
      "learning_rate": 0.00013653060521379216,
      "loss": 0.5627,
      "step": 1960
    },
    {
      "epoch": 2.225988700564972,
      "grad_norm": 0.5211389660835266,
      "learning_rate": 0.0001357950537217921,
      "loss": 0.5797,
      "step": 1970
    },
    {
      "epoch": 2.23728813559322,
      "grad_norm": 0.5209491848945618,
      "learning_rate": 0.00013505727421535892,
      "loss": 0.5644,
      "step": 1980
    },
    {
      "epoch": 2.248587570621469,
      "grad_norm": 0.40179020166397095,
      "learning_rate": 0.00013431731261657462,
      "loss": 0.5446,
      "step": 1990
    },
    {
      "epoch": 2.2598870056497176,
      "grad_norm": 0.45757922530174255,
      "learning_rate": 0.0001335752149833424,
      "loss": 0.5528,
      "step": 2000
    },
    {
      "epoch": 2.2598870056497176,
      "eval_loss": 0.731483519077301,
      "eval_runtime": 16.8378,
      "eval_samples_per_second": 88.551,
      "eval_steps_per_second": 11.106,
      "step": 2000
    },
    {
      "epoch": 2.2711864406779663,
      "grad_norm": 0.5437088012695312,
      "learning_rate": 0.00013283102750652007,
      "loss": 0.5476,
      "step": 2010
    },
    {
      "epoch": 2.2824858757062145,
      "grad_norm": 0.5207261443138123,
      "learning_rate": 0.00013208479650704503,
      "loss": 0.5839,
      "step": 2020
    },
    {
      "epoch": 2.2937853107344632,
      "grad_norm": 0.47627148032188416,
      "learning_rate": 0.00013133656843305078,
      "loss": 0.5791,
      "step": 2030
    },
    {
      "epoch": 2.305084745762712,
      "grad_norm": 0.4785435199737549,
      "learning_rate": 0.00013058638985697618,
      "loss": 0.5389,
      "step": 2040
    },
    {
      "epoch": 2.3163841807909606,
      "grad_norm": 0.40203753113746643,
      "learning_rate": 0.0001298343074726663,
      "loss": 0.5417,
      "step": 2050
    },
    {
      "epoch": 2.327683615819209,
      "grad_norm": 0.6017877459526062,
      "learning_rate": 0.00012908036809246623,
      "loss": 0.544,
      "step": 2060
    },
    {
      "epoch": 2.3389830508474576,
      "grad_norm": 0.5637748837471008,
      "learning_rate": 0.0001283246186443073,
      "loss": 0.5429,
      "step": 2070
    },
    {
      "epoch": 2.3502824858757063,
      "grad_norm": 0.4815099835395813,
      "learning_rate": 0.00012756710616878593,
      "loss": 0.5466,
      "step": 2080
    },
    {
      "epoch": 2.361581920903955,
      "grad_norm": 0.4496362805366516,
      "learning_rate": 0.00012680787781623582,
      "loss": 0.5894,
      "step": 2090
    },
    {
      "epoch": 2.3728813559322033,
      "grad_norm": 0.4290124475955963,
      "learning_rate": 0.00012604698084379322,
      "loss": 0.5504,
      "step": 2100
    },
    {
      "epoch": 2.384180790960452,
      "grad_norm": 0.5070333480834961,
      "learning_rate": 0.00012528446261245515,
      "loss": 0.5522,
      "step": 2110
    },
    {
      "epoch": 2.3954802259887007,
      "grad_norm": 0.5735118985176086,
      "learning_rate": 0.00012452037058413194,
      "loss": 0.5674,
      "step": 2120
    },
    {
      "epoch": 2.406779661016949,
      "grad_norm": 0.40572336316108704,
      "learning_rate": 0.00012375475231869249,
      "loss": 0.5814,
      "step": 2130
    },
    {
      "epoch": 2.4180790960451977,
      "grad_norm": 0.5074964165687561,
      "learning_rate": 0.00012298765547100445,
      "loss": 0.5663,
      "step": 2140
    },
    {
      "epoch": 2.4293785310734464,
      "grad_norm": 0.445563942193985,
      "learning_rate": 0.0001222191277879678,
      "loss": 0.5476,
      "step": 2150
    },
    {
      "epoch": 2.440677966101695,
      "grad_norm": 0.5029497742652893,
      "learning_rate": 0.00012144921710554278,
      "loss": 0.5818,
      "step": 2160
    },
    {
      "epoch": 2.4519774011299433,
      "grad_norm": 0.4790378212928772,
      "learning_rate": 0.00012067797134577275,
      "loss": 0.5878,
      "step": 2170
    },
    {
      "epoch": 2.463276836158192,
      "grad_norm": 0.43805941939353943,
      "learning_rate": 0.00011990543851380094,
      "loss": 0.5602,
      "step": 2180
    },
    {
      "epoch": 2.4745762711864407,
      "grad_norm": 0.49226316809654236,
      "learning_rate": 0.00011913166669488278,
      "loss": 0.5692,
      "step": 2190
    },
    {
      "epoch": 2.4858757062146895,
      "grad_norm": 0.44978803396224976,
      "learning_rate": 0.00011835670405139272,
      "loss": 0.5698,
      "step": 2200
    },
    {
      "epoch": 2.4971751412429377,
      "grad_norm": 0.4665353000164032,
      "learning_rate": 0.00011758059881982639,
      "loss": 0.567,
      "step": 2210
    },
    {
      "epoch": 2.5084745762711864,
      "grad_norm": 0.3619357943534851,
      "learning_rate": 0.00011680339930779843,
      "loss": 0.5528,
      "step": 2220
    },
    {
      "epoch": 2.519774011299435,
      "grad_norm": 0.5655903220176697,
      "learning_rate": 0.00011602515389103529,
      "loss": 0.57,
      "step": 2230
    },
    {
      "epoch": 2.5310734463276834,
      "grad_norm": 0.5191211700439453,
      "learning_rate": 0.00011524591101036446,
      "loss": 0.5731,
      "step": 2240
    },
    {
      "epoch": 2.542372881355932,
      "grad_norm": 0.4011457562446594,
      "learning_rate": 0.00011446571916869917,
      "loss": 0.5603,
      "step": 2250
    },
    {
      "epoch": 2.553672316384181,
      "grad_norm": 0.4324322044849396,
      "learning_rate": 0.00011368462692801944,
      "loss": 0.5508,
      "step": 2260
    },
    {
      "epoch": 2.5649717514124295,
      "grad_norm": 0.44678226113319397,
      "learning_rate": 0.00011290268290634943,
      "loss": 0.5695,
      "step": 2270
    },
    {
      "epoch": 2.576271186440678,
      "grad_norm": 0.7076367139816284,
      "learning_rate": 0.00011211993577473121,
      "loss": 0.5812,
      "step": 2280
    },
    {
      "epoch": 2.5875706214689265,
      "grad_norm": 0.36905479431152344,
      "learning_rate": 0.00011133643425419531,
      "loss": 0.5775,
      "step": 2290
    },
    {
      "epoch": 2.598870056497175,
      "grad_norm": 0.4640151560306549,
      "learning_rate": 0.0001105522271127283,
      "loss": 0.5932,
      "step": 2300
    },
    {
      "epoch": 2.610169491525424,
      "grad_norm": 0.541059672832489,
      "learning_rate": 0.00010976736316223698,
      "loss": 0.5803,
      "step": 2310
    },
    {
      "epoch": 2.621468926553672,
      "grad_norm": 0.46233227849006653,
      "learning_rate": 0.00010898189125551052,
      "loss": 0.5939,
      "step": 2320
    },
    {
      "epoch": 2.632768361581921,
      "grad_norm": 0.49569135904312134,
      "learning_rate": 0.0001081958602831794,
      "loss": 0.5762,
      "step": 2330
    },
    {
      "epoch": 2.6440677966101696,
      "grad_norm": 0.46241089701652527,
      "learning_rate": 0.00010740931917067231,
      "loss": 0.5455,
      "step": 2340
    },
    {
      "epoch": 2.655367231638418,
      "grad_norm": 0.4825233221054077,
      "learning_rate": 0.00010662231687517112,
      "loss": 0.5697,
      "step": 2350
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.3972626328468323,
      "learning_rate": 0.00010583490238256322,
      "loss": 0.5544,
      "step": 2360
    },
    {
      "epoch": 2.6779661016949152,
      "grad_norm": 0.458835244178772,
      "learning_rate": 0.00010504712470439266,
      "loss": 0.5603,
      "step": 2370
    },
    {
      "epoch": 2.689265536723164,
      "grad_norm": 0.5480552911758423,
      "learning_rate": 0.00010425903287480969,
      "loss": 0.5555,
      "step": 2380
    },
    {
      "epoch": 2.7005649717514126,
      "grad_norm": 0.5089293122291565,
      "learning_rate": 0.0001034706759475182,
      "loss": 0.5374,
      "step": 2390
    },
    {
      "epoch": 2.711864406779661,
      "grad_norm": 0.39211511611938477,
      "learning_rate": 0.00010268210299272294,
      "loss": 0.5499,
      "step": 2400
    },
    {
      "epoch": 2.7231638418079096,
      "grad_norm": 0.4466584324836731,
      "learning_rate": 0.00010189336309407492,
      "loss": 0.5692,
      "step": 2410
    },
    {
      "epoch": 2.7344632768361583,
      "grad_norm": 0.4684606194496155,
      "learning_rate": 0.00010110450534561629,
      "loss": 0.5858,
      "step": 2420
    },
    {
      "epoch": 2.7457627118644066,
      "grad_norm": 0.4925701320171356,
      "learning_rate": 0.00010031557884872471,
      "loss": 0.5577,
      "step": 2430
    },
    {
      "epoch": 2.7570621468926553,
      "grad_norm": 0.5621470212936401,
      "learning_rate": 9.952663270905691e-05,
      "loss": 0.543,
      "step": 2440
    },
    {
      "epoch": 2.768361581920904,
      "grad_norm": 0.5334658622741699,
      "learning_rate": 9.873771603349229e-05,
      "loss": 0.5673,
      "step": 2450
    },
    {
      "epoch": 2.7796610169491527,
      "grad_norm": 0.4116109311580658,
      "learning_rate": 9.79488779270763e-05,
      "loss": 0.5759,
      "step": 2460
    },
    {
      "epoch": 2.7909604519774014,
      "grad_norm": 0.579428493976593,
      "learning_rate": 9.716016748996397e-05,
      "loss": 0.5804,
      "step": 2470
    },
    {
      "epoch": 2.8022598870056497,
      "grad_norm": 0.6294684410095215,
      "learning_rate": 9.637163381436368e-05,
      "loss": 0.573,
      "step": 2480
    },
    {
      "epoch": 2.8135593220338984,
      "grad_norm": 0.4215356409549713,
      "learning_rate": 9.558332598148156e-05,
      "loss": 0.5702,
      "step": 2490
    },
    {
      "epoch": 2.824858757062147,
      "grad_norm": 0.6607437133789062,
      "learning_rate": 9.479529305846652e-05,
      "loss": 0.5556,
      "step": 2500
    },
    {
      "epoch": 2.8361581920903953,
      "grad_norm": 0.4144829213619232,
      "learning_rate": 9.400758409535597e-05,
      "loss": 0.5417,
      "step": 2510
    },
    {
      "epoch": 2.847457627118644,
      "grad_norm": 0.47953999042510986,
      "learning_rate": 9.322024812202295e-05,
      "loss": 0.5531,
      "step": 2520
    },
    {
      "epoch": 2.8587570621468927,
      "grad_norm": 0.4800129532814026,
      "learning_rate": 9.243333414512428e-05,
      "loss": 0.5692,
      "step": 2530
    },
    {
      "epoch": 2.870056497175141,
      "grad_norm": 0.6145460605621338,
      "learning_rate": 9.164689114505007e-05,
      "loss": 0.556,
      "step": 2540
    },
    {
      "epoch": 2.8813559322033897,
      "grad_norm": 0.45880794525146484,
      "learning_rate": 9.086096807287525e-05,
      "loss": 0.547,
      "step": 2550
    },
    {
      "epoch": 2.8926553672316384,
      "grad_norm": 0.4852420389652252,
      "learning_rate": 9.007561384731249e-05,
      "loss": 0.5843,
      "step": 2560
    },
    {
      "epoch": 2.903954802259887,
      "grad_norm": 0.4079766571521759,
      "learning_rate": 8.929087735166738e-05,
      "loss": 0.5642,
      "step": 2570
    },
    {
      "epoch": 2.915254237288136,
      "grad_norm": 0.41701099276542664,
      "learning_rate": 8.850680743079576e-05,
      "loss": 0.5191,
      "step": 2580
    },
    {
      "epoch": 2.926553672316384,
      "grad_norm": 0.51291823387146,
      "learning_rate": 8.772345288806338e-05,
      "loss": 0.5823,
      "step": 2590
    },
    {
      "epoch": 2.937853107344633,
      "grad_norm": 0.3969896137714386,
      "learning_rate": 8.694086248230836e-05,
      "loss": 0.5673,
      "step": 2600
    },
    {
      "epoch": 2.9491525423728815,
      "grad_norm": 0.5091503858566284,
      "learning_rate": 8.615908492480598e-05,
      "loss": 0.5555,
      "step": 2610
    },
    {
      "epoch": 2.9604519774011298,
      "grad_norm": 0.5286157131195068,
      "learning_rate": 8.537816887623706e-05,
      "loss": 0.5883,
      "step": 2620
    },
    {
      "epoch": 2.9717514124293785,
      "grad_norm": 0.42479345202445984,
      "learning_rate": 8.459816294365892e-05,
      "loss": 0.6002,
      "step": 2630
    },
    {
      "epoch": 2.983050847457627,
      "grad_norm": 0.6383501291275024,
      "learning_rate": 8.381911567747986e-05,
      "loss": 0.5751,
      "step": 2640
    },
    {
      "epoch": 2.994350282485876,
      "grad_norm": 0.4221038520336151,
      "learning_rate": 8.304107556843739e-05,
      "loss": 0.5583,
      "step": 2650
    },
    {
      "epoch": 3.005649717514124,
      "grad_norm": 0.46783480048179626,
      "learning_rate": 8.22640910445799e-05,
      "loss": 0.5075,
      "step": 2660
    },
    {
      "epoch": 3.016949152542373,
      "grad_norm": 0.48267218470573425,
      "learning_rate": 8.14882104682522e-05,
      "loss": 0.4768,
      "step": 2670
    },
    {
      "epoch": 3.0282485875706215,
      "grad_norm": 0.3863345682621002,
      "learning_rate": 8.07134821330855e-05,
      "loss": 0.4959,
      "step": 2680
    },
    {
      "epoch": 3.0395480225988702,
      "grad_norm": 0.5208786129951477,
      "learning_rate": 7.993995426099132e-05,
      "loss": 0.4765,
      "step": 2690
    },
    {
      "epoch": 3.0508474576271185,
      "grad_norm": 0.4619675874710083,
      "learning_rate": 7.916767499915998e-05,
      "loss": 0.4615,
      "step": 2700
    },
    {
      "epoch": 3.062146892655367,
      "grad_norm": 0.4720297157764435,
      "learning_rate": 7.839669241706368e-05,
      "loss": 0.495,
      "step": 2710
    },
    {
      "epoch": 3.073446327683616,
      "grad_norm": 0.5332727432250977,
      "learning_rate": 7.762705450346462e-05,
      "loss": 0.486,
      "step": 2720
    },
    {
      "epoch": 3.084745762711864,
      "grad_norm": 0.5170637965202332,
      "learning_rate": 7.685880916342795e-05,
      "loss": 0.4816,
      "step": 2730
    },
    {
      "epoch": 3.096045197740113,
      "grad_norm": 0.47557196021080017,
      "learning_rate": 7.609200421533991e-05,
      "loss": 0.4706,
      "step": 2740
    },
    {
      "epoch": 3.1073446327683616,
      "grad_norm": 0.7031615376472473,
      "learning_rate": 7.532668738793159e-05,
      "loss": 0.479,
      "step": 2750
    },
    {
      "epoch": 3.1186440677966103,
      "grad_norm": 0.4733956456184387,
      "learning_rate": 7.456290631730795e-05,
      "loss": 0.5023,
      "step": 2760
    },
    {
      "epoch": 3.1299435028248586,
      "grad_norm": 0.5425024032592773,
      "learning_rate": 7.380070854398292e-05,
      "loss": 0.477,
      "step": 2770
    },
    {
      "epoch": 3.1412429378531073,
      "grad_norm": 0.4106122851371765,
      "learning_rate": 7.304014150992016e-05,
      "loss": 0.511,
      "step": 2780
    },
    {
      "epoch": 3.152542372881356,
      "grad_norm": 0.580517053604126,
      "learning_rate": 7.228125255558026e-05,
      "loss": 0.4797,
      "step": 2790
    },
    {
      "epoch": 3.1638418079096047,
      "grad_norm": 0.6407458186149597,
      "learning_rate": 7.152408891697406e-05,
      "loss": 0.4908,
      "step": 2800
    },
    {
      "epoch": 3.175141242937853,
      "grad_norm": 0.6218848824501038,
      "learning_rate": 7.076869772272231e-05,
      "loss": 0.4944,
      "step": 2810
    },
    {
      "epoch": 3.1864406779661016,
      "grad_norm": 0.49968206882476807,
      "learning_rate": 7.001512599112254e-05,
      "loss": 0.4764,
      "step": 2820
    },
    {
      "epoch": 3.1977401129943503,
      "grad_norm": 0.465017706155777,
      "learning_rate": 6.926342062722223e-05,
      "loss": 0.4917,
      "step": 2830
    },
    {
      "epoch": 3.209039548022599,
      "grad_norm": 0.49751996994018555,
      "learning_rate": 6.851362841989926e-05,
      "loss": 0.5105,
      "step": 2840
    },
    {
      "epoch": 3.2203389830508473,
      "grad_norm": 0.5394355654716492,
      "learning_rate": 6.776579603894983e-05,
      "loss": 0.4888,
      "step": 2850
    },
    {
      "epoch": 3.231638418079096,
      "grad_norm": 0.628495454788208,
      "learning_rate": 6.701997003218334e-05,
      "loss": 0.4597,
      "step": 2860
    },
    {
      "epoch": 3.2429378531073447,
      "grad_norm": 0.544737696647644,
      "learning_rate": 6.627619682252514e-05,
      "loss": 0.5198,
      "step": 2870
    },
    {
      "epoch": 3.2542372881355934,
      "grad_norm": 0.3959389626979828,
      "learning_rate": 6.553452270512708e-05,
      "loss": 0.4881,
      "step": 2880
    },
    {
      "epoch": 3.2655367231638417,
      "grad_norm": 0.7966346740722656,
      "learning_rate": 6.47949938444858e-05,
      "loss": 0.4839,
      "step": 2890
    },
    {
      "epoch": 3.2768361581920904,
      "grad_norm": 0.5471003651618958,
      "learning_rate": 6.405765627156934e-05,
      "loss": 0.4953,
      "step": 2900
    },
    {
      "epoch": 3.288135593220339,
      "grad_norm": 0.6175327301025391,
      "learning_rate": 6.332255588095211e-05,
      "loss": 0.4936,
      "step": 2910
    },
    {
      "epoch": 3.2994350282485874,
      "grad_norm": 0.5062096118927002,
      "learning_rate": 6.258973842795803e-05,
      "loss": 0.4927,
      "step": 2920
    },
    {
      "epoch": 3.310734463276836,
      "grad_norm": 0.5980800986289978,
      "learning_rate": 6.185924952581286e-05,
      "loss": 0.4908,
      "step": 2930
    },
    {
      "epoch": 3.3220338983050848,
      "grad_norm": 0.47181573510169983,
      "learning_rate": 6.11311346428046e-05,
      "loss": 0.4938,
      "step": 2940
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.4729149341583252,
      "learning_rate": 6.040543909945393e-05,
      "loss": 0.4795,
      "step": 2950
    },
    {
      "epoch": 3.3446327683615817,
      "grad_norm": 0.4721032977104187,
      "learning_rate": 5.968220806569289e-05,
      "loss": 0.4821,
      "step": 2960
    },
    {
      "epoch": 3.3559322033898304,
      "grad_norm": 0.49077048897743225,
      "learning_rate": 5.89614865580534e-05,
      "loss": 0.4765,
      "step": 2970
    },
    {
      "epoch": 3.367231638418079,
      "grad_norm": 0.4721141755580902,
      "learning_rate": 5.8243319436865554e-05,
      "loss": 0.4622,
      "step": 2980
    },
    {
      "epoch": 3.378531073446328,
      "grad_norm": 0.5188248157501221,
      "learning_rate": 5.7527751403464935e-05,
      "loss": 0.4966,
      "step": 2990
    },
    {
      "epoch": 3.389830508474576,
      "grad_norm": 0.5185617208480835,
      "learning_rate": 5.681482699741052e-05,
      "loss": 0.4847,
      "step": 3000
    },
    {
      "epoch": 3.389830508474576,
      "eval_loss": 0.7709711194038391,
      "eval_runtime": 16.7631,
      "eval_samples_per_second": 88.945,
      "eval_steps_per_second": 11.155,
      "step": 3000
    },
    {
      "epoch": 3.401129943502825,
      "grad_norm": 0.44520846009254456,
      "learning_rate": 5.610459059371229e-05,
      "loss": 0.4904,
      "step": 3010
    },
    {
      "epoch": 3.4124293785310735,
      "grad_norm": 0.4945251941680908,
      "learning_rate": 5.53970864000693e-05,
      "loss": 0.486,
      "step": 3020
    },
    {
      "epoch": 3.423728813559322,
      "grad_norm": 0.45569783449172974,
      "learning_rate": 5.4692358454117596e-05,
      "loss": 0.485,
      "step": 3030
    },
    {
      "epoch": 3.4350282485875705,
      "grad_norm": 0.6237002015113831,
      "learning_rate": 5.399045062068977e-05,
      "loss": 0.5055,
      "step": 3040
    },
    {
      "epoch": 3.446327683615819,
      "grad_norm": 0.4695291519165039,
      "learning_rate": 5.329140658908423e-05,
      "loss": 0.5047,
      "step": 3050
    },
    {
      "epoch": 3.457627118644068,
      "grad_norm": 0.5210557579994202,
      "learning_rate": 5.259526987034593e-05,
      "loss": 0.5181,
      "step": 3060
    },
    {
      "epoch": 3.4689265536723166,
      "grad_norm": 0.5110927224159241,
      "learning_rate": 5.190208379455802e-05,
      "loss": 0.4901,
      "step": 3070
    },
    {
      "epoch": 3.480225988700565,
      "grad_norm": 0.5741479396820068,
      "learning_rate": 5.121189150814507e-05,
      "loss": 0.4919,
      "step": 3080
    },
    {
      "epoch": 3.4915254237288136,
      "grad_norm": 0.49151936173439026,
      "learning_rate": 5.052473597118715e-05,
      "loss": 0.4954,
      "step": 3090
    },
    {
      "epoch": 3.5028248587570623,
      "grad_norm": 0.5301800966262817,
      "learning_rate": 4.984065995474601e-05,
      "loss": 0.4966,
      "step": 3100
    },
    {
      "epoch": 3.5141242937853105,
      "grad_norm": 0.5711139440536499,
      "learning_rate": 4.915970603820293e-05,
      "loss": 0.4907,
      "step": 3110
    },
    {
      "epoch": 3.5254237288135593,
      "grad_norm": 0.6174890995025635,
      "learning_rate": 4.84819166066082e-05,
      "loss": 0.4957,
      "step": 3120
    },
    {
      "epoch": 3.536723163841808,
      "grad_norm": 0.5370630621910095,
      "learning_rate": 4.780733384804312e-05,
      "loss": 0.5131,
      "step": 3130
    },
    {
      "epoch": 3.548022598870056,
      "grad_norm": 0.4994024634361267,
      "learning_rate": 4.713599975099399e-05,
      "loss": 0.5214,
      "step": 3140
    },
    {
      "epoch": 3.559322033898305,
      "grad_norm": 0.8335681557655334,
      "learning_rate": 4.646795610173864e-05,
      "loss": 0.498,
      "step": 3150
    },
    {
      "epoch": 3.5706214689265536,
      "grad_norm": 0.5772273540496826,
      "learning_rate": 4.5803244481745275e-05,
      "loss": 0.4889,
      "step": 3160
    },
    {
      "epoch": 3.5819209039548023,
      "grad_norm": 0.725974977016449,
      "learning_rate": 4.514190626508465e-05,
      "loss": 0.4889,
      "step": 3170
    },
    {
      "epoch": 3.593220338983051,
      "grad_norm": 0.4802244305610657,
      "learning_rate": 4.448398261585459e-05,
      "loss": 0.4688,
      "step": 3180
    },
    {
      "epoch": 3.6045197740112993,
      "grad_norm": 1.078092098236084,
      "learning_rate": 4.3829514485617754e-05,
      "loss": 0.4719,
      "step": 3190
    },
    {
      "epoch": 3.615819209039548,
      "grad_norm": 0.5553079843521118,
      "learning_rate": 4.31785426108527e-05,
      "loss": 0.4816,
      "step": 3200
    },
    {
      "epoch": 3.6271186440677967,
      "grad_norm": 0.48087623715400696,
      "learning_rate": 4.253110751041846e-05,
      "loss": 0.4894,
      "step": 3210
    },
    {
      "epoch": 3.638418079096045,
      "grad_norm": 0.6121565103530884,
      "learning_rate": 4.1887249483032195e-05,
      "loss": 0.4978,
      "step": 3220
    },
    {
      "epoch": 3.6497175141242937,
      "grad_norm": 0.616296648979187,
      "learning_rate": 4.1247008604761096e-05,
      "loss": 0.4834,
      "step": 3230
    },
    {
      "epoch": 3.6610169491525424,
      "grad_norm": 0.5769548416137695,
      "learning_rate": 4.061042472652786e-05,
      "loss": 0.5177,
      "step": 3240
    },
    {
      "epoch": 3.672316384180791,
      "grad_norm": 0.5508149862289429,
      "learning_rate": 3.997753747163014e-05,
      "loss": 0.5206,
      "step": 3250
    },
    {
      "epoch": 3.68361581920904,
      "grad_norm": 0.4783056080341339,
      "learning_rate": 3.934838623327427e-05,
      "loss": 0.4996,
      "step": 3260
    },
    {
      "epoch": 3.694915254237288,
      "grad_norm": 0.9651548266410828,
      "learning_rate": 3.872301017212337e-05,
      "loss": 0.4741,
      "step": 3270
    },
    {
      "epoch": 3.7062146892655368,
      "grad_norm": 0.5215980410575867,
      "learning_rate": 3.8101448213859846e-05,
      "loss": 0.4757,
      "step": 3280
    },
    {
      "epoch": 3.7175141242937855,
      "grad_norm": 0.4626060724258423,
      "learning_rate": 3.7483739046762326e-05,
      "loss": 0.4887,
      "step": 3290
    },
    {
      "epoch": 3.7288135593220337,
      "grad_norm": 0.5218005180358887,
      "learning_rate": 3.6869921119297714e-05,
      "loss": 0.4603,
      "step": 3300
    },
    {
      "epoch": 3.7401129943502824,
      "grad_norm": 0.557963490486145,
      "learning_rate": 3.626003263772813e-05,
      "loss": 0.5009,
      "step": 3310
    },
    {
      "epoch": 3.751412429378531,
      "grad_norm": 0.7049193382263184,
      "learning_rate": 3.565411156373252e-05,
      "loss": 0.4982,
      "step": 3320
    },
    {
      "epoch": 3.7627118644067794,
      "grad_norm": 0.5267760753631592,
      "learning_rate": 3.505219561204398e-05,
      "loss": 0.4716,
      "step": 3330
    },
    {
      "epoch": 3.774011299435028,
      "grad_norm": 0.5169788002967834,
      "learning_rate": 3.445432224810232e-05,
      "loss": 0.4992,
      "step": 3340
    },
    {
      "epoch": 3.785310734463277,
      "grad_norm": 0.5268625020980835,
      "learning_rate": 3.386052868572186e-05,
      "loss": 0.5063,
      "step": 3350
    },
    {
      "epoch": 3.7966101694915255,
      "grad_norm": 0.5163342356681824,
      "learning_rate": 3.3270851884775203e-05,
      "loss": 0.5045,
      "step": 3360
    },
    {
      "epoch": 3.8079096045197742,
      "grad_norm": 0.6581404805183411,
      "learning_rate": 3.268532854889287e-05,
      "loss": 0.4818,
      "step": 3370
    },
    {
      "epoch": 3.8192090395480225,
      "grad_norm": 0.5500733256340027,
      "learning_rate": 3.210399512317849e-05,
      "loss": 0.5064,
      "step": 3380
    },
    {
      "epoch": 3.830508474576271,
      "grad_norm": 0.530192494392395,
      "learning_rate": 3.1526887791940395e-05,
      "loss": 0.4761,
      "step": 3390
    },
    {
      "epoch": 3.84180790960452,
      "grad_norm": 0.5225520133972168,
      "learning_rate": 3.095404247643955e-05,
      "loss": 0.4874,
      "step": 3400
    },
    {
      "epoch": 3.853107344632768,
      "grad_norm": 0.52415531873703,
      "learning_rate": 3.038549483265348e-05,
      "loss": 0.4748,
      "step": 3410
    },
    {
      "epoch": 3.864406779661017,
      "grad_norm": 0.6530227661132812,
      "learning_rate": 2.982128024905697e-05,
      "loss": 0.4841,
      "step": 3420
    },
    {
      "epoch": 3.8757062146892656,
      "grad_norm": 0.5648587346076965,
      "learning_rate": 2.9261433844419306e-05,
      "loss": 0.4687,
      "step": 3430
    },
    {
      "epoch": 3.8870056497175143,
      "grad_norm": 0.957992672920227,
      "learning_rate": 2.870599046561858e-05,
      "loss": 0.489,
      "step": 3440
    },
    {
      "epoch": 3.898305084745763,
      "grad_norm": 0.5714943408966064,
      "learning_rate": 2.8154984685472373e-05,
      "loss": 0.4886,
      "step": 3450
    },
    {
      "epoch": 3.9096045197740112,
      "grad_norm": 0.48912012577056885,
      "learning_rate": 2.7608450800586016e-05,
      "loss": 0.5028,
      "step": 3460
    },
    {
      "epoch": 3.92090395480226,
      "grad_norm": 0.5882663726806641,
      "learning_rate": 2.7066422829217875e-05,
      "loss": 0.5216,
      "step": 3470
    },
    {
      "epoch": 3.9322033898305087,
      "grad_norm": 0.5730474591255188,
      "learning_rate": 2.6528934509161763e-05,
      "loss": 0.4596,
      "step": 3480
    },
    {
      "epoch": 3.943502824858757,
      "grad_norm": 0.6839239001274109,
      "learning_rate": 2.599601929564709e-05,
      "loss": 0.4944,
      "step": 3490
    },
    {
      "epoch": 3.9548022598870056,
      "grad_norm": 0.5317196249961853,
      "learning_rate": 2.5467710359256557e-05,
      "loss": 0.4779,
      "step": 3500
    },
    {
      "epoch": 3.9661016949152543,
      "grad_norm": 0.5707146525382996,
      "learning_rate": 2.4944040583861284e-05,
      "loss": 0.4878,
      "step": 3510
    },
    {
      "epoch": 3.9774011299435026,
      "grad_norm": 0.781765341758728,
      "learning_rate": 2.4425042564574184e-05,
      "loss": 0.4719,
      "step": 3520
    },
    {
      "epoch": 3.9887005649717513,
      "grad_norm": 0.6404000520706177,
      "learning_rate": 2.3910748605721066e-05,
      "loss": 0.4983,
      "step": 3530
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6482042074203491,
      "learning_rate": 2.3401190718829945e-05,
      "loss": 0.4916,
      "step": 3540
    },
    {
      "epoch": 4.011299435028248,
      "grad_norm": 0.4203549027442932,
      "learning_rate": 2.289640062063839e-05,
      "loss": 0.4653,
      "step": 3550
    },
    {
      "epoch": 4.022598870056497,
      "grad_norm": 0.48745134472846985,
      "learning_rate": 2.239640973111945e-05,
      "loss": 0.4595,
      "step": 3560
    },
    {
      "epoch": 4.033898305084746,
      "grad_norm": 0.5251093506813049,
      "learning_rate": 2.1901249171526063e-05,
      "loss": 0.4471,
      "step": 3570
    },
    {
      "epoch": 4.045197740112994,
      "grad_norm": 0.6273601651191711,
      "learning_rate": 2.1410949762453735e-05,
      "loss": 0.437,
      "step": 3580
    },
    {
      "epoch": 4.056497175141243,
      "grad_norm": 0.5651510953903198,
      "learning_rate": 2.0925542021922283e-05,
      "loss": 0.4194,
      "step": 3590
    },
    {
      "epoch": 4.067796610169491,
      "grad_norm": 0.7425692081451416,
      "learning_rate": 2.0445056163476374e-05,
      "loss": 0.4568,
      "step": 3600
    },
    {
      "epoch": 4.0790960451977405,
      "grad_norm": 0.49514123797416687,
      "learning_rate": 1.9969522094304704e-05,
      "loss": 0.4673,
      "step": 3610
    },
    {
      "epoch": 4.090395480225989,
      "grad_norm": 0.7164167761802673,
      "learning_rate": 1.949896941337861e-05,
      "loss": 0.4507,
      "step": 3620
    },
    {
      "epoch": 4.101694915254237,
      "grad_norm": 0.5990516543388367,
      "learning_rate": 1.9033427409609738e-05,
      "loss": 0.4271,
      "step": 3630
    },
    {
      "epoch": 4.112994350282486,
      "grad_norm": 0.4122953414916992,
      "learning_rate": 1.8572925060026868e-05,
      "loss": 0.4331,
      "step": 3640
    },
    {
      "epoch": 4.124293785310734,
      "grad_norm": 0.5930763483047485,
      "learning_rate": 1.8117491027972476e-05,
      "loss": 0.4388,
      "step": 3650
    },
    {
      "epoch": 4.135593220338983,
      "grad_norm": 0.5768269300460815,
      "learning_rate": 1.766715366131837e-05,
      "loss": 0.4432,
      "step": 3660
    },
    {
      "epoch": 4.146892655367232,
      "grad_norm": 1.2128398418426514,
      "learning_rate": 1.722194099070148e-05,
      "loss": 0.4445,
      "step": 3670
    },
    {
      "epoch": 4.15819209039548,
      "grad_norm": 0.5718667507171631,
      "learning_rate": 1.678188072777891e-05,
      "loss": 0.4348,
      "step": 3680
    },
    {
      "epoch": 4.169491525423728,
      "grad_norm": 0.6063762903213501,
      "learning_rate": 1.6347000263503176e-05,
      "loss": 0.4293,
      "step": 3690
    },
    {
      "epoch": 4.1807909604519775,
      "grad_norm": 0.5244751572608948,
      "learning_rate": 1.591732666641732e-05,
      "loss": 0.4469,
      "step": 3700
    },
    {
      "epoch": 4.192090395480226,
      "grad_norm": 0.6261086463928223,
      "learning_rate": 1.5492886680969963e-05,
      "loss": 0.4415,
      "step": 3710
    },
    {
      "epoch": 4.203389830508475,
      "grad_norm": 1.0888699293136597,
      "learning_rate": 1.5073706725850679e-05,
      "loss": 0.4365,
      "step": 3720
    },
    {
      "epoch": 4.214689265536723,
      "grad_norm": 0.5183214545249939,
      "learning_rate": 1.4659812892345692e-05,
      "loss": 0.4717,
      "step": 3730
    },
    {
      "epoch": 4.2259887005649714,
      "grad_norm": 0.5120430588722229,
      "learning_rate": 1.425123094271369e-05,
      "loss": 0.4484,
      "step": 3740
    },
    {
      "epoch": 4.237288135593221,
      "grad_norm": 0.4748518764972687,
      "learning_rate": 1.3847986308582384e-05,
      "loss": 0.4566,
      "step": 3750
    },
    {
      "epoch": 4.248587570621469,
      "grad_norm": 0.5438849329948425,
      "learning_rate": 1.3450104089365611e-05,
      "loss": 0.4768,
      "step": 3760
    },
    {
      "epoch": 4.259887005649717,
      "grad_norm": 0.4719776213169098,
      "learning_rate": 1.3057609050700881e-05,
      "loss": 0.4657,
      "step": 3770
    },
    {
      "epoch": 4.271186440677966,
      "grad_norm": 0.5300427675247192,
      "learning_rate": 1.267052562290808e-05,
      "loss": 0.4412,
      "step": 3780
    },
    {
      "epoch": 4.2824858757062145,
      "grad_norm": 0.6340104937553406,
      "learning_rate": 1.2288877899468653e-05,
      "loss": 0.4528,
      "step": 3790
    },
    {
      "epoch": 4.293785310734464,
      "grad_norm": 0.5274319648742676,
      "learning_rate": 1.1912689635526086e-05,
      "loss": 0.4466,
      "step": 3800
    },
    {
      "epoch": 4.305084745762712,
      "grad_norm": 0.5659282803535461,
      "learning_rate": 1.1541984246407189e-05,
      "loss": 0.4332,
      "step": 3810
    },
    {
      "epoch": 4.31638418079096,
      "grad_norm": 0.5424573421478271,
      "learning_rate": 1.1176784806164676e-05,
      "loss": 0.4223,
      "step": 3820
    },
    {
      "epoch": 4.327683615819209,
      "grad_norm": 0.5446699261665344,
      "learning_rate": 1.081711404614104e-05,
      "loss": 0.4335,
      "step": 3830
    },
    {
      "epoch": 4.338983050847458,
      "grad_norm": 0.4954063594341278,
      "learning_rate": 1.0462994353553512e-05,
      "loss": 0.4221,
      "step": 3840
    },
    {
      "epoch": 4.350282485875706,
      "grad_norm": 0.6320489048957825,
      "learning_rate": 1.0114447770100688e-05,
      "loss": 0.43,
      "step": 3850
    },
    {
      "epoch": 4.361581920903955,
      "grad_norm": 0.5482968688011169,
      "learning_rate": 9.77149599059063e-06,
      "loss": 0.4407,
      "step": 3860
    },
    {
      "epoch": 4.372881355932203,
      "grad_norm": 0.6075878739356995,
      "learning_rate": 9.434160361590372e-06,
      "loss": 0.4434,
      "step": 3870
    },
    {
      "epoch": 4.3841807909604515,
      "grad_norm": 0.6049032211303711,
      "learning_rate": 9.102461880097302e-06,
      "loss": 0.4411,
      "step": 3880
    },
    {
      "epoch": 4.395480225988701,
      "grad_norm": 0.5798004865646362,
      "learning_rate": 8.776421192232298e-06,
      "loss": 0.431,
      "step": 3890
    },
    {
      "epoch": 4.406779661016949,
      "grad_norm": 0.5278260707855225,
      "learning_rate": 8.456058591954463e-06,
      "loss": 0.4422,
      "step": 3900
    },
    {
      "epoch": 4.418079096045198,
      "grad_norm": 0.568495512008667,
      "learning_rate": 8.141394019798177e-06,
      "loss": 0.4413,
      "step": 3910
    },
    {
      "epoch": 4.429378531073446,
      "grad_norm": 0.5533962845802307,
      "learning_rate": 7.832447061631688e-06,
      "loss": 0.4191,
      "step": 3920
    },
    {
      "epoch": 4.440677966101695,
      "grad_norm": 0.5724579691886902,
      "learning_rate": 7.529236947438256e-06,
      "loss": 0.4428,
      "step": 3930
    },
    {
      "epoch": 4.451977401129944,
      "grad_norm": 0.5295354723930359,
      "learning_rate": 7.231782550119015e-06,
      "loss": 0.4363,
      "step": 3940
    },
    {
      "epoch": 4.463276836158192,
      "grad_norm": 0.6352728605270386,
      "learning_rate": 6.940102384318314e-06,
      "loss": 0.4462,
      "step": 3950
    },
    {
      "epoch": 4.47457627118644,
      "grad_norm": 0.4653913080692291,
      "learning_rate": 6.654214605271414e-06,
      "loss": 0.4463,
      "step": 3960
    },
    {
      "epoch": 4.4858757062146895,
      "grad_norm": 0.5769869685173035,
      "learning_rate": 6.37413700767423e-06,
      "loss": 0.4388,
      "step": 3970
    },
    {
      "epoch": 4.497175141242938,
      "grad_norm": 0.4576704204082489,
      "learning_rate": 6.099887024575901e-06,
      "loss": 0.4466,
      "step": 3980
    },
    {
      "epoch": 4.508474576271187,
      "grad_norm": 0.5665444135665894,
      "learning_rate": 5.831481726293609e-06,
      "loss": 0.4628,
      "step": 3990
    },
    {
      "epoch": 4.519774011299435,
      "grad_norm": 0.527612566947937,
      "learning_rate": 5.568937819350084e-06,
      "loss": 0.4261,
      "step": 4000
    },
    {
      "epoch": 4.519774011299435,
      "eval_loss": 0.8213748335838318,
      "eval_runtime": 16.8138,
      "eval_samples_per_second": 88.677,
      "eval_steps_per_second": 11.122,
      "step": 4000
    },
    {
      "epoch": 4.531073446327683,
      "grad_norm": 0.668229877948761,
      "learning_rate": 5.312271645433697e-06,
      "loss": 0.4298,
      "step": 4010
    },
    {
      "epoch": 4.5423728813559325,
      "grad_norm": 0.5199172496795654,
      "learning_rate": 5.061499180381391e-06,
      "loss": 0.4547,
      "step": 4020
    },
    {
      "epoch": 4.553672316384181,
      "grad_norm": 0.5515559911727905,
      "learning_rate": 4.816636033184119e-06,
      "loss": 0.4172,
      "step": 4030
    },
    {
      "epoch": 4.564971751412429,
      "grad_norm": 0.648692786693573,
      "learning_rate": 4.577697445015472e-06,
      "loss": 0.4664,
      "step": 4040
    },
    {
      "epoch": 4.576271186440678,
      "grad_norm": 0.5000047087669373,
      "learning_rate": 4.3446982882828555e-06,
      "loss": 0.4401,
      "step": 4050
    },
    {
      "epoch": 4.5875706214689265,
      "grad_norm": 0.51460862159729,
      "learning_rate": 4.117653065701899e-06,
      "loss": 0.4297,
      "step": 4060
    },
    {
      "epoch": 4.598870056497175,
      "grad_norm": 0.5801824331283569,
      "learning_rate": 3.896575909393663e-06,
      "loss": 0.453,
      "step": 4070
    },
    {
      "epoch": 4.610169491525424,
      "grad_norm": 0.5613316297531128,
      "learning_rate": 3.6814805800050255e-06,
      "loss": 0.445,
      "step": 4080
    },
    {
      "epoch": 4.621468926553672,
      "grad_norm": 0.556430459022522,
      "learning_rate": 3.4723804658522318e-06,
      "loss": 0.4332,
      "step": 4090
    },
    {
      "epoch": 4.632768361581921,
      "grad_norm": 0.7290319204330444,
      "learning_rate": 3.269288582087493e-06,
      "loss": 0.4667,
      "step": 4100
    },
    {
      "epoch": 4.6440677966101696,
      "grad_norm": 0.5109266638755798,
      "learning_rate": 3.072217569888847e-06,
      "loss": 0.4484,
      "step": 4110
    },
    {
      "epoch": 4.655367231638418,
      "grad_norm": 0.5184862017631531,
      "learning_rate": 2.8811796956734418e-06,
      "loss": 0.4476,
      "step": 4120
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.45906636118888855,
      "learning_rate": 2.696186850333893e-06,
      "loss": 0.4556,
      "step": 4130
    },
    {
      "epoch": 4.677966101694915,
      "grad_norm": 0.6825990080833435,
      "learning_rate": 2.5172505484982733e-06,
      "loss": 0.4361,
      "step": 4140
    },
    {
      "epoch": 4.6892655367231635,
      "grad_norm": 0.6751702427864075,
      "learning_rate": 2.3443819278132996e-06,
      "loss": 0.454,
      "step": 4150
    },
    {
      "epoch": 4.700564971751413,
      "grad_norm": 0.6289049983024597,
      "learning_rate": 2.17759174825114e-06,
      "loss": 0.4438,
      "step": 4160
    },
    {
      "epoch": 4.711864406779661,
      "grad_norm": 0.6290619969367981,
      "learning_rate": 2.016890391439674e-06,
      "loss": 0.4665,
      "step": 4170
    },
    {
      "epoch": 4.72316384180791,
      "grad_norm": 0.5864591002464294,
      "learning_rate": 1.8622878600162763e-06,
      "loss": 0.4477,
      "step": 4180
    },
    {
      "epoch": 4.734463276836158,
      "grad_norm": 0.558210015296936,
      "learning_rate": 1.7137937770052349e-06,
      "loss": 0.4605,
      "step": 4190
    },
    {
      "epoch": 4.745762711864407,
      "grad_norm": 0.6048043966293335,
      "learning_rate": 1.5714173852187875e-06,
      "loss": 0.4458,
      "step": 4200
    },
    {
      "epoch": 4.757062146892656,
      "grad_norm": 0.6476958990097046,
      "learning_rate": 1.4351675466817705e-06,
      "loss": 0.4311,
      "step": 4210
    },
    {
      "epoch": 4.768361581920904,
      "grad_norm": 0.6173336505889893,
      "learning_rate": 1.3050527420800928e-06,
      "loss": 0.4526,
      "step": 4220
    },
    {
      "epoch": 4.779661016949152,
      "grad_norm": 0.5486857295036316,
      "learning_rate": 1.1810810702327701e-06,
      "loss": 0.4407,
      "step": 4230
    },
    {
      "epoch": 4.790960451977401,
      "grad_norm": 0.4905402362346649,
      "learning_rate": 1.063260247587905e-06,
      "loss": 0.4367,
      "step": 4240
    },
    {
      "epoch": 4.80225988700565,
      "grad_norm": 0.6765434741973877,
      "learning_rate": 9.515976077423716e-07,
      "loss": 0.4438,
      "step": 4250
    },
    {
      "epoch": 4.813559322033898,
      "grad_norm": 0.5942672491073608,
      "learning_rate": 8.461001009852809e-07,
      "loss": 0.4326,
      "step": 4260
    },
    {
      "epoch": 4.824858757062147,
      "grad_norm": 0.5058199167251587,
      "learning_rate": 7.46774293865471e-07,
      "loss": 0.4258,
      "step": 4270
    },
    {
      "epoch": 4.836158192090395,
      "grad_norm": 0.5257803797721863,
      "learning_rate": 6.536263687827005e-07,
      "loss": 0.4377,
      "step": 4280
    },
    {
      "epoch": 4.847457627118644,
      "grad_norm": 0.5546150207519531,
      "learning_rate": 5.66662123602879e-07,
      "loss": 0.447,
      "step": 4290
    },
    {
      "epoch": 4.858757062146893,
      "grad_norm": 0.5133059024810791,
      "learning_rate": 4.858869712971669e-07,
      "loss": 0.4536,
      "step": 4300
    },
    {
      "epoch": 4.870056497175141,
      "grad_norm": 0.6438310742378235,
      "learning_rate": 4.113059396050445e-07,
      "loss": 0.4328,
      "step": 4310
    },
    {
      "epoch": 4.88135593220339,
      "grad_norm": 0.5640572309494019,
      "learning_rate": 3.4292367072140677e-07,
      "loss": 0.4629,
      "step": 4320
    },
    {
      "epoch": 4.892655367231638,
      "grad_norm": 0.6676624417304993,
      "learning_rate": 2.807444210075616e-07,
      "loss": 0.4347,
      "step": 4330
    },
    {
      "epoch": 4.903954802259887,
      "grad_norm": 0.45671963691711426,
      "learning_rate": 2.2477206072633038e-07,
      "loss": 0.4536,
      "step": 4340
    },
    {
      "epoch": 4.915254237288136,
      "grad_norm": 0.6115604043006897,
      "learning_rate": 1.7501007380115174e-07,
      "loss": 0.4453,
      "step": 4350
    },
    {
      "epoch": 4.926553672316384,
      "grad_norm": 0.7250502109527588,
      "learning_rate": 1.3146155759923285e-07,
      "loss": 0.4454,
      "step": 4360
    },
    {
      "epoch": 4.937853107344633,
      "grad_norm": 0.5707434415817261,
      "learning_rate": 9.412922273871471e-08,
      "loss": 0.4107,
      "step": 4370
    },
    {
      "epoch": 4.9491525423728815,
      "grad_norm": 0.6125264763832092,
      "learning_rate": 6.301539292001834e-08,
      "loss": 0.4355,
      "step": 4380
    },
    {
      "epoch": 4.96045197740113,
      "grad_norm": 0.6177554130554199,
      "learning_rate": 3.8122004781149246e-08,
      "loss": 0.4502,
      "step": 4390
    },
    {
      "epoch": 4.971751412429379,
      "grad_norm": 0.7295418381690979,
      "learning_rate": 1.9450607777204978e-08,
      "loss": 0.4426,
      "step": 4400
    },
    {
      "epoch": 4.983050847457627,
      "grad_norm": 0.5891022682189941,
      "learning_rate": 7.00236408388566e-09,
      "loss": 0.449,
      "step": 4410
    },
    {
      "epoch": 4.994350282485875,
      "grad_norm": 0.5898358225822449,
      "learning_rate": 7.780485252073533e-10,
      "loss": 0.4552,
      "step": 4420
    },
    {
      "epoch": 5.0,
      "step": 4425,
      "total_flos": 1.2071933436588196e+18,
      "train_loss": 0.5702266787943867,
      "train_runtime": 4287.601,
      "train_samples_per_second": 33.024,
      "train_steps_per_second": 1.032
    }
  ],
  "logging_steps": 10,
  "max_steps": 4425,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 1.2071933436588196e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}