{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 42.10526315789474, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.021052631578947368, "grad_norm": 8.503754615783691, "learning_rate": 1.8e-06, "loss": 2.7207, "step": 10 }, { "epoch": 0.042105263157894736, "grad_norm": 5.705361366271973, "learning_rate": 3.8e-06, "loss": 2.3016, "step": 20 }, { "epoch": 0.06315789473684211, "grad_norm": 6.511943340301514, "learning_rate": 5.8e-06, "loss": 1.9608, "step": 30 }, { "epoch": 0.08421052631578947, "grad_norm": 2.6855931282043457, "learning_rate": 7.8e-06, "loss": 1.4027, "step": 40 }, { "epoch": 0.10526315789473684, "grad_norm": 3.0242934226989746, "learning_rate": 9.800000000000001e-06, "loss": 1.0584, "step": 50 }, { "epoch": 0.12631578947368421, "grad_norm": 1.1803932189941406, "learning_rate": 1.18e-05, "loss": 0.7455, "step": 60 }, { "epoch": 0.14736842105263157, "grad_norm": 0.9207198619842529, "learning_rate": 1.3800000000000002e-05, "loss": 0.6333, "step": 70 }, { "epoch": 0.16842105263157894, "grad_norm": 0.4894751012325287, "learning_rate": 1.58e-05, "loss": 0.5182, "step": 80 }, { "epoch": 0.18947368421052632, "grad_norm": 0.9047948122024536, "learning_rate": 1.78e-05, "loss": 0.4963, "step": 90 }, { "epoch": 0.21052631578947367, "grad_norm": 0.4729878008365631, "learning_rate": 1.9800000000000004e-05, "loss": 0.4302, "step": 100 }, { "epoch": 0.23157894736842105, "grad_norm": 0.5196611285209656, "learning_rate": 2.18e-05, "loss": 0.3952, "step": 110 }, { "epoch": 0.25263157894736843, "grad_norm": 0.4544428884983063, "learning_rate": 2.38e-05, "loss": 0.3413, "step": 120 }, { "epoch": 0.2736842105263158, "grad_norm": 0.6557949185371399, "learning_rate": 2.58e-05, "loss": 0.3202, "step": 130 }, { "epoch": 0.29473684210526313, "grad_norm": 0.4605613946914673, "learning_rate": 2.7800000000000005e-05, "loss": 0.2833, "step": 140 }, { "epoch": 0.3157894736842105, "grad_norm": 0.577867329120636, "learning_rate": 2.98e-05, "loss": 0.2496, "step": 150 }, { "epoch": 0.3368421052631579, "grad_norm": 0.6250572800636292, "learning_rate": 3.18e-05, "loss": 0.2349, "step": 160 }, { "epoch": 0.35789473684210527, "grad_norm": 0.5590401887893677, "learning_rate": 3.38e-05, "loss": 0.1829, "step": 170 }, { "epoch": 0.37894736842105264, "grad_norm": 0.7268416881561279, "learning_rate": 3.58e-05, "loss": 0.1772, "step": 180 }, { "epoch": 0.4, "grad_norm": 0.6964499354362488, "learning_rate": 3.7800000000000004e-05, "loss": 0.1597, "step": 190 }, { "epoch": 0.42105263157894735, "grad_norm": 0.7219511866569519, "learning_rate": 3.9800000000000005e-05, "loss": 0.1335, "step": 200 }, { "epoch": 0.4421052631578947, "grad_norm": 0.6252866983413696, "learning_rate": 4.18e-05, "loss": 0.1195, "step": 210 }, { "epoch": 0.4631578947368421, "grad_norm": 0.5906018018722534, "learning_rate": 4.38e-05, "loss": 0.1082, "step": 220 }, { "epoch": 0.4842105263157895, "grad_norm": 0.7115458250045776, "learning_rate": 4.58e-05, "loss": 0.0933, "step": 230 }, { "epoch": 0.5052631578947369, "grad_norm": 0.4811767339706421, "learning_rate": 4.78e-05, "loss": 0.1059, "step": 240 }, { "epoch": 0.5263157894736842, "grad_norm": 0.6518828868865967, "learning_rate": 4.9800000000000004e-05, "loss": 0.0851, "step": 250 }, { "epoch": 0.5473684210526316, "grad_norm": 0.5674479603767395, "learning_rate": 5.1800000000000005e-05, "loss": 0.0958, "step": 260 }, { "epoch": 0.5684210526315789, "grad_norm": 0.617078959941864, "learning_rate": 5.380000000000001e-05, "loss": 0.0811, "step": 270 }, { "epoch": 0.5894736842105263, "grad_norm": 0.6392322778701782, "learning_rate": 5.580000000000001e-05, "loss": 0.0803, "step": 280 }, { "epoch": 0.6105263157894737, "grad_norm": 0.4684285521507263, "learning_rate": 5.7799999999999995e-05, "loss": 0.0694, "step": 290 }, { "epoch": 0.631578947368421, "grad_norm": 0.5111625790596008, "learning_rate": 5.9800000000000003e-05, "loss": 0.0684, "step": 300 }, { "epoch": 0.6526315789473685, "grad_norm": 0.7169430255889893, "learning_rate": 6.18e-05, "loss": 0.0631, "step": 310 }, { "epoch": 0.6736842105263158, "grad_norm": 0.5509821772575378, "learning_rate": 6.38e-05, "loss": 0.0646, "step": 320 }, { "epoch": 0.6947368421052632, "grad_norm": 0.6498633623123169, "learning_rate": 6.58e-05, "loss": 0.0685, "step": 330 }, { "epoch": 0.7157894736842105, "grad_norm": 0.6302815079689026, "learning_rate": 6.780000000000001e-05, "loss": 0.0621, "step": 340 }, { "epoch": 0.7368421052631579, "grad_norm": 0.6227464079856873, "learning_rate": 6.98e-05, "loss": 0.059, "step": 350 }, { "epoch": 0.7578947368421053, "grad_norm": 0.4377397298812866, "learning_rate": 7.18e-05, "loss": 0.0574, "step": 360 }, { "epoch": 0.7789473684210526, "grad_norm": 0.5990102887153625, "learning_rate": 7.38e-05, "loss": 0.0652, "step": 370 }, { "epoch": 0.8, "grad_norm": 0.5051601529121399, "learning_rate": 7.58e-05, "loss": 0.0579, "step": 380 }, { "epoch": 0.8210526315789474, "grad_norm": 0.502711296081543, "learning_rate": 7.780000000000001e-05, "loss": 0.0573, "step": 390 }, { "epoch": 0.8421052631578947, "grad_norm": 0.6807025671005249, "learning_rate": 7.98e-05, "loss": 0.0542, "step": 400 }, { "epoch": 0.8631578947368421, "grad_norm": 0.8033237457275391, "learning_rate": 8.18e-05, "loss": 0.0618, "step": 410 }, { "epoch": 0.8842105263157894, "grad_norm": 0.5379871129989624, "learning_rate": 8.38e-05, "loss": 0.0542, "step": 420 }, { "epoch": 0.9052631578947369, "grad_norm": 0.6699684262275696, "learning_rate": 8.58e-05, "loss": 0.0524, "step": 430 }, { "epoch": 0.9263157894736842, "grad_norm": 0.5673244595527649, "learning_rate": 8.78e-05, "loss": 0.0507, "step": 440 }, { "epoch": 0.9473684210526315, "grad_norm": 0.9217477440834045, "learning_rate": 8.98e-05, "loss": 0.05, "step": 450 }, { "epoch": 0.968421052631579, "grad_norm": 0.9530487656593323, "learning_rate": 9.180000000000001e-05, "loss": 0.047, "step": 460 }, { "epoch": 0.9894736842105263, "grad_norm": 0.7898050546646118, "learning_rate": 9.38e-05, "loss": 0.0504, "step": 470 }, { "epoch": 1.0105263157894737, "grad_norm": 0.799554705619812, "learning_rate": 9.58e-05, "loss": 0.0479, "step": 480 }, { "epoch": 1.0315789473684212, "grad_norm": 0.5109299421310425, "learning_rate": 9.78e-05, "loss": 0.0451, "step": 490 }, { "epoch": 1.0526315789473684, "grad_norm": 0.5954387187957764, "learning_rate": 9.98e-05, "loss": 0.0436, "step": 500 }, { "epoch": 1.0736842105263158, "grad_norm": 0.6091400980949402, "learning_rate": 9.9999778549206e-05, "loss": 0.0434, "step": 510 }, { "epoch": 1.0947368421052632, "grad_norm": 0.7591428756713867, "learning_rate": 9.999901304280685e-05, "loss": 0.0427, "step": 520 }, { "epoch": 1.1157894736842104, "grad_norm": 0.6233532428741455, "learning_rate": 9.999770075521164e-05, "loss": 0.0447, "step": 530 }, { "epoch": 1.1368421052631579, "grad_norm": 0.6603638529777527, "learning_rate": 9.99958417007713e-05, "loss": 0.0484, "step": 540 }, { "epoch": 1.1578947368421053, "grad_norm": 0.6122381091117859, "learning_rate": 9.999343589981615e-05, "loss": 0.043, "step": 550 }, { "epoch": 1.1789473684210527, "grad_norm": 0.5791255831718445, "learning_rate": 9.999048337865568e-05, "loss": 0.0385, "step": 560 }, { "epoch": 1.2, "grad_norm": 0.5661384463310242, "learning_rate": 9.998698416957815e-05, "loss": 0.0398, "step": 570 }, { "epoch": 1.2210526315789474, "grad_norm": 0.5633945465087891, "learning_rate": 9.998293831085037e-05, "loss": 0.0427, "step": 580 }, { "epoch": 1.2421052631578948, "grad_norm": 0.5103133320808411, "learning_rate": 9.997834584671719e-05, "loss": 0.0399, "step": 590 }, { "epoch": 1.263157894736842, "grad_norm": 0.5546170473098755, "learning_rate": 9.997320682740107e-05, "loss": 0.044, "step": 600 }, { "epoch": 1.2842105263157895, "grad_norm": 0.6189123392105103, "learning_rate": 9.996752130910149e-05, "loss": 0.0438, "step": 610 }, { "epoch": 1.305263157894737, "grad_norm": 0.6569340825080872, "learning_rate": 9.99612893539944e-05, "loss": 0.042, "step": 620 }, { "epoch": 1.3263157894736843, "grad_norm": 0.7489569783210754, "learning_rate": 9.995451103023144e-05, "loss": 0.0452, "step": 630 }, { "epoch": 1.3473684210526315, "grad_norm": 0.626463770866394, "learning_rate": 9.994718641193928e-05, "loss": 0.0403, "step": 640 }, { "epoch": 1.368421052631579, "grad_norm": 0.5835581421852112, "learning_rate": 9.993931557921874e-05, "loss": 0.0402, "step": 650 }, { "epoch": 1.3894736842105262, "grad_norm": 0.7310763597488403, "learning_rate": 9.993089861814402e-05, "loss": 0.0347, "step": 660 }, { "epoch": 1.4105263157894736, "grad_norm": 0.5743061900138855, "learning_rate": 9.992193562076166e-05, "loss": 0.0383, "step": 670 }, { "epoch": 1.431578947368421, "grad_norm": 0.629757821559906, "learning_rate": 9.991242668508954e-05, "loss": 0.0378, "step": 680 }, { "epoch": 1.4526315789473685, "grad_norm": 0.5781970024108887, "learning_rate": 9.990237191511587e-05, "loss": 0.037, "step": 690 }, { "epoch": 1.4736842105263157, "grad_norm": 0.7783619165420532, "learning_rate": 9.989177142079802e-05, "loss": 0.0331, "step": 700 }, { "epoch": 1.4947368421052631, "grad_norm": 0.55281662940979, "learning_rate": 9.988062531806126e-05, "loss": 0.0383, "step": 710 }, { "epoch": 1.5157894736842106, "grad_norm": 0.4128439128398895, "learning_rate": 9.986893372879762e-05, "loss": 0.037, "step": 720 }, { "epoch": 1.5368421052631578, "grad_norm": 0.4584996998310089, "learning_rate": 9.985669678086443e-05, "loss": 0.0387, "step": 730 }, { "epoch": 1.5578947368421052, "grad_norm": 0.4686180353164673, "learning_rate": 9.984391460808298e-05, "loss": 0.0356, "step": 740 }, { "epoch": 1.5789473684210527, "grad_norm": 0.6640623211860657, "learning_rate": 9.983058735023709e-05, "loss": 0.0354, "step": 750 }, { "epoch": 1.6, "grad_norm": 0.5313104391098022, "learning_rate": 9.98167151530715e-05, "loss": 0.0373, "step": 760 }, { "epoch": 1.6210526315789475, "grad_norm": 0.494812935590744, "learning_rate": 9.980229816829034e-05, "loss": 0.0358, "step": 770 }, { "epoch": 1.6421052631578947, "grad_norm": 0.5269111394882202, "learning_rate": 9.978733655355544e-05, "loss": 0.0358, "step": 780 }, { "epoch": 1.663157894736842, "grad_norm": 0.48315274715423584, "learning_rate": 9.977183047248464e-05, "loss": 0.0348, "step": 790 }, { "epoch": 1.6842105263157894, "grad_norm": 0.558003306388855, "learning_rate": 9.975578009464992e-05, "loss": 0.0376, "step": 800 }, { "epoch": 1.7052631578947368, "grad_norm": 0.6481532454490662, "learning_rate": 9.97391855955757e-05, "loss": 0.0295, "step": 810 }, { "epoch": 1.7263157894736842, "grad_norm": 0.9239071607589722, "learning_rate": 9.972204715673669e-05, "loss": 0.0361, "step": 820 }, { "epoch": 1.7473684210526317, "grad_norm": 0.6654219627380371, "learning_rate": 9.970436496555617e-05, "loss": 0.0326, "step": 830 }, { "epoch": 1.768421052631579, "grad_norm": 0.5071125030517578, "learning_rate": 9.968613921540373e-05, "loss": 0.0363, "step": 840 }, { "epoch": 1.7894736842105263, "grad_norm": 0.5101365447044373, "learning_rate": 9.966737010559326e-05, "loss": 0.0357, "step": 850 }, { "epoch": 1.8105263157894735, "grad_norm": 0.5802615284919739, "learning_rate": 9.964805784138072e-05, "loss": 0.0321, "step": 860 }, { "epoch": 1.831578947368421, "grad_norm": 0.5119983553886414, "learning_rate": 9.962820263396195e-05, "loss": 0.0308, "step": 870 }, { "epoch": 1.8526315789473684, "grad_norm": 0.49391722679138184, "learning_rate": 9.960780470047033e-05, "loss": 0.0348, "step": 880 }, { "epoch": 1.8736842105263158, "grad_norm": 0.5086302161216736, "learning_rate": 9.958686426397437e-05, "loss": 0.0322, "step": 890 }, { "epoch": 1.8947368421052633, "grad_norm": 0.47594889998435974, "learning_rate": 9.956538155347534e-05, "loss": 0.0318, "step": 900 }, { "epoch": 1.9157894736842105, "grad_norm": 0.5074366927146912, "learning_rate": 9.95433568039047e-05, "loss": 0.0288, "step": 910 }, { "epoch": 1.936842105263158, "grad_norm": 0.48226410150527954, "learning_rate": 9.952079025612162e-05, "loss": 0.032, "step": 920 }, { "epoch": 1.9578947368421051, "grad_norm": 0.5096187591552734, "learning_rate": 9.949768215691022e-05, "loss": 0.0283, "step": 930 }, { "epoch": 1.9789473684210526, "grad_norm": 0.477422297000885, "learning_rate": 9.9474032758977e-05, "loss": 0.0328, "step": 940 }, { "epoch": 2.0, "grad_norm": 0.5572161078453064, "learning_rate": 9.944984232094794e-05, "loss": 0.035, "step": 950 }, { "epoch": 2.0210526315789474, "grad_norm": 0.5973727703094482, "learning_rate": 9.942511110736584e-05, "loss": 0.0309, "step": 960 }, { "epoch": 2.042105263157895, "grad_norm": 0.4117347002029419, "learning_rate": 9.939983938868726e-05, "loss": 0.0249, "step": 970 }, { "epoch": 2.0631578947368423, "grad_norm": 0.5318140387535095, "learning_rate": 9.93740274412797e-05, "loss": 0.0301, "step": 980 }, { "epoch": 2.0842105263157893, "grad_norm": 0.4426449239253998, "learning_rate": 9.934767554741846e-05, "loss": 0.0342, "step": 990 }, { "epoch": 2.1052631578947367, "grad_norm": 0.42530208826065063, "learning_rate": 9.932078399528361e-05, "loss": 0.0279, "step": 1000 }, { "epoch": 2.126315789473684, "grad_norm": 0.5296599268913269, "learning_rate": 9.929335307895689e-05, "loss": 0.0389, "step": 1010 }, { "epoch": 2.1473684210526316, "grad_norm": 0.5326554775238037, "learning_rate": 9.926538309841839e-05, "loss": 0.0301, "step": 1020 }, { "epoch": 2.168421052631579, "grad_norm": 0.521385669708252, "learning_rate": 9.923687435954334e-05, "loss": 0.0287, "step": 1030 }, { "epoch": 2.1894736842105265, "grad_norm": 0.45701244473457336, "learning_rate": 9.920782717409873e-05, "loss": 0.0278, "step": 1040 }, { "epoch": 2.2105263157894735, "grad_norm": 0.4592525064945221, "learning_rate": 9.917824185973994e-05, "loss": 0.0295, "step": 1050 }, { "epoch": 2.231578947368421, "grad_norm": 0.48061153292655945, "learning_rate": 9.914811874000723e-05, "loss": 0.029, "step": 1060 }, { "epoch": 2.2526315789473683, "grad_norm": 0.5665943026542664, "learning_rate": 9.911745814432218e-05, "loss": 0.0291, "step": 1070 }, { "epoch": 2.2736842105263158, "grad_norm": 0.36088046431541443, "learning_rate": 9.90862604079842e-05, "loss": 0.029, "step": 1080 }, { "epoch": 2.294736842105263, "grad_norm": 0.6926175951957703, "learning_rate": 9.90545258721667e-05, "loss": 0.0283, "step": 1090 }, { "epoch": 2.3157894736842106, "grad_norm": 0.47341257333755493, "learning_rate": 9.90222548839135e-05, "loss": 0.0293, "step": 1100 }, { "epoch": 2.336842105263158, "grad_norm": 0.29812091588974, "learning_rate": 9.898944779613495e-05, "loss": 0.0287, "step": 1110 }, { "epoch": 2.3578947368421055, "grad_norm": 0.3907870650291443, "learning_rate": 9.89561049676041e-05, "loss": 0.0299, "step": 1120 }, { "epoch": 2.3789473684210525, "grad_norm": 0.6614090800285339, "learning_rate": 9.89222267629528e-05, "loss": 0.0292, "step": 1130 }, { "epoch": 2.4, "grad_norm": 0.7025224566459656, "learning_rate": 9.888781355266763e-05, "loss": 0.0338, "step": 1140 }, { "epoch": 2.4210526315789473, "grad_norm": 0.5552415251731873, "learning_rate": 9.885286571308598e-05, "loss": 0.0276, "step": 1150 }, { "epoch": 2.442105263157895, "grad_norm": 0.6426026821136475, "learning_rate": 9.881738362639182e-05, "loss": 0.0307, "step": 1160 }, { "epoch": 2.463157894736842, "grad_norm": 0.6928815245628357, "learning_rate": 9.878136768061154e-05, "loss": 0.0292, "step": 1170 }, { "epoch": 2.4842105263157896, "grad_norm": 0.4774184823036194, "learning_rate": 9.874481826960979e-05, "loss": 0.0263, "step": 1180 }, { "epoch": 2.5052631578947366, "grad_norm": 0.3754744827747345, "learning_rate": 9.870773579308503e-05, "loss": 0.0301, "step": 1190 }, { "epoch": 2.526315789473684, "grad_norm": 0.4843890368938446, "learning_rate": 9.867012065656533e-05, "loss": 0.0244, "step": 1200 }, { "epoch": 2.5473684210526315, "grad_norm": 0.4003877341747284, "learning_rate": 9.863197327140376e-05, "loss": 0.0257, "step": 1210 }, { "epoch": 2.568421052631579, "grad_norm": 0.524228572845459, "learning_rate": 9.859329405477403e-05, "loss": 0.0234, "step": 1220 }, { "epoch": 2.5894736842105264, "grad_norm": 0.38531818985939026, "learning_rate": 9.855408342966585e-05, "loss": 0.0265, "step": 1230 }, { "epoch": 2.610526315789474, "grad_norm": 0.5720009803771973, "learning_rate": 9.851434182488033e-05, "loss": 0.0252, "step": 1240 }, { "epoch": 2.6315789473684212, "grad_norm": 0.5133599042892456, "learning_rate": 9.84740696750253e-05, "loss": 0.0245, "step": 1250 }, { "epoch": 2.6526315789473687, "grad_norm": 0.46227186918258667, "learning_rate": 9.843326742051055e-05, "loss": 0.0277, "step": 1260 }, { "epoch": 2.6736842105263157, "grad_norm": 0.45995837450027466, "learning_rate": 9.839193550754297e-05, "loss": 0.0274, "step": 1270 }, { "epoch": 2.694736842105263, "grad_norm": 0.5029829144477844, "learning_rate": 9.835007438812177e-05, "loss": 0.0269, "step": 1280 }, { "epoch": 2.7157894736842105, "grad_norm": 0.3601205348968506, "learning_rate": 9.830768452003341e-05, "loss": 0.0278, "step": 1290 }, { "epoch": 2.736842105263158, "grad_norm": 0.45984387397766113, "learning_rate": 9.826476636684671e-05, "loss": 0.0249, "step": 1300 }, { "epoch": 2.7578947368421054, "grad_norm": 0.48026055097579956, "learning_rate": 9.822132039790773e-05, "loss": 0.0299, "step": 1310 }, { "epoch": 2.7789473684210524, "grad_norm": 0.454712450504303, "learning_rate": 9.817734708833461e-05, "loss": 0.0305, "step": 1320 }, { "epoch": 2.8, "grad_norm": 0.49036332964897156, "learning_rate": 9.813284691901243e-05, "loss": 0.0255, "step": 1330 }, { "epoch": 2.8210526315789473, "grad_norm": 0.7132875323295593, "learning_rate": 9.808782037658792e-05, "loss": 0.0303, "step": 1340 }, { "epoch": 2.8421052631578947, "grad_norm": 0.35366278886795044, "learning_rate": 9.804226795346411e-05, "loss": 0.0251, "step": 1350 }, { "epoch": 2.863157894736842, "grad_norm": 0.5551488995552063, "learning_rate": 9.799619014779503e-05, "loss": 0.0283, "step": 1360 }, { "epoch": 2.8842105263157896, "grad_norm": 0.4735112190246582, "learning_rate": 9.794958746348013e-05, "loss": 0.0322, "step": 1370 }, { "epoch": 2.905263157894737, "grad_norm": 0.5704708695411682, "learning_rate": 9.790246041015896e-05, "loss": 0.0264, "step": 1380 }, { "epoch": 2.9263157894736844, "grad_norm": 0.49438658356666565, "learning_rate": 9.785480950320538e-05, "loss": 0.0262, "step": 1390 }, { "epoch": 2.9473684210526314, "grad_norm": 0.30881860852241516, "learning_rate": 9.78066352637221e-05, "loss": 0.0256, "step": 1400 }, { "epoch": 2.968421052631579, "grad_norm": 0.41681334376335144, "learning_rate": 9.775793821853488e-05, "loss": 0.0256, "step": 1410 }, { "epoch": 2.9894736842105263, "grad_norm": 0.611025869846344, "learning_rate": 9.77087189001868e-05, "loss": 0.0259, "step": 1420 }, { "epoch": 3.0105263157894737, "grad_norm": 0.4861544966697693, "learning_rate": 9.765897784693243e-05, "loss": 0.0238, "step": 1430 }, { "epoch": 3.031578947368421, "grad_norm": 0.4368692934513092, "learning_rate": 9.760871560273197e-05, "loss": 0.0215, "step": 1440 }, { "epoch": 3.0526315789473686, "grad_norm": 0.4631462097167969, "learning_rate": 9.755793271724526e-05, "loss": 0.0236, "step": 1450 }, { "epoch": 3.0736842105263156, "grad_norm": 0.427481472492218, "learning_rate": 9.750662974582584e-05, "loss": 0.0277, "step": 1460 }, { "epoch": 3.094736842105263, "grad_norm": 0.4549742341041565, "learning_rate": 9.745480724951473e-05, "loss": 0.025, "step": 1470 }, { "epoch": 3.1157894736842104, "grad_norm": 0.39217936992645264, "learning_rate": 9.740246579503447e-05, "loss": 0.023, "step": 1480 }, { "epoch": 3.136842105263158, "grad_norm": 0.568151593208313, "learning_rate": 9.734960595478284e-05, "loss": 0.0241, "step": 1490 }, { "epoch": 3.1578947368421053, "grad_norm": 0.6392190456390381, "learning_rate": 9.729622830682657e-05, "loss": 0.0297, "step": 1500 }, { "epoch": 3.1789473684210527, "grad_norm": 0.3805929124355316, "learning_rate": 9.724233343489504e-05, "loss": 0.0227, "step": 1510 }, { "epoch": 3.2, "grad_norm": 0.5149405598640442, "learning_rate": 9.718792192837396e-05, "loss": 0.0276, "step": 1520 }, { "epoch": 3.221052631578947, "grad_norm": 0.6292766332626343, "learning_rate": 9.713299438229886e-05, "loss": 0.0268, "step": 1530 }, { "epoch": 3.2421052631578946, "grad_norm": 0.38368725776672363, "learning_rate": 9.707755139734855e-05, "loss": 0.0247, "step": 1540 }, { "epoch": 3.263157894736842, "grad_norm": 0.45574772357940674, "learning_rate": 9.702159357983866e-05, "loss": 0.0217, "step": 1550 }, { "epoch": 3.2842105263157895, "grad_norm": 0.45282182097435, "learning_rate": 9.696512154171492e-05, "loss": 0.027, "step": 1560 }, { "epoch": 3.305263157894737, "grad_norm": 0.4796925187110901, "learning_rate": 9.690813590054645e-05, "loss": 0.0243, "step": 1570 }, { "epoch": 3.3263157894736843, "grad_norm": 0.43071457743644714, "learning_rate": 9.685063727951914e-05, "loss": 0.0299, "step": 1580 }, { "epoch": 3.3473684210526318, "grad_norm": 0.5006822347640991, "learning_rate": 9.679262630742865e-05, "loss": 0.0278, "step": 1590 }, { "epoch": 3.3684210526315788, "grad_norm": 0.7141162157058716, "learning_rate": 9.673410361867373e-05, "loss": 0.0257, "step": 1600 }, { "epoch": 3.389473684210526, "grad_norm": 0.6680638194084167, "learning_rate": 9.667506985324909e-05, "loss": 0.0254, "step": 1610 }, { "epoch": 3.4105263157894736, "grad_norm": 0.4552099108695984, "learning_rate": 9.661552565673855e-05, "loss": 0.0221, "step": 1620 }, { "epoch": 3.431578947368421, "grad_norm": 0.5017929673194885, "learning_rate": 9.655547168030789e-05, "loss": 0.024, "step": 1630 }, { "epoch": 3.4526315789473685, "grad_norm": 0.5434540510177612, "learning_rate": 9.649490858069777e-05, "loss": 0.0238, "step": 1640 }, { "epoch": 3.473684210526316, "grad_norm": 0.5149438977241516, "learning_rate": 9.643383702021658e-05, "loss": 0.027, "step": 1650 }, { "epoch": 3.4947368421052634, "grad_norm": 0.5576283931732178, "learning_rate": 9.637225766673307e-05, "loss": 0.0242, "step": 1660 }, { "epoch": 3.515789473684211, "grad_norm": 0.479753315448761, "learning_rate": 9.631017119366922e-05, "loss": 0.0233, "step": 1670 }, { "epoch": 3.536842105263158, "grad_norm": 0.4490499198436737, "learning_rate": 9.624757827999273e-05, "loss": 0.0222, "step": 1680 }, { "epoch": 3.557894736842105, "grad_norm": 0.5094167590141296, "learning_rate": 9.618447961020971e-05, "loss": 0.029, "step": 1690 }, { "epoch": 3.5789473684210527, "grad_norm": 0.508141279220581, "learning_rate": 9.612087587435707e-05, "loss": 0.0252, "step": 1700 }, { "epoch": 3.6, "grad_norm": 0.4324563145637512, "learning_rate": 9.605676776799508e-05, "loss": 0.0234, "step": 1710 }, { "epoch": 3.6210526315789475, "grad_norm": 0.5795191526412964, "learning_rate": 9.599215599219973e-05, "loss": 0.0244, "step": 1720 }, { "epoch": 3.6421052631578945, "grad_norm": 0.5809512138366699, "learning_rate": 9.592704125355505e-05, "loss": 0.0252, "step": 1730 }, { "epoch": 3.663157894736842, "grad_norm": 0.5399959683418274, "learning_rate": 9.586142426414538e-05, "loss": 0.0205, "step": 1740 }, { "epoch": 3.6842105263157894, "grad_norm": 0.541290283203125, "learning_rate": 9.57953057415476e-05, "loss": 0.0257, "step": 1750 }, { "epoch": 3.705263157894737, "grad_norm": 0.4715765118598938, "learning_rate": 9.572868640882328e-05, "loss": 0.0272, "step": 1760 }, { "epoch": 3.7263157894736842, "grad_norm": 0.4055497944355011, "learning_rate": 9.56615669945108e-05, "loss": 0.0231, "step": 1770 }, { "epoch": 3.7473684210526317, "grad_norm": 0.4754236042499542, "learning_rate": 9.55939482326173e-05, "loss": 0.0215, "step": 1780 }, { "epoch": 3.768421052631579, "grad_norm": 0.4024890065193176, "learning_rate": 9.552583086261069e-05, "loss": 0.0229, "step": 1790 }, { "epoch": 3.7894736842105265, "grad_norm": 0.5853466391563416, "learning_rate": 9.545721562941168e-05, "loss": 0.0259, "step": 1800 }, { "epoch": 3.8105263157894735, "grad_norm": 0.5559604167938232, "learning_rate": 9.538810328338543e-05, "loss": 0.027, "step": 1810 }, { "epoch": 3.831578947368421, "grad_norm": 0.4336540699005127, "learning_rate": 9.531849458033349e-05, "loss": 0.0253, "step": 1820 }, { "epoch": 3.8526315789473684, "grad_norm": 0.3995809555053711, "learning_rate": 9.524839028148547e-05, "loss": 0.0197, "step": 1830 }, { "epoch": 3.873684210526316, "grad_norm": 0.5117527842521667, "learning_rate": 9.517779115349077e-05, "loss": 0.0232, "step": 1840 }, { "epoch": 3.8947368421052633, "grad_norm": 0.5508883595466614, "learning_rate": 9.510669796841014e-05, "loss": 0.0199, "step": 1850 }, { "epoch": 3.9157894736842103, "grad_norm": 0.5789328217506409, "learning_rate": 9.503511150370727e-05, "loss": 0.0187, "step": 1860 }, { "epoch": 3.9368421052631577, "grad_norm": 0.45967409014701843, "learning_rate": 9.496303254224024e-05, "loss": 0.0199, "step": 1870 }, { "epoch": 3.957894736842105, "grad_norm": 0.46641257405281067, "learning_rate": 9.489046187225306e-05, "loss": 0.019, "step": 1880 }, { "epoch": 3.9789473684210526, "grad_norm": 0.4676485061645508, "learning_rate": 9.481740028736692e-05, "loss": 0.0218, "step": 1890 }, { "epoch": 4.0, "grad_norm": 0.4378170967102051, "learning_rate": 9.474384858657164e-05, "loss": 0.0199, "step": 1900 }, { "epoch": 4.021052631578947, "grad_norm": 0.43269822001457214, "learning_rate": 9.466980757421679e-05, "loss": 0.0255, "step": 1910 }, { "epoch": 4.042105263157895, "grad_norm": 0.5129058957099915, "learning_rate": 9.459527806000305e-05, "loss": 0.0206, "step": 1920 }, { "epoch": 4.063157894736842, "grad_norm": 0.402515709400177, "learning_rate": 9.452026085897325e-05, "loss": 0.021, "step": 1930 }, { "epoch": 4.08421052631579, "grad_norm": 0.32655131816864014, "learning_rate": 9.444475679150348e-05, "loss": 0.0242, "step": 1940 }, { "epoch": 4.105263157894737, "grad_norm": 0.3683534860610962, "learning_rate": 9.436876668329411e-05, "loss": 0.0209, "step": 1950 }, { "epoch": 4.126315789473685, "grad_norm": 0.43539005517959595, "learning_rate": 9.429229136536079e-05, "loss": 0.0211, "step": 1960 }, { "epoch": 4.147368421052631, "grad_norm": 0.2985154688358307, "learning_rate": 9.421533167402534e-05, "loss": 0.019, "step": 1970 }, { "epoch": 4.168421052631579, "grad_norm": 0.2865261733531952, "learning_rate": 9.413788845090666e-05, "loss": 0.023, "step": 1980 }, { "epoch": 4.189473684210526, "grad_norm": 0.4096466302871704, "learning_rate": 9.405996254291136e-05, "loss": 0.021, "step": 1990 }, { "epoch": 4.2105263157894735, "grad_norm": 0.3747730255126953, "learning_rate": 9.398155480222474e-05, "loss": 0.0249, "step": 2000 }, { "epoch": 4.231578947368421, "grad_norm": 0.4947047829627991, "learning_rate": 9.390266608630128e-05, "loss": 0.0209, "step": 2010 }, { "epoch": 4.252631578947368, "grad_norm": 0.5458012819290161, "learning_rate": 9.38232972578553e-05, "loss": 0.0219, "step": 2020 }, { "epoch": 4.273684210526316, "grad_norm": 0.35812586545944214, "learning_rate": 9.374344918485164e-05, "loss": 0.0216, "step": 2030 }, { "epoch": 4.294736842105263, "grad_norm": 0.40601086616516113, "learning_rate": 9.366312274049602e-05, "loss": 0.0203, "step": 2040 }, { "epoch": 4.315789473684211, "grad_norm": 0.49646010994911194, "learning_rate": 9.358231880322554e-05, "loss": 0.0225, "step": 2050 }, { "epoch": 4.336842105263158, "grad_norm": 0.3560623526573181, "learning_rate": 9.350103825669916e-05, "loss": 0.0212, "step": 2060 }, { "epoch": 4.3578947368421055, "grad_norm": 0.40576326847076416, "learning_rate": 9.341928198978787e-05, "loss": 0.025, "step": 2070 }, { "epoch": 4.378947368421053, "grad_norm": 0.6641100645065308, "learning_rate": 9.333705089656512e-05, "loss": 0.023, "step": 2080 }, { "epoch": 4.4, "grad_norm": 0.6189097166061401, "learning_rate": 9.325434587629698e-05, "loss": 0.0268, "step": 2090 }, { "epoch": 4.421052631578947, "grad_norm": 0.3435753881931305, "learning_rate": 9.31711678334323e-05, "loss": 0.0178, "step": 2100 }, { "epoch": 4.442105263157894, "grad_norm": 0.2675717771053314, "learning_rate": 9.308751767759282e-05, "loss": 0.0208, "step": 2110 }, { "epoch": 4.463157894736842, "grad_norm": 0.4262930452823639, "learning_rate": 9.300339632356325e-05, "loss": 0.0208, "step": 2120 }, { "epoch": 4.484210526315789, "grad_norm": 0.3225375711917877, "learning_rate": 9.291880469128124e-05, "loss": 0.0223, "step": 2130 }, { "epoch": 4.505263157894737, "grad_norm": 0.35266122221946716, "learning_rate": 9.283374370582732e-05, "loss": 0.0192, "step": 2140 }, { "epoch": 4.526315789473684, "grad_norm": 0.5800217986106873, "learning_rate": 9.274821429741482e-05, "loss": 0.0195, "step": 2150 }, { "epoch": 4.5473684210526315, "grad_norm": 0.4638044238090515, "learning_rate": 9.266221740137961e-05, "loss": 0.019, "step": 2160 }, { "epoch": 4.568421052631579, "grad_norm": 0.40784841775894165, "learning_rate": 9.257575395817001e-05, "loss": 0.0215, "step": 2170 }, { "epoch": 4.589473684210526, "grad_norm": 0.3112774193286896, "learning_rate": 9.248882491333637e-05, "loss": 0.0199, "step": 2180 }, { "epoch": 4.610526315789474, "grad_norm": 0.3140299618244171, "learning_rate": 9.240143121752076e-05, "loss": 0.0222, "step": 2190 }, { "epoch": 4.631578947368421, "grad_norm": 0.30427688360214233, "learning_rate": 9.23135738264467e-05, "loss": 0.0201, "step": 2200 }, { "epoch": 4.652631578947369, "grad_norm": 0.3895975351333618, "learning_rate": 9.222525370090849e-05, "loss": 0.0201, "step": 2210 }, { "epoch": 4.673684210526316, "grad_norm": 0.5227062702178955, "learning_rate": 9.213647180676088e-05, "loss": 0.0211, "step": 2220 }, { "epoch": 4.6947368421052635, "grad_norm": 0.5867322087287903, "learning_rate": 9.204722911490846e-05, "loss": 0.022, "step": 2230 }, { "epoch": 4.715789473684211, "grad_norm": 0.4475304186344147, "learning_rate": 9.1957526601295e-05, "loss": 0.0193, "step": 2240 }, { "epoch": 4.7368421052631575, "grad_norm": 0.2602757215499878, "learning_rate": 9.186736524689281e-05, "loss": 0.0183, "step": 2250 }, { "epoch": 4.757894736842105, "grad_norm": 0.4600152373313904, "learning_rate": 9.177674603769204e-05, "loss": 0.0238, "step": 2260 }, { "epoch": 4.778947368421052, "grad_norm": 0.46683645248413086, "learning_rate": 9.168566996468983e-05, "loss": 0.0257, "step": 2270 }, { "epoch": 4.8, "grad_norm": 0.40355321764945984, "learning_rate": 9.159413802387951e-05, "loss": 0.0206, "step": 2280 }, { "epoch": 4.821052631578947, "grad_norm": 0.3210936188697815, "learning_rate": 9.150215121623974e-05, "loss": 0.022, "step": 2290 }, { "epoch": 4.842105263157895, "grad_norm": 0.39786839485168457, "learning_rate": 9.140971054772349e-05, "loss": 0.0216, "step": 2300 }, { "epoch": 4.863157894736842, "grad_norm": 0.42395541071891785, "learning_rate": 9.131681702924713e-05, "loss": 0.0201, "step": 2310 }, { "epoch": 4.88421052631579, "grad_norm": 0.3191361427307129, "learning_rate": 9.122347167667926e-05, "loss": 0.0186, "step": 2320 }, { "epoch": 4.905263157894737, "grad_norm": 0.41964268684387207, "learning_rate": 9.112967551082973e-05, "loss": 0.0195, "step": 2330 }, { "epoch": 4.926315789473684, "grad_norm": 0.27586865425109863, "learning_rate": 9.103542955743835e-05, "loss": 0.018, "step": 2340 }, { "epoch": 4.947368421052632, "grad_norm": 0.28472578525543213, "learning_rate": 9.094073484716381e-05, "loss": 0.0164, "step": 2350 }, { "epoch": 4.968421052631579, "grad_norm": 0.3787604570388794, "learning_rate": 9.084559241557226e-05, "loss": 0.0267, "step": 2360 }, { "epoch": 4.989473684210527, "grad_norm": 0.3770425021648407, "learning_rate": 9.075000330312608e-05, "loss": 0.0214, "step": 2370 }, { "epoch": 5.010526315789473, "grad_norm": 0.3219016194343567, "learning_rate": 9.065396855517253e-05, "loss": 0.0217, "step": 2380 }, { "epoch": 5.031578947368421, "grad_norm": 0.3787834644317627, "learning_rate": 9.055748922193219e-05, "loss": 0.0187, "step": 2390 }, { "epoch": 5.052631578947368, "grad_norm": 0.5634660124778748, "learning_rate": 9.046056635848761e-05, "loss": 0.0211, "step": 2400 }, { "epoch": 5.073684210526316, "grad_norm": 0.6474153995513916, "learning_rate": 9.036320102477169e-05, "loss": 0.0214, "step": 2410 }, { "epoch": 5.094736842105263, "grad_norm": 0.5344454050064087, "learning_rate": 9.02653942855561e-05, "loss": 0.0205, "step": 2420 }, { "epoch": 5.11578947368421, "grad_norm": 0.45306095480918884, "learning_rate": 9.016714721043971e-05, "loss": 0.0214, "step": 2430 }, { "epoch": 5.136842105263158, "grad_norm": 0.30573728680610657, "learning_rate": 9.006846087383675e-05, "loss": 0.0192, "step": 2440 }, { "epoch": 5.157894736842105, "grad_norm": 0.4813554286956787, "learning_rate": 8.996933635496523e-05, "loss": 0.0201, "step": 2450 }, { "epoch": 5.178947368421053, "grad_norm": 0.3331827223300934, "learning_rate": 8.986977473783498e-05, "loss": 0.0194, "step": 2460 }, { "epoch": 5.2, "grad_norm": 0.4607407748699188, "learning_rate": 8.97697771112359e-05, "loss": 0.0214, "step": 2470 }, { "epoch": 5.221052631578948, "grad_norm": 0.5646907091140747, "learning_rate": 8.966934456872602e-05, "loss": 0.0184, "step": 2480 }, { "epoch": 5.242105263157895, "grad_norm": 0.3156760632991791, "learning_rate": 8.95684782086195e-05, "loss": 0.018, "step": 2490 }, { "epoch": 5.2631578947368425, "grad_norm": 0.3745982348918915, "learning_rate": 8.946717913397476e-05, "loss": 0.0186, "step": 2500 }, { "epoch": 5.284210526315789, "grad_norm": 0.361148864030838, "learning_rate": 8.93654484525822e-05, "loss": 0.0188, "step": 2510 }, { "epoch": 5.3052631578947365, "grad_norm": 0.2441021353006363, "learning_rate": 8.926328727695226e-05, "loss": 0.0171, "step": 2520 }, { "epoch": 5.326315789473684, "grad_norm": 0.4154561460018158, "learning_rate": 8.916069672430319e-05, "loss": 0.022, "step": 2530 }, { "epoch": 5.347368421052631, "grad_norm": 0.32268857955932617, "learning_rate": 8.905767791654884e-05, "loss": 0.0199, "step": 2540 }, { "epoch": 5.368421052631579, "grad_norm": 0.43453478813171387, "learning_rate": 8.895423198028638e-05, "loss": 0.0178, "step": 2550 }, { "epoch": 5.389473684210526, "grad_norm": 0.46428170800209045, "learning_rate": 8.885036004678402e-05, "loss": 0.0174, "step": 2560 }, { "epoch": 5.410526315789474, "grad_norm": 0.35074326395988464, "learning_rate": 8.874606325196857e-05, "loss": 0.0177, "step": 2570 }, { "epoch": 5.431578947368421, "grad_norm": 0.28758835792541504, "learning_rate": 8.864134273641304e-05, "loss": 0.0212, "step": 2580 }, { "epoch": 5.4526315789473685, "grad_norm": 0.3045271635055542, "learning_rate": 8.853619964532427e-05, "loss": 0.0232, "step": 2590 }, { "epoch": 5.473684210526316, "grad_norm": 0.3790994882583618, "learning_rate": 8.843063512853019e-05, "loss": 0.0179, "step": 2600 }, { "epoch": 5.494736842105263, "grad_norm": 0.39538803696632385, "learning_rate": 8.832465034046749e-05, "loss": 0.0176, "step": 2610 }, { "epoch": 5.515789473684211, "grad_norm": 0.3040083646774292, "learning_rate": 8.821824644016882e-05, "loss": 0.018, "step": 2620 }, { "epoch": 5.536842105263158, "grad_norm": 0.39079105854034424, "learning_rate": 8.811142459125019e-05, "loss": 0.0234, "step": 2630 }, { "epoch": 5.557894736842105, "grad_norm": 0.3650378882884979, "learning_rate": 8.800418596189822e-05, "loss": 0.0213, "step": 2640 }, { "epoch": 5.578947368421053, "grad_norm": 0.4253116846084595, "learning_rate": 8.789653172485737e-05, "loss": 0.0183, "step": 2650 }, { "epoch": 5.6, "grad_norm": 0.3615553677082062, "learning_rate": 8.778846305741715e-05, "loss": 0.0157, "step": 2660 }, { "epoch": 5.621052631578947, "grad_norm": 0.4618593454360962, "learning_rate": 8.767998114139918e-05, "loss": 0.02, "step": 2670 }, { "epoch": 5.6421052631578945, "grad_norm": 0.38564813137054443, "learning_rate": 8.757108716314429e-05, "loss": 0.02, "step": 2680 }, { "epoch": 5.663157894736842, "grad_norm": 0.297162801027298, "learning_rate": 8.746178231349962e-05, "loss": 0.0206, "step": 2690 }, { "epoch": 5.684210526315789, "grad_norm": 0.3215448260307312, "learning_rate": 8.735206778780549e-05, "loss": 0.0166, "step": 2700 }, { "epoch": 5.705263157894737, "grad_norm": 0.3596934676170349, "learning_rate": 8.724194478588234e-05, "loss": 0.019, "step": 2710 }, { "epoch": 5.726315789473684, "grad_norm": 0.426040381193161, "learning_rate": 8.713141451201772e-05, "loss": 0.0212, "step": 2720 }, { "epoch": 5.747368421052632, "grad_norm": 0.5354184508323669, "learning_rate": 8.702047817495295e-05, "loss": 0.0218, "step": 2730 }, { "epoch": 5.768421052631579, "grad_norm": 0.27179598808288574, "learning_rate": 8.69091369878701e-05, "loss": 0.0194, "step": 2740 }, { "epoch": 5.7894736842105265, "grad_norm": 0.37561744451522827, "learning_rate": 8.679739216837849e-05, "loss": 0.0192, "step": 2750 }, { "epoch": 5.810526315789474, "grad_norm": 0.3362484872341156, "learning_rate": 8.66852449385016e-05, "loss": 0.0203, "step": 2760 }, { "epoch": 5.831578947368421, "grad_norm": 0.3647758364677429, "learning_rate": 8.657269652466356e-05, "loss": 0.0242, "step": 2770 }, { "epoch": 5.852631578947369, "grad_norm": 0.515769362449646, "learning_rate": 8.645974815767577e-05, "loss": 0.0185, "step": 2780 }, { "epoch": 5.873684210526315, "grad_norm": 0.465288370847702, "learning_rate": 8.634640107272351e-05, "loss": 0.0195, "step": 2790 }, { "epoch": 5.894736842105263, "grad_norm": 0.4067588448524475, "learning_rate": 8.623265650935234e-05, "loss": 0.0192, "step": 2800 }, { "epoch": 5.91578947368421, "grad_norm": 0.425921767950058, "learning_rate": 8.611851571145456e-05, "loss": 0.0165, "step": 2810 }, { "epoch": 5.936842105263158, "grad_norm": 0.34170037508010864, "learning_rate": 8.600397992725566e-05, "loss": 0.0188, "step": 2820 }, { "epoch": 5.957894736842105, "grad_norm": 0.4490221440792084, "learning_rate": 8.588905040930061e-05, "loss": 0.0175, "step": 2830 }, { "epoch": 5.978947368421053, "grad_norm": 0.48505014181137085, "learning_rate": 8.577372841444022e-05, "loss": 0.0205, "step": 2840 }, { "epoch": 6.0, "grad_norm": 0.3089272677898407, "learning_rate": 8.565801520381736e-05, "loss": 0.0191, "step": 2850 }, { "epoch": 6.021052631578947, "grad_norm": 0.3596801161766052, "learning_rate": 8.554191204285313e-05, "loss": 0.0168, "step": 2860 }, { "epoch": 6.042105263157895, "grad_norm": 0.3110174536705017, "learning_rate": 8.542542020123315e-05, "loss": 0.0164, "step": 2870 }, { "epoch": 6.063157894736842, "grad_norm": 0.4730689525604248, "learning_rate": 8.530854095289347e-05, "loss": 0.0154, "step": 2880 }, { "epoch": 6.08421052631579, "grad_norm": 0.25971922278404236, "learning_rate": 8.519127557600688e-05, "loss": 0.0157, "step": 2890 }, { "epoch": 6.105263157894737, "grad_norm": 0.3299713730812073, "learning_rate": 8.507362535296871e-05, "loss": 0.0204, "step": 2900 }, { "epoch": 6.126315789473685, "grad_norm": 0.4056624174118042, "learning_rate": 8.495559157038299e-05, "loss": 0.0197, "step": 2910 }, { "epoch": 6.147368421052631, "grad_norm": 0.2801871597766876, "learning_rate": 8.483717551904823e-05, "loss": 0.0169, "step": 2920 }, { "epoch": 6.168421052631579, "grad_norm": 0.2704514265060425, "learning_rate": 8.47183784939434e-05, "loss": 0.0162, "step": 2930 }, { "epoch": 6.189473684210526, "grad_norm": 0.4430570900440216, "learning_rate": 8.459920179421374e-05, "loss": 0.0145, "step": 2940 }, { "epoch": 6.2105263157894735, "grad_norm": 0.46088963747024536, "learning_rate": 8.447964672315656e-05, "loss": 0.0177, "step": 2950 }, { "epoch": 6.231578947368421, "grad_norm": 0.4396511912345886, "learning_rate": 8.435971458820692e-05, "loss": 0.0164, "step": 2960 }, { "epoch": 6.252631578947368, "grad_norm": 0.2893114984035492, "learning_rate": 8.423940670092345e-05, "loss": 0.0163, "step": 2970 }, { "epoch": 6.273684210526316, "grad_norm": 0.4042656123638153, "learning_rate": 8.411872437697394e-05, "loss": 0.0189, "step": 2980 }, { "epoch": 6.294736842105263, "grad_norm": 0.2971465587615967, "learning_rate": 8.399766893612096e-05, "loss": 0.0193, "step": 2990 }, { "epoch": 6.315789473684211, "grad_norm": 0.38724568486213684, "learning_rate": 8.38762417022074e-05, "loss": 0.0203, "step": 3000 }, { "epoch": 6.336842105263158, "grad_norm": 0.33399614691734314, "learning_rate": 8.375444400314204e-05, "loss": 0.0205, "step": 3010 }, { "epoch": 6.3578947368421055, "grad_norm": 0.524776041507721, "learning_rate": 8.3632277170885e-05, "loss": 0.0181, "step": 3020 }, { "epoch": 6.378947368421053, "grad_norm": 0.460173636674881, "learning_rate": 8.350974254143318e-05, "loss": 0.021, "step": 3030 }, { "epoch": 6.4, "grad_norm": 0.31842440366744995, "learning_rate": 8.338684145480566e-05, "loss": 0.0189, "step": 3040 }, { "epoch": 6.421052631578947, "grad_norm": 0.3491009473800659, "learning_rate": 8.326357525502904e-05, "loss": 0.0154, "step": 3050 }, { "epoch": 6.442105263157894, "grad_norm": 0.34132686257362366, "learning_rate": 8.313994529012273e-05, "loss": 0.0198, "step": 3060 }, { "epoch": 6.463157894736842, "grad_norm": 0.4157889783382416, "learning_rate": 8.301595291208422e-05, "loss": 0.0179, "step": 3070 }, { "epoch": 6.484210526315789, "grad_norm": 0.3819684386253357, "learning_rate": 8.289159947687427e-05, "loss": 0.0186, "step": 3080 }, { "epoch": 6.505263157894737, "grad_norm": 0.43223848938941956, "learning_rate": 8.276688634440216e-05, "loss": 0.0178, "step": 3090 }, { "epoch": 6.526315789473684, "grad_norm": 0.37996047735214233, "learning_rate": 8.26418148785107e-05, "loss": 0.0167, "step": 3100 }, { "epoch": 6.5473684210526315, "grad_norm": 0.46757009625434875, "learning_rate": 8.251638644696141e-05, "loss": 0.024, "step": 3110 }, { "epoch": 6.568421052631579, "grad_norm": 0.3201007843017578, "learning_rate": 8.23906024214195e-05, "loss": 0.0173, "step": 3120 }, { "epoch": 6.589473684210526, "grad_norm": 0.3826674520969391, "learning_rate": 8.226446417743897e-05, "loss": 0.0171, "step": 3130 }, { "epoch": 6.610526315789474, "grad_norm": 0.5460965633392334, "learning_rate": 8.213797309444742e-05, "loss": 0.018, "step": 3140 }, { "epoch": 6.631578947368421, "grad_norm": 0.5551507472991943, "learning_rate": 8.201113055573105e-05, "loss": 0.0204, "step": 3150 }, { "epoch": 6.652631578947369, "grad_norm": 0.4706227481365204, "learning_rate": 8.188393794841958e-05, "loss": 0.0188, "step": 3160 }, { "epoch": 6.673684210526316, "grad_norm": 0.5014885663986206, "learning_rate": 8.175639666347094e-05, "loss": 0.0186, "step": 3170 }, { "epoch": 6.6947368421052635, "grad_norm": 0.38826411962509155, "learning_rate": 8.162850809565623e-05, "loss": 0.0179, "step": 3180 }, { "epoch": 6.715789473684211, "grad_norm": 0.3493991792201996, "learning_rate": 8.150027364354431e-05, "loss": 0.0185, "step": 3190 }, { "epoch": 6.7368421052631575, "grad_norm": 0.49658477306365967, "learning_rate": 8.137169470948662e-05, "loss": 0.0191, "step": 3200 }, { "epoch": 6.757894736842105, "grad_norm": 0.514453113079071, "learning_rate": 8.124277269960179e-05, "loss": 0.0168, "step": 3210 }, { "epoch": 6.778947368421052, "grad_norm": 0.4190099537372589, "learning_rate": 8.111350902376023e-05, "loss": 0.0167, "step": 3220 }, { "epoch": 6.8, "grad_norm": 0.4062015116214752, "learning_rate": 8.098390509556883e-05, "loss": 0.0197, "step": 3230 }, { "epoch": 6.821052631578947, "grad_norm": 0.28561830520629883, "learning_rate": 8.085396233235536e-05, "loss": 0.0157, "step": 3240 }, { "epoch": 6.842105263157895, "grad_norm": 0.35810598731040955, "learning_rate": 8.072368215515306e-05, "loss": 0.0176, "step": 3250 }, { "epoch": 6.863157894736842, "grad_norm": 0.3420398533344269, "learning_rate": 8.059306598868506e-05, "loss": 0.0183, "step": 3260 }, { "epoch": 6.88421052631579, "grad_norm": 0.43223845958709717, "learning_rate": 8.046211526134888e-05, "loss": 0.0175, "step": 3270 }, { "epoch": 6.905263157894737, "grad_norm": 0.38998815417289734, "learning_rate": 8.033083140520065e-05, "loss": 0.02, "step": 3280 }, { "epoch": 6.926315789473684, "grad_norm": 0.3556065857410431, "learning_rate": 8.019921585593962e-05, "loss": 0.0157, "step": 3290 }, { "epoch": 6.947368421052632, "grad_norm": 0.43601328134536743, "learning_rate": 8.006727005289232e-05, "loss": 0.0228, "step": 3300 }, { "epoch": 6.968421052631579, "grad_norm": 0.3192536234855652, "learning_rate": 7.993499543899692e-05, "loss": 0.0153, "step": 3310 }, { "epoch": 6.989473684210527, "grad_norm": 0.365305632352829, "learning_rate": 7.980239346078742e-05, "loss": 0.016, "step": 3320 }, { "epoch": 7.010526315789473, "grad_norm": 0.36051324009895325, "learning_rate": 7.966946556837778e-05, "loss": 0.0189, "step": 3330 }, { "epoch": 7.031578947368421, "grad_norm": 0.5628525614738464, "learning_rate": 7.953621321544616e-05, "loss": 0.0149, "step": 3340 }, { "epoch": 7.052631578947368, "grad_norm": 0.5390278697013855, "learning_rate": 7.940263785921896e-05, "loss": 0.0222, "step": 3350 }, { "epoch": 7.073684210526316, "grad_norm": 0.2605836093425751, "learning_rate": 7.926874096045482e-05, "loss": 0.0184, "step": 3360 }, { "epoch": 7.094736842105263, "grad_norm": 0.2945224642753601, "learning_rate": 7.913452398342881e-05, "loss": 0.0144, "step": 3370 }, { "epoch": 7.11578947368421, "grad_norm": 0.34929975867271423, "learning_rate": 7.89999883959163e-05, "loss": 0.0188, "step": 3380 }, { "epoch": 7.136842105263158, "grad_norm": 0.2720954418182373, "learning_rate": 7.886513566917687e-05, "loss": 0.0188, "step": 3390 }, { "epoch": 7.157894736842105, "grad_norm": 0.33141934871673584, "learning_rate": 7.872996727793838e-05, "loss": 0.019, "step": 3400 }, { "epoch": 7.178947368421053, "grad_norm": 0.26607412099838257, "learning_rate": 7.859448470038069e-05, "loss": 0.0164, "step": 3410 }, { "epoch": 7.2, "grad_norm": 0.35745716094970703, "learning_rate": 7.845868941811956e-05, "loss": 0.0167, "step": 3420 }, { "epoch": 7.221052631578948, "grad_norm": 0.3563767373561859, "learning_rate": 7.832258291619043e-05, "loss": 0.019, "step": 3430 }, { "epoch": 7.242105263157895, "grad_norm": 0.2731028199195862, "learning_rate": 7.81861666830322e-05, "loss": 0.0165, "step": 3440 }, { "epoch": 7.2631578947368425, "grad_norm": 0.3665086328983307, "learning_rate": 7.804944221047097e-05, "loss": 0.0157, "step": 3450 }, { "epoch": 7.284210526315789, "grad_norm": 0.5059680938720703, "learning_rate": 7.791241099370364e-05, "loss": 0.0162, "step": 3460 }, { "epoch": 7.3052631578947365, "grad_norm": 0.4355224370956421, "learning_rate": 7.777507453128163e-05, "loss": 0.0161, "step": 3470 }, { "epoch": 7.326315789473684, "grad_norm": 0.4263385534286499, "learning_rate": 7.763743432509451e-05, "loss": 0.0178, "step": 3480 }, { "epoch": 7.347368421052631, "grad_norm": 0.3377225697040558, "learning_rate": 7.749949188035353e-05, "loss": 0.0155, "step": 3490 }, { "epoch": 7.368421052631579, "grad_norm": 0.7231624722480774, "learning_rate": 7.736124870557516e-05, "loss": 0.017, "step": 3500 }, { "epoch": 7.389473684210526, "grad_norm": 0.38127657771110535, "learning_rate": 7.722270631256459e-05, "loss": 0.0189, "step": 3510 }, { "epoch": 7.410526315789474, "grad_norm": 0.34664636850357056, "learning_rate": 7.708386621639925e-05, "loss": 0.0153, "step": 3520 }, { "epoch": 7.431578947368421, "grad_norm": 0.410997599363327, "learning_rate": 7.694472993541219e-05, "loss": 0.0188, "step": 3530 }, { "epoch": 7.4526315789473685, "grad_norm": 0.2986126244068146, "learning_rate": 7.680529899117547e-05, "loss": 0.0159, "step": 3540 }, { "epoch": 7.473684210526316, "grad_norm": 0.409452348947525, "learning_rate": 7.666557490848358e-05, "loss": 0.0171, "step": 3550 }, { "epoch": 7.494736842105263, "grad_norm": 0.33516064286231995, "learning_rate": 7.65255592153367e-05, "loss": 0.0171, "step": 3560 }, { "epoch": 7.515789473684211, "grad_norm": 0.2987566590309143, "learning_rate": 7.638525344292402e-05, "loss": 0.0156, "step": 3570 }, { "epoch": 7.536842105263158, "grad_norm": 0.30680418014526367, "learning_rate": 7.624465912560697e-05, "loss": 0.0178, "step": 3580 }, { "epoch": 7.557894736842105, "grad_norm": 0.3499041497707367, "learning_rate": 7.610377780090249e-05, "loss": 0.0185, "step": 3590 }, { "epoch": 7.578947368421053, "grad_norm": 0.28610754013061523, "learning_rate": 7.596261100946618e-05, "loss": 0.0146, "step": 3600 }, { "epoch": 7.6, "grad_norm": 0.31157001852989197, "learning_rate": 7.582116029507542e-05, "loss": 0.0178, "step": 3610 }, { "epoch": 7.621052631578947, "grad_norm": 0.31508561968803406, "learning_rate": 7.56794272046126e-05, "loss": 0.0174, "step": 3620 }, { "epoch": 7.6421052631578945, "grad_norm": 0.2863736152648926, "learning_rate": 7.55374132880481e-05, "loss": 0.0159, "step": 3630 }, { "epoch": 7.663157894736842, "grad_norm": 0.29254570603370667, "learning_rate": 7.539512009842333e-05, "loss": 0.0154, "step": 3640 }, { "epoch": 7.684210526315789, "grad_norm": 0.5215784311294556, "learning_rate": 7.525254919183382e-05, "loss": 0.0181, "step": 3650 }, { "epoch": 7.705263157894737, "grad_norm": 0.25562772154808044, "learning_rate": 7.510970212741215e-05, "loss": 0.015, "step": 3660 }, { "epoch": 7.726315789473684, "grad_norm": 0.33486708998680115, "learning_rate": 7.496658046731096e-05, "loss": 0.0175, "step": 3670 }, { "epoch": 7.747368421052632, "grad_norm": 0.4196351170539856, "learning_rate": 7.482318577668578e-05, "loss": 0.0177, "step": 3680 }, { "epoch": 7.768421052631579, "grad_norm": 0.32597073912620544, "learning_rate": 7.467951962367796e-05, "loss": 0.0159, "step": 3690 }, { "epoch": 7.7894736842105265, "grad_norm": 0.4474583864212036, "learning_rate": 7.453558357939755e-05, "loss": 0.0148, "step": 3700 }, { "epoch": 7.810526315789474, "grad_norm": 0.28767499327659607, "learning_rate": 7.439137921790606e-05, "loss": 0.0184, "step": 3710 }, { "epoch": 7.831578947368421, "grad_norm": 0.2805885076522827, "learning_rate": 7.42469081161993e-05, "loss": 0.0201, "step": 3720 }, { "epoch": 7.852631578947369, "grad_norm": 0.34607821702957153, "learning_rate": 7.410217185419006e-05, "loss": 0.0158, "step": 3730 }, { "epoch": 7.873684210526315, "grad_norm": 0.49345099925994873, "learning_rate": 7.395717201469095e-05, "loss": 0.0146, "step": 3740 }, { "epoch": 7.894736842105263, "grad_norm": 0.24862772226333618, "learning_rate": 7.381191018339696e-05, "loss": 0.0143, "step": 3750 }, { "epoch": 7.91578947368421, "grad_norm": 0.30638623237609863, "learning_rate": 7.36663879488682e-05, "loss": 0.0134, "step": 3760 }, { "epoch": 7.936842105263158, "grad_norm": 0.35113340616226196, "learning_rate": 7.352060690251254e-05, "loss": 0.0166, "step": 3770 }, { "epoch": 7.957894736842105, "grad_norm": 0.49416929483413696, "learning_rate": 7.337456863856811e-05, "loss": 0.0188, "step": 3780 }, { "epoch": 7.978947368421053, "grad_norm": 0.2981952726840973, "learning_rate": 7.3228274754086e-05, "loss": 0.015, "step": 3790 }, { "epoch": 8.0, "grad_norm": 0.323319673538208, "learning_rate": 7.308172684891267e-05, "loss": 0.0161, "step": 3800 }, { "epoch": 8.021052631578947, "grad_norm": 0.37159842252731323, "learning_rate": 7.293492652567255e-05, "loss": 0.0137, "step": 3810 }, { "epoch": 8.042105263157895, "grad_norm": 0.48827266693115234, "learning_rate": 7.278787538975043e-05, "loss": 0.0151, "step": 3820 }, { "epoch": 8.063157894736841, "grad_norm": 0.2388990819454193, "learning_rate": 7.2640575049274e-05, "loss": 0.0156, "step": 3830 }, { "epoch": 8.08421052631579, "grad_norm": 0.271533727645874, "learning_rate": 7.249302711509616e-05, "loss": 0.0158, "step": 3840 }, { "epoch": 8.105263157894736, "grad_norm": 0.2839950621128082, "learning_rate": 7.23452332007775e-05, "loss": 0.0133, "step": 3850 }, { "epoch": 8.126315789473685, "grad_norm": 0.3670549988746643, "learning_rate": 7.219719492256858e-05, "loss": 0.0156, "step": 3860 }, { "epoch": 8.147368421052631, "grad_norm": 0.2501213848590851, "learning_rate": 7.20489138993923e-05, "loss": 0.0138, "step": 3870 }, { "epoch": 8.16842105263158, "grad_norm": 0.375550240278244, "learning_rate": 7.190039175282614e-05, "loss": 0.0166, "step": 3880 }, { "epoch": 8.189473684210526, "grad_norm": 0.2757289409637451, "learning_rate": 7.175163010708455e-05, "loss": 0.0139, "step": 3890 }, { "epoch": 8.210526315789474, "grad_norm": 0.3652801513671875, "learning_rate": 7.1602630589001e-05, "loss": 0.0163, "step": 3900 }, { "epoch": 8.23157894736842, "grad_norm": 0.23484884202480316, "learning_rate": 7.14533948280104e-05, "loss": 0.0192, "step": 3910 }, { "epoch": 8.25263157894737, "grad_norm": 0.2933192849159241, "learning_rate": 7.130392445613109e-05, "loss": 0.0173, "step": 3920 }, { "epoch": 8.273684210526316, "grad_norm": 0.4634689688682556, "learning_rate": 7.115422110794711e-05, "loss": 0.018, "step": 3930 }, { "epoch": 8.294736842105262, "grad_norm": 0.21663208305835724, "learning_rate": 7.100428642059033e-05, "loss": 0.0157, "step": 3940 }, { "epoch": 8.31578947368421, "grad_norm": 0.41367101669311523, "learning_rate": 7.08541220337224e-05, "loss": 0.0209, "step": 3950 }, { "epoch": 8.336842105263157, "grad_norm": 0.2638862133026123, "learning_rate": 7.070372958951706e-05, "loss": 0.0167, "step": 3960 }, { "epoch": 8.357894736842105, "grad_norm": 0.26983147859573364, "learning_rate": 7.055311073264194e-05, "loss": 0.0141, "step": 3970 }, { "epoch": 8.378947368421052, "grad_norm": 0.4866633117198944, "learning_rate": 7.040226711024077e-05, "loss": 0.0156, "step": 3980 }, { "epoch": 8.4, "grad_norm": 0.2690824270248413, "learning_rate": 7.02512003719152e-05, "loss": 0.0167, "step": 3990 }, { "epoch": 8.421052631578947, "grad_norm": 0.2695547640323639, "learning_rate": 7.00999121697069e-05, "loss": 0.0157, "step": 4000 }, { "epoch": 8.442105263157895, "grad_norm": 0.3926413655281067, "learning_rate": 6.99484041580794e-05, "loss": 0.0196, "step": 4010 }, { "epoch": 8.463157894736842, "grad_norm": 0.4354100227355957, "learning_rate": 6.979667799390004e-05, "loss": 0.0151, "step": 4020 }, { "epoch": 8.48421052631579, "grad_norm": 0.3218841254711151, "learning_rate": 6.964473533642185e-05, "loss": 0.0159, "step": 4030 }, { "epoch": 8.505263157894737, "grad_norm": 0.29377126693725586, "learning_rate": 6.949257784726539e-05, "loss": 0.0149, "step": 4040 }, { "epoch": 8.526315789473685, "grad_norm": 0.3483302891254425, "learning_rate": 6.934020719040056e-05, "loss": 0.013, "step": 4050 }, { "epoch": 8.547368421052632, "grad_norm": 0.2860410809516907, "learning_rate": 6.918762503212848e-05, "loss": 0.0165, "step": 4060 }, { "epoch": 8.568421052631578, "grad_norm": 0.2274821251630783, "learning_rate": 6.903483304106319e-05, "loss": 0.0137, "step": 4070 }, { "epoch": 8.589473684210526, "grad_norm": 0.33035004138946533, "learning_rate": 6.888183288811341e-05, "loss": 0.0157, "step": 4080 }, { "epoch": 8.610526315789473, "grad_norm": 0.44268885254859924, "learning_rate": 6.87286262464643e-05, "loss": 0.0148, "step": 4090 }, { "epoch": 8.631578947368421, "grad_norm": 0.3003886640071869, "learning_rate": 6.857521479155915e-05, "loss": 0.0178, "step": 4100 }, { "epoch": 8.652631578947368, "grad_norm": 0.2947339117527008, "learning_rate": 6.842160020108104e-05, "loss": 0.0161, "step": 4110 }, { "epoch": 8.673684210526316, "grad_norm": 0.31560686230659485, "learning_rate": 6.826778415493455e-05, "loss": 0.0154, "step": 4120 }, { "epoch": 8.694736842105263, "grad_norm": 0.26131710410118103, "learning_rate": 6.811376833522729e-05, "loss": 0.0162, "step": 4130 }, { "epoch": 8.715789473684211, "grad_norm": 0.3049681484699249, "learning_rate": 6.795955442625159e-05, "loss": 0.0223, "step": 4140 }, { "epoch": 8.736842105263158, "grad_norm": 0.25392788648605347, "learning_rate": 6.780514411446608e-05, "loss": 0.0153, "step": 4150 }, { "epoch": 8.757894736842106, "grad_norm": 0.20414528250694275, "learning_rate": 6.765053908847716e-05, "loss": 0.0127, "step": 4160 }, { "epoch": 8.778947368421052, "grad_norm": 0.30378004908561707, "learning_rate": 6.749574103902064e-05, "loss": 0.0181, "step": 4170 }, { "epoch": 8.8, "grad_norm": 0.4045424461364746, "learning_rate": 6.734075165894317e-05, "loss": 0.0184, "step": 4180 }, { "epoch": 8.821052631578947, "grad_norm": 0.31427350640296936, "learning_rate": 6.71855726431838e-05, "loss": 0.0153, "step": 4190 }, { "epoch": 8.842105263157894, "grad_norm": 0.25699537992477417, "learning_rate": 6.703020568875538e-05, "loss": 0.0177, "step": 4200 }, { "epoch": 8.863157894736842, "grad_norm": 0.38881710171699524, "learning_rate": 6.687465249472603e-05, "loss": 0.0147, "step": 4210 }, { "epoch": 8.884210526315789, "grad_norm": 0.37669476866722107, "learning_rate": 6.671891476220055e-05, "loss": 0.015, "step": 4220 }, { "epoch": 8.905263157894737, "grad_norm": 0.4347198009490967, "learning_rate": 6.656299419430183e-05, "loss": 0.0151, "step": 4230 }, { "epoch": 8.926315789473684, "grad_norm": 0.28435149788856506, "learning_rate": 6.640689249615223e-05, "loss": 0.0141, "step": 4240 }, { "epoch": 8.947368421052632, "grad_norm": 0.353750616312027, "learning_rate": 6.625061137485491e-05, "loss": 0.0165, "step": 4250 }, { "epoch": 8.968421052631578, "grad_norm": 0.2271672934293747, "learning_rate": 6.609415253947517e-05, "loss": 0.0124, "step": 4260 }, { "epoch": 8.989473684210527, "grad_norm": 0.36257851123809814, "learning_rate": 6.593751770102178e-05, "loss": 0.0162, "step": 4270 }, { "epoch": 9.010526315789473, "grad_norm": 0.3670196831226349, "learning_rate": 6.578070857242823e-05, "loss": 0.0142, "step": 4280 }, { "epoch": 9.031578947368422, "grad_norm": 0.36094874143600464, "learning_rate": 6.562372686853402e-05, "loss": 0.0132, "step": 4290 }, { "epoch": 9.052631578947368, "grad_norm": 0.2824079096317291, "learning_rate": 6.546657430606593e-05, "loss": 0.024, "step": 4300 }, { "epoch": 9.073684210526316, "grad_norm": 0.2186138927936554, "learning_rate": 6.530925260361918e-05, "loss": 0.0149, "step": 4310 }, { "epoch": 9.094736842105263, "grad_norm": 0.4528978765010834, "learning_rate": 6.515176348163871e-05, "loss": 0.015, "step": 4320 }, { "epoch": 9.115789473684211, "grad_norm": 0.4333273470401764, "learning_rate": 6.499410866240032e-05, "loss": 0.0156, "step": 4330 }, { "epoch": 9.136842105263158, "grad_norm": 0.26535049080848694, "learning_rate": 6.48362898699919e-05, "loss": 0.0163, "step": 4340 }, { "epoch": 9.157894736842104, "grad_norm": 0.36900678277015686, "learning_rate": 6.467830883029443e-05, "loss": 0.017, "step": 4350 }, { "epoch": 9.178947368421053, "grad_norm": 0.2458067238330841, "learning_rate": 6.452016727096326e-05, "loss": 0.0137, "step": 4360 }, { "epoch": 9.2, "grad_norm": 0.29793351888656616, "learning_rate": 6.436186692140916e-05, "loss": 0.0141, "step": 4370 }, { "epoch": 9.221052631578948, "grad_norm": 0.297637939453125, "learning_rate": 6.420340951277938e-05, "loss": 0.0144, "step": 4380 }, { "epoch": 9.242105263157894, "grad_norm": 0.285221666097641, "learning_rate": 6.404479677793874e-05, "loss": 0.0124, "step": 4390 }, { "epoch": 9.263157894736842, "grad_norm": 0.40709781646728516, "learning_rate": 6.388603045145075e-05, "loss": 0.0156, "step": 4400 }, { "epoch": 9.284210526315789, "grad_norm": 0.371254026889801, "learning_rate": 6.372711226955843e-05, "loss": 0.0123, "step": 4410 }, { "epoch": 9.305263157894737, "grad_norm": 0.1768479347229004, "learning_rate": 6.356804397016564e-05, "loss": 0.0197, "step": 4420 }, { "epoch": 9.326315789473684, "grad_norm": 0.30748406052589417, "learning_rate": 6.340882729281779e-05, "loss": 0.0171, "step": 4430 }, { "epoch": 9.347368421052632, "grad_norm": 0.22597071528434753, "learning_rate": 6.324946397868294e-05, "loss": 0.014, "step": 4440 }, { "epoch": 9.368421052631579, "grad_norm": 0.3074761629104614, "learning_rate": 6.308995577053276e-05, "loss": 0.014, "step": 4450 }, { "epoch": 9.389473684210527, "grad_norm": 0.5196573138237, "learning_rate": 6.293030441272347e-05, "loss": 0.0172, "step": 4460 }, { "epoch": 9.410526315789474, "grad_norm": 0.4442651569843292, "learning_rate": 6.277051165117677e-05, "loss": 0.0142, "step": 4470 }, { "epoch": 9.431578947368422, "grad_norm": 0.3239370584487915, "learning_rate": 6.261057923336064e-05, "loss": 0.0135, "step": 4480 }, { "epoch": 9.452631578947368, "grad_norm": 0.35525473952293396, "learning_rate": 6.245050890827042e-05, "loss": 0.0146, "step": 4490 }, { "epoch": 9.473684210526315, "grad_norm": 0.23829486966133118, "learning_rate": 6.229030242640952e-05, "loss": 0.0152, "step": 4500 }, { "epoch": 9.494736842105263, "grad_norm": 0.26216113567352295, "learning_rate": 6.212996153977037e-05, "loss": 0.0131, "step": 4510 }, { "epoch": 9.51578947368421, "grad_norm": 0.39340364933013916, "learning_rate": 6.196948800181523e-05, "loss": 0.0142, "step": 4520 }, { "epoch": 9.536842105263158, "grad_norm": 0.42373722791671753, "learning_rate": 6.180888356745695e-05, "loss": 0.0188, "step": 4530 }, { "epoch": 9.557894736842105, "grad_norm": 0.4443669319152832, "learning_rate": 6.164814999303995e-05, "loss": 0.0136, "step": 4540 }, { "epoch": 9.578947368421053, "grad_norm": 0.3581441044807434, "learning_rate": 6.148728903632081e-05, "loss": 0.0195, "step": 4550 }, { "epoch": 9.6, "grad_norm": 0.48290738463401794, "learning_rate": 6.132630245644921e-05, "loss": 0.0129, "step": 4560 }, { "epoch": 9.621052631578948, "grad_norm": 0.36384713649749756, "learning_rate": 6.116519201394857e-05, "loss": 0.0135, "step": 4570 }, { "epoch": 9.642105263157895, "grad_norm": 0.3014262020587921, "learning_rate": 6.10039594706969e-05, "loss": 0.0142, "step": 4580 }, { "epoch": 9.663157894736843, "grad_norm": 0.33191978931427, "learning_rate": 6.084260658990744e-05, "loss": 0.0121, "step": 4590 }, { "epoch": 9.68421052631579, "grad_norm": 0.2311135232448578, "learning_rate": 6.068113513610943e-05, "loss": 0.0127, "step": 4600 }, { "epoch": 9.705263157894738, "grad_norm": 0.3405277132987976, "learning_rate": 6.0519546875128876e-05, "loss": 0.0167, "step": 4610 }, { "epoch": 9.726315789473684, "grad_norm": 0.25871551036834717, "learning_rate": 6.035784357406906e-05, "loss": 0.0157, "step": 4620 }, { "epoch": 9.74736842105263, "grad_norm": 0.3758455812931061, "learning_rate": 6.01960270012914e-05, "loss": 0.0162, "step": 4630 }, { "epoch": 9.76842105263158, "grad_norm": 0.27979034185409546, "learning_rate": 6.003409892639599e-05, "loss": 0.0149, "step": 4640 }, { "epoch": 9.789473684210526, "grad_norm": 0.3277779817581177, "learning_rate": 5.9872061120202336e-05, "loss": 0.0134, "step": 4650 }, { "epoch": 9.810526315789474, "grad_norm": 0.33733683824539185, "learning_rate": 5.9709915354729914e-05, "loss": 0.0132, "step": 4660 }, { "epoch": 9.83157894736842, "grad_norm": 0.40389183163642883, "learning_rate": 5.9547663403178824e-05, "loss": 0.014, "step": 4670 }, { "epoch": 9.852631578947369, "grad_norm": 0.3807900846004486, "learning_rate": 5.9385307039910445e-05, "loss": 0.0146, "step": 4680 }, { "epoch": 9.873684210526315, "grad_norm": 0.2574458718299866, "learning_rate": 5.922284804042792e-05, "loss": 0.0159, "step": 4690 }, { "epoch": 9.894736842105264, "grad_norm": 0.2770082950592041, "learning_rate": 5.906028818135687e-05, "loss": 0.0143, "step": 4700 }, { "epoch": 9.91578947368421, "grad_norm": 0.23428484797477722, "learning_rate": 5.889762924042585e-05, "loss": 0.0127, "step": 4710 }, { "epoch": 9.936842105263159, "grad_norm": 0.2577557861804962, "learning_rate": 5.873487299644699e-05, "loss": 0.0134, "step": 4720 }, { "epoch": 9.957894736842105, "grad_norm": 0.33987632393836975, "learning_rate": 5.857202122929649e-05, "loss": 0.0143, "step": 4730 }, { "epoch": 9.978947368421053, "grad_norm": 0.39755430817604065, "learning_rate": 5.840907571989518e-05, "loss": 0.0142, "step": 4740 }, { "epoch": 10.0, "grad_norm": 0.381356418132782, "learning_rate": 5.824603825018904e-05, "loss": 0.0134, "step": 4750 }, { "epoch": 10.021052631578947, "grad_norm": 0.5197859406471252, "learning_rate": 5.808291060312975e-05, "loss": 0.0148, "step": 4760 }, { "epoch": 10.042105263157895, "grad_norm": 0.3720296025276184, "learning_rate": 5.7919694562655083e-05, "loss": 0.0152, "step": 4770 }, { "epoch": 10.063157894736841, "grad_norm": 0.24121274054050446, "learning_rate": 5.775639191366954e-05, "loss": 0.0142, "step": 4780 }, { "epoch": 10.08421052631579, "grad_norm": 0.26737508177757263, "learning_rate": 5.75930044420247e-05, "loss": 0.0129, "step": 4790 }, { "epoch": 10.105263157894736, "grad_norm": 0.22389690577983856, "learning_rate": 5.74295339344998e-05, "loss": 0.0136, "step": 4800 }, { "epoch": 10.126315789473685, "grad_norm": 0.28643742203712463, "learning_rate": 5.726598217878211e-05, "loss": 0.0142, "step": 4810 }, { "epoch": 10.147368421052631, "grad_norm": 0.3776261508464813, "learning_rate": 5.71023509634474e-05, "loss": 0.0117, "step": 4820 }, { "epoch": 10.16842105263158, "grad_norm": 0.3595387041568756, "learning_rate": 5.693864207794049e-05, "loss": 0.0138, "step": 4830 }, { "epoch": 10.189473684210526, "grad_norm": 0.3222644031047821, "learning_rate": 5.677485731255545e-05, "loss": 0.017, "step": 4840 }, { "epoch": 10.210526315789474, "grad_norm": 0.32698315382003784, "learning_rate": 5.6610998458416296e-05, "loss": 0.012, "step": 4850 }, { "epoch": 10.23157894736842, "grad_norm": 0.2258709967136383, "learning_rate": 5.644706730745716e-05, "loss": 0.0135, "step": 4860 }, { "epoch": 10.25263157894737, "grad_norm": 0.44267264008522034, "learning_rate": 5.628306565240287e-05, "loss": 0.0113, "step": 4870 }, { "epoch": 10.273684210526316, "grad_norm": 0.250570148229599, "learning_rate": 5.611899528674923e-05, "loss": 0.0109, "step": 4880 }, { "epoch": 10.294736842105262, "grad_norm": 0.2860018014907837, "learning_rate": 5.595485800474349e-05, "loss": 0.0118, "step": 4890 }, { "epoch": 10.31578947368421, "grad_norm": 0.4406115412712097, "learning_rate": 5.579065560136467e-05, "loss": 0.0145, "step": 4900 }, { "epoch": 10.336842105263157, "grad_norm": 0.3966688811779022, "learning_rate": 5.562638987230392e-05, "loss": 0.013, "step": 4910 }, { "epoch": 10.357894736842105, "grad_norm": 0.48601025342941284, "learning_rate": 5.546206261394498e-05, "loss": 0.018, "step": 4920 }, { "epoch": 10.378947368421052, "grad_norm": 0.38568636775016785, "learning_rate": 5.529767562334437e-05, "loss": 0.0107, "step": 4930 }, { "epoch": 10.4, "grad_norm": 0.360707551240921, "learning_rate": 5.5133230698211926e-05, "loss": 0.0127, "step": 4940 }, { "epoch": 10.421052631578947, "grad_norm": 0.3719407320022583, "learning_rate": 5.496872963689096e-05, "loss": 0.0127, "step": 4950 }, { "epoch": 10.442105263157895, "grad_norm": 0.2505301833152771, "learning_rate": 5.4804174238338756e-05, "loss": 0.0148, "step": 4960 }, { "epoch": 10.463157894736842, "grad_norm": 0.29948848485946655, "learning_rate": 5.463956630210678e-05, "loss": 0.0146, "step": 4970 }, { "epoch": 10.48421052631579, "grad_norm": 0.4698536992073059, "learning_rate": 5.4474907628321046e-05, "loss": 0.0173, "step": 4980 }, { "epoch": 10.505263157894737, "grad_norm": 0.3408210575580597, "learning_rate": 5.431020001766244e-05, "loss": 0.0137, "step": 4990 }, { "epoch": 10.526315789473685, "grad_norm": 0.2353430837392807, "learning_rate": 5.4145445271346986e-05, "loss": 0.0127, "step": 5000 }, { "epoch": 10.547368421052632, "grad_norm": 0.31311601400375366, "learning_rate": 5.398064519110622e-05, "loss": 0.0144, "step": 5010 }, { "epoch": 10.568421052631578, "grad_norm": 0.3829793632030487, "learning_rate": 5.3815801579167394e-05, "loss": 0.0142, "step": 5020 }, { "epoch": 10.589473684210526, "grad_norm": 0.33505240082740784, "learning_rate": 5.365091623823382e-05, "loss": 0.0127, "step": 5030 }, { "epoch": 10.610526315789473, "grad_norm": 0.27066004276275635, "learning_rate": 5.348599097146521e-05, "loss": 0.0131, "step": 5040 }, { "epoch": 10.631578947368421, "grad_norm": 0.25077617168426514, "learning_rate": 5.3321027582457836e-05, "loss": 0.0133, "step": 5050 }, { "epoch": 10.652631578947368, "grad_norm": 0.24700289964675903, "learning_rate": 5.315602787522491e-05, "loss": 0.0149, "step": 5060 }, { "epoch": 10.673684210526316, "grad_norm": 0.3117201030254364, "learning_rate": 5.299099365417678e-05, "loss": 0.0128, "step": 5070 }, { "epoch": 10.694736842105263, "grad_norm": 0.34067851305007935, "learning_rate": 5.2825926724101236e-05, "loss": 0.0168, "step": 5080 }, { "epoch": 10.715789473684211, "grad_norm": 0.3709547817707062, "learning_rate": 5.26608288901438e-05, "loss": 0.0167, "step": 5090 }, { "epoch": 10.736842105263158, "grad_norm": 0.2576051354408264, "learning_rate": 5.24957019577879e-05, "loss": 0.0143, "step": 5100 }, { "epoch": 10.757894736842106, "grad_norm": 0.36624500155448914, "learning_rate": 5.2330547732835266e-05, "loss": 0.0122, "step": 5110 }, { "epoch": 10.778947368421052, "grad_norm": 0.28927966952323914, "learning_rate": 5.2165368021385996e-05, "loss": 0.0132, "step": 5120 }, { "epoch": 10.8, "grad_norm": 0.3085300624370575, "learning_rate": 5.200016462981897e-05, "loss": 0.0135, "step": 5130 }, { "epoch": 10.821052631578947, "grad_norm": 0.3308599889278412, "learning_rate": 5.1834939364772015e-05, "loss": 0.0144, "step": 5140 }, { "epoch": 10.842105263157894, "grad_norm": 0.26345300674438477, "learning_rate": 5.166969403312214e-05, "loss": 0.014, "step": 5150 }, { "epoch": 10.863157894736842, "grad_norm": 0.2755007743835449, "learning_rate": 5.1504430441965844e-05, "loss": 0.0128, "step": 5160 }, { "epoch": 10.884210526315789, "grad_norm": 0.2243039906024933, "learning_rate": 5.133915039859923e-05, "loss": 0.0118, "step": 5170 }, { "epoch": 10.905263157894737, "grad_norm": 0.3064967095851898, "learning_rate": 5.1173855710498444e-05, "loss": 0.0162, "step": 5180 }, { "epoch": 10.926315789473684, "grad_norm": 0.16886362433433533, "learning_rate": 5.100854818529967e-05, "loss": 0.0125, "step": 5190 }, { "epoch": 10.947368421052632, "grad_norm": 0.2918197810649872, "learning_rate": 5.084322963077951e-05, "loss": 0.0187, "step": 5200 }, { "epoch": 10.968421052631578, "grad_norm": 0.1947583109140396, "learning_rate": 5.067790185483522e-05, "loss": 0.0124, "step": 5210 }, { "epoch": 10.989473684210527, "grad_norm": 0.49756118655204773, "learning_rate": 5.0512566665464844e-05, "loss": 0.0128, "step": 5220 }, { "epoch": 11.010526315789473, "grad_norm": 0.20627428591251373, "learning_rate": 5.034722587074755e-05, "loss": 0.0174, "step": 5230 }, { "epoch": 11.031578947368422, "grad_norm": 0.23628760874271393, "learning_rate": 5.018188127882375e-05, "loss": 0.0115, "step": 5240 }, { "epoch": 11.052631578947368, "grad_norm": 0.1982555091381073, "learning_rate": 5.0016534697875417e-05, "loss": 0.0146, "step": 5250 }, { "epoch": 11.073684210526316, "grad_norm": 0.22756166756153107, "learning_rate": 4.9851187936106294e-05, "loss": 0.013, "step": 5260 }, { "epoch": 11.094736842105263, "grad_norm": 0.3595881164073944, "learning_rate": 4.968584280172206e-05, "loss": 0.0124, "step": 5270 }, { "epoch": 11.115789473684211, "grad_norm": 0.32685622572898865, "learning_rate": 4.95205011029106e-05, "loss": 0.0131, "step": 5280 }, { "epoch": 11.136842105263158, "grad_norm": 0.2763034999370575, "learning_rate": 4.935516464782227e-05, "loss": 0.0119, "step": 5290 }, { "epoch": 11.157894736842104, "grad_norm": 0.28238409757614136, "learning_rate": 4.918983524455003e-05, "loss": 0.0125, "step": 5300 }, { "epoch": 11.178947368421053, "grad_norm": 0.23236745595932007, "learning_rate": 4.9024514701109766e-05, "loss": 0.0117, "step": 5310 }, { "epoch": 11.2, "grad_norm": 0.3133922517299652, "learning_rate": 4.885920482542043e-05, "loss": 0.012, "step": 5320 }, { "epoch": 11.221052631578948, "grad_norm": 0.3018359839916229, "learning_rate": 4.869390742528438e-05, "loss": 0.0137, "step": 5330 }, { "epoch": 11.242105263157894, "grad_norm": 0.3999684154987335, "learning_rate": 4.852862430836744e-05, "loss": 0.0101, "step": 5340 }, { "epoch": 11.263157894736842, "grad_norm": 0.35241612792015076, "learning_rate": 4.836335728217933e-05, "loss": 0.0132, "step": 5350 }, { "epoch": 11.284210526315789, "grad_norm": 0.4391363263130188, "learning_rate": 4.819810815405379e-05, "loss": 0.0106, "step": 5360 }, { "epoch": 11.305263157894737, "grad_norm": 0.3460538685321808, "learning_rate": 4.803287873112877e-05, "loss": 0.0132, "step": 5370 }, { "epoch": 11.326315789473684, "grad_norm": 0.39460355043411255, "learning_rate": 4.786767082032681e-05, "loss": 0.0107, "step": 5380 }, { "epoch": 11.347368421052632, "grad_norm": 0.2866694927215576, "learning_rate": 4.77024862283351e-05, "loss": 0.0119, "step": 5390 }, { "epoch": 11.368421052631579, "grad_norm": 0.3136466443538666, "learning_rate": 4.753732676158593e-05, "loss": 0.0137, "step": 5400 }, { "epoch": 11.389473684210527, "grad_norm": 0.25927698612213135, "learning_rate": 4.737219422623672e-05, "loss": 0.0117, "step": 5410 }, { "epoch": 11.410526315789474, "grad_norm": 0.19059674441814423, "learning_rate": 4.720709042815044e-05, "loss": 0.0106, "step": 5420 }, { "epoch": 11.431578947368422, "grad_norm": 0.30297887325286865, "learning_rate": 4.704201717287578e-05, "loss": 0.0123, "step": 5430 }, { "epoch": 11.452631578947368, "grad_norm": 0.30885186791419983, "learning_rate": 4.6876976265627404e-05, "loss": 0.0128, "step": 5440 }, { "epoch": 11.473684210526315, "grad_norm": 0.2597584128379822, "learning_rate": 4.671196951126626e-05, "loss": 0.0125, "step": 5450 }, { "epoch": 11.494736842105263, "grad_norm": 0.29008573293685913, "learning_rate": 4.654699871427971e-05, "loss": 0.0115, "step": 5460 }, { "epoch": 11.51578947368421, "grad_norm": 0.25596117973327637, "learning_rate": 4.6382065678762034e-05, "loss": 0.0114, "step": 5470 }, { "epoch": 11.536842105263158, "grad_norm": 0.3098960816860199, "learning_rate": 4.6217172208394424e-05, "loss": 0.013, "step": 5480 }, { "epoch": 11.557894736842105, "grad_norm": 0.3481617867946625, "learning_rate": 4.605232010642549e-05, "loss": 0.0141, "step": 5490 }, { "epoch": 11.578947368421053, "grad_norm": 0.2176370471715927, "learning_rate": 4.588751117565142e-05, "loss": 0.0131, "step": 5500 }, { "epoch": 11.6, "grad_norm": 0.20991674065589905, "learning_rate": 4.5722747218396214e-05, "loss": 0.0138, "step": 5510 }, { "epoch": 11.621052631578948, "grad_norm": 0.33867937326431274, "learning_rate": 4.5558030036492194e-05, "loss": 0.0123, "step": 5520 }, { "epoch": 11.642105263157895, "grad_norm": 0.37924477458000183, "learning_rate": 4.539336143125999e-05, "loss": 0.0127, "step": 5530 }, { "epoch": 11.663157894736843, "grad_norm": 0.3857584595680237, "learning_rate": 4.522874320348916e-05, "loss": 0.0113, "step": 5540 }, { "epoch": 11.68421052631579, "grad_norm": 0.35021325945854187, "learning_rate": 4.506417715341821e-05, "loss": 0.012, "step": 5550 }, { "epoch": 11.705263157894738, "grad_norm": 0.21999479830265045, "learning_rate": 4.489966508071511e-05, "loss": 0.014, "step": 5560 }, { "epoch": 11.726315789473684, "grad_norm": 0.29077979922294617, "learning_rate": 4.4735208784457575e-05, "loss": 0.0145, "step": 5570 }, { "epoch": 11.74736842105263, "grad_norm": 0.34818798303604126, "learning_rate": 4.457081006311325e-05, "loss": 0.0126, "step": 5580 }, { "epoch": 11.76842105263158, "grad_norm": 0.300442636013031, "learning_rate": 4.440647071452027e-05, "loss": 0.0156, "step": 5590 }, { "epoch": 11.789473684210526, "grad_norm": 0.2729852497577667, "learning_rate": 4.424219253586737e-05, "loss": 0.0105, "step": 5600 }, { "epoch": 11.810526315789474, "grad_norm": 0.24548235535621643, "learning_rate": 4.407797732367443e-05, "loss": 0.0119, "step": 5610 }, { "epoch": 11.83157894736842, "grad_norm": 0.28283464908599854, "learning_rate": 4.391382687377268e-05, "loss": 0.0109, "step": 5620 }, { "epoch": 11.852631578947369, "grad_norm": 0.26903975009918213, "learning_rate": 4.374974298128512e-05, "loss": 0.0123, "step": 5630 }, { "epoch": 11.873684210526315, "grad_norm": 0.26440849900245667, "learning_rate": 4.358572744060699e-05, "loss": 0.0112, "step": 5640 }, { "epoch": 11.894736842105264, "grad_norm": 0.40656521916389465, "learning_rate": 4.342178204538588e-05, "loss": 0.0108, "step": 5650 }, { "epoch": 11.91578947368421, "grad_norm": 0.2077610045671463, "learning_rate": 4.325790858850241e-05, "loss": 0.0138, "step": 5660 }, { "epoch": 11.936842105263159, "grad_norm": 0.29773902893066406, "learning_rate": 4.309410886205043e-05, "loss": 0.0104, "step": 5670 }, { "epoch": 11.957894736842105, "grad_norm": 0.29168128967285156, "learning_rate": 4.293038465731752e-05, "loss": 0.013, "step": 5680 }, { "epoch": 11.978947368421053, "grad_norm": 0.27413210272789, "learning_rate": 4.276673776476533e-05, "loss": 0.0117, "step": 5690 }, { "epoch": 12.0, "grad_norm": 0.2525549530982971, "learning_rate": 4.260316997401007e-05, "loss": 0.012, "step": 5700 }, { "epoch": 12.021052631578947, "grad_norm": 0.2853563129901886, "learning_rate": 4.243968307380293e-05, "loss": 0.011, "step": 5710 }, { "epoch": 12.042105263157895, "grad_norm": 0.30066773295402527, "learning_rate": 4.22762788520104e-05, "loss": 0.0105, "step": 5720 }, { "epoch": 12.063157894736841, "grad_norm": 0.31023097038269043, "learning_rate": 4.211295909559491e-05, "loss": 0.0104, "step": 5730 }, { "epoch": 12.08421052631579, "grad_norm": 0.27115997672080994, "learning_rate": 4.194972559059511e-05, "loss": 0.0096, "step": 5740 }, { "epoch": 12.105263157894736, "grad_norm": 0.314968079328537, "learning_rate": 4.178658012210651e-05, "loss": 0.0134, "step": 5750 }, { "epoch": 12.126315789473685, "grad_norm": 0.3072914779186249, "learning_rate": 4.162352447426177e-05, "loss": 0.0114, "step": 5760 }, { "epoch": 12.147368421052631, "grad_norm": 0.22562859952449799, "learning_rate": 4.146056043021135e-05, "loss": 0.0108, "step": 5770 }, { "epoch": 12.16842105263158, "grad_norm": 0.29691118001937866, "learning_rate": 4.1297689772103944e-05, "loss": 0.0131, "step": 5780 }, { "epoch": 12.189473684210526, "grad_norm": 0.25910550355911255, "learning_rate": 4.113491428106694e-05, "loss": 0.0113, "step": 5790 }, { "epoch": 12.210526315789474, "grad_norm": 0.24949148297309875, "learning_rate": 4.0972235737187055e-05, "loss": 0.0168, "step": 5800 }, { "epoch": 12.23157894736842, "grad_norm": 0.22757770121097565, "learning_rate": 4.080965591949076e-05, "loss": 0.0117, "step": 5810 }, { "epoch": 12.25263157894737, "grad_norm": 0.18142646551132202, "learning_rate": 4.0647176605924924e-05, "loss": 0.0101, "step": 5820 }, { "epoch": 12.273684210526316, "grad_norm": 0.33148688077926636, "learning_rate": 4.0484799573337255e-05, "loss": 0.0122, "step": 5830 }, { "epoch": 12.294736842105262, "grad_norm": 0.37832051515579224, "learning_rate": 4.032252659745699e-05, "loss": 0.0147, "step": 5840 }, { "epoch": 12.31578947368421, "grad_norm": 0.26507580280303955, "learning_rate": 4.016035945287539e-05, "loss": 0.0163, "step": 5850 }, { "epoch": 12.336842105263157, "grad_norm": 0.23577500879764557, "learning_rate": 3.999829991302635e-05, "loss": 0.0107, "step": 5860 }, { "epoch": 12.357894736842105, "grad_norm": 0.1889563798904419, "learning_rate": 3.983634975016707e-05, "loss": 0.0108, "step": 5870 }, { "epoch": 12.378947368421052, "grad_norm": 0.3043462932109833, "learning_rate": 3.967451073535854e-05, "loss": 0.0132, "step": 5880 }, { "epoch": 12.4, "grad_norm": 0.26580101251602173, "learning_rate": 3.951278463844633e-05, "loss": 0.0136, "step": 5890 }, { "epoch": 12.421052631578947, "grad_norm": 0.3457507789134979, "learning_rate": 3.935117322804111e-05, "loss": 0.0109, "step": 5900 }, { "epoch": 12.442105263157895, "grad_norm": 0.3290192782878876, "learning_rate": 3.918967827149938e-05, "loss": 0.0119, "step": 5910 }, { "epoch": 12.463157894736842, "grad_norm": 0.300097793340683, "learning_rate": 3.9028301534904094e-05, "loss": 0.0125, "step": 5920 }, { "epoch": 12.48421052631579, "grad_norm": 0.21364746987819672, "learning_rate": 3.88670447830454e-05, "loss": 0.0108, "step": 5930 }, { "epoch": 12.505263157894737, "grad_norm": 0.21573108434677124, "learning_rate": 3.870590977940132e-05, "loss": 0.0127, "step": 5940 }, { "epoch": 12.526315789473685, "grad_norm": 0.3635398745536804, "learning_rate": 3.8544898286118404e-05, "loss": 0.0112, "step": 5950 }, { "epoch": 12.547368421052632, "grad_norm": 0.30178016424179077, "learning_rate": 3.838401206399257e-05, "loss": 0.0107, "step": 5960 }, { "epoch": 12.568421052631578, "grad_norm": 0.278812438249588, "learning_rate": 3.822325287244975e-05, "loss": 0.012, "step": 5970 }, { "epoch": 12.589473684210526, "grad_norm": 0.34470126032829285, "learning_rate": 3.8062622469526725e-05, "loss": 0.0119, "step": 5980 }, { "epoch": 12.610526315789473, "grad_norm": 0.38831329345703125, "learning_rate": 3.790212261185183e-05, "loss": 0.0143, "step": 5990 }, { "epoch": 12.631578947368421, "grad_norm": 0.34272217750549316, "learning_rate": 3.7741755054625794e-05, "loss": 0.0107, "step": 6000 }, { "epoch": 12.652631578947368, "grad_norm": 0.2135208249092102, "learning_rate": 3.758152155160255e-05, "loss": 0.0106, "step": 6010 }, { "epoch": 12.673684210526316, "grad_norm": 0.271282821893692, "learning_rate": 3.742142385506999e-05, "loss": 0.0101, "step": 6020 }, { "epoch": 12.694736842105263, "grad_norm": 0.26926925778388977, "learning_rate": 3.72614637158309e-05, "loss": 0.0109, "step": 6030 }, { "epoch": 12.715789473684211, "grad_norm": 0.29975637793540955, "learning_rate": 3.710164288318371e-05, "loss": 0.0115, "step": 6040 }, { "epoch": 12.736842105263158, "grad_norm": 0.1963178515434265, "learning_rate": 3.694196310490345e-05, "loss": 0.0125, "step": 6050 }, { "epoch": 12.757894736842106, "grad_norm": 0.17428891360759735, "learning_rate": 3.678242612722259e-05, "loss": 0.0107, "step": 6060 }, { "epoch": 12.778947368421052, "grad_norm": 0.20561233162879944, "learning_rate": 3.6623033694811953e-05, "loss": 0.01, "step": 6070 }, { "epoch": 12.8, "grad_norm": 0.3465796709060669, "learning_rate": 3.6463787550761665e-05, "loss": 0.0109, "step": 6080 }, { "epoch": 12.821052631578947, "grad_norm": 0.3859802782535553, "learning_rate": 3.630468943656202e-05, "loss": 0.0099, "step": 6090 }, { "epoch": 12.842105263157894, "grad_norm": 0.39540693163871765, "learning_rate": 3.6145741092084523e-05, "loss": 0.0126, "step": 6100 }, { "epoch": 12.863157894736842, "grad_norm": 0.19579145312309265, "learning_rate": 3.598694425556278e-05, "loss": 0.0123, "step": 6110 }, { "epoch": 12.884210526315789, "grad_norm": 0.27704334259033203, "learning_rate": 3.58283006635736e-05, "loss": 0.0142, "step": 6120 }, { "epoch": 12.905263157894737, "grad_norm": 0.2573539614677429, "learning_rate": 3.566981205101781e-05, "loss": 0.0107, "step": 6130 }, { "epoch": 12.926315789473684, "grad_norm": 0.2780447006225586, "learning_rate": 3.5511480151101556e-05, "loss": 0.0117, "step": 6140 }, { "epoch": 12.947368421052632, "grad_norm": 0.30854132771492004, "learning_rate": 3.5353306695317104e-05, "loss": 0.0155, "step": 6150 }, { "epoch": 12.968421052631578, "grad_norm": 0.24485301971435547, "learning_rate": 3.519529341342402e-05, "loss": 0.0099, "step": 6160 }, { "epoch": 12.989473684210527, "grad_norm": 0.2743643522262573, "learning_rate": 3.503744203343026e-05, "loss": 0.0102, "step": 6170 }, { "epoch": 13.010526315789473, "grad_norm": 0.37031930685043335, "learning_rate": 3.487975428157318e-05, "loss": 0.0116, "step": 6180 }, { "epoch": 13.031578947368422, "grad_norm": 0.2035238891839981, "learning_rate": 3.472223188230083e-05, "loss": 0.0092, "step": 6190 }, { "epoch": 13.052631578947368, "grad_norm": 0.25985631346702576, "learning_rate": 3.4564876558252866e-05, "loss": 0.0134, "step": 6200 }, { "epoch": 13.073684210526316, "grad_norm": 0.27489617466926575, "learning_rate": 3.440769003024195e-05, "loss": 0.011, "step": 6210 }, { "epoch": 13.094736842105263, "grad_norm": 0.3478982746601105, "learning_rate": 3.425067401723477e-05, "loss": 0.0109, "step": 6220 }, { "epoch": 13.115789473684211, "grad_norm": 0.2666727602481842, "learning_rate": 3.409383023633325e-05, "loss": 0.0118, "step": 6230 }, { "epoch": 13.136842105263158, "grad_norm": 0.1723157912492752, "learning_rate": 3.3937160402755894e-05, "loss": 0.0092, "step": 6240 }, { "epoch": 13.157894736842104, "grad_norm": 0.20668773353099823, "learning_rate": 3.378066622981885e-05, "loss": 0.0111, "step": 6250 }, { "epoch": 13.178947368421053, "grad_norm": 0.251992791891098, "learning_rate": 3.362434942891738e-05, "loss": 0.0123, "step": 6260 }, { "epoch": 13.2, "grad_norm": 0.36864086985588074, "learning_rate": 3.346821170950693e-05, "loss": 0.0152, "step": 6270 }, { "epoch": 13.221052631578948, "grad_norm": 0.3657408356666565, "learning_rate": 3.3312254779084585e-05, "loss": 0.0115, "step": 6280 }, { "epoch": 13.242105263157894, "grad_norm": 0.44476518034935, "learning_rate": 3.315648034317039e-05, "loss": 0.009, "step": 6290 }, { "epoch": 13.263157894736842, "grad_norm": 0.28683358430862427, "learning_rate": 3.3000890105288564e-05, "loss": 0.0098, "step": 6300 }, { "epoch": 13.284210526315789, "grad_norm": 0.39317336678504944, "learning_rate": 3.284548576694908e-05, "loss": 0.0104, "step": 6310 }, { "epoch": 13.305263157894737, "grad_norm": 0.1946222186088562, "learning_rate": 3.2690269027628815e-05, "loss": 0.0106, "step": 6320 }, { "epoch": 13.326315789473684, "grad_norm": 0.2779843509197235, "learning_rate": 3.253524158475324e-05, "loss": 0.0111, "step": 6330 }, { "epoch": 13.347368421052632, "grad_norm": 0.262795090675354, "learning_rate": 3.238040513367757e-05, "loss": 0.0106, "step": 6340 }, { "epoch": 13.368421052631579, "grad_norm": 0.4126349687576294, "learning_rate": 3.222576136766843e-05, "loss": 0.0138, "step": 6350 }, { "epoch": 13.389473684210527, "grad_norm": 0.23619171977043152, "learning_rate": 3.2071311977885324e-05, "loss": 0.0114, "step": 6360 }, { "epoch": 13.410526315789474, "grad_norm": 0.14706963300704956, "learning_rate": 3.191705865336197e-05, "loss": 0.011, "step": 6370 }, { "epoch": 13.431578947368422, "grad_norm": 0.22374007105827332, "learning_rate": 3.1763003080988075e-05, "loss": 0.0102, "step": 6380 }, { "epoch": 13.452631578947368, "grad_norm": 0.35418787598609924, "learning_rate": 3.160914694549063e-05, "loss": 0.0147, "step": 6390 }, { "epoch": 13.473684210526315, "grad_norm": 0.24454770982265472, "learning_rate": 3.145549192941573e-05, "loss": 0.0103, "step": 6400 }, { "epoch": 13.494736842105263, "grad_norm": 0.21057283878326416, "learning_rate": 3.130203971310999e-05, "loss": 0.008, "step": 6410 }, { "epoch": 13.51578947368421, "grad_norm": 0.22345757484436035, "learning_rate": 3.114879197470225e-05, "loss": 0.0116, "step": 6420 }, { "epoch": 13.536842105263158, "grad_norm": 0.18965598940849304, "learning_rate": 3.0995750390085285e-05, "loss": 0.009, "step": 6430 }, { "epoch": 13.557894736842105, "grad_norm": 0.1750851571559906, "learning_rate": 3.084291663289728e-05, "loss": 0.0108, "step": 6440 }, { "epoch": 13.578947368421053, "grad_norm": 0.2309773713350296, "learning_rate": 3.069029237450375e-05, "loss": 0.0098, "step": 6450 }, { "epoch": 13.6, "grad_norm": 0.2105448991060257, "learning_rate": 3.053787928397911e-05, "loss": 0.0108, "step": 6460 }, { "epoch": 13.621052631578948, "grad_norm": 0.2749241590499878, "learning_rate": 3.0385679028088526e-05, "loss": 0.0091, "step": 6470 }, { "epoch": 13.642105263157895, "grad_norm": 0.31380677223205566, "learning_rate": 3.023369327126959e-05, "loss": 0.0111, "step": 6480 }, { "epoch": 13.663157894736843, "grad_norm": 0.24529504776000977, "learning_rate": 3.0081923675614198e-05, "loss": 0.0112, "step": 6490 }, { "epoch": 13.68421052631579, "grad_norm": 0.44318637251853943, "learning_rate": 2.993037190085034e-05, "loss": 0.0134, "step": 6500 }, { "epoch": 13.705263157894738, "grad_norm": 0.2532472014427185, "learning_rate": 2.977903960432392e-05, "loss": 0.0106, "step": 6510 }, { "epoch": 13.726315789473684, "grad_norm": 0.2415039986371994, "learning_rate": 2.9627928440980722e-05, "loss": 0.0108, "step": 6520 }, { "epoch": 13.74736842105263, "grad_norm": 0.1994156837463379, "learning_rate": 2.9477040063348183e-05, "loss": 0.0108, "step": 6530 }, { "epoch": 13.76842105263158, "grad_norm": 0.19320830702781677, "learning_rate": 2.9326376121517456e-05, "loss": 0.0089, "step": 6540 }, { "epoch": 13.789473684210526, "grad_norm": 0.29363948106765747, "learning_rate": 2.9175938263125236e-05, "loss": 0.011, "step": 6550 }, { "epoch": 13.810526315789474, "grad_norm": 0.17705342173576355, "learning_rate": 2.9025728133335873e-05, "loss": 0.0098, "step": 6560 }, { "epoch": 13.83157894736842, "grad_norm": 0.34992334246635437, "learning_rate": 2.8875747374823288e-05, "loss": 0.0121, "step": 6570 }, { "epoch": 13.852631578947369, "grad_norm": 0.2314222902059555, "learning_rate": 2.872599762775298e-05, "loss": 0.0091, "step": 6580 }, { "epoch": 13.873684210526315, "grad_norm": 0.4391345977783203, "learning_rate": 2.857648052976425e-05, "loss": 0.0107, "step": 6590 }, { "epoch": 13.894736842105264, "grad_norm": 0.33117741346359253, "learning_rate": 2.8427197715952047e-05, "loss": 0.0108, "step": 6600 }, { "epoch": 13.91578947368421, "grad_norm": 0.21777181327342987, "learning_rate": 2.8278150818849393e-05, "loss": 0.0117, "step": 6610 }, { "epoch": 13.936842105263159, "grad_norm": 0.2570987641811371, "learning_rate": 2.812934146840922e-05, "loss": 0.01, "step": 6620 }, { "epoch": 13.957894736842105, "grad_norm": 0.36871713399887085, "learning_rate": 2.7980771291986764e-05, "loss": 0.0111, "step": 6630 }, { "epoch": 13.978947368421053, "grad_norm": 0.27420464158058167, "learning_rate": 2.783244191432167e-05, "loss": 0.0097, "step": 6640 }, { "epoch": 14.0, "grad_norm": 0.23203907907009125, "learning_rate": 2.768435495752022e-05, "loss": 0.0102, "step": 6650 }, { "epoch": 14.021052631578947, "grad_norm": 0.25504323840141296, "learning_rate": 2.753651204103771e-05, "loss": 0.0111, "step": 6660 }, { "epoch": 14.042105263157895, "grad_norm": 0.24600966274738312, "learning_rate": 2.7388914781660523e-05, "loss": 0.0098, "step": 6670 }, { "epoch": 14.063157894736841, "grad_norm": 0.3143404424190521, "learning_rate": 2.7241564793488693e-05, "loss": 0.0111, "step": 6680 }, { "epoch": 14.08421052631579, "grad_norm": 0.40884312987327576, "learning_rate": 2.7094463687918037e-05, "loss": 0.0109, "step": 6690 }, { "epoch": 14.105263157894736, "grad_norm": 0.1468818634748459, "learning_rate": 2.694761307362268e-05, "loss": 0.0104, "step": 6700 }, { "epoch": 14.126315789473685, "grad_norm": 0.18425166606903076, "learning_rate": 2.6801014556537467e-05, "loss": 0.0089, "step": 6710 }, { "epoch": 14.147368421052631, "grad_norm": 0.2821352481842041, "learning_rate": 2.6654669739840243e-05, "loss": 0.0091, "step": 6720 }, { "epoch": 14.16842105263158, "grad_norm": 0.18080680072307587, "learning_rate": 2.650858022393451e-05, "loss": 0.0111, "step": 6730 }, { "epoch": 14.189473684210526, "grad_norm": 0.25704097747802734, "learning_rate": 2.6362747606431747e-05, "loss": 0.0115, "step": 6740 }, { "epoch": 14.210526315789474, "grad_norm": 0.21547827124595642, "learning_rate": 2.6217173482134172e-05, "loss": 0.011, "step": 6750 }, { "epoch": 14.23157894736842, "grad_norm": 0.24751166999340057, "learning_rate": 2.6071859443017044e-05, "loss": 0.009, "step": 6760 }, { "epoch": 14.25263157894737, "grad_norm": 0.20818237960338593, "learning_rate": 2.5926807078211414e-05, "loss": 0.009, "step": 6770 }, { "epoch": 14.273684210526316, "grad_norm": 0.17736408114433289, "learning_rate": 2.5782017973986728e-05, "loss": 0.0093, "step": 6780 }, { "epoch": 14.294736842105262, "grad_norm": 0.22523564100265503, "learning_rate": 2.5637493713733374e-05, "loss": 0.0097, "step": 6790 }, { "epoch": 14.31578947368421, "grad_norm": 0.33863598108291626, "learning_rate": 2.549323587794559e-05, "loss": 0.0126, "step": 6800 }, { "epoch": 14.336842105263157, "grad_norm": 0.18106992542743683, "learning_rate": 2.5349246044203895e-05, "loss": 0.0101, "step": 6810 }, { "epoch": 14.357894736842105, "grad_norm": 0.2817424535751343, "learning_rate": 2.520552578715808e-05, "loss": 0.009, "step": 6820 }, { "epoch": 14.378947368421052, "grad_norm": 0.22960305213928223, "learning_rate": 2.506207667850981e-05, "loss": 0.0109, "step": 6830 }, { "epoch": 14.4, "grad_norm": 0.25113645195961, "learning_rate": 2.4918900286995555e-05, "loss": 0.0088, "step": 6840 }, { "epoch": 14.421052631578947, "grad_norm": 0.19736681878566742, "learning_rate": 2.4775998178369458e-05, "loss": 0.0087, "step": 6850 }, { "epoch": 14.442105263157895, "grad_norm": 0.22692395746707916, "learning_rate": 2.4633371915386017e-05, "loss": 0.0109, "step": 6860 }, { "epoch": 14.463157894736842, "grad_norm": 0.47821295261383057, "learning_rate": 2.4491023057783235e-05, "loss": 0.0122, "step": 6870 }, { "epoch": 14.48421052631579, "grad_norm": 0.2852095365524292, "learning_rate": 2.4348953162265375e-05, "loss": 0.0088, "step": 6880 }, { "epoch": 14.505263157894737, "grad_norm": 0.2301289141178131, "learning_rate": 2.420716378248607e-05, "loss": 0.0098, "step": 6890 }, { "epoch": 14.526315789473685, "grad_norm": 0.2752452790737152, "learning_rate": 2.4065656469031266e-05, "loss": 0.0089, "step": 6900 }, { "epoch": 14.547368421052632, "grad_norm": 0.15273694694042206, "learning_rate": 2.3924432769402268e-05, "loss": 0.0089, "step": 6910 }, { "epoch": 14.568421052631578, "grad_norm": 0.3534073829650879, "learning_rate": 2.3783494227998844e-05, "loss": 0.0102, "step": 6920 }, { "epoch": 14.589473684210526, "grad_norm": 0.1326676905155182, "learning_rate": 2.3642842386102264e-05, "loss": 0.0094, "step": 6930 }, { "epoch": 14.610526315789473, "grad_norm": 0.18266354501247406, "learning_rate": 2.3502478781858567e-05, "loss": 0.009, "step": 6940 }, { "epoch": 14.631578947368421, "grad_norm": 0.262503445148468, "learning_rate": 2.3362404950261628e-05, "loss": 0.0128, "step": 6950 }, { "epoch": 14.652631578947368, "grad_norm": 0.22765569388866425, "learning_rate": 2.3222622423136458e-05, "loss": 0.0154, "step": 6960 }, { "epoch": 14.673684210526316, "grad_norm": 0.26077789068222046, "learning_rate": 2.3083132729122332e-05, "loss": 0.01, "step": 6970 }, { "epoch": 14.694736842105263, "grad_norm": 0.16915956139564514, "learning_rate": 2.294393739365621e-05, "loss": 0.0091, "step": 6980 }, { "epoch": 14.715789473684211, "grad_norm": 0.38230687379837036, "learning_rate": 2.2805037938956e-05, "loss": 0.0088, "step": 6990 }, { "epoch": 14.736842105263158, "grad_norm": 0.22908318042755127, "learning_rate": 2.266643588400386e-05, "loss": 0.0103, "step": 7000 }, { "epoch": 14.757894736842106, "grad_norm": 0.18986192345619202, "learning_rate": 2.252813274452969e-05, "loss": 0.0104, "step": 7010 }, { "epoch": 14.778947368421052, "grad_norm": 0.27457815408706665, "learning_rate": 2.2390130032994427e-05, "loss": 0.0101, "step": 7020 }, { "epoch": 14.8, "grad_norm": 0.26034972071647644, "learning_rate": 2.2252429258573633e-05, "loss": 0.0101, "step": 7030 }, { "epoch": 14.821052631578947, "grad_norm": 0.2530904710292816, "learning_rate": 2.2115031927140904e-05, "loss": 0.0098, "step": 7040 }, { "epoch": 14.842105263157894, "grad_norm": 0.31196683645248413, "learning_rate": 2.1977939541251463e-05, "loss": 0.0097, "step": 7050 }, { "epoch": 14.863157894736842, "grad_norm": 0.17124992609024048, "learning_rate": 2.1841153600125684e-05, "loss": 0.0113, "step": 7060 }, { "epoch": 14.884210526315789, "grad_norm": 0.2542293071746826, "learning_rate": 2.170467559963267e-05, "loss": 0.0144, "step": 7070 }, { "epoch": 14.905263157894737, "grad_norm": 0.26144328713417053, "learning_rate": 2.1568507032273982e-05, "loss": 0.0111, "step": 7080 }, { "epoch": 14.926315789473684, "grad_norm": 0.19823256134986877, "learning_rate": 2.1432649387167264e-05, "loss": 0.0117, "step": 7090 }, { "epoch": 14.947368421052632, "grad_norm": 0.306387722492218, "learning_rate": 2.1297104150029973e-05, "loss": 0.0136, "step": 7100 }, { "epoch": 14.968421052631578, "grad_norm": 0.19889692962169647, "learning_rate": 2.116187280316307e-05, "loss": 0.0106, "step": 7110 }, { "epoch": 14.989473684210527, "grad_norm": 0.3079199492931366, "learning_rate": 2.1026956825434908e-05, "loss": 0.0106, "step": 7120 }, { "epoch": 15.010526315789473, "grad_norm": 0.25304290652275085, "learning_rate": 2.0892357692265017e-05, "loss": 0.0105, "step": 7130 }, { "epoch": 15.031578947368422, "grad_norm": 0.22999891638755798, "learning_rate": 2.0758076875607947e-05, "loss": 0.0092, "step": 7140 }, { "epoch": 15.052631578947368, "grad_norm": 0.22952020168304443, "learning_rate": 2.0624115843937207e-05, "loss": 0.008, "step": 7150 }, { "epoch": 15.073684210526316, "grad_norm": 0.205112487077713, "learning_rate": 2.0490476062229157e-05, "loss": 0.0086, "step": 7160 }, { "epoch": 15.094736842105263, "grad_norm": 0.23687493801116943, "learning_rate": 2.035715899194704e-05, "loss": 0.0098, "step": 7170 }, { "epoch": 15.115789473684211, "grad_norm": 0.1891774833202362, "learning_rate": 2.022416609102499e-05, "loss": 0.0088, "step": 7180 }, { "epoch": 15.136842105263158, "grad_norm": 0.2020113617181778, "learning_rate": 2.009149881385205e-05, "loss": 0.0076, "step": 7190 }, { "epoch": 15.157894736842104, "grad_norm": 0.18823060393333435, "learning_rate": 1.995915861125634e-05, "loss": 0.0085, "step": 7200 }, { "epoch": 15.178947368421053, "grad_norm": 0.29947078227996826, "learning_rate": 1.9827146930489065e-05, "loss": 0.0085, "step": 7210 }, { "epoch": 15.2, "grad_norm": 0.26559212803840637, "learning_rate": 1.9695465215208848e-05, "loss": 0.0093, "step": 7220 }, { "epoch": 15.221052631578948, "grad_norm": 0.13796184957027435, "learning_rate": 1.9564114905465813e-05, "loss": 0.0069, "step": 7230 }, { "epoch": 15.242105263157894, "grad_norm": 0.1861109584569931, "learning_rate": 1.9433097437685936e-05, "loss": 0.0105, "step": 7240 }, { "epoch": 15.263157894736842, "grad_norm": 0.22382038831710815, "learning_rate": 1.930241424465521e-05, "loss": 0.0092, "step": 7250 }, { "epoch": 15.284210526315789, "grad_norm": 0.1356709748506546, "learning_rate": 1.9172066755504115e-05, "loss": 0.0084, "step": 7260 }, { "epoch": 15.305263157894737, "grad_norm": 0.17231476306915283, "learning_rate": 1.9042056395691914e-05, "loss": 0.0102, "step": 7270 }, { "epoch": 15.326315789473684, "grad_norm": 0.19465982913970947, "learning_rate": 1.8912384586991066e-05, "loss": 0.0093, "step": 7280 }, { "epoch": 15.347368421052632, "grad_norm": 0.30636411905288696, "learning_rate": 1.8783052747471717e-05, "loss": 0.009, "step": 7290 }, { "epoch": 15.368421052631579, "grad_norm": 0.299232542514801, "learning_rate": 1.865406229148611e-05, "loss": 0.0113, "step": 7300 }, { "epoch": 15.389473684210527, "grad_norm": 0.3965524435043335, "learning_rate": 1.8525414629653233e-05, "loss": 0.0092, "step": 7310 }, { "epoch": 15.410526315789474, "grad_norm": 0.16469106078147888, "learning_rate": 1.8397111168843255e-05, "loss": 0.0091, "step": 7320 }, { "epoch": 15.431578947368422, "grad_norm": 0.330039381980896, "learning_rate": 1.8269153312162323e-05, "loss": 0.009, "step": 7330 }, { "epoch": 15.452631578947368, "grad_norm": 0.2373582273721695, "learning_rate": 1.8141542458937054e-05, "loss": 0.0085, "step": 7340 }, { "epoch": 15.473684210526315, "grad_norm": 0.2247738093137741, "learning_rate": 1.8014280004699268e-05, "loss": 0.0111, "step": 7350 }, { "epoch": 15.494736842105263, "grad_norm": 0.2236502468585968, "learning_rate": 1.788736734117078e-05, "loss": 0.0107, "step": 7360 }, { "epoch": 15.51578947368421, "grad_norm": 0.17109239101409912, "learning_rate": 1.7760805856248152e-05, "loss": 0.0089, "step": 7370 }, { "epoch": 15.536842105263158, "grad_norm": 0.3020864427089691, "learning_rate": 1.7634596933987518e-05, "loss": 0.0092, "step": 7380 }, { "epoch": 15.557894736842105, "grad_norm": 0.25554168224334717, "learning_rate": 1.7508741954589404e-05, "loss": 0.0121, "step": 7390 }, { "epoch": 15.578947368421053, "grad_norm": 0.4176688492298126, "learning_rate": 1.7383242294383717e-05, "loss": 0.0093, "step": 7400 }, { "epoch": 15.6, "grad_norm": 0.14629438519477844, "learning_rate": 1.7258099325814632e-05, "loss": 0.0092, "step": 7410 }, { "epoch": 15.621052631578948, "grad_norm": 0.2520493268966675, "learning_rate": 1.7133314417425594e-05, "loss": 0.0102, "step": 7420 }, { "epoch": 15.642105263157895, "grad_norm": 0.1684764176607132, "learning_rate": 1.7008888933844408e-05, "loss": 0.008, "step": 7430 }, { "epoch": 15.663157894736843, "grad_norm": 0.2696071267127991, "learning_rate": 1.6884824235768172e-05, "loss": 0.0073, "step": 7440 }, { "epoch": 15.68421052631579, "grad_norm": 0.20125557482242584, "learning_rate": 1.6761121679948592e-05, "loss": 0.0091, "step": 7450 }, { "epoch": 15.705263157894738, "grad_norm": 0.1749448925256729, "learning_rate": 1.663778261917695e-05, "loss": 0.01, "step": 7460 }, { "epoch": 15.726315789473684, "grad_norm": 0.25218579173088074, "learning_rate": 1.651480840226952e-05, "loss": 0.0068, "step": 7470 }, { "epoch": 15.74736842105263, "grad_norm": 0.16348335146903992, "learning_rate": 1.639220037405258e-05, "loss": 0.0082, "step": 7480 }, { "epoch": 15.76842105263158, "grad_norm": 0.2954403758049011, "learning_rate": 1.6269959875347906e-05, "loss": 0.0089, "step": 7490 }, { "epoch": 15.789473684210526, "grad_norm": 0.17681606113910675, "learning_rate": 1.614808824295802e-05, "loss": 0.0076, "step": 7500 }, { "epoch": 15.810526315789474, "grad_norm": 0.192280575633049, "learning_rate": 1.602658680965152e-05, "loss": 0.0081, "step": 7510 }, { "epoch": 15.83157894736842, "grad_norm": 0.21202149987220764, "learning_rate": 1.5905456904148686e-05, "loss": 0.0069, "step": 7520 }, { "epoch": 15.852631578947369, "grad_norm": 0.11812349408864975, "learning_rate": 1.57846998511067e-05, "loss": 0.0073, "step": 7530 }, { "epoch": 15.873684210526315, "grad_norm": 0.3178236782550812, "learning_rate": 1.566431697110538e-05, "loss": 0.014, "step": 7540 }, { "epoch": 15.894736842105264, "grad_norm": 0.2718460261821747, "learning_rate": 1.554430958063259e-05, "loss": 0.0095, "step": 7550 }, { "epoch": 15.91578947368421, "grad_norm": 0.1963137686252594, "learning_rate": 1.5424678992069912e-05, "loss": 0.0084, "step": 7560 }, { "epoch": 15.936842105263159, "grad_norm": 0.17500977218151093, "learning_rate": 1.5305426513678362e-05, "loss": 0.0089, "step": 7570 }, { "epoch": 15.957894736842105, "grad_norm": 0.2482364922761917, "learning_rate": 1.518655344958388e-05, "loss": 0.0076, "step": 7580 }, { "epoch": 15.978947368421053, "grad_norm": 0.17356327176094055, "learning_rate": 1.5068061099763275e-05, "loss": 0.007, "step": 7590 }, { "epoch": 16.0, "grad_norm": 0.3485445976257324, "learning_rate": 1.494995076002988e-05, "loss": 0.01, "step": 7600 }, { "epoch": 16.021052631578947, "grad_norm": 0.22509033977985382, "learning_rate": 1.4832223722019456e-05, "loss": 0.0076, "step": 7610 }, { "epoch": 16.042105263157893, "grad_norm": 0.2686956822872162, "learning_rate": 1.4714881273176035e-05, "loss": 0.0095, "step": 7620 }, { "epoch": 16.063157894736843, "grad_norm": 0.19264329969882965, "learning_rate": 1.4597924696737835e-05, "loss": 0.008, "step": 7630 }, { "epoch": 16.08421052631579, "grad_norm": 0.5003637075424194, "learning_rate": 1.4481355271723252e-05, "loss": 0.0097, "step": 7640 }, { "epoch": 16.105263157894736, "grad_norm": 0.2948462963104248, "learning_rate": 1.4365174272916809e-05, "loss": 0.0094, "step": 7650 }, { "epoch": 16.126315789473683, "grad_norm": 0.5390455722808838, "learning_rate": 1.4249382970855319e-05, "loss": 0.0111, "step": 7660 }, { "epoch": 16.147368421052633, "grad_norm": 0.22467181086540222, "learning_rate": 1.4133982631813903e-05, "loss": 0.0075, "step": 7670 }, { "epoch": 16.16842105263158, "grad_norm": 0.275263249874115, "learning_rate": 1.4018974517792194e-05, "loss": 0.0089, "step": 7680 }, { "epoch": 16.189473684210526, "grad_norm": 0.30138248205184937, "learning_rate": 1.390435988650048e-05, "loss": 0.0098, "step": 7690 }, { "epoch": 16.210526315789473, "grad_norm": 0.45287588238716125, "learning_rate": 1.3790139991346006e-05, "loss": 0.0094, "step": 7700 }, { "epoch": 16.231578947368423, "grad_norm": 0.30403706431388855, "learning_rate": 1.367631608141926e-05, "loss": 0.0076, "step": 7710 }, { "epoch": 16.25263157894737, "grad_norm": 0.2147553265094757, "learning_rate": 1.3562889401480278e-05, "loss": 0.0093, "step": 7720 }, { "epoch": 16.273684210526316, "grad_norm": 0.38180047273635864, "learning_rate": 1.3449861191945074e-05, "loss": 0.0093, "step": 7730 }, { "epoch": 16.294736842105262, "grad_norm": 0.21284984052181244, "learning_rate": 1.3337232688872009e-05, "loss": 0.0071, "step": 7740 }, { "epoch": 16.31578947368421, "grad_norm": 0.3071005344390869, "learning_rate": 1.3225005123948364e-05, "loss": 0.0081, "step": 7750 }, { "epoch": 16.33684210526316, "grad_norm": 0.18478229641914368, "learning_rate": 1.311317972447681e-05, "loss": 0.0088, "step": 7760 }, { "epoch": 16.357894736842105, "grad_norm": 0.1683579981327057, "learning_rate": 1.3001757713361996e-05, "loss": 0.0085, "step": 7770 }, { "epoch": 16.378947368421052, "grad_norm": 0.17573809623718262, "learning_rate": 1.2890740309097204e-05, "loss": 0.0091, "step": 7780 }, { "epoch": 16.4, "grad_norm": 0.2427658885717392, "learning_rate": 1.2780128725750944e-05, "loss": 0.0097, "step": 7790 }, { "epoch": 16.42105263157895, "grad_norm": 0.21274468302726746, "learning_rate": 1.266992417295379e-05, "loss": 0.0097, "step": 7800 }, { "epoch": 16.442105263157895, "grad_norm": 0.7105726003646851, "learning_rate": 1.2560127855885073e-05, "loss": 0.0112, "step": 7810 }, { "epoch": 16.46315789473684, "grad_norm": 0.17152060568332672, "learning_rate": 1.2450740975259745e-05, "loss": 0.0081, "step": 7820 }, { "epoch": 16.48421052631579, "grad_norm": 0.28981897234916687, "learning_rate": 1.234176472731517e-05, "loss": 0.0102, "step": 7830 }, { "epoch": 16.50526315789474, "grad_norm": 0.2669426202774048, "learning_rate": 1.2233200303798158e-05, "loss": 0.0093, "step": 7840 }, { "epoch": 16.526315789473685, "grad_norm": 0.21747006475925446, "learning_rate": 1.2125048891951846e-05, "loss": 0.0092, "step": 7850 }, { "epoch": 16.54736842105263, "grad_norm": 0.20006047189235687, "learning_rate": 1.2017311674502745e-05, "loss": 0.0073, "step": 7860 }, { "epoch": 16.568421052631578, "grad_norm": 0.2669523358345032, "learning_rate": 1.1909989829647822e-05, "loss": 0.0102, "step": 7870 }, { "epoch": 16.589473684210525, "grad_norm": 0.24711818993091583, "learning_rate": 1.1803084531041553e-05, "loss": 0.0092, "step": 7880 }, { "epoch": 16.610526315789475, "grad_norm": 0.3917536735534668, "learning_rate": 1.1696596947783162e-05, "loss": 0.0073, "step": 7890 }, { "epoch": 16.63157894736842, "grad_norm": 0.18167628347873688, "learning_rate": 1.1590528244403803e-05, "loss": 0.0086, "step": 7900 }, { "epoch": 16.652631578947368, "grad_norm": 0.5472114682197571, "learning_rate": 1.148487958085382e-05, "loss": 0.0085, "step": 7910 }, { "epoch": 16.673684210526314, "grad_norm": 0.2348708063364029, "learning_rate": 1.1379652112490086e-05, "loss": 0.0083, "step": 7920 }, { "epoch": 16.694736842105264, "grad_norm": 0.18810661137104034, "learning_rate": 1.1274846990063315e-05, "loss": 0.0077, "step": 7930 }, { "epoch": 16.71578947368421, "grad_norm": 0.3286520838737488, "learning_rate": 1.117046535970554e-05, "loss": 0.007, "step": 7940 }, { "epoch": 16.736842105263158, "grad_norm": 0.2243700921535492, "learning_rate": 1.106650836291755e-05, "loss": 0.0086, "step": 7950 }, { "epoch": 16.757894736842104, "grad_norm": 0.31743335723876953, "learning_rate": 1.0962977136556418e-05, "loss": 0.0087, "step": 7960 }, { "epoch": 16.778947368421054, "grad_norm": 0.2021445333957672, "learning_rate": 1.0859872812823024e-05, "loss": 0.0088, "step": 7970 }, { "epoch": 16.8, "grad_norm": 0.22925220429897308, "learning_rate": 1.0757196519249747e-05, "loss": 0.0097, "step": 7980 }, { "epoch": 16.821052631578947, "grad_norm": 0.13621430099010468, "learning_rate": 1.0654949378688077e-05, "loss": 0.0085, "step": 7990 }, { "epoch": 16.842105263157894, "grad_norm": 0.19834457337856293, "learning_rate": 1.0553132509296376e-05, "loss": 0.0102, "step": 8000 }, { "epoch": 16.863157894736844, "grad_norm": 0.17377321422100067, "learning_rate": 1.0451747024527613e-05, "loss": 0.0081, "step": 8010 }, { "epoch": 16.88421052631579, "grad_norm": 0.16337136924266815, "learning_rate": 1.0350794033117189e-05, "loss": 0.0073, "step": 8020 }, { "epoch": 16.905263157894737, "grad_norm": 0.18713410198688507, "learning_rate": 1.0250274639070856e-05, "loss": 0.0069, "step": 8030 }, { "epoch": 16.926315789473684, "grad_norm": 0.18635889887809753, "learning_rate": 1.0150189941652599e-05, "loss": 0.0064, "step": 8040 }, { "epoch": 16.94736842105263, "grad_norm": 0.24890416860580444, "learning_rate": 1.0050541035372635e-05, "loss": 0.0092, "step": 8050 }, { "epoch": 16.96842105263158, "grad_norm": 0.1714644432067871, "learning_rate": 9.951329009975458e-06, "loss": 0.0063, "step": 8060 }, { "epoch": 16.989473684210527, "grad_norm": 0.1188599169254303, "learning_rate": 9.852554950427845e-06, "loss": 0.0087, "step": 8070 }, { "epoch": 17.010526315789473, "grad_norm": 0.10455501079559326, "learning_rate": 9.754219936907105e-06, "loss": 0.0076, "step": 8080 }, { "epoch": 17.03157894736842, "grad_norm": 0.18457181751728058, "learning_rate": 9.656325044789194e-06, "loss": 0.0084, "step": 8090 }, { "epoch": 17.05263157894737, "grad_norm": 0.1763952672481537, "learning_rate": 9.55887134463697e-06, "loss": 0.0086, "step": 8100 }, { "epoch": 17.073684210526316, "grad_norm": 0.20308321714401245, "learning_rate": 9.461859902188475e-06, "loss": 0.0074, "step": 8110 }, { "epoch": 17.094736842105263, "grad_norm": 0.2186141163110733, "learning_rate": 9.365291778345303e-06, "loss": 0.0065, "step": 8120 }, { "epoch": 17.11578947368421, "grad_norm": 0.1747424453496933, "learning_rate": 9.269168029160991e-06, "loss": 0.0104, "step": 8130 }, { "epoch": 17.13684210526316, "grad_norm": 0.33370327949523926, "learning_rate": 9.173489705829447e-06, "loss": 0.0083, "step": 8140 }, { "epoch": 17.157894736842106, "grad_norm": 0.2662794888019562, "learning_rate": 9.078257854673516e-06, "loss": 0.0094, "step": 8150 }, { "epoch": 17.178947368421053, "grad_norm": 0.28646308183670044, "learning_rate": 8.983473517133429e-06, "loss": 0.0099, "step": 8160 }, { "epoch": 17.2, "grad_norm": 0.2722373604774475, "learning_rate": 8.889137729755537e-06, "loss": 0.0078, "step": 8170 }, { "epoch": 17.221052631578946, "grad_norm": 0.2046920359134674, "learning_rate": 8.79525152418087e-06, "loss": 0.0094, "step": 8180 }, { "epoch": 17.242105263157896, "grad_norm": 0.27062350511550903, "learning_rate": 8.701815927133961e-06, "loss": 0.0091, "step": 8190 }, { "epoch": 17.263157894736842, "grad_norm": 0.23174948990345, "learning_rate": 8.608831960411534e-06, "loss": 0.0091, "step": 8200 }, { "epoch": 17.28421052631579, "grad_norm": 0.13621558248996735, "learning_rate": 8.516300640871321e-06, "loss": 0.0081, "step": 8210 }, { "epoch": 17.305263157894736, "grad_norm": 0.211971253156662, "learning_rate": 8.424222980421038e-06, "loss": 0.0095, "step": 8220 }, { "epoch": 17.326315789473686, "grad_norm": 0.18213477730751038, "learning_rate": 8.332599986007184e-06, "loss": 0.0082, "step": 8230 }, { "epoch": 17.347368421052632, "grad_norm": 0.16762478649616241, "learning_rate": 8.241432659604203e-06, "loss": 0.0083, "step": 8240 }, { "epoch": 17.36842105263158, "grad_norm": 0.2141171097755432, "learning_rate": 8.150721998203331e-06, "loss": 0.0067, "step": 8250 }, { "epoch": 17.389473684210525, "grad_norm": 0.1684076339006424, "learning_rate": 8.06046899380184e-06, "loss": 0.0077, "step": 8260 }, { "epoch": 17.410526315789475, "grad_norm": 0.21785205602645874, "learning_rate": 7.970674633392133e-06, "loss": 0.0098, "step": 8270 }, { "epoch": 17.431578947368422, "grad_norm": 0.29006505012512207, "learning_rate": 7.881339898950924e-06, "loss": 0.0088, "step": 8280 }, { "epoch": 17.45263157894737, "grad_norm": 0.19461551308631897, "learning_rate": 7.792465767428597e-06, "loss": 0.0088, "step": 8290 }, { "epoch": 17.473684210526315, "grad_norm": 0.2775125503540039, "learning_rate": 7.704053210738376e-06, "loss": 0.0095, "step": 8300 }, { "epoch": 17.49473684210526, "grad_norm": 0.13282199203968048, "learning_rate": 7.6161031957458494e-06, "loss": 0.007, "step": 8310 }, { "epoch": 17.51578947368421, "grad_norm": 0.251899778842926, "learning_rate": 7.5286166842582605e-06, "loss": 0.0099, "step": 8320 }, { "epoch": 17.53684210526316, "grad_norm": 0.409236878156662, "learning_rate": 7.4415946330140814e-06, "loss": 0.0118, "step": 8330 }, { "epoch": 17.557894736842105, "grad_norm": 0.30573299527168274, "learning_rate": 7.3550379936725644e-06, "loss": 0.0099, "step": 8340 }, { "epoch": 17.57894736842105, "grad_norm": 0.44984927773475647, "learning_rate": 7.2689477128032035e-06, "loss": 0.0088, "step": 8350 }, { "epoch": 17.6, "grad_norm": 0.14957810938358307, "learning_rate": 7.183324731875551e-06, "loss": 0.0065, "step": 8360 }, { "epoch": 17.621052631578948, "grad_norm": 0.11822706460952759, "learning_rate": 7.098169987248782e-06, "loss": 0.0088, "step": 8370 }, { "epoch": 17.642105263157895, "grad_norm": 0.22821851074695587, "learning_rate": 7.013484410161553e-06, "loss": 0.007, "step": 8380 }, { "epoch": 17.66315789473684, "grad_norm": 0.2048841118812561, "learning_rate": 6.92926892672176e-06, "loss": 0.0098, "step": 8390 }, { "epoch": 17.68421052631579, "grad_norm": 0.16232773661613464, "learning_rate": 6.845524457896446e-06, "loss": 0.0079, "step": 8400 }, { "epoch": 17.705263157894738, "grad_norm": 0.1672491729259491, "learning_rate": 6.7622519195017165e-06, "loss": 0.0084, "step": 8410 }, { "epoch": 17.726315789473684, "grad_norm": 0.29104024171829224, "learning_rate": 6.679452222192684e-06, "loss": 0.0082, "step": 8420 }, { "epoch": 17.74736842105263, "grad_norm": 0.2120663970708847, "learning_rate": 6.597126271453579e-06, "loss": 0.0078, "step": 8430 }, { "epoch": 17.768421052631577, "grad_norm": 0.1816309243440628, "learning_rate": 6.51527496758782e-06, "loss": 0.0086, "step": 8440 }, { "epoch": 17.789473684210527, "grad_norm": 0.2621295750141144, "learning_rate": 6.433899205708155e-06, "loss": 0.0112, "step": 8450 }, { "epoch": 17.810526315789474, "grad_norm": 0.11127659678459167, "learning_rate": 6.352999875726856e-06, "loss": 0.0082, "step": 8460 }, { "epoch": 17.83157894736842, "grad_norm": 0.18613527715206146, "learning_rate": 6.272577862346052e-06, "loss": 0.0092, "step": 8470 }, { "epoch": 17.852631578947367, "grad_norm": 0.2400396317243576, "learning_rate": 6.192634045047996e-06, "loss": 0.0096, "step": 8480 }, { "epoch": 17.873684210526317, "grad_norm": 0.17899535596370697, "learning_rate": 6.113169298085458e-06, "loss": 0.0097, "step": 8490 }, { "epoch": 17.894736842105264, "grad_norm": 0.2727416157722473, "learning_rate": 6.034184490472195e-06, "loss": 0.0088, "step": 8500 }, { "epoch": 17.91578947368421, "grad_norm": 0.3036291003227234, "learning_rate": 5.955680485973386e-06, "loss": 0.0086, "step": 8510 }, { "epoch": 17.936842105263157, "grad_norm": 0.22371500730514526, "learning_rate": 5.877658143096265e-06, "loss": 0.0084, "step": 8520 }, { "epoch": 17.957894736842107, "grad_norm": 0.1316075325012207, "learning_rate": 5.800118315080661e-06, "loss": 0.0058, "step": 8530 }, { "epoch": 17.978947368421053, "grad_norm": 0.26276010274887085, "learning_rate": 5.723061849889716e-06, "loss": 0.0103, "step": 8540 }, { "epoch": 18.0, "grad_norm": 0.23892588913440704, "learning_rate": 5.646489590200604e-06, "loss": 0.0096, "step": 8550 }, { "epoch": 18.021052631578947, "grad_norm": 0.15599054098129272, "learning_rate": 5.570402373395256e-06, "loss": 0.0091, "step": 8560 }, { "epoch": 18.042105263157893, "grad_norm": 0.10391727834939957, "learning_rate": 5.494801031551305e-06, "loss": 0.0063, "step": 8570 }, { "epoch": 18.063157894736843, "grad_norm": 0.15484397113323212, "learning_rate": 5.41968639143291e-06, "loss": 0.0095, "step": 8580 }, { "epoch": 18.08421052631579, "grad_norm": 0.23186784982681274, "learning_rate": 5.345059274481751e-06, "loss": 0.0079, "step": 8590 }, { "epoch": 18.105263157894736, "grad_norm": 0.21175621449947357, "learning_rate": 5.270920496808002e-06, "loss": 0.0076, "step": 8600 }, { "epoch": 18.126315789473683, "grad_norm": 0.2248217612504959, "learning_rate": 5.1972708691814695e-06, "loss": 0.0084, "step": 8610 }, { "epoch": 18.147368421052633, "grad_norm": 0.15544597804546356, "learning_rate": 5.124111197022674e-06, "loss": 0.0073, "step": 8620 }, { "epoch": 18.16842105263158, "grad_norm": 0.46217983961105347, "learning_rate": 5.051442280394081e-06, "loss": 0.0085, "step": 8630 }, { "epoch": 18.189473684210526, "grad_norm": 0.2669810950756073, "learning_rate": 4.979264913991322e-06, "loss": 0.0082, "step": 8640 }, { "epoch": 18.210526315789473, "grad_norm": 0.1743437945842743, "learning_rate": 4.907579887134489e-06, "loss": 0.0088, "step": 8650 }, { "epoch": 18.231578947368423, "grad_norm": 0.18000270426273346, "learning_rate": 4.836387983759572e-06, "loss": 0.0066, "step": 8660 }, { "epoch": 18.25263157894737, "grad_norm": 0.15320366621017456, "learning_rate": 4.765689982409816e-06, "loss": 0.0071, "step": 8670 }, { "epoch": 18.273684210526316, "grad_norm": 0.13704019784927368, "learning_rate": 4.695486656227233e-06, "loss": 0.0071, "step": 8680 }, { "epoch": 18.294736842105262, "grad_norm": 0.1702423244714737, "learning_rate": 4.625778772944156e-06, "loss": 0.0078, "step": 8690 }, { "epoch": 18.31578947368421, "grad_norm": 0.3772662580013275, "learning_rate": 4.556567094874825e-06, "loss": 0.0092, "step": 8700 }, { "epoch": 18.33684210526316, "grad_norm": 0.1366453766822815, "learning_rate": 4.487852378907059e-06, "loss": 0.0063, "step": 8710 }, { "epoch": 18.357894736842105, "grad_norm": 0.24812857806682587, "learning_rate": 4.419635376493986e-06, "loss": 0.0081, "step": 8720 }, { "epoch": 18.378947368421052, "grad_norm": 0.31375062465667725, "learning_rate": 4.351916833645825e-06, "loss": 0.0089, "step": 8730 }, { "epoch": 18.4, "grad_norm": 0.286811500787735, "learning_rate": 4.284697490921691e-06, "loss": 0.0087, "step": 8740 }, { "epoch": 18.42105263157895, "grad_norm": 0.3487447202205658, "learning_rate": 4.2179780834215585e-06, "loss": 0.0087, "step": 8750 }, { "epoch": 18.442105263157895, "grad_norm": 0.2674311697483063, "learning_rate": 4.151759340778178e-06, "loss": 0.0112, "step": 8760 }, { "epoch": 18.46315789473684, "grad_norm": 0.09591855108737946, "learning_rate": 4.086041987149109e-06, "loss": 0.0076, "step": 8770 }, { "epoch": 18.48421052631579, "grad_norm": 0.13226757943630219, "learning_rate": 4.020826741208811e-06, "loss": 0.0092, "step": 8780 }, { "epoch": 18.50526315789474, "grad_norm": 0.17560096085071564, "learning_rate": 3.956114316140746e-06, "loss": 0.0093, "step": 8790 }, { "epoch": 18.526315789473685, "grad_norm": 0.16507865488529205, "learning_rate": 3.891905419629643e-06, "loss": 0.0091, "step": 8800 }, { "epoch": 18.54736842105263, "grad_norm": 0.2131287157535553, "learning_rate": 3.8282007538536946e-06, "loss": 0.0115, "step": 8810 }, { "epoch": 18.568421052631578, "grad_norm": 0.2253255397081375, "learning_rate": 3.7650010154769265e-06, "loss": 0.0066, "step": 8820 }, { "epoch": 18.589473684210525, "grad_norm": 0.21640874445438385, "learning_rate": 3.7023068956415608e-06, "loss": 0.0078, "step": 8830 }, { "epoch": 18.610526315789475, "grad_norm": 0.57985919713974, "learning_rate": 3.6401190799604303e-06, "loss": 0.0085, "step": 8840 }, { "epoch": 18.63157894736842, "grad_norm": 0.10613711178302765, "learning_rate": 3.578438248509536e-06, "loss": 0.0091, "step": 8850 }, { "epoch": 18.652631578947368, "grad_norm": 0.2695419192314148, "learning_rate": 3.5172650758205583e-06, "loss": 0.0065, "step": 8860 }, { "epoch": 18.673684210526314, "grad_norm": 0.20468224585056305, "learning_rate": 3.45660023087353e-06, "loss": 0.0084, "step": 8870 }, { "epoch": 18.694736842105264, "grad_norm": 0.22441209852695465, "learning_rate": 3.3964443770894528e-06, "loss": 0.0084, "step": 8880 }, { "epoch": 18.71578947368421, "grad_norm": 0.15253819525241852, "learning_rate": 3.3367981723231245e-06, "loss": 0.0086, "step": 8890 }, { "epoch": 18.736842105263158, "grad_norm": 0.2988612949848175, "learning_rate": 3.2776622688558746e-06, "loss": 0.0095, "step": 8900 }, { "epoch": 18.757894736842104, "grad_norm": 0.170147106051445, "learning_rate": 3.2190373133884677e-06, "loss": 0.009, "step": 8910 }, { "epoch": 18.778947368421054, "grad_norm": 0.32086387276649475, "learning_rate": 3.1609239470340446e-06, "loss": 0.0081, "step": 8920 }, { "epoch": 18.8, "grad_norm": 0.24304763972759247, "learning_rate": 3.1033228053110373e-06, "loss": 0.0085, "step": 8930 }, { "epoch": 18.821052631578947, "grad_norm": 0.14699025452136993, "learning_rate": 3.0462345181363314e-06, "loss": 0.0068, "step": 8940 }, { "epoch": 18.842105263157894, "grad_norm": 0.2285974770784378, "learning_rate": 2.9896597098182654e-06, "loss": 0.0086, "step": 8950 }, { "epoch": 18.863157894736844, "grad_norm": 0.1616249531507492, "learning_rate": 2.933598999049891e-06, "loss": 0.0073, "step": 8960 }, { "epoch": 18.88421052631579, "grad_norm": 0.15154878795146942, "learning_rate": 2.8780529989021697e-06, "loss": 0.0074, "step": 8970 }, { "epoch": 18.905263157894737, "grad_norm": 0.20797893404960632, "learning_rate": 2.823022316817242e-06, "loss": 0.0073, "step": 8980 }, { "epoch": 18.926315789473684, "grad_norm": 0.07916927337646484, "learning_rate": 2.7685075546018456e-06, "loss": 0.0077, "step": 8990 }, { "epoch": 18.94736842105263, "grad_norm": 0.2022639364004135, "learning_rate": 2.7145093084206598e-06, "loss": 0.0076, "step": 9000 }, { "epoch": 18.96842105263158, "grad_norm": 0.19936202466487885, "learning_rate": 2.661028168789892e-06, "loss": 0.0071, "step": 9010 }, { "epoch": 18.989473684210527, "grad_norm": 0.3514852225780487, "learning_rate": 2.6080647205706855e-06, "loss": 0.0069, "step": 9020 }, { "epoch": 19.010526315789473, "grad_norm": 0.17553842067718506, "learning_rate": 2.555619542962834e-06, "loss": 0.0072, "step": 9030 }, { "epoch": 19.03157894736842, "grad_norm": 0.09492363035678864, "learning_rate": 2.503693209498409e-06, "loss": 0.0081, "step": 9040 }, { "epoch": 19.05263157894737, "grad_norm": 0.2049422711133957, "learning_rate": 2.452286288035449e-06, "loss": 0.0078, "step": 9050 }, { "epoch": 19.073684210526316, "grad_norm": 0.22294217348098755, "learning_rate": 2.4013993407518363e-06, "loss": 0.0073, "step": 9060 }, { "epoch": 19.094736842105263, "grad_norm": 0.19313672184944153, "learning_rate": 2.351032924139063e-06, "loss": 0.0097, "step": 9070 }, { "epoch": 19.11578947368421, "grad_norm": 0.17759540677070618, "learning_rate": 2.30118758899619e-06, "loss": 0.0067, "step": 9080 }, { "epoch": 19.13684210526316, "grad_norm": 0.22824054956436157, "learning_rate": 2.2518638804238157e-06, "loss": 0.0093, "step": 9090 }, { "epoch": 19.157894736842106, "grad_norm": 0.130801722407341, "learning_rate": 2.203062337818118e-06, "loss": 0.0065, "step": 9100 }, { "epoch": 19.178947368421053, "grad_norm": 0.14103183150291443, "learning_rate": 2.1547834948649483e-06, "loss": 0.006, "step": 9110 }, { "epoch": 19.2, "grad_norm": 0.29217180609703064, "learning_rate": 2.1070278795340017e-06, "loss": 0.0084, "step": 9120 }, { "epoch": 19.221052631578946, "grad_norm": 0.19553405046463013, "learning_rate": 2.059796014073029e-06, "loss": 0.0078, "step": 9130 }, { "epoch": 19.242105263157896, "grad_norm": 0.12673620879650116, "learning_rate": 2.01308841500214e-06, "loss": 0.009, "step": 9140 }, { "epoch": 19.263157894736842, "grad_norm": 0.23371729254722595, "learning_rate": 1.9669055931081704e-06, "loss": 0.0089, "step": 9150 }, { "epoch": 19.28421052631579, "grad_norm": 0.3160805404186249, "learning_rate": 1.9212480534390507e-06, "loss": 0.0061, "step": 9160 }, { "epoch": 19.305263157894736, "grad_norm": 0.1390615552663803, "learning_rate": 1.8761162952983246e-06, "loss": 0.0105, "step": 9170 }, { "epoch": 19.326315789473686, "grad_norm": 0.20343443751335144, "learning_rate": 1.8315108122396618e-06, "loss": 0.0069, "step": 9180 }, { "epoch": 19.347368421052632, "grad_norm": 0.1566215604543686, "learning_rate": 1.787432092061475e-06, "loss": 0.0062, "step": 9190 }, { "epoch": 19.36842105263158, "grad_norm": 0.3467118442058563, "learning_rate": 1.743880616801602e-06, "loss": 0.0074, "step": 9200 }, { "epoch": 19.389473684210525, "grad_norm": 0.24993333220481873, "learning_rate": 1.7008568627319865e-06, "loss": 0.0061, "step": 9210 }, { "epoch": 19.410526315789475, "grad_norm": 0.12367149442434311, "learning_rate": 1.6583613003535226e-06, "loss": 0.008, "step": 9220 }, { "epoch": 19.431578947368422, "grad_norm": 0.17127743363380432, "learning_rate": 1.6163943943908522e-06, "loss": 0.0081, "step": 9230 }, { "epoch": 19.45263157894737, "grad_norm": 0.30250033736228943, "learning_rate": 1.5749566037873476e-06, "loss": 0.0085, "step": 9240 }, { "epoch": 19.473684210526315, "grad_norm": 0.1750367134809494, "learning_rate": 1.5340483817000428e-06, "loss": 0.0079, "step": 9250 }, { "epoch": 19.49473684210526, "grad_norm": 0.19316816329956055, "learning_rate": 1.4936701754947101e-06, "loss": 0.0069, "step": 9260 }, { "epoch": 19.51578947368421, "grad_norm": 0.15851515531539917, "learning_rate": 1.4538224267409361e-06, "loss": 0.0062, "step": 9270 }, { "epoch": 19.53684210526316, "grad_norm": 0.22913870215415955, "learning_rate": 1.414505571207314e-06, "loss": 0.0087, "step": 9280 }, { "epoch": 19.557894736842105, "grad_norm": 0.20993894338607788, "learning_rate": 1.3757200388566816e-06, "loss": 0.0076, "step": 9290 }, { "epoch": 19.57894736842105, "grad_norm": 0.16939908266067505, "learning_rate": 1.3374662538414074e-06, "loss": 0.007, "step": 9300 }, { "epoch": 19.6, "grad_norm": 0.2342451512813568, "learning_rate": 1.2997446344987617e-06, "loss": 0.0087, "step": 9310 }, { "epoch": 19.621052631578948, "grad_norm": 0.1497613787651062, "learning_rate": 1.262555593346315e-06, "loss": 0.0071, "step": 9320 }, { "epoch": 19.642105263157895, "grad_norm": 0.11424117535352707, "learning_rate": 1.2258995370774685e-06, "loss": 0.0079, "step": 9330 }, { "epoch": 19.66315789473684, "grad_norm": 0.2616402208805084, "learning_rate": 1.1897768665569798e-06, "loss": 0.0072, "step": 9340 }, { "epoch": 19.68421052631579, "grad_norm": 0.17460066080093384, "learning_rate": 1.1541879768165954e-06, "loss": 0.0098, "step": 9350 }, { "epoch": 19.705263157894738, "grad_norm": 0.22212785482406616, "learning_rate": 1.1191332570507085e-06, "loss": 0.0097, "step": 9360 }, { "epoch": 19.726315789473684, "grad_norm": 0.21805289387702942, "learning_rate": 1.0846130906121132e-06, "loss": 0.0072, "step": 9370 }, { "epoch": 19.74736842105263, "grad_norm": 0.15638454258441925, "learning_rate": 1.0506278550078131e-06, "loss": 0.0066, "step": 9380 }, { "epoch": 19.768421052631577, "grad_norm": 0.1968940645456314, "learning_rate": 1.0171779218949185e-06, "loss": 0.0076, "step": 9390 }, { "epoch": 19.789473684210527, "grad_norm": 0.1103457659482956, "learning_rate": 9.842636570765174e-07, "loss": 0.0072, "step": 9400 }, { "epoch": 19.810526315789474, "grad_norm": 0.15305237472057343, "learning_rate": 9.518854204977612e-07, "loss": 0.008, "step": 9410 }, { "epoch": 19.83157894736842, "grad_norm": 0.17306384444236755, "learning_rate": 9.200435662418349e-07, "loss": 0.0075, "step": 9420 }, { "epoch": 19.852631578947367, "grad_norm": 0.23455674946308136, "learning_rate": 8.887384425261658e-07, "loss": 0.0072, "step": 9430 }, { "epoch": 19.873684210526317, "grad_norm": 0.35442811250686646, "learning_rate": 8.579703916985648e-07, "loss": 0.0067, "step": 9440 }, { "epoch": 19.894736842105264, "grad_norm": 0.10996224731206894, "learning_rate": 8.277397502335194e-07, "loss": 0.0082, "step": 9450 }, { "epoch": 19.91578947368421, "grad_norm": 0.13578014075756073, "learning_rate": 7.980468487284675e-07, "loss": 0.0066, "step": 9460 }, { "epoch": 19.936842105263157, "grad_norm": 0.3439953327178955, "learning_rate": 7.688920119002297e-07, "loss": 0.0075, "step": 9470 }, { "epoch": 19.957894736842107, "grad_norm": 0.14722110331058502, "learning_rate": 7.402755585814269e-07, "loss": 0.007, "step": 9480 }, { "epoch": 19.978947368421053, "grad_norm": 0.3389034867286682, "learning_rate": 7.121978017170073e-07, "loss": 0.0079, "step": 9490 }, { "epoch": 20.0, "grad_norm": 0.139041930437088, "learning_rate": 6.846590483608306e-07, "loss": 0.0081, "step": 9500 }, { "epoch": 20.021052631578947, "grad_norm": 0.12726765871047974, "learning_rate": 6.576595996722834e-07, "loss": 0.0071, "step": 9510 }, { "epoch": 20.042105263157893, "grad_norm": 0.20909473299980164, "learning_rate": 6.311997509130141e-07, "loss": 0.0092, "step": 9520 }, { "epoch": 20.063157894736843, "grad_norm": 0.13614410161972046, "learning_rate": 6.052797914436803e-07, "loss": 0.0062, "step": 9530 }, { "epoch": 20.08421052631579, "grad_norm": 0.41237279772758484, "learning_rate": 5.799000047208181e-07, "loss": 0.0076, "step": 9540 }, { "epoch": 20.105263157894736, "grad_norm": 0.46575409173965454, "learning_rate": 5.550606682937054e-07, "loss": 0.0071, "step": 9550 }, { "epoch": 20.126315789473683, "grad_norm": 0.17044775187969208, "learning_rate": 5.307620538013481e-07, "loss": 0.0068, "step": 9560 }, { "epoch": 20.147368421052633, "grad_norm": 0.14108993113040924, "learning_rate": 5.070044269694874e-07, "loss": 0.0076, "step": 9570 }, { "epoch": 20.16842105263158, "grad_norm": 0.20842072367668152, "learning_rate": 4.837880476077417e-07, "loss": 0.008, "step": 9580 }, { "epoch": 20.189473684210526, "grad_norm": 0.3221527338027954, "learning_rate": 4.6111316960670835e-07, "loss": 0.0081, "step": 9590 }, { "epoch": 20.210526315789473, "grad_norm": 0.17323972284793854, "learning_rate": 4.389800409352218e-07, "loss": 0.0072, "step": 9600 }, { "epoch": 20.231578947368423, "grad_norm": 0.15687398612499237, "learning_rate": 4.173889036376277e-07, "loss": 0.0066, "step": 9610 }, { "epoch": 20.25263157894737, "grad_norm": 0.14287979900836945, "learning_rate": 3.963399938311463e-07, "loss": 0.0084, "step": 9620 }, { "epoch": 20.273684210526316, "grad_norm": 0.24779008328914642, "learning_rate": 3.7583354170328545e-07, "loss": 0.0074, "step": 9630 }, { "epoch": 20.294736842105262, "grad_norm": 0.2461765706539154, "learning_rate": 3.558697715093207e-07, "loss": 0.0078, "step": 9640 }, { "epoch": 20.31578947368421, "grad_norm": 0.204090416431427, "learning_rate": 3.3644890156983576e-07, "loss": 0.0057, "step": 9650 }, { "epoch": 20.33684210526316, "grad_norm": 0.08027578145265579, "learning_rate": 3.175711442683638e-07, "loss": 0.0074, "step": 9660 }, { "epoch": 20.357894736842105, "grad_norm": 0.18983590602874756, "learning_rate": 2.9923670604902197e-07, "loss": 0.0062, "step": 9670 }, { "epoch": 20.378947368421052, "grad_norm": 0.19346703588962555, "learning_rate": 2.814457874143028e-07, "loss": 0.0073, "step": 9680 }, { "epoch": 20.4, "grad_norm": 0.3468160927295685, "learning_rate": 2.641985829228366e-07, "loss": 0.0071, "step": 9690 }, { "epoch": 20.42105263157895, "grad_norm": 0.1895330250263214, "learning_rate": 2.474952811872877e-07, "loss": 0.0064, "step": 9700 }, { "epoch": 20.442105263157895, "grad_norm": 0.1430523842573166, "learning_rate": 2.3133606487228397e-07, "loss": 0.0064, "step": 9710 }, { "epoch": 20.46315789473684, "grad_norm": 0.26815056800842285, "learning_rate": 2.157211106924295e-07, "loss": 0.0071, "step": 9720 }, { "epoch": 20.48421052631579, "grad_norm": 0.29149240255355835, "learning_rate": 2.006505894103672e-07, "loss": 0.0073, "step": 9730 }, { "epoch": 20.50526315789474, "grad_norm": 0.2159319669008255, "learning_rate": 1.8612466583489696e-07, "loss": 0.0072, "step": 9740 }, { "epoch": 20.526315789473685, "grad_norm": 0.1789724975824356, "learning_rate": 1.7214349881918834e-07, "loss": 0.0064, "step": 9750 }, { "epoch": 20.54736842105263, "grad_norm": 0.26327988505363464, "learning_rate": 1.5870724125904845e-07, "loss": 0.0082, "step": 9760 }, { "epoch": 20.568421052631578, "grad_norm": 0.19998528063297272, "learning_rate": 1.4581604009124006e-07, "loss": 0.0088, "step": 9770 }, { "epoch": 20.589473684210525, "grad_norm": 0.1405046284198761, "learning_rate": 1.334700362918717e-07, "loss": 0.0093, "step": 9780 }, { "epoch": 20.610526315789475, "grad_norm": 0.3699970245361328, "learning_rate": 1.2166936487486015e-07, "loss": 0.0064, "step": 9790 }, { "epoch": 20.63157894736842, "grad_norm": 0.2263661026954651, "learning_rate": 1.1041415489045914e-07, "loss": 0.0087, "step": 9800 }, { "epoch": 20.652631578947368, "grad_norm": 0.1515847146511078, "learning_rate": 9.970452942384412e-08, "loss": 0.0062, "step": 9810 }, { "epoch": 20.673684210526314, "grad_norm": 0.19275744259357452, "learning_rate": 8.954060559375754e-08, "loss": 0.0072, "step": 9820 }, { "epoch": 20.694736842105264, "grad_norm": 0.2573884427547455, "learning_rate": 7.99224945512489e-08, "loss": 0.0061, "step": 9830 }, { "epoch": 20.71578947368421, "grad_norm": 0.28303226828575134, "learning_rate": 7.085030147843675e-08, "loss": 0.008, "step": 9840 }, { "epoch": 20.736842105263158, "grad_norm": 0.37284648418426514, "learning_rate": 6.232412558736523e-08, "loss": 0.0079, "step": 9850 }, { "epoch": 20.757894736842104, "grad_norm": 0.12823373079299927, "learning_rate": 5.434406011893822e-08, "loss": 0.0074, "step": 9860 }, { "epoch": 20.778947368421054, "grad_norm": 0.20485688745975494, "learning_rate": 4.6910192341864664e-08, "loss": 0.0113, "step": 9870 }, { "epoch": 20.8, "grad_norm": 0.4243900179862976, "learning_rate": 4.0022603551737035e-08, "loss": 0.0072, "step": 9880 }, { "epoch": 20.821052631578947, "grad_norm": 0.1457342654466629, "learning_rate": 3.3681369070120985e-08, "loss": 0.0068, "step": 9890 }, { "epoch": 20.842105263157894, "grad_norm": 0.16468758881092072, "learning_rate": 2.7886558243744866e-08, "loss": 0.0103, "step": 9900 }, { "epoch": 20.863157894736844, "grad_norm": 0.3381127417087555, "learning_rate": 2.2638234443722596e-08, "loss": 0.0114, "step": 9910 }, { "epoch": 20.88421052631579, "grad_norm": 0.12577301263809204, "learning_rate": 1.7936455064887504e-08, "loss": 0.0092, "step": 9920 }, { "epoch": 20.905263157894737, "grad_norm": 0.2844722270965576, "learning_rate": 1.378127152514841e-08, "loss": 0.0078, "step": 9930 }, { "epoch": 20.926315789473684, "grad_norm": 0.22686950862407684, "learning_rate": 1.0172729264917857e-08, "loss": 0.0069, "step": 9940 }, { "epoch": 20.94736842105263, "grad_norm": 0.18211308121681213, "learning_rate": 7.1108677466458215e-09, "loss": 0.0111, "step": 9950 }, { "epoch": 20.96842105263158, "grad_norm": 0.4254113435745239, "learning_rate": 4.595720454353414e-09, "loss": 0.0101, "step": 9960 }, { "epoch": 20.989473684210527, "grad_norm": 0.17721208930015564, "learning_rate": 2.627314893294264e-09, "loss": 0.0083, "step": 9970 }, { "epoch": 21.010526315789473, "grad_norm": 0.285002201795578, "learning_rate": 1.2056725896270048e-09, "loss": 0.0075, "step": 9980 }, { "epoch": 21.03157894736842, "grad_norm": 0.0966964140534401, "learning_rate": 3.308090902098826e-10, "loss": 0.0084, "step": 9990 }, { "epoch": 21.05263157894737, "grad_norm": 0.20568659901618958, "learning_rate": 2.7339624120159555e-12, "loss": 0.0098, "step": 10000 }, { "epoch": 21.073684210526316, "grad_norm": 0.18844188749790192, "learning_rate": 5.405481072206917e-05, "loss": 0.0074, "step": 10010 }, { "epoch": 21.094736842105263, "grad_norm": 0.4040485918521881, "learning_rate": 5.397240402940402e-05, "loss": 0.0098, "step": 10020 }, { "epoch": 21.11578947368421, "grad_norm": 0.3954405188560486, "learning_rate": 5.388998647633794e-05, "loss": 0.0117, "step": 10030 }, { "epoch": 21.13684210526316, "grad_norm": 0.26791974902153015, "learning_rate": 5.380755828819737e-05, "loss": 0.012, "step": 10040 }, { "epoch": 21.157894736842106, "grad_norm": 0.416629433631897, "learning_rate": 5.3725119690337846e-05, "loss": 0.0115, "step": 10050 }, { "epoch": 21.178947368421053, "grad_norm": 0.24902381002902985, "learning_rate": 5.3642670908143324e-05, "loss": 0.0118, "step": 10060 }, { "epoch": 21.2, "grad_norm": 0.4256839156150818, "learning_rate": 5.356021216702562e-05, "loss": 0.0154, "step": 10070 }, { "epoch": 21.221052631578946, "grad_norm": 0.3614608943462372, "learning_rate": 5.347774369242381e-05, "loss": 0.0144, "step": 10080 }, { "epoch": 21.242105263157896, "grad_norm": 0.47240909934043884, "learning_rate": 5.3395265709803545e-05, "loss": 0.0123, "step": 10090 }, { "epoch": 21.263157894736842, "grad_norm": 0.3689449429512024, "learning_rate": 5.331277844465647e-05, "loss": 0.0132, "step": 10100 }, { "epoch": 21.28421052631579, "grad_norm": 0.36580732464790344, "learning_rate": 5.323028212249963e-05, "loss": 0.0153, "step": 10110 }, { "epoch": 21.305263157894736, "grad_norm": 0.21029132604599, "learning_rate": 5.314777696887481e-05, "loss": 0.0113, "step": 10120 }, { "epoch": 21.326315789473686, "grad_norm": 0.2312794327735901, "learning_rate": 5.306526320934796e-05, "loss": 0.0144, "step": 10130 }, { "epoch": 21.347368421052632, "grad_norm": 0.2962414026260376, "learning_rate": 5.298274106950854e-05, "loss": 0.0129, "step": 10140 }, { "epoch": 21.36842105263158, "grad_norm": 0.3565818965435028, "learning_rate": 5.290021077496893e-05, "loss": 0.0134, "step": 10150 }, { "epoch": 21.389473684210525, "grad_norm": 0.3331308662891388, "learning_rate": 5.2817672551363816e-05, "loss": 0.0132, "step": 10160 }, { "epoch": 21.410526315789475, "grad_norm": 0.2898876368999481, "learning_rate": 5.273512662434952e-05, "loss": 0.0121, "step": 10170 }, { "epoch": 21.431578947368422, "grad_norm": 0.2981153726577759, "learning_rate": 5.265257321960349e-05, "loss": 0.0141, "step": 10180 }, { "epoch": 21.45263157894737, "grad_norm": 0.3445838391780853, "learning_rate": 5.257001256282357e-05, "loss": 0.0174, "step": 10190 }, { "epoch": 21.473684210526315, "grad_norm": 0.2669433653354645, "learning_rate": 5.248744487972742e-05, "loss": 0.0141, "step": 10200 }, { "epoch": 21.49473684210526, "grad_norm": 0.3064207136631012, "learning_rate": 5.240487039605196e-05, "loss": 0.0104, "step": 10210 }, { "epoch": 21.51578947368421, "grad_norm": 0.2684043049812317, "learning_rate": 5.232228933755267e-05, "loss": 0.0126, "step": 10220 }, { "epoch": 21.53684210526316, "grad_norm": 0.40115877985954285, "learning_rate": 5.2239701930003006e-05, "loss": 0.0123, "step": 10230 }, { "epoch": 21.557894736842105, "grad_norm": 0.30079665780067444, "learning_rate": 5.215710839919379e-05, "loss": 0.0127, "step": 10240 }, { "epoch": 21.57894736842105, "grad_norm": 0.3688611090183258, "learning_rate": 5.207450897093257e-05, "loss": 0.0127, "step": 10250 }, { "epoch": 21.6, "grad_norm": 0.3397323489189148, "learning_rate": 5.1991903871043046e-05, "loss": 0.0123, "step": 10260 }, { "epoch": 21.621052631578948, "grad_norm": 0.24516400694847107, "learning_rate": 5.190929332536439e-05, "loss": 0.0125, "step": 10270 }, { "epoch": 21.642105263157895, "grad_norm": 0.3106752634048462, "learning_rate": 5.182667755975071e-05, "loss": 0.0125, "step": 10280 }, { "epoch": 21.66315789473684, "grad_norm": 0.2998864948749542, "learning_rate": 5.1744056800070315e-05, "loss": 0.0102, "step": 10290 }, { "epoch": 21.68421052631579, "grad_norm": 0.34219709038734436, "learning_rate": 5.166143127220524e-05, "loss": 0.0129, "step": 10300 }, { "epoch": 21.705263157894738, "grad_norm": 0.20866811275482178, "learning_rate": 5.1578801202050485e-05, "loss": 0.011, "step": 10310 }, { "epoch": 21.726315789473684, "grad_norm": 0.2867657244205475, "learning_rate": 5.149616681551355e-05, "loss": 0.0114, "step": 10320 }, { "epoch": 21.74736842105263, "grad_norm": 0.2084168642759323, "learning_rate": 5.141352833851367e-05, "loss": 0.0137, "step": 10330 }, { "epoch": 21.768421052631577, "grad_norm": 0.2571122348308563, "learning_rate": 5.1330885996981285e-05, "loss": 0.0151, "step": 10340 }, { "epoch": 21.789473684210527, "grad_norm": 0.46845701336860657, "learning_rate": 5.124824001685741e-05, "loss": 0.0107, "step": 10350 }, { "epoch": 21.810526315789474, "grad_norm": 0.298113077878952, "learning_rate": 5.116559062409298e-05, "loss": 0.0103, "step": 10360 }, { "epoch": 21.83157894736842, "grad_norm": 0.28352585434913635, "learning_rate": 5.10829380446483e-05, "loss": 0.011, "step": 10370 }, { "epoch": 21.852631578947367, "grad_norm": 0.27861908078193665, "learning_rate": 5.100028250449235e-05, "loss": 0.0124, "step": 10380 }, { "epoch": 21.873684210526317, "grad_norm": 0.45851364731788635, "learning_rate": 5.0917624229602234e-05, "loss": 0.0117, "step": 10390 }, { "epoch": 21.894736842105264, "grad_norm": 0.30498942732810974, "learning_rate": 5.0834963445962524e-05, "loss": 0.0129, "step": 10400 }, { "epoch": 21.91578947368421, "grad_norm": 0.30296874046325684, "learning_rate": 5.075230037956461e-05, "loss": 0.0139, "step": 10410 }, { "epoch": 21.936842105263157, "grad_norm": 0.2135019302368164, "learning_rate": 5.0669635256406213e-05, "loss": 0.0106, "step": 10420 }, { "epoch": 21.957894736842107, "grad_norm": 0.22030536830425262, "learning_rate": 5.058696830249058e-05, "loss": 0.0113, "step": 10430 }, { "epoch": 21.978947368421053, "grad_norm": 0.17459960281848907, "learning_rate": 5.050429974382602e-05, "loss": 0.0111, "step": 10440 }, { "epoch": 22.0, "grad_norm": 0.320172518491745, "learning_rate": 5.042162980642523e-05, "loss": 0.0114, "step": 10450 }, { "epoch": 22.021052631578947, "grad_norm": 0.48593223094940186, "learning_rate": 5.033895871630462e-05, "loss": 0.0106, "step": 10460 }, { "epoch": 22.042105263157893, "grad_norm": 0.37920939922332764, "learning_rate": 5.025628669948386e-05, "loss": 0.0093, "step": 10470 }, { "epoch": 22.063157894736843, "grad_norm": 0.35398560762405396, "learning_rate": 5.017361398198502e-05, "loss": 0.0121, "step": 10480 }, { "epoch": 22.08421052631579, "grad_norm": 0.45290523767471313, "learning_rate": 5.009094078983221e-05, "loss": 0.0117, "step": 10490 }, { "epoch": 22.105263157894736, "grad_norm": 0.32007813453674316, "learning_rate": 5.000826734905073e-05, "loss": 0.012, "step": 10500 }, { "epoch": 22.126315789473683, "grad_norm": 0.33983558416366577, "learning_rate": 4.9925593885666645e-05, "loss": 0.015, "step": 10510 }, { "epoch": 22.147368421052633, "grad_norm": 0.24503061175346375, "learning_rate": 4.984292062570602e-05, "loss": 0.0103, "step": 10520 }, { "epoch": 22.16842105263158, "grad_norm": 0.32504063844680786, "learning_rate": 4.976024779519442e-05, "loss": 0.0143, "step": 10530 }, { "epoch": 22.189473684210526, "grad_norm": 0.16512326896190643, "learning_rate": 4.9677575620156194e-05, "loss": 0.0123, "step": 10540 }, { "epoch": 22.210526315789473, "grad_norm": 0.3169327974319458, "learning_rate": 4.959490432661391e-05, "loss": 0.0168, "step": 10550 }, { "epoch": 22.231578947368423, "grad_norm": 0.3156251013278961, "learning_rate": 4.9512234140587726e-05, "loss": 0.0102, "step": 10560 }, { "epoch": 22.25263157894737, "grad_norm": 0.22277295589447021, "learning_rate": 4.942956528809477e-05, "loss": 0.0127, "step": 10570 }, { "epoch": 22.273684210526316, "grad_norm": 0.2656615376472473, "learning_rate": 4.934689799514854e-05, "loss": 0.0135, "step": 10580 }, { "epoch": 22.294736842105262, "grad_norm": 0.3086012005805969, "learning_rate": 4.926423248775827e-05, "loss": 0.0126, "step": 10590 }, { "epoch": 22.31578947368421, "grad_norm": 0.33681073784828186, "learning_rate": 4.918156899192826e-05, "loss": 0.011, "step": 10600 }, { "epoch": 22.33684210526316, "grad_norm": 0.3496444523334503, "learning_rate": 4.909890773365738e-05, "loss": 0.0121, "step": 10610 }, { "epoch": 22.357894736842105, "grad_norm": 0.24939432740211487, "learning_rate": 4.9016248938938344e-05, "loss": 0.0124, "step": 10620 }, { "epoch": 22.378947368421052, "grad_norm": 0.31838706135749817, "learning_rate": 4.8933592833757156e-05, "loss": 0.0118, "step": 10630 }, { "epoch": 22.4, "grad_norm": 0.2896694540977478, "learning_rate": 4.8850939644092435e-05, "loss": 0.013, "step": 10640 }, { "epoch": 22.42105263157895, "grad_norm": 0.3095000088214874, "learning_rate": 4.876828959591485e-05, "loss": 0.0104, "step": 10650 }, { "epoch": 22.442105263157895, "grad_norm": 0.22305376827716827, "learning_rate": 4.8685642915186474e-05, "loss": 0.0107, "step": 10660 }, { "epoch": 22.46315789473684, "grad_norm": 0.251404345035553, "learning_rate": 4.860299982786018e-05, "loss": 0.0111, "step": 10670 }, { "epoch": 22.48421052631579, "grad_norm": 0.37422317266464233, "learning_rate": 4.852036055987901e-05, "loss": 0.011, "step": 10680 }, { "epoch": 22.50526315789474, "grad_norm": 0.2603485584259033, "learning_rate": 4.843772533717558e-05, "loss": 0.015, "step": 10690 }, { "epoch": 22.526315789473685, "grad_norm": 0.26270291209220886, "learning_rate": 4.835509438567142e-05, "loss": 0.0098, "step": 10700 }, { "epoch": 22.54736842105263, "grad_norm": 0.2557259500026703, "learning_rate": 4.827246793127639e-05, "loss": 0.0124, "step": 10710 }, { "epoch": 22.568421052631578, "grad_norm": 0.5568275451660156, "learning_rate": 4.818984619988807e-05, "loss": 0.0138, "step": 10720 }, { "epoch": 22.589473684210525, "grad_norm": 0.2866614758968353, "learning_rate": 4.810722941739115e-05, "loss": 0.0111, "step": 10730 }, { "epoch": 22.610526315789475, "grad_norm": 0.2378075271844864, "learning_rate": 4.8024617809656684e-05, "loss": 0.0094, "step": 10740 }, { "epoch": 22.63157894736842, "grad_norm": 0.36355796456336975, "learning_rate": 4.794201160254171e-05, "loss": 0.0116, "step": 10750 }, { "epoch": 22.652631578947368, "grad_norm": 0.20442569255828857, "learning_rate": 4.785941102188844e-05, "loss": 0.0107, "step": 10760 }, { "epoch": 22.673684210526314, "grad_norm": 0.3297772705554962, "learning_rate": 4.7776816293523686e-05, "loss": 0.0116, "step": 10770 }, { "epoch": 22.694736842105264, "grad_norm": 0.3343777060508728, "learning_rate": 4.769422764325832e-05, "loss": 0.0109, "step": 10780 }, { "epoch": 22.71578947368421, "grad_norm": 0.3144363462924957, "learning_rate": 4.76116452968865e-05, "loss": 0.0105, "step": 10790 }, { "epoch": 22.736842105263158, "grad_norm": 0.20294076204299927, "learning_rate": 4.752906948018525e-05, "loss": 0.0112, "step": 10800 }, { "epoch": 22.757894736842104, "grad_norm": 0.3381933867931366, "learning_rate": 4.7446500418913684e-05, "loss": 0.0169, "step": 10810 }, { "epoch": 22.778947368421054, "grad_norm": 0.324470579624176, "learning_rate": 4.736393833881247e-05, "loss": 0.0123, "step": 10820 }, { "epoch": 22.8, "grad_norm": 0.5833401083946228, "learning_rate": 4.7281383465603194e-05, "loss": 0.013, "step": 10830 }, { "epoch": 22.821052631578947, "grad_norm": 0.22370722889900208, "learning_rate": 4.71988360249877e-05, "loss": 0.0103, "step": 10840 }, { "epoch": 22.842105263157894, "grad_norm": 0.2853754460811615, "learning_rate": 4.7116296242647554e-05, "loss": 0.011, "step": 10850 }, { "epoch": 22.863157894736844, "grad_norm": 0.33950573205947876, "learning_rate": 4.703376434424336e-05, "loss": 0.0127, "step": 10860 }, { "epoch": 22.88421052631579, "grad_norm": 0.2670488953590393, "learning_rate": 4.695124055541421e-05, "loss": 0.0139, "step": 10870 }, { "epoch": 22.905263157894737, "grad_norm": 0.2606572210788727, "learning_rate": 4.6868725101776934e-05, "loss": 0.0102, "step": 10880 }, { "epoch": 22.926315789473684, "grad_norm": 0.18392933905124664, "learning_rate": 4.678621820892567e-05, "loss": 0.0108, "step": 10890 }, { "epoch": 22.94736842105263, "grad_norm": 0.37060001492500305, "learning_rate": 4.670372010243111e-05, "loss": 0.0115, "step": 10900 }, { "epoch": 22.96842105263158, "grad_norm": 0.2832513451576233, "learning_rate": 4.662123100783992e-05, "loss": 0.012, "step": 10910 }, { "epoch": 22.989473684210527, "grad_norm": 0.3530137538909912, "learning_rate": 4.653875115067415e-05, "loss": 0.0113, "step": 10920 }, { "epoch": 23.010526315789473, "grad_norm": 0.212493434548378, "learning_rate": 4.6456280756430545e-05, "loss": 0.0094, "step": 10930 }, { "epoch": 23.03157894736842, "grad_norm": 0.1956809163093567, "learning_rate": 4.637382005058004e-05, "loss": 0.0099, "step": 10940 }, { "epoch": 23.05263157894737, "grad_norm": 0.314688116312027, "learning_rate": 4.629136925856705e-05, "loss": 0.0109, "step": 10950 }, { "epoch": 23.073684210526316, "grad_norm": 0.336753785610199, "learning_rate": 4.6208928605808895e-05, "loss": 0.0095, "step": 10960 }, { "epoch": 23.094736842105263, "grad_norm": 0.2907612919807434, "learning_rate": 4.612649831769519e-05, "loss": 0.0097, "step": 10970 }, { "epoch": 23.11578947368421, "grad_norm": 0.2436918318271637, "learning_rate": 4.604407861958715e-05, "loss": 0.0113, "step": 10980 }, { "epoch": 23.13684210526316, "grad_norm": 0.16787075996398926, "learning_rate": 4.5961669736817114e-05, "loss": 0.0093, "step": 10990 }, { "epoch": 23.157894736842106, "grad_norm": 0.2536000609397888, "learning_rate": 4.5879271894687814e-05, "loss": 0.0111, "step": 11000 }, { "epoch": 23.178947368421053, "grad_norm": 0.2601369321346283, "learning_rate": 4.5796885318471826e-05, "loss": 0.0101, "step": 11010 }, { "epoch": 23.2, "grad_norm": 0.23308065533638, "learning_rate": 4.571451023341086e-05, "loss": 0.0121, "step": 11020 }, { "epoch": 23.221052631578946, "grad_norm": 0.20177806913852692, "learning_rate": 4.563214686471527e-05, "loss": 0.0113, "step": 11030 }, { "epoch": 23.242105263157896, "grad_norm": 0.2421198934316635, "learning_rate": 4.5549795437563365e-05, "loss": 0.0116, "step": 11040 }, { "epoch": 23.263157894736842, "grad_norm": 0.2715374231338501, "learning_rate": 4.546745617710081e-05, "loss": 0.0114, "step": 11050 }, { "epoch": 23.28421052631579, "grad_norm": 0.32884082198143005, "learning_rate": 4.5385129308440014e-05, "loss": 0.0099, "step": 11060 }, { "epoch": 23.305263157894736, "grad_norm": 0.2899945080280304, "learning_rate": 4.530281505665944e-05, "loss": 0.0099, "step": 11070 }, { "epoch": 23.326315789473686, "grad_norm": 0.2816093862056732, "learning_rate": 4.5220513646803134e-05, "loss": 0.0121, "step": 11080 }, { "epoch": 23.347368421052632, "grad_norm": 0.290099561214447, "learning_rate": 4.513822530388003e-05, "loss": 0.0112, "step": 11090 }, { "epoch": 23.36842105263158, "grad_norm": 0.312424898147583, "learning_rate": 4.5055950252863296e-05, "loss": 0.0109, "step": 11100 }, { "epoch": 23.389473684210525, "grad_norm": 0.2681463956832886, "learning_rate": 4.4973688718689803e-05, "loss": 0.0092, "step": 11110 }, { "epoch": 23.410526315789475, "grad_norm": 0.24714350700378418, "learning_rate": 4.4891440926259406e-05, "loss": 0.0106, "step": 11120 }, { "epoch": 23.431578947368422, "grad_norm": 0.4299587309360504, "learning_rate": 4.480920710043443e-05, "loss": 0.0106, "step": 11130 }, { "epoch": 23.45263157894737, "grad_norm": 0.20013949275016785, "learning_rate": 4.4726987466039044e-05, "loss": 0.0095, "step": 11140 }, { "epoch": 23.473684210526315, "grad_norm": 0.2793571352958679, "learning_rate": 4.46447822478586e-05, "loss": 0.009, "step": 11150 }, { "epoch": 23.49473684210526, "grad_norm": 0.2682344317436218, "learning_rate": 4.4562591670638974e-05, "loss": 0.0113, "step": 11160 }, { "epoch": 23.51578947368421, "grad_norm": 0.3045579791069031, "learning_rate": 4.4480415959086105e-05, "loss": 0.0078, "step": 11170 }, { "epoch": 23.53684210526316, "grad_norm": 0.38914862275123596, "learning_rate": 4.439825533786522e-05, "loss": 0.0129, "step": 11180 }, { "epoch": 23.557894736842105, "grad_norm": 0.23205167055130005, "learning_rate": 4.431611003160035e-05, "loss": 0.0104, "step": 11190 }, { "epoch": 23.57894736842105, "grad_norm": 0.3094616234302521, "learning_rate": 4.4233980264873636e-05, "loss": 0.0102, "step": 11200 }, { "epoch": 23.6, "grad_norm": 0.2169589251279831, "learning_rate": 4.4151866262224684e-05, "loss": 0.0097, "step": 11210 }, { "epoch": 23.621052631578948, "grad_norm": 0.29721760749816895, "learning_rate": 4.406976824815006e-05, "loss": 0.0092, "step": 11220 }, { "epoch": 23.642105263157895, "grad_norm": 0.30239713191986084, "learning_rate": 4.3987686447102595e-05, "loss": 0.0117, "step": 11230 }, { "epoch": 23.66315789473684, "grad_norm": 0.4493488073348999, "learning_rate": 4.3905621083490804e-05, "loss": 0.0098, "step": 11240 }, { "epoch": 23.68421052631579, "grad_norm": 0.2181037962436676, "learning_rate": 4.3823572381678286e-05, "loss": 0.0098, "step": 11250 }, { "epoch": 23.705263157894738, "grad_norm": 0.23776207864284515, "learning_rate": 4.374154056598301e-05, "loss": 0.012, "step": 11260 }, { "epoch": 23.726315789473684, "grad_norm": 0.25478363037109375, "learning_rate": 4.3659525860676845e-05, "loss": 0.0102, "step": 11270 }, { "epoch": 23.74736842105263, "grad_norm": 0.3424645960330963, "learning_rate": 4.3577528489984854e-05, "loss": 0.0126, "step": 11280 }, { "epoch": 23.768421052631577, "grad_norm": 0.2040155977010727, "learning_rate": 4.349554867808476e-05, "loss": 0.0126, "step": 11290 }, { "epoch": 23.789473684210527, "grad_norm": 0.2485988289117813, "learning_rate": 4.34135866491062e-05, "loss": 0.0094, "step": 11300 }, { "epoch": 23.810526315789474, "grad_norm": 0.3788502514362335, "learning_rate": 4.333164262713022e-05, "loss": 0.0103, "step": 11310 }, { "epoch": 23.83157894736842, "grad_norm": 0.3531363308429718, "learning_rate": 4.324971683618868e-05, "loss": 0.0105, "step": 11320 }, { "epoch": 23.852631578947367, "grad_norm": 0.28572991490364075, "learning_rate": 4.316780950026354e-05, "loss": 0.0107, "step": 11330 }, { "epoch": 23.873684210526317, "grad_norm": 0.3336009383201599, "learning_rate": 4.308592084328637e-05, "loss": 0.0101, "step": 11340 }, { "epoch": 23.894736842105264, "grad_norm": 0.2713831663131714, "learning_rate": 4.3004051089137576e-05, "loss": 0.0107, "step": 11350 }, { "epoch": 23.91578947368421, "grad_norm": 0.217068612575531, "learning_rate": 4.292220046164597e-05, "loss": 0.0141, "step": 11360 }, { "epoch": 23.936842105263157, "grad_norm": 0.2655010521411896, "learning_rate": 4.2840369184588035e-05, "loss": 0.0129, "step": 11370 }, { "epoch": 23.957894736842107, "grad_norm": 0.27458834648132324, "learning_rate": 4.2758557481687345e-05, "loss": 0.0136, "step": 11380 }, { "epoch": 23.978947368421053, "grad_norm": 0.20840920507907867, "learning_rate": 4.267676557661403e-05, "loss": 0.0107, "step": 11390 }, { "epoch": 24.0, "grad_norm": 0.2673669755458832, "learning_rate": 4.2594993692983955e-05, "loss": 0.0101, "step": 11400 }, { "epoch": 24.021052631578947, "grad_norm": 0.36723533272743225, "learning_rate": 4.251324205435837e-05, "loss": 0.0121, "step": 11410 }, { "epoch": 24.042105263157893, "grad_norm": 0.3276667892932892, "learning_rate": 4.243151088424312e-05, "loss": 0.0084, "step": 11420 }, { "epoch": 24.063157894736843, "grad_norm": 0.4714983105659485, "learning_rate": 4.234980040608813e-05, "loss": 0.0103, "step": 11430 }, { "epoch": 24.08421052631579, "grad_norm": 0.3111778199672699, "learning_rate": 4.22681108432867e-05, "loss": 0.0109, "step": 11440 }, { "epoch": 24.105263157894736, "grad_norm": 0.3208775222301483, "learning_rate": 4.2186442419174984e-05, "loss": 0.0111, "step": 11450 }, { "epoch": 24.126315789473683, "grad_norm": 0.40515458583831787, "learning_rate": 4.210479535703133e-05, "loss": 0.0112, "step": 11460 }, { "epoch": 24.147368421052633, "grad_norm": 0.16872435808181763, "learning_rate": 4.202316988007567e-05, "loss": 0.0084, "step": 11470 }, { "epoch": 24.16842105263158, "grad_norm": 0.23413999378681183, "learning_rate": 4.194156621146901e-05, "loss": 0.0102, "step": 11480 }, { "epoch": 24.189473684210526, "grad_norm": 0.24059568345546722, "learning_rate": 4.1859984574312596e-05, "loss": 0.0098, "step": 11490 }, { "epoch": 24.210526315789473, "grad_norm": 0.5043342113494873, "learning_rate": 4.177842519164752e-05, "loss": 0.008, "step": 11500 }, { "epoch": 24.231578947368423, "grad_norm": 0.22272741794586182, "learning_rate": 4.169688828645404e-05, "loss": 0.009, "step": 11510 }, { "epoch": 24.25263157894737, "grad_norm": 0.29004842042922974, "learning_rate": 4.161537408165092e-05, "loss": 0.0098, "step": 11520 }, { "epoch": 24.273684210526316, "grad_norm": 0.22499515116214752, "learning_rate": 4.1533882800094924e-05, "loss": 0.0113, "step": 11530 }, { "epoch": 24.294736842105262, "grad_norm": 0.2771598696708679, "learning_rate": 4.145241466458005e-05, "loss": 0.0117, "step": 11540 }, { "epoch": 24.31578947368421, "grad_norm": 0.45312973856925964, "learning_rate": 4.13709698978371e-05, "loss": 0.0103, "step": 11550 }, { "epoch": 24.33684210526316, "grad_norm": 0.33605921268463135, "learning_rate": 4.1289548722532944e-05, "loss": 0.013, "step": 11560 }, { "epoch": 24.357894736842105, "grad_norm": 0.24979981780052185, "learning_rate": 4.120815136126999e-05, "loss": 0.0116, "step": 11570 }, { "epoch": 24.378947368421052, "grad_norm": 0.21812357008457184, "learning_rate": 4.112677803658548e-05, "loss": 0.0107, "step": 11580 }, { "epoch": 24.4, "grad_norm": 0.18380087614059448, "learning_rate": 4.1045428970951e-05, "loss": 0.0102, "step": 11590 }, { "epoch": 24.42105263157895, "grad_norm": 0.3362160623073578, "learning_rate": 4.0964104386771785e-05, "loss": 0.0104, "step": 11600 }, { "epoch": 24.442105263157895, "grad_norm": 0.21821846067905426, "learning_rate": 4.0882804506386144e-05, "loss": 0.0108, "step": 11610 }, { "epoch": 24.46315789473684, "grad_norm": 0.27546316385269165, "learning_rate": 4.080152955206485e-05, "loss": 0.0076, "step": 11620 }, { "epoch": 24.48421052631579, "grad_norm": 0.20614686608314514, "learning_rate": 4.0720279746010505e-05, "loss": 0.0101, "step": 11630 }, { "epoch": 24.50526315789474, "grad_norm": 0.22117513418197632, "learning_rate": 4.063905531035699e-05, "loss": 0.01, "step": 11640 }, { "epoch": 24.526315789473685, "grad_norm": 0.21417807042598724, "learning_rate": 4.055785646716882e-05, "loss": 0.0092, "step": 11650 }, { "epoch": 24.54736842105263, "grad_norm": 0.28614434599876404, "learning_rate": 4.047668343844051e-05, "loss": 0.0104, "step": 11660 }, { "epoch": 24.568421052631578, "grad_norm": 0.22346588969230652, "learning_rate": 4.039553644609604e-05, "loss": 0.0119, "step": 11670 }, { "epoch": 24.589473684210525, "grad_norm": 0.19713829457759857, "learning_rate": 4.0314415711988176e-05, "loss": 0.0097, "step": 11680 }, { "epoch": 24.610526315789475, "grad_norm": 0.2163500189781189, "learning_rate": 4.023332145789792e-05, "loss": 0.0091, "step": 11690 }, { "epoch": 24.63157894736842, "grad_norm": 0.727031946182251, "learning_rate": 4.015225390553385e-05, "loss": 0.0086, "step": 11700 }, { "epoch": 24.652631578947368, "grad_norm": 0.30268704891204834, "learning_rate": 4.007121327653158e-05, "loss": 0.0101, "step": 11710 }, { "epoch": 24.673684210526314, "grad_norm": 0.3467192053794861, "learning_rate": 3.9990199792453064e-05, "loss": 0.0092, "step": 11720 }, { "epoch": 24.694736842105264, "grad_norm": 0.29946407675743103, "learning_rate": 3.9909213674786103e-05, "loss": 0.012, "step": 11730 }, { "epoch": 24.71578947368421, "grad_norm": 0.2599567770957947, "learning_rate": 3.982825514494363e-05, "loss": 0.0115, "step": 11740 }, { "epoch": 24.736842105263158, "grad_norm": 0.2873029112815857, "learning_rate": 3.974732442426319e-05, "loss": 0.0116, "step": 11750 }, { "epoch": 24.757894736842104, "grad_norm": 0.2502472698688507, "learning_rate": 3.966642173400629e-05, "loss": 0.0109, "step": 11760 }, { "epoch": 24.778947368421054, "grad_norm": 0.2317812591791153, "learning_rate": 3.9585547295357764e-05, "loss": 0.0124, "step": 11770 }, { "epoch": 24.8, "grad_norm": 0.5006920099258423, "learning_rate": 3.950470132942526e-05, "loss": 0.012, "step": 11780 }, { "epoch": 24.821052631578947, "grad_norm": 0.21395473182201385, "learning_rate": 3.942388405723856e-05, "loss": 0.0128, "step": 11790 }, { "epoch": 24.842105263157894, "grad_norm": 0.2722606658935547, "learning_rate": 3.9343095699749e-05, "loss": 0.0138, "step": 11800 }, { "epoch": 24.863157894736844, "grad_norm": 0.3033861219882965, "learning_rate": 3.9262336477828874e-05, "loss": 0.0095, "step": 11810 }, { "epoch": 24.88421052631579, "grad_norm": 0.25664418935775757, "learning_rate": 3.9181606612270794e-05, "loss": 0.0144, "step": 11820 }, { "epoch": 24.905263157894737, "grad_norm": 0.3629527688026428, "learning_rate": 3.910090632378713e-05, "loss": 0.0085, "step": 11830 }, { "epoch": 24.926315789473684, "grad_norm": 0.2363012433052063, "learning_rate": 3.90202358330094e-05, "loss": 0.0095, "step": 11840 }, { "epoch": 24.94736842105263, "grad_norm": 0.19582737982273102, "learning_rate": 3.8939595360487656e-05, "loss": 0.0112, "step": 11850 }, { "epoch": 24.96842105263158, "grad_norm": 0.32092416286468506, "learning_rate": 3.885898512668984e-05, "loss": 0.009, "step": 11860 }, { "epoch": 24.989473684210527, "grad_norm": 0.23234784603118896, "learning_rate": 3.877840535200127e-05, "loss": 0.0099, "step": 11870 }, { "epoch": 25.010526315789473, "grad_norm": 0.2076670527458191, "learning_rate": 3.869785625672397e-05, "loss": 0.0095, "step": 11880 }, { "epoch": 25.03157894736842, "grad_norm": 0.23365065455436707, "learning_rate": 3.8617338061076094e-05, "loss": 0.0085, "step": 11890 }, { "epoch": 25.05263157894737, "grad_norm": 0.23059524595737457, "learning_rate": 3.853685098519132e-05, "loss": 0.0106, "step": 11900 }, { "epoch": 25.073684210526316, "grad_norm": 0.21644051373004913, "learning_rate": 3.845639524911823e-05, "loss": 0.008, "step": 11910 }, { "epoch": 25.094736842105263, "grad_norm": 0.3645325303077698, "learning_rate": 3.837597107281974e-05, "loss": 0.0094, "step": 11920 }, { "epoch": 25.11578947368421, "grad_norm": 0.3703218698501587, "learning_rate": 3.829557867617247e-05, "loss": 0.0106, "step": 11930 }, { "epoch": 25.13684210526316, "grad_norm": 0.29417309165000916, "learning_rate": 3.821521827896618e-05, "loss": 0.0124, "step": 11940 }, { "epoch": 25.157894736842106, "grad_norm": 0.2348426729440689, "learning_rate": 3.81348901009031e-05, "loss": 0.0124, "step": 11950 }, { "epoch": 25.178947368421053, "grad_norm": 0.2611774206161499, "learning_rate": 3.805459436159741e-05, "loss": 0.0104, "step": 11960 }, { "epoch": 25.2, "grad_norm": 0.24029992520809174, "learning_rate": 3.797433128057461e-05, "loss": 0.0113, "step": 11970 }, { "epoch": 25.221052631578946, "grad_norm": 0.23153045773506165, "learning_rate": 3.789410107727089e-05, "loss": 0.0096, "step": 11980 }, { "epoch": 25.242105263157896, "grad_norm": 0.26612886786460876, "learning_rate": 3.781390397103257e-05, "loss": 0.0119, "step": 11990 }, { "epoch": 25.263157894736842, "grad_norm": 0.28183913230895996, "learning_rate": 3.7733740181115455e-05, "loss": 0.0114, "step": 12000 }, { "epoch": 25.28421052631579, "grad_norm": 0.17333750426769257, "learning_rate": 3.7653609926684306e-05, "loss": 0.0119, "step": 12010 }, { "epoch": 25.305263157894736, "grad_norm": 0.17891530692577362, "learning_rate": 3.757351342681217e-05, "loss": 0.0084, "step": 12020 }, { "epoch": 25.326315789473686, "grad_norm": 0.16984692215919495, "learning_rate": 3.749345090047982e-05, "loss": 0.0101, "step": 12030 }, { "epoch": 25.347368421052632, "grad_norm": 0.19870617985725403, "learning_rate": 3.741342256657515e-05, "loss": 0.0133, "step": 12040 }, { "epoch": 25.36842105263158, "grad_norm": 0.22995954751968384, "learning_rate": 3.7333428643892567e-05, "loss": 0.0104, "step": 12050 }, { "epoch": 25.389473684210525, "grad_norm": 0.3034108877182007, "learning_rate": 3.725346935113239e-05, "loss": 0.0091, "step": 12060 }, { "epoch": 25.410526315789475, "grad_norm": 0.23518213629722595, "learning_rate": 3.717354490690029e-05, "loss": 0.0127, "step": 12070 }, { "epoch": 25.431578947368422, "grad_norm": 0.15414737164974213, "learning_rate": 3.709365552970664e-05, "loss": 0.0092, "step": 12080 }, { "epoch": 25.45263157894737, "grad_norm": 0.23784324526786804, "learning_rate": 3.7013801437965945e-05, "loss": 0.0108, "step": 12090 }, { "epoch": 25.473684210526315, "grad_norm": 0.2581232190132141, "learning_rate": 3.693398284999623e-05, "loss": 0.01, "step": 12100 }, { "epoch": 25.49473684210526, "grad_norm": 0.32191982865333557, "learning_rate": 3.6854199984018484e-05, "loss": 0.011, "step": 12110 }, { "epoch": 25.51578947368421, "grad_norm": 0.2644596993923187, "learning_rate": 3.677445305815601e-05, "loss": 0.0092, "step": 12120 }, { "epoch": 25.53684210526316, "grad_norm": 0.19325336813926697, "learning_rate": 3.669474229043387e-05, "loss": 0.0108, "step": 12130 }, { "epoch": 25.557894736842105, "grad_norm": 0.2987186312675476, "learning_rate": 3.6615067898778235e-05, "loss": 0.0108, "step": 12140 }, { "epoch": 25.57894736842105, "grad_norm": 0.24576792120933533, "learning_rate": 3.6535430101015866e-05, "loss": 0.0079, "step": 12150 }, { "epoch": 25.6, "grad_norm": 0.2080317586660385, "learning_rate": 3.645582911487345e-05, "loss": 0.0113, "step": 12160 }, { "epoch": 25.621052631578948, "grad_norm": 0.23942561447620392, "learning_rate": 3.637626515797706e-05, "loss": 0.0102, "step": 12170 }, { "epoch": 25.642105263157895, "grad_norm": 0.3159041404724121, "learning_rate": 3.629673844785152e-05, "loss": 0.0098, "step": 12180 }, { "epoch": 25.66315789473684, "grad_norm": 0.288010835647583, "learning_rate": 3.621724920191979e-05, "loss": 0.0111, "step": 12190 }, { "epoch": 25.68421052631579, "grad_norm": 0.22122125327587128, "learning_rate": 3.6137797637502444e-05, "loss": 0.0097, "step": 12200 }, { "epoch": 25.705263157894738, "grad_norm": 0.29228949546813965, "learning_rate": 3.6058383971817035e-05, "loss": 0.0088, "step": 12210 }, { "epoch": 25.726315789473684, "grad_norm": 0.28815481066703796, "learning_rate": 3.59790084219775e-05, "loss": 0.0131, "step": 12220 }, { "epoch": 25.74736842105263, "grad_norm": 0.2932353913784027, "learning_rate": 3.589967120499353e-05, "loss": 0.01, "step": 12230 }, { "epoch": 25.768421052631577, "grad_norm": 0.22336548566818237, "learning_rate": 3.5820372537770075e-05, "loss": 0.009, "step": 12240 }, { "epoch": 25.789473684210527, "grad_norm": 0.3290993571281433, "learning_rate": 3.5741112637106655e-05, "loss": 0.0093, "step": 12250 }, { "epoch": 25.810526315789474, "grad_norm": 0.3294926583766937, "learning_rate": 3.5661891719696804e-05, "loss": 0.0091, "step": 12260 }, { "epoch": 25.83157894736842, "grad_norm": 0.2741188406944275, "learning_rate": 3.5582710002127504e-05, "loss": 0.0113, "step": 12270 }, { "epoch": 25.852631578947367, "grad_norm": 0.3542085289955139, "learning_rate": 3.550356770087853e-05, "loss": 0.0105, "step": 12280 }, { "epoch": 25.873684210526317, "grad_norm": 0.36358657479286194, "learning_rate": 3.5424465032321914e-05, "loss": 0.0115, "step": 12290 }, { "epoch": 25.894736842105264, "grad_norm": 0.33219897747039795, "learning_rate": 3.5345402212721335e-05, "loss": 0.0099, "step": 12300 }, { "epoch": 25.91578947368421, "grad_norm": 0.26769760251045227, "learning_rate": 3.526637945823152e-05, "loss": 0.0082, "step": 12310 }, { "epoch": 25.936842105263157, "grad_norm": 0.23652595281600952, "learning_rate": 3.518739698489767e-05, "loss": 0.0106, "step": 12320 }, { "epoch": 25.957894736842107, "grad_norm": 0.36284926533699036, "learning_rate": 3.510845500865485e-05, "loss": 0.0085, "step": 12330 }, { "epoch": 25.978947368421053, "grad_norm": 0.25974729657173157, "learning_rate": 3.502955374532739e-05, "loss": 0.0085, "step": 12340 }, { "epoch": 26.0, "grad_norm": 0.26233789324760437, "learning_rate": 3.495069341062836e-05, "loss": 0.0094, "step": 12350 }, { "epoch": 26.021052631578947, "grad_norm": 0.24767759442329407, "learning_rate": 3.4871874220158896e-05, "loss": 0.0095, "step": 12360 }, { "epoch": 26.042105263157893, "grad_norm": 0.2951067388057709, "learning_rate": 3.479309638940762e-05, "loss": 0.0096, "step": 12370 }, { "epoch": 26.063157894736843, "grad_norm": 0.21949627995491028, "learning_rate": 3.4714360133750146e-05, "loss": 0.01, "step": 12380 }, { "epoch": 26.08421052631579, "grad_norm": 0.3150234520435333, "learning_rate": 3.463566566844839e-05, "loss": 0.0095, "step": 12390 }, { "epoch": 26.105263157894736, "grad_norm": 0.22582416236400604, "learning_rate": 3.4557013208650016e-05, "loss": 0.0107, "step": 12400 }, { "epoch": 26.126315789473683, "grad_norm": 0.31235894560813904, "learning_rate": 3.4478402969387857e-05, "loss": 0.0095, "step": 12410 }, { "epoch": 26.147368421052633, "grad_norm": 0.25391313433647156, "learning_rate": 3.4399835165579266e-05, "loss": 0.0091, "step": 12420 }, { "epoch": 26.16842105263158, "grad_norm": 0.31153202056884766, "learning_rate": 3.4321310012025645e-05, "loss": 0.0097, "step": 12430 }, { "epoch": 26.189473684210526, "grad_norm": 0.23246295750141144, "learning_rate": 3.424282772341176e-05, "loss": 0.008, "step": 12440 }, { "epoch": 26.210526315789473, "grad_norm": 0.7248786091804504, "learning_rate": 3.416438851430519e-05, "loss": 0.0095, "step": 12450 }, { "epoch": 26.231578947368423, "grad_norm": 0.2315547615289688, "learning_rate": 3.408599259915577e-05, "loss": 0.0094, "step": 12460 }, { "epoch": 26.25263157894737, "grad_norm": 0.19536492228507996, "learning_rate": 3.400764019229487e-05, "loss": 0.0088, "step": 12470 }, { "epoch": 26.273684210526316, "grad_norm": 0.30122825503349304, "learning_rate": 3.3929331507935035e-05, "loss": 0.0091, "step": 12480 }, { "epoch": 26.294736842105262, "grad_norm": 0.18936626613140106, "learning_rate": 3.3851066760169196e-05, "loss": 0.0079, "step": 12490 }, { "epoch": 26.31578947368421, "grad_norm": 0.25034573674201965, "learning_rate": 3.377284616297021e-05, "loss": 0.0087, "step": 12500 }, { "epoch": 26.33684210526316, "grad_norm": 0.18255548179149628, "learning_rate": 3.3694669930190166e-05, "loss": 0.01, "step": 12510 }, { "epoch": 26.357894736842105, "grad_norm": 0.21346361935138702, "learning_rate": 3.36165382755599e-05, "loss": 0.0069, "step": 12520 }, { "epoch": 26.378947368421052, "grad_norm": 0.28867852687835693, "learning_rate": 3.35384514126884e-05, "loss": 0.0089, "step": 12530 }, { "epoch": 26.4, "grad_norm": 0.1456679403781891, "learning_rate": 3.3460409555062154e-05, "loss": 0.009, "step": 12540 }, { "epoch": 26.42105263157895, "grad_norm": 0.20177412033081055, "learning_rate": 3.3382412916044645e-05, "loss": 0.0088, "step": 12550 }, { "epoch": 26.442105263157895, "grad_norm": 0.2798839211463928, "learning_rate": 3.330446170887566e-05, "loss": 0.0103, "step": 12560 }, { "epoch": 26.46315789473684, "grad_norm": 0.19058318436145782, "learning_rate": 3.3226556146670834e-05, "loss": 0.0111, "step": 12570 }, { "epoch": 26.48421052631579, "grad_norm": 0.28373706340789795, "learning_rate": 3.314869644242102e-05, "loss": 0.0106, "step": 12580 }, { "epoch": 26.50526315789474, "grad_norm": 0.22390972077846527, "learning_rate": 3.3070882808991674e-05, "loss": 0.0075, "step": 12590 }, { "epoch": 26.526315789473685, "grad_norm": 0.24949587881565094, "learning_rate": 3.2993115459122305e-05, "loss": 0.0109, "step": 12600 }, { "epoch": 26.54736842105263, "grad_norm": 0.30132320523262024, "learning_rate": 3.2915394605425835e-05, "loss": 0.0098, "step": 12610 }, { "epoch": 26.568421052631578, "grad_norm": 0.35436466336250305, "learning_rate": 3.283772046038816e-05, "loss": 0.0079, "step": 12620 }, { "epoch": 26.589473684210525, "grad_norm": 0.21300280094146729, "learning_rate": 3.276009323636739e-05, "loss": 0.0086, "step": 12630 }, { "epoch": 26.610526315789475, "grad_norm": 0.20357239246368408, "learning_rate": 3.268251314559344e-05, "loss": 0.0112, "step": 12640 }, { "epoch": 26.63157894736842, "grad_norm": 0.27617037296295166, "learning_rate": 3.2604980400167254e-05, "loss": 0.0109, "step": 12650 }, { "epoch": 26.652631578947368, "grad_norm": 0.17612224817276, "learning_rate": 3.252749521206042e-05, "loss": 0.0081, "step": 12660 }, { "epoch": 26.673684210526314, "grad_norm": 0.2737526297569275, "learning_rate": 3.2450057793114494e-05, "loss": 0.0109, "step": 12670 }, { "epoch": 26.694736842105264, "grad_norm": 0.1760866492986679, "learning_rate": 3.2372668355040435e-05, "loss": 0.0083, "step": 12680 }, { "epoch": 26.71578947368421, "grad_norm": 0.27277809381484985, "learning_rate": 3.2295327109418005e-05, "loss": 0.0073, "step": 12690 }, { "epoch": 26.736842105263158, "grad_norm": 0.13685613870620728, "learning_rate": 3.221803426769518e-05, "loss": 0.008, "step": 12700 }, { "epoch": 26.757894736842104, "grad_norm": 0.1561344563961029, "learning_rate": 3.214079004118768e-05, "loss": 0.0086, "step": 12710 }, { "epoch": 26.778947368421054, "grad_norm": 0.23012271523475647, "learning_rate": 3.2063594641078234e-05, "loss": 0.0078, "step": 12720 }, { "epoch": 26.8, "grad_norm": 0.2333366572856903, "learning_rate": 3.198644827841616e-05, "loss": 0.0111, "step": 12730 }, { "epoch": 26.821052631578947, "grad_norm": 0.21872606873512268, "learning_rate": 3.1909351164116654e-05, "loss": 0.0104, "step": 12740 }, { "epoch": 26.842105263157894, "grad_norm": 0.22063520550727844, "learning_rate": 3.183230350896026e-05, "loss": 0.0107, "step": 12750 }, { "epoch": 26.863157894736844, "grad_norm": 0.20807285606861115, "learning_rate": 3.1755305523592337e-05, "loss": 0.0088, "step": 12760 }, { "epoch": 26.88421052631579, "grad_norm": 0.21177072823047638, "learning_rate": 3.167835741852245e-05, "loss": 0.012, "step": 12770 }, { "epoch": 26.905263157894737, "grad_norm": 0.38845720887184143, "learning_rate": 3.160145940412378e-05, "loss": 0.0106, "step": 12780 }, { "epoch": 26.926315789473684, "grad_norm": 0.16690467298030853, "learning_rate": 3.1524611690632545e-05, "loss": 0.0098, "step": 12790 }, { "epoch": 26.94736842105263, "grad_norm": 0.16756059229373932, "learning_rate": 3.144781448814746e-05, "loss": 0.0089, "step": 12800 }, { "epoch": 26.96842105263158, "grad_norm": 0.2189628779888153, "learning_rate": 3.1371068006629145e-05, "loss": 0.0095, "step": 12810 }, { "epoch": 26.989473684210527, "grad_norm": 0.21110323071479797, "learning_rate": 3.129437245589956e-05, "loss": 0.0074, "step": 12820 }, { "epoch": 27.010526315789473, "grad_norm": 0.1583942025899887, "learning_rate": 3.121772804564143e-05, "loss": 0.0065, "step": 12830 }, { "epoch": 27.03157894736842, "grad_norm": 0.29040881991386414, "learning_rate": 3.11411349853976e-05, "loss": 0.0086, "step": 12840 }, { "epoch": 27.05263157894737, "grad_norm": 0.2939377427101135, "learning_rate": 3.10645934845706e-05, "loss": 0.0114, "step": 12850 }, { "epoch": 27.073684210526316, "grad_norm": 0.21462659537792206, "learning_rate": 3.098810375242196e-05, "loss": 0.0097, "step": 12860 }, { "epoch": 27.094736842105263, "grad_norm": 0.3005073666572571, "learning_rate": 3.0911665998071704e-05, "loss": 0.0122, "step": 12870 }, { "epoch": 27.11578947368421, "grad_norm": 0.18820872902870178, "learning_rate": 3.083528043049774e-05, "loss": 0.009, "step": 12880 }, { "epoch": 27.13684210526316, "grad_norm": 0.1723090559244156, "learning_rate": 3.0758947258535255e-05, "loss": 0.0066, "step": 12890 }, { "epoch": 27.157894736842106, "grad_norm": 0.22065721452236176, "learning_rate": 3.068266669087625e-05, "loss": 0.0082, "step": 12900 }, { "epoch": 27.178947368421053, "grad_norm": 0.3580736815929413, "learning_rate": 3.060643893606887e-05, "loss": 0.0082, "step": 12910 }, { "epoch": 27.2, "grad_norm": 0.34314003586769104, "learning_rate": 3.053026420251693e-05, "loss": 0.0091, "step": 12920 }, { "epoch": 27.221052631578946, "grad_norm": 0.25226277112960815, "learning_rate": 3.0454142698479183e-05, "loss": 0.0077, "step": 12930 }, { "epoch": 27.242105263157896, "grad_norm": 0.3002634048461914, "learning_rate": 3.0378074632068954e-05, "loss": 0.01, "step": 12940 }, { "epoch": 27.263157894736842, "grad_norm": 0.22496157884597778, "learning_rate": 3.0302060211253408e-05, "loss": 0.0092, "step": 12950 }, { "epoch": 27.28421052631579, "grad_norm": 0.20467554032802582, "learning_rate": 3.0226099643853073e-05, "loss": 0.0091, "step": 12960 }, { "epoch": 27.305263157894736, "grad_norm": 0.3891584575176239, "learning_rate": 3.0150193137541283e-05, "loss": 0.0083, "step": 12970 }, { "epoch": 27.326315789473686, "grad_norm": 0.20552760362625122, "learning_rate": 3.0074340899843467e-05, "loss": 0.0086, "step": 12980 }, { "epoch": 27.347368421052632, "grad_norm": 0.2844048738479614, "learning_rate": 2.999854313813677e-05, "loss": 0.0099, "step": 12990 }, { "epoch": 27.36842105263158, "grad_norm": 0.2878531217575073, "learning_rate": 2.9922800059649382e-05, "loss": 0.0117, "step": 13000 }, { "epoch": 27.389473684210525, "grad_norm": 0.23985274136066437, "learning_rate": 2.9847111871459976e-05, "loss": 0.0098, "step": 13010 }, { "epoch": 27.410526315789475, "grad_norm": 0.1885942816734314, "learning_rate": 2.977147878049721e-05, "loss": 0.0087, "step": 13020 }, { "epoch": 27.431578947368422, "grad_norm": 0.18704041838645935, "learning_rate": 2.9695900993539006e-05, "loss": 0.0083, "step": 13030 }, { "epoch": 27.45263157894737, "grad_norm": 0.1678140014410019, "learning_rate": 2.9620378717212183e-05, "loss": 0.0074, "step": 13040 }, { "epoch": 27.473684210526315, "grad_norm": 0.22837752103805542, "learning_rate": 2.9544912157991745e-05, "loss": 0.009, "step": 13050 }, { "epoch": 27.49473684210526, "grad_norm": 0.1571223884820938, "learning_rate": 2.9469501522200405e-05, "loss": 0.0083, "step": 13060 }, { "epoch": 27.51578947368421, "grad_norm": 0.24090372025966644, "learning_rate": 2.9394147016007946e-05, "loss": 0.0128, "step": 13070 }, { "epoch": 27.53684210526316, "grad_norm": 0.2780013680458069, "learning_rate": 2.9318848845430702e-05, "loss": 0.0112, "step": 13080 }, { "epoch": 27.557894736842105, "grad_norm": 0.2570752203464508, "learning_rate": 2.9243607216331013e-05, "loss": 0.0107, "step": 13090 }, { "epoch": 27.57894736842105, "grad_norm": 0.21750429272651672, "learning_rate": 2.916842233441661e-05, "loss": 0.0129, "step": 13100 }, { "epoch": 27.6, "grad_norm": 0.34446167945861816, "learning_rate": 2.90932944052401e-05, "loss": 0.0092, "step": 13110 }, { "epoch": 27.621052631578948, "grad_norm": 0.18531137704849243, "learning_rate": 2.9018223634198354e-05, "loss": 0.0118, "step": 13120 }, { "epoch": 27.642105263157895, "grad_norm": 0.2086588740348816, "learning_rate": 2.8943210226532025e-05, "loss": 0.0115, "step": 13130 }, { "epoch": 27.66315789473684, "grad_norm": 0.3561864495277405, "learning_rate": 2.8868254387324857e-05, "loss": 0.009, "step": 13140 }, { "epoch": 27.68421052631579, "grad_norm": 0.1328297108411789, "learning_rate": 2.8793356321503306e-05, "loss": 0.0091, "step": 13150 }, { "epoch": 27.705263157894738, "grad_norm": 0.3327886164188385, "learning_rate": 2.87185162338358e-05, "loss": 0.007, "step": 13160 }, { "epoch": 27.726315789473684, "grad_norm": 0.1689920276403427, "learning_rate": 2.8643734328932253e-05, "loss": 0.008, "step": 13170 }, { "epoch": 27.74736842105263, "grad_norm": 0.21940167248249054, "learning_rate": 2.856901081124359e-05, "loss": 0.0089, "step": 13180 }, { "epoch": 27.768421052631577, "grad_norm": 0.2455526441335678, "learning_rate": 2.8494345885061002e-05, "loss": 0.0073, "step": 13190 }, { "epoch": 27.789473684210527, "grad_norm": 0.34249722957611084, "learning_rate": 2.8419739754515616e-05, "loss": 0.0083, "step": 13200 }, { "epoch": 27.810526315789474, "grad_norm": 0.23199450969696045, "learning_rate": 2.8345192623577666e-05, "loss": 0.0093, "step": 13210 }, { "epoch": 27.83157894736842, "grad_norm": 0.3347368538379669, "learning_rate": 2.8270704696056193e-05, "loss": 0.008, "step": 13220 }, { "epoch": 27.852631578947367, "grad_norm": 0.24476131796836853, "learning_rate": 2.8196276175598367e-05, "loss": 0.0074, "step": 13230 }, { "epoch": 27.873684210526317, "grad_norm": 0.21885205805301666, "learning_rate": 2.8121907265688884e-05, "loss": 0.0102, "step": 13240 }, { "epoch": 27.894736842105264, "grad_norm": 0.28656816482543945, "learning_rate": 2.804759816964957e-05, "loss": 0.0073, "step": 13250 }, { "epoch": 27.91578947368421, "grad_norm": 0.2536468803882599, "learning_rate": 2.797334909063857e-05, "loss": 0.0109, "step": 13260 }, { "epoch": 27.936842105263157, "grad_norm": 0.20811447501182556, "learning_rate": 2.7899160231650056e-05, "loss": 0.0092, "step": 13270 }, { "epoch": 27.957894736842107, "grad_norm": 0.3432919383049011, "learning_rate": 2.7825031795513585e-05, "loss": 0.0082, "step": 13280 }, { "epoch": 27.978947368421053, "grad_norm": 0.19303379952907562, "learning_rate": 2.775096398489341e-05, "loss": 0.0086, "step": 13290 }, { "epoch": 28.0, "grad_norm": 0.27213549613952637, "learning_rate": 2.7676957002288163e-05, "loss": 0.007, "step": 13300 }, { "epoch": 28.021052631578947, "grad_norm": 0.25984033942222595, "learning_rate": 2.760301105003003e-05, "loss": 0.0069, "step": 13310 }, { "epoch": 28.042105263157893, "grad_norm": 0.1738375723361969, "learning_rate": 2.752912633028446e-05, "loss": 0.0081, "step": 13320 }, { "epoch": 28.063157894736843, "grad_norm": 0.3305702209472656, "learning_rate": 2.7455303045049474e-05, "loss": 0.0092, "step": 13330 }, { "epoch": 28.08421052631579, "grad_norm": 0.2580360770225525, "learning_rate": 2.7381541396155098e-05, "loss": 0.0071, "step": 13340 }, { "epoch": 28.105263157894736, "grad_norm": 0.1943882256746292, "learning_rate": 2.730784158526286e-05, "loss": 0.0078, "step": 13350 }, { "epoch": 28.126315789473683, "grad_norm": 0.18848343193531036, "learning_rate": 2.723420381386521e-05, "loss": 0.009, "step": 13360 }, { "epoch": 28.147368421052633, "grad_norm": 0.23332719504833221, "learning_rate": 2.7160628283285018e-05, "loss": 0.0092, "step": 13370 }, { "epoch": 28.16842105263158, "grad_norm": 0.30231228470802307, "learning_rate": 2.7087115194675007e-05, "loss": 0.0079, "step": 13380 }, { "epoch": 28.189473684210526, "grad_norm": 0.26833462715148926, "learning_rate": 2.701366474901712e-05, "loss": 0.0085, "step": 13390 }, { "epoch": 28.210526315789473, "grad_norm": 0.31022223830223083, "learning_rate": 2.6940277147122085e-05, "loss": 0.0079, "step": 13400 }, { "epoch": 28.231578947368423, "grad_norm": 0.14095735549926758, "learning_rate": 2.686695258962878e-05, "loss": 0.0071, "step": 13410 }, { "epoch": 28.25263157894737, "grad_norm": 0.15236598253250122, "learning_rate": 2.679369127700375e-05, "loss": 0.0075, "step": 13420 }, { "epoch": 28.273684210526316, "grad_norm": 0.22839903831481934, "learning_rate": 2.672049340954067e-05, "loss": 0.0093, "step": 13430 }, { "epoch": 28.294736842105262, "grad_norm": 0.1991535872220993, "learning_rate": 2.6647359187359676e-05, "loss": 0.0108, "step": 13440 }, { "epoch": 28.31578947368421, "grad_norm": 0.23363997042179108, "learning_rate": 2.6574288810406946e-05, "loss": 0.0083, "step": 13450 }, { "epoch": 28.33684210526316, "grad_norm": 0.21993620693683624, "learning_rate": 2.6501282478454083e-05, "loss": 0.0075, "step": 13460 }, { "epoch": 28.357894736842105, "grad_norm": 0.14352431893348694, "learning_rate": 2.6428340391097618e-05, "loss": 0.0081, "step": 13470 }, { "epoch": 28.378947368421052, "grad_norm": 0.1982034295797348, "learning_rate": 2.6355462747758485e-05, "loss": 0.0064, "step": 13480 }, { "epoch": 28.4, "grad_norm": 0.29882416129112244, "learning_rate": 2.6282649747681304e-05, "loss": 0.0086, "step": 13490 }, { "epoch": 28.42105263157895, "grad_norm": 0.16924378275871277, "learning_rate": 2.620990158993406e-05, "loss": 0.009, "step": 13500 }, { "epoch": 28.442105263157895, "grad_norm": 0.2445613145828247, "learning_rate": 2.6137218473407477e-05, "loss": 0.0094, "step": 13510 }, { "epoch": 28.46315789473684, "grad_norm": 0.20527969300746918, "learning_rate": 2.606460059681436e-05, "loss": 0.0078, "step": 13520 }, { "epoch": 28.48421052631579, "grad_norm": 0.43657386302948, "learning_rate": 2.599204815868928e-05, "loss": 0.0077, "step": 13530 }, { "epoch": 28.50526315789474, "grad_norm": 0.30783429741859436, "learning_rate": 2.5919561357387756e-05, "loss": 0.0085, "step": 13540 }, { "epoch": 28.526315789473685, "grad_norm": 0.24516229331493378, "learning_rate": 2.5847140391085972e-05, "loss": 0.0099, "step": 13550 }, { "epoch": 28.54736842105263, "grad_norm": 0.1881500482559204, "learning_rate": 2.5774785457780103e-05, "loss": 0.0076, "step": 13560 }, { "epoch": 28.568421052631578, "grad_norm": 0.2716509997844696, "learning_rate": 2.5702496755285753e-05, "loss": 0.0097, "step": 13570 }, { "epoch": 28.589473684210525, "grad_norm": 0.1896287202835083, "learning_rate": 2.5630274481237483e-05, "loss": 0.0077, "step": 13580 }, { "epoch": 28.610526315789475, "grad_norm": 0.42379677295684814, "learning_rate": 2.5558118833088197e-05, "loss": 0.0095, "step": 13590 }, { "epoch": 28.63157894736842, "grad_norm": 0.1949995905160904, "learning_rate": 2.548603000810872e-05, "loss": 0.0083, "step": 13600 }, { "epoch": 28.652631578947368, "grad_norm": 0.3533550798892975, "learning_rate": 2.5414008203387152e-05, "loss": 0.0105, "step": 13610 }, { "epoch": 28.673684210526314, "grad_norm": 0.1749424934387207, "learning_rate": 2.534205361582834e-05, "loss": 0.0132, "step": 13620 }, { "epoch": 28.694736842105264, "grad_norm": 0.15182329714298248, "learning_rate": 2.527016644215338e-05, "loss": 0.0086, "step": 13630 }, { "epoch": 28.71578947368421, "grad_norm": 0.22687606513500214, "learning_rate": 2.519834687889905e-05, "loss": 0.0078, "step": 13640 }, { "epoch": 28.736842105263158, "grad_norm": 0.2531571388244629, "learning_rate": 2.5126595122417295e-05, "loss": 0.0077, "step": 13650 }, { "epoch": 28.757894736842104, "grad_norm": 0.20138174295425415, "learning_rate": 2.5054911368874713e-05, "loss": 0.0083, "step": 13660 }, { "epoch": 28.778947368421054, "grad_norm": 0.278076171875, "learning_rate": 2.4983295814251916e-05, "loss": 0.0079, "step": 13670 }, { "epoch": 28.8, "grad_norm": 0.32636088132858276, "learning_rate": 2.4911748654343105e-05, "loss": 0.0111, "step": 13680 }, { "epoch": 28.821052631578947, "grad_norm": 0.2022264003753662, "learning_rate": 2.4840270084755463e-05, "loss": 0.01, "step": 13690 }, { "epoch": 28.842105263157894, "grad_norm": 0.25392743945121765, "learning_rate": 2.4768860300908685e-05, "loss": 0.0085, "step": 13700 }, { "epoch": 28.863157894736844, "grad_norm": 0.20970629155635834, "learning_rate": 2.469751949803443e-05, "loss": 0.0091, "step": 13710 }, { "epoch": 28.88421052631579, "grad_norm": 0.3875325322151184, "learning_rate": 2.4626247871175666e-05, "loss": 0.008, "step": 13720 }, { "epoch": 28.905263157894737, "grad_norm": 0.3806176781654358, "learning_rate": 2.4555045615186346e-05, "loss": 0.0084, "step": 13730 }, { "epoch": 28.926315789473684, "grad_norm": 0.13991738855838776, "learning_rate": 2.4483912924730677e-05, "loss": 0.0076, "step": 13740 }, { "epoch": 28.94736842105263, "grad_norm": 0.20290987193584442, "learning_rate": 2.4412849994282742e-05, "loss": 0.0082, "step": 13750 }, { "epoch": 28.96842105263158, "grad_norm": 0.21654534339904785, "learning_rate": 2.434185701812592e-05, "loss": 0.0065, "step": 13760 }, { "epoch": 28.989473684210527, "grad_norm": 0.20341575145721436, "learning_rate": 2.4270934190352218e-05, "loss": 0.0093, "step": 13770 }, { "epoch": 29.010526315789473, "grad_norm": 0.2032337486743927, "learning_rate": 2.4200081704861998e-05, "loss": 0.0074, "step": 13780 }, { "epoch": 29.03157894736842, "grad_norm": 0.20646920800209045, "learning_rate": 2.412929975536321e-05, "loss": 0.0079, "step": 13790 }, { "epoch": 29.05263157894737, "grad_norm": 0.2738656997680664, "learning_rate": 2.4058588535371017e-05, "loss": 0.0114, "step": 13800 }, { "epoch": 29.073684210526316, "grad_norm": 0.23966056108474731, "learning_rate": 2.3987948238207243e-05, "loss": 0.0085, "step": 13810 }, { "epoch": 29.094736842105263, "grad_norm": 0.18627813458442688, "learning_rate": 2.3917379056999678e-05, "loss": 0.0088, "step": 13820 }, { "epoch": 29.11578947368421, "grad_norm": 0.20841464400291443, "learning_rate": 2.3846881184681824e-05, "loss": 0.0088, "step": 13830 }, { "epoch": 29.13684210526316, "grad_norm": 0.19140899181365967, "learning_rate": 2.377645481399214e-05, "loss": 0.0072, "step": 13840 }, { "epoch": 29.157894736842106, "grad_norm": 0.28309258818626404, "learning_rate": 2.3706100137473667e-05, "loss": 0.0083, "step": 13850 }, { "epoch": 29.178947368421053, "grad_norm": 0.26309069991111755, "learning_rate": 2.3635817347473394e-05, "loss": 0.0082, "step": 13860 }, { "epoch": 29.2, "grad_norm": 0.20435458421707153, "learning_rate": 2.3565606636141757e-05, "loss": 0.0069, "step": 13870 }, { "epoch": 29.221052631578946, "grad_norm": 0.1827511191368103, "learning_rate": 2.3495468195432203e-05, "loss": 0.0079, "step": 13880 }, { "epoch": 29.242105263157896, "grad_norm": 0.1750650405883789, "learning_rate": 2.3425402217100507e-05, "loss": 0.006, "step": 13890 }, { "epoch": 29.263157894736842, "grad_norm": 0.22454121708869934, "learning_rate": 2.3355408892704424e-05, "loss": 0.0068, "step": 13900 }, { "epoch": 29.28421052631579, "grad_norm": 0.1806291937828064, "learning_rate": 2.3285488413603003e-05, "loss": 0.0101, "step": 13910 }, { "epoch": 29.305263157894736, "grad_norm": 0.15026022493839264, "learning_rate": 2.321564097095615e-05, "loss": 0.009, "step": 13920 }, { "epoch": 29.326315789473686, "grad_norm": 0.3588649332523346, "learning_rate": 2.3145866755724142e-05, "loss": 0.0082, "step": 13930 }, { "epoch": 29.347368421052632, "grad_norm": 0.2120971828699112, "learning_rate": 2.307616595866699e-05, "loss": 0.0067, "step": 13940 }, { "epoch": 29.36842105263158, "grad_norm": 0.2924186885356903, "learning_rate": 2.3006538770344032e-05, "loss": 0.0089, "step": 13950 }, { "epoch": 29.389473684210525, "grad_norm": 0.26052024960517883, "learning_rate": 2.293698538111334e-05, "loss": 0.009, "step": 13960 }, { "epoch": 29.410526315789475, "grad_norm": 0.3897743225097656, "learning_rate": 2.28675059811312e-05, "loss": 0.0083, "step": 13970 }, { "epoch": 29.431578947368422, "grad_norm": 0.16744349896907806, "learning_rate": 2.279810076035167e-05, "loss": 0.0072, "step": 13980 }, { "epoch": 29.45263157894737, "grad_norm": 0.3183251619338989, "learning_rate": 2.272876990852596e-05, "loss": 0.01, "step": 13990 }, { "epoch": 29.473684210526315, "grad_norm": 0.2224886268377304, "learning_rate": 2.265951361520195e-05, "loss": 0.0072, "step": 14000 }, { "epoch": 29.49473684210526, "grad_norm": 0.27266162633895874, "learning_rate": 2.2590332069723748e-05, "loss": 0.0085, "step": 14010 }, { "epoch": 29.51578947368421, "grad_norm": 0.17823873460292816, "learning_rate": 2.2521225461231004e-05, "loss": 0.0082, "step": 14020 }, { "epoch": 29.53684210526316, "grad_norm": 0.1824556291103363, "learning_rate": 2.2452193978658597e-05, "loss": 0.0088, "step": 14030 }, { "epoch": 29.557894736842105, "grad_norm": 0.2138897329568863, "learning_rate": 2.238323781073594e-05, "loss": 0.0084, "step": 14040 }, { "epoch": 29.57894736842105, "grad_norm": 0.43306055665016174, "learning_rate": 2.2314357145986552e-05, "loss": 0.0083, "step": 14050 }, { "epoch": 29.6, "grad_norm": 0.1324135810136795, "learning_rate": 2.224555217272757e-05, "loss": 0.0095, "step": 14060 }, { "epoch": 29.621052631578948, "grad_norm": 0.22888098657131195, "learning_rate": 2.2176823079069127e-05, "loss": 0.008, "step": 14070 }, { "epoch": 29.642105263157895, "grad_norm": 0.16154105961322784, "learning_rate": 2.210817005291398e-05, "loss": 0.0074, "step": 14080 }, { "epoch": 29.66315789473684, "grad_norm": 0.28925132751464844, "learning_rate": 2.203959328195686e-05, "loss": 0.0094, "step": 14090 }, { "epoch": 29.68421052631579, "grad_norm": 0.1975063532590866, "learning_rate": 2.1971092953684026e-05, "loss": 0.0069, "step": 14100 }, { "epoch": 29.705263157894738, "grad_norm": 0.2896670401096344, "learning_rate": 2.1902669255372788e-05, "loss": 0.0085, "step": 14110 }, { "epoch": 29.726315789473684, "grad_norm": 0.13104401528835297, "learning_rate": 2.1834322374090897e-05, "loss": 0.0087, "step": 14120 }, { "epoch": 29.74736842105263, "grad_norm": 0.40553852915763855, "learning_rate": 2.1766052496696153e-05, "loss": 0.0106, "step": 14130 }, { "epoch": 29.768421052631577, "grad_norm": 0.2427341789007187, "learning_rate": 2.169785980983577e-05, "loss": 0.0105, "step": 14140 }, { "epoch": 29.789473684210527, "grad_norm": 0.2214793860912323, "learning_rate": 2.162974449994593e-05, "loss": 0.0079, "step": 14150 }, { "epoch": 29.810526315789474, "grad_norm": 0.16957376897335052, "learning_rate": 2.1561706753251337e-05, "loss": 0.013, "step": 14160 }, { "epoch": 29.83157894736842, "grad_norm": 0.2877258360385895, "learning_rate": 2.1493746755764544e-05, "loss": 0.0067, "step": 14170 }, { "epoch": 29.852631578947367, "grad_norm": 0.24668973684310913, "learning_rate": 2.1425864693285635e-05, "loss": 0.0088, "step": 14180 }, { "epoch": 29.873684210526317, "grad_norm": 0.23060107231140137, "learning_rate": 2.1358060751401547e-05, "loss": 0.0094, "step": 14190 }, { "epoch": 29.894736842105264, "grad_norm": 0.12060017883777618, "learning_rate": 2.129033511548566e-05, "loss": 0.0069, "step": 14200 }, { "epoch": 29.91578947368421, "grad_norm": 0.21695490181446075, "learning_rate": 2.1222687970697315e-05, "loss": 0.008, "step": 14210 }, { "epoch": 29.936842105263157, "grad_norm": 0.26187774538993835, "learning_rate": 2.1155119501981173e-05, "loss": 0.0072, "step": 14220 }, { "epoch": 29.957894736842107, "grad_norm": 0.21658729016780853, "learning_rate": 2.1087629894066895e-05, "loss": 0.0062, "step": 14230 }, { "epoch": 29.978947368421053, "grad_norm": 0.24550166726112366, "learning_rate": 2.1020219331468473e-05, "loss": 0.0058, "step": 14240 }, { "epoch": 30.0, "grad_norm": 0.16428810358047485, "learning_rate": 2.095288799848379e-05, "loss": 0.0078, "step": 14250 }, { "epoch": 30.021052631578947, "grad_norm": 0.2207103818655014, "learning_rate": 2.088563607919417e-05, "loss": 0.0075, "step": 14260 }, { "epoch": 30.042105263157893, "grad_norm": 0.19754497706890106, "learning_rate": 2.0818463757463786e-05, "loss": 0.0073, "step": 14270 }, { "epoch": 30.063157894736843, "grad_norm": 0.18510282039642334, "learning_rate": 2.0751371216939175e-05, "loss": 0.0065, "step": 14280 }, { "epoch": 30.08421052631579, "grad_norm": 0.3847796618938446, "learning_rate": 2.068435864104882e-05, "loss": 0.0083, "step": 14290 }, { "epoch": 30.105263157894736, "grad_norm": 0.7218855023384094, "learning_rate": 2.0617426213002506e-05, "loss": 0.0082, "step": 14300 }, { "epoch": 30.126315789473683, "grad_norm": 0.2656722962856293, "learning_rate": 2.055057411579097e-05, "loss": 0.0074, "step": 14310 }, { "epoch": 30.147368421052633, "grad_norm": 0.25196540355682373, "learning_rate": 2.0483802532185286e-05, "loss": 0.0076, "step": 14320 }, { "epoch": 30.16842105263158, "grad_norm": 0.2755727767944336, "learning_rate": 2.041711164473638e-05, "loss": 0.0078, "step": 14330 }, { "epoch": 30.189473684210526, "grad_norm": 0.18422053754329681, "learning_rate": 2.0350501635774637e-05, "loss": 0.0067, "step": 14340 }, { "epoch": 30.210526315789473, "grad_norm": 0.21088512241840363, "learning_rate": 2.0283972687409247e-05, "loss": 0.0099, "step": 14350 }, { "epoch": 30.231578947368423, "grad_norm": 0.2056790292263031, "learning_rate": 2.021752498152784e-05, "loss": 0.0082, "step": 14360 }, { "epoch": 30.25263157894737, "grad_norm": 0.1726762056350708, "learning_rate": 2.015115869979589e-05, "loss": 0.0072, "step": 14370 }, { "epoch": 30.273684210526316, "grad_norm": 0.15846243500709534, "learning_rate": 2.0084874023656265e-05, "loss": 0.0099, "step": 14380 }, { "epoch": 30.294736842105262, "grad_norm": 0.2252965271472931, "learning_rate": 2.001867113432877e-05, "loss": 0.0079, "step": 14390 }, { "epoch": 30.31578947368421, "grad_norm": 0.28469809889793396, "learning_rate": 1.995255021280954e-05, "loss": 0.0096, "step": 14400 }, { "epoch": 30.33684210526316, "grad_norm": 0.3162561058998108, "learning_rate": 1.9886511439870688e-05, "loss": 0.0058, "step": 14410 }, { "epoch": 30.357894736842105, "grad_norm": 0.1415104866027832, "learning_rate": 1.9820554996059675e-05, "loss": 0.0078, "step": 14420 }, { "epoch": 30.378947368421052, "grad_norm": 0.13662178814411163, "learning_rate": 1.9754681061698893e-05, "loss": 0.0075, "step": 14430 }, { "epoch": 30.4, "grad_norm": 0.22984516620635986, "learning_rate": 1.9688889816885185e-05, "loss": 0.0105, "step": 14440 }, { "epoch": 30.42105263157895, "grad_norm": 0.24426791071891785, "learning_rate": 1.962318144148928e-05, "loss": 0.0068, "step": 14450 }, { "epoch": 30.442105263157895, "grad_norm": 0.14705131947994232, "learning_rate": 1.955755611515539e-05, "loss": 0.0087, "step": 14460 }, { "epoch": 30.46315789473684, "grad_norm": 0.16464659571647644, "learning_rate": 1.9492014017300642e-05, "loss": 0.006, "step": 14470 }, { "epoch": 30.48421052631579, "grad_norm": 0.2616230547428131, "learning_rate": 1.942655532711461e-05, "loss": 0.0096, "step": 14480 }, { "epoch": 30.50526315789474, "grad_norm": 0.41326475143432617, "learning_rate": 1.9361180223558882e-05, "loss": 0.0077, "step": 14490 }, { "epoch": 30.526315789473685, "grad_norm": 0.18032829463481903, "learning_rate": 1.929588888536647e-05, "loss": 0.008, "step": 14500 }, { "epoch": 30.54736842105263, "grad_norm": 0.2000184953212738, "learning_rate": 1.9230681491041425e-05, "loss": 0.0081, "step": 14510 }, { "epoch": 30.568421052631578, "grad_norm": 0.32803457975387573, "learning_rate": 1.9165558218858264e-05, "loss": 0.0093, "step": 14520 }, { "epoch": 30.589473684210525, "grad_norm": 0.1497115194797516, "learning_rate": 1.9100519246861505e-05, "loss": 0.0069, "step": 14530 }, { "epoch": 30.610526315789475, "grad_norm": 0.22089125216007233, "learning_rate": 1.9035564752865248e-05, "loss": 0.008, "step": 14540 }, { "epoch": 30.63157894736842, "grad_norm": 0.2232479453086853, "learning_rate": 1.897069491445258e-05, "loss": 0.0071, "step": 14550 }, { "epoch": 30.652631578947368, "grad_norm": 0.1714835911989212, "learning_rate": 1.890590990897515e-05, "loss": 0.0093, "step": 14560 }, { "epoch": 30.673684210526314, "grad_norm": 0.28117090463638306, "learning_rate": 1.884120991355272e-05, "loss": 0.0081, "step": 14570 }, { "epoch": 30.694736842105264, "grad_norm": 0.1617913693189621, "learning_rate": 1.8776595105072576e-05, "loss": 0.0076, "step": 14580 }, { "epoch": 30.71578947368421, "grad_norm": 0.2576236426830292, "learning_rate": 1.8712065660189166e-05, "loss": 0.0086, "step": 14590 }, { "epoch": 30.736842105263158, "grad_norm": 0.2295185774564743, "learning_rate": 1.8647621755323513e-05, "loss": 0.0064, "step": 14600 }, { "epoch": 30.757894736842104, "grad_norm": 0.1837611049413681, "learning_rate": 1.858326356666278e-05, "loss": 0.0069, "step": 14610 }, { "epoch": 30.778947368421054, "grad_norm": 0.14770616590976715, "learning_rate": 1.851899127015983e-05, "loss": 0.0068, "step": 14620 }, { "epoch": 30.8, "grad_norm": 0.2873314321041107, "learning_rate": 1.8454805041532626e-05, "loss": 0.0063, "step": 14630 }, { "epoch": 30.821052631578947, "grad_norm": 0.2073104828596115, "learning_rate": 1.8390705056263906e-05, "loss": 0.0088, "step": 14640 }, { "epoch": 30.842105263157894, "grad_norm": 0.26032280921936035, "learning_rate": 1.832669148960057e-05, "loss": 0.008, "step": 14650 }, { "epoch": 30.863157894736844, "grad_norm": 0.1355925053358078, "learning_rate": 1.8262764516553233e-05, "loss": 0.0088, "step": 14660 }, { "epoch": 30.88421052631579, "grad_norm": 0.10953214019536972, "learning_rate": 1.8198924311895843e-05, "loss": 0.0064, "step": 14670 }, { "epoch": 30.905263157894737, "grad_norm": 0.43499886989593506, "learning_rate": 1.813517105016505e-05, "loss": 0.007, "step": 14680 }, { "epoch": 30.926315789473684, "grad_norm": 0.11693327128887177, "learning_rate": 1.8071504905659888e-05, "loss": 0.0082, "step": 14690 }, { "epoch": 30.94736842105263, "grad_norm": 0.182923823595047, "learning_rate": 1.800792605244109e-05, "loss": 0.0068, "step": 14700 }, { "epoch": 30.96842105263158, "grad_norm": 0.25011950731277466, "learning_rate": 1.7944434664330844e-05, "loss": 0.0075, "step": 14710 }, { "epoch": 30.989473684210527, "grad_norm": 0.2805038392543793, "learning_rate": 1.7881030914912212e-05, "loss": 0.0064, "step": 14720 }, { "epoch": 31.010526315789473, "grad_norm": 0.2884676456451416, "learning_rate": 1.7817714977528577e-05, "loss": 0.008, "step": 14730 }, { "epoch": 31.03157894736842, "grad_norm": 0.2757003605365753, "learning_rate": 1.7754487025283332e-05, "loss": 0.0085, "step": 14740 }, { "epoch": 31.05263157894737, "grad_norm": 0.29432958364486694, "learning_rate": 1.7691347231039275e-05, "loss": 0.0066, "step": 14750 }, { "epoch": 31.073684210526316, "grad_norm": 0.20237314701080322, "learning_rate": 1.7628295767418164e-05, "loss": 0.0078, "step": 14760 }, { "epoch": 31.094736842105263, "grad_norm": 0.19100914895534515, "learning_rate": 1.7565332806800333e-05, "loss": 0.009, "step": 14770 }, { "epoch": 31.11578947368421, "grad_norm": 0.17795927822589874, "learning_rate": 1.750245852132408e-05, "loss": 0.0076, "step": 14780 }, { "epoch": 31.13684210526316, "grad_norm": 0.2295130342245102, "learning_rate": 1.7439673082885323e-05, "loss": 0.0061, "step": 14790 }, { "epoch": 31.157894736842106, "grad_norm": 0.1838303804397583, "learning_rate": 1.7376976663137047e-05, "loss": 0.008, "step": 14800 }, { "epoch": 31.178947368421053, "grad_norm": 0.2718106508255005, "learning_rate": 1.7314369433488853e-05, "loss": 0.0066, "step": 14810 }, { "epoch": 31.2, "grad_norm": 0.16362103819847107, "learning_rate": 1.7251851565106548e-05, "loss": 0.0071, "step": 14820 }, { "epoch": 31.221052631578946, "grad_norm": 0.17834098637104034, "learning_rate": 1.7189423228911574e-05, "loss": 0.0063, "step": 14830 }, { "epoch": 31.242105263157896, "grad_norm": 0.14919085800647736, "learning_rate": 1.7127084595580606e-05, "loss": 0.0069, "step": 14840 }, { "epoch": 31.263157894736842, "grad_norm": 0.15990301966667175, "learning_rate": 1.706483583554513e-05, "loss": 0.009, "step": 14850 }, { "epoch": 31.28421052631579, "grad_norm": 0.4708520770072937, "learning_rate": 1.700267711899083e-05, "loss": 0.0088, "step": 14860 }, { "epoch": 31.305263157894736, "grad_norm": 0.1293952912092209, "learning_rate": 1.69406086158573e-05, "loss": 0.0065, "step": 14870 }, { "epoch": 31.326315789473686, "grad_norm": 0.1065206229686737, "learning_rate": 1.6878630495837455e-05, "loss": 0.0058, "step": 14880 }, { "epoch": 31.347368421052632, "grad_norm": 0.5305003523826599, "learning_rate": 1.681674292837707e-05, "loss": 0.0062, "step": 14890 }, { "epoch": 31.36842105263158, "grad_norm": 0.2643958032131195, "learning_rate": 1.6754946082674444e-05, "loss": 0.0084, "step": 14900 }, { "epoch": 31.389473684210525, "grad_norm": 0.19130836427211761, "learning_rate": 1.6693240127679748e-05, "loss": 0.0069, "step": 14910 }, { "epoch": 31.410526315789475, "grad_norm": 0.3049776256084442, "learning_rate": 1.663162523209475e-05, "loss": 0.0085, "step": 14920 }, { "epoch": 31.431578947368422, "grad_norm": 0.21087656915187836, "learning_rate": 1.6570101564372193e-05, "loss": 0.0082, "step": 14930 }, { "epoch": 31.45263157894737, "grad_norm": 0.20998533070087433, "learning_rate": 1.650866929271543e-05, "loss": 0.0094, "step": 14940 }, { "epoch": 31.473684210526315, "grad_norm": 0.20435744524002075, "learning_rate": 1.644732858507797e-05, "loss": 0.007, "step": 14950 }, { "epoch": 31.49473684210526, "grad_norm": 0.20268464088439941, "learning_rate": 1.6386079609162943e-05, "loss": 0.0068, "step": 14960 }, { "epoch": 31.51578947368421, "grad_norm": 0.1465202420949936, "learning_rate": 1.6324922532422742e-05, "loss": 0.007, "step": 14970 }, { "epoch": 31.53684210526316, "grad_norm": 0.1669362634420395, "learning_rate": 1.6263857522058434e-05, "loss": 0.0066, "step": 14980 }, { "epoch": 31.557894736842105, "grad_norm": 0.2680119276046753, "learning_rate": 1.6202884745019443e-05, "loss": 0.0063, "step": 14990 }, { "epoch": 31.57894736842105, "grad_norm": 0.2759992480278015, "learning_rate": 1.614200436800304e-05, "loss": 0.0081, "step": 15000 }, { "epoch": 31.6, "grad_norm": 0.13312698900699615, "learning_rate": 1.6081216557453814e-05, "loss": 0.0084, "step": 15010 }, { "epoch": 31.621052631578948, "grad_norm": 0.20523647964000702, "learning_rate": 1.6020521479563367e-05, "loss": 0.0108, "step": 15020 }, { "epoch": 31.642105263157895, "grad_norm": 0.27351808547973633, "learning_rate": 1.5959919300269654e-05, "loss": 0.0085, "step": 15030 }, { "epoch": 31.66315789473684, "grad_norm": 0.1839187890291214, "learning_rate": 1.5899410185256764e-05, "loss": 0.0076, "step": 15040 }, { "epoch": 31.68421052631579, "grad_norm": 0.28489264845848083, "learning_rate": 1.583899429995431e-05, "loss": 0.0067, "step": 15050 }, { "epoch": 31.705263157894738, "grad_norm": 0.21566076576709747, "learning_rate": 1.5778671809536993e-05, "loss": 0.0068, "step": 15060 }, { "epoch": 31.726315789473684, "grad_norm": 0.13545149564743042, "learning_rate": 1.5718442878924246e-05, "loss": 0.0066, "step": 15070 }, { "epoch": 31.74736842105263, "grad_norm": 0.21678216755390167, "learning_rate": 1.5658307672779593e-05, "loss": 0.0103, "step": 15080 }, { "epoch": 31.768421052631577, "grad_norm": 0.20695659518241882, "learning_rate": 1.5598266355510427e-05, "loss": 0.0062, "step": 15090 }, { "epoch": 31.789473684210527, "grad_norm": 0.27032315731048584, "learning_rate": 1.553831909126744e-05, "loss": 0.0092, "step": 15100 }, { "epoch": 31.810526315789474, "grad_norm": 0.19598978757858276, "learning_rate": 1.5478466043944135e-05, "loss": 0.0083, "step": 15110 }, { "epoch": 31.83157894736842, "grad_norm": 0.2019454538822174, "learning_rate": 1.5418707377176468e-05, "loss": 0.0075, "step": 15120 }, { "epoch": 31.852631578947367, "grad_norm": 0.24870608747005463, "learning_rate": 1.535904325434233e-05, "loss": 0.009, "step": 15130 }, { "epoch": 31.873684210526317, "grad_norm": 0.25429317355155945, "learning_rate": 1.529947383856118e-05, "loss": 0.0069, "step": 15140 }, { "epoch": 31.894736842105264, "grad_norm": 0.2184925377368927, "learning_rate": 1.5239999292693524e-05, "loss": 0.007, "step": 15150 }, { "epoch": 31.91578947368421, "grad_norm": 0.16922415792942047, "learning_rate": 1.5180619779340505e-05, "loss": 0.0063, "step": 15160 }, { "epoch": 31.936842105263157, "grad_norm": 0.28436383605003357, "learning_rate": 1.5121335460843428e-05, "loss": 0.0065, "step": 15170 }, { "epoch": 31.957894736842107, "grad_norm": 0.2527836859226227, "learning_rate": 1.5062146499283347e-05, "loss": 0.0105, "step": 15180 }, { "epoch": 31.978947368421053, "grad_norm": 0.18763695657253265, "learning_rate": 1.5003053056480643e-05, "loss": 0.0053, "step": 15190 }, { "epoch": 32.0, "grad_norm": 0.1743103265762329, "learning_rate": 1.4944055293994551e-05, "loss": 0.0079, "step": 15200 }, { "epoch": 32.02105263157895, "grad_norm": 0.20997509360313416, "learning_rate": 1.4885153373122656e-05, "loss": 0.0078, "step": 15210 }, { "epoch": 32.04210526315789, "grad_norm": 0.2153230756521225, "learning_rate": 1.482634745490059e-05, "loss": 0.0085, "step": 15220 }, { "epoch": 32.06315789473684, "grad_norm": 0.2588895559310913, "learning_rate": 1.4767637700101466e-05, "loss": 0.0117, "step": 15230 }, { "epoch": 32.084210526315786, "grad_norm": 0.2758863568305969, "learning_rate": 1.4709024269235528e-05, "loss": 0.0064, "step": 15240 }, { "epoch": 32.10526315789474, "grad_norm": 0.2897651791572571, "learning_rate": 1.4650507322549684e-05, "loss": 0.0075, "step": 15250 }, { "epoch": 32.126315789473686, "grad_norm": 0.25971776247024536, "learning_rate": 1.4592087020026972e-05, "loss": 0.0073, "step": 15260 }, { "epoch": 32.14736842105263, "grad_norm": 0.23889927566051483, "learning_rate": 1.4533763521386318e-05, "loss": 0.0079, "step": 15270 }, { "epoch": 32.16842105263158, "grad_norm": 0.1666051298379898, "learning_rate": 1.44755369860819e-05, "loss": 0.0069, "step": 15280 }, { "epoch": 32.189473684210526, "grad_norm": 0.3269045054912567, "learning_rate": 1.441740757330287e-05, "loss": 0.0093, "step": 15290 }, { "epoch": 32.21052631578947, "grad_norm": 0.25456225872039795, "learning_rate": 1.4359375441972844e-05, "loss": 0.0077, "step": 15300 }, { "epoch": 32.23157894736842, "grad_norm": 0.17168349027633667, "learning_rate": 1.4301440750749395e-05, "loss": 0.0076, "step": 15310 }, { "epoch": 32.252631578947366, "grad_norm": 0.1563490927219391, "learning_rate": 1.4243603658023808e-05, "loss": 0.0052, "step": 15320 }, { "epoch": 32.27368421052632, "grad_norm": 0.25623518228530884, "learning_rate": 1.4185864321920444e-05, "loss": 0.01, "step": 15330 }, { "epoch": 32.294736842105266, "grad_norm": 0.18644222617149353, "learning_rate": 1.4128222900296485e-05, "loss": 0.0064, "step": 15340 }, { "epoch": 32.31578947368421, "grad_norm": 0.22770662605762482, "learning_rate": 1.407067955074135e-05, "loss": 0.0078, "step": 15350 }, { "epoch": 32.33684210526316, "grad_norm": 0.24341079592704773, "learning_rate": 1.4013234430576356e-05, "loss": 0.0085, "step": 15360 }, { "epoch": 32.357894736842105, "grad_norm": 0.22318828105926514, "learning_rate": 1.3955887696854286e-05, "loss": 0.0073, "step": 15370 }, { "epoch": 32.37894736842105, "grad_norm": 0.16290991008281708, "learning_rate": 1.38986395063589e-05, "loss": 0.0089, "step": 15380 }, { "epoch": 32.4, "grad_norm": 0.1288612335920334, "learning_rate": 1.3841490015604597e-05, "loss": 0.0094, "step": 15390 }, { "epoch": 32.421052631578945, "grad_norm": 0.4018467366695404, "learning_rate": 1.3784439380835879e-05, "loss": 0.0055, "step": 15400 }, { "epoch": 32.44210526315789, "grad_norm": 0.21378478407859802, "learning_rate": 1.3727487758026986e-05, "loss": 0.0101, "step": 15410 }, { "epoch": 32.463157894736845, "grad_norm": 0.12224799394607544, "learning_rate": 1.3670635302881525e-05, "loss": 0.0078, "step": 15420 }, { "epoch": 32.48421052631579, "grad_norm": 0.24979203939437866, "learning_rate": 1.3613882170831888e-05, "loss": 0.0073, "step": 15430 }, { "epoch": 32.50526315789474, "grad_norm": 0.1486126184463501, "learning_rate": 1.355722851703901e-05, "loss": 0.0068, "step": 15440 }, { "epoch": 32.526315789473685, "grad_norm": 0.30153027176856995, "learning_rate": 1.3500674496391814e-05, "loss": 0.0059, "step": 15450 }, { "epoch": 32.54736842105263, "grad_norm": 0.10784848034381866, "learning_rate": 1.3444220263506795e-05, "loss": 0.0056, "step": 15460 }, { "epoch": 32.56842105263158, "grad_norm": 0.1558651477098465, "learning_rate": 1.3387865972727714e-05, "loss": 0.0062, "step": 15470 }, { "epoch": 32.589473684210525, "grad_norm": 0.13509288430213928, "learning_rate": 1.3331611778125036e-05, "loss": 0.0074, "step": 15480 }, { "epoch": 32.61052631578947, "grad_norm": 0.2354826033115387, "learning_rate": 1.3275457833495564e-05, "loss": 0.0091, "step": 15490 }, { "epoch": 32.63157894736842, "grad_norm": 0.16576234996318817, "learning_rate": 1.3219404292362065e-05, "loss": 0.0081, "step": 15500 }, { "epoch": 32.65263157894737, "grad_norm": 0.24482505023479462, "learning_rate": 1.3163451307972751e-05, "loss": 0.0062, "step": 15510 }, { "epoch": 32.67368421052632, "grad_norm": 0.2360767126083374, "learning_rate": 1.3107599033300977e-05, "loss": 0.0065, "step": 15520 }, { "epoch": 32.694736842105264, "grad_norm": 0.1606118232011795, "learning_rate": 1.305184762104471e-05, "loss": 0.0068, "step": 15530 }, { "epoch": 32.71578947368421, "grad_norm": 0.22887492179870605, "learning_rate": 1.2996197223626178e-05, "loss": 0.0077, "step": 15540 }, { "epoch": 32.73684210526316, "grad_norm": 0.1967179775238037, "learning_rate": 1.2940647993191457e-05, "loss": 0.0059, "step": 15550 }, { "epoch": 32.757894736842104, "grad_norm": 0.23685483634471893, "learning_rate": 1.2885200081610005e-05, "loss": 0.0059, "step": 15560 }, { "epoch": 32.77894736842105, "grad_norm": 0.2852192521095276, "learning_rate": 1.2829853640474316e-05, "loss": 0.0065, "step": 15570 }, { "epoch": 32.8, "grad_norm": 0.32896214723587036, "learning_rate": 1.2774608821099438e-05, "loss": 0.0064, "step": 15580 }, { "epoch": 32.82105263157895, "grad_norm": 0.2111087441444397, "learning_rate": 1.2719465774522577e-05, "loss": 0.0095, "step": 15590 }, { "epoch": 32.8421052631579, "grad_norm": 0.232244074344635, "learning_rate": 1.2664424651502755e-05, "loss": 0.0082, "step": 15600 }, { "epoch": 32.863157894736844, "grad_norm": 0.14995408058166504, "learning_rate": 1.260948560252026e-05, "loss": 0.0072, "step": 15610 }, { "epoch": 32.88421052631579, "grad_norm": 0.16271407902240753, "learning_rate": 1.2554648777776396e-05, "loss": 0.0079, "step": 15620 }, { "epoch": 32.90526315789474, "grad_norm": 0.18830609321594238, "learning_rate": 1.2499914327192919e-05, "loss": 0.007, "step": 15630 }, { "epoch": 32.92631578947368, "grad_norm": 0.15559962391853333, "learning_rate": 1.2445282400411722e-05, "loss": 0.0053, "step": 15640 }, { "epoch": 32.94736842105263, "grad_norm": 0.1678699254989624, "learning_rate": 1.2390753146794437e-05, "loss": 0.0052, "step": 15650 }, { "epoch": 32.96842105263158, "grad_norm": 0.13225388526916504, "learning_rate": 1.2336326715421925e-05, "loss": 0.0078, "step": 15660 }, { "epoch": 32.98947368421052, "grad_norm": 0.15565164387226105, "learning_rate": 1.2282003255094005e-05, "loss": 0.0064, "step": 15670 }, { "epoch": 33.01052631578948, "grad_norm": 0.15805837512016296, "learning_rate": 1.2227782914328928e-05, "loss": 0.0077, "step": 15680 }, { "epoch": 33.03157894736842, "grad_norm": 0.1343039572238922, "learning_rate": 1.2173665841363018e-05, "loss": 0.0056, "step": 15690 }, { "epoch": 33.05263157894737, "grad_norm": 0.22710512578487396, "learning_rate": 1.211965218415032e-05, "loss": 0.0071, "step": 15700 }, { "epoch": 33.07368421052632, "grad_norm": 0.13804076611995697, "learning_rate": 1.2065742090362082e-05, "loss": 0.0085, "step": 15710 }, { "epoch": 33.09473684210526, "grad_norm": 0.20180857181549072, "learning_rate": 1.2011935707386457e-05, "loss": 0.006, "step": 15720 }, { "epoch": 33.11578947368421, "grad_norm": 0.2238907665014267, "learning_rate": 1.1958233182328044e-05, "loss": 0.0096, "step": 15730 }, { "epoch": 33.136842105263156, "grad_norm": 0.19997446238994598, "learning_rate": 1.1904634662007474e-05, "loss": 0.0069, "step": 15740 }, { "epoch": 33.1578947368421, "grad_norm": 0.17497725784778595, "learning_rate": 1.1851140292961088e-05, "loss": 0.0068, "step": 15750 }, { "epoch": 33.17894736842105, "grad_norm": 0.2541259825229645, "learning_rate": 1.1797750221440424e-05, "loss": 0.0086, "step": 15760 }, { "epoch": 33.2, "grad_norm": 0.2602000832557678, "learning_rate": 1.1744464593411897e-05, "loss": 0.0071, "step": 15770 }, { "epoch": 33.22105263157895, "grad_norm": 0.2533060610294342, "learning_rate": 1.1691283554556399e-05, "loss": 0.0057, "step": 15780 }, { "epoch": 33.242105263157896, "grad_norm": 0.28924787044525146, "learning_rate": 1.1638207250268834e-05, "loss": 0.0086, "step": 15790 }, { "epoch": 33.26315789473684, "grad_norm": 0.13425567746162415, "learning_rate": 1.158523582565782e-05, "loss": 0.005, "step": 15800 }, { "epoch": 33.28421052631579, "grad_norm": 0.30303871631622314, "learning_rate": 1.1532369425545192e-05, "loss": 0.0082, "step": 15810 }, { "epoch": 33.305263157894736, "grad_norm": 0.1883496791124344, "learning_rate": 1.1479608194465662e-05, "loss": 0.0074, "step": 15820 }, { "epoch": 33.32631578947368, "grad_norm": 0.17850254476070404, "learning_rate": 1.1426952276666442e-05, "loss": 0.0078, "step": 15830 }, { "epoch": 33.34736842105263, "grad_norm": 0.3247033655643463, "learning_rate": 1.1374401816106778e-05, "loss": 0.006, "step": 15840 }, { "epoch": 33.36842105263158, "grad_norm": 0.21902047097682953, "learning_rate": 1.1321956956457646e-05, "loss": 0.0066, "step": 15850 }, { "epoch": 33.38947368421053, "grad_norm": 0.38270804286003113, "learning_rate": 1.1269617841101277e-05, "loss": 0.0086, "step": 15860 }, { "epoch": 33.410526315789475, "grad_norm": 0.21657057106494904, "learning_rate": 1.1217384613130804e-05, "loss": 0.0123, "step": 15870 }, { "epoch": 33.43157894736842, "grad_norm": 0.2505341172218323, "learning_rate": 1.11652574153499e-05, "loss": 0.0063, "step": 15880 }, { "epoch": 33.45263157894737, "grad_norm": 0.21321310102939606, "learning_rate": 1.1113236390272303e-05, "loss": 0.0065, "step": 15890 }, { "epoch": 33.473684210526315, "grad_norm": 0.1688096672296524, "learning_rate": 1.106132168012155e-05, "loss": 0.0064, "step": 15900 }, { "epoch": 33.49473684210526, "grad_norm": 0.1778864562511444, "learning_rate": 1.1009513426830448e-05, "loss": 0.0076, "step": 15910 }, { "epoch": 33.51578947368421, "grad_norm": 0.17639154195785522, "learning_rate": 1.0957811772040777e-05, "loss": 0.0056, "step": 15920 }, { "epoch": 33.536842105263155, "grad_norm": 0.2643498480319977, "learning_rate": 1.0906216857102913e-05, "loss": 0.011, "step": 15930 }, { "epoch": 33.55789473684211, "grad_norm": 0.17843037843704224, "learning_rate": 1.0854728823075355e-05, "loss": 0.0082, "step": 15940 }, { "epoch": 33.578947368421055, "grad_norm": 0.19937655329704285, "learning_rate": 1.0803347810724452e-05, "loss": 0.0082, "step": 15950 }, { "epoch": 33.6, "grad_norm": 0.23345668613910675, "learning_rate": 1.0752073960523911e-05, "loss": 0.0063, "step": 15960 }, { "epoch": 33.62105263157895, "grad_norm": 0.2531355917453766, "learning_rate": 1.070090741265447e-05, "loss": 0.0092, "step": 15970 }, { "epoch": 33.642105263157895, "grad_norm": 0.26380395889282227, "learning_rate": 1.0649848307003547e-05, "loss": 0.0074, "step": 15980 }, { "epoch": 33.66315789473684, "grad_norm": 0.19202816486358643, "learning_rate": 1.0598896783164757e-05, "loss": 0.0081, "step": 15990 }, { "epoch": 33.68421052631579, "grad_norm": 0.1270727813243866, "learning_rate": 1.0548052980437645e-05, "loss": 0.0075, "step": 16000 }, { "epoch": 33.705263157894734, "grad_norm": 0.2557709813117981, "learning_rate": 1.049731703782722e-05, "loss": 0.009, "step": 16010 }, { "epoch": 33.72631578947369, "grad_norm": 0.26593950390815735, "learning_rate": 1.0446689094043587e-05, "loss": 0.0084, "step": 16020 }, { "epoch": 33.747368421052634, "grad_norm": 0.20001785457134247, "learning_rate": 1.039616928750165e-05, "loss": 0.0081, "step": 16030 }, { "epoch": 33.76842105263158, "grad_norm": 0.22243507206439972, "learning_rate": 1.0345757756320612e-05, "loss": 0.0086, "step": 16040 }, { "epoch": 33.78947368421053, "grad_norm": 0.21397337317466736, "learning_rate": 1.0295454638323666e-05, "loss": 0.0093, "step": 16050 }, { "epoch": 33.810526315789474, "grad_norm": 0.13067622482776642, "learning_rate": 1.0245260071037632e-05, "loss": 0.0054, "step": 16060 }, { "epoch": 33.83157894736842, "grad_norm": 0.19286282360553741, "learning_rate": 1.0195174191692518e-05, "loss": 0.0069, "step": 16070 }, { "epoch": 33.85263157894737, "grad_norm": 0.1560581773519516, "learning_rate": 1.014519713722124e-05, "loss": 0.0059, "step": 16080 }, { "epoch": 33.873684210526314, "grad_norm": 0.35694295167922974, "learning_rate": 1.0095329044259132e-05, "loss": 0.0081, "step": 16090 }, { "epoch": 33.89473684210526, "grad_norm": 0.16174718737602234, "learning_rate": 1.004557004914365e-05, "loss": 0.0063, "step": 16100 }, { "epoch": 33.915789473684214, "grad_norm": 0.1908571422100067, "learning_rate": 9.995920287914007e-06, "loss": 0.0058, "step": 16110 }, { "epoch": 33.93684210526316, "grad_norm": 0.3351208567619324, "learning_rate": 9.946379896310737e-06, "loss": 0.0059, "step": 16120 }, { "epoch": 33.95789473684211, "grad_norm": 0.187356099486351, "learning_rate": 9.896949009775396e-06, "loss": 0.0071, "step": 16130 }, { "epoch": 33.97894736842105, "grad_norm": 0.20981526374816895, "learning_rate": 9.847627763450134e-06, "loss": 0.0067, "step": 16140 }, { "epoch": 34.0, "grad_norm": 0.18283602595329285, "learning_rate": 9.798416292177337e-06, "loss": 0.0059, "step": 16150 }, { "epoch": 34.02105263157895, "grad_norm": 0.212486132979393, "learning_rate": 9.74931473049932e-06, "loss": 0.0075, "step": 16160 }, { "epoch": 34.04210526315789, "grad_norm": 0.22119209170341492, "learning_rate": 9.700323212657847e-06, "loss": 0.0069, "step": 16170 }, { "epoch": 34.06315789473684, "grad_norm": 0.2188841849565506, "learning_rate": 9.65144187259388e-06, "loss": 0.0052, "step": 16180 }, { "epoch": 34.084210526315786, "grad_norm": 0.3244289755821228, "learning_rate": 9.602670843947132e-06, "loss": 0.0056, "step": 16190 }, { "epoch": 34.10526315789474, "grad_norm": 0.20056390762329102, "learning_rate": 9.554010260055713e-06, "loss": 0.0064, "step": 16200 }, { "epoch": 34.126315789473686, "grad_norm": 0.23310799896717072, "learning_rate": 9.505460253955834e-06, "loss": 0.0087, "step": 16210 }, { "epoch": 34.14736842105263, "grad_norm": 0.14263103902339935, "learning_rate": 9.457020958381324e-06, "loss": 0.0069, "step": 16220 }, { "epoch": 34.16842105263158, "grad_norm": 0.1621190309524536, "learning_rate": 9.408692505763395e-06, "loss": 0.0073, "step": 16230 }, { "epoch": 34.189473684210526, "grad_norm": 0.19166207313537598, "learning_rate": 9.360475028230181e-06, "loss": 0.0057, "step": 16240 }, { "epoch": 34.21052631578947, "grad_norm": 0.21739833056926727, "learning_rate": 9.312368657606412e-06, "loss": 0.0061, "step": 16250 }, { "epoch": 34.23157894736842, "grad_norm": 0.14224664866924286, "learning_rate": 9.264373525413096e-06, "loss": 0.0079, "step": 16260 }, { "epoch": 34.252631578947366, "grad_norm": 0.16238942742347717, "learning_rate": 9.216489762867058e-06, "loss": 0.0063, "step": 16270 }, { "epoch": 34.27368421052632, "grad_norm": 0.1567145586013794, "learning_rate": 9.168717500880708e-06, "loss": 0.0061, "step": 16280 }, { "epoch": 34.294736842105266, "grad_norm": 0.12805266678333282, "learning_rate": 9.121056870061574e-06, "loss": 0.0054, "step": 16290 }, { "epoch": 34.31578947368421, "grad_norm": 0.19360177218914032, "learning_rate": 9.073508000711983e-06, "loss": 0.0079, "step": 16300 }, { "epoch": 34.33684210526316, "grad_norm": 0.40272781252861023, "learning_rate": 9.026071022828758e-06, "loss": 0.0056, "step": 16310 }, { "epoch": 34.357894736842105, "grad_norm": 0.1367650181055069, "learning_rate": 8.978746066102771e-06, "loss": 0.0067, "step": 16320 }, { "epoch": 34.37894736842105, "grad_norm": 0.12401112914085388, "learning_rate": 8.931533259918634e-06, "loss": 0.0087, "step": 16330 }, { "epoch": 34.4, "grad_norm": 0.11796879768371582, "learning_rate": 8.884432733354382e-06, "loss": 0.0064, "step": 16340 }, { "epoch": 34.421052631578945, "grad_norm": 0.18151667714118958, "learning_rate": 8.837444615181029e-06, "loss": 0.0088, "step": 16350 }, { "epoch": 34.44210526315789, "grad_norm": 0.1514277160167694, "learning_rate": 8.790569033862323e-06, "loss": 0.0077, "step": 16360 }, { "epoch": 34.463157894736845, "grad_norm": 0.5123454332351685, "learning_rate": 8.7438061175543e-06, "loss": 0.0082, "step": 16370 }, { "epoch": 34.48421052631579, "grad_norm": 0.35679662227630615, "learning_rate": 8.697155994104978e-06, "loss": 0.0061, "step": 16380 }, { "epoch": 34.50526315789474, "grad_norm": 0.19193775951862335, "learning_rate": 8.650618791054033e-06, "loss": 0.0053, "step": 16390 }, { "epoch": 34.526315789473685, "grad_norm": 0.11048384010791779, "learning_rate": 8.604194635632373e-06, "loss": 0.0066, "step": 16400 }, { "epoch": 34.54736842105263, "grad_norm": 0.15789823234081268, "learning_rate": 8.557883654761906e-06, "loss": 0.0087, "step": 16410 }, { "epoch": 34.56842105263158, "grad_norm": 0.1315511167049408, "learning_rate": 8.511685975055061e-06, "loss": 0.0055, "step": 16420 }, { "epoch": 34.589473684210525, "grad_norm": 0.20763583481311798, "learning_rate": 8.46560172281452e-06, "loss": 0.0061, "step": 16430 }, { "epoch": 34.61052631578947, "grad_norm": 0.26897597312927246, "learning_rate": 8.419631024032893e-06, "loss": 0.0065, "step": 16440 }, { "epoch": 34.63157894736842, "grad_norm": 0.16659536957740784, "learning_rate": 8.373774004392293e-06, "loss": 0.0073, "step": 16450 }, { "epoch": 34.65263157894737, "grad_norm": 0.1896999031305313, "learning_rate": 8.32803078926409e-06, "loss": 0.0088, "step": 16460 }, { "epoch": 34.67368421052632, "grad_norm": 0.2654164433479309, "learning_rate": 8.282401503708454e-06, "loss": 0.0077, "step": 16470 }, { "epoch": 34.694736842105264, "grad_norm": 0.2373175024986267, "learning_rate": 8.23688627247412e-06, "loss": 0.0099, "step": 16480 }, { "epoch": 34.71578947368421, "grad_norm": 0.07604005932807922, "learning_rate": 8.191485219998007e-06, "loss": 0.0051, "step": 16490 }, { "epoch": 34.73684210526316, "grad_norm": 0.1973443180322647, "learning_rate": 8.146198470404843e-06, "loss": 0.0091, "step": 16500 }, { "epoch": 34.757894736842104, "grad_norm": 0.12699267268180847, "learning_rate": 8.101026147506897e-06, "loss": 0.0059, "step": 16510 }, { "epoch": 34.77894736842105, "grad_norm": 0.1474410742521286, "learning_rate": 8.05596837480353e-06, "loss": 0.0074, "step": 16520 }, { "epoch": 34.8, "grad_norm": 0.18893833458423615, "learning_rate": 8.011025275480998e-06, "loss": 0.0073, "step": 16530 }, { "epoch": 34.82105263157895, "grad_norm": 0.11666407436132431, "learning_rate": 7.966196972412027e-06, "loss": 0.0072, "step": 16540 }, { "epoch": 34.8421052631579, "grad_norm": 0.18330778181552887, "learning_rate": 7.92148358815547e-06, "loss": 0.0062, "step": 16550 }, { "epoch": 34.863157894736844, "grad_norm": 0.22751784324645996, "learning_rate": 7.87688524495604e-06, "loss": 0.0068, "step": 16560 }, { "epoch": 34.88421052631579, "grad_norm": 0.2337121069431305, "learning_rate": 7.83240206474386e-06, "loss": 0.0052, "step": 16570 }, { "epoch": 34.90526315789474, "grad_norm": 0.2381645143032074, "learning_rate": 7.788034169134272e-06, "loss": 0.0054, "step": 16580 }, { "epoch": 34.92631578947368, "grad_norm": 0.15580588579177856, "learning_rate": 7.743781679427414e-06, "loss": 0.0078, "step": 16590 }, { "epoch": 34.94736842105263, "grad_norm": 0.21617160737514496, "learning_rate": 7.699644716607895e-06, "loss": 0.0132, "step": 16600 }, { "epoch": 34.96842105263158, "grad_norm": 0.2272576540708542, "learning_rate": 7.655623401344486e-06, "loss": 0.0051, "step": 16610 }, { "epoch": 34.98947368421052, "grad_norm": 0.8391486406326294, "learning_rate": 7.611717853989775e-06, "loss": 0.0079, "step": 16620 }, { "epoch": 35.01052631578948, "grad_norm": 0.14792363345623016, "learning_rate": 7.567928194579854e-06, "loss": 0.0081, "step": 16630 }, { "epoch": 35.03157894736842, "grad_norm": 0.15669766068458557, "learning_rate": 7.524254542833997e-06, "loss": 0.0054, "step": 16640 }, { "epoch": 35.05263157894737, "grad_norm": 0.24068182706832886, "learning_rate": 7.480697018154286e-06, "loss": 0.0055, "step": 16650 }, { "epoch": 35.07368421052632, "grad_norm": 0.12318270653486252, "learning_rate": 7.437255739625332e-06, "loss": 0.0063, "step": 16660 }, { "epoch": 35.09473684210526, "grad_norm": 0.21222592890262604, "learning_rate": 7.393930826013923e-06, "loss": 0.0053, "step": 16670 }, { "epoch": 35.11578947368421, "grad_norm": 0.18709376454353333, "learning_rate": 7.350722395768722e-06, "loss": 0.0061, "step": 16680 }, { "epoch": 35.136842105263156, "grad_norm": 0.34556302428245544, "learning_rate": 7.307630567019963e-06, "loss": 0.0057, "step": 16690 }, { "epoch": 35.1578947368421, "grad_norm": 0.17476578056812286, "learning_rate": 7.264655457579e-06, "loss": 0.0073, "step": 16700 }, { "epoch": 35.17894736842105, "grad_norm": 0.15282899141311646, "learning_rate": 7.221797184938184e-06, "loss": 0.0073, "step": 16710 }, { "epoch": 35.2, "grad_norm": 0.3032562732696533, "learning_rate": 7.179055866270373e-06, "loss": 0.0069, "step": 16720 }, { "epoch": 35.22105263157895, "grad_norm": 0.19090840220451355, "learning_rate": 7.136431618428707e-06, "loss": 0.0064, "step": 16730 }, { "epoch": 35.242105263157896, "grad_norm": 0.1088152602314949, "learning_rate": 7.09392455794628e-06, "loss": 0.0045, "step": 16740 }, { "epoch": 35.26315789473684, "grad_norm": 0.19404193758964539, "learning_rate": 7.051534801035725e-06, "loss": 0.0058, "step": 16750 }, { "epoch": 35.28421052631579, "grad_norm": 0.20368295907974243, "learning_rate": 7.00926246358905e-06, "loss": 0.0059, "step": 16760 }, { "epoch": 35.305263157894736, "grad_norm": 0.18923069536685944, "learning_rate": 6.967107661177191e-06, "loss": 0.0055, "step": 16770 }, { "epoch": 35.32631578947368, "grad_norm": 0.17283311486244202, "learning_rate": 6.925070509049786e-06, "loss": 0.0066, "step": 16780 }, { "epoch": 35.34736842105263, "grad_norm": 0.15812484920024872, "learning_rate": 6.883151122134812e-06, "loss": 0.0062, "step": 16790 }, { "epoch": 35.36842105263158, "grad_norm": 0.25983771681785583, "learning_rate": 6.8413496150382394e-06, "loss": 0.0076, "step": 16800 }, { "epoch": 35.38947368421053, "grad_norm": 0.18617279827594757, "learning_rate": 6.7996661020438165e-06, "loss": 0.0057, "step": 16810 }, { "epoch": 35.410526315789475, "grad_norm": 0.25430840253829956, "learning_rate": 6.758100697112662e-06, "loss": 0.0084, "step": 16820 }, { "epoch": 35.43157894736842, "grad_norm": 0.19629605114459991, "learning_rate": 6.716653513883026e-06, "loss": 0.0067, "step": 16830 }, { "epoch": 35.45263157894737, "grad_norm": 0.2564931809902191, "learning_rate": 6.675324665669913e-06, "loss": 0.0057, "step": 16840 }, { "epoch": 35.473684210526315, "grad_norm": 0.24880260229110718, "learning_rate": 6.634114265464803e-06, "loss": 0.0067, "step": 16850 }, { "epoch": 35.49473684210526, "grad_norm": 0.13853061199188232, "learning_rate": 6.59302242593538e-06, "loss": 0.0063, "step": 16860 }, { "epoch": 35.51578947368421, "grad_norm": 0.1486787348985672, "learning_rate": 6.552049259425141e-06, "loss": 0.0054, "step": 16870 }, { "epoch": 35.536842105263155, "grad_norm": 0.1714516133069992, "learning_rate": 6.511194877953181e-06, "loss": 0.0044, "step": 16880 }, { "epoch": 35.55789473684211, "grad_norm": 0.2381514459848404, "learning_rate": 6.470459393213813e-06, "loss": 0.0059, "step": 16890 }, { "epoch": 35.578947368421055, "grad_norm": 0.3203369975090027, "learning_rate": 6.429842916576279e-06, "loss": 0.0058, "step": 16900 }, { "epoch": 35.6, "grad_norm": 0.12113554030656815, "learning_rate": 6.389345559084503e-06, "loss": 0.0073, "step": 16910 }, { "epoch": 35.62105263157895, "grad_norm": 0.22297073900699615, "learning_rate": 6.348967431456682e-06, "loss": 0.0064, "step": 16920 }, { "epoch": 35.642105263157895, "grad_norm": 0.09199515730142593, "learning_rate": 6.30870864408511e-06, "loss": 0.0057, "step": 16930 }, { "epoch": 35.66315789473684, "grad_norm": 0.14955952763557434, "learning_rate": 6.268569307035754e-06, "loss": 0.005, "step": 16940 }, { "epoch": 35.68421052631579, "grad_norm": 0.21413952112197876, "learning_rate": 6.228549530048022e-06, "loss": 0.0067, "step": 16950 }, { "epoch": 35.705263157894734, "grad_norm": 0.19949617981910706, "learning_rate": 6.1886494225344814e-06, "loss": 0.0061, "step": 16960 }, { "epoch": 35.72631578947369, "grad_norm": 0.12478101253509521, "learning_rate": 6.148869093580479e-06, "loss": 0.0059, "step": 16970 }, { "epoch": 35.747368421052634, "grad_norm": 0.13965237140655518, "learning_rate": 6.109208651943921e-06, "loss": 0.0068, "step": 16980 }, { "epoch": 35.76842105263158, "grad_norm": 0.22288350760936737, "learning_rate": 6.069668206054946e-06, "loss": 0.0055, "step": 16990 }, { "epoch": 35.78947368421053, "grad_norm": 0.30123525857925415, "learning_rate": 6.0302478640156145e-06, "loss": 0.007, "step": 17000 }, { "epoch": 35.810526315789474, "grad_norm": 0.16799479722976685, "learning_rate": 5.990947733599644e-06, "loss": 0.0072, "step": 17010 }, { "epoch": 35.83157894736842, "grad_norm": 0.2649695575237274, "learning_rate": 5.951767922252105e-06, "loss": 0.0062, "step": 17020 }, { "epoch": 35.85263157894737, "grad_norm": 0.19558314979076385, "learning_rate": 5.912708537089068e-06, "loss": 0.0059, "step": 17030 }, { "epoch": 35.873684210526314, "grad_norm": 0.30635038018226624, "learning_rate": 5.873769684897434e-06, "loss": 0.0098, "step": 17040 }, { "epoch": 35.89473684210526, "grad_norm": 0.11542361974716187, "learning_rate": 5.834951472134514e-06, "loss": 0.006, "step": 17050 }, { "epoch": 35.915789473684214, "grad_norm": 0.13567988574504852, "learning_rate": 5.796254004927832e-06, "loss": 0.0051, "step": 17060 }, { "epoch": 35.93684210526316, "grad_norm": 0.20508268475532532, "learning_rate": 5.757677389074806e-06, "loss": 0.0056, "step": 17070 }, { "epoch": 35.95789473684211, "grad_norm": 0.14656208455562592, "learning_rate": 5.719221730042385e-06, "loss": 0.0052, "step": 17080 }, { "epoch": 35.97894736842105, "grad_norm": 0.20173552632331848, "learning_rate": 5.680887132966911e-06, "loss": 0.0065, "step": 17090 }, { "epoch": 36.0, "grad_norm": 0.15530571341514587, "learning_rate": 5.642673702653683e-06, "loss": 0.0054, "step": 17100 }, { "epoch": 36.02105263157895, "grad_norm": 0.13736867904663086, "learning_rate": 5.604581543576781e-06, "loss": 0.007, "step": 17110 }, { "epoch": 36.04210526315789, "grad_norm": 0.21100950241088867, "learning_rate": 5.566610759878704e-06, "loss": 0.0063, "step": 17120 }, { "epoch": 36.06315789473684, "grad_norm": 0.3151254653930664, "learning_rate": 5.528761455370119e-06, "loss": 0.0053, "step": 17130 }, { "epoch": 36.084210526315786, "grad_norm": 0.27163761854171753, "learning_rate": 5.491033733529594e-06, "loss": 0.0076, "step": 17140 }, { "epoch": 36.10526315789474, "grad_norm": 0.15246763825416565, "learning_rate": 5.453427697503255e-06, "loss": 0.0056, "step": 17150 }, { "epoch": 36.126315789473686, "grad_norm": 0.24492502212524414, "learning_rate": 5.415943450104599e-06, "loss": 0.0053, "step": 17160 }, { "epoch": 36.14736842105263, "grad_norm": 0.1335715800523758, "learning_rate": 5.378581093814111e-06, "loss": 0.0059, "step": 17170 }, { "epoch": 36.16842105263158, "grad_norm": 0.35286378860473633, "learning_rate": 5.3413407307790375e-06, "loss": 0.0058, "step": 17180 }, { "epoch": 36.189473684210526, "grad_norm": 0.2539844512939453, "learning_rate": 5.30422246281313e-06, "loss": 0.0062, "step": 17190 }, { "epoch": 36.21052631578947, "grad_norm": 0.25194138288497925, "learning_rate": 5.267226391396296e-06, "loss": 0.0067, "step": 17200 }, { "epoch": 36.23157894736842, "grad_norm": 0.14362867176532745, "learning_rate": 5.2303526176744e-06, "loss": 0.005, "step": 17210 }, { "epoch": 36.252631578947366, "grad_norm": 0.17609809339046478, "learning_rate": 5.193601242458929e-06, "loss": 0.0057, "step": 17220 }, { "epoch": 36.27368421052632, "grad_norm": 0.19929446280002594, "learning_rate": 5.156972366226714e-06, "loss": 0.0053, "step": 17230 }, { "epoch": 36.294736842105266, "grad_norm": 0.20093467831611633, "learning_rate": 5.120466089119735e-06, "loss": 0.0068, "step": 17240 }, { "epoch": 36.31578947368421, "grad_norm": 0.13467498123645782, "learning_rate": 5.084082510944749e-06, "loss": 0.006, "step": 17250 }, { "epoch": 36.33684210526316, "grad_norm": 0.13269397616386414, "learning_rate": 5.047821731173058e-06, "loss": 0.0068, "step": 17260 }, { "epoch": 36.357894736842105, "grad_norm": 0.19329935312271118, "learning_rate": 5.011683848940274e-06, "loss": 0.0058, "step": 17270 }, { "epoch": 36.37894736842105, "grad_norm": 0.20316052436828613, "learning_rate": 4.975668963045954e-06, "loss": 0.0064, "step": 17280 }, { "epoch": 36.4, "grad_norm": 0.7139684557914734, "learning_rate": 4.9397771719534525e-06, "loss": 0.0058, "step": 17290 }, { "epoch": 36.421052631578945, "grad_norm": 0.11600442230701447, "learning_rate": 4.904008573789548e-06, "loss": 0.0048, "step": 17300 }, { "epoch": 36.44210526315789, "grad_norm": 0.31775158643722534, "learning_rate": 4.8683632663442005e-06, "loss": 0.0089, "step": 17310 }, { "epoch": 36.463157894736845, "grad_norm": 0.14351940155029297, "learning_rate": 4.832841347070343e-06, "loss": 0.0101, "step": 17320 }, { "epoch": 36.48421052631579, "grad_norm": 0.0865265503525734, "learning_rate": 4.797442913083539e-06, "loss": 0.0055, "step": 17330 }, { "epoch": 36.50526315789474, "grad_norm": 0.12475805729627609, "learning_rate": 4.7621680611617596e-06, "loss": 0.0059, "step": 17340 }, { "epoch": 36.526315789473685, "grad_norm": 0.13018088042736053, "learning_rate": 4.727016887745095e-06, "loss": 0.0066, "step": 17350 }, { "epoch": 36.54736842105263, "grad_norm": 0.24857190251350403, "learning_rate": 4.691989488935511e-06, "loss": 0.007, "step": 17360 }, { "epoch": 36.56842105263158, "grad_norm": 0.15572337806224823, "learning_rate": 4.657085960496588e-06, "loss": 0.0054, "step": 17370 }, { "epoch": 36.589473684210525, "grad_norm": 0.13048893213272095, "learning_rate": 4.6223063978532265e-06, "loss": 0.0091, "step": 17380 }, { "epoch": 36.61052631578947, "grad_norm": 0.31207430362701416, "learning_rate": 4.587650896091439e-06, "loss": 0.0085, "step": 17390 }, { "epoch": 36.63157894736842, "grad_norm": 0.18984505534172058, "learning_rate": 4.553119549958035e-06, "loss": 0.0078, "step": 17400 }, { "epoch": 36.65263157894737, "grad_norm": 0.21136991679668427, "learning_rate": 4.518712453860385e-06, "loss": 0.0051, "step": 17410 }, { "epoch": 36.67368421052632, "grad_norm": 0.20009778439998627, "learning_rate": 4.484429701866205e-06, "loss": 0.0053, "step": 17420 }, { "epoch": 36.694736842105264, "grad_norm": 0.37487050890922546, "learning_rate": 4.4502713877031975e-06, "loss": 0.0052, "step": 17430 }, { "epoch": 36.71578947368421, "grad_norm": 0.3514719605445862, "learning_rate": 4.416237604758911e-06, "loss": 0.0051, "step": 17440 }, { "epoch": 36.73684210526316, "grad_norm": 0.21781769394874573, "learning_rate": 4.3823284460804025e-06, "loss": 0.0062, "step": 17450 }, { "epoch": 36.757894736842104, "grad_norm": 0.15801717340946198, "learning_rate": 4.348544004374011e-06, "loss": 0.0044, "step": 17460 }, { "epoch": 36.77894736842105, "grad_norm": 0.15864868462085724, "learning_rate": 4.314884372005123e-06, "loss": 0.0064, "step": 17470 }, { "epoch": 36.8, "grad_norm": 0.26204487681388855, "learning_rate": 4.281349640997867e-06, "loss": 0.007, "step": 17480 }, { "epoch": 36.82105263157895, "grad_norm": 0.18630720674991608, "learning_rate": 4.247939903034942e-06, "loss": 0.0058, "step": 17490 }, { "epoch": 36.8421052631579, "grad_norm": 0.1479000300168991, "learning_rate": 4.214655249457284e-06, "loss": 0.0054, "step": 17500 }, { "epoch": 36.863157894736844, "grad_norm": 0.221022829413414, "learning_rate": 4.181495771263855e-06, "loss": 0.0083, "step": 17510 }, { "epoch": 36.88421052631579, "grad_norm": 0.09618619829416275, "learning_rate": 4.148461559111427e-06, "loss": 0.0084, "step": 17520 }, { "epoch": 36.90526315789474, "grad_norm": 0.2048938125371933, "learning_rate": 4.115552703314252e-06, "loss": 0.0055, "step": 17530 }, { "epoch": 36.92631578947368, "grad_norm": 0.19127421081066132, "learning_rate": 4.082769293843886e-06, "loss": 0.0081, "step": 17540 }, { "epoch": 36.94736842105263, "grad_norm": 0.21639379858970642, "learning_rate": 4.050111420328939e-06, "loss": 0.0058, "step": 17550 }, { "epoch": 36.96842105263158, "grad_norm": 0.17874519526958466, "learning_rate": 4.017579172054764e-06, "loss": 0.0068, "step": 17560 }, { "epoch": 36.98947368421052, "grad_norm": 0.29711785912513733, "learning_rate": 3.985172637963308e-06, "loss": 0.0058, "step": 17570 }, { "epoch": 37.01052631578948, "grad_norm": 0.17080585658550262, "learning_rate": 3.952891906652784e-06, "loss": 0.0075, "step": 17580 }, { "epoch": 37.03157894736842, "grad_norm": 0.10525134205818176, "learning_rate": 3.920737066377478e-06, "loss": 0.0069, "step": 17590 }, { "epoch": 37.05263157894737, "grad_norm": 0.18288913369178772, "learning_rate": 3.888708205047509e-06, "loss": 0.006, "step": 17600 }, { "epoch": 37.07368421052632, "grad_norm": 0.22854484617710114, "learning_rate": 3.856805410228542e-06, "loss": 0.0062, "step": 17610 }, { "epoch": 37.09473684210526, "grad_norm": 0.23764319717884064, "learning_rate": 3.82502876914162e-06, "loss": 0.0071, "step": 17620 }, { "epoch": 37.11578947368421, "grad_norm": 0.12854120135307312, "learning_rate": 3.7933783686628586e-06, "loss": 0.0081, "step": 17630 }, { "epoch": 37.136842105263156, "grad_norm": 0.28885728120803833, "learning_rate": 3.7618542953232306e-06, "loss": 0.0067, "step": 17640 }, { "epoch": 37.1578947368421, "grad_norm": 0.20770415663719177, "learning_rate": 3.7304566353083658e-06, "loss": 0.009, "step": 17650 }, { "epoch": 37.17894736842105, "grad_norm": 0.34234145283699036, "learning_rate": 3.6991854744582555e-06, "loss": 0.0061, "step": 17660 }, { "epoch": 37.2, "grad_norm": 0.18601037561893463, "learning_rate": 3.6680408982670777e-06, "loss": 0.0057, "step": 17670 }, { "epoch": 37.22105263157895, "grad_norm": 0.1311025619506836, "learning_rate": 3.637022991882899e-06, "loss": 0.0061, "step": 17680 }, { "epoch": 37.242105263157896, "grad_norm": 0.1946367621421814, "learning_rate": 3.606131840107485e-06, "loss": 0.0057, "step": 17690 }, { "epoch": 37.26315789473684, "grad_norm": 0.15771858394145966, "learning_rate": 3.575367527396084e-06, "loss": 0.0047, "step": 17700 }, { "epoch": 37.28421052631579, "grad_norm": 0.13776181638240814, "learning_rate": 3.5447301378571386e-06, "loss": 0.0054, "step": 17710 }, { "epoch": 37.305263157894736, "grad_norm": 0.26486194133758545, "learning_rate": 3.514219755252113e-06, "loss": 0.0047, "step": 17720 }, { "epoch": 37.32631578947368, "grad_norm": 0.17364691197872162, "learning_rate": 3.4838364629952213e-06, "loss": 0.0061, "step": 17730 }, { "epoch": 37.34736842105263, "grad_norm": 0.15215320885181427, "learning_rate": 3.4535803441532123e-06, "loss": 0.0058, "step": 17740 }, { "epoch": 37.36842105263158, "grad_norm": 0.22190314531326294, "learning_rate": 3.4234514814451836e-06, "loss": 0.0062, "step": 17750 }, { "epoch": 37.38947368421053, "grad_norm": 0.17768807709217072, "learning_rate": 3.393449957242273e-06, "loss": 0.0046, "step": 17760 }, { "epoch": 37.410526315789475, "grad_norm": 0.14600935578346252, "learning_rate": 3.363575853567524e-06, "loss": 0.0068, "step": 17770 }, { "epoch": 37.43157894736842, "grad_norm": 0.29287049174308777, "learning_rate": 3.3338292520955826e-06, "loss": 0.0055, "step": 17780 }, { "epoch": 37.45263157894737, "grad_norm": 0.3140420913696289, "learning_rate": 3.304210234152516e-06, "loss": 0.0058, "step": 17790 }, { "epoch": 37.473684210526315, "grad_norm": 0.20797590911388397, "learning_rate": 3.2747188807155993e-06, "loss": 0.0054, "step": 17800 }, { "epoch": 37.49473684210526, "grad_norm": 0.27216771245002747, "learning_rate": 3.2453552724130643e-06, "loss": 0.0057, "step": 17810 }, { "epoch": 37.51578947368421, "grad_norm": 0.25675395131111145, "learning_rate": 3.216119489523889e-06, "loss": 0.0059, "step": 17820 }, { "epoch": 37.536842105263155, "grad_norm": 0.18905606865882874, "learning_rate": 3.1870116119775917e-06, "loss": 0.0071, "step": 17830 }, { "epoch": 37.55789473684211, "grad_norm": 0.36775752902030945, "learning_rate": 3.158031719353999e-06, "loss": 0.0054, "step": 17840 }, { "epoch": 37.578947368421055, "grad_norm": 0.5587309002876282, "learning_rate": 3.1291798908830273e-06, "loss": 0.0051, "step": 17850 }, { "epoch": 37.6, "grad_norm": 0.24791088700294495, "learning_rate": 3.1004562054444853e-06, "loss": 0.0058, "step": 17860 }, { "epoch": 37.62105263157895, "grad_norm": 0.12946897745132446, "learning_rate": 3.071860741567806e-06, "loss": 0.0067, "step": 17870 }, { "epoch": 37.642105263157895, "grad_norm": 0.16378577053546906, "learning_rate": 3.04339357743193e-06, "loss": 0.0038, "step": 17880 }, { "epoch": 37.66315789473684, "grad_norm": 0.14784088730812073, "learning_rate": 3.0150547908649628e-06, "loss": 0.005, "step": 17890 }, { "epoch": 37.68421052631579, "grad_norm": 0.10730218887329102, "learning_rate": 2.9868444593440957e-06, "loss": 0.0064, "step": 17900 }, { "epoch": 37.705263157894734, "grad_norm": 0.14930342137813568, "learning_rate": 2.9587626599952846e-06, "loss": 0.0074, "step": 17910 }, { "epoch": 37.72631578947369, "grad_norm": 0.21251827478408813, "learning_rate": 2.930809469593082e-06, "loss": 0.0074, "step": 17920 }, { "epoch": 37.747368421052634, "grad_norm": 0.12328379601240158, "learning_rate": 2.9029849645604733e-06, "loss": 0.0048, "step": 17930 }, { "epoch": 37.76842105263158, "grad_norm": 0.13373564183712006, "learning_rate": 2.8752892209685632e-06, "loss": 0.0083, "step": 17940 }, { "epoch": 37.78947368421053, "grad_norm": 0.1978817880153656, "learning_rate": 2.847722314536483e-06, "loss": 0.0055, "step": 17950 }, { "epoch": 37.810526315789474, "grad_norm": 0.1324198693037033, "learning_rate": 2.820284320631078e-06, "loss": 0.0105, "step": 17960 }, { "epoch": 37.83157894736842, "grad_norm": 0.22388894855976105, "learning_rate": 2.792975314266788e-06, "loss": 0.0048, "step": 17970 }, { "epoch": 37.85263157894737, "grad_norm": 0.14556686580181122, "learning_rate": 2.7657953701054007e-06, "loss": 0.0057, "step": 17980 }, { "epoch": 37.873684210526314, "grad_norm": 0.22952602803707123, "learning_rate": 2.7387445624558306e-06, "loss": 0.0059, "step": 17990 }, { "epoch": 37.89473684210526, "grad_norm": 0.1546177715063095, "learning_rate": 2.7118229652739747e-06, "loss": 0.0044, "step": 18000 }, { "epoch": 37.915789473684214, "grad_norm": 0.19021856784820557, "learning_rate": 2.6850306521624236e-06, "loss": 0.0064, "step": 18010 }, { "epoch": 37.93684210526316, "grad_norm": 0.17553071677684784, "learning_rate": 2.6583676963703507e-06, "loss": 0.0057, "step": 18020 }, { "epoch": 37.95789473684211, "grad_norm": 0.11338918656110764, "learning_rate": 2.631834170793268e-06, "loss": 0.0044, "step": 18030 }, { "epoch": 37.97894736842105, "grad_norm": 0.2741971015930176, "learning_rate": 2.6054301479728036e-06, "loss": 0.0061, "step": 18040 }, { "epoch": 38.0, "grad_norm": 0.09937745332717896, "learning_rate": 2.579155700096575e-06, "loss": 0.0044, "step": 18050 }, { "epoch": 38.02105263157895, "grad_norm": 0.13788659870624542, "learning_rate": 2.5530108989978873e-06, "loss": 0.0052, "step": 18060 }, { "epoch": 38.04210526315789, "grad_norm": 0.1968238353729248, "learning_rate": 2.5269958161556416e-06, "loss": 0.0059, "step": 18070 }, { "epoch": 38.06315789473684, "grad_norm": 0.18085478246212006, "learning_rate": 2.5011105226940888e-06, "loss": 0.01, "step": 18080 }, { "epoch": 38.084210526315786, "grad_norm": 0.26330244541168213, "learning_rate": 2.4753550893826248e-06, "loss": 0.0056, "step": 18090 }, { "epoch": 38.10526315789474, "grad_norm": 0.08467608690261841, "learning_rate": 2.4497295866356296e-06, "loss": 0.0052, "step": 18100 }, { "epoch": 38.126315789473686, "grad_norm": 0.14706610143184662, "learning_rate": 2.424234084512228e-06, "loss": 0.0048, "step": 18110 }, { "epoch": 38.14736842105263, "grad_norm": 0.11437925696372986, "learning_rate": 2.3988686527161687e-06, "loss": 0.006, "step": 18120 }, { "epoch": 38.16842105263158, "grad_norm": 0.18339034914970398, "learning_rate": 2.373633360595573e-06, "loss": 0.0046, "step": 18130 }, { "epoch": 38.189473684210526, "grad_norm": 0.3321571946144104, "learning_rate": 2.3485282771427585e-06, "loss": 0.0047, "step": 18140 }, { "epoch": 38.21052631578947, "grad_norm": 0.06080101057887077, "learning_rate": 2.3235534709940665e-06, "loss": 0.0064, "step": 18150 }, { "epoch": 38.23157894736842, "grad_norm": 0.1721467524766922, "learning_rate": 2.2987090104296617e-06, "loss": 0.0055, "step": 18160 }, { "epoch": 38.252631578947366, "grad_norm": 0.11653546988964081, "learning_rate": 2.273994963373355e-06, "loss": 0.0067, "step": 18170 }, { "epoch": 38.27368421052632, "grad_norm": 0.13548722863197327, "learning_rate": 2.249411397392409e-06, "loss": 0.0049, "step": 18180 }, { "epoch": 38.294736842105266, "grad_norm": 0.212895929813385, "learning_rate": 2.2249583796973506e-06, "loss": 0.0065, "step": 18190 }, { "epoch": 38.31578947368421, "grad_norm": 0.23653659224510193, "learning_rate": 2.200635977141796e-06, "loss": 0.0052, "step": 18200 }, { "epoch": 38.33684210526316, "grad_norm": 0.10319563746452332, "learning_rate": 2.17644425622226e-06, "loss": 0.005, "step": 18210 }, { "epoch": 38.357894736842105, "grad_norm": 0.16314345598220825, "learning_rate": 2.152383283077991e-06, "loss": 0.0048, "step": 18220 }, { "epoch": 38.37894736842105, "grad_norm": 0.1677694320678711, "learning_rate": 2.128453123490781e-06, "loss": 0.0085, "step": 18230 }, { "epoch": 38.4, "grad_norm": 0.22153127193450928, "learning_rate": 2.1046538428847462e-06, "loss": 0.0059, "step": 18240 }, { "epoch": 38.421052631578945, "grad_norm": 0.2457754760980606, "learning_rate": 2.0809855063262273e-06, "loss": 0.0054, "step": 18250 }, { "epoch": 38.44210526315789, "grad_norm": 0.24883964657783508, "learning_rate": 2.057448178523558e-06, "loss": 0.0064, "step": 18260 }, { "epoch": 38.463157894736845, "grad_norm": 0.14141437411308289, "learning_rate": 2.034041923826885e-06, "loss": 0.0054, "step": 18270 }, { "epoch": 38.48421052631579, "grad_norm": 0.16457584500312805, "learning_rate": 2.0107668062280204e-06, "loss": 0.0068, "step": 18280 }, { "epoch": 38.50526315789474, "grad_norm": 0.15389791131019592, "learning_rate": 1.9876228893602357e-06, "loss": 0.0071, "step": 18290 }, { "epoch": 38.526315789473685, "grad_norm": 0.1827876716852188, "learning_rate": 1.9646102364981266e-06, "loss": 0.0059, "step": 18300 }, { "epoch": 38.54736842105263, "grad_norm": 0.1543724089860916, "learning_rate": 1.9417289105574053e-06, "loss": 0.0064, "step": 18310 }, { "epoch": 38.56842105263158, "grad_norm": 0.25029608607292175, "learning_rate": 1.9189789740947427e-06, "loss": 0.007, "step": 18320 }, { "epoch": 38.589473684210525, "grad_norm": 0.10255102813243866, "learning_rate": 1.896360489307597e-06, "loss": 0.005, "step": 18330 }, { "epoch": 38.61052631578947, "grad_norm": 0.1962905079126358, "learning_rate": 1.8738735180340362e-06, "loss": 0.0044, "step": 18340 }, { "epoch": 38.63157894736842, "grad_norm": 0.18482111394405365, "learning_rate": 1.8515181217525824e-06, "loss": 0.0048, "step": 18350 }, { "epoch": 38.65263157894737, "grad_norm": 0.1062772274017334, "learning_rate": 1.8292943615820457e-06, "loss": 0.0066, "step": 18360 }, { "epoch": 38.67368421052632, "grad_norm": 0.1845751851797104, "learning_rate": 1.8072022982813296e-06, "loss": 0.0072, "step": 18370 }, { "epoch": 38.694736842105264, "grad_norm": 0.1854289174079895, "learning_rate": 1.7852419922492925e-06, "loss": 0.005, "step": 18380 }, { "epoch": 38.71578947368421, "grad_norm": 0.10939481109380722, "learning_rate": 1.763413503524569e-06, "loss": 0.0063, "step": 18390 }, { "epoch": 38.73684210526316, "grad_norm": 0.18425236642360687, "learning_rate": 1.7417168917854165e-06, "loss": 0.0056, "step": 18400 }, { "epoch": 38.757894736842104, "grad_norm": 0.1699085831642151, "learning_rate": 1.720152216349552e-06, "loss": 0.0053, "step": 18410 }, { "epoch": 38.77894736842105, "grad_norm": 0.22808070480823517, "learning_rate": 1.6987195361739595e-06, "loss": 0.0056, "step": 18420 }, { "epoch": 38.8, "grad_norm": 0.1845967173576355, "learning_rate": 1.6774189098547832e-06, "loss": 0.0062, "step": 18430 }, { "epoch": 38.82105263157895, "grad_norm": 0.13927826285362244, "learning_rate": 1.6562503956271069e-06, "loss": 0.0045, "step": 18440 }, { "epoch": 38.8421052631579, "grad_norm": 0.40052419900894165, "learning_rate": 1.6352140513648417e-06, "loss": 0.0042, "step": 18450 }, { "epoch": 38.863157894736844, "grad_norm": 0.17272958159446716, "learning_rate": 1.6143099345805712e-06, "loss": 0.0061, "step": 18460 }, { "epoch": 38.88421052631579, "grad_norm": 0.13425906002521515, "learning_rate": 1.5935381024253293e-06, "loss": 0.0069, "step": 18470 }, { "epoch": 38.90526315789474, "grad_norm": 0.21603135764598846, "learning_rate": 1.572898611688517e-06, "loss": 0.0065, "step": 18480 }, { "epoch": 38.92631578947368, "grad_norm": 0.09027724713087082, "learning_rate": 1.5523915187977133e-06, "loss": 0.0062, "step": 18490 }, { "epoch": 38.94736842105263, "grad_norm": 0.12289512902498245, "learning_rate": 1.532016879818532e-06, "loss": 0.0064, "step": 18500 }, { "epoch": 38.96842105263158, "grad_norm": 0.19144630432128906, "learning_rate": 1.51177475045447e-06, "loss": 0.0051, "step": 18510 }, { "epoch": 38.98947368421052, "grad_norm": 0.24771396815776825, "learning_rate": 1.4916651860467035e-06, "loss": 0.0053, "step": 18520 }, { "epoch": 39.01052631578948, "grad_norm": 0.12960097193717957, "learning_rate": 1.471688241574043e-06, "loss": 0.0064, "step": 18530 }, { "epoch": 39.03157894736842, "grad_norm": 0.10911896824836731, "learning_rate": 1.451843971652672e-06, "loss": 0.0045, "step": 18540 }, { "epoch": 39.05263157894737, "grad_norm": 0.22111719846725464, "learning_rate": 1.432132430536076e-06, "loss": 0.0044, "step": 18550 }, { "epoch": 39.07368421052632, "grad_norm": 0.1767210215330124, "learning_rate": 1.412553672114869e-06, "loss": 0.0052, "step": 18560 }, { "epoch": 39.09473684210526, "grad_norm": 0.11228210479021072, "learning_rate": 1.3931077499166056e-06, "loss": 0.0054, "step": 18570 }, { "epoch": 39.11578947368421, "grad_norm": 0.34514155983924866, "learning_rate": 1.3737947171057085e-06, "loss": 0.0051, "step": 18580 }, { "epoch": 39.136842105263156, "grad_norm": 0.27800098061561584, "learning_rate": 1.3546146264832582e-06, "loss": 0.0064, "step": 18590 }, { "epoch": 39.1578947368421, "grad_norm": 0.2222786247730255, "learning_rate": 1.3355675304869086e-06, "loss": 0.007, "step": 18600 }, { "epoch": 39.17894736842105, "grad_norm": 0.18191252648830414, "learning_rate": 1.3166534811906827e-06, "loss": 0.0046, "step": 18610 }, { "epoch": 39.2, "grad_norm": 0.3747563064098358, "learning_rate": 1.2978725303048666e-06, "loss": 0.0056, "step": 18620 }, { "epoch": 39.22105263157895, "grad_norm": 0.09123487770557404, "learning_rate": 1.2792247291758762e-06, "loss": 0.0062, "step": 18630 }, { "epoch": 39.242105263157896, "grad_norm": 0.12069083005189896, "learning_rate": 1.2607101287860635e-06, "loss": 0.0063, "step": 18640 }, { "epoch": 39.26315789473684, "grad_norm": 0.15651407837867737, "learning_rate": 1.2423287797536654e-06, "loss": 0.0063, "step": 18650 }, { "epoch": 39.28421052631579, "grad_norm": 0.1645941585302353, "learning_rate": 1.2240807323325776e-06, "loss": 0.0066, "step": 18660 }, { "epoch": 39.305263157894736, "grad_norm": 0.11339884996414185, "learning_rate": 1.205966036412254e-06, "loss": 0.0083, "step": 18670 }, { "epoch": 39.32631578947368, "grad_norm": 0.17816181480884552, "learning_rate": 1.1879847415175949e-06, "loss": 0.0048, "step": 18680 }, { "epoch": 39.34736842105263, "grad_norm": 0.17298206686973572, "learning_rate": 1.1701368968087712e-06, "loss": 0.006, "step": 18690 }, { "epoch": 39.36842105263158, "grad_norm": 0.1495581567287445, "learning_rate": 1.1524225510811116e-06, "loss": 0.0042, "step": 18700 }, { "epoch": 39.38947368421053, "grad_norm": 0.2083953320980072, "learning_rate": 1.1348417527649535e-06, "loss": 0.0056, "step": 18710 }, { "epoch": 39.410526315789475, "grad_norm": 0.1353476494550705, "learning_rate": 1.1173945499255268e-06, "loss": 0.0073, "step": 18720 }, { "epoch": 39.43157894736842, "grad_norm": 0.2228955328464508, "learning_rate": 1.1000809902628307e-06, "loss": 0.0049, "step": 18730 }, { "epoch": 39.45263157894737, "grad_norm": 0.2307243049144745, "learning_rate": 1.082901121111468e-06, "loss": 0.0092, "step": 18740 }, { "epoch": 39.473684210526315, "grad_norm": 0.14222300052642822, "learning_rate": 1.0658549894405456e-06, "loss": 0.0054, "step": 18750 }, { "epoch": 39.49473684210526, "grad_norm": 0.31763318181037903, "learning_rate": 1.0489426418535342e-06, "loss": 0.006, "step": 18760 }, { "epoch": 39.51578947368421, "grad_norm": 0.19707821309566498, "learning_rate": 1.0321641245881474e-06, "loss": 0.0053, "step": 18770 }, { "epoch": 39.536842105263155, "grad_norm": 0.1618211269378662, "learning_rate": 1.015519483516214e-06, "loss": 0.0059, "step": 18780 }, { "epoch": 39.55789473684211, "grad_norm": 0.08212044090032578, "learning_rate": 9.990087641435443e-07, "loss": 0.0061, "step": 18790 }, { "epoch": 39.578947368421055, "grad_norm": 0.1942002922296524, "learning_rate": 9.826320116098132e-07, "loss": 0.0059, "step": 18800 }, { "epoch": 39.6, "grad_norm": 0.24003362655639648, "learning_rate": 9.663892706884447e-07, "loss": 0.005, "step": 18810 }, { "epoch": 39.62105263157895, "grad_norm": 0.07871801406145096, "learning_rate": 9.502805857864616e-07, "loss": 0.0061, "step": 18820 }, { "epoch": 39.642105263157895, "grad_norm": 0.176169291138649, "learning_rate": 9.34306000944396e-07, "loss": 0.0046, "step": 18830 }, { "epoch": 39.66315789473684, "grad_norm": 0.2820357382297516, "learning_rate": 9.184655598361624e-07, "loss": 0.0061, "step": 18840 }, { "epoch": 39.68421052631579, "grad_norm": 0.0815410390496254, "learning_rate": 9.027593057689076e-07, "loss": 0.0051, "step": 18850 }, { "epoch": 39.705263157894734, "grad_norm": 0.3055027723312378, "learning_rate": 8.871872816829441e-07, "loss": 0.0055, "step": 18860 }, { "epoch": 39.72631578947369, "grad_norm": 0.1715899109840393, "learning_rate": 8.717495301515777e-07, "loss": 0.007, "step": 18870 }, { "epoch": 39.747368421052634, "grad_norm": 0.19522921741008759, "learning_rate": 8.564460933810415e-07, "loss": 0.006, "step": 18880 }, { "epoch": 39.76842105263158, "grad_norm": 0.11580485105514526, "learning_rate": 8.412770132103453e-07, "loss": 0.0057, "step": 18890 }, { "epoch": 39.78947368421053, "grad_norm": 0.12208366394042969, "learning_rate": 8.262423311111711e-07, "loss": 0.0072, "step": 18900 }, { "epoch": 39.810526315789474, "grad_norm": 0.4677560329437256, "learning_rate": 8.113420881877665e-07, "loss": 0.008, "step": 18910 }, { "epoch": 39.83157894736842, "grad_norm": 0.1980770230293274, "learning_rate": 7.965763251768288e-07, "loss": 0.0049, "step": 18920 }, { "epoch": 39.85263157894737, "grad_norm": 0.21428321301937103, "learning_rate": 7.819450824473995e-07, "loss": 0.0071, "step": 18930 }, { "epoch": 39.873684210526314, "grad_norm": 0.18382854759693146, "learning_rate": 7.674484000007198e-07, "loss": 0.0048, "step": 18940 }, { "epoch": 39.89473684210526, "grad_norm": 0.1342938393354416, "learning_rate": 7.530863174701752e-07, "loss": 0.0053, "step": 18950 }, { "epoch": 39.915789473684214, "grad_norm": 0.23706510663032532, "learning_rate": 7.38858874121151e-07, "loss": 0.0053, "step": 18960 }, { "epoch": 39.93684210526316, "grad_norm": 0.24547156691551208, "learning_rate": 7.247661088509328e-07, "loss": 0.0054, "step": 18970 }, { "epoch": 39.95789473684211, "grad_norm": 0.11005083471536636, "learning_rate": 7.108080601886002e-07, "loss": 0.0052, "step": 18980 }, { "epoch": 39.97894736842105, "grad_norm": 0.10332196950912476, "learning_rate": 6.969847662949336e-07, "loss": 0.0081, "step": 18990 }, { "epoch": 40.0, "grad_norm": 0.17723864316940308, "learning_rate": 6.832962649622798e-07, "loss": 0.0059, "step": 19000 }, { "epoch": 40.02105263157895, "grad_norm": 0.14079649746418, "learning_rate": 6.697425936144863e-07, "loss": 0.0048, "step": 19010 }, { "epoch": 40.04210526315789, "grad_norm": 0.1651502400636673, "learning_rate": 6.563237893067731e-07, "loss": 0.0062, "step": 19020 }, { "epoch": 40.06315789473684, "grad_norm": 0.303705096244812, "learning_rate": 6.430398887256328e-07, "loss": 0.0057, "step": 19030 }, { "epoch": 40.084210526315786, "grad_norm": 0.11680895835161209, "learning_rate": 6.298909281887478e-07, "loss": 0.0055, "step": 19040 }, { "epoch": 40.10526315789474, "grad_norm": 0.15815600752830505, "learning_rate": 6.168769436448673e-07, "loss": 0.0058, "step": 19050 }, { "epoch": 40.126315789473686, "grad_norm": 0.1406935751438141, "learning_rate": 6.03997970673742e-07, "loss": 0.0072, "step": 19060 }, { "epoch": 40.14736842105263, "grad_norm": 0.18783733248710632, "learning_rate": 5.912540444859782e-07, "loss": 0.0072, "step": 19070 }, { "epoch": 40.16842105263158, "grad_norm": 0.13626421988010406, "learning_rate": 5.786451999229837e-07, "loss": 0.0058, "step": 19080 }, { "epoch": 40.189473684210526, "grad_norm": 0.29903435707092285, "learning_rate": 5.661714714568722e-07, "loss": 0.0058, "step": 19090 }, { "epoch": 40.21052631578947, "grad_norm": 0.20697127282619476, "learning_rate": 5.538328931903259e-07, "loss": 0.0059, "step": 19100 }, { "epoch": 40.23157894736842, "grad_norm": 0.16280770301818848, "learning_rate": 5.416294988565551e-07, "loss": 0.0054, "step": 19110 }, { "epoch": 40.252631578947366, "grad_norm": 0.1765798181295395, "learning_rate": 5.29561321819172e-07, "loss": 0.0051, "step": 19120 }, { "epoch": 40.27368421052632, "grad_norm": 0.1348101794719696, "learning_rate": 5.176283950721061e-07, "loss": 0.0057, "step": 19130 }, { "epoch": 40.294736842105266, "grad_norm": 0.1282510906457901, "learning_rate": 5.058307512395332e-07, "loss": 0.0051, "step": 19140 }, { "epoch": 40.31578947368421, "grad_norm": 0.5281218886375427, "learning_rate": 4.941684225757526e-07, "loss": 0.0056, "step": 19150 }, { "epoch": 40.33684210526316, "grad_norm": 0.15402410924434662, "learning_rate": 4.826414409651314e-07, "loss": 0.0044, "step": 19160 }, { "epoch": 40.357894736842105, "grad_norm": 0.09039716422557831, "learning_rate": 4.712498379219943e-07, "loss": 0.005, "step": 19170 }, { "epoch": 40.37894736842105, "grad_norm": 0.1483982503414154, "learning_rate": 4.599936445905506e-07, "loss": 0.0059, "step": 19180 }, { "epoch": 40.4, "grad_norm": 0.6021777987480164, "learning_rate": 4.4887289174480594e-07, "loss": 0.0071, "step": 19190 }, { "epoch": 40.421052631578945, "grad_norm": 0.282133013010025, "learning_rate": 4.378876097884621e-07, "loss": 0.005, "step": 19200 }, { "epoch": 40.44210526315789, "grad_norm": 0.27955183386802673, "learning_rate": 4.2703782875487264e-07, "loss": 0.0077, "step": 19210 }, { "epoch": 40.463157894736845, "grad_norm": 0.20672127604484558, "learning_rate": 4.163235783069208e-07, "loss": 0.0077, "step": 19220 }, { "epoch": 40.48421052631579, "grad_norm": 0.1930103302001953, "learning_rate": 4.057448877369585e-07, "loss": 0.0042, "step": 19230 }, { "epoch": 40.50526315789474, "grad_norm": 0.08194991201162338, "learning_rate": 3.9530178596672295e-07, "loss": 0.0051, "step": 19240 }, { "epoch": 40.526315789473685, "grad_norm": 0.23267321288585663, "learning_rate": 3.849943015472479e-07, "loss": 0.005, "step": 19250 }, { "epoch": 40.54736842105263, "grad_norm": 0.12487190216779709, "learning_rate": 3.748224626588137e-07, "loss": 0.0049, "step": 19260 }, { "epoch": 40.56842105263158, "grad_norm": 0.12444762140512466, "learning_rate": 3.647862971108307e-07, "loss": 0.0054, "step": 19270 }, { "epoch": 40.589473684210525, "grad_norm": 0.12371847778558731, "learning_rate": 3.5488583234179473e-07, "loss": 0.0059, "step": 19280 }, { "epoch": 40.61052631578947, "grad_norm": 0.16967153549194336, "learning_rate": 3.4512109541920413e-07, "loss": 0.0065, "step": 19290 }, { "epoch": 40.63157894736842, "grad_norm": 0.19704501330852509, "learning_rate": 3.354921130394706e-07, "loss": 0.0054, "step": 19300 }, { "epoch": 40.65263157894737, "grad_norm": 0.17244797945022583, "learning_rate": 3.259989115278639e-07, "loss": 0.0053, "step": 19310 }, { "epoch": 40.67368421052632, "grad_norm": 0.15910956263542175, "learning_rate": 3.1664151683843403e-07, "loss": 0.0067, "step": 19320 }, { "epoch": 40.694736842105264, "grad_norm": 0.33466586470603943, "learning_rate": 3.074199545539447e-07, "loss": 0.0058, "step": 19330 }, { "epoch": 40.71578947368421, "grad_norm": 0.11034300923347473, "learning_rate": 2.983342498857955e-07, "loss": 0.005, "step": 19340 }, { "epoch": 40.73684210526316, "grad_norm": 0.2962793707847595, "learning_rate": 2.893844276739499e-07, "loss": 0.0056, "step": 19350 }, { "epoch": 40.757894736842104, "grad_norm": 0.1948065608739853, "learning_rate": 2.8057051238688514e-07, "loss": 0.0058, "step": 19360 }, { "epoch": 40.77894736842105, "grad_norm": 0.1473235934972763, "learning_rate": 2.71892528121509e-07, "loss": 0.0055, "step": 19370 }, { "epoch": 40.8, "grad_norm": 0.47820261120796204, "learning_rate": 2.633504986030988e-07, "loss": 0.0051, "step": 19380 }, { "epoch": 40.82105263157895, "grad_norm": 0.07083901017904282, "learning_rate": 2.549444471852347e-07, "loss": 0.0073, "step": 19390 }, { "epoch": 40.8421052631579, "grad_norm": 0.19570326805114746, "learning_rate": 2.4667439684974423e-07, "loss": 0.0046, "step": 19400 }, { "epoch": 40.863157894736844, "grad_norm": 0.17362263798713684, "learning_rate": 2.3854037020662467e-07, "loss": 0.005, "step": 19410 }, { "epoch": 40.88421052631579, "grad_norm": 0.18181286752223969, "learning_rate": 2.3054238949399288e-07, "loss": 0.006, "step": 19420 }, { "epoch": 40.90526315789474, "grad_norm": 0.29303258657455444, "learning_rate": 2.2268047657802993e-07, "loss": 0.0058, "step": 19430 }, { "epoch": 40.92631578947368, "grad_norm": 0.30490928888320923, "learning_rate": 2.149546529529034e-07, "loss": 0.0057, "step": 19440 }, { "epoch": 40.94736842105263, "grad_norm": 0.11637499183416367, "learning_rate": 2.0736493974071736e-07, "loss": 0.005, "step": 19450 }, { "epoch": 40.96842105263158, "grad_norm": 0.23569583892822266, "learning_rate": 1.9991135769145686e-07, "loss": 0.0054, "step": 19460 }, { "epoch": 40.98947368421052, "grad_norm": 0.16460788249969482, "learning_rate": 1.9259392718293245e-07, "loss": 0.0057, "step": 19470 }, { "epoch": 41.01052631578948, "grad_norm": 0.07794702053070068, "learning_rate": 1.8541266822072467e-07, "loss": 0.0054, "step": 19480 }, { "epoch": 41.03157894736842, "grad_norm": 0.21711046993732452, "learning_rate": 1.7836760043811184e-07, "loss": 0.0061, "step": 19490 }, { "epoch": 41.05263157894737, "grad_norm": 0.16037379205226898, "learning_rate": 1.7145874309604792e-07, "loss": 0.0065, "step": 19500 }, { "epoch": 41.07368421052632, "grad_norm": 0.17098471522331238, "learning_rate": 1.6468611508308474e-07, "loss": 0.0059, "step": 19510 }, { "epoch": 41.09473684210526, "grad_norm": 0.09000364691019058, "learning_rate": 1.5804973491532204e-07, "loss": 0.0048, "step": 19520 }, { "epoch": 41.11578947368421, "grad_norm": 0.09484422951936722, "learning_rate": 1.5154962073637424e-07, "loss": 0.0039, "step": 19530 }, { "epoch": 41.136842105263156, "grad_norm": 0.23400329053401947, "learning_rate": 1.4518579031730372e-07, "loss": 0.0099, "step": 19540 }, { "epoch": 41.1578947368421, "grad_norm": 0.473676860332489, "learning_rate": 1.389582610565876e-07, "loss": 0.0054, "step": 19550 }, { "epoch": 41.17894736842105, "grad_norm": 0.1450461745262146, "learning_rate": 1.3286704998003995e-07, "loss": 0.0052, "step": 19560 }, { "epoch": 41.2, "grad_norm": 0.20966008305549622, "learning_rate": 1.2691217374080632e-07, "loss": 0.0067, "step": 19570 }, { "epoch": 41.22105263157895, "grad_norm": 0.08345336467027664, "learning_rate": 1.2109364861929705e-07, "loss": 0.0057, "step": 19580 }, { "epoch": 41.242105263157896, "grad_norm": 0.13087333738803864, "learning_rate": 1.1541149052312628e-07, "loss": 0.0052, "step": 19590 }, { "epoch": 41.26315789473684, "grad_norm": 0.11097513884305954, "learning_rate": 1.0986571498710074e-07, "loss": 0.0056, "step": 19600 }, { "epoch": 41.28421052631579, "grad_norm": 0.10127098858356476, "learning_rate": 1.0445633717316438e-07, "loss": 0.0045, "step": 19610 }, { "epoch": 41.305263157894736, "grad_norm": 0.15606486797332764, "learning_rate": 9.918337187034277e-08, "loss": 0.005, "step": 19620 }, { "epoch": 41.32631578947368, "grad_norm": 0.12066517770290375, "learning_rate": 9.404683349472643e-08, "loss": 0.0059, "step": 19630 }, { "epoch": 41.34736842105263, "grad_norm": 0.19478879868984222, "learning_rate": 8.904673608940983e-08, "loss": 0.0049, "step": 19640 }, { "epoch": 41.36842105263158, "grad_norm": 0.32228004932403564, "learning_rate": 8.418309332447471e-08, "loss": 0.0059, "step": 19650 }, { "epoch": 41.38947368421053, "grad_norm": 0.18889617919921875, "learning_rate": 7.945591849692902e-08, "loss": 0.0064, "step": 19660 }, { "epoch": 41.410526315789475, "grad_norm": 0.12168541550636292, "learning_rate": 7.486522453069578e-08, "loss": 0.0054, "step": 19670 }, { "epoch": 41.43157894736842, "grad_norm": 0.13789477944374084, "learning_rate": 7.041102397655208e-08, "loss": 0.0051, "step": 19680 }, { "epoch": 41.45263157894737, "grad_norm": 0.27351823449134827, "learning_rate": 6.609332901210685e-08, "loss": 0.0058, "step": 19690 }, { "epoch": 41.473684210526315, "grad_norm": 0.06374403834342957, "learning_rate": 6.191215144178419e-08, "loss": 0.0054, "step": 19700 }, { "epoch": 41.49473684210526, "grad_norm": 0.12469975650310516, "learning_rate": 5.786750269675678e-08, "loss": 0.005, "step": 19710 }, { "epoch": 41.51578947368421, "grad_norm": 0.1977238655090332, "learning_rate": 5.395939383494031e-08, "loss": 0.0062, "step": 19720 }, { "epoch": 41.536842105263155, "grad_norm": 0.14408008754253387, "learning_rate": 5.018783554095463e-08, "loss": 0.0044, "step": 19730 }, { "epoch": 41.55789473684211, "grad_norm": 0.23956133425235748, "learning_rate": 4.655283812610156e-08, "loss": 0.0055, "step": 19740 }, { "epoch": 41.578947368421055, "grad_norm": 0.3781638741493225, "learning_rate": 4.305441152831491e-08, "loss": 0.0068, "step": 19750 }, { "epoch": 41.6, "grad_norm": 0.19598640501499176, "learning_rate": 3.9692565312171584e-08, "loss": 0.0058, "step": 19760 }, { "epoch": 41.62105263157895, "grad_norm": 0.15229447185993195, "learning_rate": 3.6467308668824975e-08, "loss": 0.0065, "step": 19770 }, { "epoch": 41.642105263157895, "grad_norm": 0.2514556348323822, "learning_rate": 3.3378650416004964e-08, "loss": 0.0059, "step": 19780 }, { "epoch": 41.66315789473684, "grad_norm": 0.19993579387664795, "learning_rate": 3.042659899797906e-08, "loss": 0.0059, "step": 19790 }, { "epoch": 41.68421052631579, "grad_norm": 0.14011704921722412, "learning_rate": 2.76111624855524e-08, "loss": 0.006, "step": 19800 }, { "epoch": 41.705263157894734, "grad_norm": 0.08230623602867126, "learning_rate": 2.4932348576017784e-08, "loss": 0.0039, "step": 19810 }, { "epoch": 41.72631578947369, "grad_norm": 0.10957983881235123, "learning_rate": 2.239016459314458e-08, "loss": 0.0047, "step": 19820 }, { "epoch": 41.747368421052634, "grad_norm": 0.15484783053398132, "learning_rate": 1.9984617487173174e-08, "loss": 0.0048, "step": 19830 }, { "epoch": 41.76842105263158, "grad_norm": 0.18863263726234436, "learning_rate": 1.7715713834776105e-08, "loss": 0.0066, "step": 19840 }, { "epoch": 41.78947368421053, "grad_norm": 0.10231838375329971, "learning_rate": 1.5583459839046964e-08, "loss": 0.0059, "step": 19850 }, { "epoch": 41.810526315789474, "grad_norm": 0.14678262174129486, "learning_rate": 1.3587861329489304e-08, "loss": 0.0057, "step": 19860 }, { "epoch": 41.83157894736842, "grad_norm": 0.19383388757705688, "learning_rate": 1.1728923761994415e-08, "loss": 0.0044, "step": 19870 }, { "epoch": 41.85263157894737, "grad_norm": 0.14286305010318756, "learning_rate": 1.0006652218819135e-08, "loss": 0.0052, "step": 19880 }, { "epoch": 41.873684210526314, "grad_norm": 0.29257792234420776, "learning_rate": 8.421051408596947e-09, "loss": 0.0083, "step": 19890 }, { "epoch": 41.89473684210526, "grad_norm": 0.05898323282599449, "learning_rate": 6.972125666299123e-09, "loss": 0.0055, "step": 19900 }, { "epoch": 41.915789473684214, "grad_norm": 0.17864054441452026, "learning_rate": 5.659878953229169e-09, "loss": 0.0083, "step": 19910 }, { "epoch": 41.93684210526316, "grad_norm": 0.3817465305328369, "learning_rate": 4.48431485701728e-09, "loss": 0.0076, "step": 19920 }, { "epoch": 41.95789473684211, "grad_norm": 0.3704585134983063, "learning_rate": 3.4454365916203322e-09, "loss": 0.005, "step": 19930 }, { "epoch": 41.97894736842105, "grad_norm": 0.21493758261203766, "learning_rate": 2.5432469972830332e-09, "loss": 0.0054, "step": 19940 }, { "epoch": 42.0, "grad_norm": 0.21132296323776245, "learning_rate": 1.7777485405601203e-09, "loss": 0.0085, "step": 19950 }, { "epoch": 42.02105263157895, "grad_norm": 0.15939442813396454, "learning_rate": 1.1489433142941597e-09, "loss": 0.007, "step": 19960 }, { "epoch": 42.04210526315789, "grad_norm": 0.29529860615730286, "learning_rate": 6.568330376210963e-10, "loss": 0.0057, "step": 19970 }, { "epoch": 42.06315789473684, "grad_norm": 0.12582182884216309, "learning_rate": 3.0141905594249787e-10, "loss": 0.0054, "step": 19980 }, { "epoch": 42.084210526315786, "grad_norm": 0.2676975131034851, "learning_rate": 8.270234094776008e-11, "loss": 0.0067, "step": 19990 }, { "epoch": 42.10526315789474, "grad_norm": 0.136400505900383, "learning_rate": 6.834906085551041e-13, "loss": 0.0059, "step": 20000 } ], "logging_steps": 10, "max_steps": 20000, "num_input_tokens_seen": 0, "num_train_epochs": 43, "save_steps": 1000, "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": 36, "trial_name": null, "trial_params": null }