| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.21769102387344896, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "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 | |
| }, | |
| { | |
| "epoch": 0.1095711486829693, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 4.870098278139394e-05, | |
| "loss": 0.2262, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.11029678542921413, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 4.8682655785432975e-05, | |
| "loss": 0.2247, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.11102242217545896, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 4.866420390907476e-05, | |
| "loss": 0.2009, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.1117480589217038, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 4.8645627249617406e-05, | |
| "loss": 0.1592, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.11247369566794863, | |
| "grad_norm": 2.84375, | |
| "learning_rate": 4.8626925905017017e-05, | |
| "loss": 0.1411, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.11319933241419346, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 4.8608099973887167e-05, | |
| "loss": 0.2083, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.11392496916043829, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 4.858914955549839e-05, | |
| "loss": 0.1602, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.11465060590668312, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 4.857007474977764e-05, | |
| "loss": 0.345, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.11537624265292794, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 4.855087565730778e-05, | |
| "loss": 0.2072, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.11610187939917277, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 4.853155237932705e-05, | |
| "loss": 0.2014, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.1168275161454176, | |
| "grad_norm": 0.92578125, | |
| "learning_rate": 4.8512105017728514e-05, | |
| "loss": 0.2079, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.11755315289166243, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 4.8492533675059564e-05, | |
| "loss": 0.2249, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.11827878963790726, | |
| "grad_norm": 5.71875, | |
| "learning_rate": 4.847283845452133e-05, | |
| "loss": 0.217, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.1190044263841521, | |
| "grad_norm": 3.734375, | |
| "learning_rate": 4.8453019459968174e-05, | |
| "loss": 0.1916, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.11973006313039693, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 4.843307679590711e-05, | |
| "loss": 0.2316, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.12045569987664176, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 4.8413010567497295e-05, | |
| "loss": 0.142, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.12118133662288658, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 4.8392820880549434e-05, | |
| "loss": 0.1559, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.12190697336913141, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 4.837250784152524e-05, | |
| "loss": 0.2895, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.12263261011537624, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 4.835207155753686e-05, | |
| "loss": 0.1535, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.12335824686162107, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 4.833151213634635e-05, | |
| "loss": 0.1885, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1240838836078659, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 4.831082968636505e-05, | |
| "loss": 0.1253, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.12480952035411073, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 4.8290024316653055e-05, | |
| "loss": 0.1217, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.12553515710035557, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 4.826909613691861e-05, | |
| "loss": 0.1156, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.12626079384660038, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 4.824804525751757e-05, | |
| "loss": 0.1962, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.12698643059284523, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 4.822687178945278e-05, | |
| "loss": 0.2423, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.12771206733909005, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 4.820557584437349e-05, | |
| "loss": 0.2332, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.1284377040853349, | |
| "grad_norm": 7.03125, | |
| "learning_rate": 4.8184157534574806e-05, | |
| "loss": 0.2688, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.1291633408315797, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 4.816261697299706e-05, | |
| "loss": 0.164, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.12988897757782455, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 4.814095427322522e-05, | |
| "loss": 0.2155, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.13061461432406937, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 4.811916954948832e-05, | |
| "loss": 0.257, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.1313402510703142, | |
| "grad_norm": 5.625, | |
| "learning_rate": 4.809726291665881e-05, | |
| "loss": 0.1876, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.13206588781655904, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 4.8075234490251994e-05, | |
| "loss": 0.2021, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.13279152456280385, | |
| "grad_norm": 6.3125, | |
| "learning_rate": 4.805308438642539e-05, | |
| "loss": 0.2199, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.1335171613090487, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 4.8030812721978135e-05, | |
| "loss": 0.2505, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.13424279805529352, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.800841961435037e-05, | |
| "loss": 0.1716, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.13496843480153836, | |
| "grad_norm": 5.125, | |
| "learning_rate": 4.7985905181622605e-05, | |
| "loss": 0.1753, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.13569407154778318, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 4.7963269542515106e-05, | |
| "loss": 0.1421, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.136419708294028, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 4.794051281638729e-05, | |
| "loss": 0.2325, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.13714534504027284, | |
| "grad_norm": 2.625, | |
| "learning_rate": 4.7917635123237025e-05, | |
| "loss": 0.2494, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.13787098178651766, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 4.789463658370013e-05, | |
| "loss": 0.1902, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.1385966185327625, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 4.7871517319049566e-05, | |
| "loss": 0.2068, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.13932225527900732, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 4.784827745119493e-05, | |
| "loss": 0.2449, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.14004789202525217, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.7824917102681796e-05, | |
| "loss": 0.1809, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.140773528771497, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 4.780143639669098e-05, | |
| "loss": 0.1704, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.14149916551774183, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 4.7777835457038e-05, | |
| "loss": 0.2107, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.14222480226398665, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.775411440817236e-05, | |
| "loss": 0.2472, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.14295043901023147, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 4.7730273375176885e-05, | |
| "loss": 0.2565, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.1436760757564763, | |
| "grad_norm": 6.125, | |
| "learning_rate": 4.770631248376713e-05, | |
| "loss": 0.2251, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.14440171250272113, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 4.768223186029064e-05, | |
| "loss": 0.2618, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.14512734924896598, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 4.765803163172632e-05, | |
| "loss": 0.3018, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1458529859952108, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 4.7633711925683786e-05, | |
| "loss": 0.1203, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.14657862274145564, | |
| "grad_norm": 4.0, | |
| "learning_rate": 4.760927287040262e-05, | |
| "loss": 0.2229, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.14730425948770046, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 4.75847145947518e-05, | |
| "loss": 0.1754, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.14802989623394527, | |
| "grad_norm": 3.125, | |
| "learning_rate": 4.756003722822891e-05, | |
| "loss": 0.2028, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.14875553298019012, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 4.753524090095953e-05, | |
| "loss": 0.1875, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.14948116972643494, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 4.751032574369655e-05, | |
| "loss": 0.2235, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.15020680647267978, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 4.748529188781942e-05, | |
| "loss": 0.1848, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.1509324432189246, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 4.746013946533352e-05, | |
| "loss": 0.1657, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.15165807996516945, | |
| "grad_norm": 3.375, | |
| "learning_rate": 4.7434868608869435e-05, | |
| "loss": 0.2041, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.15238371671141426, | |
| "grad_norm": 2.5, | |
| "learning_rate": 4.740947945168227e-05, | |
| "loss": 0.1752, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.1531093534576591, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.7383972127650924e-05, | |
| "loss": 0.1863, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.15383499020390393, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 4.73583467712774e-05, | |
| "loss": 0.1921, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.15456062695014874, | |
| "grad_norm": 4.8125, | |
| "learning_rate": 4.733260351768611e-05, | |
| "loss": 0.2554, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.1552862636963936, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 4.7306742502623124e-05, | |
| "loss": 0.1769, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.1560119004426384, | |
| "grad_norm": 3.59375, | |
| "learning_rate": 4.72807638624555e-05, | |
| "loss": 0.1624, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.15673753718888325, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 4.7254667734170526e-05, | |
| "loss": 0.1537, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.15746317393512807, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 4.7228454255375024e-05, | |
| "loss": 0.2632, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.15818881068137292, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 4.72021235642946e-05, | |
| "loss": 0.2272, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.15891444742761773, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 4.7175675799772935e-05, | |
| "loss": 0.167, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.15964008417386255, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 4.714911110127106e-05, | |
| "loss": 0.1569, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.1603657209201074, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 4.712242960886659e-05, | |
| "loss": 0.2176, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.16109135766635221, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 4.709563146325301e-05, | |
| "loss": 0.144, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.16181699441259706, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 4.7068716805738935e-05, | |
| "loss": 0.2054, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.16254263115884188, | |
| "grad_norm": 3.6875, | |
| "learning_rate": 4.704168577824733e-05, | |
| "loss": 0.2055, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.16326826790508672, | |
| "grad_norm": 4.625, | |
| "learning_rate": 4.701453852331482e-05, | |
| "loss": 0.1583, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.16399390465133154, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 4.6987275184090876e-05, | |
| "loss": 0.2128, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.16471954139757639, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 4.6959895904337115e-05, | |
| "loss": 0.1703, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.1654451781438212, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 4.6932400828426505e-05, | |
| "loss": 0.1396, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.16617081489006602, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 4.690479010134261e-05, | |
| "loss": 0.1687, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.16689645163631087, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 4.687706386867885e-05, | |
| "loss": 0.2448, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.16762208838255568, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 4.6849222276637686e-05, | |
| "loss": 0.1783, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.16834772512880053, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 4.682126547202991e-05, | |
| "loss": 0.2612, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.16907336187504535, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.679319360227381e-05, | |
| "loss": 0.1674, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.1697989986212902, | |
| "grad_norm": 0.79296875, | |
| "learning_rate": 4.676500681539444e-05, | |
| "loss": 0.199, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.170524635367535, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 4.673670526002281e-05, | |
| "loss": 0.237, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.17125027211377986, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 4.6708289085395115e-05, | |
| "loss": 0.204, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.17197590886002467, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 4.6679758441351964e-05, | |
| "loss": 0.1338, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.1727015456062695, | |
| "grad_norm": 4.625, | |
| "learning_rate": 4.665111347833754e-05, | |
| "loss": 0.217, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.17342718235251434, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 4.662235434739885e-05, | |
| "loss": 0.1559, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.17415281909875915, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 4.659348120018493e-05, | |
| "loss": 0.2078, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.174878455845004, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 4.6564494188946015e-05, | |
| "loss": 0.2176, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.17560409259124882, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 4.653539346653278e-05, | |
| "loss": 0.1632, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.17632972933749366, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 4.650617918639546e-05, | |
| "loss": 0.2434, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.17705536608373848, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 4.647685150258314e-05, | |
| "loss": 0.19, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.1777810028299833, | |
| "grad_norm": 3.75, | |
| "learning_rate": 4.644741056974287e-05, | |
| "loss": 0.2335, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.17850663957622814, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 4.641785654311884e-05, | |
| "loss": 0.1898, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.17923227632247296, | |
| "grad_norm": 3.0625, | |
| "learning_rate": 4.6388189578551646e-05, | |
| "loss": 0.2144, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.1799579130687178, | |
| "grad_norm": 3.953125, | |
| "learning_rate": 4.635840983247738e-05, | |
| "loss": 0.2199, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.18068354981496262, | |
| "grad_norm": 3.671875, | |
| "learning_rate": 4.632851746192685e-05, | |
| "loss": 0.2206, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.18140918656120747, | |
| "grad_norm": 4.875, | |
| "learning_rate": 4.6298512624524734e-05, | |
| "loss": 0.1457, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.1821348233074523, | |
| "grad_norm": 6.125, | |
| "learning_rate": 4.626839547848877e-05, | |
| "loss": 0.1616, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.18286046005369713, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 4.62381661826289e-05, | |
| "loss": 0.1201, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.18358609679994195, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 4.620782489634642e-05, | |
| "loss": 0.1118, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.18431173354618677, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 4.617737177963319e-05, | |
| "loss": 0.1446, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.1850373702924316, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 4.614680699307076e-05, | |
| "loss": 0.1634, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.18576300703867643, | |
| "grad_norm": 4.96875, | |
| "learning_rate": 4.6116130697829495e-05, | |
| "loss": 0.193, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.18648864378492128, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 4.6085343055667786e-05, | |
| "loss": 0.2006, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.1872142805311661, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 4.6054444228931136e-05, | |
| "loss": 0.2664, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.18793991727741094, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 4.6023434380551354e-05, | |
| "loss": 0.1798, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.18866555402365576, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 4.5992313674045664e-05, | |
| "loss": 0.1751, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.18939119076990057, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 4.5961082273515856e-05, | |
| "loss": 0.0962, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.19011682751614542, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 4.592974034364741e-05, | |
| "loss": 0.3025, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.19084246426239024, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 4.589828804970865e-05, | |
| "loss": 0.2085, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.19156810100863508, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 4.5866725557549864e-05, | |
| "loss": 0.1526, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.1922937377548799, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 4.583505303360238e-05, | |
| "loss": 0.1638, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.19301937450112475, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 4.580327064487778e-05, | |
| "loss": 0.1828, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.19374501124736956, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 4.577137855896695e-05, | |
| "loss": 0.1708, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.1944706479936144, | |
| "grad_norm": 3.015625, | |
| "learning_rate": 4.5739376944039225e-05, | |
| "loss": 0.244, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.19519628473985923, | |
| "grad_norm": 3.5, | |
| "learning_rate": 4.5707265968841475e-05, | |
| "loss": 0.193, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.19592192148610404, | |
| "grad_norm": 5.25, | |
| "learning_rate": 4.5675045802697266e-05, | |
| "loss": 0.2082, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.1966475582323489, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 4.564271661550591e-05, | |
| "loss": 0.2354, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.1973731949785937, | |
| "grad_norm": 3.75, | |
| "learning_rate": 4.56102785777416e-05, | |
| "loss": 0.1798, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.19809883172483855, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 4.5577731860452525e-05, | |
| "loss": 0.1521, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.19882446847108337, | |
| "grad_norm": 4.53125, | |
| "learning_rate": 4.5545076635259924e-05, | |
| "loss": 0.1715, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.19955010521732822, | |
| "grad_norm": 3.25, | |
| "learning_rate": 4.5512313074357225e-05, | |
| "loss": 0.1588, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.20027574196357303, | |
| "grad_norm": 4.625, | |
| "learning_rate": 4.54794413505091e-05, | |
| "loss": 0.1447, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.20100137870981785, | |
| "grad_norm": 4.90625, | |
| "learning_rate": 4.544646163705059e-05, | |
| "loss": 0.1705, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.2017270154560627, | |
| "grad_norm": 1.75, | |
| "learning_rate": 4.541337410788616e-05, | |
| "loss": 0.1254, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.20245265220230751, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.538017893748881e-05, | |
| "loss": 0.1214, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.20317828894855236, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 4.534687630089912e-05, | |
| "loss": 0.1694, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.20390392569479718, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 4.5313466373724366e-05, | |
| "loss": 0.2304, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.20462956244104202, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 4.5279949332137564e-05, | |
| "loss": 0.0711, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.20535519918728684, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 4.524632535287655e-05, | |
| "loss": 0.1899, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.2060808359335317, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 4.521259461324308e-05, | |
| "loss": 0.1557, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.2068064726797765, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 4.51787572911018e-05, | |
| "loss": 0.1584, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.20753210942602132, | |
| "grad_norm": 0.81640625, | |
| "learning_rate": 4.514481356487945e-05, | |
| "loss": 0.1092, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.20825774617226617, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 4.51107636135638e-05, | |
| "loss": 0.1223, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.20898338291851098, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 4.507660761670277e-05, | |
| "loss": 0.1717, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.20970901966475583, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 4.504234575440345e-05, | |
| "loss": 0.2018, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.21043465641100065, | |
| "grad_norm": 2.890625, | |
| "learning_rate": 4.500797820733117e-05, | |
| "loss": 0.1408, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.2111602931572455, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 4.4973505156708576e-05, | |
| "loss": 0.2373, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.2118859299034903, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 4.4938926784314585e-05, | |
| "loss": 0.2772, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.21261156664973516, | |
| "grad_norm": 3.484375, | |
| "learning_rate": 4.490424327248354e-05, | |
| "loss": 0.1638, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.21333720339597997, | |
| "grad_norm": 1.375, | |
| "learning_rate": 4.486945480410415e-05, | |
| "loss": 0.1424, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.2140628401422248, | |
| "grad_norm": 5.53125, | |
| "learning_rate": 4.4834561562618595e-05, | |
| "loss": 0.1922, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.21478847688846964, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 4.479956373202151e-05, | |
| "loss": 0.1118, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.21551411363471445, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 4.476446149685906e-05, | |
| "loss": 0.18, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.2162397503809593, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 4.472925504222792e-05, | |
| "loss": 0.2539, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.21696538712720412, | |
| "grad_norm": 4.875, | |
| "learning_rate": 4.4693944553774335e-05, | |
| "loss": 0.1941, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.21769102387344896, | |
| "grad_norm": 4.25, | |
| "learning_rate": 4.465853021769313e-05, | |
| "loss": 0.2244, | |
| "step": 3000 | |
| } | |
| ], | |
| "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 | |
| } | |