|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 0.10884551193672448, |
|
"eval_steps": 500, |
|
"global_step": 1500, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0007256367462448298, |
|
"grad_norm": 15.6875, |
|
"learning_rate": 5e-06, |
|
"loss": 3.2562, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.0014512734924896596, |
|
"grad_norm": 17.5, |
|
"learning_rate": 1e-05, |
|
"loss": 2.9076, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.0021769102387344894, |
|
"grad_norm": 19.125, |
|
"learning_rate": 1.5e-05, |
|
"loss": 3.0281, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.0029025469849793192, |
|
"grad_norm": 17.875, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6225, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.003628183731224149, |
|
"grad_norm": 13.9375, |
|
"learning_rate": 2.5e-05, |
|
"loss": 2.5137, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.004353820477468979, |
|
"grad_norm": 10.6875, |
|
"learning_rate": 3e-05, |
|
"loss": 1.994, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.005079457223713809, |
|
"grad_norm": 7.53125, |
|
"learning_rate": 3.5e-05, |
|
"loss": 1.5519, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.0058050939699586385, |
|
"grad_norm": 7.09375, |
|
"learning_rate": 4e-05, |
|
"loss": 1.2358, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.006530730716203468, |
|
"grad_norm": 8.0625, |
|
"learning_rate": 4.5e-05, |
|
"loss": 0.953, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.007256367462448298, |
|
"grad_norm": 6.65625, |
|
"learning_rate": 5e-05, |
|
"loss": 0.625, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.007982004208693128, |
|
"grad_norm": 5.0625, |
|
"learning_rate": 4.9999934086574596e-05, |
|
"loss": 0.6595, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.008707640954937958, |
|
"grad_norm": 5.59375, |
|
"learning_rate": 4.9999736346645943e-05, |
|
"loss": 0.6584, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.009433277701182788, |
|
"grad_norm": 3.953125, |
|
"learning_rate": 4.999940678125673e-05, |
|
"loss": 0.5113, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.010158914447427617, |
|
"grad_norm": 5.4375, |
|
"learning_rate": 4.9998945392144796e-05, |
|
"loss": 0.5528, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.010884551193672447, |
|
"grad_norm": 3.984375, |
|
"learning_rate": 4.999835218174307e-05, |
|
"loss": 0.5527, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.011610187939917277, |
|
"grad_norm": 3.234375, |
|
"learning_rate": 4.99976271531796e-05, |
|
"loss": 0.5342, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.012335824686162107, |
|
"grad_norm": 4.53125, |
|
"learning_rate": 4.9996770310277506e-05, |
|
"loss": 0.4445, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.013061461432406937, |
|
"grad_norm": 4.0625, |
|
"learning_rate": 4.9995781657555e-05, |
|
"loss": 0.4823, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.013787098178651766, |
|
"grad_norm": 4.78125, |
|
"learning_rate": 4.99946612002253e-05, |
|
"loss": 0.5318, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.014512734924896596, |
|
"grad_norm": 2.40625, |
|
"learning_rate": 4.9993408944196676e-05, |
|
"loss": 0.3257, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 0.015238371671141426, |
|
"grad_norm": 4.3125, |
|
"learning_rate": 4.9992024896072364e-05, |
|
"loss": 0.3893, |
|
"step": 210 |
|
}, |
|
{ |
|
"epoch": 0.015964008417386256, |
|
"grad_norm": 5.75, |
|
"learning_rate": 4.999050906315055e-05, |
|
"loss": 0.3641, |
|
"step": 220 |
|
}, |
|
{ |
|
"epoch": 0.016689645163631087, |
|
"grad_norm": 1.8359375, |
|
"learning_rate": 4.998886145342434e-05, |
|
"loss": 0.4811, |
|
"step": 230 |
|
}, |
|
{ |
|
"epoch": 0.017415281909875915, |
|
"grad_norm": 3.671875, |
|
"learning_rate": 4.9987082075581684e-05, |
|
"loss": 0.4718, |
|
"step": 240 |
|
}, |
|
{ |
|
"epoch": 0.018140918656120747, |
|
"grad_norm": 3.046875, |
|
"learning_rate": 4.9985170939005386e-05, |
|
"loss": 0.3431, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 0.018866555402365575, |
|
"grad_norm": 4.6875, |
|
"learning_rate": 4.998312805377302e-05, |
|
"loss": 0.3574, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 0.019592192148610407, |
|
"grad_norm": 4.5, |
|
"learning_rate": 4.998095343065685e-05, |
|
"loss": 0.3829, |
|
"step": 270 |
|
}, |
|
{ |
|
"epoch": 0.020317828894855235, |
|
"grad_norm": 3.96875, |
|
"learning_rate": 4.997864708112384e-05, |
|
"loss": 0.2996, |
|
"step": 280 |
|
}, |
|
{ |
|
"epoch": 0.021043465641100066, |
|
"grad_norm": 6.125, |
|
"learning_rate": 4.997620901733554e-05, |
|
"loss": 0.3374, |
|
"step": 290 |
|
}, |
|
{ |
|
"epoch": 0.021769102387344894, |
|
"grad_norm": 2.671875, |
|
"learning_rate": 4.997363925214803e-05, |
|
"loss": 0.3309, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 0.022494739133589726, |
|
"grad_norm": 3.125, |
|
"learning_rate": 4.9970937799111896e-05, |
|
"loss": 0.2993, |
|
"step": 310 |
|
}, |
|
{ |
|
"epoch": 0.023220375879834554, |
|
"grad_norm": 6.25, |
|
"learning_rate": 4.996810467247207e-05, |
|
"loss": 0.3885, |
|
"step": 320 |
|
}, |
|
{ |
|
"epoch": 0.023946012626079385, |
|
"grad_norm": 2.03125, |
|
"learning_rate": 4.9965139887167856e-05, |
|
"loss": 0.273, |
|
"step": 330 |
|
}, |
|
{ |
|
"epoch": 0.024671649372324213, |
|
"grad_norm": 3.0625, |
|
"learning_rate": 4.996204345883278e-05, |
|
"loss": 0.3373, |
|
"step": 340 |
|
}, |
|
{ |
|
"epoch": 0.025397286118569045, |
|
"grad_norm": 2.484375, |
|
"learning_rate": 4.9958815403794546e-05, |
|
"loss": 0.3743, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 0.026122922864813873, |
|
"grad_norm": 5.09375, |
|
"learning_rate": 4.995545573907492e-05, |
|
"loss": 0.3469, |
|
"step": 360 |
|
}, |
|
{ |
|
"epoch": 0.026848559611058705, |
|
"grad_norm": 4.90625, |
|
"learning_rate": 4.995196448238966e-05, |
|
"loss": 0.3278, |
|
"step": 370 |
|
}, |
|
{ |
|
"epoch": 0.027574196357303533, |
|
"grad_norm": 4.9375, |
|
"learning_rate": 4.9948341652148436e-05, |
|
"loss": 0.2599, |
|
"step": 380 |
|
}, |
|
{ |
|
"epoch": 0.028299833103548364, |
|
"grad_norm": 5.40625, |
|
"learning_rate": 4.994458726745468e-05, |
|
"loss": 0.3683, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 0.029025469849793192, |
|
"grad_norm": 0.671875, |
|
"learning_rate": 4.9940701348105554e-05, |
|
"loss": 0.3141, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 0.029751106596038024, |
|
"grad_norm": 3.53125, |
|
"learning_rate": 4.99366839145918e-05, |
|
"loss": 0.2115, |
|
"step": 410 |
|
}, |
|
{ |
|
"epoch": 0.030476743342282852, |
|
"grad_norm": 6.09375, |
|
"learning_rate": 4.993253498809762e-05, |
|
"loss": 0.3558, |
|
"step": 420 |
|
}, |
|
{ |
|
"epoch": 0.031202380088527683, |
|
"grad_norm": 5.15625, |
|
"learning_rate": 4.9928254590500646e-05, |
|
"loss": 0.3509, |
|
"step": 430 |
|
}, |
|
{ |
|
"epoch": 0.03192801683477251, |
|
"grad_norm": 5.375, |
|
"learning_rate": 4.9923842744371707e-05, |
|
"loss": 0.2839, |
|
"step": 440 |
|
}, |
|
{ |
|
"epoch": 0.03265365358101734, |
|
"grad_norm": 2.453125, |
|
"learning_rate": 4.99192994729748e-05, |
|
"loss": 0.2205, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 0.033379290327262175, |
|
"grad_norm": 4.09375, |
|
"learning_rate": 4.991462480026693e-05, |
|
"loss": 0.3745, |
|
"step": 460 |
|
}, |
|
{ |
|
"epoch": 0.034104927073507, |
|
"grad_norm": 1.171875, |
|
"learning_rate": 4.9909818750898e-05, |
|
"loss": 0.2963, |
|
"step": 470 |
|
}, |
|
{ |
|
"epoch": 0.03483056381975183, |
|
"grad_norm": 5.90625, |
|
"learning_rate": 4.990488135021065e-05, |
|
"loss": 0.2691, |
|
"step": 480 |
|
}, |
|
{ |
|
"epoch": 0.03555620056599666, |
|
"grad_norm": 1.0390625, |
|
"learning_rate": 4.989981262424017e-05, |
|
"loss": 0.2942, |
|
"step": 490 |
|
}, |
|
{ |
|
"epoch": 0.036281837312241494, |
|
"grad_norm": 5.3125, |
|
"learning_rate": 4.989461259971432e-05, |
|
"loss": 0.2185, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.03700747405848632, |
|
"grad_norm": 2.78125, |
|
"learning_rate": 4.988928130405323e-05, |
|
"loss": 0.2557, |
|
"step": 510 |
|
}, |
|
{ |
|
"epoch": 0.03773311080473115, |
|
"grad_norm": 3.421875, |
|
"learning_rate": 4.9883818765369194e-05, |
|
"loss": 0.3066, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 0.03845874755097598, |
|
"grad_norm": 3.609375, |
|
"learning_rate": 4.98782250124666e-05, |
|
"loss": 0.2938, |
|
"step": 530 |
|
}, |
|
{ |
|
"epoch": 0.03918438429722081, |
|
"grad_norm": 2.3125, |
|
"learning_rate": 4.987250007484172e-05, |
|
"loss": 0.2382, |
|
"step": 540 |
|
}, |
|
{ |
|
"epoch": 0.03991002104346564, |
|
"grad_norm": 4.03125, |
|
"learning_rate": 4.986664398268256e-05, |
|
"loss": 0.2569, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 0.04063565778971047, |
|
"grad_norm": 3.59375, |
|
"learning_rate": 4.986065676686874e-05, |
|
"loss": 0.2508, |
|
"step": 560 |
|
}, |
|
{ |
|
"epoch": 0.0413612945359553, |
|
"grad_norm": 5.8125, |
|
"learning_rate": 4.98545384589713e-05, |
|
"loss": 0.2815, |
|
"step": 570 |
|
}, |
|
{ |
|
"epoch": 0.04208693128220013, |
|
"grad_norm": 2.65625, |
|
"learning_rate": 4.984828909125251e-05, |
|
"loss": 0.2459, |
|
"step": 580 |
|
}, |
|
{ |
|
"epoch": 0.042812568028444964, |
|
"grad_norm": 1.9609375, |
|
"learning_rate": 4.9841908696665764e-05, |
|
"loss": 0.1962, |
|
"step": 590 |
|
}, |
|
{ |
|
"epoch": 0.04353820477468979, |
|
"grad_norm": 1.765625, |
|
"learning_rate": 4.9835397308855344e-05, |
|
"loss": 0.2863, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 0.04426384152093462, |
|
"grad_norm": 2.96875, |
|
"learning_rate": 4.9828754962156286e-05, |
|
"loss": 0.291, |
|
"step": 610 |
|
}, |
|
{ |
|
"epoch": 0.04498947826717945, |
|
"grad_norm": 6.71875, |
|
"learning_rate": 4.9821981691594175e-05, |
|
"loss": 0.3332, |
|
"step": 620 |
|
}, |
|
{ |
|
"epoch": 0.04571511501342428, |
|
"grad_norm": 4.21875, |
|
"learning_rate": 4.981507753288497e-05, |
|
"loss": 0.2352, |
|
"step": 630 |
|
}, |
|
{ |
|
"epoch": 0.04644075175966911, |
|
"grad_norm": 3.265625, |
|
"learning_rate": 4.9808042522434814e-05, |
|
"loss": 0.2336, |
|
"step": 640 |
|
}, |
|
{ |
|
"epoch": 0.04716638850591394, |
|
"grad_norm": 5.0625, |
|
"learning_rate": 4.9800876697339824e-05, |
|
"loss": 0.2952, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 0.04789202525215877, |
|
"grad_norm": 4.21875, |
|
"learning_rate": 4.979358009538594e-05, |
|
"loss": 0.3213, |
|
"step": 660 |
|
}, |
|
{ |
|
"epoch": 0.0486176619984036, |
|
"grad_norm": 4.6875, |
|
"learning_rate": 4.978615275504869e-05, |
|
"loss": 0.2433, |
|
"step": 670 |
|
}, |
|
{ |
|
"epoch": 0.04934329874464843, |
|
"grad_norm": 5.03125, |
|
"learning_rate": 4.977859471549297e-05, |
|
"loss": 0.2746, |
|
"step": 680 |
|
}, |
|
{ |
|
"epoch": 0.05006893549089326, |
|
"grad_norm": 1.6640625, |
|
"learning_rate": 4.977090601657289e-05, |
|
"loss": 0.1977, |
|
"step": 690 |
|
}, |
|
{ |
|
"epoch": 0.05079457223713809, |
|
"grad_norm": 2.859375, |
|
"learning_rate": 4.976308669883153e-05, |
|
"loss": 0.2929, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 0.05152020898338292, |
|
"grad_norm": 2.171875, |
|
"learning_rate": 4.975513680350073e-05, |
|
"loss": 0.2365, |
|
"step": 710 |
|
}, |
|
{ |
|
"epoch": 0.052245845729627746, |
|
"grad_norm": 3.484375, |
|
"learning_rate": 4.974705637250089e-05, |
|
"loss": 0.2096, |
|
"step": 720 |
|
}, |
|
{ |
|
"epoch": 0.05297148247587258, |
|
"grad_norm": 2.046875, |
|
"learning_rate": 4.97388454484407e-05, |
|
"loss": 0.2366, |
|
"step": 730 |
|
}, |
|
{ |
|
"epoch": 0.05369711922211741, |
|
"grad_norm": 2.5625, |
|
"learning_rate": 4.973050407461698e-05, |
|
"loss": 0.2849, |
|
"step": 740 |
|
}, |
|
{ |
|
"epoch": 0.05442275596836224, |
|
"grad_norm": 1.1484375, |
|
"learning_rate": 4.972203229501441e-05, |
|
"loss": 0.2053, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 0.055148392714607065, |
|
"grad_norm": 4.96875, |
|
"learning_rate": 4.971343015430532e-05, |
|
"loss": 0.2244, |
|
"step": 760 |
|
}, |
|
{ |
|
"epoch": 0.0558740294608519, |
|
"grad_norm": 5.25, |
|
"learning_rate": 4.970469769784941e-05, |
|
"loss": 0.2291, |
|
"step": 770 |
|
}, |
|
{ |
|
"epoch": 0.05659966620709673, |
|
"grad_norm": 4.875, |
|
"learning_rate": 4.96958349716936e-05, |
|
"loss": 0.2903, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 0.05732530295334156, |
|
"grad_norm": 5.125, |
|
"learning_rate": 4.968684202257169e-05, |
|
"loss": 0.3105, |
|
"step": 790 |
|
}, |
|
{ |
|
"epoch": 0.058050939699586385, |
|
"grad_norm": 3.84375, |
|
"learning_rate": 4.967771889790416e-05, |
|
"loss": 0.235, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 0.058776576445831216, |
|
"grad_norm": 2.90625, |
|
"learning_rate": 4.966846564579792e-05, |
|
"loss": 0.1795, |
|
"step": 810 |
|
}, |
|
{ |
|
"epoch": 0.05950221319207605, |
|
"grad_norm": 3.28125, |
|
"learning_rate": 4.965908231504607e-05, |
|
"loss": 0.2627, |
|
"step": 820 |
|
}, |
|
{ |
|
"epoch": 0.06022784993832088, |
|
"grad_norm": 3.859375, |
|
"learning_rate": 4.964956895512759e-05, |
|
"loss": 0.236, |
|
"step": 830 |
|
}, |
|
{ |
|
"epoch": 0.060953486684565704, |
|
"grad_norm": 3.140625, |
|
"learning_rate": 4.963992561620714e-05, |
|
"loss": 0.2544, |
|
"step": 840 |
|
}, |
|
{ |
|
"epoch": 0.061679123430810535, |
|
"grad_norm": 2.703125, |
|
"learning_rate": 4.963015234913475e-05, |
|
"loss": 0.2296, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 0.06240476017705537, |
|
"grad_norm": 2.046875, |
|
"learning_rate": 4.96202492054456e-05, |
|
"loss": 0.1725, |
|
"step": 860 |
|
}, |
|
{ |
|
"epoch": 0.06313039692330019, |
|
"grad_norm": 5.59375, |
|
"learning_rate": 4.9610216237359684e-05, |
|
"loss": 0.3872, |
|
"step": 870 |
|
}, |
|
{ |
|
"epoch": 0.06385603366954502, |
|
"grad_norm": 7.0, |
|
"learning_rate": 4.960005349778159e-05, |
|
"loss": 0.2456, |
|
"step": 880 |
|
}, |
|
{ |
|
"epoch": 0.06458167041578985, |
|
"grad_norm": 4.25, |
|
"learning_rate": 4.95897610403002e-05, |
|
"loss": 0.2577, |
|
"step": 890 |
|
}, |
|
{ |
|
"epoch": 0.06530730716203469, |
|
"grad_norm": 2.296875, |
|
"learning_rate": 4.95793389191884e-05, |
|
"loss": 0.213, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 0.06603294390827952, |
|
"grad_norm": 3.828125, |
|
"learning_rate": 4.95687871894028e-05, |
|
"loss": 0.1865, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 0.06675858065452435, |
|
"grad_norm": 0.8203125, |
|
"learning_rate": 4.9558105906583466e-05, |
|
"loss": 0.1707, |
|
"step": 920 |
|
}, |
|
{ |
|
"epoch": 0.06748421740076918, |
|
"grad_norm": 4.84375, |
|
"learning_rate": 4.9547295127053586e-05, |
|
"loss": 0.3426, |
|
"step": 930 |
|
}, |
|
{ |
|
"epoch": 0.068209854147014, |
|
"grad_norm": 3.859375, |
|
"learning_rate": 4.95363549078192e-05, |
|
"loss": 0.2181, |
|
"step": 940 |
|
}, |
|
{ |
|
"epoch": 0.06893549089325883, |
|
"grad_norm": 3.765625, |
|
"learning_rate": 4.952528530656889e-05, |
|
"loss": 0.2281, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 0.06966112763950366, |
|
"grad_norm": 4.21875, |
|
"learning_rate": 4.9514086381673496e-05, |
|
"loss": 0.2399, |
|
"step": 960 |
|
}, |
|
{ |
|
"epoch": 0.0703867643857485, |
|
"grad_norm": 4.625, |
|
"learning_rate": 4.9502758192185774e-05, |
|
"loss": 0.2499, |
|
"step": 970 |
|
}, |
|
{ |
|
"epoch": 0.07111240113199332, |
|
"grad_norm": 5.6875, |
|
"learning_rate": 4.949130079784009e-05, |
|
"loss": 0.2576, |
|
"step": 980 |
|
}, |
|
{ |
|
"epoch": 0.07183803787823816, |
|
"grad_norm": 3.984375, |
|
"learning_rate": 4.9479714259052143e-05, |
|
"loss": 0.2581, |
|
"step": 990 |
|
}, |
|
{ |
|
"epoch": 0.07256367462448299, |
|
"grad_norm": 5.5625, |
|
"learning_rate": 4.946799863691862e-05, |
|
"loss": 0.2237, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.07328931137072782, |
|
"grad_norm": 0.77734375, |
|
"learning_rate": 4.9456153993216855e-05, |
|
"loss": 0.2626, |
|
"step": 1010 |
|
}, |
|
{ |
|
"epoch": 0.07401494811697264, |
|
"grad_norm": 2.265625, |
|
"learning_rate": 4.944418039040453e-05, |
|
"loss": 0.22, |
|
"step": 1020 |
|
}, |
|
{ |
|
"epoch": 0.07474058486321747, |
|
"grad_norm": 2.8125, |
|
"learning_rate": 4.943207789161934e-05, |
|
"loss": 0.1587, |
|
"step": 1030 |
|
}, |
|
{ |
|
"epoch": 0.0754662216094623, |
|
"grad_norm": 0.349609375, |
|
"learning_rate": 4.9419846560678657e-05, |
|
"loss": 0.2245, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 0.07619185835570713, |
|
"grad_norm": 5.0625, |
|
"learning_rate": 4.940748646207919e-05, |
|
"loss": 0.2629, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 0.07691749510195196, |
|
"grad_norm": 4.6875, |
|
"learning_rate": 4.9394997660996664e-05, |
|
"loss": 0.2444, |
|
"step": 1060 |
|
}, |
|
{ |
|
"epoch": 0.0776431318481968, |
|
"grad_norm": 1.484375, |
|
"learning_rate": 4.938238022328545e-05, |
|
"loss": 0.1888, |
|
"step": 1070 |
|
}, |
|
{ |
|
"epoch": 0.07836876859444163, |
|
"grad_norm": 5.5625, |
|
"learning_rate": 4.9369634215478225e-05, |
|
"loss": 0.1678, |
|
"step": 1080 |
|
}, |
|
{ |
|
"epoch": 0.07909440534068646, |
|
"grad_norm": 2.359375, |
|
"learning_rate": 4.935675970478563e-05, |
|
"loss": 0.1647, |
|
"step": 1090 |
|
}, |
|
{ |
|
"epoch": 0.07982004208693128, |
|
"grad_norm": 2.921875, |
|
"learning_rate": 4.9343756759095924e-05, |
|
"loss": 0.2022, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 0.08054567883317611, |
|
"grad_norm": 6.15625, |
|
"learning_rate": 4.933062544697459e-05, |
|
"loss": 0.2638, |
|
"step": 1110 |
|
}, |
|
{ |
|
"epoch": 0.08127131557942094, |
|
"grad_norm": 4.09375, |
|
"learning_rate": 4.931736583766402e-05, |
|
"loss": 0.2955, |
|
"step": 1120 |
|
}, |
|
{ |
|
"epoch": 0.08199695232566577, |
|
"grad_norm": 1.6953125, |
|
"learning_rate": 4.9303978001083105e-05, |
|
"loss": 0.2066, |
|
"step": 1130 |
|
}, |
|
{ |
|
"epoch": 0.0827225890719106, |
|
"grad_norm": 4.84375, |
|
"learning_rate": 4.929046200782691e-05, |
|
"loss": 0.203, |
|
"step": 1140 |
|
}, |
|
{ |
|
"epoch": 0.08344822581815543, |
|
"grad_norm": 3.578125, |
|
"learning_rate": 4.927681792916626e-05, |
|
"loss": 0.3354, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 0.08417386256440026, |
|
"grad_norm": 3.203125, |
|
"learning_rate": 4.926304583704739e-05, |
|
"loss": 0.2382, |
|
"step": 1160 |
|
}, |
|
{ |
|
"epoch": 0.0848994993106451, |
|
"grad_norm": 1.75, |
|
"learning_rate": 4.924914580409157e-05, |
|
"loss": 0.2661, |
|
"step": 1170 |
|
}, |
|
{ |
|
"epoch": 0.08562513605688993, |
|
"grad_norm": 2.828125, |
|
"learning_rate": 4.9235117903594696e-05, |
|
"loss": 0.202, |
|
"step": 1180 |
|
}, |
|
{ |
|
"epoch": 0.08635077280313475, |
|
"grad_norm": 15.5625, |
|
"learning_rate": 4.922096220952692e-05, |
|
"loss": 0.2572, |
|
"step": 1190 |
|
}, |
|
{ |
|
"epoch": 0.08707640954937958, |
|
"grad_norm": 5.8125, |
|
"learning_rate": 4.9206678796532285e-05, |
|
"loss": 0.2071, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 0.08780204629562441, |
|
"grad_norm": 5.125, |
|
"learning_rate": 4.919226773992826e-05, |
|
"loss": 0.2708, |
|
"step": 1210 |
|
}, |
|
{ |
|
"epoch": 0.08852768304186924, |
|
"grad_norm": 5.28125, |
|
"learning_rate": 4.917772911570544e-05, |
|
"loss": 0.1818, |
|
"step": 1220 |
|
}, |
|
{ |
|
"epoch": 0.08925331978811407, |
|
"grad_norm": 1.9140625, |
|
"learning_rate": 4.916306300052704e-05, |
|
"loss": 0.1904, |
|
"step": 1230 |
|
}, |
|
{ |
|
"epoch": 0.0899789565343589, |
|
"grad_norm": 1.5, |
|
"learning_rate": 4.91482694717286e-05, |
|
"loss": 0.2974, |
|
"step": 1240 |
|
}, |
|
{ |
|
"epoch": 0.09070459328060373, |
|
"grad_norm": 4.15625, |
|
"learning_rate": 4.9133348607317464e-05, |
|
"loss": 0.2587, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 0.09143023002684857, |
|
"grad_norm": 1.8046875, |
|
"learning_rate": 4.911830048597247e-05, |
|
"loss": 0.1893, |
|
"step": 1260 |
|
}, |
|
{ |
|
"epoch": 0.09215586677309338, |
|
"grad_norm": 3.0, |
|
"learning_rate": 4.910312518704348e-05, |
|
"loss": 0.1727, |
|
"step": 1270 |
|
}, |
|
{ |
|
"epoch": 0.09288150351933822, |
|
"grad_norm": 5.375, |
|
"learning_rate": 4.908782279055095e-05, |
|
"loss": 0.1781, |
|
"step": 1280 |
|
}, |
|
{ |
|
"epoch": 0.09360714026558305, |
|
"grad_norm": 3.5625, |
|
"learning_rate": 4.907239337718557e-05, |
|
"loss": 0.2121, |
|
"step": 1290 |
|
}, |
|
{ |
|
"epoch": 0.09433277701182788, |
|
"grad_norm": 3.375, |
|
"learning_rate": 4.9056837028307755e-05, |
|
"loss": 0.2192, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 0.09505841375807271, |
|
"grad_norm": 2.0625, |
|
"learning_rate": 4.9041153825947314e-05, |
|
"loss": 0.1038, |
|
"step": 1310 |
|
}, |
|
{ |
|
"epoch": 0.09578405050431754, |
|
"grad_norm": 3.0625, |
|
"learning_rate": 4.902534385280292e-05, |
|
"loss": 0.2037, |
|
"step": 1320 |
|
}, |
|
{ |
|
"epoch": 0.09650968725056237, |
|
"grad_norm": 6.5, |
|
"learning_rate": 4.900940719224173e-05, |
|
"loss": 0.2828, |
|
"step": 1330 |
|
}, |
|
{ |
|
"epoch": 0.0972353239968072, |
|
"grad_norm": 10.0, |
|
"learning_rate": 4.899334392829893e-05, |
|
"loss": 0.332, |
|
"step": 1340 |
|
}, |
|
{ |
|
"epoch": 0.09796096074305202, |
|
"grad_norm": 6.4375, |
|
"learning_rate": 4.897715414567732e-05, |
|
"loss": 0.2783, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 0.09868659748929685, |
|
"grad_norm": 4.46875, |
|
"learning_rate": 4.8960837929746794e-05, |
|
"loss": 0.2361, |
|
"step": 1360 |
|
}, |
|
{ |
|
"epoch": 0.09941223423554169, |
|
"grad_norm": 0.94140625, |
|
"learning_rate": 4.894439536654399e-05, |
|
"loss": 0.2502, |
|
"step": 1370 |
|
}, |
|
{ |
|
"epoch": 0.10013787098178652, |
|
"grad_norm": 4.0625, |
|
"learning_rate": 4.892782654277175e-05, |
|
"loss": 0.2139, |
|
"step": 1380 |
|
}, |
|
{ |
|
"epoch": 0.10086350772803135, |
|
"grad_norm": 1.75, |
|
"learning_rate": 4.8911131545798715e-05, |
|
"loss": 0.2648, |
|
"step": 1390 |
|
}, |
|
{ |
|
"epoch": 0.10158914447427618, |
|
"grad_norm": 4.46875, |
|
"learning_rate": 4.889431046365883e-05, |
|
"loss": 0.1607, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 0.10231478122052101, |
|
"grad_norm": 4.9375, |
|
"learning_rate": 4.887736338505092e-05, |
|
"loss": 0.2305, |
|
"step": 1410 |
|
}, |
|
{ |
|
"epoch": 0.10304041796676584, |
|
"grad_norm": 2.53125, |
|
"learning_rate": 4.886029039933817e-05, |
|
"loss": 0.2019, |
|
"step": 1420 |
|
}, |
|
{ |
|
"epoch": 0.10376605471301066, |
|
"grad_norm": 2.421875, |
|
"learning_rate": 4.884309159654771e-05, |
|
"loss": 0.2753, |
|
"step": 1430 |
|
}, |
|
{ |
|
"epoch": 0.10449169145925549, |
|
"grad_norm": 5.25, |
|
"learning_rate": 4.8825767067370086e-05, |
|
"loss": 0.1891, |
|
"step": 1440 |
|
}, |
|
{ |
|
"epoch": 0.10521732820550032, |
|
"grad_norm": 4.5, |
|
"learning_rate": 4.880831690315884e-05, |
|
"loss": 0.1899, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 0.10594296495174516, |
|
"grad_norm": 3.984375, |
|
"learning_rate": 4.879074119592998e-05, |
|
"loss": 0.2918, |
|
"step": 1460 |
|
}, |
|
{ |
|
"epoch": 0.10666860169798999, |
|
"grad_norm": 4.40625, |
|
"learning_rate": 4.877304003836149e-05, |
|
"loss": 0.2286, |
|
"step": 1470 |
|
}, |
|
{ |
|
"epoch": 0.10739423844423482, |
|
"grad_norm": 2.859375, |
|
"learning_rate": 4.87552135237929e-05, |
|
"loss": 0.201, |
|
"step": 1480 |
|
}, |
|
{ |
|
"epoch": 0.10811987519047965, |
|
"grad_norm": 3.609375, |
|
"learning_rate": 4.873726174622475e-05, |
|
"loss": 0.3009, |
|
"step": 1490 |
|
}, |
|
{ |
|
"epoch": 0.10884551193672448, |
|
"grad_norm": 4.0625, |
|
"learning_rate": 4.8719184800318066e-05, |
|
"loss": 0.2848, |
|
"step": 1500 |
|
} |
|
], |
|
"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 |
|
} |
|
|