trapoom555's picture
Upload checkpoints
d453b94
raw
history blame
No virus
73.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.32653653581017344,
"eval_steps": 500,
"global_step": 4500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007256367462448298,
"grad_norm": 13.1875,
"learning_rate": 5e-06,
"loss": 3.553,
"step": 10
},
{
"epoch": 0.0014512734924896596,
"grad_norm": 9.375,
"learning_rate": 1e-05,
"loss": 3.5658,
"step": 20
},
{
"epoch": 0.0021769102387344894,
"grad_norm": 8.8125,
"learning_rate": 1.5e-05,
"loss": 3.4669,
"step": 30
},
{
"epoch": 0.0029025469849793192,
"grad_norm": 13.25,
"learning_rate": 2e-05,
"loss": 3.5007,
"step": 40
},
{
"epoch": 0.003628183731224149,
"grad_norm": 9.3125,
"learning_rate": 2.5e-05,
"loss": 3.5062,
"step": 50
},
{
"epoch": 0.004353820477468979,
"grad_norm": 9.9375,
"learning_rate": 3e-05,
"loss": 3.4178,
"step": 60
},
{
"epoch": 0.005079457223713809,
"grad_norm": 17.125,
"learning_rate": 3.5e-05,
"loss": 3.4136,
"step": 70
},
{
"epoch": 0.0058050939699586385,
"grad_norm": 17.5,
"learning_rate": 4e-05,
"loss": 3.3792,
"step": 80
},
{
"epoch": 0.006530730716203468,
"grad_norm": 17.0,
"learning_rate": 4.5e-05,
"loss": 3.3684,
"step": 90
},
{
"epoch": 0.007256367462448298,
"grad_norm": 21.875,
"learning_rate": 5e-05,
"loss": 3.2698,
"step": 100
},
{
"epoch": 0.007982004208693128,
"grad_norm": 32.5,
"learning_rate": 4.9999934086574596e-05,
"loss": 3.1523,
"step": 110
},
{
"epoch": 0.008707640954937958,
"grad_norm": 73.5,
"learning_rate": 4.9999736346645943e-05,
"loss": 3.0884,
"step": 120
},
{
"epoch": 0.009433277701182788,
"grad_norm": 34.75,
"learning_rate": 4.999940678125673e-05,
"loss": 2.9507,
"step": 130
},
{
"epoch": 0.010158914447427617,
"grad_norm": 47.5,
"learning_rate": 4.9998945392144796e-05,
"loss": 2.8659,
"step": 140
},
{
"epoch": 0.010884551193672447,
"grad_norm": 35.25,
"learning_rate": 4.999835218174307e-05,
"loss": 2.9846,
"step": 150
},
{
"epoch": 0.011610187939917277,
"grad_norm": 27.5,
"learning_rate": 4.99976271531796e-05,
"loss": 2.8913,
"step": 160
},
{
"epoch": 0.012335824686162107,
"grad_norm": 38.5,
"learning_rate": 4.9996770310277506e-05,
"loss": 2.8982,
"step": 170
},
{
"epoch": 0.013061461432406937,
"grad_norm": 52.0,
"learning_rate": 4.9995781657555e-05,
"loss": 2.7165,
"step": 180
},
{
"epoch": 0.013787098178651766,
"grad_norm": 34.75,
"learning_rate": 4.99946612002253e-05,
"loss": 2.4166,
"step": 190
},
{
"epoch": 0.014512734924896596,
"grad_norm": 40.25,
"learning_rate": 4.9993408944196676e-05,
"loss": 2.2161,
"step": 200
},
{
"epoch": 0.015238371671141426,
"grad_norm": 39.25,
"learning_rate": 4.9992024896072364e-05,
"loss": 1.7825,
"step": 210
},
{
"epoch": 0.015964008417386256,
"grad_norm": 37.25,
"learning_rate": 4.999050906315055e-05,
"loss": 1.46,
"step": 220
},
{
"epoch": 0.016689645163631087,
"grad_norm": 13.625,
"learning_rate": 4.998886145342434e-05,
"loss": 1.3157,
"step": 230
},
{
"epoch": 0.017415281909875915,
"grad_norm": 10.4375,
"learning_rate": 4.9987082075581684e-05,
"loss": 1.1408,
"step": 240
},
{
"epoch": 0.018140918656120747,
"grad_norm": 14.0,
"learning_rate": 4.9985170939005386e-05,
"loss": 0.8782,
"step": 250
},
{
"epoch": 0.018866555402365575,
"grad_norm": 22.0,
"learning_rate": 4.998312805377302e-05,
"loss": 0.9109,
"step": 260
},
{
"epoch": 0.019592192148610407,
"grad_norm": 12.625,
"learning_rate": 4.998095343065685e-05,
"loss": 0.7414,
"step": 270
},
{
"epoch": 0.020317828894855235,
"grad_norm": 15.8125,
"learning_rate": 4.997864708112384e-05,
"loss": 0.7926,
"step": 280
},
{
"epoch": 0.021043465641100066,
"grad_norm": 16.25,
"learning_rate": 4.997620901733554e-05,
"loss": 0.7382,
"step": 290
},
{
"epoch": 0.021769102387344894,
"grad_norm": 22.0,
"learning_rate": 4.997363925214803e-05,
"loss": 0.7158,
"step": 300
},
{
"epoch": 0.022494739133589726,
"grad_norm": 13.375,
"learning_rate": 4.9970937799111896e-05,
"loss": 0.526,
"step": 310
},
{
"epoch": 0.023220375879834554,
"grad_norm": 12.9375,
"learning_rate": 4.996810467247207e-05,
"loss": 0.6137,
"step": 320
},
{
"epoch": 0.023946012626079385,
"grad_norm": 6.78125,
"learning_rate": 4.9965139887167856e-05,
"loss": 0.4864,
"step": 330
},
{
"epoch": 0.024671649372324213,
"grad_norm": 10.25,
"learning_rate": 4.996204345883278e-05,
"loss": 0.6921,
"step": 340
},
{
"epoch": 0.025397286118569045,
"grad_norm": 6.3125,
"learning_rate": 4.9958815403794546e-05,
"loss": 0.5826,
"step": 350
},
{
"epoch": 0.026122922864813873,
"grad_norm": 10.625,
"learning_rate": 4.995545573907492e-05,
"loss": 0.5838,
"step": 360
},
{
"epoch": 0.026848559611058705,
"grad_norm": 17.25,
"learning_rate": 4.995196448238966e-05,
"loss": 0.5941,
"step": 370
},
{
"epoch": 0.027574196357303533,
"grad_norm": 9.875,
"learning_rate": 4.9948341652148436e-05,
"loss": 0.5055,
"step": 380
},
{
"epoch": 0.028299833103548364,
"grad_norm": 21.25,
"learning_rate": 4.994458726745468e-05,
"loss": 0.6114,
"step": 390
},
{
"epoch": 0.029025469849793192,
"grad_norm": 3.375,
"learning_rate": 4.9940701348105554e-05,
"loss": 0.5372,
"step": 400
},
{
"epoch": 0.029751106596038024,
"grad_norm": 12.625,
"learning_rate": 4.99366839145918e-05,
"loss": 0.4815,
"step": 410
},
{
"epoch": 0.030476743342282852,
"grad_norm": 10.8125,
"learning_rate": 4.993253498809762e-05,
"loss": 0.4903,
"step": 420
},
{
"epoch": 0.031202380088527683,
"grad_norm": 9.0,
"learning_rate": 4.9928254590500646e-05,
"loss": 0.5927,
"step": 430
},
{
"epoch": 0.03192801683477251,
"grad_norm": 6.71875,
"learning_rate": 4.9923842744371707e-05,
"loss": 0.4424,
"step": 440
},
{
"epoch": 0.03265365358101734,
"grad_norm": 5.0,
"learning_rate": 4.99192994729748e-05,
"loss": 0.362,
"step": 450
},
{
"epoch": 0.033379290327262175,
"grad_norm": 8.625,
"learning_rate": 4.991462480026693e-05,
"loss": 0.593,
"step": 460
},
{
"epoch": 0.034104927073507,
"grad_norm": 4.53125,
"learning_rate": 4.9909818750898e-05,
"loss": 0.4769,
"step": 470
},
{
"epoch": 0.03483056381975183,
"grad_norm": 11.125,
"learning_rate": 4.990488135021065e-05,
"loss": 0.4455,
"step": 480
},
{
"epoch": 0.03555620056599666,
"grad_norm": 8.375,
"learning_rate": 4.989981262424017e-05,
"loss": 0.4713,
"step": 490
},
{
"epoch": 0.036281837312241494,
"grad_norm": 7.75,
"learning_rate": 4.989461259971432e-05,
"loss": 0.2778,
"step": 500
},
{
"epoch": 0.03700747405848632,
"grad_norm": 5.65625,
"learning_rate": 4.988928130405323e-05,
"loss": 0.3604,
"step": 510
},
{
"epoch": 0.03773311080473115,
"grad_norm": 7.5625,
"learning_rate": 4.9883818765369194e-05,
"loss": 0.4234,
"step": 520
},
{
"epoch": 0.03845874755097598,
"grad_norm": 8.9375,
"learning_rate": 4.98782250124666e-05,
"loss": 0.3531,
"step": 530
},
{
"epoch": 0.03918438429722081,
"grad_norm": 6.5,
"learning_rate": 4.987250007484172e-05,
"loss": 0.3281,
"step": 540
},
{
"epoch": 0.03991002104346564,
"grad_norm": 4.59375,
"learning_rate": 4.986664398268256e-05,
"loss": 0.3477,
"step": 550
},
{
"epoch": 0.04063565778971047,
"grad_norm": 6.4375,
"learning_rate": 4.986065676686874e-05,
"loss": 0.369,
"step": 560
},
{
"epoch": 0.0413612945359553,
"grad_norm": 8.0625,
"learning_rate": 4.98545384589713e-05,
"loss": 0.3068,
"step": 570
},
{
"epoch": 0.04208693128220013,
"grad_norm": 6.28125,
"learning_rate": 4.984828909125251e-05,
"loss": 0.2926,
"step": 580
},
{
"epoch": 0.042812568028444964,
"grad_norm": 4.4375,
"learning_rate": 4.9841908696665764e-05,
"loss": 0.2771,
"step": 590
},
{
"epoch": 0.04353820477468979,
"grad_norm": 4.15625,
"learning_rate": 4.9835397308855344e-05,
"loss": 0.3676,
"step": 600
},
{
"epoch": 0.04426384152093462,
"grad_norm": 4.4375,
"learning_rate": 4.9828754962156286e-05,
"loss": 0.3307,
"step": 610
},
{
"epoch": 0.04498947826717945,
"grad_norm": 13.75,
"learning_rate": 4.9821981691594175e-05,
"loss": 0.3601,
"step": 620
},
{
"epoch": 0.04571511501342428,
"grad_norm": 4.59375,
"learning_rate": 4.981507753288497e-05,
"loss": 0.2939,
"step": 630
},
{
"epoch": 0.04644075175966911,
"grad_norm": 6.1875,
"learning_rate": 4.9808042522434814e-05,
"loss": 0.293,
"step": 640
},
{
"epoch": 0.04716638850591394,
"grad_norm": 4.65625,
"learning_rate": 4.9800876697339824e-05,
"loss": 0.3213,
"step": 650
},
{
"epoch": 0.04789202525215877,
"grad_norm": 7.65625,
"learning_rate": 4.979358009538594e-05,
"loss": 0.3903,
"step": 660
},
{
"epoch": 0.0486176619984036,
"grad_norm": 9.3125,
"learning_rate": 4.978615275504869e-05,
"loss": 0.2936,
"step": 670
},
{
"epoch": 0.04934329874464843,
"grad_norm": 8.4375,
"learning_rate": 4.977859471549297e-05,
"loss": 0.3681,
"step": 680
},
{
"epoch": 0.05006893549089326,
"grad_norm": 6.28125,
"learning_rate": 4.977090601657289e-05,
"loss": 0.2591,
"step": 690
},
{
"epoch": 0.05079457223713809,
"grad_norm": 10.8125,
"learning_rate": 4.976308669883153e-05,
"loss": 0.3351,
"step": 700
},
{
"epoch": 0.05152020898338292,
"grad_norm": 6.65625,
"learning_rate": 4.975513680350073e-05,
"loss": 0.3042,
"step": 710
},
{
"epoch": 0.052245845729627746,
"grad_norm": 6.65625,
"learning_rate": 4.974705637250089e-05,
"loss": 0.2396,
"step": 720
},
{
"epoch": 0.05297148247587258,
"grad_norm": 6.1875,
"learning_rate": 4.97388454484407e-05,
"loss": 0.2892,
"step": 730
},
{
"epoch": 0.05369711922211741,
"grad_norm": 3.78125,
"learning_rate": 4.973050407461698e-05,
"loss": 0.3032,
"step": 740
},
{
"epoch": 0.05442275596836224,
"grad_norm": 2.65625,
"learning_rate": 4.972203229501441e-05,
"loss": 0.2522,
"step": 750
},
{
"epoch": 0.055148392714607065,
"grad_norm": 7.84375,
"learning_rate": 4.971343015430532e-05,
"loss": 0.2502,
"step": 760
},
{
"epoch": 0.0558740294608519,
"grad_norm": 10.6875,
"learning_rate": 4.970469769784941e-05,
"loss": 0.2664,
"step": 770
},
{
"epoch": 0.05659966620709673,
"grad_norm": 7.90625,
"learning_rate": 4.96958349716936e-05,
"loss": 0.332,
"step": 780
},
{
"epoch": 0.05732530295334156,
"grad_norm": 10.75,
"learning_rate": 4.968684202257169e-05,
"loss": 0.3737,
"step": 790
},
{
"epoch": 0.058050939699586385,
"grad_norm": 6.5,
"learning_rate": 4.967771889790416e-05,
"loss": 0.2794,
"step": 800
},
{
"epoch": 0.058776576445831216,
"grad_norm": 2.921875,
"learning_rate": 4.966846564579792e-05,
"loss": 0.2375,
"step": 810
},
{
"epoch": 0.05950221319207605,
"grad_norm": 6.09375,
"learning_rate": 4.965908231504607e-05,
"loss": 0.365,
"step": 820
},
{
"epoch": 0.06022784993832088,
"grad_norm": 9.0625,
"learning_rate": 4.964956895512759e-05,
"loss": 0.3469,
"step": 830
},
{
"epoch": 0.060953486684565704,
"grad_norm": 5.25,
"learning_rate": 4.963992561620714e-05,
"loss": 0.2904,
"step": 840
},
{
"epoch": 0.061679123430810535,
"grad_norm": 5.5,
"learning_rate": 4.963015234913475e-05,
"loss": 0.2779,
"step": 850
},
{
"epoch": 0.06240476017705537,
"grad_norm": 5.6875,
"learning_rate": 4.96202492054456e-05,
"loss": 0.2658,
"step": 860
},
{
"epoch": 0.06313039692330019,
"grad_norm": 8.8125,
"learning_rate": 4.9610216237359684e-05,
"loss": 0.3973,
"step": 870
},
{
"epoch": 0.06385603366954502,
"grad_norm": 10.0,
"learning_rate": 4.960005349778159e-05,
"loss": 0.3169,
"step": 880
},
{
"epoch": 0.06458167041578985,
"grad_norm": 11.9375,
"learning_rate": 4.95897610403002e-05,
"loss": 0.3172,
"step": 890
},
{
"epoch": 0.06530730716203469,
"grad_norm": 3.0,
"learning_rate": 4.95793389191884e-05,
"loss": 0.2664,
"step": 900
},
{
"epoch": 0.06603294390827952,
"grad_norm": 6.90625,
"learning_rate": 4.95687871894028e-05,
"loss": 0.2507,
"step": 910
},
{
"epoch": 0.06675858065452435,
"grad_norm": 3.796875,
"learning_rate": 4.9558105906583466e-05,
"loss": 0.2383,
"step": 920
},
{
"epoch": 0.06748421740076918,
"grad_norm": 6.28125,
"learning_rate": 4.9547295127053586e-05,
"loss": 0.3942,
"step": 930
},
{
"epoch": 0.068209854147014,
"grad_norm": 9.0,
"learning_rate": 4.95363549078192e-05,
"loss": 0.2427,
"step": 940
},
{
"epoch": 0.06893549089325883,
"grad_norm": 5.59375,
"learning_rate": 4.952528530656889e-05,
"loss": 0.2351,
"step": 950
},
{
"epoch": 0.06966112763950366,
"grad_norm": 9.5625,
"learning_rate": 4.9514086381673496e-05,
"loss": 0.282,
"step": 960
},
{
"epoch": 0.0703867643857485,
"grad_norm": 10.0625,
"learning_rate": 4.9502758192185774e-05,
"loss": 0.2687,
"step": 970
},
{
"epoch": 0.07111240113199332,
"grad_norm": 9.0625,
"learning_rate": 4.949130079784009e-05,
"loss": 0.3135,
"step": 980
},
{
"epoch": 0.07183803787823816,
"grad_norm": 5.59375,
"learning_rate": 4.9479714259052143e-05,
"loss": 0.3207,
"step": 990
},
{
"epoch": 0.07256367462448299,
"grad_norm": 8.3125,
"learning_rate": 4.946799863691862e-05,
"loss": 0.2572,
"step": 1000
},
{
"epoch": 0.07328931137072782,
"grad_norm": 10.75,
"learning_rate": 4.9456153993216855e-05,
"loss": 0.3406,
"step": 1010
},
{
"epoch": 0.07401494811697264,
"grad_norm": 5.1875,
"learning_rate": 4.944418039040453e-05,
"loss": 0.288,
"step": 1020
},
{
"epoch": 0.07474058486321747,
"grad_norm": 5.375,
"learning_rate": 4.943207789161934e-05,
"loss": 0.2024,
"step": 1030
},
{
"epoch": 0.0754662216094623,
"grad_norm": 1.671875,
"learning_rate": 4.9419846560678657e-05,
"loss": 0.2623,
"step": 1040
},
{
"epoch": 0.07619185835570713,
"grad_norm": 8.5625,
"learning_rate": 4.940748646207919e-05,
"loss": 0.3,
"step": 1050
},
{
"epoch": 0.07691749510195196,
"grad_norm": 8.8125,
"learning_rate": 4.9394997660996664e-05,
"loss": 0.2651,
"step": 1060
},
{
"epoch": 0.0776431318481968,
"grad_norm": 4.0,
"learning_rate": 4.938238022328545e-05,
"loss": 0.2841,
"step": 1070
},
{
"epoch": 0.07836876859444163,
"grad_norm": 9.5625,
"learning_rate": 4.9369634215478225e-05,
"loss": 0.256,
"step": 1080
},
{
"epoch": 0.07909440534068646,
"grad_norm": 3.421875,
"learning_rate": 4.935675970478563e-05,
"loss": 0.1948,
"step": 1090
},
{
"epoch": 0.07982004208693128,
"grad_norm": 6.46875,
"learning_rate": 4.9343756759095924e-05,
"loss": 0.2309,
"step": 1100
},
{
"epoch": 0.08054567883317611,
"grad_norm": 8.6875,
"learning_rate": 4.933062544697459e-05,
"loss": 0.2647,
"step": 1110
},
{
"epoch": 0.08127131557942094,
"grad_norm": 4.71875,
"learning_rate": 4.931736583766402e-05,
"loss": 0.3253,
"step": 1120
},
{
"epoch": 0.08199695232566577,
"grad_norm": 6.125,
"learning_rate": 4.9303978001083105e-05,
"loss": 0.2177,
"step": 1130
},
{
"epoch": 0.0827225890719106,
"grad_norm": 6.78125,
"learning_rate": 4.929046200782691e-05,
"loss": 0.2809,
"step": 1140
},
{
"epoch": 0.08344822581815543,
"grad_norm": 8.9375,
"learning_rate": 4.927681792916626e-05,
"loss": 0.3722,
"step": 1150
},
{
"epoch": 0.08417386256440026,
"grad_norm": 5.71875,
"learning_rate": 4.926304583704739e-05,
"loss": 0.2601,
"step": 1160
},
{
"epoch": 0.0848994993106451,
"grad_norm": 5.21875,
"learning_rate": 4.924914580409157e-05,
"loss": 0.3125,
"step": 1170
},
{
"epoch": 0.08562513605688993,
"grad_norm": 4.125,
"learning_rate": 4.9235117903594696e-05,
"loss": 0.2442,
"step": 1180
},
{
"epoch": 0.08635077280313475,
"grad_norm": 6.6875,
"learning_rate": 4.922096220952692e-05,
"loss": 0.2837,
"step": 1190
},
{
"epoch": 0.08707640954937958,
"grad_norm": 7.5625,
"learning_rate": 4.9206678796532285e-05,
"loss": 0.2564,
"step": 1200
},
{
"epoch": 0.08780204629562441,
"grad_norm": 12.0625,
"learning_rate": 4.919226773992826e-05,
"loss": 0.2853,
"step": 1210
},
{
"epoch": 0.08852768304186924,
"grad_norm": 8.0,
"learning_rate": 4.917772911570544e-05,
"loss": 0.2109,
"step": 1220
},
{
"epoch": 0.08925331978811407,
"grad_norm": 4.90625,
"learning_rate": 4.916306300052704e-05,
"loss": 0.281,
"step": 1230
},
{
"epoch": 0.0899789565343589,
"grad_norm": 3.9375,
"learning_rate": 4.91482694717286e-05,
"loss": 0.3385,
"step": 1240
},
{
"epoch": 0.09070459328060373,
"grad_norm": 2.5625,
"learning_rate": 4.9133348607317464e-05,
"loss": 0.2559,
"step": 1250
},
{
"epoch": 0.09143023002684857,
"grad_norm": 1.796875,
"learning_rate": 4.911830048597247e-05,
"loss": 0.2376,
"step": 1260
},
{
"epoch": 0.09215586677309338,
"grad_norm": 6.25,
"learning_rate": 4.910312518704348e-05,
"loss": 0.2155,
"step": 1270
},
{
"epoch": 0.09288150351933822,
"grad_norm": 6.96875,
"learning_rate": 4.908782279055095e-05,
"loss": 0.2329,
"step": 1280
},
{
"epoch": 0.09360714026558305,
"grad_norm": 7.59375,
"learning_rate": 4.907239337718557e-05,
"loss": 0.2468,
"step": 1290
},
{
"epoch": 0.09433277701182788,
"grad_norm": 5.875,
"learning_rate": 4.9056837028307755e-05,
"loss": 0.2792,
"step": 1300
},
{
"epoch": 0.09505841375807271,
"grad_norm": 3.78125,
"learning_rate": 4.9041153825947314e-05,
"loss": 0.1919,
"step": 1310
},
{
"epoch": 0.09578405050431754,
"grad_norm": 5.65625,
"learning_rate": 4.902534385280292e-05,
"loss": 0.2775,
"step": 1320
},
{
"epoch": 0.09650968725056237,
"grad_norm": 8.625,
"learning_rate": 4.900940719224173e-05,
"loss": 0.3544,
"step": 1330
},
{
"epoch": 0.0972353239968072,
"grad_norm": 11.8125,
"learning_rate": 4.899334392829893e-05,
"loss": 0.3571,
"step": 1340
},
{
"epoch": 0.09796096074305202,
"grad_norm": 10.25,
"learning_rate": 4.897715414567732e-05,
"loss": 0.309,
"step": 1350
},
{
"epoch": 0.09868659748929685,
"grad_norm": 9.6875,
"learning_rate": 4.8960837929746794e-05,
"loss": 0.2712,
"step": 1360
},
{
"epoch": 0.09941223423554169,
"grad_norm": 2.265625,
"learning_rate": 4.894439536654399e-05,
"loss": 0.2685,
"step": 1370
},
{
"epoch": 0.10013787098178652,
"grad_norm": 6.53125,
"learning_rate": 4.892782654277175e-05,
"loss": 0.266,
"step": 1380
},
{
"epoch": 0.10086350772803135,
"grad_norm": 4.25,
"learning_rate": 4.8911131545798715e-05,
"loss": 0.2936,
"step": 1390
},
{
"epoch": 0.10158914447427618,
"grad_norm": 10.9375,
"learning_rate": 4.889431046365883e-05,
"loss": 0.1927,
"step": 1400
},
{
"epoch": 0.10231478122052101,
"grad_norm": 5.875,
"learning_rate": 4.887736338505092e-05,
"loss": 0.2864,
"step": 1410
},
{
"epoch": 0.10304041796676584,
"grad_norm": 3.875,
"learning_rate": 4.886029039933817e-05,
"loss": 0.2369,
"step": 1420
},
{
"epoch": 0.10376605471301066,
"grad_norm": 0.91015625,
"learning_rate": 4.884309159654771e-05,
"loss": 0.2904,
"step": 1430
},
{
"epoch": 0.10449169145925549,
"grad_norm": 9.0,
"learning_rate": 4.8825767067370086e-05,
"loss": 0.2294,
"step": 1440
},
{
"epoch": 0.10521732820550032,
"grad_norm": 6.84375,
"learning_rate": 4.880831690315884e-05,
"loss": 0.2302,
"step": 1450
},
{
"epoch": 0.10594296495174516,
"grad_norm": 8.25,
"learning_rate": 4.879074119592998e-05,
"loss": 0.3332,
"step": 1460
},
{
"epoch": 0.10666860169798999,
"grad_norm": 7.9375,
"learning_rate": 4.877304003836149e-05,
"loss": 0.2783,
"step": 1470
},
{
"epoch": 0.10739423844423482,
"grad_norm": 3.84375,
"learning_rate": 4.87552135237929e-05,
"loss": 0.2884,
"step": 1480
},
{
"epoch": 0.10811987519047965,
"grad_norm": 6.4375,
"learning_rate": 4.873726174622475e-05,
"loss": 0.3063,
"step": 1490
},
{
"epoch": 0.10884551193672448,
"grad_norm": 7.78125,
"learning_rate": 4.8719184800318066e-05,
"loss": 0.2767,
"step": 1500
},
{
"epoch": 0.1095711486829693,
"grad_norm": 6.59375,
"learning_rate": 4.870098278139394e-05,
"loss": 0.2703,
"step": 1510
},
{
"epoch": 0.11029678542921413,
"grad_norm": 9.3125,
"learning_rate": 4.8682655785432975e-05,
"loss": 0.2799,
"step": 1520
},
{
"epoch": 0.11102242217545896,
"grad_norm": 3.984375,
"learning_rate": 4.866420390907476e-05,
"loss": 0.2277,
"step": 1530
},
{
"epoch": 0.1117480589217038,
"grad_norm": 3.59375,
"learning_rate": 4.8645627249617406e-05,
"loss": 0.2159,
"step": 1540
},
{
"epoch": 0.11247369566794863,
"grad_norm": 7.0625,
"learning_rate": 4.8626925905017017e-05,
"loss": 0.1614,
"step": 1550
},
{
"epoch": 0.11319933241419346,
"grad_norm": 13.6875,
"learning_rate": 4.8608099973887167e-05,
"loss": 0.2458,
"step": 1560
},
{
"epoch": 0.11392496916043829,
"grad_norm": 6.1875,
"learning_rate": 4.858914955549839e-05,
"loss": 0.2246,
"step": 1570
},
{
"epoch": 0.11465060590668312,
"grad_norm": 8.0,
"learning_rate": 4.857007474977764e-05,
"loss": 0.3887,
"step": 1580
},
{
"epoch": 0.11537624265292794,
"grad_norm": 1.296875,
"learning_rate": 4.855087565730778e-05,
"loss": 0.2589,
"step": 1590
},
{
"epoch": 0.11610187939917277,
"grad_norm": 6.09375,
"learning_rate": 4.853155237932705e-05,
"loss": 0.276,
"step": 1600
},
{
"epoch": 0.1168275161454176,
"grad_norm": 2.71875,
"learning_rate": 4.8512105017728514e-05,
"loss": 0.2841,
"step": 1610
},
{
"epoch": 0.11755315289166243,
"grad_norm": 3.484375,
"learning_rate": 4.8492533675059564e-05,
"loss": 0.2426,
"step": 1620
},
{
"epoch": 0.11827878963790726,
"grad_norm": 10.375,
"learning_rate": 4.847283845452133e-05,
"loss": 0.2559,
"step": 1630
},
{
"epoch": 0.1190044263841521,
"grad_norm": 5.4375,
"learning_rate": 4.8453019459968174e-05,
"loss": 0.2093,
"step": 1640
},
{
"epoch": 0.11973006313039693,
"grad_norm": 9.75,
"learning_rate": 4.843307679590711e-05,
"loss": 0.3073,
"step": 1650
},
{
"epoch": 0.12045569987664176,
"grad_norm": 6.84375,
"learning_rate": 4.8413010567497295e-05,
"loss": 0.2148,
"step": 1660
},
{
"epoch": 0.12118133662288658,
"grad_norm": 3.765625,
"learning_rate": 4.8392820880549434e-05,
"loss": 0.1865,
"step": 1670
},
{
"epoch": 0.12190697336913141,
"grad_norm": 4.1875,
"learning_rate": 4.837250784152524e-05,
"loss": 0.3362,
"step": 1680
},
{
"epoch": 0.12263261011537624,
"grad_norm": 1.578125,
"learning_rate": 4.835207155753686e-05,
"loss": 0.1977,
"step": 1690
},
{
"epoch": 0.12335824686162107,
"grad_norm": 5.71875,
"learning_rate": 4.833151213634635e-05,
"loss": 0.2128,
"step": 1700
},
{
"epoch": 0.1240838836078659,
"grad_norm": 4.375,
"learning_rate": 4.831082968636505e-05,
"loss": 0.1508,
"step": 1710
},
{
"epoch": 0.12480952035411073,
"grad_norm": 3.28125,
"learning_rate": 4.8290024316653055e-05,
"loss": 0.1477,
"step": 1720
},
{
"epoch": 0.12553515710035557,
"grad_norm": 4.34375,
"learning_rate": 4.826909613691861e-05,
"loss": 0.164,
"step": 1730
},
{
"epoch": 0.12626079384660038,
"grad_norm": 5.25,
"learning_rate": 4.824804525751757e-05,
"loss": 0.2322,
"step": 1740
},
{
"epoch": 0.12698643059284523,
"grad_norm": 4.8125,
"learning_rate": 4.822687178945278e-05,
"loss": 0.2295,
"step": 1750
},
{
"epoch": 0.12771206733909005,
"grad_norm": 6.6875,
"learning_rate": 4.820557584437349e-05,
"loss": 0.2771,
"step": 1760
},
{
"epoch": 0.1284377040853349,
"grad_norm": 8.5,
"learning_rate": 4.8184157534574806e-05,
"loss": 0.3078,
"step": 1770
},
{
"epoch": 0.1291633408315797,
"grad_norm": 5.0625,
"learning_rate": 4.816261697299706e-05,
"loss": 0.2051,
"step": 1780
},
{
"epoch": 0.12988897757782455,
"grad_norm": 13.1875,
"learning_rate": 4.814095427322522e-05,
"loss": 0.2311,
"step": 1790
},
{
"epoch": 0.13061461432406937,
"grad_norm": 5.59375,
"learning_rate": 4.811916954948832e-05,
"loss": 0.2827,
"step": 1800
},
{
"epoch": 0.1313402510703142,
"grad_norm": 7.90625,
"learning_rate": 4.809726291665881e-05,
"loss": 0.2158,
"step": 1810
},
{
"epoch": 0.13206588781655904,
"grad_norm": 6.875,
"learning_rate": 4.8075234490251994e-05,
"loss": 0.2128,
"step": 1820
},
{
"epoch": 0.13279152456280385,
"grad_norm": 12.5625,
"learning_rate": 4.805308438642539e-05,
"loss": 0.2479,
"step": 1830
},
{
"epoch": 0.1335171613090487,
"grad_norm": 6.46875,
"learning_rate": 4.8030812721978135e-05,
"loss": 0.2439,
"step": 1840
},
{
"epoch": 0.13424279805529352,
"grad_norm": 10.125,
"learning_rate": 4.800841961435037e-05,
"loss": 0.2566,
"step": 1850
},
{
"epoch": 0.13496843480153836,
"grad_norm": 9.6875,
"learning_rate": 4.7985905181622605e-05,
"loss": 0.2342,
"step": 1860
},
{
"epoch": 0.13569407154778318,
"grad_norm": 2.25,
"learning_rate": 4.7963269542515106e-05,
"loss": 0.1505,
"step": 1870
},
{
"epoch": 0.136419708294028,
"grad_norm": 3.640625,
"learning_rate": 4.794051281638729e-05,
"loss": 0.2658,
"step": 1880
},
{
"epoch": 0.13714534504027284,
"grad_norm": 7.21875,
"learning_rate": 4.7917635123237025e-05,
"loss": 0.3074,
"step": 1890
},
{
"epoch": 0.13787098178651766,
"grad_norm": 2.65625,
"learning_rate": 4.789463658370013e-05,
"loss": 0.2701,
"step": 1900
},
{
"epoch": 0.1385966185327625,
"grad_norm": 5.84375,
"learning_rate": 4.7871517319049566e-05,
"loss": 0.2645,
"step": 1910
},
{
"epoch": 0.13932225527900732,
"grad_norm": 3.6875,
"learning_rate": 4.784827745119493e-05,
"loss": 0.2724,
"step": 1920
},
{
"epoch": 0.14004789202525217,
"grad_norm": 0.9375,
"learning_rate": 4.7824917102681796e-05,
"loss": 0.21,
"step": 1930
},
{
"epoch": 0.140773528771497,
"grad_norm": 5.96875,
"learning_rate": 4.780143639669098e-05,
"loss": 0.1734,
"step": 1940
},
{
"epoch": 0.14149916551774183,
"grad_norm": 6.84375,
"learning_rate": 4.7777835457038e-05,
"loss": 0.2173,
"step": 1950
},
{
"epoch": 0.14222480226398665,
"grad_norm": 4.5,
"learning_rate": 4.775411440817236e-05,
"loss": 0.2905,
"step": 1960
},
{
"epoch": 0.14295043901023147,
"grad_norm": 3.03125,
"learning_rate": 4.7730273375176885e-05,
"loss": 0.2289,
"step": 1970
},
{
"epoch": 0.1436760757564763,
"grad_norm": 12.4375,
"learning_rate": 4.770631248376713e-05,
"loss": 0.2817,
"step": 1980
},
{
"epoch": 0.14440171250272113,
"grad_norm": 3.8125,
"learning_rate": 4.768223186029064e-05,
"loss": 0.3234,
"step": 1990
},
{
"epoch": 0.14512734924896598,
"grad_norm": 11.9375,
"learning_rate": 4.765803163172632e-05,
"loss": 0.3516,
"step": 2000
},
{
"epoch": 0.1458529859952108,
"grad_norm": 4.8125,
"learning_rate": 4.7633711925683786e-05,
"loss": 0.1774,
"step": 2010
},
{
"epoch": 0.14657862274145564,
"grad_norm": 4.875,
"learning_rate": 4.760927287040262e-05,
"loss": 0.2695,
"step": 2020
},
{
"epoch": 0.14730425948770046,
"grad_norm": 7.5,
"learning_rate": 4.75847145947518e-05,
"loss": 0.2207,
"step": 2030
},
{
"epoch": 0.14802989623394527,
"grad_norm": 6.03125,
"learning_rate": 4.756003722822891e-05,
"loss": 0.2679,
"step": 2040
},
{
"epoch": 0.14875553298019012,
"grad_norm": 5.8125,
"learning_rate": 4.753524090095953e-05,
"loss": 0.2325,
"step": 2050
},
{
"epoch": 0.14948116972643494,
"grad_norm": 5.4375,
"learning_rate": 4.751032574369655e-05,
"loss": 0.2521,
"step": 2060
},
{
"epoch": 0.15020680647267978,
"grad_norm": 1.8515625,
"learning_rate": 4.748529188781942e-05,
"loss": 0.2213,
"step": 2070
},
{
"epoch": 0.1509324432189246,
"grad_norm": 3.1875,
"learning_rate": 4.746013946533352e-05,
"loss": 0.2019,
"step": 2080
},
{
"epoch": 0.15165807996516945,
"grad_norm": 6.75,
"learning_rate": 4.7434868608869435e-05,
"loss": 0.2303,
"step": 2090
},
{
"epoch": 0.15238371671141426,
"grad_norm": 2.34375,
"learning_rate": 4.740947945168227e-05,
"loss": 0.2411,
"step": 2100
},
{
"epoch": 0.1531093534576591,
"grad_norm": 5.28125,
"learning_rate": 4.7383972127650924e-05,
"loss": 0.2548,
"step": 2110
},
{
"epoch": 0.15383499020390393,
"grad_norm": 2.640625,
"learning_rate": 4.73583467712774e-05,
"loss": 0.1987,
"step": 2120
},
{
"epoch": 0.15456062695014874,
"grad_norm": 10.3125,
"learning_rate": 4.733260351768611e-05,
"loss": 0.294,
"step": 2130
},
{
"epoch": 0.1552862636963936,
"grad_norm": 6.15625,
"learning_rate": 4.7306742502623124e-05,
"loss": 0.178,
"step": 2140
},
{
"epoch": 0.1560119004426384,
"grad_norm": 5.5,
"learning_rate": 4.72807638624555e-05,
"loss": 0.2031,
"step": 2150
},
{
"epoch": 0.15673753718888325,
"grad_norm": 3.296875,
"learning_rate": 4.7254667734170526e-05,
"loss": 0.1921,
"step": 2160
},
{
"epoch": 0.15746317393512807,
"grad_norm": 2.4375,
"learning_rate": 4.7228454255375024e-05,
"loss": 0.2581,
"step": 2170
},
{
"epoch": 0.15818881068137292,
"grad_norm": 7.125,
"learning_rate": 4.72021235642946e-05,
"loss": 0.2175,
"step": 2180
},
{
"epoch": 0.15891444742761773,
"grad_norm": 3.609375,
"learning_rate": 4.7175675799772935e-05,
"loss": 0.212,
"step": 2190
},
{
"epoch": 0.15964008417386255,
"grad_norm": 8.0625,
"learning_rate": 4.714911110127106e-05,
"loss": 0.1766,
"step": 2200
},
{
"epoch": 0.1603657209201074,
"grad_norm": 2.265625,
"learning_rate": 4.712242960886659e-05,
"loss": 0.2605,
"step": 2210
},
{
"epoch": 0.16109135766635221,
"grad_norm": 0.87890625,
"learning_rate": 4.709563146325301e-05,
"loss": 0.211,
"step": 2220
},
{
"epoch": 0.16181699441259706,
"grad_norm": 6.625,
"learning_rate": 4.7068716805738935e-05,
"loss": 0.2391,
"step": 2230
},
{
"epoch": 0.16254263115884188,
"grad_norm": 5.15625,
"learning_rate": 4.704168577824733e-05,
"loss": 0.259,
"step": 2240
},
{
"epoch": 0.16326826790508672,
"grad_norm": 8.5,
"learning_rate": 4.701453852331482e-05,
"loss": 0.2316,
"step": 2250
},
{
"epoch": 0.16399390465133154,
"grad_norm": 5.125,
"learning_rate": 4.6987275184090876e-05,
"loss": 0.2271,
"step": 2260
},
{
"epoch": 0.16471954139757639,
"grad_norm": 7.15625,
"learning_rate": 4.6959895904337115e-05,
"loss": 0.2099,
"step": 2270
},
{
"epoch": 0.1654451781438212,
"grad_norm": 4.03125,
"learning_rate": 4.6932400828426505e-05,
"loss": 0.1646,
"step": 2280
},
{
"epoch": 0.16617081489006602,
"grad_norm": 1.796875,
"learning_rate": 4.690479010134261e-05,
"loss": 0.2043,
"step": 2290
},
{
"epoch": 0.16689645163631087,
"grad_norm": 4.625,
"learning_rate": 4.687706386867885e-05,
"loss": 0.2699,
"step": 2300
},
{
"epoch": 0.16762208838255568,
"grad_norm": 5.1875,
"learning_rate": 4.6849222276637686e-05,
"loss": 0.2132,
"step": 2310
},
{
"epoch": 0.16834772512880053,
"grad_norm": 12.0625,
"learning_rate": 4.682126547202991e-05,
"loss": 0.2614,
"step": 2320
},
{
"epoch": 0.16907336187504535,
"grad_norm": 2.484375,
"learning_rate": 4.679319360227381e-05,
"loss": 0.1741,
"step": 2330
},
{
"epoch": 0.1697989986212902,
"grad_norm": 1.0625,
"learning_rate": 4.676500681539444e-05,
"loss": 0.2116,
"step": 2340
},
{
"epoch": 0.170524635367535,
"grad_norm": 5.0625,
"learning_rate": 4.673670526002281e-05,
"loss": 0.2688,
"step": 2350
},
{
"epoch": 0.17125027211377986,
"grad_norm": 2.15625,
"learning_rate": 4.6708289085395115e-05,
"loss": 0.2458,
"step": 2360
},
{
"epoch": 0.17197590886002467,
"grad_norm": 8.625,
"learning_rate": 4.6679758441351964e-05,
"loss": 0.1767,
"step": 2370
},
{
"epoch": 0.1727015456062695,
"grad_norm": 7.5,
"learning_rate": 4.665111347833754e-05,
"loss": 0.2735,
"step": 2380
},
{
"epoch": 0.17342718235251434,
"grad_norm": 5.0625,
"learning_rate": 4.662235434739885e-05,
"loss": 0.1716,
"step": 2390
},
{
"epoch": 0.17415281909875915,
"grad_norm": 8.625,
"learning_rate": 4.659348120018493e-05,
"loss": 0.2536,
"step": 2400
},
{
"epoch": 0.174878455845004,
"grad_norm": 2.046875,
"learning_rate": 4.6564494188946015e-05,
"loss": 0.2293,
"step": 2410
},
{
"epoch": 0.17560409259124882,
"grad_norm": 8.6875,
"learning_rate": 4.653539346653278e-05,
"loss": 0.2266,
"step": 2420
},
{
"epoch": 0.17632972933749366,
"grad_norm": 5.875,
"learning_rate": 4.650617918639546e-05,
"loss": 0.2485,
"step": 2430
},
{
"epoch": 0.17705536608373848,
"grad_norm": 2.953125,
"learning_rate": 4.647685150258314e-05,
"loss": 0.2105,
"step": 2440
},
{
"epoch": 0.1777810028299833,
"grad_norm": 5.34375,
"learning_rate": 4.644741056974287e-05,
"loss": 0.2731,
"step": 2450
},
{
"epoch": 0.17850663957622814,
"grad_norm": 8.1875,
"learning_rate": 4.641785654311884e-05,
"loss": 0.2572,
"step": 2460
},
{
"epoch": 0.17923227632247296,
"grad_norm": 7.71875,
"learning_rate": 4.6388189578551646e-05,
"loss": 0.2705,
"step": 2470
},
{
"epoch": 0.1799579130687178,
"grad_norm": 1.234375,
"learning_rate": 4.635840983247738e-05,
"loss": 0.2216,
"step": 2480
},
{
"epoch": 0.18068354981496262,
"grad_norm": 3.796875,
"learning_rate": 4.632851746192685e-05,
"loss": 0.2416,
"step": 2490
},
{
"epoch": 0.18140918656120747,
"grad_norm": 8.9375,
"learning_rate": 4.6298512624524734e-05,
"loss": 0.0987,
"step": 2500
},
{
"epoch": 0.1821348233074523,
"grad_norm": 7.09375,
"learning_rate": 4.626839547848877e-05,
"loss": 0.1668,
"step": 2510
},
{
"epoch": 0.18286046005369713,
"grad_norm": 1.5546875,
"learning_rate": 4.62381661826289e-05,
"loss": 0.1733,
"step": 2520
},
{
"epoch": 0.18358609679994195,
"grad_norm": 5.0625,
"learning_rate": 4.620782489634642e-05,
"loss": 0.2451,
"step": 2530
},
{
"epoch": 0.18431173354618677,
"grad_norm": 8.3125,
"learning_rate": 4.617737177963319e-05,
"loss": 0.1714,
"step": 2540
},
{
"epoch": 0.1850373702924316,
"grad_norm": 3.796875,
"learning_rate": 4.614680699307076e-05,
"loss": 0.173,
"step": 2550
},
{
"epoch": 0.18576300703867643,
"grad_norm": 7.9375,
"learning_rate": 4.6116130697829495e-05,
"loss": 0.2125,
"step": 2560
},
{
"epoch": 0.18648864378492128,
"grad_norm": 3.5625,
"learning_rate": 4.6085343055667786e-05,
"loss": 0.2366,
"step": 2570
},
{
"epoch": 0.1872142805311661,
"grad_norm": 1.1875,
"learning_rate": 4.6054444228931136e-05,
"loss": 0.2829,
"step": 2580
},
{
"epoch": 0.18793991727741094,
"grad_norm": 3.734375,
"learning_rate": 4.6023434380551354e-05,
"loss": 0.2183,
"step": 2590
},
{
"epoch": 0.18866555402365576,
"grad_norm": 6.46875,
"learning_rate": 4.5992313674045664e-05,
"loss": 0.2144,
"step": 2600
},
{
"epoch": 0.18939119076990057,
"grad_norm": 5.53125,
"learning_rate": 4.5961082273515856e-05,
"loss": 0.1141,
"step": 2610
},
{
"epoch": 0.19011682751614542,
"grad_norm": 4.125,
"learning_rate": 4.592974034364741e-05,
"loss": 0.3297,
"step": 2620
},
{
"epoch": 0.19084246426239024,
"grad_norm": 3.734375,
"learning_rate": 4.589828804970865e-05,
"loss": 0.2677,
"step": 2630
},
{
"epoch": 0.19156810100863508,
"grad_norm": 6.125,
"learning_rate": 4.5866725557549864e-05,
"loss": 0.1603,
"step": 2640
},
{
"epoch": 0.1922937377548799,
"grad_norm": 3.25,
"learning_rate": 4.583505303360238e-05,
"loss": 0.2041,
"step": 2650
},
{
"epoch": 0.19301937450112475,
"grad_norm": 4.71875,
"learning_rate": 4.580327064487778e-05,
"loss": 0.2155,
"step": 2660
},
{
"epoch": 0.19374501124736956,
"grad_norm": 3.328125,
"learning_rate": 4.577137855896695e-05,
"loss": 0.1948,
"step": 2670
},
{
"epoch": 0.1944706479936144,
"grad_norm": 10.375,
"learning_rate": 4.5739376944039225e-05,
"loss": 0.286,
"step": 2680
},
{
"epoch": 0.19519628473985923,
"grad_norm": 9.8125,
"learning_rate": 4.5707265968841475e-05,
"loss": 0.1766,
"step": 2690
},
{
"epoch": 0.19592192148610404,
"grad_norm": 8.625,
"learning_rate": 4.5675045802697266e-05,
"loss": 0.1973,
"step": 2700
},
{
"epoch": 0.1966475582323489,
"grad_norm": 7.5625,
"learning_rate": 4.564271661550591e-05,
"loss": 0.2644,
"step": 2710
},
{
"epoch": 0.1973731949785937,
"grad_norm": 6.09375,
"learning_rate": 4.56102785777416e-05,
"loss": 0.2265,
"step": 2720
},
{
"epoch": 0.19809883172483855,
"grad_norm": 3.890625,
"learning_rate": 4.5577731860452525e-05,
"loss": 0.2042,
"step": 2730
},
{
"epoch": 0.19882446847108337,
"grad_norm": 5.625,
"learning_rate": 4.5545076635259924e-05,
"loss": 0.1945,
"step": 2740
},
{
"epoch": 0.19955010521732822,
"grad_norm": 7.625,
"learning_rate": 4.5512313074357225e-05,
"loss": 0.2212,
"step": 2750
},
{
"epoch": 0.20027574196357303,
"grad_norm": 7.6875,
"learning_rate": 4.54794413505091e-05,
"loss": 0.1836,
"step": 2760
},
{
"epoch": 0.20100137870981785,
"grad_norm": 5.53125,
"learning_rate": 4.544646163705059e-05,
"loss": 0.2268,
"step": 2770
},
{
"epoch": 0.2017270154560627,
"grad_norm": 1.796875,
"learning_rate": 4.541337410788616e-05,
"loss": 0.1336,
"step": 2780
},
{
"epoch": 0.20245265220230751,
"grad_norm": 9.4375,
"learning_rate": 4.538017893748881e-05,
"loss": 0.161,
"step": 2790
},
{
"epoch": 0.20317828894855236,
"grad_norm": 4.59375,
"learning_rate": 4.534687630089912e-05,
"loss": 0.2052,
"step": 2800
},
{
"epoch": 0.20390392569479718,
"grad_norm": 9.25,
"learning_rate": 4.5313466373724366e-05,
"loss": 0.288,
"step": 2810
},
{
"epoch": 0.20462956244104202,
"grad_norm": 0.87109375,
"learning_rate": 4.5279949332137564e-05,
"loss": 0.1063,
"step": 2820
},
{
"epoch": 0.20535519918728684,
"grad_norm": 9.625,
"learning_rate": 4.524632535287655e-05,
"loss": 0.2642,
"step": 2830
},
{
"epoch": 0.2060808359335317,
"grad_norm": 7.21875,
"learning_rate": 4.521259461324308e-05,
"loss": 0.1667,
"step": 2840
},
{
"epoch": 0.2068064726797765,
"grad_norm": 5.78125,
"learning_rate": 4.51787572911018e-05,
"loss": 0.1666,
"step": 2850
},
{
"epoch": 0.20753210942602132,
"grad_norm": 3.140625,
"learning_rate": 4.514481356487945e-05,
"loss": 0.1352,
"step": 2860
},
{
"epoch": 0.20825774617226617,
"grad_norm": 4.53125,
"learning_rate": 4.51107636135638e-05,
"loss": 0.1691,
"step": 2870
},
{
"epoch": 0.20898338291851098,
"grad_norm": 3.140625,
"learning_rate": 4.507660761670277e-05,
"loss": 0.1735,
"step": 2880
},
{
"epoch": 0.20970901966475583,
"grad_norm": 6.8125,
"learning_rate": 4.504234575440345e-05,
"loss": 0.2212,
"step": 2890
},
{
"epoch": 0.21043465641100065,
"grad_norm": 8.4375,
"learning_rate": 4.500797820733117e-05,
"loss": 0.1927,
"step": 2900
},
{
"epoch": 0.2111602931572455,
"grad_norm": 5.0,
"learning_rate": 4.4973505156708576e-05,
"loss": 0.2425,
"step": 2910
},
{
"epoch": 0.2118859299034903,
"grad_norm": 4.375,
"learning_rate": 4.4938926784314585e-05,
"loss": 0.2642,
"step": 2920
},
{
"epoch": 0.21261156664973516,
"grad_norm": 6.4375,
"learning_rate": 4.490424327248354e-05,
"loss": 0.1719,
"step": 2930
},
{
"epoch": 0.21333720339597997,
"grad_norm": 3.390625,
"learning_rate": 4.486945480410415e-05,
"loss": 0.1807,
"step": 2940
},
{
"epoch": 0.2140628401422248,
"grad_norm": 11.0,
"learning_rate": 4.4834561562618595e-05,
"loss": 0.2051,
"step": 2950
},
{
"epoch": 0.21478847688846964,
"grad_norm": 5.78125,
"learning_rate": 4.479956373202151e-05,
"loss": 0.1103,
"step": 2960
},
{
"epoch": 0.21551411363471445,
"grad_norm": 3.8125,
"learning_rate": 4.476446149685906e-05,
"loss": 0.2068,
"step": 2970
},
{
"epoch": 0.2162397503809593,
"grad_norm": 3.140625,
"learning_rate": 4.472925504222792e-05,
"loss": 0.3126,
"step": 2980
},
{
"epoch": 0.21696538712720412,
"grad_norm": 24.875,
"learning_rate": 4.4693944553774335e-05,
"loss": 0.2166,
"step": 2990
},
{
"epoch": 0.21769102387344896,
"grad_norm": 7.15625,
"learning_rate": 4.465853021769313e-05,
"loss": 0.277,
"step": 3000
},
{
"epoch": 0.21841666061969378,
"grad_norm": 5.9375,
"learning_rate": 4.4623012220726705e-05,
"loss": 0.2262,
"step": 3010
},
{
"epoch": 0.2191422973659386,
"grad_norm": 6.90625,
"learning_rate": 4.45873907501641e-05,
"loss": 0.1627,
"step": 3020
},
{
"epoch": 0.21986793411218344,
"grad_norm": 8.3125,
"learning_rate": 4.455166599383997e-05,
"loss": 0.1624,
"step": 3030
},
{
"epoch": 0.22059357085842826,
"grad_norm": 3.875,
"learning_rate": 4.4515838140133594e-05,
"loss": 0.2401,
"step": 3040
},
{
"epoch": 0.2213192076046731,
"grad_norm": 2.953125,
"learning_rate": 4.447990737796789e-05,
"loss": 0.1749,
"step": 3050
},
{
"epoch": 0.22204484435091792,
"grad_norm": 2.46875,
"learning_rate": 4.4443873896808434e-05,
"loss": 0.2449,
"step": 3060
},
{
"epoch": 0.22277048109716277,
"grad_norm": 9.125,
"learning_rate": 4.440773788666244e-05,
"loss": 0.2794,
"step": 3070
},
{
"epoch": 0.2234961178434076,
"grad_norm": 8.5625,
"learning_rate": 4.437149953807777e-05,
"loss": 0.2724,
"step": 3080
},
{
"epoch": 0.22422175458965243,
"grad_norm": 4.1875,
"learning_rate": 4.433515904214191e-05,
"loss": 0.1973,
"step": 3090
},
{
"epoch": 0.22494739133589725,
"grad_norm": 3.34375,
"learning_rate": 4.429871659048099e-05,
"loss": 0.3198,
"step": 3100
},
{
"epoch": 0.22567302808214207,
"grad_norm": 7.96875,
"learning_rate": 4.426217237525875e-05,
"loss": 0.275,
"step": 3110
},
{
"epoch": 0.22639866482838691,
"grad_norm": 7.09375,
"learning_rate": 4.422552658917556e-05,
"loss": 0.1821,
"step": 3120
},
{
"epoch": 0.22712430157463173,
"grad_norm": 4.9375,
"learning_rate": 4.418877942546733e-05,
"loss": 0.2716,
"step": 3130
},
{
"epoch": 0.22784993832087658,
"grad_norm": 1.796875,
"learning_rate": 4.4151931077904606e-05,
"loss": 0.1725,
"step": 3140
},
{
"epoch": 0.2285755750671214,
"grad_norm": 2.375,
"learning_rate": 4.4114981740791436e-05,
"loss": 0.1706,
"step": 3150
},
{
"epoch": 0.22930121181336624,
"grad_norm": 5.09375,
"learning_rate": 4.407793160896442e-05,
"loss": 0.1741,
"step": 3160
},
{
"epoch": 0.23002684855961106,
"grad_norm": 3.078125,
"learning_rate": 4.404078087779164e-05,
"loss": 0.1467,
"step": 3170
},
{
"epoch": 0.23075248530585588,
"grad_norm": 2.921875,
"learning_rate": 4.400352974317166e-05,
"loss": 0.2124,
"step": 3180
},
{
"epoch": 0.23147812205210072,
"grad_norm": 5.15625,
"learning_rate": 4.396617840153245e-05,
"loss": 0.197,
"step": 3190
},
{
"epoch": 0.23220375879834554,
"grad_norm": 2.09375,
"learning_rate": 4.392872704983042e-05,
"loss": 0.2052,
"step": 3200
},
{
"epoch": 0.23292939554459038,
"grad_norm": 6.5,
"learning_rate": 4.389117588554932e-05,
"loss": 0.2435,
"step": 3210
},
{
"epoch": 0.2336550322908352,
"grad_norm": 3.578125,
"learning_rate": 4.38535251066992e-05,
"loss": 0.1877,
"step": 3220
},
{
"epoch": 0.23438066903708005,
"grad_norm": 7.625,
"learning_rate": 4.381577491181542e-05,
"loss": 0.3165,
"step": 3230
},
{
"epoch": 0.23510630578332486,
"grad_norm": 2.171875,
"learning_rate": 4.377792549995755e-05,
"loss": 0.1709,
"step": 3240
},
{
"epoch": 0.2358319425295697,
"grad_norm": 3.8125,
"learning_rate": 4.373997707070835e-05,
"loss": 0.161,
"step": 3250
},
{
"epoch": 0.23655757927581453,
"grad_norm": 4.8125,
"learning_rate": 4.370192982417268e-05,
"loss": 0.233,
"step": 3260
},
{
"epoch": 0.23728321602205935,
"grad_norm": 7.46875,
"learning_rate": 4.366378396097649e-05,
"loss": 0.1879,
"step": 3270
},
{
"epoch": 0.2380088527683042,
"grad_norm": 2.25,
"learning_rate": 4.362553968226575e-05,
"loss": 0.2066,
"step": 3280
},
{
"epoch": 0.238734489514549,
"grad_norm": 7.53125,
"learning_rate": 4.3587197189705375e-05,
"loss": 0.1907,
"step": 3290
},
{
"epoch": 0.23946012626079385,
"grad_norm": 5.09375,
"learning_rate": 4.354875668547815e-05,
"loss": 0.138,
"step": 3300
},
{
"epoch": 0.24018576300703867,
"grad_norm": 8.375,
"learning_rate": 4.351021837228372e-05,
"loss": 0.2027,
"step": 3310
},
{
"epoch": 0.24091139975328352,
"grad_norm": 9.375,
"learning_rate": 4.347158245333744e-05,
"loss": 0.1607,
"step": 3320
},
{
"epoch": 0.24163703649952833,
"grad_norm": 3.890625,
"learning_rate": 4.343284913236939e-05,
"loss": 0.1537,
"step": 3330
},
{
"epoch": 0.24236267324577315,
"grad_norm": 10.0625,
"learning_rate": 4.3394018613623235e-05,
"loss": 0.2282,
"step": 3340
},
{
"epoch": 0.243088309992018,
"grad_norm": 6.53125,
"learning_rate": 4.335509110185517e-05,
"loss": 0.1904,
"step": 3350
},
{
"epoch": 0.24381394673826282,
"grad_norm": 2.078125,
"learning_rate": 4.3316066802332854e-05,
"loss": 0.1727,
"step": 3360
},
{
"epoch": 0.24453958348450766,
"grad_norm": 7.59375,
"learning_rate": 4.327694592083431e-05,
"loss": 0.2824,
"step": 3370
},
{
"epoch": 0.24526522023075248,
"grad_norm": 3.046875,
"learning_rate": 4.323772866364683e-05,
"loss": 0.295,
"step": 3380
},
{
"epoch": 0.24599085697699732,
"grad_norm": 3.625,
"learning_rate": 4.319841523756591e-05,
"loss": 0.232,
"step": 3390
},
{
"epoch": 0.24671649372324214,
"grad_norm": 5.09375,
"learning_rate": 4.3159005849894175e-05,
"loss": 0.2444,
"step": 3400
},
{
"epoch": 0.247442130469487,
"grad_norm": 4.4375,
"learning_rate": 4.3119500708440244e-05,
"loss": 0.1459,
"step": 3410
},
{
"epoch": 0.2481677672157318,
"grad_norm": 4.46875,
"learning_rate": 4.3079900021517626e-05,
"loss": 0.2661,
"step": 3420
},
{
"epoch": 0.24889340396197662,
"grad_norm": 1.0859375,
"learning_rate": 4.304020399794371e-05,
"loss": 0.3032,
"step": 3430
},
{
"epoch": 0.24961904070822147,
"grad_norm": 7.40625,
"learning_rate": 4.300041284703853e-05,
"loss": 0.2178,
"step": 3440
},
{
"epoch": 0.2503446774544663,
"grad_norm": 5.90625,
"learning_rate": 4.29605267786238e-05,
"loss": 0.2001,
"step": 3450
},
{
"epoch": 0.25107031420071113,
"grad_norm": 8.0625,
"learning_rate": 4.292054600302171e-05,
"loss": 0.1923,
"step": 3460
},
{
"epoch": 0.25179595094695595,
"grad_norm": 4.125,
"learning_rate": 4.288047073105384e-05,
"loss": 0.1961,
"step": 3470
},
{
"epoch": 0.25252158769320077,
"grad_norm": 8.25,
"learning_rate": 4.2840301174040056e-05,
"loss": 0.2374,
"step": 3480
},
{
"epoch": 0.25324722443944564,
"grad_norm": 6.125,
"learning_rate": 4.280003754379742e-05,
"loss": 0.2464,
"step": 3490
},
{
"epoch": 0.25397286118569046,
"grad_norm": 9.25,
"learning_rate": 4.275968005263904e-05,
"loss": 0.2261,
"step": 3500
},
{
"epoch": 0.2546984979319353,
"grad_norm": 1.8125,
"learning_rate": 4.2719228913372946e-05,
"loss": 0.1982,
"step": 3510
},
{
"epoch": 0.2554241346781801,
"grad_norm": 1.9921875,
"learning_rate": 4.2678684339300987e-05,
"loss": 0.1902,
"step": 3520
},
{
"epoch": 0.2561497714244249,
"grad_norm": 4.46875,
"learning_rate": 4.263804654421771e-05,
"loss": 0.239,
"step": 3530
},
{
"epoch": 0.2568754081706698,
"grad_norm": 8.375,
"learning_rate": 4.2597315742409214e-05,
"loss": 0.2097,
"step": 3540
},
{
"epoch": 0.2576010449169146,
"grad_norm": 6.3125,
"learning_rate": 4.2556492148652036e-05,
"loss": 0.234,
"step": 3550
},
{
"epoch": 0.2583266816631594,
"grad_norm": 4.875,
"learning_rate": 4.2515575978212006e-05,
"loss": 0.3032,
"step": 3560
},
{
"epoch": 0.25905231840940424,
"grad_norm": 3.5625,
"learning_rate": 4.2474567446843115e-05,
"loss": 0.285,
"step": 3570
},
{
"epoch": 0.2597779551556491,
"grad_norm": 3.625,
"learning_rate": 4.243346677078639e-05,
"loss": 0.1276,
"step": 3580
},
{
"epoch": 0.2605035919018939,
"grad_norm": 5.40625,
"learning_rate": 4.239227416676874e-05,
"loss": 0.1852,
"step": 3590
},
{
"epoch": 0.26122922864813874,
"grad_norm": 6.6875,
"learning_rate": 4.2350989852001824e-05,
"loss": 0.2314,
"step": 3600
},
{
"epoch": 0.26195486539438356,
"grad_norm": 3.8125,
"learning_rate": 4.2309614044180856e-05,
"loss": 0.1953,
"step": 3610
},
{
"epoch": 0.2626805021406284,
"grad_norm": 6.90625,
"learning_rate": 4.2268146961483573e-05,
"loss": 0.2408,
"step": 3620
},
{
"epoch": 0.26340613888687325,
"grad_norm": 4.40625,
"learning_rate": 4.2226588822568956e-05,
"loss": 0.184,
"step": 3630
},
{
"epoch": 0.26413177563311807,
"grad_norm": 5.1875,
"learning_rate": 4.218493984657614e-05,
"loss": 0.1992,
"step": 3640
},
{
"epoch": 0.2648574123793629,
"grad_norm": 5.71875,
"learning_rate": 4.214320025312328e-05,
"loss": 0.1468,
"step": 3650
},
{
"epoch": 0.2655830491256077,
"grad_norm": 5.65625,
"learning_rate": 4.2101370262306325e-05,
"loss": 0.233,
"step": 3660
},
{
"epoch": 0.2663086858718525,
"grad_norm": 6.84375,
"learning_rate": 4.2059450094697913e-05,
"loss": 0.1435,
"step": 3670
},
{
"epoch": 0.2670343226180974,
"grad_norm": 5.90625,
"learning_rate": 4.2017439971346195e-05,
"loss": 0.2085,
"step": 3680
},
{
"epoch": 0.2677599593643422,
"grad_norm": 3.125,
"learning_rate": 4.197534011377367e-05,
"loss": 0.1725,
"step": 3690
},
{
"epoch": 0.26848559611058703,
"grad_norm": 6.0,
"learning_rate": 4.1933150743976e-05,
"loss": 0.1852,
"step": 3700
},
{
"epoch": 0.26921123285683185,
"grad_norm": 3.578125,
"learning_rate": 4.189087208442084e-05,
"loss": 0.2231,
"step": 3710
},
{
"epoch": 0.2699368696030767,
"grad_norm": 4.15625,
"learning_rate": 4.184850435804672e-05,
"loss": 0.2219,
"step": 3720
},
{
"epoch": 0.27066250634932154,
"grad_norm": 6.375,
"learning_rate": 4.180604778826176e-05,
"loss": 0.2189,
"step": 3730
},
{
"epoch": 0.27138814309556636,
"grad_norm": 6.25,
"learning_rate": 4.176350259894263e-05,
"loss": 0.1797,
"step": 3740
},
{
"epoch": 0.2721137798418112,
"grad_norm": 6.65625,
"learning_rate": 4.172086901443325e-05,
"loss": 0.2735,
"step": 3750
},
{
"epoch": 0.272839416588056,
"grad_norm": 5.46875,
"learning_rate": 4.167814725954367e-05,
"loss": 0.2413,
"step": 3760
},
{
"epoch": 0.27356505333430087,
"grad_norm": 8.625,
"learning_rate": 4.163533755954885e-05,
"loss": 0.2151,
"step": 3770
},
{
"epoch": 0.2742906900805457,
"grad_norm": 1.0703125,
"learning_rate": 4.159244014018754e-05,
"loss": 0.1747,
"step": 3780
},
{
"epoch": 0.2750163268267905,
"grad_norm": 5.84375,
"learning_rate": 4.154945522766098e-05,
"loss": 0.2007,
"step": 3790
},
{
"epoch": 0.2757419635730353,
"grad_norm": 6.0,
"learning_rate": 4.15063830486318e-05,
"loss": 0.1511,
"step": 3800
},
{
"epoch": 0.2764676003192802,
"grad_norm": 5.5625,
"learning_rate": 4.14632238302228e-05,
"loss": 0.1725,
"step": 3810
},
{
"epoch": 0.277193237065525,
"grad_norm": 6.375,
"learning_rate": 4.141997780001573e-05,
"loss": 0.1666,
"step": 3820
},
{
"epoch": 0.27791887381176983,
"grad_norm": 5.96875,
"learning_rate": 4.1376645186050105e-05,
"loss": 0.2431,
"step": 3830
},
{
"epoch": 0.27864451055801465,
"grad_norm": 6.15625,
"learning_rate": 4.133322621682201e-05,
"loss": 0.2868,
"step": 3840
},
{
"epoch": 0.27937014730425946,
"grad_norm": 1.625,
"learning_rate": 4.128972112128288e-05,
"loss": 0.2607,
"step": 3850
},
{
"epoch": 0.28009578405050434,
"grad_norm": 6.71875,
"learning_rate": 4.124613012883831e-05,
"loss": 0.1274,
"step": 3860
},
{
"epoch": 0.28082142079674915,
"grad_norm": 5.15625,
"learning_rate": 4.1202453469346826e-05,
"loss": 0.1535,
"step": 3870
},
{
"epoch": 0.281547057542994,
"grad_norm": 3.625,
"learning_rate": 4.115869137311869e-05,
"loss": 0.1437,
"step": 3880
},
{
"epoch": 0.2822726942892388,
"grad_norm": 4.125,
"learning_rate": 4.111484407091467e-05,
"loss": 0.2132,
"step": 3890
},
{
"epoch": 0.28299833103548366,
"grad_norm": 5.84375,
"learning_rate": 4.107091179394484e-05,
"loss": 0.2423,
"step": 3900
},
{
"epoch": 0.2837239677817285,
"grad_norm": 2.28125,
"learning_rate": 4.1026894773867355e-05,
"loss": 0.2818,
"step": 3910
},
{
"epoch": 0.2844496045279733,
"grad_norm": 1.640625,
"learning_rate": 4.098279324278721e-05,
"loss": 0.2143,
"step": 3920
},
{
"epoch": 0.2851752412742181,
"grad_norm": 7.375,
"learning_rate": 4.0938607433255056e-05,
"loss": 0.2649,
"step": 3930
},
{
"epoch": 0.28590087802046293,
"grad_norm": 5.34375,
"learning_rate": 4.0894337578265924e-05,
"loss": 0.1729,
"step": 3940
},
{
"epoch": 0.2866265147667078,
"grad_norm": 1.2890625,
"learning_rate": 4.084998391125804e-05,
"loss": 0.1692,
"step": 3950
},
{
"epoch": 0.2873521515129526,
"grad_norm": 3.28125,
"learning_rate": 4.0805546666111575e-05,
"loss": 0.2736,
"step": 3960
},
{
"epoch": 0.28807778825919744,
"grad_norm": 11.0625,
"learning_rate": 4.0761026077147425e-05,
"loss": 0.2362,
"step": 3970
},
{
"epoch": 0.28880342500544226,
"grad_norm": 4.21875,
"learning_rate": 4.071642237912593e-05,
"loss": 0.1906,
"step": 3980
},
{
"epoch": 0.2895290617516871,
"grad_norm": 5.03125,
"learning_rate": 4.0671735807245704e-05,
"loss": 0.1604,
"step": 3990
},
{
"epoch": 0.29025469849793195,
"grad_norm": 6.53125,
"learning_rate": 4.0626966597142354e-05,
"loss": 0.172,
"step": 4000
},
{
"epoch": 0.29098033524417677,
"grad_norm": 8.0,
"learning_rate": 4.0582114984887225e-05,
"loss": 0.2528,
"step": 4010
},
{
"epoch": 0.2917059719904216,
"grad_norm": 2.65625,
"learning_rate": 4.05371812069862e-05,
"loss": 0.199,
"step": 4020
},
{
"epoch": 0.2924316087366664,
"grad_norm": 5.25,
"learning_rate": 4.049216550037841e-05,
"loss": 0.1516,
"step": 4030
},
{
"epoch": 0.2931572454829113,
"grad_norm": 2.40625,
"learning_rate": 4.0447068102435015e-05,
"loss": 0.1504,
"step": 4040
},
{
"epoch": 0.2938828822291561,
"grad_norm": 6.9375,
"learning_rate": 4.0401889250957926e-05,
"loss": 0.2432,
"step": 4050
},
{
"epoch": 0.2946085189754009,
"grad_norm": 8.8125,
"learning_rate": 4.035662918417857e-05,
"loss": 0.244,
"step": 4060
},
{
"epoch": 0.29533415572164573,
"grad_norm": 2.25,
"learning_rate": 4.031128814075665e-05,
"loss": 0.1402,
"step": 4070
},
{
"epoch": 0.29605979246789055,
"grad_norm": 8.625,
"learning_rate": 4.026586635977882e-05,
"loss": 0.2012,
"step": 4080
},
{
"epoch": 0.2967854292141354,
"grad_norm": 10.75,
"learning_rate": 4.02203640807575e-05,
"loss": 0.1722,
"step": 4090
},
{
"epoch": 0.29751106596038024,
"grad_norm": 6.625,
"learning_rate": 4.017478154362959e-05,
"loss": 0.1729,
"step": 4100
},
{
"epoch": 0.29823670270662506,
"grad_norm": 2.609375,
"learning_rate": 4.012911898875516e-05,
"loss": 0.1613,
"step": 4110
},
{
"epoch": 0.2989623394528699,
"grad_norm": 6.625,
"learning_rate": 4.008337665691626e-05,
"loss": 0.1887,
"step": 4120
},
{
"epoch": 0.29968797619911475,
"grad_norm": 5.90625,
"learning_rate": 4.0037554789315585e-05,
"loss": 0.1853,
"step": 4130
},
{
"epoch": 0.30041361294535956,
"grad_norm": 2.234375,
"learning_rate": 3.999165362757524e-05,
"loss": 0.2209,
"step": 4140
},
{
"epoch": 0.3011392496916044,
"grad_norm": 3.859375,
"learning_rate": 3.994567341373544e-05,
"loss": 0.256,
"step": 4150
},
{
"epoch": 0.3018648864378492,
"grad_norm": 1.5625,
"learning_rate": 3.9899614390253264e-05,
"loss": 0.1986,
"step": 4160
},
{
"epoch": 0.302590523184094,
"grad_norm": 3.25,
"learning_rate": 3.9853476800001356e-05,
"loss": 0.1695,
"step": 4170
},
{
"epoch": 0.3033161599303389,
"grad_norm": 3.125,
"learning_rate": 3.980726088626664e-05,
"loss": 0.1863,
"step": 4180
},
{
"epoch": 0.3040417966765837,
"grad_norm": 6.5,
"learning_rate": 3.976096689274904e-05,
"loss": 0.1599,
"step": 4190
},
{
"epoch": 0.3047674334228285,
"grad_norm": 4.65625,
"learning_rate": 3.9714595063560235e-05,
"loss": 0.1489,
"step": 4200
},
{
"epoch": 0.30549307016907334,
"grad_norm": 0.74609375,
"learning_rate": 3.9668145643222296e-05,
"loss": 0.1681,
"step": 4210
},
{
"epoch": 0.3062187069153182,
"grad_norm": 6.3125,
"learning_rate": 3.962161887666645e-05,
"loss": 0.1861,
"step": 4220
},
{
"epoch": 0.30694434366156303,
"grad_norm": 9.9375,
"learning_rate": 3.9575015009231806e-05,
"loss": 0.3147,
"step": 4230
},
{
"epoch": 0.30766998040780785,
"grad_norm": 2.03125,
"learning_rate": 3.952833428666398e-05,
"loss": 0.2436,
"step": 4240
},
{
"epoch": 0.30839561715405267,
"grad_norm": 6.25,
"learning_rate": 3.9481576955113897e-05,
"loss": 0.2173,
"step": 4250
},
{
"epoch": 0.3091212539002975,
"grad_norm": 4.78125,
"learning_rate": 3.943474326113642e-05,
"loss": 0.2422,
"step": 4260
},
{
"epoch": 0.30984689064654236,
"grad_norm": 6.96875,
"learning_rate": 3.938783345168909e-05,
"loss": 0.2201,
"step": 4270
},
{
"epoch": 0.3105725273927872,
"grad_norm": 3.703125,
"learning_rate": 3.9340847774130796e-05,
"loss": 0.1594,
"step": 4280
},
{
"epoch": 0.311298164139032,
"grad_norm": 3.390625,
"learning_rate": 3.92937864762205e-05,
"loss": 0.1946,
"step": 4290
},
{
"epoch": 0.3120238008852768,
"grad_norm": 2.515625,
"learning_rate": 3.92466498061159e-05,
"loss": 0.1694,
"step": 4300
},
{
"epoch": 0.3127494376315217,
"grad_norm": 2.140625,
"learning_rate": 3.919943801237217e-05,
"loss": 0.2164,
"step": 4310
},
{
"epoch": 0.3134750743777665,
"grad_norm": 11.0625,
"learning_rate": 3.915215134394057e-05,
"loss": 0.2352,
"step": 4320
},
{
"epoch": 0.3142007111240113,
"grad_norm": 6.34375,
"learning_rate": 3.910479005016722e-05,
"loss": 0.2234,
"step": 4330
},
{
"epoch": 0.31492634787025614,
"grad_norm": 12.125,
"learning_rate": 3.905735438079171e-05,
"loss": 0.1806,
"step": 4340
},
{
"epoch": 0.31565198461650096,
"grad_norm": 5.8125,
"learning_rate": 3.900984458594586e-05,
"loss": 0.197,
"step": 4350
},
{
"epoch": 0.31637762136274583,
"grad_norm": 1.6484375,
"learning_rate": 3.8962260916152315e-05,
"loss": 0.143,
"step": 4360
},
{
"epoch": 0.31710325810899065,
"grad_norm": 6.59375,
"learning_rate": 3.8914603622323295e-05,
"loss": 0.2315,
"step": 4370
},
{
"epoch": 0.31782889485523547,
"grad_norm": 0.6640625,
"learning_rate": 3.886687295575925e-05,
"loss": 0.1557,
"step": 4380
},
{
"epoch": 0.3185545316014803,
"grad_norm": 4.8125,
"learning_rate": 3.88190691681475e-05,
"loss": 0.1443,
"step": 4390
},
{
"epoch": 0.3192801683477251,
"grad_norm": 4.65625,
"learning_rate": 3.877119251156097e-05,
"loss": 0.1125,
"step": 4400
},
{
"epoch": 0.32000580509397,
"grad_norm": 9.375,
"learning_rate": 3.872324323845681e-05,
"loss": 0.2663,
"step": 4410
},
{
"epoch": 0.3207314418402148,
"grad_norm": 1.84375,
"learning_rate": 3.8675221601675085e-05,
"loss": 0.2016,
"step": 4420
},
{
"epoch": 0.3214570785864596,
"grad_norm": 2.28125,
"learning_rate": 3.862712785443744e-05,
"loss": 0.2524,
"step": 4430
},
{
"epoch": 0.32218271533270443,
"grad_norm": 7.875,
"learning_rate": 3.8578962250345765e-05,
"loss": 0.2159,
"step": 4440
},
{
"epoch": 0.3229083520789493,
"grad_norm": 7.125,
"learning_rate": 3.853072504338087e-05,
"loss": 0.1617,
"step": 4450
},
{
"epoch": 0.3236339888251941,
"grad_norm": 7.5,
"learning_rate": 3.848241648790109e-05,
"loss": 0.2983,
"step": 4460
},
{
"epoch": 0.32435962557143894,
"grad_norm": 2.125,
"learning_rate": 3.843403683864104e-05,
"loss": 0.1618,
"step": 4470
},
{
"epoch": 0.32508526231768375,
"grad_norm": 2.46875,
"learning_rate": 3.838558635071018e-05,
"loss": 0.1574,
"step": 4480
},
{
"epoch": 0.32581089906392857,
"grad_norm": 7.5625,
"learning_rate": 3.8337065279591536e-05,
"loss": 0.1802,
"step": 4490
},
{
"epoch": 0.32653653581017344,
"grad_norm": 3.34375,
"learning_rate": 3.828847388114027e-05,
"loss": 0.1615,
"step": 4500
}
],
"logging_steps": 10,
"max_steps": 13781,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"total_flos": 0.0,
"train_batch_size": 5,
"trial_name": null,
"trial_params": null
}