{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1535,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03257328990228013,
      "grad_norm": 10.164939880371094,
      "learning_rate": 2.5974025974025972e-05,
      "loss": 3.0074,
      "step": 10
    },
    {
      "epoch": 0.06514657980456026,
      "grad_norm": 11.759410858154297,
      "learning_rate": 5.1948051948051944e-05,
      "loss": 1.1608,
      "step": 20
    },
    {
      "epoch": 0.09771986970684039,
      "grad_norm": 2.336061716079712,
      "learning_rate": 7.792207792207793e-05,
      "loss": 0.8385,
      "step": 30
    },
    {
      "epoch": 0.13029315960912052,
      "grad_norm": 1.3015629053115845,
      "learning_rate": 0.00010389610389610389,
      "loss": 0.2802,
      "step": 40
    },
    {
      "epoch": 0.16286644951140064,
      "grad_norm": 1.0599477291107178,
      "learning_rate": 0.00012987012987012987,
      "loss": 0.1576,
      "step": 50
    },
    {
      "epoch": 0.19543973941368079,
      "grad_norm": 3.056833267211914,
      "learning_rate": 0.00015584415584415587,
      "loss": 0.2403,
      "step": 60
    },
    {
      "epoch": 0.2280130293159609,
      "grad_norm": 2.162693738937378,
      "learning_rate": 0.00018181818181818183,
      "loss": 0.251,
      "step": 70
    },
    {
      "epoch": 0.26058631921824105,
      "grad_norm": 3.423180341720581,
      "learning_rate": 0.0001999979107245606,
      "loss": 0.2418,
      "step": 80
    },
    {
      "epoch": 0.2931596091205212,
      "grad_norm": 1.267954707145691,
      "learning_rate": 0.0001999607704786645,
      "loss": 0.157,
      "step": 90
    },
    {
      "epoch": 0.3257328990228013,
      "grad_norm": 2.345654249191284,
      "learning_rate": 0.00019987722173728587,
      "loss": 0.2589,
      "step": 100
    },
    {
      "epoch": 0.3583061889250814,
      "grad_norm": 3.0315651893615723,
      "learning_rate": 0.00019974730328935534,
      "loss": 0.2621,
      "step": 110
    },
    {
      "epoch": 0.39087947882736157,
      "grad_norm": 0.543853223323822,
      "learning_rate": 0.00019957107545173192,
      "loss": 0.1612,
      "step": 120
    },
    {
      "epoch": 0.4234527687296417,
      "grad_norm": 0.9071061611175537,
      "learning_rate": 0.00019934862004120004,
      "loss": 0.2302,
      "step": 130
    },
    {
      "epoch": 0.4560260586319218,
      "grad_norm": 2.0897481441497803,
      "learning_rate": 0.00019908004033648453,
      "loss": 0.1522,
      "step": 140
    },
    {
      "epoch": 0.48859934853420195,
      "grad_norm": 0.8175027370452881,
      "learning_rate": 0.00019876546103030196,
      "loss": 0.0943,
      "step": 150
    },
    {
      "epoch": 0.5211726384364821,
      "grad_norm": 0.5068450570106506,
      "learning_rate": 0.00019840502817146966,
      "loss": 0.0865,
      "step": 160
    },
    {
      "epoch": 0.5537459283387622,
      "grad_norm": 1.0016964673995972,
      "learning_rate": 0.00019799890909710013,
      "loss": 0.1229,
      "step": 170
    },
    {
      "epoch": 0.5863192182410424,
      "grad_norm": 2.739112377166748,
      "learning_rate": 0.00019754729235491207,
      "loss": 0.1327,
      "step": 180
    },
    {
      "epoch": 0.6188925081433225,
      "grad_norm": 1.9622081518173218,
      "learning_rate": 0.0001970503876156937,
      "loss": 0.1361,
      "step": 190
    },
    {
      "epoch": 0.6514657980456026,
      "grad_norm": 3.4183578491210938,
      "learning_rate": 0.00019650842557595967,
      "loss": 0.1363,
      "step": 200
    },
    {
      "epoch": 0.6840390879478827,
      "grad_norm": 0.8435428738594055,
      "learning_rate": 0.00019592165785084603,
      "loss": 0.1079,
      "step": 210
    },
    {
      "epoch": 0.7166123778501629,
      "grad_norm": 2.0992093086242676,
      "learning_rate": 0.00019529035685729391,
      "loss": 0.2401,
      "step": 220
    },
    {
      "epoch": 0.749185667752443,
      "grad_norm": 2.749847650527954,
      "learning_rate": 0.00019461481568757506,
      "loss": 0.1147,
      "step": 230
    },
    {
      "epoch": 0.7817589576547231,
      "grad_norm": 2.5013184547424316,
      "learning_rate": 0.00019389534797321884,
      "loss": 0.1661,
      "step": 240
    },
    {
      "epoch": 0.8143322475570033,
      "grad_norm": 0.8295263648033142,
      "learning_rate": 0.00019313228773940345,
      "loss": 0.1267,
      "step": 250
    },
    {
      "epoch": 0.8469055374592834,
      "grad_norm": 2.893522262573242,
      "learning_rate": 0.00019232598924987903,
      "loss": 0.0937,
      "step": 260
    },
    {
      "epoch": 0.8794788273615635,
      "grad_norm": 3.0416808128356934,
      "learning_rate": 0.0001914768268424946,
      "loss": 0.1318,
      "step": 270
    },
    {
      "epoch": 0.9120521172638436,
      "grad_norm": 0.6984137892723083,
      "learning_rate": 0.00019058519475540538,
      "loss": 0.1003,
      "step": 280
    },
    {
      "epoch": 0.9446254071661238,
      "grad_norm": 1.0821751356124878,
      "learning_rate": 0.00018965150694404094,
      "loss": 0.0888,
      "step": 290
    },
    {
      "epoch": 0.9771986970684039,
      "grad_norm": 0.9735985994338989,
      "learning_rate": 0.00018867619688891937,
      "loss": 0.207,
      "step": 300
    },
    {
      "epoch": 1.009771986970684,
      "grad_norm": 0.6221311688423157,
      "learning_rate": 0.0001876597173943965,
      "loss": 0.1472,
      "step": 310
    },
    {
      "epoch": 1.0423452768729642,
      "grad_norm": 2.335994243621826,
      "learning_rate": 0.00018660254037844388,
      "loss": 0.071,
      "step": 320
    },
    {
      "epoch": 1.0749185667752443,
      "grad_norm": 5.066413402557373,
      "learning_rate": 0.00018550515665355247,
      "loss": 0.0781,
      "step": 330
    },
    {
      "epoch": 1.1074918566775245,
      "grad_norm": 1.0136305093765259,
      "learning_rate": 0.000184368075698865,
      "loss": 0.0879,
      "step": 340
    },
    {
      "epoch": 1.1400651465798046,
      "grad_norm": 1.4384146928787231,
      "learning_rate": 0.00018319182542364117,
      "loss": 0.1173,
      "step": 350
    },
    {
      "epoch": 1.1726384364820848,
      "grad_norm": 1.0204399824142456,
      "learning_rate": 0.00018197695192216702,
      "loss": 0.0738,
      "step": 360
    },
    {
      "epoch": 1.205211726384365,
      "grad_norm": 0.8423169255256653,
      "learning_rate": 0.00018072401922022117,
      "loss": 0.0999,
      "step": 370
    },
    {
      "epoch": 1.237785016286645,
      "grad_norm": 1.6605499982833862,
      "learning_rate": 0.0001794336090132164,
      "loss": 0.0888,
      "step": 380
    },
    {
      "epoch": 1.2703583061889252,
      "grad_norm": 3.043694019317627,
      "learning_rate": 0.00017810632039613736,
      "loss": 0.1041,
      "step": 390
    },
    {
      "epoch": 1.3029315960912053,
      "grad_norm": 0.49020448327064514,
      "learning_rate": 0.00017674276958540072,
      "loss": 0.1021,
      "step": 400
    },
    {
      "epoch": 1.3355048859934853,
      "grad_norm": 2.167153835296631,
      "learning_rate": 0.00017534358963276607,
      "loss": 0.0971,
      "step": 410
    },
    {
      "epoch": 1.3680781758957654,
      "grad_norm": 0.8787184953689575,
      "learning_rate": 0.00017390943013143083,
      "loss": 0.0941,
      "step": 420
    },
    {
      "epoch": 1.4006514657980456,
      "grad_norm": 1.053727626800537,
      "learning_rate": 0.0001724409569144455,
      "loss": 0.0756,
      "step": 430
    },
    {
      "epoch": 1.4332247557003257,
      "grad_norm": 0.5299842357635498,
      "learning_rate": 0.0001709388517455893,
      "loss": 0.0927,
      "step": 440
    },
    {
      "epoch": 1.4657980456026058,
      "grad_norm": 1.387248158454895,
      "learning_rate": 0.00016940381200284972,
      "loss": 0.0828,
      "step": 450
    },
    {
      "epoch": 1.498371335504886,
      "grad_norm": 0.581618070602417,
      "learning_rate": 0.0001678365503546528,
      "loss": 0.0697,
      "step": 460
    },
    {
      "epoch": 1.5309446254071661,
      "grad_norm": 0.41632965207099915,
      "learning_rate": 0.0001662377944289948,
      "loss": 0.1533,
      "step": 470
    },
    {
      "epoch": 1.5635179153094463,
      "grad_norm": 1.1524041891098022,
      "learning_rate": 0.0001646082864756282,
      "loss": 0.0879,
      "step": 480
    },
    {
      "epoch": 1.5960912052117264,
      "grad_norm": 0.3662007451057434,
      "learning_rate": 0.00016294878302145987,
      "loss": 0.0421,
      "step": 490
    },
    {
      "epoch": 1.6286644951140063,
      "grad_norm": 0.5205959677696228,
      "learning_rate": 0.0001612600545193203,
      "loss": 0.1127,
      "step": 500
    },
    {
      "epoch": 1.6612377850162865,
      "grad_norm": 0.5683305263519287,
      "learning_rate": 0.00015954288499026782,
      "loss": 0.0697,
      "step": 510
    },
    {
      "epoch": 1.6938110749185666,
      "grad_norm": 1.1733250617980957,
      "learning_rate": 0.0001577980716595934,
      "loss": 0.0724,
      "step": 520
    },
    {
      "epoch": 1.7263843648208468,
      "grad_norm": 3.809410810470581,
      "learning_rate": 0.00015602642458669528,
      "loss": 0.0993,
      "step": 530
    },
    {
      "epoch": 1.758957654723127,
      "grad_norm": 0.48794323205947876,
      "learning_rate": 0.0001542287662889948,
      "loss": 0.0576,
      "step": 540
    },
    {
      "epoch": 1.791530944625407,
      "grad_norm": 1.401726245880127,
      "learning_rate": 0.00015240593136006897,
      "loss": 0.1039,
      "step": 550
    },
    {
      "epoch": 1.8241042345276872,
      "grad_norm": 0.33422133326530457,
      "learning_rate": 0.0001505587660821759,
      "loss": 0.0458,
      "step": 560
    },
    {
      "epoch": 1.8566775244299674,
      "grad_norm": 0.5557196140289307,
      "learning_rate": 0.0001486881280333539,
      "loss": 0.0489,
      "step": 570
    },
    {
      "epoch": 1.8892508143322475,
      "grad_norm": 2.7507164478302,
      "learning_rate": 0.00014679488568927616,
      "loss": 0.0873,
      "step": 580
    },
    {
      "epoch": 1.9218241042345277,
      "grad_norm": 0.8317222595214844,
      "learning_rate": 0.00014487991802004623,
      "loss": 0.0402,
      "step": 590
    },
    {
      "epoch": 1.9543973941368078,
      "grad_norm": 2.6091113090515137,
      "learning_rate": 0.0001429441140821209,
      "loss": 0.0853,
      "step": 600
    },
    {
      "epoch": 1.986970684039088,
      "grad_norm": 0.29834985733032227,
      "learning_rate": 0.00014098837260555084,
      "loss": 0.0544,
      "step": 610
    },
    {
      "epoch": 2.019543973941368,
      "grad_norm": 0.7023916840553284,
      "learning_rate": 0.0001390136015767295,
      "loss": 0.0976,
      "step": 620
    },
    {
      "epoch": 2.0521172638436482,
      "grad_norm": 0.34739574790000916,
      "learning_rate": 0.00013702071781684517,
      "loss": 0.0612,
      "step": 630
    },
    {
      "epoch": 2.0846905537459284,
      "grad_norm": 0.43586787581443787,
      "learning_rate": 0.00013501064655623094,
      "loss": 0.0614,
      "step": 640
    },
    {
      "epoch": 2.1172638436482085,
      "grad_norm": 0.27919432520866394,
      "learning_rate": 0.00013298432100481079,
      "loss": 0.0539,
      "step": 650
    },
    {
      "epoch": 2.1498371335504887,
      "grad_norm": 0.5438246130943298,
      "learning_rate": 0.0001309426819188409,
      "loss": 0.0401,
      "step": 660
    },
    {
      "epoch": 2.182410423452769,
      "grad_norm": 0.6277980804443359,
      "learning_rate": 0.0001288866771641474,
      "loss": 0.0744,
      "step": 670
    },
    {
      "epoch": 2.214983713355049,
      "grad_norm": 0.9506388306617737,
      "learning_rate": 0.00012681726127606376,
      "loss": 0.073,
      "step": 680
    },
    {
      "epoch": 2.247557003257329,
      "grad_norm": 1.6029094457626343,
      "learning_rate": 0.000124735395016271,
      "loss": 0.0329,
      "step": 690
    },
    {
      "epoch": 2.2801302931596092,
      "grad_norm": 2.2171266078948975,
      "learning_rate": 0.00012264204492674815,
      "loss": 0.0885,
      "step": 700
    },
    {
      "epoch": 2.3127035830618894,
      "grad_norm": 1.0621176958084106,
      "learning_rate": 0.0001205381828810382,
      "loss": 0.0518,
      "step": 710
    },
    {
      "epoch": 2.3452768729641695,
      "grad_norm": 1.3911964893341064,
      "learning_rate": 0.00011842478563303952,
      "loss": 0.0694,
      "step": 720
    },
    {
      "epoch": 2.3778501628664497,
      "grad_norm": 0.586409330368042,
      "learning_rate": 0.00011630283436353098,
      "loss": 0.1295,
      "step": 730
    },
    {
      "epoch": 2.41042345276873,
      "grad_norm": 0.4753137230873108,
      "learning_rate": 0.00011417331422464205,
      "loss": 0.0438,
      "step": 740
    },
    {
      "epoch": 2.44299674267101,
      "grad_norm": 0.2599440813064575,
      "learning_rate": 0.00011203721388247923,
      "loss": 0.0446,
      "step": 750
    },
    {
      "epoch": 2.47557003257329,
      "grad_norm": 0.43588873744010925,
      "learning_rate": 0.00010989552505812072,
      "loss": 0.1583,
      "step": 760
    },
    {
      "epoch": 2.5081433224755703,
      "grad_norm": 0.36574748158454895,
      "learning_rate": 0.0001077492420671931,
      "loss": 0.0968,
      "step": 770
    },
    {
      "epoch": 2.5407166123778504,
      "grad_norm": 0.5202295780181885,
      "learning_rate": 0.00010559936135824322,
      "loss": 0.048,
      "step": 780
    },
    {
      "epoch": 2.5732899022801305,
      "grad_norm": 0.40310633182525635,
      "learning_rate": 0.00010344688105012005,
      "loss": 0.0422,
      "step": 790
    },
    {
      "epoch": 2.6058631921824107,
      "grad_norm": 0.2391999065876007,
      "learning_rate": 0.00010129280046858086,
      "loss": 0.092,
      "step": 800
    },
    {
      "epoch": 2.6384364820846904,
      "grad_norm": 0.390307754278183,
      "learning_rate": 9.913811968233716e-05,
      "loss": 0.0432,
      "step": 810
    },
    {
      "epoch": 2.6710097719869705,
      "grad_norm": 0.5276124477386475,
      "learning_rate": 9.69838390387558e-05,
      "loss": 0.0606,
      "step": 820
    },
    {
      "epoch": 2.7035830618892507,
      "grad_norm": 0.44748738408088684,
      "learning_rate": 9.483095869943055e-05,
      "loss": 0.0445,
      "step": 830
    },
    {
      "epoch": 2.736156351791531,
      "grad_norm": 1.4012900590896606,
      "learning_rate": 9.268047817583998e-05,
      "loss": 0.0372,
      "step": 840
    },
    {
      "epoch": 2.768729641693811,
      "grad_norm": 0.78389972448349,
      "learning_rate": 9.053339586530723e-05,
      "loss": 0.0623,
      "step": 850
    },
    {
      "epoch": 2.801302931596091,
      "grad_norm": 0.5246724486351013,
      "learning_rate": 8.839070858747697e-05,
      "loss": 0.0499,
      "step": 860
    },
    {
      "epoch": 2.8338762214983713,
      "grad_norm": 0.2494039684534073,
      "learning_rate": 8.625341112152487e-05,
      "loss": 0.0462,
      "step": 870
    },
    {
      "epoch": 2.8664495114006514,
      "grad_norm": 0.3887605667114258,
      "learning_rate": 8.412249574431428e-05,
      "loss": 0.0319,
      "step": 880
    },
    {
      "epoch": 2.8990228013029316,
      "grad_norm": 0.3090686500072479,
      "learning_rate": 8.199895176971488e-05,
      "loss": 0.0804,
      "step": 890
    },
    {
      "epoch": 2.9315960912052117,
      "grad_norm": 0.47201988101005554,
      "learning_rate": 7.988376508929676e-05,
      "loss": 0.0466,
      "step": 900
    },
    {
      "epoch": 2.964169381107492,
      "grad_norm": 0.4059586822986603,
      "learning_rate": 7.777791771461332e-05,
      "loss": 0.0337,
      "step": 910
    },
    {
      "epoch": 2.996742671009772,
      "grad_norm": 0.4746284782886505,
      "learning_rate": 7.568238732128585e-05,
      "loss": 0.0489,
      "step": 920
    },
    {
      "epoch": 3.029315960912052,
      "grad_norm": 0.24269038438796997,
      "learning_rate": 7.359814679510065e-05,
      "loss": 0.0408,
      "step": 930
    },
    {
      "epoch": 3.0618892508143323,
      "grad_norm": 0.6225023865699768,
      "learning_rate": 7.152616378033042e-05,
      "loss": 0.1245,
      "step": 940
    },
    {
      "epoch": 3.0944625407166124,
      "grad_norm": 0.3462587296962738,
      "learning_rate": 6.94674002304887e-05,
      "loss": 0.0369,
      "step": 950
    },
    {
      "epoch": 3.1270358306188926,
      "grad_norm": 1.3566789627075195,
      "learning_rate": 6.742281196172663e-05,
      "loss": 0.0499,
      "step": 960
    },
    {
      "epoch": 3.1596091205211727,
      "grad_norm": 0.45816901326179504,
      "learning_rate": 6.539334820907888e-05,
      "loss": 0.0368,
      "step": 970
    },
    {
      "epoch": 3.192182410423453,
      "grad_norm": 0.3949735760688782,
      "learning_rate": 6.337995118576521e-05,
      "loss": 0.0542,
      "step": 980
    },
    {
      "epoch": 3.224755700325733,
      "grad_norm": 0.7988837361335754,
      "learning_rate": 6.138355564575169e-05,
      "loss": 0.0415,
      "step": 990
    },
    {
      "epoch": 3.257328990228013,
      "grad_norm": 0.2154189944267273,
      "learning_rate": 5.940508844977537e-05,
      "loss": 0.0322,
      "step": 1000
    },
    {
      "epoch": 3.2899022801302933,
      "grad_norm": 0.7597559690475464,
      "learning_rate": 5.744546813503328e-05,
      "loss": 0.0924,
      "step": 1010
    },
    {
      "epoch": 3.3224755700325734,
      "grad_norm": 0.26576346158981323,
      "learning_rate": 5.550560448873575e-05,
      "loss": 0.0344,
      "step": 1020
    },
    {
      "epoch": 3.3550488599348536,
      "grad_norm": 2.2215240001678467,
      "learning_rate": 5.358639812572244e-05,
      "loss": 0.0429,
      "step": 1030
    },
    {
      "epoch": 3.3876221498371337,
      "grad_norm": 0.7066490650177002,
      "learning_rate": 5.168874007033615e-05,
      "loss": 0.0419,
      "step": 1040
    },
    {
      "epoch": 3.420195439739414,
      "grad_norm": 0.23159299790859222,
      "learning_rate": 4.9813511342749805e-05,
      "loss": 0.0253,
      "step": 1050
    },
    {
      "epoch": 3.4527687296416936,
      "grad_norm": 0.5614166259765625,
      "learning_rate": 4.7961582549937675e-05,
      "loss": 0.0264,
      "step": 1060
    },
    {
      "epoch": 3.4853420195439737,
      "grad_norm": 0.4681215286254883,
      "learning_rate": 4.6133813481481246e-05,
      "loss": 0.0234,
      "step": 1070
    },
    {
      "epoch": 3.517915309446254,
      "grad_norm": 0.3369932770729065,
      "learning_rate": 4.433105271039721e-05,
      "loss": 0.0441,
      "step": 1080
    },
    {
      "epoch": 3.550488599348534,
      "grad_norm": 0.30148735642433167,
      "learning_rate": 4.255413719917294e-05,
      "loss": 0.0492,
      "step": 1090
    },
    {
      "epoch": 3.583061889250814,
      "grad_norm": 0.3371269106864929,
      "learning_rate": 4.080389191119241e-05,
      "loss": 0.0325,
      "step": 1100
    },
    {
      "epoch": 3.6156351791530943,
      "grad_norm": 0.8490313291549683,
      "learning_rate": 3.9081129427732774e-05,
      "loss": 0.0264,
      "step": 1110
    },
    {
      "epoch": 3.6482084690553744,
      "grad_norm": 0.30010277032852173,
      "learning_rate": 3.7386649570709644e-05,
      "loss": 0.0276,
      "step": 1120
    },
    {
      "epoch": 3.6807817589576546,
      "grad_norm": 0.2943625748157501,
      "learning_rate": 3.5721239031346066e-05,
      "loss": 0.0945,
      "step": 1130
    },
    {
      "epoch": 3.7133550488599347,
      "grad_norm": 0.19849584996700287,
      "learning_rate": 3.408567100493787e-05,
      "loss": 0.0543,
      "step": 1140
    },
    {
      "epoch": 3.745928338762215,
      "grad_norm": 0.24954350292682648,
      "learning_rate": 3.248070483188426e-05,
      "loss": 0.0423,
      "step": 1150
    },
    {
      "epoch": 3.778501628664495,
      "grad_norm": 0.425689697265625,
      "learning_rate": 3.090708564515124e-05,
      "loss": 0.0556,
      "step": 1160
    },
    {
      "epoch": 3.811074918566775,
      "grad_norm": 0.18508583307266235,
      "learning_rate": 2.936554402433087e-05,
      "loss": 0.0217,
      "step": 1170
    },
    {
      "epoch": 3.8436482084690553,
      "grad_norm": 4.1349196434021,
      "learning_rate": 2.7856795656457257e-05,
      "loss": 0.0729,
      "step": 1180
    },
    {
      "epoch": 3.8762214983713354,
      "grad_norm": 0.4305671155452728,
      "learning_rate": 2.6381541003736486e-05,
      "loss": 0.0327,
      "step": 1190
    },
    {
      "epoch": 3.9087947882736156,
      "grad_norm": 0.16706599295139313,
      "learning_rate": 2.494046497834518e-05,
      "loss": 0.0202,
      "step": 1200
    },
    {
      "epoch": 3.9413680781758957,
      "grad_norm": 0.40168455243110657,
      "learning_rate": 2.3534236624448302e-05,
      "loss": 0.0428,
      "step": 1210
    },
    {
      "epoch": 3.973941368078176,
      "grad_norm": 0.33089303970336914,
      "learning_rate": 2.2163508807583998e-05,
      "loss": 0.0289,
      "step": 1220
    },
    {
      "epoch": 4.006514657980456,
      "grad_norm": 0.1787663847208023,
      "learning_rate": 2.082891791155954e-05,
      "loss": 0.0152,
      "step": 1230
    },
    {
      "epoch": 4.039087947882736,
      "grad_norm": 0.7322351932525635,
      "learning_rate": 1.9531083542999317e-05,
      "loss": 0.0303,
      "step": 1240
    },
    {
      "epoch": 4.071661237785016,
      "grad_norm": 0.34002748131752014,
      "learning_rate": 1.8270608243681953e-05,
      "loss": 0.0144,
      "step": 1250
    },
    {
      "epoch": 4.1042345276872965,
      "grad_norm": 0.2633480131626129,
      "learning_rate": 1.7048077210799772e-05,
      "loss": 0.0162,
      "step": 1260
    },
    {
      "epoch": 4.136807817589577,
      "grad_norm": 0.32320234179496765,
      "learning_rate": 1.5864058025271246e-05,
      "loss": 0.0158,
      "step": 1270
    },
    {
      "epoch": 4.169381107491857,
      "grad_norm": 0.2887153625488281,
      "learning_rate": 1.47191003882317e-05,
      "loss": 0.0275,
      "step": 1280
    },
    {
      "epoch": 4.201954397394137,
      "grad_norm": 0.2007487416267395,
      "learning_rate": 1.3613735865825305e-05,
      "loss": 0.0192,
      "step": 1290
    },
    {
      "epoch": 4.234527687296417,
      "grad_norm": 0.6144905090332031,
      "learning_rate": 1.2548477642416256e-05,
      "loss": 0.0225,
      "step": 1300
    },
    {
      "epoch": 4.267100977198697,
      "grad_norm": 0.2809244692325592,
      "learning_rate": 1.1523820282334219e-05,
      "loss": 0.0201,
      "step": 1310
    },
    {
      "epoch": 4.299674267100977,
      "grad_norm": 0.2203732132911682,
      "learning_rate": 1.0540239500264516e-05,
      "loss": 0.0213,
      "step": 1320
    },
    {
      "epoch": 4.3322475570032575,
      "grad_norm": 0.17500793933868408,
      "learning_rate": 9.598191940389256e-06,
      "loss": 0.0193,
      "step": 1330
    },
    {
      "epoch": 4.364820846905538,
      "grad_norm": 0.45830363035202026,
      "learning_rate": 8.698114964382598e-06,
      "loss": 0.0661,
      "step": 1340
    },
    {
      "epoch": 4.397394136807818,
      "grad_norm": 0.34097039699554443,
      "learning_rate": 7.840426448358085e-06,
      "loss": 0.0155,
      "step": 1350
    },
    {
      "epoch": 4.429967426710098,
      "grad_norm": 2.330317497253418,
      "learning_rate": 7.025524588862542e-06,
      "loss": 0.0291,
      "step": 1360
    },
    {
      "epoch": 4.462540716612378,
      "grad_norm": 0.39467132091522217,
      "learning_rate": 6.253787718006498e-06,
      "loss": 0.0568,
      "step": 1370
    },
    {
      "epoch": 4.495114006514658,
      "grad_norm": 0.28935298323631287,
      "learning_rate": 5.525574127817046e-06,
      "loss": 0.0302,
      "step": 1380
    },
    {
      "epoch": 4.527687296416938,
      "grad_norm": 0.3630921542644501,
      "learning_rate": 4.841221903894633e-06,
      "loss": 0.027,
      "step": 1390
    },
    {
      "epoch": 4.5602605863192185,
      "grad_norm": 0.1575985699892044,
      "learning_rate": 4.20104876845111e-06,
      "loss": 0.0208,
      "step": 1400
    },
    {
      "epoch": 4.592833876221499,
      "grad_norm": 0.3565620481967926,
      "learning_rate": 3.605351932801693e-06,
      "loss": 0.0725,
      "step": 1410
    },
    {
      "epoch": 4.625407166123779,
      "grad_norm": 0.17438632249832153,
      "learning_rate": 3.0544079593795573e-06,
      "loss": 0.0167,
      "step": 1420
    },
    {
      "epoch": 4.657980456026059,
      "grad_norm": 0.15625396370887756,
      "learning_rate": 2.548472633337007e-06,
      "loss": 0.0162,
      "step": 1430
    },
    {
      "epoch": 4.690553745928339,
      "grad_norm": 0.21613460779190063,
      "learning_rate": 2.0877808437928637e-06,
      "loss": 0.02,
      "step": 1440
    },
    {
      "epoch": 4.723127035830619,
      "grad_norm": 0.25940829515457153,
      "learning_rate": 1.6725464747811447e-06,
      "loss": 0.0327,
      "step": 1450
    },
    {
      "epoch": 4.755700325732899,
      "grad_norm": 0.6820478439331055,
      "learning_rate": 1.3029623059517493e-06,
      "loss": 0.0182,
      "step": 1460
    },
    {
      "epoch": 4.7882736156351795,
      "grad_norm": 0.4143361747264862,
      "learning_rate": 9.791999230692629e-07,
      "loss": 0.0236,
      "step": 1470
    },
    {
      "epoch": 4.82084690553746,
      "grad_norm": 0.18389806151390076,
      "learning_rate": 7.014096383512802e-07,
      "loss": 0.0245,
      "step": 1480
    },
    {
      "epoch": 4.85342019543974,
      "grad_norm": 0.1316615343093872,
      "learning_rate": 4.6972042068341714e-07,
      "loss": 0.0162,
      "step": 1490
    },
    {
      "epoch": 4.88599348534202,
      "grad_norm": 0.7174280881881714,
      "learning_rate": 2.8423983574328295e-07,
      "loss": 0.0292,
      "step": 1500
    },
    {
      "epoch": 4.918566775244299,
      "grad_norm": 0.36342909932136536,
      "learning_rate": 1.4505399606130621e-07,
      "loss": 0.0269,
      "step": 1510
    },
    {
      "epoch": 4.95114006514658,
      "grad_norm": 0.12144973129034042,
      "learning_rate": 5.2227521041470216e-08,
      "loss": 0.0193,
      "step": 1520
    },
    {
      "epoch": 4.9837133550488595,
      "grad_norm": 0.2756155729293823,
      "learning_rate": 5.803506960722072e-09,
      "loss": 0.0375,
      "step": 1530
    },
    {
      "epoch": 5.0,
      "step": 1535,
      "total_flos": 5.197521391244208e+16,
      "train_loss": 0.10585675648980886,
      "train_runtime": 733.8736,
      "train_samples_per_second": 33.466,
      "train_steps_per_second": 2.092
    }
  ],
  "logging_steps": 10,
  "max_steps": 1535,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 5.197521391244208e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}