| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 26.666666666666668, |
| "eval_steps": 500, |
| "global_step": 6000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.044444444444444446, |
| "grad_norm": 8.389141082763672, |
| "learning_rate": 3e-06, |
| "loss": 1.6669, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.08888888888888889, |
| "grad_norm": 5.795680999755859, |
| "learning_rate": 6.333333333333334e-06, |
| "loss": 1.2184, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.13333333333333333, |
| "grad_norm": 1.3343589305877686, |
| "learning_rate": 9.666666666666667e-06, |
| "loss": 0.5803, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.17777777777777778, |
| "grad_norm": 1.2480086088180542, |
| "learning_rate": 1.3000000000000001e-05, |
| "loss": 0.3342, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.2222222222222222, |
| "grad_norm": 1.8908542394638062, |
| "learning_rate": 1.6333333333333335e-05, |
| "loss": 0.2242, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.26666666666666666, |
| "grad_norm": 1.1537951231002808, |
| "learning_rate": 1.9666666666666666e-05, |
| "loss": 0.2021, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.3111111111111111, |
| "grad_norm": 0.7749633193016052, |
| "learning_rate": 2.3000000000000003e-05, |
| "loss": 0.1238, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.35555555555555557, |
| "grad_norm": 0.9088910818099976, |
| "learning_rate": 2.633333333333333e-05, |
| "loss": 0.1099, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.4, |
| "grad_norm": 0.8467608690261841, |
| "learning_rate": 2.9666666666666672e-05, |
| "loss": 0.0949, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.4444444444444444, |
| "grad_norm": 0.8593980073928833, |
| "learning_rate": 3.3e-05, |
| "loss": 0.0801, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.4888888888888889, |
| "grad_norm": 0.5598143935203552, |
| "learning_rate": 3.633333333333333e-05, |
| "loss": 0.0651, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.5333333333333333, |
| "grad_norm": 0.8410921692848206, |
| "learning_rate": 3.966666666666667e-05, |
| "loss": 0.0813, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.5777777777777777, |
| "grad_norm": 0.7393496632575989, |
| "learning_rate": 4.3e-05, |
| "loss": 0.0511, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.6222222222222222, |
| "grad_norm": 0.5598465800285339, |
| "learning_rate": 4.633333333333333e-05, |
| "loss": 0.0517, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.6666666666666666, |
| "grad_norm": 0.5250635743141174, |
| "learning_rate": 4.966666666666667e-05, |
| "loss": 0.0551, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.7111111111111111, |
| "grad_norm": 0.3859585225582123, |
| "learning_rate": 5.300000000000001e-05, |
| "loss": 0.0448, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.7555555555555555, |
| "grad_norm": 0.2931934893131256, |
| "learning_rate": 5.633333333333334e-05, |
| "loss": 0.0413, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.8, |
| "grad_norm": 0.3726317584514618, |
| "learning_rate": 5.966666666666667e-05, |
| "loss": 0.0368, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.8444444444444444, |
| "grad_norm": 0.4870269298553467, |
| "learning_rate": 6.3e-05, |
| "loss": 0.0439, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.8888888888888888, |
| "grad_norm": 0.6991150379180908, |
| "learning_rate": 6.633333333333334e-05, |
| "loss": 0.0449, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.9333333333333333, |
| "grad_norm": 0.8043137192726135, |
| "learning_rate": 6.966666666666668e-05, |
| "loss": 0.0428, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.9777777777777777, |
| "grad_norm": 0.5199294686317444, |
| "learning_rate": 7.3e-05, |
| "loss": 0.0497, |
| "step": 220 |
| }, |
| { |
| "epoch": 1.0222222222222221, |
| "grad_norm": 0.5832147598266602, |
| "learning_rate": 7.633333333333334e-05, |
| "loss": 0.043, |
| "step": 230 |
| }, |
| { |
| "epoch": 1.0666666666666667, |
| "grad_norm": 0.5762455463409424, |
| "learning_rate": 7.966666666666666e-05, |
| "loss": 0.042, |
| "step": 240 |
| }, |
| { |
| "epoch": 1.1111111111111112, |
| "grad_norm": 0.5296523571014404, |
| "learning_rate": 8.3e-05, |
| "loss": 0.0455, |
| "step": 250 |
| }, |
| { |
| "epoch": 1.1555555555555554, |
| "grad_norm": 0.5995281338691711, |
| "learning_rate": 8.633333333333334e-05, |
| "loss": 0.0446, |
| "step": 260 |
| }, |
| { |
| "epoch": 1.2, |
| "grad_norm": 0.44988754391670227, |
| "learning_rate": 8.966666666666666e-05, |
| "loss": 0.0409, |
| "step": 270 |
| }, |
| { |
| "epoch": 1.2444444444444445, |
| "grad_norm": 0.8164713978767395, |
| "learning_rate": 9.300000000000001e-05, |
| "loss": 0.0534, |
| "step": 280 |
| }, |
| { |
| "epoch": 1.2888888888888888, |
| "grad_norm": 0.7666190266609192, |
| "learning_rate": 9.633333333333335e-05, |
| "loss": 0.0402, |
| "step": 290 |
| }, |
| { |
| "epoch": 1.3333333333333333, |
| "grad_norm": 0.4470829963684082, |
| "learning_rate": 9.966666666666667e-05, |
| "loss": 0.049, |
| "step": 300 |
| }, |
| { |
| "epoch": 1.3777777777777778, |
| "grad_norm": 0.6112174987792969, |
| "learning_rate": 9.999938485971279e-05, |
| "loss": 0.0428, |
| "step": 310 |
| }, |
| { |
| "epoch": 1.4222222222222223, |
| "grad_norm": 0.6660336256027222, |
| "learning_rate": 9.999725846827562e-05, |
| "loss": 0.0436, |
| "step": 320 |
| }, |
| { |
| "epoch": 1.4666666666666668, |
| "grad_norm": 0.6819430589675903, |
| "learning_rate": 9.999361329594254e-05, |
| "loss": 0.0486, |
| "step": 330 |
| }, |
| { |
| "epoch": 1.511111111111111, |
| "grad_norm": 0.4355473518371582, |
| "learning_rate": 9.998844945344405e-05, |
| "loss": 0.0402, |
| "step": 340 |
| }, |
| { |
| "epoch": 1.5555555555555556, |
| "grad_norm": 0.46765050292015076, |
| "learning_rate": 9.99817670976436e-05, |
| "loss": 0.04, |
| "step": 350 |
| }, |
| { |
| "epoch": 1.6, |
| "grad_norm": 0.6269946098327637, |
| "learning_rate": 9.997356643153303e-05, |
| "loss": 0.04, |
| "step": 360 |
| }, |
| { |
| "epoch": 1.6444444444444444, |
| "grad_norm": 0.36605480313301086, |
| "learning_rate": 9.996384770422629e-05, |
| "loss": 0.0309, |
| "step": 370 |
| }, |
| { |
| "epoch": 1.6888888888888889, |
| "grad_norm": 0.4317551255226135, |
| "learning_rate": 9.995261121095194e-05, |
| "loss": 0.0325, |
| "step": 380 |
| }, |
| { |
| "epoch": 1.7333333333333334, |
| "grad_norm": 0.3933548927307129, |
| "learning_rate": 9.993985729304408e-05, |
| "loss": 0.0431, |
| "step": 390 |
| }, |
| { |
| "epoch": 1.7777777777777777, |
| "grad_norm": 0.3174668252468109, |
| "learning_rate": 9.992558633793212e-05, |
| "loss": 0.0278, |
| "step": 400 |
| }, |
| { |
| "epoch": 1.8222222222222222, |
| "grad_norm": 0.5716902613639832, |
| "learning_rate": 9.990979877912891e-05, |
| "loss": 0.0337, |
| "step": 410 |
| }, |
| { |
| "epoch": 1.8666666666666667, |
| "grad_norm": 0.43925878405570984, |
| "learning_rate": 9.989249509621759e-05, |
| "loss": 0.0403, |
| "step": 420 |
| }, |
| { |
| "epoch": 1.911111111111111, |
| "grad_norm": 0.5131371021270752, |
| "learning_rate": 9.987367581483705e-05, |
| "loss": 0.0315, |
| "step": 430 |
| }, |
| { |
| "epoch": 1.9555555555555557, |
| "grad_norm": 0.5429732203483582, |
| "learning_rate": 9.985334150666592e-05, |
| "loss": 0.0356, |
| "step": 440 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 0.6750060319900513, |
| "learning_rate": 9.983149278940526e-05, |
| "loss": 0.0331, |
| "step": 450 |
| }, |
| { |
| "epoch": 2.0444444444444443, |
| "grad_norm": 0.37390217185020447, |
| "learning_rate": 9.980813032675974e-05, |
| "loss": 0.0269, |
| "step": 460 |
| }, |
| { |
| "epoch": 2.088888888888889, |
| "grad_norm": 0.42575696110725403, |
| "learning_rate": 9.978325482841753e-05, |
| "loss": 0.0344, |
| "step": 470 |
| }, |
| { |
| "epoch": 2.1333333333333333, |
| "grad_norm": 0.3201421797275543, |
| "learning_rate": 9.975686705002867e-05, |
| "loss": 0.0253, |
| "step": 480 |
| }, |
| { |
| "epoch": 2.1777777777777776, |
| "grad_norm": 0.35483041405677795, |
| "learning_rate": 9.972896779318219e-05, |
| "loss": 0.031, |
| "step": 490 |
| }, |
| { |
| "epoch": 2.2222222222222223, |
| "grad_norm": 0.48134100437164307, |
| "learning_rate": 9.969955790538175e-05, |
| "loss": 0.0272, |
| "step": 500 |
| }, |
| { |
| "epoch": 2.2666666666666666, |
| "grad_norm": 0.3745223879814148, |
| "learning_rate": 9.966863828001982e-05, |
| "loss": 0.0316, |
| "step": 510 |
| }, |
| { |
| "epoch": 2.311111111111111, |
| "grad_norm": 0.4058675467967987, |
| "learning_rate": 9.963620985635065e-05, |
| "loss": 0.0223, |
| "step": 520 |
| }, |
| { |
| "epoch": 2.3555555555555556, |
| "grad_norm": 0.33446767926216125, |
| "learning_rate": 9.960227361946164e-05, |
| "loss": 0.0313, |
| "step": 530 |
| }, |
| { |
| "epoch": 2.4, |
| "grad_norm": 0.4397954046726227, |
| "learning_rate": 9.95668306002435e-05, |
| "loss": 0.0347, |
| "step": 540 |
| }, |
| { |
| "epoch": 2.4444444444444446, |
| "grad_norm": 0.34488892555236816, |
| "learning_rate": 9.952988187535886e-05, |
| "loss": 0.0207, |
| "step": 550 |
| }, |
| { |
| "epoch": 2.488888888888889, |
| "grad_norm": 0.2571009695529938, |
| "learning_rate": 9.949142856720961e-05, |
| "loss": 0.0286, |
| "step": 560 |
| }, |
| { |
| "epoch": 2.533333333333333, |
| "grad_norm": 0.5211164355278015, |
| "learning_rate": 9.945147184390278e-05, |
| "loss": 0.0449, |
| "step": 570 |
| }, |
| { |
| "epoch": 2.5777777777777775, |
| "grad_norm": 0.32125112414360046, |
| "learning_rate": 9.941001291921512e-05, |
| "loss": 0.0288, |
| "step": 580 |
| }, |
| { |
| "epoch": 2.6222222222222222, |
| "grad_norm": 0.4309665858745575, |
| "learning_rate": 9.936705305255612e-05, |
| "loss": 0.0261, |
| "step": 590 |
| }, |
| { |
| "epoch": 2.6666666666666665, |
| "grad_norm": 0.6445382833480835, |
| "learning_rate": 9.932259354892984e-05, |
| "loss": 0.0346, |
| "step": 600 |
| }, |
| { |
| "epoch": 2.7111111111111112, |
| "grad_norm": 0.4063079357147217, |
| "learning_rate": 9.927663575889521e-05, |
| "loss": 0.038, |
| "step": 610 |
| }, |
| { |
| "epoch": 2.7555555555555555, |
| "grad_norm": 0.5310010313987732, |
| "learning_rate": 9.922918107852504e-05, |
| "loss": 0.0336, |
| "step": 620 |
| }, |
| { |
| "epoch": 2.8, |
| "grad_norm": 0.448752224445343, |
| "learning_rate": 9.918023094936363e-05, |
| "loss": 0.0296, |
| "step": 630 |
| }, |
| { |
| "epoch": 2.8444444444444446, |
| "grad_norm": 0.4404497742652893, |
| "learning_rate": 9.912978685838294e-05, |
| "loss": 0.0272, |
| "step": 640 |
| }, |
| { |
| "epoch": 2.888888888888889, |
| "grad_norm": 0.30450695753097534, |
| "learning_rate": 9.90778503379374e-05, |
| "loss": 0.0276, |
| "step": 650 |
| }, |
| { |
| "epoch": 2.9333333333333336, |
| "grad_norm": 0.4000938832759857, |
| "learning_rate": 9.902442296571743e-05, |
| "loss": 0.0257, |
| "step": 660 |
| }, |
| { |
| "epoch": 2.977777777777778, |
| "grad_norm": 0.3830060660839081, |
| "learning_rate": 9.896950636470147e-05, |
| "loss": 0.018, |
| "step": 670 |
| }, |
| { |
| "epoch": 3.022222222222222, |
| "grad_norm": 0.3050785958766937, |
| "learning_rate": 9.891310220310666e-05, |
| "loss": 0.0291, |
| "step": 680 |
| }, |
| { |
| "epoch": 3.066666666666667, |
| "grad_norm": 0.3223566710948944, |
| "learning_rate": 9.885521219433823e-05, |
| "loss": 0.0162, |
| "step": 690 |
| }, |
| { |
| "epoch": 3.111111111111111, |
| "grad_norm": 0.27647706866264343, |
| "learning_rate": 9.879583809693738e-05, |
| "loss": 0.0202, |
| "step": 700 |
| }, |
| { |
| "epoch": 3.1555555555555554, |
| "grad_norm": 0.32375937700271606, |
| "learning_rate": 9.873498171452789e-05, |
| "loss": 0.0255, |
| "step": 710 |
| }, |
| { |
| "epoch": 3.2, |
| "grad_norm": 0.34824398159980774, |
| "learning_rate": 9.867264489576135e-05, |
| "loss": 0.0204, |
| "step": 720 |
| }, |
| { |
| "epoch": 3.2444444444444445, |
| "grad_norm": 0.3028803765773773, |
| "learning_rate": 9.860882953426099e-05, |
| "loss": 0.0261, |
| "step": 730 |
| }, |
| { |
| "epoch": 3.2888888888888888, |
| "grad_norm": 0.4095816910266876, |
| "learning_rate": 9.854353756856412e-05, |
| "loss": 0.0249, |
| "step": 740 |
| }, |
| { |
| "epoch": 3.3333333333333335, |
| "grad_norm": 0.2636725604534149, |
| "learning_rate": 9.847677098206332e-05, |
| "loss": 0.0246, |
| "step": 750 |
| }, |
| { |
| "epoch": 3.3777777777777778, |
| "grad_norm": 0.3920441269874573, |
| "learning_rate": 9.840853180294608e-05, |
| "loss": 0.0212, |
| "step": 760 |
| }, |
| { |
| "epoch": 3.422222222222222, |
| "grad_norm": 0.28506380319595337, |
| "learning_rate": 9.833882210413332e-05, |
| "loss": 0.0247, |
| "step": 770 |
| }, |
| { |
| "epoch": 3.466666666666667, |
| "grad_norm": 0.43021294474601746, |
| "learning_rate": 9.826764400321633e-05, |
| "loss": 0.028, |
| "step": 780 |
| }, |
| { |
| "epoch": 3.511111111111111, |
| "grad_norm": 0.42594388127326965, |
| "learning_rate": 9.819499966239243e-05, |
| "loss": 0.0205, |
| "step": 790 |
| }, |
| { |
| "epoch": 3.5555555555555554, |
| "grad_norm": 0.3029635548591614, |
| "learning_rate": 9.812089128839938e-05, |
| "loss": 0.0234, |
| "step": 800 |
| }, |
| { |
| "epoch": 3.6, |
| "grad_norm": 0.35575440526008606, |
| "learning_rate": 9.804532113244828e-05, |
| "loss": 0.0205, |
| "step": 810 |
| }, |
| { |
| "epoch": 3.6444444444444444, |
| "grad_norm": 0.39676201343536377, |
| "learning_rate": 9.796829149015517e-05, |
| "loss": 0.0342, |
| "step": 820 |
| }, |
| { |
| "epoch": 3.688888888888889, |
| "grad_norm": 0.36034658551216125, |
| "learning_rate": 9.788980470147132e-05, |
| "loss": 0.0235, |
| "step": 830 |
| }, |
| { |
| "epoch": 3.7333333333333334, |
| "grad_norm": 0.36659449338912964, |
| "learning_rate": 9.780986315061218e-05, |
| "loss": 0.0188, |
| "step": 840 |
| }, |
| { |
| "epoch": 3.7777777777777777, |
| "grad_norm": 0.3578703999519348, |
| "learning_rate": 9.772846926598491e-05, |
| "loss": 0.0206, |
| "step": 850 |
| }, |
| { |
| "epoch": 3.822222222222222, |
| "grad_norm": 0.430593341588974, |
| "learning_rate": 9.76456255201146e-05, |
| "loss": 0.0278, |
| "step": 860 |
| }, |
| { |
| "epoch": 3.8666666666666667, |
| "grad_norm": 0.3125564157962799, |
| "learning_rate": 9.756133442956923e-05, |
| "loss": 0.0195, |
| "step": 870 |
| }, |
| { |
| "epoch": 3.911111111111111, |
| "grad_norm": 0.33247822523117065, |
| "learning_rate": 9.747559855488313e-05, |
| "loss": 0.0222, |
| "step": 880 |
| }, |
| { |
| "epoch": 3.9555555555555557, |
| "grad_norm": 0.4072255790233612, |
| "learning_rate": 9.73884205004793e-05, |
| "loss": 0.0239, |
| "step": 890 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 0.4470653533935547, |
| "learning_rate": 9.729980291459019e-05, |
| "loss": 0.0218, |
| "step": 900 |
| }, |
| { |
| "epoch": 4.044444444444444, |
| "grad_norm": 0.3318852186203003, |
| "learning_rate": 9.720974848917735e-05, |
| "loss": 0.0222, |
| "step": 910 |
| }, |
| { |
| "epoch": 4.088888888888889, |
| "grad_norm": 0.35080021619796753, |
| "learning_rate": 9.711825995984957e-05, |
| "loss": 0.0288, |
| "step": 920 |
| }, |
| { |
| "epoch": 4.133333333333334, |
| "grad_norm": 0.28936517238616943, |
| "learning_rate": 9.702534010577991e-05, |
| "loss": 0.032, |
| "step": 930 |
| }, |
| { |
| "epoch": 4.177777777777778, |
| "grad_norm": 0.354633092880249, |
| "learning_rate": 9.693099174962103e-05, |
| "loss": 0.0151, |
| "step": 940 |
| }, |
| { |
| "epoch": 4.222222222222222, |
| "grad_norm": 0.13268741965293884, |
| "learning_rate": 9.683521775741977e-05, |
| "loss": 0.0181, |
| "step": 950 |
| }, |
| { |
| "epoch": 4.266666666666667, |
| "grad_norm": 0.3944114148616791, |
| "learning_rate": 9.673802103852979e-05, |
| "loss": 0.0253, |
| "step": 960 |
| }, |
| { |
| "epoch": 4.311111111111111, |
| "grad_norm": 0.35363683104515076, |
| "learning_rate": 9.663940454552342e-05, |
| "loss": 0.0174, |
| "step": 970 |
| }, |
| { |
| "epoch": 4.355555555555555, |
| "grad_norm": 0.30592408776283264, |
| "learning_rate": 9.65393712741018e-05, |
| "loss": 0.0213, |
| "step": 980 |
| }, |
| { |
| "epoch": 4.4, |
| "grad_norm": 0.46146953105926514, |
| "learning_rate": 9.6437924263004e-05, |
| "loss": 0.0224, |
| "step": 990 |
| }, |
| { |
| "epoch": 4.444444444444445, |
| "grad_norm": 0.3364277184009552, |
| "learning_rate": 9.63350665939146e-05, |
| "loss": 0.0185, |
| "step": 1000 |
| }, |
| { |
| "epoch": 4.488888888888889, |
| "grad_norm": 0.4292885363101959, |
| "learning_rate": 9.623080139137023e-05, |
| "loss": 0.0298, |
| "step": 1010 |
| }, |
| { |
| "epoch": 4.533333333333333, |
| "grad_norm": 0.2547819912433624, |
| "learning_rate": 9.612513182266447e-05, |
| "loss": 0.0214, |
| "step": 1020 |
| }, |
| { |
| "epoch": 4.5777777777777775, |
| "grad_norm": 0.3421243131160736, |
| "learning_rate": 9.601806109775179e-05, |
| "loss": 0.0209, |
| "step": 1030 |
| }, |
| { |
| "epoch": 4.622222222222222, |
| "grad_norm": 0.4412442147731781, |
| "learning_rate": 9.590959246914995e-05, |
| "loss": 0.0183, |
| "step": 1040 |
| }, |
| { |
| "epoch": 4.666666666666667, |
| "grad_norm": 0.32375088334083557, |
| "learning_rate": 9.579972923184122e-05, |
| "loss": 0.0207, |
| "step": 1050 |
| }, |
| { |
| "epoch": 4.711111111111111, |
| "grad_norm": 0.2944181561470032, |
| "learning_rate": 9.568847472317232e-05, |
| "loss": 0.0263, |
| "step": 1060 |
| }, |
| { |
| "epoch": 4.7555555555555555, |
| "grad_norm": 0.2564533054828644, |
| "learning_rate": 9.557583232275303e-05, |
| "loss": 0.016, |
| "step": 1070 |
| }, |
| { |
| "epoch": 4.8, |
| "grad_norm": 0.402987539768219, |
| "learning_rate": 9.546180545235344e-05, |
| "loss": 0.0225, |
| "step": 1080 |
| }, |
| { |
| "epoch": 4.844444444444444, |
| "grad_norm": 0.40538668632507324, |
| "learning_rate": 9.534639757580013e-05, |
| "loss": 0.0228, |
| "step": 1090 |
| }, |
| { |
| "epoch": 4.888888888888889, |
| "grad_norm": 0.3121398091316223, |
| "learning_rate": 9.522961219887092e-05, |
| "loss": 0.0181, |
| "step": 1100 |
| }, |
| { |
| "epoch": 4.933333333333334, |
| "grad_norm": 0.31277185678482056, |
| "learning_rate": 9.511145286918828e-05, |
| "loss": 0.0184, |
| "step": 1110 |
| }, |
| { |
| "epoch": 4.977777777777778, |
| "grad_norm": 0.26053231954574585, |
| "learning_rate": 9.499192317611167e-05, |
| "loss": 0.0178, |
| "step": 1120 |
| }, |
| { |
| "epoch": 5.022222222222222, |
| "grad_norm": 0.3141275942325592, |
| "learning_rate": 9.487102675062851e-05, |
| "loss": 0.0184, |
| "step": 1130 |
| }, |
| { |
| "epoch": 5.066666666666666, |
| "grad_norm": 0.3041296899318695, |
| "learning_rate": 9.474876726524374e-05, |
| "loss": 0.014, |
| "step": 1140 |
| }, |
| { |
| "epoch": 5.111111111111111, |
| "grad_norm": 0.2884008586406708, |
| "learning_rate": 9.462514843386845e-05, |
| "loss": 0.0222, |
| "step": 1150 |
| }, |
| { |
| "epoch": 5.155555555555556, |
| "grad_norm": 0.22763675451278687, |
| "learning_rate": 9.450017401170689e-05, |
| "loss": 0.0152, |
| "step": 1160 |
| }, |
| { |
| "epoch": 5.2, |
| "grad_norm": 0.2472584992647171, |
| "learning_rate": 9.437384779514256e-05, |
| "loss": 0.0185, |
| "step": 1170 |
| }, |
| { |
| "epoch": 5.2444444444444445, |
| "grad_norm": 0.30222344398498535, |
| "learning_rate": 9.424617362162271e-05, |
| "loss": 0.0197, |
| "step": 1180 |
| }, |
| { |
| "epoch": 5.288888888888889, |
| "grad_norm": 0.2808185815811157, |
| "learning_rate": 9.411715536954196e-05, |
| "loss": 0.0192, |
| "step": 1190 |
| }, |
| { |
| "epoch": 5.333333333333333, |
| "grad_norm": 0.24729570746421814, |
| "learning_rate": 9.39867969581243e-05, |
| "loss": 0.0202, |
| "step": 1200 |
| }, |
| { |
| "epoch": 5.377777777777778, |
| "grad_norm": 0.2555364966392517, |
| "learning_rate": 9.385510234730415e-05, |
| "loss": 0.026, |
| "step": 1210 |
| }, |
| { |
| "epoch": 5.4222222222222225, |
| "grad_norm": 0.42294108867645264, |
| "learning_rate": 9.372207553760603e-05, |
| "loss": 0.0204, |
| "step": 1220 |
| }, |
| { |
| "epoch": 5.466666666666667, |
| "grad_norm": 0.30665698647499084, |
| "learning_rate": 9.358772057002312e-05, |
| "loss": 0.0159, |
| "step": 1230 |
| }, |
| { |
| "epoch": 5.511111111111111, |
| "grad_norm": 0.3128688633441925, |
| "learning_rate": 9.345204152589428e-05, |
| "loss": 0.0239, |
| "step": 1240 |
| }, |
| { |
| "epoch": 5.555555555555555, |
| "grad_norm": 0.2368372231721878, |
| "learning_rate": 9.331504252678037e-05, |
| "loss": 0.014, |
| "step": 1250 |
| }, |
| { |
| "epoch": 5.6, |
| "grad_norm": 0.26079320907592773, |
| "learning_rate": 9.317672773433876e-05, |
| "loss": 0.015, |
| "step": 1260 |
| }, |
| { |
| "epoch": 5.644444444444445, |
| "grad_norm": 0.23744435608386993, |
| "learning_rate": 9.30371013501972e-05, |
| "loss": 0.0151, |
| "step": 1270 |
| }, |
| { |
| "epoch": 5.688888888888889, |
| "grad_norm": 0.2587639093399048, |
| "learning_rate": 9.289616761582587e-05, |
| "loss": 0.0181, |
| "step": 1280 |
| }, |
| { |
| "epoch": 5.733333333333333, |
| "grad_norm": 0.2486286461353302, |
| "learning_rate": 9.275393081240882e-05, |
| "loss": 0.0157, |
| "step": 1290 |
| }, |
| { |
| "epoch": 5.777777777777778, |
| "grad_norm": 0.2125803828239441, |
| "learning_rate": 9.261039526071374e-05, |
| "loss": 0.0168, |
| "step": 1300 |
| }, |
| { |
| "epoch": 5.822222222222222, |
| "grad_norm": 0.21079698204994202, |
| "learning_rate": 9.246556532096078e-05, |
| "loss": 0.0139, |
| "step": 1310 |
| }, |
| { |
| "epoch": 5.866666666666667, |
| "grad_norm": 0.27687445282936096, |
| "learning_rate": 9.231944539269009e-05, |
| "loss": 0.0168, |
| "step": 1320 |
| }, |
| { |
| "epoch": 5.911111111111111, |
| "grad_norm": 0.3301202058792114, |
| "learning_rate": 9.217203991462815e-05, |
| "loss": 0.0206, |
| "step": 1330 |
| }, |
| { |
| "epoch": 5.955555555555556, |
| "grad_norm": 0.3505961298942566, |
| "learning_rate": 9.202335336455296e-05, |
| "loss": 0.0275, |
| "step": 1340 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 0.29252493381500244, |
| "learning_rate": 9.187339025915802e-05, |
| "loss": 0.0177, |
| "step": 1350 |
| }, |
| { |
| "epoch": 6.044444444444444, |
| "grad_norm": 0.36544620990753174, |
| "learning_rate": 9.17221551539151e-05, |
| "loss": 0.0261, |
| "step": 1360 |
| }, |
| { |
| "epoch": 6.088888888888889, |
| "grad_norm": 0.41433846950531006, |
| "learning_rate": 9.156965264293586e-05, |
| "loss": 0.018, |
| "step": 1370 |
| }, |
| { |
| "epoch": 6.133333333333334, |
| "grad_norm": 0.3925895094871521, |
| "learning_rate": 9.141588735883232e-05, |
| "loss": 0.0172, |
| "step": 1380 |
| }, |
| { |
| "epoch": 6.177777777777778, |
| "grad_norm": 0.29030516743659973, |
| "learning_rate": 9.126086397257612e-05, |
| "loss": 0.0221, |
| "step": 1390 |
| }, |
| { |
| "epoch": 6.222222222222222, |
| "grad_norm": 0.29653820395469666, |
| "learning_rate": 9.110458719335659e-05, |
| "loss": 0.0182, |
| "step": 1400 |
| }, |
| { |
| "epoch": 6.266666666666667, |
| "grad_norm": 0.47055333852767944, |
| "learning_rate": 9.094706176843777e-05, |
| "loss": 0.031, |
| "step": 1410 |
| }, |
| { |
| "epoch": 6.311111111111111, |
| "grad_norm": 0.4297561049461365, |
| "learning_rate": 9.078829248301417e-05, |
| "loss": 0.0205, |
| "step": 1420 |
| }, |
| { |
| "epoch": 6.355555555555555, |
| "grad_norm": 0.4727896749973297, |
| "learning_rate": 9.062828416006539e-05, |
| "loss": 0.022, |
| "step": 1430 |
| }, |
| { |
| "epoch": 6.4, |
| "grad_norm": 0.47195112705230713, |
| "learning_rate": 9.046704166020961e-05, |
| "loss": 0.0168, |
| "step": 1440 |
| }, |
| { |
| "epoch": 6.444444444444445, |
| "grad_norm": 0.2805466055870056, |
| "learning_rate": 9.030456988155596e-05, |
| "loss": 0.0245, |
| "step": 1450 |
| }, |
| { |
| "epoch": 6.488888888888889, |
| "grad_norm": 0.24627427756786346, |
| "learning_rate": 9.014087375955573e-05, |
| "loss": 0.0183, |
| "step": 1460 |
| }, |
| { |
| "epoch": 6.533333333333333, |
| "grad_norm": 0.4139058589935303, |
| "learning_rate": 8.997595826685243e-05, |
| "loss": 0.0143, |
| "step": 1470 |
| }, |
| { |
| "epoch": 6.5777777777777775, |
| "grad_norm": 0.37815412878990173, |
| "learning_rate": 8.980982841313074e-05, |
| "loss": 0.0196, |
| "step": 1480 |
| }, |
| { |
| "epoch": 6.622222222222222, |
| "grad_norm": 0.36660775542259216, |
| "learning_rate": 8.964248924496435e-05, |
| "loss": 0.0166, |
| "step": 1490 |
| }, |
| { |
| "epoch": 6.666666666666667, |
| "grad_norm": 0.21890142560005188, |
| "learning_rate": 8.947394584566258e-05, |
| "loss": 0.0192, |
| "step": 1500 |
| }, |
| { |
| "epoch": 6.711111111111111, |
| "grad_norm": 0.26794126629829407, |
| "learning_rate": 8.930420333511606e-05, |
| "loss": 0.0137, |
| "step": 1510 |
| }, |
| { |
| "epoch": 6.7555555555555555, |
| "grad_norm": 0.34494853019714355, |
| "learning_rate": 8.913326686964117e-05, |
| "loss": 0.0174, |
| "step": 1520 |
| }, |
| { |
| "epoch": 6.8, |
| "grad_norm": 0.22623825073242188, |
| "learning_rate": 8.89611416418234e-05, |
| "loss": 0.0166, |
| "step": 1530 |
| }, |
| { |
| "epoch": 6.844444444444444, |
| "grad_norm": 0.2902779281139374, |
| "learning_rate": 8.878783288035957e-05, |
| "loss": 0.014, |
| "step": 1540 |
| }, |
| { |
| "epoch": 6.888888888888889, |
| "grad_norm": 0.2804957330226898, |
| "learning_rate": 8.86133458498991e-05, |
| "loss": 0.0207, |
| "step": 1550 |
| }, |
| { |
| "epoch": 6.933333333333334, |
| "grad_norm": 0.2749207615852356, |
| "learning_rate": 8.843768585088393e-05, |
| "loss": 0.0129, |
| "step": 1560 |
| }, |
| { |
| "epoch": 6.977777777777778, |
| "grad_norm": 0.2836652100086212, |
| "learning_rate": 8.82608582193877e-05, |
| "loss": 0.0151, |
| "step": 1570 |
| }, |
| { |
| "epoch": 7.022222222222222, |
| "grad_norm": 0.3585423231124878, |
| "learning_rate": 8.80828683269535e-05, |
| "loss": 0.0157, |
| "step": 1580 |
| }, |
| { |
| "epoch": 7.066666666666666, |
| "grad_norm": 0.18756495416164398, |
| "learning_rate": 8.790372158043074e-05, |
| "loss": 0.018, |
| "step": 1590 |
| }, |
| { |
| "epoch": 7.111111111111111, |
| "grad_norm": 0.30908632278442383, |
| "learning_rate": 8.772342342181095e-05, |
| "loss": 0.0191, |
| "step": 1600 |
| }, |
| { |
| "epoch": 7.155555555555556, |
| "grad_norm": 0.3190164268016815, |
| "learning_rate": 8.75419793280624e-05, |
| "loss": 0.0185, |
| "step": 1610 |
| }, |
| { |
| "epoch": 7.2, |
| "grad_norm": 0.20416675508022308, |
| "learning_rate": 8.735939481096378e-05, |
| "loss": 0.0165, |
| "step": 1620 |
| }, |
| { |
| "epoch": 7.2444444444444445, |
| "grad_norm": 0.27324119210243225, |
| "learning_rate": 8.717567541693673e-05, |
| "loss": 0.0224, |
| "step": 1630 |
| }, |
| { |
| "epoch": 7.288888888888889, |
| "grad_norm": 0.2618747651576996, |
| "learning_rate": 8.699082672687734e-05, |
| "loss": 0.015, |
| "step": 1640 |
| }, |
| { |
| "epoch": 7.333333333333333, |
| "grad_norm": 0.2985251545906067, |
| "learning_rate": 8.680485435598673e-05, |
| "loss": 0.0149, |
| "step": 1650 |
| }, |
| { |
| "epoch": 7.377777777777778, |
| "grad_norm": 0.40937545895576477, |
| "learning_rate": 8.661776395360029e-05, |
| "loss": 0.0254, |
| "step": 1660 |
| }, |
| { |
| "epoch": 7.4222222222222225, |
| "grad_norm": 0.4289396405220032, |
| "learning_rate": 8.642956120301626e-05, |
| "loss": 0.018, |
| "step": 1670 |
| }, |
| { |
| "epoch": 7.466666666666667, |
| "grad_norm": 0.3917348086833954, |
| "learning_rate": 8.624025182132292e-05, |
| "loss": 0.0167, |
| "step": 1680 |
| }, |
| { |
| "epoch": 7.511111111111111, |
| "grad_norm": 0.3585996925830841, |
| "learning_rate": 8.604984155922506e-05, |
| "loss": 0.0301, |
| "step": 1690 |
| }, |
| { |
| "epoch": 7.555555555555555, |
| "grad_norm": 0.2999572157859802, |
| "learning_rate": 8.585833620086918e-05, |
| "loss": 0.0158, |
| "step": 1700 |
| }, |
| { |
| "epoch": 7.6, |
| "grad_norm": 0.31374847888946533, |
| "learning_rate": 8.566574156366784e-05, |
| "loss": 0.0157, |
| "step": 1710 |
| }, |
| { |
| "epoch": 7.644444444444445, |
| "grad_norm": 0.3261179029941559, |
| "learning_rate": 8.547206349812298e-05, |
| "loss": 0.0192, |
| "step": 1720 |
| }, |
| { |
| "epoch": 7.688888888888889, |
| "grad_norm": 0.28447800874710083, |
| "learning_rate": 8.527730788764805e-05, |
| "loss": 0.0143, |
| "step": 1730 |
| }, |
| { |
| "epoch": 7.733333333333333, |
| "grad_norm": 0.2870975136756897, |
| "learning_rate": 8.508148064838948e-05, |
| "loss": 0.0204, |
| "step": 1740 |
| }, |
| { |
| "epoch": 7.777777777777778, |
| "grad_norm": 0.24093402922153473, |
| "learning_rate": 8.488458772904684e-05, |
| "loss": 0.0136, |
| "step": 1750 |
| }, |
| { |
| "epoch": 7.822222222222222, |
| "grad_norm": 0.2736014127731323, |
| "learning_rate": 8.468663511069217e-05, |
| "loss": 0.0143, |
| "step": 1760 |
| }, |
| { |
| "epoch": 7.866666666666667, |
| "grad_norm": 0.2989179193973541, |
| "learning_rate": 8.448762880658825e-05, |
| "loss": 0.0147, |
| "step": 1770 |
| }, |
| { |
| "epoch": 7.911111111111111, |
| "grad_norm": 0.19723202288150787, |
| "learning_rate": 8.428757486200603e-05, |
| "loss": 0.0113, |
| "step": 1780 |
| }, |
| { |
| "epoch": 7.955555555555556, |
| "grad_norm": 0.36469390988349915, |
| "learning_rate": 8.40864793540409e-05, |
| "loss": 0.02, |
| "step": 1790 |
| }, |
| { |
| "epoch": 8.0, |
| "grad_norm": 0.23067057132720947, |
| "learning_rate": 8.388434839142813e-05, |
| "loss": 0.0203, |
| "step": 1800 |
| }, |
| { |
| "epoch": 8.044444444444444, |
| "grad_norm": 0.27546632289886475, |
| "learning_rate": 8.368118811435726e-05, |
| "loss": 0.0137, |
| "step": 1810 |
| }, |
| { |
| "epoch": 8.088888888888889, |
| "grad_norm": 0.26730671525001526, |
| "learning_rate": 8.347700469428564e-05, |
| "loss": 0.0119, |
| "step": 1820 |
| }, |
| { |
| "epoch": 8.133333333333333, |
| "grad_norm": 0.240224227309227, |
| "learning_rate": 8.327180433375091e-05, |
| "loss": 0.0169, |
| "step": 1830 |
| }, |
| { |
| "epoch": 8.177777777777777, |
| "grad_norm": 0.374720960855484, |
| "learning_rate": 8.306559326618259e-05, |
| "loss": 0.0174, |
| "step": 1840 |
| }, |
| { |
| "epoch": 8.222222222222221, |
| "grad_norm": 0.28660672903060913, |
| "learning_rate": 8.285837775571276e-05, |
| "loss": 0.0159, |
| "step": 1850 |
| }, |
| { |
| "epoch": 8.266666666666667, |
| "grad_norm": 0.21417830884456635, |
| "learning_rate": 8.265016409698573e-05, |
| "loss": 0.0145, |
| "step": 1860 |
| }, |
| { |
| "epoch": 8.311111111111112, |
| "grad_norm": 0.2233651578426361, |
| "learning_rate": 8.244095861496686e-05, |
| "loss": 0.0128, |
| "step": 1870 |
| }, |
| { |
| "epoch": 8.355555555555556, |
| "grad_norm": 0.25084781646728516, |
| "learning_rate": 8.223076766475035e-05, |
| "loss": 0.0165, |
| "step": 1880 |
| }, |
| { |
| "epoch": 8.4, |
| "grad_norm": 0.2701694369316101, |
| "learning_rate": 8.201959763136633e-05, |
| "loss": 0.0181, |
| "step": 1890 |
| }, |
| { |
| "epoch": 8.444444444444445, |
| "grad_norm": 0.20661218464374542, |
| "learning_rate": 8.180745492958674e-05, |
| "loss": 0.0195, |
| "step": 1900 |
| }, |
| { |
| "epoch": 8.488888888888889, |
| "grad_norm": 0.3604136109352112, |
| "learning_rate": 8.159434600373061e-05, |
| "loss": 0.0139, |
| "step": 1910 |
| }, |
| { |
| "epoch": 8.533333333333333, |
| "grad_norm": 0.230348140001297, |
| "learning_rate": 8.138027732746818e-05, |
| "loss": 0.0107, |
| "step": 1920 |
| }, |
| { |
| "epoch": 8.577777777777778, |
| "grad_norm": 0.27103084325790405, |
| "learning_rate": 8.116525540362434e-05, |
| "loss": 0.0146, |
| "step": 1930 |
| }, |
| { |
| "epoch": 8.622222222222222, |
| "grad_norm": 0.26219263672828674, |
| "learning_rate": 8.094928676398101e-05, |
| "loss": 0.0104, |
| "step": 1940 |
| }, |
| { |
| "epoch": 8.666666666666666, |
| "grad_norm": 0.24252575635910034, |
| "learning_rate": 8.073237796907882e-05, |
| "loss": 0.0151, |
| "step": 1950 |
| }, |
| { |
| "epoch": 8.71111111111111, |
| "grad_norm": 0.3279937207698822, |
| "learning_rate": 8.051453560801772e-05, |
| "loss": 0.0147, |
| "step": 1960 |
| }, |
| { |
| "epoch": 8.755555555555556, |
| "grad_norm": 0.24254244565963745, |
| "learning_rate": 8.029576629825687e-05, |
| "loss": 0.0154, |
| "step": 1970 |
| }, |
| { |
| "epoch": 8.8, |
| "grad_norm": 0.2968604266643524, |
| "learning_rate": 8.007607668541362e-05, |
| "loss": 0.0185, |
| "step": 1980 |
| }, |
| { |
| "epoch": 8.844444444444445, |
| "grad_norm": 0.28566935658454895, |
| "learning_rate": 7.985547344306161e-05, |
| "loss": 0.0139, |
| "step": 1990 |
| }, |
| { |
| "epoch": 8.88888888888889, |
| "grad_norm": 0.2821260988712311, |
| "learning_rate": 7.963396327252812e-05, |
| "loss": 0.0171, |
| "step": 2000 |
| }, |
| { |
| "epoch": 8.933333333333334, |
| "grad_norm": 0.23841512203216553, |
| "learning_rate": 7.941155290269038e-05, |
| "loss": 0.015, |
| "step": 2010 |
| }, |
| { |
| "epoch": 8.977777777777778, |
| "grad_norm": 0.24885499477386475, |
| "learning_rate": 7.918824908977123e-05, |
| "loss": 0.0159, |
| "step": 2020 |
| }, |
| { |
| "epoch": 9.022222222222222, |
| "grad_norm": 0.30475321412086487, |
| "learning_rate": 7.896405861713394e-05, |
| "loss": 0.014, |
| "step": 2030 |
| }, |
| { |
| "epoch": 9.066666666666666, |
| "grad_norm": 0.27069583535194397, |
| "learning_rate": 7.873898829507606e-05, |
| "loss": 0.0147, |
| "step": 2040 |
| }, |
| { |
| "epoch": 9.11111111111111, |
| "grad_norm": 0.3272053897380829, |
| "learning_rate": 7.851304496062254e-05, |
| "loss": 0.0141, |
| "step": 2050 |
| }, |
| { |
| "epoch": 9.155555555555555, |
| "grad_norm": 0.4056951701641083, |
| "learning_rate": 7.828623547731818e-05, |
| "loss": 0.02, |
| "step": 2060 |
| }, |
| { |
| "epoch": 9.2, |
| "grad_norm": 0.2933352291584015, |
| "learning_rate": 7.80585667350189e-05, |
| "loss": 0.0167, |
| "step": 2070 |
| }, |
| { |
| "epoch": 9.244444444444444, |
| "grad_norm": 0.255327969789505, |
| "learning_rate": 7.783004564968263e-05, |
| "loss": 0.0169, |
| "step": 2080 |
| }, |
| { |
| "epoch": 9.28888888888889, |
| "grad_norm": 0.2668040990829468, |
| "learning_rate": 7.760067916315921e-05, |
| "loss": 0.0136, |
| "step": 2090 |
| }, |
| { |
| "epoch": 9.333333333333334, |
| "grad_norm": 0.30625119805336, |
| "learning_rate": 7.737047424297941e-05, |
| "loss": 0.0143, |
| "step": 2100 |
| }, |
| { |
| "epoch": 9.377777777777778, |
| "grad_norm": 0.22706979513168335, |
| "learning_rate": 7.713943788214337e-05, |
| "loss": 0.0163, |
| "step": 2110 |
| }, |
| { |
| "epoch": 9.422222222222222, |
| "grad_norm": 0.3837350904941559, |
| "learning_rate": 7.690757709890812e-05, |
| "loss": 0.0224, |
| "step": 2120 |
| }, |
| { |
| "epoch": 9.466666666666667, |
| "grad_norm": 0.30685484409332275, |
| "learning_rate": 7.66748989365744e-05, |
| "loss": 0.0191, |
| "step": 2130 |
| }, |
| { |
| "epoch": 9.511111111111111, |
| "grad_norm": 0.25118085741996765, |
| "learning_rate": 7.644141046327271e-05, |
| "loss": 0.0136, |
| "step": 2140 |
| }, |
| { |
| "epoch": 9.555555555555555, |
| "grad_norm": 0.19412904977798462, |
| "learning_rate": 7.620711877174866e-05, |
| "loss": 0.0101, |
| "step": 2150 |
| }, |
| { |
| "epoch": 9.6, |
| "grad_norm": 0.24385957419872284, |
| "learning_rate": 7.597203097914732e-05, |
| "loss": 0.021, |
| "step": 2160 |
| }, |
| { |
| "epoch": 9.644444444444444, |
| "grad_norm": 0.16870255768299103, |
| "learning_rate": 7.573615422679726e-05, |
| "loss": 0.0106, |
| "step": 2170 |
| }, |
| { |
| "epoch": 9.688888888888888, |
| "grad_norm": 0.21056927740573883, |
| "learning_rate": 7.549949567999345e-05, |
| "loss": 0.0095, |
| "step": 2180 |
| }, |
| { |
| "epoch": 9.733333333333333, |
| "grad_norm": 0.25866520404815674, |
| "learning_rate": 7.526206252777968e-05, |
| "loss": 0.0132, |
| "step": 2190 |
| }, |
| { |
| "epoch": 9.777777777777779, |
| "grad_norm": 0.26797500252723694, |
| "learning_rate": 7.50238619827301e-05, |
| "loss": 0.0138, |
| "step": 2200 |
| }, |
| { |
| "epoch": 9.822222222222223, |
| "grad_norm": 0.23290124535560608, |
| "learning_rate": 7.478490128073022e-05, |
| "loss": 0.0171, |
| "step": 2210 |
| }, |
| { |
| "epoch": 9.866666666666667, |
| "grad_norm": 0.3036246597766876, |
| "learning_rate": 7.454518768075704e-05, |
| "loss": 0.0145, |
| "step": 2220 |
| }, |
| { |
| "epoch": 9.911111111111111, |
| "grad_norm": 0.1924067586660385, |
| "learning_rate": 7.430472846465856e-05, |
| "loss": 0.0164, |
| "step": 2230 |
| }, |
| { |
| "epoch": 9.955555555555556, |
| "grad_norm": 0.23308129608631134, |
| "learning_rate": 7.406353093693253e-05, |
| "loss": 0.0149, |
| "step": 2240 |
| }, |
| { |
| "epoch": 10.0, |
| "grad_norm": 0.27332085371017456, |
| "learning_rate": 7.382160242450469e-05, |
| "loss": 0.0151, |
| "step": 2250 |
| }, |
| { |
| "epoch": 10.044444444444444, |
| "grad_norm": 0.2550460696220398, |
| "learning_rate": 7.357895027650598e-05, |
| "loss": 0.0237, |
| "step": 2260 |
| }, |
| { |
| "epoch": 10.088888888888889, |
| "grad_norm": 0.29789426922798157, |
| "learning_rate": 7.333558186404958e-05, |
| "loss": 0.0189, |
| "step": 2270 |
| }, |
| { |
| "epoch": 10.133333333333333, |
| "grad_norm": 0.2446967363357544, |
| "learning_rate": 7.309150458000668e-05, |
| "loss": 0.0168, |
| "step": 2280 |
| }, |
| { |
| "epoch": 10.177777777777777, |
| "grad_norm": 0.3303419053554535, |
| "learning_rate": 7.284672583878219e-05, |
| "loss": 0.0163, |
| "step": 2290 |
| }, |
| { |
| "epoch": 10.222222222222221, |
| "grad_norm": 0.21081838011741638, |
| "learning_rate": 7.260125307608929e-05, |
| "loss": 0.0151, |
| "step": 2300 |
| }, |
| { |
| "epoch": 10.266666666666667, |
| "grad_norm": 0.3426179885864258, |
| "learning_rate": 7.235509374872373e-05, |
| "loss": 0.0131, |
| "step": 2310 |
| }, |
| { |
| "epoch": 10.311111111111112, |
| "grad_norm": 0.3267004191875458, |
| "learning_rate": 7.210825533433719e-05, |
| "loss": 0.0169, |
| "step": 2320 |
| }, |
| { |
| "epoch": 10.355555555555556, |
| "grad_norm": 0.230835422873497, |
| "learning_rate": 7.186074533121013e-05, |
| "loss": 0.0122, |
| "step": 2330 |
| }, |
| { |
| "epoch": 10.4, |
| "grad_norm": 0.20230941474437714, |
| "learning_rate": 7.161257125802413e-05, |
| "loss": 0.0154, |
| "step": 2340 |
| }, |
| { |
| "epoch": 10.444444444444445, |
| "grad_norm": 0.3438013195991516, |
| "learning_rate": 7.136374065363334e-05, |
| "loss": 0.0137, |
| "step": 2350 |
| }, |
| { |
| "epoch": 10.488888888888889, |
| "grad_norm": 0.22763001918792725, |
| "learning_rate": 7.11142610768356e-05, |
| "loss": 0.0112, |
| "step": 2360 |
| }, |
| { |
| "epoch": 10.533333333333333, |
| "grad_norm": 0.18584167957305908, |
| "learning_rate": 7.086414010614276e-05, |
| "loss": 0.0159, |
| "step": 2370 |
| }, |
| { |
| "epoch": 10.577777777777778, |
| "grad_norm": 0.3445107638835907, |
| "learning_rate": 7.061338533955043e-05, |
| "loss": 0.0141, |
| "step": 2380 |
| }, |
| { |
| "epoch": 10.622222222222222, |
| "grad_norm": 0.3785816729068756, |
| "learning_rate": 7.036200439430725e-05, |
| "loss": 0.0205, |
| "step": 2390 |
| }, |
| { |
| "epoch": 10.666666666666666, |
| "grad_norm": 0.27487748861312866, |
| "learning_rate": 7.01100049066835e-05, |
| "loss": 0.0137, |
| "step": 2400 |
| }, |
| { |
| "epoch": 10.71111111111111, |
| "grad_norm": 0.27125370502471924, |
| "learning_rate": 6.985739453173903e-05, |
| "loss": 0.0155, |
| "step": 2410 |
| }, |
| { |
| "epoch": 10.755555555555556, |
| "grad_norm": 0.29975003004074097, |
| "learning_rate": 6.960418094309085e-05, |
| "loss": 0.0135, |
| "step": 2420 |
| }, |
| { |
| "epoch": 10.8, |
| "grad_norm": 0.2853398323059082, |
| "learning_rate": 6.93503718326799e-05, |
| "loss": 0.0125, |
| "step": 2430 |
| }, |
| { |
| "epoch": 10.844444444444445, |
| "grad_norm": 0.2614617943763733, |
| "learning_rate": 6.909597491053751e-05, |
| "loss": 0.0113, |
| "step": 2440 |
| }, |
| { |
| "epoch": 10.88888888888889, |
| "grad_norm": 0.20357394218444824, |
| "learning_rate": 6.884099790455113e-05, |
| "loss": 0.0163, |
| "step": 2450 |
| }, |
| { |
| "epoch": 10.933333333333334, |
| "grad_norm": 0.21524812281131744, |
| "learning_rate": 6.858544856022952e-05, |
| "loss": 0.0135, |
| "step": 2460 |
| }, |
| { |
| "epoch": 10.977777777777778, |
| "grad_norm": 0.238930344581604, |
| "learning_rate": 6.83293346404676e-05, |
| "loss": 0.0131, |
| "step": 2470 |
| }, |
| { |
| "epoch": 11.022222222222222, |
| "grad_norm": 0.19882909953594208, |
| "learning_rate": 6.80726639253105e-05, |
| "loss": 0.0131, |
| "step": 2480 |
| }, |
| { |
| "epoch": 11.066666666666666, |
| "grad_norm": 0.19018594920635223, |
| "learning_rate": 6.781544421171732e-05, |
| "loss": 0.0171, |
| "step": 2490 |
| }, |
| { |
| "epoch": 11.11111111111111, |
| "grad_norm": 0.205998957157135, |
| "learning_rate": 6.755768331332424e-05, |
| "loss": 0.0176, |
| "step": 2500 |
| }, |
| { |
| "epoch": 11.155555555555555, |
| "grad_norm": 0.38071829080581665, |
| "learning_rate": 6.729938906020713e-05, |
| "loss": 0.014, |
| "step": 2510 |
| }, |
| { |
| "epoch": 11.2, |
| "grad_norm": 0.20081113278865814, |
| "learning_rate": 6.704056929864376e-05, |
| "loss": 0.0108, |
| "step": 2520 |
| }, |
| { |
| "epoch": 11.244444444444444, |
| "grad_norm": 0.2780754268169403, |
| "learning_rate": 6.67812318908754e-05, |
| "loss": 0.0126, |
| "step": 2530 |
| }, |
| { |
| "epoch": 11.28888888888889, |
| "grad_norm": 0.22811779379844666, |
| "learning_rate": 6.6521384714868e-05, |
| "loss": 0.0148, |
| "step": 2540 |
| }, |
| { |
| "epoch": 11.333333333333334, |
| "grad_norm": 0.22134119272232056, |
| "learning_rate": 6.626103566407295e-05, |
| "loss": 0.0113, |
| "step": 2550 |
| }, |
| { |
| "epoch": 11.377777777777778, |
| "grad_norm": 0.253707617521286, |
| "learning_rate": 6.600019264718713e-05, |
| "loss": 0.0193, |
| "step": 2560 |
| }, |
| { |
| "epoch": 11.422222222222222, |
| "grad_norm": 0.289699912071228, |
| "learning_rate": 6.573886358791285e-05, |
| "loss": 0.0181, |
| "step": 2570 |
| }, |
| { |
| "epoch": 11.466666666666667, |
| "grad_norm": 0.27764225006103516, |
| "learning_rate": 6.547705642471703e-05, |
| "loss": 0.0193, |
| "step": 2580 |
| }, |
| { |
| "epoch": 11.511111111111111, |
| "grad_norm": 0.2252022624015808, |
| "learning_rate": 6.521477911059008e-05, |
| "loss": 0.0139, |
| "step": 2590 |
| }, |
| { |
| "epoch": 11.555555555555555, |
| "grad_norm": 0.23172208666801453, |
| "learning_rate": 6.495203961280434e-05, |
| "loss": 0.0167, |
| "step": 2600 |
| }, |
| { |
| "epoch": 11.6, |
| "grad_norm": 0.2703852951526642, |
| "learning_rate": 6.468884591267204e-05, |
| "loss": 0.0132, |
| "step": 2610 |
| }, |
| { |
| "epoch": 11.644444444444444, |
| "grad_norm": 0.1801518201828003, |
| "learning_rate": 6.44252060053028e-05, |
| "loss": 0.0119, |
| "step": 2620 |
| }, |
| { |
| "epoch": 11.688888888888888, |
| "grad_norm": 0.33298370242118835, |
| "learning_rate": 6.416112789936086e-05, |
| "loss": 0.0156, |
| "step": 2630 |
| }, |
| { |
| "epoch": 11.733333333333333, |
| "grad_norm": 0.2229069471359253, |
| "learning_rate": 6.389661961682173e-05, |
| "loss": 0.0167, |
| "step": 2640 |
| }, |
| { |
| "epoch": 11.777777777777779, |
| "grad_norm": 0.16075904667377472, |
| "learning_rate": 6.363168919272846e-05, |
| "loss": 0.0138, |
| "step": 2650 |
| }, |
| { |
| "epoch": 11.822222222222223, |
| "grad_norm": 0.23325996100902557, |
| "learning_rate": 6.336634467494768e-05, |
| "loss": 0.0123, |
| "step": 2660 |
| }, |
| { |
| "epoch": 11.866666666666667, |
| "grad_norm": 0.2903526723384857, |
| "learning_rate": 6.310059412392505e-05, |
| "loss": 0.0168, |
| "step": 2670 |
| }, |
| { |
| "epoch": 11.911111111111111, |
| "grad_norm": 0.18903379142284393, |
| "learning_rate": 6.283444561244042e-05, |
| "loss": 0.0123, |
| "step": 2680 |
| }, |
| { |
| "epoch": 11.955555555555556, |
| "grad_norm": 0.19427189230918884, |
| "learning_rate": 6.256790722536251e-05, |
| "loss": 0.0103, |
| "step": 2690 |
| }, |
| { |
| "epoch": 12.0, |
| "grad_norm": 0.3297659158706665, |
| "learning_rate": 6.230098705940354e-05, |
| "loss": 0.0117, |
| "step": 2700 |
| }, |
| { |
| "epoch": 12.044444444444444, |
| "grad_norm": 0.3722420632839203, |
| "learning_rate": 6.203369322287306e-05, |
| "loss": 0.0129, |
| "step": 2710 |
| }, |
| { |
| "epoch": 12.088888888888889, |
| "grad_norm": 0.2268306463956833, |
| "learning_rate": 6.17660338354317e-05, |
| "loss": 0.015, |
| "step": 2720 |
| }, |
| { |
| "epoch": 12.133333333333333, |
| "grad_norm": 0.37547385692596436, |
| "learning_rate": 6.149801702784456e-05, |
| "loss": 0.0098, |
| "step": 2730 |
| }, |
| { |
| "epoch": 12.177777777777777, |
| "grad_norm": 0.27735385298728943, |
| "learning_rate": 6.122965094173424e-05, |
| "loss": 0.0105, |
| "step": 2740 |
| }, |
| { |
| "epoch": 12.222222222222221, |
| "grad_norm": 0.20112404227256775, |
| "learning_rate": 6.0960943729333374e-05, |
| "loss": 0.0109, |
| "step": 2750 |
| }, |
| { |
| "epoch": 12.266666666666667, |
| "grad_norm": 0.17389725148677826, |
| "learning_rate": 6.069190355323717e-05, |
| "loss": 0.0126, |
| "step": 2760 |
| }, |
| { |
| "epoch": 12.311111111111112, |
| "grad_norm": 0.19448718428611755, |
| "learning_rate": 6.042253858615532e-05, |
| "loss": 0.0086, |
| "step": 2770 |
| }, |
| { |
| "epoch": 12.355555555555556, |
| "grad_norm": 0.2483740746974945, |
| "learning_rate": 6.015285701066382e-05, |
| "loss": 0.0102, |
| "step": 2780 |
| }, |
| { |
| "epoch": 12.4, |
| "grad_norm": 0.30692949891090393, |
| "learning_rate": 5.988286701895631e-05, |
| "loss": 0.0147, |
| "step": 2790 |
| }, |
| { |
| "epoch": 12.444444444444445, |
| "grad_norm": 0.23687846958637238, |
| "learning_rate": 5.961257681259535e-05, |
| "loss": 0.0106, |
| "step": 2800 |
| }, |
| { |
| "epoch": 12.488888888888889, |
| "grad_norm": 0.1735508292913437, |
| "learning_rate": 5.934199460226317e-05, |
| "loss": 0.0101, |
| "step": 2810 |
| }, |
| { |
| "epoch": 12.533333333333333, |
| "grad_norm": 0.1609802097082138, |
| "learning_rate": 5.9071128607512285e-05, |
| "loss": 0.0085, |
| "step": 2820 |
| }, |
| { |
| "epoch": 12.577777777777778, |
| "grad_norm": 0.26658353209495544, |
| "learning_rate": 5.8799987056515804e-05, |
| "loss": 0.0174, |
| "step": 2830 |
| }, |
| { |
| "epoch": 12.622222222222222, |
| "grad_norm": 0.29022836685180664, |
| "learning_rate": 5.8528578185817514e-05, |
| "loss": 0.0133, |
| "step": 2840 |
| }, |
| { |
| "epoch": 12.666666666666666, |
| "grad_norm": 0.12686531245708466, |
| "learning_rate": 5.825691024008162e-05, |
| "loss": 0.0109, |
| "step": 2850 |
| }, |
| { |
| "epoch": 12.71111111111111, |
| "grad_norm": 0.18447507917881012, |
| "learning_rate": 5.798499147184233e-05, |
| "loss": 0.0112, |
| "step": 2860 |
| }, |
| { |
| "epoch": 12.755555555555556, |
| "grad_norm": 0.2927151620388031, |
| "learning_rate": 5.771283014125317e-05, |
| "loss": 0.0127, |
| "step": 2870 |
| }, |
| { |
| "epoch": 12.8, |
| "grad_norm": 0.18832269310951233, |
| "learning_rate": 5.7440434515836064e-05, |
| "loss": 0.0106, |
| "step": 2880 |
| }, |
| { |
| "epoch": 12.844444444444445, |
| "grad_norm": 0.2521161735057831, |
| "learning_rate": 5.7167812870230094e-05, |
| "loss": 0.0132, |
| "step": 2890 |
| }, |
| { |
| "epoch": 12.88888888888889, |
| "grad_norm": 0.29403573274612427, |
| "learning_rate": 5.689497348594035e-05, |
| "loss": 0.0123, |
| "step": 2900 |
| }, |
| { |
| "epoch": 12.933333333333334, |
| "grad_norm": 0.33810940384864807, |
| "learning_rate": 5.662192465108613e-05, |
| "loss": 0.0125, |
| "step": 2910 |
| }, |
| { |
| "epoch": 12.977777777777778, |
| "grad_norm": 0.17549262940883636, |
| "learning_rate": 5.634867466014932e-05, |
| "loss": 0.0121, |
| "step": 2920 |
| }, |
| { |
| "epoch": 13.022222222222222, |
| "grad_norm": 0.1700422614812851, |
| "learning_rate": 5.607523181372234e-05, |
| "loss": 0.0111, |
| "step": 2930 |
| }, |
| { |
| "epoch": 13.066666666666666, |
| "grad_norm": 0.1775393933057785, |
| "learning_rate": 5.5801604418256117e-05, |
| "loss": 0.0121, |
| "step": 2940 |
| }, |
| { |
| "epoch": 13.11111111111111, |
| "grad_norm": 0.23550347983837128, |
| "learning_rate": 5.552780078580756e-05, |
| "loss": 0.019, |
| "step": 2950 |
| }, |
| { |
| "epoch": 13.155555555555555, |
| "grad_norm": 0.2833026647567749, |
| "learning_rate": 5.525382923378728e-05, |
| "loss": 0.0123, |
| "step": 2960 |
| }, |
| { |
| "epoch": 13.2, |
| "grad_norm": 0.1993987113237381, |
| "learning_rate": 5.49796980847068e-05, |
| "loss": 0.0104, |
| "step": 2970 |
| }, |
| { |
| "epoch": 13.244444444444444, |
| "grad_norm": 0.18256790935993195, |
| "learning_rate": 5.470541566592573e-05, |
| "loss": 0.0103, |
| "step": 2980 |
| }, |
| { |
| "epoch": 13.28888888888889, |
| "grad_norm": 0.2355622947216034, |
| "learning_rate": 5.443099030939887e-05, |
| "loss": 0.0138, |
| "step": 2990 |
| }, |
| { |
| "epoch": 13.333333333333334, |
| "grad_norm": 0.21295620501041412, |
| "learning_rate": 5.415643035142309e-05, |
| "loss": 0.0202, |
| "step": 3000 |
| }, |
| { |
| "epoch": 13.377777777777778, |
| "grad_norm": 0.20920953154563904, |
| "learning_rate": 5.3881744132384104e-05, |
| "loss": 0.0111, |
| "step": 3010 |
| }, |
| { |
| "epoch": 13.422222222222222, |
| "grad_norm": 0.19399991631507874, |
| "learning_rate": 5.360693999650303e-05, |
| "loss": 0.0115, |
| "step": 3020 |
| }, |
| { |
| "epoch": 13.466666666666667, |
| "grad_norm": 0.25338056683540344, |
| "learning_rate": 5.3332026291583016e-05, |
| "loss": 0.0117, |
| "step": 3030 |
| }, |
| { |
| "epoch": 13.511111111111111, |
| "grad_norm": 0.24487930536270142, |
| "learning_rate": 5.305701136875566e-05, |
| "loss": 0.0124, |
| "step": 3040 |
| }, |
| { |
| "epoch": 13.555555555555555, |
| "grad_norm": 0.2577228844165802, |
| "learning_rate": 5.278190358222721e-05, |
| "loss": 0.0129, |
| "step": 3050 |
| }, |
| { |
| "epoch": 13.6, |
| "grad_norm": 0.2954920828342438, |
| "learning_rate": 5.25067112890249e-05, |
| "loss": 0.0107, |
| "step": 3060 |
| }, |
| { |
| "epoch": 13.644444444444444, |
| "grad_norm": 0.24685540795326233, |
| "learning_rate": 5.2231442848743064e-05, |
| "loss": 0.0087, |
| "step": 3070 |
| }, |
| { |
| "epoch": 13.688888888888888, |
| "grad_norm": 0.3384992182254791, |
| "learning_rate": 5.1956106623289145e-05, |
| "loss": 0.0209, |
| "step": 3080 |
| }, |
| { |
| "epoch": 13.733333333333333, |
| "grad_norm": 0.32244497537612915, |
| "learning_rate": 5.168071097662972e-05, |
| "loss": 0.0142, |
| "step": 3090 |
| }, |
| { |
| "epoch": 13.777777777777779, |
| "grad_norm": 0.21630807220935822, |
| "learning_rate": 5.1405264274536445e-05, |
| "loss": 0.014, |
| "step": 3100 |
| }, |
| { |
| "epoch": 13.822222222222223, |
| "grad_norm": 0.26491034030914307, |
| "learning_rate": 5.112977488433188e-05, |
| "loss": 0.0107, |
| "step": 3110 |
| }, |
| { |
| "epoch": 13.866666666666667, |
| "grad_norm": 0.19515036046504974, |
| "learning_rate": 5.085425117463533e-05, |
| "loss": 0.0125, |
| "step": 3120 |
| }, |
| { |
| "epoch": 13.911111111111111, |
| "grad_norm": 0.2637917697429657, |
| "learning_rate": 5.057870151510864e-05, |
| "loss": 0.0108, |
| "step": 3130 |
| }, |
| { |
| "epoch": 13.955555555555556, |
| "grad_norm": 0.19765052199363708, |
| "learning_rate": 5.030313427620197e-05, |
| "loss": 0.0104, |
| "step": 3140 |
| }, |
| { |
| "epoch": 14.0, |
| "grad_norm": 0.17635421454906464, |
| "learning_rate": 5.0027557828899426e-05, |
| "loss": 0.0145, |
| "step": 3150 |
| }, |
| { |
| "epoch": 14.044444444444444, |
| "grad_norm": 0.16387489438056946, |
| "learning_rate": 4.975198054446492e-05, |
| "loss": 0.0078, |
| "step": 3160 |
| }, |
| { |
| "epoch": 14.088888888888889, |
| "grad_norm": 0.19314999878406525, |
| "learning_rate": 4.947641079418773e-05, |
| "loss": 0.0124, |
| "step": 3170 |
| }, |
| { |
| "epoch": 14.133333333333333, |
| "grad_norm": 0.18981797993183136, |
| "learning_rate": 4.920085694912828e-05, |
| "loss": 0.011, |
| "step": 3180 |
| }, |
| { |
| "epoch": 14.177777777777777, |
| "grad_norm": 0.224662646651268, |
| "learning_rate": 4.892532737986387e-05, |
| "loss": 0.0082, |
| "step": 3190 |
| }, |
| { |
| "epoch": 14.222222222222221, |
| "grad_norm": 0.2058078795671463, |
| "learning_rate": 4.864983045623434e-05, |
| "loss": 0.0132, |
| "step": 3200 |
| }, |
| { |
| "epoch": 14.266666666666667, |
| "grad_norm": 0.2257150411605835, |
| "learning_rate": 4.837437454708784e-05, |
| "loss": 0.0084, |
| "step": 3210 |
| }, |
| { |
| "epoch": 14.311111111111112, |
| "grad_norm": 0.20257611572742462, |
| "learning_rate": 4.809896802002662e-05, |
| "loss": 0.0091, |
| "step": 3220 |
| }, |
| { |
| "epoch": 14.355555555555556, |
| "grad_norm": 0.17713074386119843, |
| "learning_rate": 4.7823619241152854e-05, |
| "loss": 0.0086, |
| "step": 3230 |
| }, |
| { |
| "epoch": 14.4, |
| "grad_norm": 0.18521209061145782, |
| "learning_rate": 4.754833657481445e-05, |
| "loss": 0.0087, |
| "step": 3240 |
| }, |
| { |
| "epoch": 14.444444444444445, |
| "grad_norm": 0.14204886555671692, |
| "learning_rate": 4.7273128383351015e-05, |
| "loss": 0.0103, |
| "step": 3250 |
| }, |
| { |
| "epoch": 14.488888888888889, |
| "grad_norm": 0.28341996669769287, |
| "learning_rate": 4.699800302683981e-05, |
| "loss": 0.0084, |
| "step": 3260 |
| }, |
| { |
| "epoch": 14.533333333333333, |
| "grad_norm": 0.2390553206205368, |
| "learning_rate": 4.6722968862841806e-05, |
| "loss": 0.0125, |
| "step": 3270 |
| }, |
| { |
| "epoch": 14.577777777777778, |
| "grad_norm": 0.20710928738117218, |
| "learning_rate": 4.6448034246147754e-05, |
| "loss": 0.0094, |
| "step": 3280 |
| }, |
| { |
| "epoch": 14.622222222222222, |
| "grad_norm": 0.19245918095111847, |
| "learning_rate": 4.6173207528524476e-05, |
| "loss": 0.0119, |
| "step": 3290 |
| }, |
| { |
| "epoch": 14.666666666666666, |
| "grad_norm": 0.14937560260295868, |
| "learning_rate": 4.58984970584611e-05, |
| "loss": 0.0139, |
| "step": 3300 |
| }, |
| { |
| "epoch": 14.71111111111111, |
| "grad_norm": 0.14270950853824615, |
| "learning_rate": 4.562391118091544e-05, |
| "loss": 0.0082, |
| "step": 3310 |
| }, |
| { |
| "epoch": 14.755555555555556, |
| "grad_norm": 0.1969487965106964, |
| "learning_rate": 4.534945823706056e-05, |
| "loss": 0.0079, |
| "step": 3320 |
| }, |
| { |
| "epoch": 14.8, |
| "grad_norm": 0.22423358261585236, |
| "learning_rate": 4.507514656403137e-05, |
| "loss": 0.0112, |
| "step": 3330 |
| }, |
| { |
| "epoch": 14.844444444444445, |
| "grad_norm": 0.18679030239582062, |
| "learning_rate": 4.480098449467132e-05, |
| "loss": 0.0102, |
| "step": 3340 |
| }, |
| { |
| "epoch": 14.88888888888889, |
| "grad_norm": 0.28495559096336365, |
| "learning_rate": 4.452698035727929e-05, |
| "loss": 0.0116, |
| "step": 3350 |
| }, |
| { |
| "epoch": 14.933333333333334, |
| "grad_norm": 0.1513061821460724, |
| "learning_rate": 4.425314247535668e-05, |
| "loss": 0.0142, |
| "step": 3360 |
| }, |
| { |
| "epoch": 14.977777777777778, |
| "grad_norm": 0.22088178992271423, |
| "learning_rate": 4.3979479167354477e-05, |
| "loss": 0.0157, |
| "step": 3370 |
| }, |
| { |
| "epoch": 15.022222222222222, |
| "grad_norm": 0.19653438031673431, |
| "learning_rate": 4.370599874642055e-05, |
| "loss": 0.0096, |
| "step": 3380 |
| }, |
| { |
| "epoch": 15.066666666666666, |
| "grad_norm": 0.21249467134475708, |
| "learning_rate": 4.3432709520147205e-05, |
| "loss": 0.0108, |
| "step": 3390 |
| }, |
| { |
| "epoch": 15.11111111111111, |
| "grad_norm": 0.1530034840106964, |
| "learning_rate": 4.315961979031875e-05, |
| "loss": 0.0099, |
| "step": 3400 |
| }, |
| { |
| "epoch": 15.155555555555555, |
| "grad_norm": 0.25806236267089844, |
| "learning_rate": 4.2886737852659325e-05, |
| "loss": 0.0107, |
| "step": 3410 |
| }, |
| { |
| "epoch": 15.2, |
| "grad_norm": 0.18777354061603546, |
| "learning_rate": 4.261407199658093e-05, |
| "loss": 0.0095, |
| "step": 3420 |
| }, |
| { |
| "epoch": 15.244444444444444, |
| "grad_norm": 0.14671854674816132, |
| "learning_rate": 4.234163050493158e-05, |
| "loss": 0.0078, |
| "step": 3430 |
| }, |
| { |
| "epoch": 15.28888888888889, |
| "grad_norm": 0.1849425733089447, |
| "learning_rate": 4.2069421653743706e-05, |
| "loss": 0.0139, |
| "step": 3440 |
| }, |
| { |
| "epoch": 15.333333333333334, |
| "grad_norm": 0.18728889524936676, |
| "learning_rate": 4.179745371198276e-05, |
| "loss": 0.0104, |
| "step": 3450 |
| }, |
| { |
| "epoch": 15.377777777777778, |
| "grad_norm": 0.20069687068462372, |
| "learning_rate": 4.1525734941296026e-05, |
| "loss": 0.0103, |
| "step": 3460 |
| }, |
| { |
| "epoch": 15.422222222222222, |
| "grad_norm": 0.24841096997261047, |
| "learning_rate": 4.125427359576162e-05, |
| "loss": 0.0099, |
| "step": 3470 |
| }, |
| { |
| "epoch": 15.466666666666667, |
| "grad_norm": 0.18070349097251892, |
| "learning_rate": 4.0983077921637815e-05, |
| "loss": 0.0089, |
| "step": 3480 |
| }, |
| { |
| "epoch": 15.511111111111111, |
| "grad_norm": 0.22775743901729584, |
| "learning_rate": 4.07121561571125e-05, |
| "loss": 0.0151, |
| "step": 3490 |
| }, |
| { |
| "epoch": 15.555555555555555, |
| "grad_norm": 0.16708751022815704, |
| "learning_rate": 4.044151653205292e-05, |
| "loss": 0.0083, |
| "step": 3500 |
| }, |
| { |
| "epoch": 15.6, |
| "grad_norm": 0.18300403654575348, |
| "learning_rate": 4.0171167267755696e-05, |
| "loss": 0.014, |
| "step": 3510 |
| }, |
| { |
| "epoch": 15.644444444444444, |
| "grad_norm": 0.15873435139656067, |
| "learning_rate": 3.9901116576697083e-05, |
| "loss": 0.0126, |
| "step": 3520 |
| }, |
| { |
| "epoch": 15.688888888888888, |
| "grad_norm": 0.20646321773529053, |
| "learning_rate": 3.963137266228349e-05, |
| "loss": 0.0068, |
| "step": 3530 |
| }, |
| { |
| "epoch": 15.733333333333333, |
| "grad_norm": 0.2832168638706207, |
| "learning_rate": 3.93619437186023e-05, |
| "loss": 0.0146, |
| "step": 3540 |
| }, |
| { |
| "epoch": 15.777777777777779, |
| "grad_norm": 0.2857988774776459, |
| "learning_rate": 3.9092837930172884e-05, |
| "loss": 0.0091, |
| "step": 3550 |
| }, |
| { |
| "epoch": 15.822222222222223, |
| "grad_norm": 0.1261863112449646, |
| "learning_rate": 3.8824063471698105e-05, |
| "loss": 0.011, |
| "step": 3560 |
| }, |
| { |
| "epoch": 15.866666666666667, |
| "grad_norm": 0.15000951290130615, |
| "learning_rate": 3.855562850781589e-05, |
| "loss": 0.0154, |
| "step": 3570 |
| }, |
| { |
| "epoch": 15.911111111111111, |
| "grad_norm": 0.22501187026500702, |
| "learning_rate": 3.828754119285123e-05, |
| "loss": 0.0133, |
| "step": 3580 |
| }, |
| { |
| "epoch": 15.955555555555556, |
| "grad_norm": 0.21757589280605316, |
| "learning_rate": 3.801980967056851e-05, |
| "loss": 0.0103, |
| "step": 3590 |
| }, |
| { |
| "epoch": 16.0, |
| "grad_norm": 0.2032909393310547, |
| "learning_rate": 3.77524420739241e-05, |
| "loss": 0.0093, |
| "step": 3600 |
| }, |
| { |
| "epoch": 16.044444444444444, |
| "grad_norm": 0.18988990783691406, |
| "learning_rate": 3.748544652481927e-05, |
| "loss": 0.0096, |
| "step": 3610 |
| }, |
| { |
| "epoch": 16.08888888888889, |
| "grad_norm": 0.22926317155361176, |
| "learning_rate": 3.721883113385353e-05, |
| "loss": 0.0082, |
| "step": 3620 |
| }, |
| { |
| "epoch": 16.133333333333333, |
| "grad_norm": 0.16686010360717773, |
| "learning_rate": 3.695260400007819e-05, |
| "loss": 0.008, |
| "step": 3630 |
| }, |
| { |
| "epoch": 16.177777777777777, |
| "grad_norm": 0.23341767489910126, |
| "learning_rate": 3.6686773210750385e-05, |
| "loss": 0.0069, |
| "step": 3640 |
| }, |
| { |
| "epoch": 16.22222222222222, |
| "grad_norm": 0.15173080563545227, |
| "learning_rate": 3.642134684108737e-05, |
| "loss": 0.0088, |
| "step": 3650 |
| }, |
| { |
| "epoch": 16.266666666666666, |
| "grad_norm": 0.15917883813381195, |
| "learning_rate": 3.615633295402123e-05, |
| "loss": 0.0076, |
| "step": 3660 |
| }, |
| { |
| "epoch": 16.31111111111111, |
| "grad_norm": 0.2568972408771515, |
| "learning_rate": 3.5891739599953945e-05, |
| "loss": 0.0087, |
| "step": 3670 |
| }, |
| { |
| "epoch": 16.355555555555554, |
| "grad_norm": 0.17996510863304138, |
| "learning_rate": 3.5627574816512846e-05, |
| "loss": 0.0063, |
| "step": 3680 |
| }, |
| { |
| "epoch": 16.4, |
| "grad_norm": 0.17340408265590668, |
| "learning_rate": 3.536384662830648e-05, |
| "loss": 0.0077, |
| "step": 3690 |
| }, |
| { |
| "epoch": 16.444444444444443, |
| "grad_norm": 0.22471307218074799, |
| "learning_rate": 3.5100563046680764e-05, |
| "loss": 0.0085, |
| "step": 3700 |
| }, |
| { |
| "epoch": 16.488888888888887, |
| "grad_norm": 0.21206466853618622, |
| "learning_rate": 3.483773206947572e-05, |
| "loss": 0.0103, |
| "step": 3710 |
| }, |
| { |
| "epoch": 16.533333333333335, |
| "grad_norm": 0.17338991165161133, |
| "learning_rate": 3.457536168078247e-05, |
| "loss": 0.0094, |
| "step": 3720 |
| }, |
| { |
| "epoch": 16.57777777777778, |
| "grad_norm": 0.1438867598772049, |
| "learning_rate": 3.431345985070067e-05, |
| "loss": 0.008, |
| "step": 3730 |
| }, |
| { |
| "epoch": 16.622222222222224, |
| "grad_norm": 0.2796429395675659, |
| "learning_rate": 3.40520345350965e-05, |
| "loss": 0.0135, |
| "step": 3740 |
| }, |
| { |
| "epoch": 16.666666666666668, |
| "grad_norm": 0.19102472066879272, |
| "learning_rate": 3.379109367536089e-05, |
| "loss": 0.0084, |
| "step": 3750 |
| }, |
| { |
| "epoch": 16.711111111111112, |
| "grad_norm": 0.1319800764322281, |
| "learning_rate": 3.3530645198168295e-05, |
| "loss": 0.0061, |
| "step": 3760 |
| }, |
| { |
| "epoch": 16.755555555555556, |
| "grad_norm": 0.15621867775917053, |
| "learning_rate": 3.327069701523595e-05, |
| "loss": 0.0072, |
| "step": 3770 |
| }, |
| { |
| "epoch": 16.8, |
| "grad_norm": 0.12455292046070099, |
| "learning_rate": 3.301125702308353e-05, |
| "loss": 0.0104, |
| "step": 3780 |
| }, |
| { |
| "epoch": 16.844444444444445, |
| "grad_norm": 0.1910143941640854, |
| "learning_rate": 3.275233310279321e-05, |
| "loss": 0.0128, |
| "step": 3790 |
| }, |
| { |
| "epoch": 16.88888888888889, |
| "grad_norm": 0.1920808106660843, |
| "learning_rate": 3.249393311977037e-05, |
| "loss": 0.0139, |
| "step": 3800 |
| }, |
| { |
| "epoch": 16.933333333333334, |
| "grad_norm": 0.18581479787826538, |
| "learning_rate": 3.223606492350451e-05, |
| "loss": 0.0112, |
| "step": 3810 |
| }, |
| { |
| "epoch": 16.977777777777778, |
| "grad_norm": 0.20907874405384064, |
| "learning_rate": 3.197873634733096e-05, |
| "loss": 0.0097, |
| "step": 3820 |
| }, |
| { |
| "epoch": 17.022222222222222, |
| "grad_norm": 0.12354215979576111, |
| "learning_rate": 3.172195520819285e-05, |
| "loss": 0.0068, |
| "step": 3830 |
| }, |
| { |
| "epoch": 17.066666666666666, |
| "grad_norm": 0.1491164118051529, |
| "learning_rate": 3.146572930640362e-05, |
| "loss": 0.0093, |
| "step": 3840 |
| }, |
| { |
| "epoch": 17.11111111111111, |
| "grad_norm": 0.1822177916765213, |
| "learning_rate": 3.121006642541014e-05, |
| "loss": 0.0078, |
| "step": 3850 |
| }, |
| { |
| "epoch": 17.155555555555555, |
| "grad_norm": 0.24089230597019196, |
| "learning_rate": 3.095497433155626e-05, |
| "loss": 0.0107, |
| "step": 3860 |
| }, |
| { |
| "epoch": 17.2, |
| "grad_norm": 0.17053531110286713, |
| "learning_rate": 3.070046077384682e-05, |
| "loss": 0.0091, |
| "step": 3870 |
| }, |
| { |
| "epoch": 17.244444444444444, |
| "grad_norm": 0.1506698876619339, |
| "learning_rate": 3.0446533483712304e-05, |
| "loss": 0.0077, |
| "step": 3880 |
| }, |
| { |
| "epoch": 17.288888888888888, |
| "grad_norm": 0.1598861813545227, |
| "learning_rate": 3.0193200174774038e-05, |
| "loss": 0.0102, |
| "step": 3890 |
| }, |
| { |
| "epoch": 17.333333333333332, |
| "grad_norm": 0.18909277021884918, |
| "learning_rate": 2.994046854260974e-05, |
| "loss": 0.0089, |
| "step": 3900 |
| }, |
| { |
| "epoch": 17.377777777777776, |
| "grad_norm": 0.2107781022787094, |
| "learning_rate": 2.9688346264519866e-05, |
| "loss": 0.0066, |
| "step": 3910 |
| }, |
| { |
| "epoch": 17.42222222222222, |
| "grad_norm": 0.29426008462905884, |
| "learning_rate": 2.943684099929436e-05, |
| "loss": 0.0073, |
| "step": 3920 |
| }, |
| { |
| "epoch": 17.466666666666665, |
| "grad_norm": 0.21497812867164612, |
| "learning_rate": 2.918596038697995e-05, |
| "loss": 0.0089, |
| "step": 3930 |
| }, |
| { |
| "epoch": 17.511111111111113, |
| "grad_norm": 0.16140930354595184, |
| "learning_rate": 2.8935712048648112e-05, |
| "loss": 0.0097, |
| "step": 3940 |
| }, |
| { |
| "epoch": 17.555555555555557, |
| "grad_norm": 0.1837870329618454, |
| "learning_rate": 2.8686103586163626e-05, |
| "loss": 0.0118, |
| "step": 3950 |
| }, |
| { |
| "epoch": 17.6, |
| "grad_norm": 0.1896575689315796, |
| "learning_rate": 2.843714258195346e-05, |
| "loss": 0.0128, |
| "step": 3960 |
| }, |
| { |
| "epoch": 17.644444444444446, |
| "grad_norm": 0.2173689901828766, |
| "learning_rate": 2.8188836598776662e-05, |
| "loss": 0.0093, |
| "step": 3970 |
| }, |
| { |
| "epoch": 17.68888888888889, |
| "grad_norm": 0.15574978291988373, |
| "learning_rate": 2.7941193179494484e-05, |
| "loss": 0.0069, |
| "step": 3980 |
| }, |
| { |
| "epoch": 17.733333333333334, |
| "grad_norm": 0.19700747728347778, |
| "learning_rate": 2.7694219846841262e-05, |
| "loss": 0.0079, |
| "step": 3990 |
| }, |
| { |
| "epoch": 17.77777777777778, |
| "grad_norm": 0.1698417365550995, |
| "learning_rate": 2.7447924103195976e-05, |
| "loss": 0.0071, |
| "step": 4000 |
| }, |
| { |
| "epoch": 17.822222222222223, |
| "grad_norm": 0.17947453260421753, |
| "learning_rate": 2.7202313430354253e-05, |
| "loss": 0.0071, |
| "step": 4010 |
| }, |
| { |
| "epoch": 17.866666666666667, |
| "grad_norm": 0.14212481677532196, |
| "learning_rate": 2.695739528930111e-05, |
| "loss": 0.0056, |
| "step": 4020 |
| }, |
| { |
| "epoch": 17.91111111111111, |
| "grad_norm": 0.10642435401678085, |
| "learning_rate": 2.67131771199844e-05, |
| "loss": 0.0064, |
| "step": 4030 |
| }, |
| { |
| "epoch": 17.955555555555556, |
| "grad_norm": 0.29085785150527954, |
| "learning_rate": 2.6469666341088677e-05, |
| "loss": 0.0116, |
| "step": 4040 |
| }, |
| { |
| "epoch": 18.0, |
| "grad_norm": 0.16551056504249573, |
| "learning_rate": 2.6226870349809885e-05, |
| "loss": 0.0093, |
| "step": 4050 |
| }, |
| { |
| "epoch": 18.044444444444444, |
| "grad_norm": 0.15443584322929382, |
| "learning_rate": 2.5984796521630737e-05, |
| "loss": 0.01, |
| "step": 4060 |
| }, |
| { |
| "epoch": 18.08888888888889, |
| "grad_norm": 0.10544218122959137, |
| "learning_rate": 2.574345221009653e-05, |
| "loss": 0.0146, |
| "step": 4070 |
| }, |
| { |
| "epoch": 18.133333333333333, |
| "grad_norm": 0.19458246231079102, |
| "learning_rate": 2.5502844746591804e-05, |
| "loss": 0.0074, |
| "step": 4080 |
| }, |
| { |
| "epoch": 18.177777777777777, |
| "grad_norm": 0.12523303925991058, |
| "learning_rate": 2.526298144011775e-05, |
| "loss": 0.007, |
| "step": 4090 |
| }, |
| { |
| "epoch": 18.22222222222222, |
| "grad_norm": 0.17862561345100403, |
| "learning_rate": 2.5023869577070013e-05, |
| "loss": 0.0084, |
| "step": 4100 |
| }, |
| { |
| "epoch": 18.266666666666666, |
| "grad_norm": 0.15415054559707642, |
| "learning_rate": 2.478551642101743e-05, |
| "loss": 0.0075, |
| "step": 4110 |
| }, |
| { |
| "epoch": 18.31111111111111, |
| "grad_norm": 0.16622138023376465, |
| "learning_rate": 2.4547929212481435e-05, |
| "loss": 0.0072, |
| "step": 4120 |
| }, |
| { |
| "epoch": 18.355555555555554, |
| "grad_norm": 0.10168420523405075, |
| "learning_rate": 2.4311115168716013e-05, |
| "loss": 0.0058, |
| "step": 4130 |
| }, |
| { |
| "epoch": 18.4, |
| "grad_norm": 0.2314300835132599, |
| "learning_rate": 2.4075081483488494e-05, |
| "loss": 0.0102, |
| "step": 4140 |
| }, |
| { |
| "epoch": 18.444444444444443, |
| "grad_norm": 0.16795116662979126, |
| "learning_rate": 2.3839835326861104e-05, |
| "loss": 0.0059, |
| "step": 4150 |
| }, |
| { |
| "epoch": 18.488888888888887, |
| "grad_norm": 0.09173907339572906, |
| "learning_rate": 2.3605383844972966e-05, |
| "loss": 0.006, |
| "step": 4160 |
| }, |
| { |
| "epoch": 18.533333333333335, |
| "grad_norm": 0.1433173269033432, |
| "learning_rate": 2.3371734159823284e-05, |
| "loss": 0.0072, |
| "step": 4170 |
| }, |
| { |
| "epoch": 18.57777777777778, |
| "grad_norm": 0.2237769067287445, |
| "learning_rate": 2.3138893369054766e-05, |
| "loss": 0.007, |
| "step": 4180 |
| }, |
| { |
| "epoch": 18.622222222222224, |
| "grad_norm": 0.1855582743883133, |
| "learning_rate": 2.2906868545738102e-05, |
| "loss": 0.0074, |
| "step": 4190 |
| }, |
| { |
| "epoch": 18.666666666666668, |
| "grad_norm": 0.15935784578323364, |
| "learning_rate": 2.2675666738157186e-05, |
| "loss": 0.0056, |
| "step": 4200 |
| }, |
| { |
| "epoch": 18.711111111111112, |
| "grad_norm": 0.09136050939559937, |
| "learning_rate": 2.2445294969594844e-05, |
| "loss": 0.006, |
| "step": 4210 |
| }, |
| { |
| "epoch": 18.755555555555556, |
| "grad_norm": 0.13247030973434448, |
| "learning_rate": 2.22157602381196e-05, |
| "loss": 0.0068, |
| "step": 4220 |
| }, |
| { |
| "epoch": 18.8, |
| "grad_norm": 0.12561368942260742, |
| "learning_rate": 2.1987069516373098e-05, |
| "loss": 0.0061, |
| "step": 4230 |
| }, |
| { |
| "epoch": 18.844444444444445, |
| "grad_norm": 0.20298150181770325, |
| "learning_rate": 2.1759229751358217e-05, |
| "loss": 0.0062, |
| "step": 4240 |
| }, |
| { |
| "epoch": 18.88888888888889, |
| "grad_norm": 0.1889529973268509, |
| "learning_rate": 2.1532247864228084e-05, |
| "loss": 0.009, |
| "step": 4250 |
| }, |
| { |
| "epoch": 18.933333333333334, |
| "grad_norm": 0.12197969108819962, |
| "learning_rate": 2.1306130750075865e-05, |
| "loss": 0.006, |
| "step": 4260 |
| }, |
| { |
| "epoch": 18.977777777777778, |
| "grad_norm": 0.1984909474849701, |
| "learning_rate": 2.1080885277725236e-05, |
| "loss": 0.007, |
| "step": 4270 |
| }, |
| { |
| "epoch": 19.022222222222222, |
| "grad_norm": 0.18631795048713684, |
| "learning_rate": 2.085651828952175e-05, |
| "loss": 0.008, |
| "step": 4280 |
| }, |
| { |
| "epoch": 19.066666666666666, |
| "grad_norm": 0.2290038764476776, |
| "learning_rate": 2.063303660112506e-05, |
| "loss": 0.0147, |
| "step": 4290 |
| }, |
| { |
| "epoch": 19.11111111111111, |
| "grad_norm": 0.1833907514810562, |
| "learning_rate": 2.0410447001301753e-05, |
| "loss": 0.0059, |
| "step": 4300 |
| }, |
| { |
| "epoch": 19.155555555555555, |
| "grad_norm": 0.2564699053764343, |
| "learning_rate": 2.0188756251719203e-05, |
| "loss": 0.0076, |
| "step": 4310 |
| }, |
| { |
| "epoch": 19.2, |
| "grad_norm": 0.19490176439285278, |
| "learning_rate": 1.9967971086740195e-05, |
| "loss": 0.0056, |
| "step": 4320 |
| }, |
| { |
| "epoch": 19.244444444444444, |
| "grad_norm": 0.16607746481895447, |
| "learning_rate": 1.974809821321827e-05, |
| "loss": 0.0079, |
| "step": 4330 |
| }, |
| { |
| "epoch": 19.288888888888888, |
| "grad_norm": 0.10978394746780396, |
| "learning_rate": 1.9529144310294023e-05, |
| "loss": 0.0045, |
| "step": 4340 |
| }, |
| { |
| "epoch": 19.333333333333332, |
| "grad_norm": 0.08454827219247818, |
| "learning_rate": 1.9311116029192278e-05, |
| "loss": 0.0056, |
| "step": 4350 |
| }, |
| { |
| "epoch": 19.377777777777776, |
| "grad_norm": 0.22791458666324615, |
| "learning_rate": 1.909401999301993e-05, |
| "loss": 0.0094, |
| "step": 4360 |
| }, |
| { |
| "epoch": 19.42222222222222, |
| "grad_norm": 0.12988129258155823, |
| "learning_rate": 1.887786279656482e-05, |
| "loss": 0.0053, |
| "step": 4370 |
| }, |
| { |
| "epoch": 19.466666666666665, |
| "grad_norm": 0.1537473052740097, |
| "learning_rate": 1.8662651006095387e-05, |
| "loss": 0.0065, |
| "step": 4380 |
| }, |
| { |
| "epoch": 19.511111111111113, |
| "grad_norm": 0.12037132680416107, |
| "learning_rate": 1.8448391159161204e-05, |
| "loss": 0.0093, |
| "step": 4390 |
| }, |
| { |
| "epoch": 19.555555555555557, |
| "grad_norm": 0.10276351124048233, |
| "learning_rate": 1.8235089764394408e-05, |
| "loss": 0.004, |
| "step": 4400 |
| }, |
| { |
| "epoch": 19.6, |
| "grad_norm": 0.13345229625701904, |
| "learning_rate": 1.8022753301311935e-05, |
| "loss": 0.0049, |
| "step": 4410 |
| }, |
| { |
| "epoch": 19.644444444444446, |
| "grad_norm": 0.14785456657409668, |
| "learning_rate": 1.7811388220118707e-05, |
| "loss": 0.0055, |
| "step": 4420 |
| }, |
| { |
| "epoch": 19.68888888888889, |
| "grad_norm": 0.15009242296218872, |
| "learning_rate": 1.7601000941511757e-05, |
| "loss": 0.0042, |
| "step": 4430 |
| }, |
| { |
| "epoch": 19.733333333333334, |
| "grad_norm": 0.09384661912918091, |
| "learning_rate": 1.7391597856485083e-05, |
| "loss": 0.0061, |
| "step": 4440 |
| }, |
| { |
| "epoch": 19.77777777777778, |
| "grad_norm": 0.08937369287014008, |
| "learning_rate": 1.7183185326135543e-05, |
| "loss": 0.0068, |
| "step": 4450 |
| }, |
| { |
| "epoch": 19.822222222222223, |
| "grad_norm": 0.19312533736228943, |
| "learning_rate": 1.6975769681469705e-05, |
| "loss": 0.0086, |
| "step": 4460 |
| }, |
| { |
| "epoch": 19.866666666666667, |
| "grad_norm": 0.11678431928157806, |
| "learning_rate": 1.676935722321139e-05, |
| "loss": 0.0065, |
| "step": 4470 |
| }, |
| { |
| "epoch": 19.91111111111111, |
| "grad_norm": 0.1437138468027115, |
| "learning_rate": 1.6563954221610355e-05, |
| "loss": 0.0046, |
| "step": 4480 |
| }, |
| { |
| "epoch": 19.955555555555556, |
| "grad_norm": 0.15808168053627014, |
| "learning_rate": 1.6359566916251845e-05, |
| "loss": 0.0063, |
| "step": 4490 |
| }, |
| { |
| "epoch": 20.0, |
| "grad_norm": 0.10884489119052887, |
| "learning_rate": 1.615620151586697e-05, |
| "loss": 0.0101, |
| "step": 4500 |
| }, |
| { |
| "epoch": 20.044444444444444, |
| "grad_norm": 0.1854449212551117, |
| "learning_rate": 1.5953864198144135e-05, |
| "loss": 0.0057, |
| "step": 4510 |
| }, |
| { |
| "epoch": 20.08888888888889, |
| "grad_norm": 0.18170243501663208, |
| "learning_rate": 1.5752561109541447e-05, |
| "loss": 0.0085, |
| "step": 4520 |
| }, |
| { |
| "epoch": 20.133333333333333, |
| "grad_norm": 0.0976211279630661, |
| "learning_rate": 1.5552298365099882e-05, |
| "loss": 0.0074, |
| "step": 4530 |
| }, |
| { |
| "epoch": 20.177777777777777, |
| "grad_norm": 0.11999211460351944, |
| "learning_rate": 1.5353082048257596e-05, |
| "loss": 0.0044, |
| "step": 4540 |
| }, |
| { |
| "epoch": 20.22222222222222, |
| "grad_norm": 0.20437392592430115, |
| "learning_rate": 1.5154918210665148e-05, |
| "loss": 0.0088, |
| "step": 4550 |
| }, |
| { |
| "epoch": 20.266666666666666, |
| "grad_norm": 0.12560054659843445, |
| "learning_rate": 1.4957812872001614e-05, |
| "loss": 0.0056, |
| "step": 4560 |
| }, |
| { |
| "epoch": 20.31111111111111, |
| "grad_norm": 0.23118850588798523, |
| "learning_rate": 1.4761772019791748e-05, |
| "loss": 0.0065, |
| "step": 4570 |
| }, |
| { |
| "epoch": 20.355555555555554, |
| "grad_norm": 0.10507191717624664, |
| "learning_rate": 1.4566801609224096e-05, |
| "loss": 0.007, |
| "step": 4580 |
| }, |
| { |
| "epoch": 20.4, |
| "grad_norm": 0.11237312853336334, |
| "learning_rate": 1.4372907562970079e-05, |
| "loss": 0.0051, |
| "step": 4590 |
| }, |
| { |
| "epoch": 20.444444444444443, |
| "grad_norm": 0.10997062176465988, |
| "learning_rate": 1.4180095771004154e-05, |
| "loss": 0.0044, |
| "step": 4600 |
| }, |
| { |
| "epoch": 20.488888888888887, |
| "grad_norm": 0.14824159443378448, |
| "learning_rate": 1.3988372090424773e-05, |
| "loss": 0.0093, |
| "step": 4610 |
| }, |
| { |
| "epoch": 20.533333333333335, |
| "grad_norm": 0.13127201795578003, |
| "learning_rate": 1.3797742345276521e-05, |
| "loss": 0.0106, |
| "step": 4620 |
| }, |
| { |
| "epoch": 20.57777777777778, |
| "grad_norm": 0.1299627423286438, |
| "learning_rate": 1.3608212326373249e-05, |
| "loss": 0.0049, |
| "step": 4630 |
| }, |
| { |
| "epoch": 20.622222222222224, |
| "grad_norm": 0.11514117568731308, |
| "learning_rate": 1.3419787791122062e-05, |
| "loss": 0.0058, |
| "step": 4640 |
| }, |
| { |
| "epoch": 20.666666666666668, |
| "grad_norm": 0.22919079661369324, |
| "learning_rate": 1.323247446334847e-05, |
| "loss": 0.0153, |
| "step": 4650 |
| }, |
| { |
| "epoch": 20.711111111111112, |
| "grad_norm": 0.21730785071849823, |
| "learning_rate": 1.3046278033122577e-05, |
| "loss": 0.0103, |
| "step": 4660 |
| }, |
| { |
| "epoch": 20.755555555555556, |
| "grad_norm": 0.14439047873020172, |
| "learning_rate": 1.286120415658611e-05, |
| "loss": 0.0064, |
| "step": 4670 |
| }, |
| { |
| "epoch": 20.8, |
| "grad_norm": 0.11379679292440414, |
| "learning_rate": 1.2677258455780683e-05, |
| "loss": 0.0042, |
| "step": 4680 |
| }, |
| { |
| "epoch": 20.844444444444445, |
| "grad_norm": 0.09024758636951447, |
| "learning_rate": 1.2494446518477022e-05, |
| "loss": 0.0106, |
| "step": 4690 |
| }, |
| { |
| "epoch": 20.88888888888889, |
| "grad_norm": 0.1456061750650406, |
| "learning_rate": 1.2312773898005175e-05, |
| "loss": 0.0089, |
| "step": 4700 |
| }, |
| { |
| "epoch": 20.933333333333334, |
| "grad_norm": 0.16650421917438507, |
| "learning_rate": 1.2132246113085822e-05, |
| "loss": 0.0072, |
| "step": 4710 |
| }, |
| { |
| "epoch": 20.977777777777778, |
| "grad_norm": 0.18011432886123657, |
| "learning_rate": 1.1952868647662696e-05, |
| "loss": 0.0047, |
| "step": 4720 |
| }, |
| { |
| "epoch": 21.022222222222222, |
| "grad_norm": 0.0823548436164856, |
| "learning_rate": 1.1774646950735913e-05, |
| "loss": 0.0039, |
| "step": 4730 |
| }, |
| { |
| "epoch": 21.066666666666666, |
| "grad_norm": 0.0739578828215599, |
| "learning_rate": 1.1597586436196473e-05, |
| "loss": 0.0092, |
| "step": 4740 |
| }, |
| { |
| "epoch": 21.11111111111111, |
| "grad_norm": 0.10134830325841904, |
| "learning_rate": 1.1421692482661856e-05, |
| "loss": 0.0042, |
| "step": 4750 |
| }, |
| { |
| "epoch": 21.155555555555555, |
| "grad_norm": 0.13486601412296295, |
| "learning_rate": 1.124697043331256e-05, |
| "loss": 0.006, |
| "step": 4760 |
| }, |
| { |
| "epoch": 21.2, |
| "grad_norm": 0.21013858914375305, |
| "learning_rate": 1.107342559572977e-05, |
| "loss": 0.0067, |
| "step": 4770 |
| }, |
| { |
| "epoch": 21.244444444444444, |
| "grad_norm": 0.18283604085445404, |
| "learning_rate": 1.090106324173426e-05, |
| "loss": 0.007, |
| "step": 4780 |
| }, |
| { |
| "epoch": 21.288888888888888, |
| "grad_norm": 0.23748941719532013, |
| "learning_rate": 1.0729888607226113e-05, |
| "loss": 0.0092, |
| "step": 4790 |
| }, |
| { |
| "epoch": 21.333333333333332, |
| "grad_norm": 0.1013626754283905, |
| "learning_rate": 1.0559906892025745e-05, |
| "loss": 0.0049, |
| "step": 4800 |
| }, |
| { |
| "epoch": 21.377777777777776, |
| "grad_norm": 0.09716016054153442, |
| "learning_rate": 1.0391123259715906e-05, |
| "loss": 0.005, |
| "step": 4810 |
| }, |
| { |
| "epoch": 21.42222222222222, |
| "grad_norm": 0.12395866215229034, |
| "learning_rate": 1.0223542837484839e-05, |
| "loss": 0.0049, |
| "step": 4820 |
| }, |
| { |
| "epoch": 21.466666666666665, |
| "grad_norm": 0.1267317533493042, |
| "learning_rate": 1.0057170715970559e-05, |
| "loss": 0.0055, |
| "step": 4830 |
| }, |
| { |
| "epoch": 21.511111111111113, |
| "grad_norm": 0.08411135524511337, |
| "learning_rate": 9.892011949106172e-06, |
| "loss": 0.0057, |
| "step": 4840 |
| }, |
| { |
| "epoch": 21.555555555555557, |
| "grad_norm": 0.10536860674619675, |
| "learning_rate": 9.728071553966339e-06, |
| "loss": 0.0037, |
| "step": 4850 |
| }, |
| { |
| "epoch": 21.6, |
| "grad_norm": 0.14186663925647736, |
| "learning_rate": 9.56535451061496e-06, |
| "loss": 0.0045, |
| "step": 4860 |
| }, |
| { |
| "epoch": 21.644444444444446, |
| "grad_norm": 0.11132841557264328, |
| "learning_rate": 9.403865761953779e-06, |
| "loss": 0.0061, |
| "step": 4870 |
| }, |
| { |
| "epoch": 21.68888888888889, |
| "grad_norm": 0.08628375083208084, |
| "learning_rate": 9.243610213572285e-06, |
| "loss": 0.0048, |
| "step": 4880 |
| }, |
| { |
| "epoch": 21.733333333333334, |
| "grad_norm": 0.10845037549734116, |
| "learning_rate": 9.084592733598735e-06, |
| "loss": 0.0049, |
| "step": 4890 |
| }, |
| { |
| "epoch": 21.77777777777778, |
| "grad_norm": 0.1388201117515564, |
| "learning_rate": 8.92681815255219e-06, |
| "loss": 0.005, |
| "step": 4900 |
| }, |
| { |
| "epoch": 21.822222222222223, |
| "grad_norm": 0.13397830724716187, |
| "learning_rate": 8.770291263195819e-06, |
| "loss": 0.006, |
| "step": 4910 |
| }, |
| { |
| "epoch": 21.866666666666667, |
| "grad_norm": 0.1188044473528862, |
| "learning_rate": 8.615016820391342e-06, |
| "loss": 0.007, |
| "step": 4920 |
| }, |
| { |
| "epoch": 21.91111111111111, |
| "grad_norm": 0.13798382878303528, |
| "learning_rate": 8.460999540954517e-06, |
| "loss": 0.0054, |
| "step": 4930 |
| }, |
| { |
| "epoch": 21.955555555555556, |
| "grad_norm": 0.12165193259716034, |
| "learning_rate": 8.308244103511909e-06, |
| "loss": 0.006, |
| "step": 4940 |
| }, |
| { |
| "epoch": 22.0, |
| "grad_norm": 0.12081032246351242, |
| "learning_rate": 8.156755148358764e-06, |
| "loss": 0.0045, |
| "step": 4950 |
| }, |
| { |
| "epoch": 22.044444444444444, |
| "grad_norm": 0.24020034074783325, |
| "learning_rate": 8.00653727731801e-06, |
| "loss": 0.0062, |
| "step": 4960 |
| }, |
| { |
| "epoch": 22.08888888888889, |
| "grad_norm": 0.09440013021230698, |
| "learning_rate": 7.857595053600513e-06, |
| "loss": 0.0042, |
| "step": 4970 |
| }, |
| { |
| "epoch": 22.133333333333333, |
| "grad_norm": 0.07173329591751099, |
| "learning_rate": 7.709933001666431e-06, |
| "loss": 0.0033, |
| "step": 4980 |
| }, |
| { |
| "epoch": 22.177777777777777, |
| "grad_norm": 0.08374012261629105, |
| "learning_rate": 7.56355560708778e-06, |
| "loss": 0.006, |
| "step": 4990 |
| }, |
| { |
| "epoch": 22.22222222222222, |
| "grad_norm": 0.1737222671508789, |
| "learning_rate": 7.418467316412158e-06, |
| "loss": 0.0075, |
| "step": 5000 |
| }, |
| { |
| "epoch": 22.266666666666666, |
| "grad_norm": 0.15025857090950012, |
| "learning_rate": 7.2746725370277435e-06, |
| "loss": 0.0054, |
| "step": 5010 |
| }, |
| { |
| "epoch": 22.31111111111111, |
| "grad_norm": 0.21231794357299805, |
| "learning_rate": 7.132175637029293e-06, |
| "loss": 0.0045, |
| "step": 5020 |
| }, |
| { |
| "epoch": 22.355555555555554, |
| "grad_norm": 0.1831008791923523, |
| "learning_rate": 6.9909809450855345e-06, |
| "loss": 0.0065, |
| "step": 5030 |
| }, |
| { |
| "epoch": 22.4, |
| "grad_norm": 0.12319664657115936, |
| "learning_rate": 6.851092750307686e-06, |
| "loss": 0.0047, |
| "step": 5040 |
| }, |
| { |
| "epoch": 22.444444444444443, |
| "grad_norm": 0.12710854411125183, |
| "learning_rate": 6.712515302119077e-06, |
| "loss": 0.0042, |
| "step": 5050 |
| }, |
| { |
| "epoch": 22.488888888888887, |
| "grad_norm": 0.11153990775346756, |
| "learning_rate": 6.575252810126143e-06, |
| "loss": 0.0037, |
| "step": 5060 |
| }, |
| { |
| "epoch": 22.533333333333335, |
| "grad_norm": 0.0981452539563179, |
| "learning_rate": 6.439309443990532e-06, |
| "loss": 0.0053, |
| "step": 5070 |
| }, |
| { |
| "epoch": 22.57777777777778, |
| "grad_norm": 0.11913459002971649, |
| "learning_rate": 6.304689333302416e-06, |
| "loss": 0.0057, |
| "step": 5080 |
| }, |
| { |
| "epoch": 22.622222222222224, |
| "grad_norm": 0.11172778904438019, |
| "learning_rate": 6.171396567455051e-06, |
| "loss": 0.0071, |
| "step": 5090 |
| }, |
| { |
| "epoch": 22.666666666666668, |
| "grad_norm": 0.153991237282753, |
| "learning_rate": 6.039435195520604e-06, |
| "loss": 0.0075, |
| "step": 5100 |
| }, |
| { |
| "epoch": 22.711111111111112, |
| "grad_norm": 0.10709087550640106, |
| "learning_rate": 5.908809226127054e-06, |
| "loss": 0.0048, |
| "step": 5110 |
| }, |
| { |
| "epoch": 22.755555555555556, |
| "grad_norm": 0.09339303523302078, |
| "learning_rate": 5.779522627336537e-06, |
| "loss": 0.0061, |
| "step": 5120 |
| }, |
| { |
| "epoch": 22.8, |
| "grad_norm": 0.09950139373540878, |
| "learning_rate": 5.651579326524709e-06, |
| "loss": 0.004, |
| "step": 5130 |
| }, |
| { |
| "epoch": 22.844444444444445, |
| "grad_norm": 0.09303288161754608, |
| "learning_rate": 5.524983210261481e-06, |
| "loss": 0.006, |
| "step": 5140 |
| }, |
| { |
| "epoch": 22.88888888888889, |
| "grad_norm": 0.20597153902053833, |
| "learning_rate": 5.399738124192988e-06, |
| "loss": 0.007, |
| "step": 5150 |
| }, |
| { |
| "epoch": 22.933333333333334, |
| "grad_norm": 0.1472003161907196, |
| "learning_rate": 5.2758478729247164e-06, |
| "loss": 0.0054, |
| "step": 5160 |
| }, |
| { |
| "epoch": 22.977777777777778, |
| "grad_norm": 0.11046073585748672, |
| "learning_rate": 5.153316219905946e-06, |
| "loss": 0.0047, |
| "step": 5170 |
| }, |
| { |
| "epoch": 23.022222222222222, |
| "grad_norm": 0.09678430110216141, |
| "learning_rate": 5.032146887315448e-06, |
| "loss": 0.0052, |
| "step": 5180 |
| }, |
| { |
| "epoch": 23.066666666666666, |
| "grad_norm": 0.0868353396654129, |
| "learning_rate": 4.91234355594839e-06, |
| "loss": 0.0042, |
| "step": 5190 |
| }, |
| { |
| "epoch": 23.11111111111111, |
| "grad_norm": 0.10893691331148148, |
| "learning_rate": 4.7939098651045235e-06, |
| "loss": 0.0044, |
| "step": 5200 |
| }, |
| { |
| "epoch": 23.155555555555555, |
| "grad_norm": 0.08827574551105499, |
| "learning_rate": 4.67684941247768e-06, |
| "loss": 0.0057, |
| "step": 5210 |
| }, |
| { |
| "epoch": 23.2, |
| "grad_norm": 0.1452520340681076, |
| "learning_rate": 4.5611657540464036e-06, |
| "loss": 0.0031, |
| "step": 5220 |
| }, |
| { |
| "epoch": 23.244444444444444, |
| "grad_norm": 0.10263426601886749, |
| "learning_rate": 4.446862403965984e-06, |
| "loss": 0.0055, |
| "step": 5230 |
| }, |
| { |
| "epoch": 23.288888888888888, |
| "grad_norm": 0.0701296329498291, |
| "learning_rate": 4.333942834461702e-06, |
| "loss": 0.0036, |
| "step": 5240 |
| }, |
| { |
| "epoch": 23.333333333333332, |
| "grad_norm": 0.12015524506568909, |
| "learning_rate": 4.222410475723326e-06, |
| "loss": 0.0037, |
| "step": 5250 |
| }, |
| { |
| "epoch": 23.377777777777776, |
| "grad_norm": 0.07371048629283905, |
| "learning_rate": 4.112268715800943e-06, |
| "loss": 0.0053, |
| "step": 5260 |
| }, |
| { |
| "epoch": 23.42222222222222, |
| "grad_norm": 0.07205082476139069, |
| "learning_rate": 4.003520900502028e-06, |
| "loss": 0.0052, |
| "step": 5270 |
| }, |
| { |
| "epoch": 23.466666666666665, |
| "grad_norm": 0.09509725123643875, |
| "learning_rate": 3.8961703332898e-06, |
| "loss": 0.0031, |
| "step": 5280 |
| }, |
| { |
| "epoch": 23.511111111111113, |
| "grad_norm": 0.12674923241138458, |
| "learning_rate": 3.790220275182854e-06, |
| "loss": 0.0041, |
| "step": 5290 |
| }, |
| { |
| "epoch": 23.555555555555557, |
| "grad_norm": 0.07842233031988144, |
| "learning_rate": 3.685673944656176e-06, |
| "loss": 0.0069, |
| "step": 5300 |
| }, |
| { |
| "epoch": 23.6, |
| "grad_norm": 0.21345362067222595, |
| "learning_rate": 3.582534517543268e-06, |
| "loss": 0.0133, |
| "step": 5310 |
| }, |
| { |
| "epoch": 23.644444444444446, |
| "grad_norm": 0.09635423123836517, |
| "learning_rate": 3.4808051269397512e-06, |
| "loss": 0.0049, |
| "step": 5320 |
| }, |
| { |
| "epoch": 23.68888888888889, |
| "grad_norm": 0.08894622325897217, |
| "learning_rate": 3.380488863108183e-06, |
| "loss": 0.0068, |
| "step": 5330 |
| }, |
| { |
| "epoch": 23.733333333333334, |
| "grad_norm": 0.07938189059495926, |
| "learning_rate": 3.2815887733841365e-06, |
| "loss": 0.0045, |
| "step": 5340 |
| }, |
| { |
| "epoch": 23.77777777777778, |
| "grad_norm": 0.10738229751586914, |
| "learning_rate": 3.1841078620836683e-06, |
| "loss": 0.0121, |
| "step": 5350 |
| }, |
| { |
| "epoch": 23.822222222222223, |
| "grad_norm": 0.06322488188743591, |
| "learning_rate": 3.0880490904120874e-06, |
| "loss": 0.006, |
| "step": 5360 |
| }, |
| { |
| "epoch": 23.866666666666667, |
| "grad_norm": 0.11578896641731262, |
| "learning_rate": 2.9934153763739205e-06, |
| "loss": 0.0032, |
| "step": 5370 |
| }, |
| { |
| "epoch": 23.91111111111111, |
| "grad_norm": 0.05660170689225197, |
| "learning_rate": 2.9002095946843277e-06, |
| "loss": 0.0056, |
| "step": 5380 |
| }, |
| { |
| "epoch": 23.955555555555556, |
| "grad_norm": 0.11994928866624832, |
| "learning_rate": 2.8084345766817676e-06, |
| "loss": 0.0071, |
| "step": 5390 |
| }, |
| { |
| "epoch": 24.0, |
| "grad_norm": 0.08521348983049393, |
| "learning_rate": 2.718093110241976e-06, |
| "loss": 0.0057, |
| "step": 5400 |
| }, |
| { |
| "epoch": 24.044444444444444, |
| "grad_norm": 0.07565505802631378, |
| "learning_rate": 2.6291879396933004e-06, |
| "loss": 0.0047, |
| "step": 5410 |
| }, |
| { |
| "epoch": 24.08888888888889, |
| "grad_norm": 0.08669985085725784, |
| "learning_rate": 2.541721765733318e-06, |
| "loss": 0.0068, |
| "step": 5420 |
| }, |
| { |
| "epoch": 24.133333333333333, |
| "grad_norm": 0.062804214656353, |
| "learning_rate": 2.455697245346783e-06, |
| "loss": 0.0039, |
| "step": 5430 |
| }, |
| { |
| "epoch": 24.177777777777777, |
| "grad_norm": 0.09471374750137329, |
| "learning_rate": 2.371116991724953e-06, |
| "loss": 0.0055, |
| "step": 5440 |
| }, |
| { |
| "epoch": 24.22222222222222, |
| "grad_norm": 0.09806012362241745, |
| "learning_rate": 2.2879835741861586e-06, |
| "loss": 0.0048, |
| "step": 5450 |
| }, |
| { |
| "epoch": 24.266666666666666, |
| "grad_norm": 0.14481781423091888, |
| "learning_rate": 2.206299518097804e-06, |
| "loss": 0.0035, |
| "step": 5460 |
| }, |
| { |
| "epoch": 24.31111111111111, |
| "grad_norm": 0.23285327851772308, |
| "learning_rate": 2.1260673047996227e-06, |
| "loss": 0.0065, |
| "step": 5470 |
| }, |
| { |
| "epoch": 24.355555555555554, |
| "grad_norm": 0.07071023434400558, |
| "learning_rate": 2.047289371528299e-06, |
| "loss": 0.0032, |
| "step": 5480 |
| }, |
| { |
| "epoch": 24.4, |
| "grad_norm": 0.07084526866674423, |
| "learning_rate": 1.96996811134344e-06, |
| "loss": 0.0036, |
| "step": 5490 |
| }, |
| { |
| "epoch": 24.444444444444443, |
| "grad_norm": 0.08298473805189133, |
| "learning_rate": 1.8941058730549132e-06, |
| "loss": 0.0038, |
| "step": 5500 |
| }, |
| { |
| "epoch": 24.488888888888887, |
| "grad_norm": 0.09886433184146881, |
| "learning_rate": 1.8197049611514194e-06, |
| "loss": 0.0052, |
| "step": 5510 |
| }, |
| { |
| "epoch": 24.533333333333335, |
| "grad_norm": 0.08648228645324707, |
| "learning_rate": 1.7467676357305561e-06, |
| "loss": 0.0039, |
| "step": 5520 |
| }, |
| { |
| "epoch": 24.57777777777778, |
| "grad_norm": 0.17427398264408112, |
| "learning_rate": 1.6752961124301415e-06, |
| "loss": 0.0043, |
| "step": 5530 |
| }, |
| { |
| "epoch": 24.622222222222224, |
| "grad_norm": 0.16165824234485626, |
| "learning_rate": 1.6052925623609049e-06, |
| "loss": 0.0042, |
| "step": 5540 |
| }, |
| { |
| "epoch": 24.666666666666668, |
| "grad_norm": 0.2243923842906952, |
| "learning_rate": 1.5367591120405256e-06, |
| "loss": 0.0044, |
| "step": 5550 |
| }, |
| { |
| "epoch": 24.711111111111112, |
| "grad_norm": 0.1862805336713791, |
| "learning_rate": 1.4696978433290653e-06, |
| "loss": 0.0044, |
| "step": 5560 |
| }, |
| { |
| "epoch": 24.755555555555556, |
| "grad_norm": 0.07909306138753891, |
| "learning_rate": 1.4041107933656928e-06, |
| "loss": 0.0047, |
| "step": 5570 |
| }, |
| { |
| "epoch": 24.8, |
| "grad_norm": 0.10913144052028656, |
| "learning_rate": 1.339999954506821e-06, |
| "loss": 0.0034, |
| "step": 5580 |
| }, |
| { |
| "epoch": 24.844444444444445, |
| "grad_norm": 0.07438622415065765, |
| "learning_rate": 1.2773672742655784e-06, |
| "loss": 0.0043, |
| "step": 5590 |
| }, |
| { |
| "epoch": 24.88888888888889, |
| "grad_norm": 0.09885593503713608, |
| "learning_rate": 1.2162146552526399e-06, |
| "loss": 0.0051, |
| "step": 5600 |
| }, |
| { |
| "epoch": 24.933333333333334, |
| "grad_norm": 0.10799284279346466, |
| "learning_rate": 1.1565439551184664e-06, |
| "loss": 0.0046, |
| "step": 5610 |
| }, |
| { |
| "epoch": 24.977777777777778, |
| "grad_norm": 0.10602659732103348, |
| "learning_rate": 1.0983569864968346e-06, |
| "loss": 0.0071, |
| "step": 5620 |
| }, |
| { |
| "epoch": 25.022222222222222, |
| "grad_norm": 0.12287070602178574, |
| "learning_rate": 1.0416555169497688e-06, |
| "loss": 0.0028, |
| "step": 5630 |
| }, |
| { |
| "epoch": 25.066666666666666, |
| "grad_norm": 0.1970548778772354, |
| "learning_rate": 9.864412689139123e-07, |
| "loss": 0.005, |
| "step": 5640 |
| }, |
| { |
| "epoch": 25.11111111111111, |
| "grad_norm": 0.10620978474617004, |
| "learning_rate": 9.327159196481138e-07, |
| "loss": 0.0041, |
| "step": 5650 |
| }, |
| { |
| "epoch": 25.155555555555555, |
| "grad_norm": 0.09525087475776672, |
| "learning_rate": 8.804811011825398e-07, |
| "loss": 0.0041, |
| "step": 5660 |
| }, |
| { |
| "epoch": 25.2, |
| "grad_norm": 0.16586178541183472, |
| "learning_rate": 8.297384002690866e-07, |
| "loss": 0.0032, |
| "step": 5670 |
| }, |
| { |
| "epoch": 25.244444444444444, |
| "grad_norm": 0.09201638400554657, |
| "learning_rate": 7.804893583331696e-07, |
| "loss": 0.0084, |
| "step": 5680 |
| }, |
| { |
| "epoch": 25.288888888888888, |
| "grad_norm": 0.09468468278646469, |
| "learning_rate": 7.32735471426893e-07, |
| "loss": 0.0029, |
| "step": 5690 |
| }, |
| { |
| "epoch": 25.333333333333332, |
| "grad_norm": 0.21121849119663239, |
| "learning_rate": 6.864781901836259e-07, |
| "loss": 0.0063, |
| "step": 5700 |
| }, |
| { |
| "epoch": 25.377777777777776, |
| "grad_norm": 0.06916485726833344, |
| "learning_rate": 6.417189197739093e-07, |
| "loss": 0.0058, |
| "step": 5710 |
| }, |
| { |
| "epoch": 25.42222222222222, |
| "grad_norm": 0.11040303111076355, |
| "learning_rate": 5.984590198627849e-07, |
| "loss": 0.0035, |
| "step": 5720 |
| }, |
| { |
| "epoch": 25.466666666666665, |
| "grad_norm": 0.10241199284791946, |
| "learning_rate": 5.566998045685112e-07, |
| "loss": 0.0053, |
| "step": 5730 |
| }, |
| { |
| "epoch": 25.511111111111113, |
| "grad_norm": 0.1048598363995552, |
| "learning_rate": 5.164425424226016e-07, |
| "loss": 0.0031, |
| "step": 5740 |
| }, |
| { |
| "epoch": 25.555555555555557, |
| "grad_norm": 0.06997186690568924, |
| "learning_rate": 4.776884563313266e-07, |
| "loss": 0.0037, |
| "step": 5750 |
| }, |
| { |
| "epoch": 25.6, |
| "grad_norm": 0.06315027177333832, |
| "learning_rate": 4.404387235385443e-07, |
| "loss": 0.0038, |
| "step": 5760 |
| }, |
| { |
| "epoch": 25.644444444444446, |
| "grad_norm": 0.07675505429506302, |
| "learning_rate": 4.0469447558995065e-07, |
| "loss": 0.003, |
| "step": 5770 |
| }, |
| { |
| "epoch": 25.68888888888889, |
| "grad_norm": 0.08786843717098236, |
| "learning_rate": 3.7045679829870175e-07, |
| "loss": 0.0043, |
| "step": 5780 |
| }, |
| { |
| "epoch": 25.733333333333334, |
| "grad_norm": 0.09154848009347916, |
| "learning_rate": 3.377267317124233e-07, |
| "loss": 0.0042, |
| "step": 5790 |
| }, |
| { |
| "epoch": 25.77777777777778, |
| "grad_norm": 0.07057943940162659, |
| "learning_rate": 3.0650527008162513e-07, |
| "loss": 0.0039, |
| "step": 5800 |
| }, |
| { |
| "epoch": 25.822222222222223, |
| "grad_norm": 0.21723133325576782, |
| "learning_rate": 2.767933618295082e-07, |
| "loss": 0.008, |
| "step": 5810 |
| }, |
| { |
| "epoch": 25.866666666666667, |
| "grad_norm": 0.08126044273376465, |
| "learning_rate": 2.485919095231326e-07, |
| "loss": 0.008, |
| "step": 5820 |
| }, |
| { |
| "epoch": 25.91111111111111, |
| "grad_norm": 0.12574568390846252, |
| "learning_rate": 2.219017698460002e-07, |
| "loss": 0.0044, |
| "step": 5830 |
| }, |
| { |
| "epoch": 25.955555555555556, |
| "grad_norm": 0.13723288476467133, |
| "learning_rate": 1.9672375357206452e-07, |
| "loss": 0.0039, |
| "step": 5840 |
| }, |
| { |
| "epoch": 26.0, |
| "grad_norm": 0.08475471287965775, |
| "learning_rate": 1.73058625541056e-07, |
| "loss": 0.004, |
| "step": 5850 |
| }, |
| { |
| "epoch": 26.044444444444444, |
| "grad_norm": 0.07984427362680435, |
| "learning_rate": 1.5090710463527836e-07, |
| "loss": 0.0052, |
| "step": 5860 |
| }, |
| { |
| "epoch": 26.08888888888889, |
| "grad_norm": 0.15684092044830322, |
| "learning_rate": 1.3026986375776485e-07, |
| "loss": 0.0036, |
| "step": 5870 |
| }, |
| { |
| "epoch": 26.133333333333333, |
| "grad_norm": 0.049164798110723495, |
| "learning_rate": 1.1114752981183917e-07, |
| "loss": 0.0053, |
| "step": 5880 |
| }, |
| { |
| "epoch": 26.177777777777777, |
| "grad_norm": 0.14477813243865967, |
| "learning_rate": 9.354068368204739e-08, |
| "loss": 0.0041, |
| "step": 5890 |
| }, |
| { |
| "epoch": 26.22222222222222, |
| "grad_norm": 0.18416763842105865, |
| "learning_rate": 7.744986021656076e-08, |
| "loss": 0.0039, |
| "step": 5900 |
| }, |
| { |
| "epoch": 26.266666666666666, |
| "grad_norm": 0.14764051139354706, |
| "learning_rate": 6.287554821087783e-08, |
| "loss": 0.0054, |
| "step": 5910 |
| }, |
| { |
| "epoch": 26.31111111111111, |
| "grad_norm": 0.17087344825267792, |
| "learning_rate": 4.981819039300284e-08, |
| "loss": 0.0049, |
| "step": 5920 |
| }, |
| { |
| "epoch": 26.355555555555554, |
| "grad_norm": 0.09899858385324478, |
| "learning_rate": 3.827818341000655e-08, |
| "loss": 0.005, |
| "step": 5930 |
| }, |
| { |
| "epoch": 26.4, |
| "grad_norm": 0.15714667737483978, |
| "learning_rate": 2.8255877815946963e-08, |
| "loss": 0.0062, |
| "step": 5940 |
| }, |
| { |
| "epoch": 26.444444444444443, |
| "grad_norm": 0.09364058822393417, |
| "learning_rate": 1.9751578061244504e-08, |
| "loss": 0.0032, |
| "step": 5950 |
| }, |
| { |
| "epoch": 26.488888888888887, |
| "grad_norm": 0.07059767097234726, |
| "learning_rate": 1.2765542483417214e-08, |
| "loss": 0.0042, |
| "step": 5960 |
| }, |
| { |
| "epoch": 26.533333333333335, |
| "grad_norm": 0.09034106135368347, |
| "learning_rate": 7.2979832992592365e-09, |
| "loss": 0.0042, |
| "step": 5970 |
| }, |
| { |
| "epoch": 26.57777777777778, |
| "grad_norm": 0.11712371557950974, |
| "learning_rate": 3.349066598362649e-09, |
| "loss": 0.0027, |
| "step": 5980 |
| }, |
| { |
| "epoch": 26.622222222222224, |
| "grad_norm": 0.2007165253162384, |
| "learning_rate": 9.189123380826114e-10, |
| "loss": 0.005, |
| "step": 5990 |
| }, |
| { |
| "epoch": 26.666666666666668, |
| "grad_norm": 0.1573939174413681, |
| "learning_rate": 7.594339912486703e-12, |
| "loss": 0.0062, |
| "step": 6000 |
| }, |
| { |
| "epoch": 26.666666666666668, |
| "step": 6000, |
| "total_flos": 0.0, |
| "train_loss": 0.021676936464073756, |
| "train_runtime": 9299.1722, |
| "train_samples_per_second": 20.647, |
| "train_steps_per_second": 0.645 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 6000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 27, |
| "save_steps": 3000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 0.0, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
