{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 24.390243902439025, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.024390243902439025, "grad_norm": 11.438058853149414, "learning_rate": 1.8e-06, "loss": 0.9244, "step": 10 }, { "epoch": 0.04878048780487805, "grad_norm": 3.859607458114624, "learning_rate": 3.8e-06, "loss": 0.7746, "step": 20 }, { "epoch": 0.07317073170731707, "grad_norm": 3.6816930770874023, "learning_rate": 5.8e-06, "loss": 0.3506, "step": 30 }, { "epoch": 0.0975609756097561, "grad_norm": 1.9363960027694702, "learning_rate": 7.8e-06, "loss": 0.2545, "step": 40 }, { "epoch": 0.12195121951219512, "grad_norm": 1.725611925125122, "learning_rate": 9.800000000000001e-06, "loss": 0.2034, "step": 50 }, { "epoch": 0.14634146341463414, "grad_norm": 1.262165904045105, "learning_rate": 1.18e-05, "loss": 0.1751, "step": 60 }, { "epoch": 0.17073170731707318, "grad_norm": 1.2210197448730469, "learning_rate": 1.3800000000000002e-05, "loss": 0.1427, "step": 70 }, { "epoch": 0.1951219512195122, "grad_norm": 1.128469467163086, "learning_rate": 1.58e-05, "loss": 0.1394, "step": 80 }, { "epoch": 0.21951219512195122, "grad_norm": 1.1902098655700684, "learning_rate": 1.78e-05, "loss": 0.1293, "step": 90 }, { "epoch": 0.24390243902439024, "grad_norm": 1.2753090858459473, "learning_rate": 1.9800000000000004e-05, "loss": 0.1307, "step": 100 }, { "epoch": 0.2682926829268293, "grad_norm": 1.369084119796753, "learning_rate": 2.18e-05, "loss": 0.1234, "step": 110 }, { "epoch": 0.2926829268292683, "grad_norm": 0.7994917631149292, "learning_rate": 2.38e-05, "loss": 0.1044, "step": 120 }, { "epoch": 0.3170731707317073, "grad_norm": 1.5699591636657715, "learning_rate": 2.58e-05, "loss": 0.1128, "step": 130 }, { "epoch": 0.34146341463414637, "grad_norm": 1.277710199356079, "learning_rate": 2.7800000000000005e-05, "loss": 0.1047, "step": 140 }, { "epoch": 0.36585365853658536, "grad_norm": 1.2865667343139648, "learning_rate": 2.98e-05, "loss": 0.104, "step": 150 }, { "epoch": 0.3902439024390244, "grad_norm": 1.2443100214004517, "learning_rate": 3.18e-05, "loss": 0.1058, "step": 160 }, { "epoch": 0.4146341463414634, "grad_norm": 1.0365393161773682, "learning_rate": 3.38e-05, "loss": 0.0979, "step": 170 }, { "epoch": 0.43902439024390244, "grad_norm": 0.96555095911026, "learning_rate": 3.58e-05, "loss": 0.0973, "step": 180 }, { "epoch": 0.4634146341463415, "grad_norm": 1.4485164880752563, "learning_rate": 3.7800000000000004e-05, "loss": 0.0901, "step": 190 }, { "epoch": 0.4878048780487805, "grad_norm": 1.753981590270996, "learning_rate": 3.9800000000000005e-05, "loss": 0.109, "step": 200 }, { "epoch": 0.5121951219512195, "grad_norm": 1.2610169649124146, "learning_rate": 4.18e-05, "loss": 0.0927, "step": 210 }, { "epoch": 0.5365853658536586, "grad_norm": 0.8480499982833862, "learning_rate": 4.38e-05, "loss": 0.0938, "step": 220 }, { "epoch": 0.5609756097560976, "grad_norm": 0.9105013608932495, "learning_rate": 4.58e-05, "loss": 0.0856, "step": 230 }, { "epoch": 0.5853658536585366, "grad_norm": 0.8315329551696777, "learning_rate": 4.78e-05, "loss": 0.0808, "step": 240 }, { "epoch": 0.6097560975609756, "grad_norm": 0.90700763463974, "learning_rate": 4.9800000000000004e-05, "loss": 0.0768, "step": 250 }, { "epoch": 0.6341463414634146, "grad_norm": 0.830387532711029, "learning_rate": 5.1800000000000005e-05, "loss": 0.0755, "step": 260 }, { "epoch": 0.6585365853658537, "grad_norm": 0.694616436958313, "learning_rate": 5.380000000000001e-05, "loss": 0.0833, "step": 270 }, { "epoch": 0.6829268292682927, "grad_norm": 1.2810885906219482, "learning_rate": 5.580000000000001e-05, "loss": 0.0706, "step": 280 }, { "epoch": 0.7073170731707317, "grad_norm": 0.8029444217681885, "learning_rate": 5.7799999999999995e-05, "loss": 0.0785, "step": 290 }, { "epoch": 0.7317073170731707, "grad_norm": 0.8832963705062866, "learning_rate": 5.9800000000000003e-05, "loss": 0.0734, "step": 300 }, { "epoch": 0.7560975609756098, "grad_norm": 0.8371597528457642, "learning_rate": 6.18e-05, "loss": 0.0777, "step": 310 }, { "epoch": 0.7804878048780488, "grad_norm": 1.166608214378357, "learning_rate": 6.38e-05, "loss": 0.0735, "step": 320 }, { "epoch": 0.8048780487804879, "grad_norm": 1.1829246282577515, "learning_rate": 6.58e-05, "loss": 0.0711, "step": 330 }, { "epoch": 0.8292682926829268, "grad_norm": 1.5263547897338867, "learning_rate": 6.780000000000001e-05, "loss": 0.0668, "step": 340 }, { "epoch": 0.8536585365853658, "grad_norm": 1.212733507156372, "learning_rate": 6.98e-05, "loss": 0.0609, "step": 350 }, { "epoch": 0.8780487804878049, "grad_norm": 0.7210371494293213, "learning_rate": 7.18e-05, "loss": 0.0606, "step": 360 }, { "epoch": 0.9024390243902439, "grad_norm": 0.7223049998283386, "learning_rate": 7.38e-05, "loss": 0.0589, "step": 370 }, { "epoch": 0.926829268292683, "grad_norm": 0.660465657711029, "learning_rate": 7.58e-05, "loss": 0.0635, "step": 380 }, { "epoch": 0.9512195121951219, "grad_norm": 0.7829694151878357, "learning_rate": 7.780000000000001e-05, "loss": 0.0627, "step": 390 }, { "epoch": 0.975609756097561, "grad_norm": 0.7635610103607178, "learning_rate": 7.98e-05, "loss": 0.0598, "step": 400 }, { "epoch": 1.0, "grad_norm": 0.7895731925964355, "learning_rate": 8.18e-05, "loss": 0.0659, "step": 410 }, { "epoch": 1.024390243902439, "grad_norm": 0.7271248698234558, "learning_rate": 8.38e-05, "loss": 0.0631, "step": 420 }, { "epoch": 1.048780487804878, "grad_norm": 0.9169553518295288, "learning_rate": 8.58e-05, "loss": 0.0623, "step": 430 }, { "epoch": 1.0731707317073171, "grad_norm": 0.9604327082633972, "learning_rate": 8.78e-05, "loss": 0.0599, "step": 440 }, { "epoch": 1.0975609756097562, "grad_norm": 1.0450021028518677, "learning_rate": 8.98e-05, "loss": 0.0661, "step": 450 }, { "epoch": 1.1219512195121952, "grad_norm": 0.8175938725471497, "learning_rate": 9.180000000000001e-05, "loss": 0.064, "step": 460 }, { "epoch": 1.146341463414634, "grad_norm": 0.8619034886360168, "learning_rate": 9.38e-05, "loss": 0.0661, "step": 470 }, { "epoch": 1.170731707317073, "grad_norm": 0.5424306988716125, "learning_rate": 9.58e-05, "loss": 0.061, "step": 480 }, { "epoch": 1.1951219512195121, "grad_norm": 0.6567324995994568, "learning_rate": 9.78e-05, "loss": 0.0509, "step": 490 }, { "epoch": 1.2195121951219512, "grad_norm": 0.6517794728279114, "learning_rate": 9.98e-05, "loss": 0.0547, "step": 500 }, { "epoch": 1.2439024390243902, "grad_norm": 0.6785968542098999, "learning_rate": 9.9999778549206e-05, "loss": 0.0568, "step": 510 }, { "epoch": 1.2682926829268293, "grad_norm": 0.8715964555740356, "learning_rate": 9.999901304280685e-05, "loss": 0.0543, "step": 520 }, { "epoch": 1.2926829268292683, "grad_norm": 0.7329784035682678, "learning_rate": 9.999770075521164e-05, "loss": 0.0599, "step": 530 }, { "epoch": 1.3170731707317074, "grad_norm": 0.5132158398628235, "learning_rate": 9.99958417007713e-05, "loss": 0.049, "step": 540 }, { "epoch": 1.3414634146341464, "grad_norm": 0.5929908156394958, "learning_rate": 9.999343589981615e-05, "loss": 0.0468, "step": 550 }, { "epoch": 1.3658536585365852, "grad_norm": 0.5250375270843506, "learning_rate": 9.999048337865568e-05, "loss": 0.0502, "step": 560 }, { "epoch": 1.3902439024390243, "grad_norm": 0.5170222520828247, "learning_rate": 9.998698416957815e-05, "loss": 0.0508, "step": 570 }, { "epoch": 1.4146341463414633, "grad_norm": 0.5049208998680115, "learning_rate": 9.998293831085037e-05, "loss": 0.0479, "step": 580 }, { "epoch": 1.4390243902439024, "grad_norm": 0.6433089971542358, "learning_rate": 9.997834584671719e-05, "loss": 0.0477, "step": 590 }, { "epoch": 1.4634146341463414, "grad_norm": 0.4838939309120178, "learning_rate": 9.997320682740107e-05, "loss": 0.0528, "step": 600 }, { "epoch": 1.4878048780487805, "grad_norm": 0.49200132489204407, "learning_rate": 9.996752130910149e-05, "loss": 0.0544, "step": 610 }, { "epoch": 1.5121951219512195, "grad_norm": 0.6802639961242676, "learning_rate": 9.99612893539944e-05, "loss": 0.0547, "step": 620 }, { "epoch": 1.5365853658536586, "grad_norm": 0.7216886281967163, "learning_rate": 9.995451103023144e-05, "loss": 0.0454, "step": 630 }, { "epoch": 1.5609756097560976, "grad_norm": 0.45784440636634827, "learning_rate": 9.994718641193928e-05, "loss": 0.0447, "step": 640 }, { "epoch": 1.5853658536585367, "grad_norm": 0.6383759379386902, "learning_rate": 9.993931557921874e-05, "loss": 0.0511, "step": 650 }, { "epoch": 1.6097560975609757, "grad_norm": 0.4707896411418915, "learning_rate": 9.993089861814402e-05, "loss": 0.0504, "step": 660 }, { "epoch": 1.6341463414634148, "grad_norm": 0.6331726312637329, "learning_rate": 9.992193562076166e-05, "loss": 0.0604, "step": 670 }, { "epoch": 1.6585365853658538, "grad_norm": 0.421948105096817, "learning_rate": 9.991242668508954e-05, "loss": 0.0495, "step": 680 }, { "epoch": 1.6829268292682928, "grad_norm": 0.5101801753044128, "learning_rate": 9.990237191511587e-05, "loss": 0.0472, "step": 690 }, { "epoch": 1.7073170731707317, "grad_norm": 0.45487090945243835, "learning_rate": 9.989177142079802e-05, "loss": 0.049, "step": 700 }, { "epoch": 1.7317073170731707, "grad_norm": 0.6539069414138794, "learning_rate": 9.988062531806126e-05, "loss": 0.0526, "step": 710 }, { "epoch": 1.7560975609756098, "grad_norm": 0.7236000895500183, "learning_rate": 9.986893372879762e-05, "loss": 0.0446, "step": 720 }, { "epoch": 1.7804878048780488, "grad_norm": 0.6140279769897461, "learning_rate": 9.985669678086443e-05, "loss": 0.0469, "step": 730 }, { "epoch": 1.8048780487804879, "grad_norm": 0.8803598880767822, "learning_rate": 9.984391460808298e-05, "loss": 0.041, "step": 740 }, { "epoch": 1.8292682926829267, "grad_norm": 0.5929242968559265, "learning_rate": 9.983058735023709e-05, "loss": 0.0422, "step": 750 }, { "epoch": 1.8536585365853657, "grad_norm": 0.6037493944168091, "learning_rate": 9.98167151530715e-05, "loss": 0.0381, "step": 760 }, { "epoch": 1.8780487804878048, "grad_norm": 0.5405340194702148, "learning_rate": 9.980229816829034e-05, "loss": 0.0447, "step": 770 }, { "epoch": 1.9024390243902438, "grad_norm": 0.519241988658905, "learning_rate": 9.978733655355544e-05, "loss": 0.0446, "step": 780 }, { "epoch": 1.9268292682926829, "grad_norm": 0.5540395975112915, "learning_rate": 9.977183047248464e-05, "loss": 0.0396, "step": 790 }, { "epoch": 1.951219512195122, "grad_norm": 0.3942469656467438, "learning_rate": 9.975578009464992e-05, "loss": 0.0387, "step": 800 }, { "epoch": 1.975609756097561, "grad_norm": 0.44967520236968994, "learning_rate": 9.97391855955757e-05, "loss": 0.0469, "step": 810 }, { "epoch": 2.0, "grad_norm": 0.972531259059906, "learning_rate": 9.972204715673669e-05, "loss": 0.0448, "step": 820 }, { "epoch": 2.024390243902439, "grad_norm": 0.5751144886016846, "learning_rate": 9.970436496555617e-05, "loss": 0.0436, "step": 830 }, { "epoch": 2.048780487804878, "grad_norm": 0.6377841234207153, "learning_rate": 9.968613921540373e-05, "loss": 0.0451, "step": 840 }, { "epoch": 2.073170731707317, "grad_norm": 0.4763578474521637, "learning_rate": 9.966737010559326e-05, "loss": 0.0532, "step": 850 }, { "epoch": 2.097560975609756, "grad_norm": 0.4867108464241028, "learning_rate": 9.964805784138072e-05, "loss": 0.041, "step": 860 }, { "epoch": 2.1219512195121952, "grad_norm": 0.5727500915527344, "learning_rate": 9.962820263396195e-05, "loss": 0.0454, "step": 870 }, { "epoch": 2.1463414634146343, "grad_norm": 0.5406850576400757, "learning_rate": 9.960780470047033e-05, "loss": 0.0396, "step": 880 }, { "epoch": 2.1707317073170733, "grad_norm": 0.4212709963321686, "learning_rate": 9.958686426397437e-05, "loss": 0.0386, "step": 890 }, { "epoch": 2.1951219512195124, "grad_norm": 0.44725343585014343, "learning_rate": 9.956538155347534e-05, "loss": 0.0432, "step": 900 }, { "epoch": 2.2195121951219514, "grad_norm": 0.38483870029449463, "learning_rate": 9.95433568039047e-05, "loss": 0.0383, "step": 910 }, { "epoch": 2.2439024390243905, "grad_norm": 0.4076084792613983, "learning_rate": 9.952079025612162e-05, "loss": 0.0364, "step": 920 }, { "epoch": 2.2682926829268295, "grad_norm": 0.39651069045066833, "learning_rate": 9.949768215691022e-05, "loss": 0.0404, "step": 930 }, { "epoch": 2.292682926829268, "grad_norm": 0.6808133125305176, "learning_rate": 9.9474032758977e-05, "loss": 0.0353, "step": 940 }, { "epoch": 2.317073170731707, "grad_norm": 0.46602684259414673, "learning_rate": 9.944984232094794e-05, "loss": 0.0456, "step": 950 }, { "epoch": 2.341463414634146, "grad_norm": 0.5331060290336609, "learning_rate": 9.942511110736584e-05, "loss": 0.0392, "step": 960 }, { "epoch": 2.3658536585365852, "grad_norm": 0.5745082497596741, "learning_rate": 9.939983938868726e-05, "loss": 0.0373, "step": 970 }, { "epoch": 2.3902439024390243, "grad_norm": 0.33656513690948486, "learning_rate": 9.93740274412797e-05, "loss": 0.035, "step": 980 }, { "epoch": 2.4146341463414633, "grad_norm": 0.44869112968444824, "learning_rate": 9.934767554741846e-05, "loss": 0.0394, "step": 990 }, { "epoch": 2.4390243902439024, "grad_norm": 0.39117568731307983, "learning_rate": 9.932078399528361e-05, "loss": 0.0373, "step": 1000 }, { "epoch": 2.4634146341463414, "grad_norm": 0.5011258125305176, "learning_rate": 9.929335307895689e-05, "loss": 0.0402, "step": 1010 }, { "epoch": 2.4878048780487805, "grad_norm": 0.5513408780097961, "learning_rate": 9.926538309841839e-05, "loss": 0.0357, "step": 1020 }, { "epoch": 2.5121951219512195, "grad_norm": 0.38576608896255493, "learning_rate": 9.923687435954334e-05, "loss": 0.0346, "step": 1030 }, { "epoch": 2.5365853658536586, "grad_norm": 0.5330151915550232, "learning_rate": 9.920782717409873e-05, "loss": 0.0428, "step": 1040 }, { "epoch": 2.5609756097560976, "grad_norm": 0.5135587453842163, "learning_rate": 9.917824185973994e-05, "loss": 0.0352, "step": 1050 }, { "epoch": 2.5853658536585367, "grad_norm": 0.5289296507835388, "learning_rate": 9.914811874000723e-05, "loss": 0.0446, "step": 1060 }, { "epoch": 2.6097560975609757, "grad_norm": 0.4024636149406433, "learning_rate": 9.911745814432218e-05, "loss": 0.0434, "step": 1070 }, { "epoch": 2.6341463414634148, "grad_norm": 0.3980396091938019, "learning_rate": 9.90862604079842e-05, "loss": 0.0429, "step": 1080 }, { "epoch": 2.658536585365854, "grad_norm": 0.554757297039032, "learning_rate": 9.90545258721667e-05, "loss": 0.0325, "step": 1090 }, { "epoch": 2.682926829268293, "grad_norm": 0.37143391370773315, "learning_rate": 9.90222548839135e-05, "loss": 0.0345, "step": 1100 }, { "epoch": 2.7073170731707314, "grad_norm": 0.48759201169013977, "learning_rate": 9.898944779613495e-05, "loss": 0.0374, "step": 1110 }, { "epoch": 2.7317073170731705, "grad_norm": 0.4527494013309479, "learning_rate": 9.89561049676041e-05, "loss": 0.0395, "step": 1120 }, { "epoch": 2.7560975609756095, "grad_norm": 0.3580166697502136, "learning_rate": 9.89222267629528e-05, "loss": 0.0462, "step": 1130 }, { "epoch": 2.7804878048780486, "grad_norm": 0.563718318939209, "learning_rate": 9.888781355266763e-05, "loss": 0.0415, "step": 1140 }, { "epoch": 2.8048780487804876, "grad_norm": 0.3706037402153015, "learning_rate": 9.885286571308598e-05, "loss": 0.0472, "step": 1150 }, { "epoch": 2.8292682926829267, "grad_norm": 0.430159330368042, "learning_rate": 9.881738362639182e-05, "loss": 0.0414, "step": 1160 }, { "epoch": 2.8536585365853657, "grad_norm": 0.45605069398880005, "learning_rate": 9.878136768061154e-05, "loss": 0.0365, "step": 1170 }, { "epoch": 2.8780487804878048, "grad_norm": 0.40683606266975403, "learning_rate": 9.874481826960979e-05, "loss": 0.0422, "step": 1180 }, { "epoch": 2.902439024390244, "grad_norm": 0.4990503787994385, "learning_rate": 9.870773579308503e-05, "loss": 0.0379, "step": 1190 }, { "epoch": 2.926829268292683, "grad_norm": 0.3867078721523285, "learning_rate": 9.867012065656533e-05, "loss": 0.0432, "step": 1200 }, { "epoch": 2.951219512195122, "grad_norm": 0.40396395325660706, "learning_rate": 9.863197327140376e-05, "loss": 0.0353, "step": 1210 }, { "epoch": 2.975609756097561, "grad_norm": 0.3166995942592621, "learning_rate": 9.859329405477403e-05, "loss": 0.035, "step": 1220 }, { "epoch": 3.0, "grad_norm": 0.5780623555183411, "learning_rate": 9.855408342966585e-05, "loss": 0.0382, "step": 1230 }, { "epoch": 3.024390243902439, "grad_norm": 0.4846343398094177, "learning_rate": 9.851434182488033e-05, "loss": 0.034, "step": 1240 }, { "epoch": 3.048780487804878, "grad_norm": 0.3104378879070282, "learning_rate": 9.84740696750253e-05, "loss": 0.0346, "step": 1250 }, { "epoch": 3.073170731707317, "grad_norm": 0.49481409788131714, "learning_rate": 9.843326742051055e-05, "loss": 0.0408, "step": 1260 }, { "epoch": 3.097560975609756, "grad_norm": 0.6306561827659607, "learning_rate": 9.839193550754297e-05, "loss": 0.0393, "step": 1270 }, { "epoch": 3.1219512195121952, "grad_norm": 0.3917936682701111, "learning_rate": 9.835007438812177e-05, "loss": 0.0367, "step": 1280 }, { "epoch": 3.1463414634146343, "grad_norm": 0.34476983547210693, "learning_rate": 9.830768452003341e-05, "loss": 0.0388, "step": 1290 }, { "epoch": 3.1707317073170733, "grad_norm": 0.5854437351226807, "learning_rate": 9.826476636684671e-05, "loss": 0.0432, "step": 1300 }, { "epoch": 3.1951219512195124, "grad_norm": 0.5620570182800293, "learning_rate": 9.822132039790773e-05, "loss": 0.0356, "step": 1310 }, { "epoch": 3.2195121951219514, "grad_norm": 0.32234442234039307, "learning_rate": 9.817734708833461e-05, "loss": 0.0375, "step": 1320 }, { "epoch": 3.2439024390243905, "grad_norm": 0.4901784360408783, "learning_rate": 9.813284691901243e-05, "loss": 0.0396, "step": 1330 }, { "epoch": 3.2682926829268295, "grad_norm": 0.4899320900440216, "learning_rate": 9.808782037658792e-05, "loss": 0.0331, "step": 1340 }, { "epoch": 3.292682926829268, "grad_norm": 0.3954959213733673, "learning_rate": 9.804226795346411e-05, "loss": 0.0365, "step": 1350 }, { "epoch": 3.317073170731707, "grad_norm": 0.43908506631851196, "learning_rate": 9.799619014779503e-05, "loss": 0.0334, "step": 1360 }, { "epoch": 3.341463414634146, "grad_norm": 0.5613491535186768, "learning_rate": 9.794958746348013e-05, "loss": 0.0285, "step": 1370 }, { "epoch": 3.3658536585365852, "grad_norm": 0.7104533910751343, "learning_rate": 9.790246041015896e-05, "loss": 0.0348, "step": 1380 }, { "epoch": 3.3902439024390243, "grad_norm": 0.46149489283561707, "learning_rate": 9.785480950320538e-05, "loss": 0.034, "step": 1390 }, { "epoch": 3.4146341463414633, "grad_norm": 0.5235434770584106, "learning_rate": 9.78066352637221e-05, "loss": 0.0365, "step": 1400 }, { "epoch": 3.4390243902439024, "grad_norm": 0.46233052015304565, "learning_rate": 9.775793821853488e-05, "loss": 0.0334, "step": 1410 }, { "epoch": 3.4634146341463414, "grad_norm": 0.32053062319755554, "learning_rate": 9.77087189001868e-05, "loss": 0.033, "step": 1420 }, { "epoch": 3.4878048780487805, "grad_norm": 0.5225449204444885, "learning_rate": 9.765897784693243e-05, "loss": 0.036, "step": 1430 }, { "epoch": 3.5121951219512195, "grad_norm": 0.5758381485939026, "learning_rate": 9.760871560273197e-05, "loss": 0.0393, "step": 1440 }, { "epoch": 3.5365853658536586, "grad_norm": 0.3062208890914917, "learning_rate": 9.755793271724526e-05, "loss": 0.0361, "step": 1450 }, { "epoch": 3.5609756097560976, "grad_norm": 0.5471229553222656, "learning_rate": 9.750662974582584e-05, "loss": 0.0345, "step": 1460 }, { "epoch": 3.5853658536585367, "grad_norm": 0.5146205425262451, "learning_rate": 9.745480724951473e-05, "loss": 0.0325, "step": 1470 }, { "epoch": 3.6097560975609757, "grad_norm": 0.40823036432266235, "learning_rate": 9.740246579503447e-05, "loss": 0.04, "step": 1480 }, { "epoch": 3.6341463414634148, "grad_norm": 0.36182716488838196, "learning_rate": 9.734960595478284e-05, "loss": 0.0293, "step": 1490 }, { "epoch": 3.658536585365854, "grad_norm": 0.5093255639076233, "learning_rate": 9.729622830682657e-05, "loss": 0.0335, "step": 1500 }, { "epoch": 3.682926829268293, "grad_norm": 0.47144684195518494, "learning_rate": 9.724233343489504e-05, "loss": 0.037, "step": 1510 }, { "epoch": 3.7073170731707314, "grad_norm": 0.350436806678772, "learning_rate": 9.718792192837396e-05, "loss": 0.0324, "step": 1520 }, { "epoch": 3.7317073170731705, "grad_norm": 0.3493197560310364, "learning_rate": 9.713299438229886e-05, "loss": 0.0368, "step": 1530 }, { "epoch": 3.7560975609756095, "grad_norm": 0.3492034673690796, "learning_rate": 9.707755139734855e-05, "loss": 0.0359, "step": 1540 }, { "epoch": 3.7804878048780486, "grad_norm": 0.21749389171600342, "learning_rate": 9.702159357983866e-05, "loss": 0.0378, "step": 1550 }, { "epoch": 3.8048780487804876, "grad_norm": 0.3634694814682007, "learning_rate": 9.696512154171492e-05, "loss": 0.0335, "step": 1560 }, { "epoch": 3.8292682926829267, "grad_norm": 0.37129950523376465, "learning_rate": 9.690813590054645e-05, "loss": 0.0321, "step": 1570 }, { "epoch": 3.8536585365853657, "grad_norm": 0.39070361852645874, "learning_rate": 9.685063727951914e-05, "loss": 0.0321, "step": 1580 }, { "epoch": 3.8780487804878048, "grad_norm": 0.3660708963871002, "learning_rate": 9.679262630742865e-05, "loss": 0.0357, "step": 1590 }, { "epoch": 3.902439024390244, "grad_norm": 0.5420876145362854, "learning_rate": 9.673410361867373e-05, "loss": 0.0301, "step": 1600 }, { "epoch": 3.926829268292683, "grad_norm": 0.7411645650863647, "learning_rate": 9.667506985324909e-05, "loss": 0.0323, "step": 1610 }, { "epoch": 3.951219512195122, "grad_norm": 0.47086405754089355, "learning_rate": 9.661552565673855e-05, "loss": 0.0312, "step": 1620 }, { "epoch": 3.975609756097561, "grad_norm": 0.6783720850944519, "learning_rate": 9.655547168030789e-05, "loss": 0.0359, "step": 1630 }, { "epoch": 4.0, "grad_norm": 0.5262812972068787, "learning_rate": 9.649490858069777e-05, "loss": 0.0294, "step": 1640 }, { "epoch": 4.024390243902439, "grad_norm": 0.2980291247367859, "learning_rate": 9.643383702021658e-05, "loss": 0.0384, "step": 1650 }, { "epoch": 4.048780487804878, "grad_norm": 0.2296736240386963, "learning_rate": 9.637225766673307e-05, "loss": 0.0343, "step": 1660 }, { "epoch": 4.073170731707317, "grad_norm": 0.3653576374053955, "learning_rate": 9.631017119366922e-05, "loss": 0.0324, "step": 1670 }, { "epoch": 4.097560975609756, "grad_norm": 0.2953051030635834, "learning_rate": 9.624757827999273e-05, "loss": 0.0332, "step": 1680 }, { "epoch": 4.121951219512195, "grad_norm": 0.34009554982185364, "learning_rate": 9.618447961020971e-05, "loss": 0.0355, "step": 1690 }, { "epoch": 4.146341463414634, "grad_norm": 0.5569214820861816, "learning_rate": 9.612087587435707e-05, "loss": 0.0321, "step": 1700 }, { "epoch": 4.170731707317073, "grad_norm": 0.6192038655281067, "learning_rate": 9.605676776799508e-05, "loss": 0.0344, "step": 1710 }, { "epoch": 4.195121951219512, "grad_norm": 0.4145585000514984, "learning_rate": 9.599215599219973e-05, "loss": 0.0332, "step": 1720 }, { "epoch": 4.219512195121951, "grad_norm": 0.46267589926719666, "learning_rate": 9.592704125355505e-05, "loss": 0.0325, "step": 1730 }, { "epoch": 4.2439024390243905, "grad_norm": 0.3412143290042877, "learning_rate": 9.586142426414538e-05, "loss": 0.0327, "step": 1740 }, { "epoch": 4.2682926829268295, "grad_norm": 0.29737308621406555, "learning_rate": 9.57953057415476e-05, "loss": 0.0333, "step": 1750 }, { "epoch": 4.2926829268292686, "grad_norm": 0.3923538625240326, "learning_rate": 9.572868640882328e-05, "loss": 0.0378, "step": 1760 }, { "epoch": 4.317073170731708, "grad_norm": 0.5407876372337341, "learning_rate": 9.56615669945108e-05, "loss": 0.0319, "step": 1770 }, { "epoch": 4.341463414634147, "grad_norm": 0.38840898871421814, "learning_rate": 9.55939482326173e-05, "loss": 0.0369, "step": 1780 }, { "epoch": 4.365853658536586, "grad_norm": 0.4214266240596771, "learning_rate": 9.552583086261069e-05, "loss": 0.0343, "step": 1790 }, { "epoch": 4.390243902439025, "grad_norm": 0.3901469111442566, "learning_rate": 9.545721562941168e-05, "loss": 0.0372, "step": 1800 }, { "epoch": 4.414634146341464, "grad_norm": 0.49115118384361267, "learning_rate": 9.538810328338543e-05, "loss": 0.0299, "step": 1810 }, { "epoch": 4.439024390243903, "grad_norm": 0.3093148469924927, "learning_rate": 9.531849458033349e-05, "loss": 0.0335, "step": 1820 }, { "epoch": 4.463414634146342, "grad_norm": 0.32012274861335754, "learning_rate": 9.524839028148547e-05, "loss": 0.0309, "step": 1830 }, { "epoch": 4.487804878048781, "grad_norm": 0.40676629543304443, "learning_rate": 9.517779115349077e-05, "loss": 0.0298, "step": 1840 }, { "epoch": 4.512195121951219, "grad_norm": 0.4764292538166046, "learning_rate": 9.510669796841014e-05, "loss": 0.0353, "step": 1850 }, { "epoch": 4.536585365853659, "grad_norm": 0.5055551528930664, "learning_rate": 9.503511150370727e-05, "loss": 0.0406, "step": 1860 }, { "epoch": 4.560975609756097, "grad_norm": 0.3942221403121948, "learning_rate": 9.496303254224024e-05, "loss": 0.0335, "step": 1870 }, { "epoch": 4.585365853658536, "grad_norm": 0.4020695686340332, "learning_rate": 9.489046187225306e-05, "loss": 0.0361, "step": 1880 }, { "epoch": 4.609756097560975, "grad_norm": 0.6468827128410339, "learning_rate": 9.481740028736692e-05, "loss": 0.0312, "step": 1890 }, { "epoch": 4.634146341463414, "grad_norm": 0.48698127269744873, "learning_rate": 9.474384858657164e-05, "loss": 0.0354, "step": 1900 }, { "epoch": 4.658536585365853, "grad_norm": 0.48748353123664856, "learning_rate": 9.466980757421679e-05, "loss": 0.0363, "step": 1910 }, { "epoch": 4.682926829268292, "grad_norm": 0.3960422873497009, "learning_rate": 9.459527806000305e-05, "loss": 0.0313, "step": 1920 }, { "epoch": 4.7073170731707314, "grad_norm": 0.2521972358226776, "learning_rate": 9.452026085897325e-05, "loss": 0.0315, "step": 1930 }, { "epoch": 4.7317073170731705, "grad_norm": 0.31679585576057434, "learning_rate": 9.444475679150348e-05, "loss": 0.0312, "step": 1940 }, { "epoch": 4.7560975609756095, "grad_norm": 0.4537126123905182, "learning_rate": 9.436876668329411e-05, "loss": 0.0334, "step": 1950 }, { "epoch": 4.780487804878049, "grad_norm": 0.3576773405075073, "learning_rate": 9.429229136536079e-05, "loss": 0.0275, "step": 1960 }, { "epoch": 4.804878048780488, "grad_norm": 0.33921948075294495, "learning_rate": 9.421533167402534e-05, "loss": 0.0335, "step": 1970 }, { "epoch": 4.829268292682927, "grad_norm": 0.46315258741378784, "learning_rate": 9.413788845090666e-05, "loss": 0.0343, "step": 1980 }, { "epoch": 4.853658536585366, "grad_norm": 0.3120940327644348, "learning_rate": 9.405996254291136e-05, "loss": 0.0366, "step": 1990 }, { "epoch": 4.878048780487805, "grad_norm": 0.3575790226459503, "learning_rate": 9.398155480222474e-05, "loss": 0.0296, "step": 2000 }, { "epoch": 4.902439024390244, "grad_norm": 0.29846304655075073, "learning_rate": 9.390266608630128e-05, "loss": 0.0272, "step": 2010 }, { "epoch": 4.926829268292683, "grad_norm": 0.26798102259635925, "learning_rate": 9.38232972578553e-05, "loss": 0.0287, "step": 2020 }, { "epoch": 4.951219512195122, "grad_norm": 0.43504130840301514, "learning_rate": 9.374344918485164e-05, "loss": 0.0389, "step": 2030 }, { "epoch": 4.975609756097561, "grad_norm": 0.32923924922943115, "learning_rate": 9.366312274049602e-05, "loss": 0.0356, "step": 2040 }, { "epoch": 5.0, "grad_norm": 0.7288613319396973, "learning_rate": 9.358231880322554e-05, "loss": 0.0352, "step": 2050 }, { "epoch": 5.024390243902439, "grad_norm": 0.5139240026473999, "learning_rate": 9.350103825669916e-05, "loss": 0.0288, "step": 2060 }, { "epoch": 5.048780487804878, "grad_norm": 0.45391201972961426, "learning_rate": 9.341928198978787e-05, "loss": 0.0291, "step": 2070 }, { "epoch": 5.073170731707317, "grad_norm": 0.4723019003868103, "learning_rate": 9.333705089656512e-05, "loss": 0.0322, "step": 2080 }, { "epoch": 5.097560975609756, "grad_norm": 0.5119431614875793, "learning_rate": 9.325434587629698e-05, "loss": 0.0304, "step": 2090 }, { "epoch": 5.121951219512195, "grad_norm": 0.6561365723609924, "learning_rate": 9.31711678334323e-05, "loss": 0.0292, "step": 2100 }, { "epoch": 5.146341463414634, "grad_norm": 0.44049131870269775, "learning_rate": 9.308751767759282e-05, "loss": 0.0277, "step": 2110 }, { "epoch": 5.170731707317073, "grad_norm": 0.3453926146030426, "learning_rate": 9.300339632356325e-05, "loss": 0.0293, "step": 2120 }, { "epoch": 5.195121951219512, "grad_norm": 0.36750051379203796, "learning_rate": 9.291880469128124e-05, "loss": 0.0298, "step": 2130 }, { "epoch": 5.219512195121951, "grad_norm": 0.500914454460144, "learning_rate": 9.283374370582732e-05, "loss": 0.0291, "step": 2140 }, { "epoch": 5.2439024390243905, "grad_norm": 0.5060284733772278, "learning_rate": 9.274821429741482e-05, "loss": 0.0382, "step": 2150 }, { "epoch": 5.2682926829268295, "grad_norm": 0.38774365186691284, "learning_rate": 9.266221740137961e-05, "loss": 0.0291, "step": 2160 }, { "epoch": 5.2926829268292686, "grad_norm": 0.30381283164024353, "learning_rate": 9.257575395817001e-05, "loss": 0.0296, "step": 2170 }, { "epoch": 5.317073170731708, "grad_norm": 0.43252184987068176, "learning_rate": 9.248882491333637e-05, "loss": 0.0334, "step": 2180 }, { "epoch": 5.341463414634147, "grad_norm": 0.3869166076183319, "learning_rate": 9.240143121752076e-05, "loss": 0.0357, "step": 2190 }, { "epoch": 5.365853658536586, "grad_norm": 0.4016537368297577, "learning_rate": 9.23135738264467e-05, "loss": 0.0335, "step": 2200 }, { "epoch": 5.390243902439025, "grad_norm": 0.46729952096939087, "learning_rate": 9.222525370090849e-05, "loss": 0.0265, "step": 2210 }, { "epoch": 5.414634146341464, "grad_norm": 0.2949901521205902, "learning_rate": 9.213647180676088e-05, "loss": 0.0241, "step": 2220 }, { "epoch": 5.439024390243903, "grad_norm": 0.5333520770072937, "learning_rate": 9.204722911490846e-05, "loss": 0.0332, "step": 2230 }, { "epoch": 5.463414634146342, "grad_norm": 0.36306658387184143, "learning_rate": 9.1957526601295e-05, "loss": 0.0332, "step": 2240 }, { "epoch": 5.487804878048781, "grad_norm": 0.45980194211006165, "learning_rate": 9.186736524689281e-05, "loss": 0.029, "step": 2250 }, { "epoch": 5.512195121951219, "grad_norm": 0.4012170135974884, "learning_rate": 9.177674603769204e-05, "loss": 0.0289, "step": 2260 }, { "epoch": 5.536585365853659, "grad_norm": 0.3718886971473694, "learning_rate": 9.168566996468983e-05, "loss": 0.028, "step": 2270 }, { "epoch": 5.560975609756097, "grad_norm": 0.28512880206108093, "learning_rate": 9.159413802387951e-05, "loss": 0.0334, "step": 2280 }, { "epoch": 5.585365853658536, "grad_norm": 0.3869372606277466, "learning_rate": 9.150215121623974e-05, "loss": 0.0316, "step": 2290 }, { "epoch": 5.609756097560975, "grad_norm": 0.30027449131011963, "learning_rate": 9.140971054772349e-05, "loss": 0.0231, "step": 2300 }, { "epoch": 5.634146341463414, "grad_norm": 0.41158419847488403, "learning_rate": 9.131681702924713e-05, "loss": 0.0348, "step": 2310 }, { "epoch": 5.658536585365853, "grad_norm": 0.4597727656364441, "learning_rate": 9.122347167667926e-05, "loss": 0.0254, "step": 2320 }, { "epoch": 5.682926829268292, "grad_norm": 0.32041415572166443, "learning_rate": 9.112967551082973e-05, "loss": 0.0352, "step": 2330 }, { "epoch": 5.7073170731707314, "grad_norm": 0.4481387138366699, "learning_rate": 9.103542955743835e-05, "loss": 0.0361, "step": 2340 }, { "epoch": 5.7317073170731705, "grad_norm": 0.5182772874832153, "learning_rate": 9.094073484716381e-05, "loss": 0.0306, "step": 2350 }, { "epoch": 5.7560975609756095, "grad_norm": 0.29152920842170715, "learning_rate": 9.084559241557226e-05, "loss": 0.0304, "step": 2360 }, { "epoch": 5.780487804878049, "grad_norm": 0.35041344165802, "learning_rate": 9.075000330312608e-05, "loss": 0.0278, "step": 2370 }, { "epoch": 5.804878048780488, "grad_norm": 0.38795459270477295, "learning_rate": 9.065396855517253e-05, "loss": 0.0237, "step": 2380 }, { "epoch": 5.829268292682927, "grad_norm": 0.28562822937965393, "learning_rate": 9.055748922193219e-05, "loss": 0.0283, "step": 2390 }, { "epoch": 5.853658536585366, "grad_norm": 0.29710128903388977, "learning_rate": 9.046056635848761e-05, "loss": 0.0315, "step": 2400 }, { "epoch": 5.878048780487805, "grad_norm": 0.3948493003845215, "learning_rate": 9.036320102477169e-05, "loss": 0.0277, "step": 2410 }, { "epoch": 5.902439024390244, "grad_norm": 0.2916833758354187, "learning_rate": 9.02653942855561e-05, "loss": 0.0281, "step": 2420 }, { "epoch": 5.926829268292683, "grad_norm": 0.24540220201015472, "learning_rate": 9.016714721043971e-05, "loss": 0.0272, "step": 2430 }, { "epoch": 5.951219512195122, "grad_norm": 0.3842625021934509, "learning_rate": 9.006846087383675e-05, "loss": 0.0324, "step": 2440 }, { "epoch": 5.975609756097561, "grad_norm": 0.23952610790729523, "learning_rate": 8.996933635496523e-05, "loss": 0.03, "step": 2450 }, { "epoch": 6.0, "grad_norm": 0.43705177307128906, "learning_rate": 8.986977473783498e-05, "loss": 0.0282, "step": 2460 }, { "epoch": 6.024390243902439, "grad_norm": 0.35932600498199463, "learning_rate": 8.97697771112359e-05, "loss": 0.0324, "step": 2470 }, { "epoch": 6.048780487804878, "grad_norm": 0.3238205313682556, "learning_rate": 8.966934456872602e-05, "loss": 0.0261, "step": 2480 }, { "epoch": 6.073170731707317, "grad_norm": 0.3258366286754608, "learning_rate": 8.95684782086195e-05, "loss": 0.0255, "step": 2490 }, { "epoch": 6.097560975609756, "grad_norm": 0.3935689926147461, "learning_rate": 8.946717913397476e-05, "loss": 0.0308, "step": 2500 }, { "epoch": 6.121951219512195, "grad_norm": 0.5110504627227783, "learning_rate": 8.93654484525822e-05, "loss": 0.0302, "step": 2510 }, { "epoch": 6.146341463414634, "grad_norm": 0.34262439608573914, "learning_rate": 8.926328727695226e-05, "loss": 0.0265, "step": 2520 }, { "epoch": 6.170731707317073, "grad_norm": 0.29374760389328003, "learning_rate": 8.916069672430319e-05, "loss": 0.0249, "step": 2530 }, { "epoch": 6.195121951219512, "grad_norm": 0.39232876896858215, "learning_rate": 8.905767791654884e-05, "loss": 0.0302, "step": 2540 }, { "epoch": 6.219512195121951, "grad_norm": 0.3207443654537201, "learning_rate": 8.895423198028638e-05, "loss": 0.0289, "step": 2550 }, { "epoch": 6.2439024390243905, "grad_norm": 0.3136545419692993, "learning_rate": 8.885036004678402e-05, "loss": 0.0318, "step": 2560 }, { "epoch": 6.2682926829268295, "grad_norm": 0.21263915300369263, "learning_rate": 8.874606325196857e-05, "loss": 0.0255, "step": 2570 }, { "epoch": 6.2926829268292686, "grad_norm": 0.26851847767829895, "learning_rate": 8.864134273641304e-05, "loss": 0.0287, "step": 2580 }, { "epoch": 6.317073170731708, "grad_norm": 0.3174366354942322, "learning_rate": 8.853619964532427e-05, "loss": 0.0276, "step": 2590 }, { "epoch": 6.341463414634147, "grad_norm": 0.30073943734169006, "learning_rate": 8.843063512853019e-05, "loss": 0.0254, "step": 2600 }, { "epoch": 6.365853658536586, "grad_norm": 0.28819018602371216, "learning_rate": 8.832465034046749e-05, "loss": 0.027, "step": 2610 }, { "epoch": 6.390243902439025, "grad_norm": 0.3695346415042877, "learning_rate": 8.821824644016882e-05, "loss": 0.0276, "step": 2620 }, { "epoch": 6.414634146341464, "grad_norm": 0.3733421862125397, "learning_rate": 8.811142459125019e-05, "loss": 0.0242, "step": 2630 }, { "epoch": 6.439024390243903, "grad_norm": 0.22292673587799072, "learning_rate": 8.800418596189822e-05, "loss": 0.0278, "step": 2640 }, { "epoch": 6.463414634146342, "grad_norm": 0.2964411973953247, "learning_rate": 8.789653172485737e-05, "loss": 0.0257, "step": 2650 }, { "epoch": 6.487804878048781, "grad_norm": 0.3321440815925598, "learning_rate": 8.778846305741715e-05, "loss": 0.0392, "step": 2660 }, { "epoch": 6.512195121951219, "grad_norm": 0.4788553714752197, "learning_rate": 8.767998114139918e-05, "loss": 0.0323, "step": 2670 }, { "epoch": 6.536585365853659, "grad_norm": 0.3286356031894684, "learning_rate": 8.757108716314429e-05, "loss": 0.0317, "step": 2680 }, { "epoch": 6.560975609756097, "grad_norm": 0.45678022503852844, "learning_rate": 8.746178231349962e-05, "loss": 0.025, "step": 2690 }, { "epoch": 6.585365853658536, "grad_norm": 0.2545185685157776, "learning_rate": 8.735206778780549e-05, "loss": 0.0296, "step": 2700 }, { "epoch": 6.609756097560975, "grad_norm": 0.3429373800754547, "learning_rate": 8.724194478588234e-05, "loss": 0.0306, "step": 2710 }, { "epoch": 6.634146341463414, "grad_norm": 0.2924957275390625, "learning_rate": 8.713141451201772e-05, "loss": 0.0284, "step": 2720 }, { "epoch": 6.658536585365853, "grad_norm": 0.23084889352321625, "learning_rate": 8.702047817495295e-05, "loss": 0.0303, "step": 2730 }, { "epoch": 6.682926829268292, "grad_norm": 0.2847777307033539, "learning_rate": 8.69091369878701e-05, "loss": 0.0292, "step": 2740 }, { "epoch": 6.7073170731707314, "grad_norm": 0.3510938584804535, "learning_rate": 8.679739216837849e-05, "loss": 0.0235, "step": 2750 }, { "epoch": 6.7317073170731705, "grad_norm": 0.2585855722427368, "learning_rate": 8.66852449385016e-05, "loss": 0.0236, "step": 2760 }, { "epoch": 6.7560975609756095, "grad_norm": 0.4169798493385315, "learning_rate": 8.657269652466356e-05, "loss": 0.0266, "step": 2770 }, { "epoch": 6.780487804878049, "grad_norm": 0.2194824367761612, "learning_rate": 8.645974815767577e-05, "loss": 0.0246, "step": 2780 }, { "epoch": 6.804878048780488, "grad_norm": 0.2582210302352905, "learning_rate": 8.634640107272351e-05, "loss": 0.0255, "step": 2790 }, { "epoch": 6.829268292682927, "grad_norm": 0.3835674226284027, "learning_rate": 8.623265650935234e-05, "loss": 0.025, "step": 2800 }, { "epoch": 6.853658536585366, "grad_norm": 0.29900819063186646, "learning_rate": 8.611851571145456e-05, "loss": 0.0288, "step": 2810 }, { "epoch": 6.878048780487805, "grad_norm": 0.28595781326293945, "learning_rate": 8.600397992725566e-05, "loss": 0.0198, "step": 2820 }, { "epoch": 6.902439024390244, "grad_norm": 0.2295774519443512, "learning_rate": 8.588905040930061e-05, "loss": 0.028, "step": 2830 }, { "epoch": 6.926829268292683, "grad_norm": 0.39176326990127563, "learning_rate": 8.577372841444022e-05, "loss": 0.0276, "step": 2840 }, { "epoch": 6.951219512195122, "grad_norm": 0.24291186034679413, "learning_rate": 8.565801520381736e-05, "loss": 0.0244, "step": 2850 }, { "epoch": 6.975609756097561, "grad_norm": 0.4400140643119812, "learning_rate": 8.554191204285313e-05, "loss": 0.0268, "step": 2860 }, { "epoch": 7.0, "grad_norm": 0.3001735508441925, "learning_rate": 8.542542020123315e-05, "loss": 0.0235, "step": 2870 }, { "epoch": 7.024390243902439, "grad_norm": 0.27095648646354675, "learning_rate": 8.530854095289347e-05, "loss": 0.0234, "step": 2880 }, { "epoch": 7.048780487804878, "grad_norm": 0.31923484802246094, "learning_rate": 8.519127557600688e-05, "loss": 0.0225, "step": 2890 }, { "epoch": 7.073170731707317, "grad_norm": 0.297625333070755, "learning_rate": 8.507362535296871e-05, "loss": 0.0243, "step": 2900 }, { "epoch": 7.097560975609756, "grad_norm": 0.4184678792953491, "learning_rate": 8.495559157038299e-05, "loss": 0.0307, "step": 2910 }, { "epoch": 7.121951219512195, "grad_norm": 0.3742824196815491, "learning_rate": 8.483717551904823e-05, "loss": 0.0258, "step": 2920 }, { "epoch": 7.146341463414634, "grad_norm": 0.23552247881889343, "learning_rate": 8.47183784939434e-05, "loss": 0.0253, "step": 2930 }, { "epoch": 7.170731707317073, "grad_norm": 0.20580260455608368, "learning_rate": 8.459920179421374e-05, "loss": 0.0221, "step": 2940 }, { "epoch": 7.195121951219512, "grad_norm": 0.5326234102249146, "learning_rate": 8.447964672315656e-05, "loss": 0.0249, "step": 2950 }, { "epoch": 7.219512195121951, "grad_norm": 0.2974367141723633, "learning_rate": 8.435971458820692e-05, "loss": 0.0264, "step": 2960 }, { "epoch": 7.2439024390243905, "grad_norm": 0.3622567355632782, "learning_rate": 8.423940670092345e-05, "loss": 0.0246, "step": 2970 }, { "epoch": 7.2682926829268295, "grad_norm": 0.27394846081733704, "learning_rate": 8.411872437697394e-05, "loss": 0.0264, "step": 2980 }, { "epoch": 7.2926829268292686, "grad_norm": 0.3948928117752075, "learning_rate": 8.399766893612096e-05, "loss": 0.0239, "step": 2990 }, { "epoch": 7.317073170731708, "grad_norm": 0.4600905478000641, "learning_rate": 8.38762417022074e-05, "loss": 0.0242, "step": 3000 }, { "epoch": 7.341463414634147, "grad_norm": 0.40093496441841125, "learning_rate": 8.375444400314204e-05, "loss": 0.0245, "step": 3010 }, { "epoch": 7.365853658536586, "grad_norm": 0.2586691975593567, "learning_rate": 8.3632277170885e-05, "loss": 0.0289, "step": 3020 }, { "epoch": 7.390243902439025, "grad_norm": 0.2621522843837738, "learning_rate": 8.350974254143318e-05, "loss": 0.0281, "step": 3030 }, { "epoch": 7.414634146341464, "grad_norm": 0.4712395668029785, "learning_rate": 8.338684145480566e-05, "loss": 0.0248, "step": 3040 }, { "epoch": 7.439024390243903, "grad_norm": 0.34933534264564514, "learning_rate": 8.326357525502904e-05, "loss": 0.0253, "step": 3050 }, { "epoch": 7.463414634146342, "grad_norm": 0.28874891996383667, "learning_rate": 8.313994529012273e-05, "loss": 0.0319, "step": 3060 }, { "epoch": 7.487804878048781, "grad_norm": 0.3775670528411865, "learning_rate": 8.301595291208422e-05, "loss": 0.0303, "step": 3070 }, { "epoch": 7.512195121951219, "grad_norm": 0.44247958064079285, "learning_rate": 8.289159947687427e-05, "loss": 0.0236, "step": 3080 }, { "epoch": 7.536585365853659, "grad_norm": 0.28774020075798035, "learning_rate": 8.276688634440216e-05, "loss": 0.0276, "step": 3090 }, { "epoch": 7.560975609756097, "grad_norm": 0.4511764347553253, "learning_rate": 8.26418148785107e-05, "loss": 0.0262, "step": 3100 }, { "epoch": 7.585365853658536, "grad_norm": 0.46926388144493103, "learning_rate": 8.251638644696141e-05, "loss": 0.0319, "step": 3110 }, { "epoch": 7.609756097560975, "grad_norm": 0.289802223443985, "learning_rate": 8.23906024214195e-05, "loss": 0.0342, "step": 3120 }, { "epoch": 7.634146341463414, "grad_norm": 0.21628691256046295, "learning_rate": 8.226446417743897e-05, "loss": 0.0253, "step": 3130 }, { "epoch": 7.658536585365853, "grad_norm": 0.2442767471075058, "learning_rate": 8.213797309444742e-05, "loss": 0.0309, "step": 3140 }, { "epoch": 7.682926829268292, "grad_norm": 0.27982255816459656, "learning_rate": 8.201113055573105e-05, "loss": 0.0295, "step": 3150 }, { "epoch": 7.7073170731707314, "grad_norm": 0.32631781697273254, "learning_rate": 8.188393794841958e-05, "loss": 0.0216, "step": 3160 }, { "epoch": 7.7317073170731705, "grad_norm": 0.3465834856033325, "learning_rate": 8.175639666347094e-05, "loss": 0.0282, "step": 3170 }, { "epoch": 7.7560975609756095, "grad_norm": 0.34934505820274353, "learning_rate": 8.162850809565623e-05, "loss": 0.0267, "step": 3180 }, { "epoch": 7.780487804878049, "grad_norm": 0.36410462856292725, "learning_rate": 8.150027364354431e-05, "loss": 0.0241, "step": 3190 }, { "epoch": 7.804878048780488, "grad_norm": 0.30558130145072937, "learning_rate": 8.137169470948662e-05, "loss": 0.0292, "step": 3200 }, { "epoch": 7.829268292682927, "grad_norm": 0.2770102918148041, "learning_rate": 8.124277269960179e-05, "loss": 0.0249, "step": 3210 }, { "epoch": 7.853658536585366, "grad_norm": 0.3832545280456543, "learning_rate": 8.111350902376023e-05, "loss": 0.0287, "step": 3220 }, { "epoch": 7.878048780487805, "grad_norm": 0.3744196891784668, "learning_rate": 8.098390509556883e-05, "loss": 0.0254, "step": 3230 }, { "epoch": 7.902439024390244, "grad_norm": 0.31735700368881226, "learning_rate": 8.085396233235536e-05, "loss": 0.0209, "step": 3240 }, { "epoch": 7.926829268292683, "grad_norm": 0.23986178636550903, "learning_rate": 8.072368215515306e-05, "loss": 0.0206, "step": 3250 }, { "epoch": 7.951219512195122, "grad_norm": 0.3069005608558655, "learning_rate": 8.059306598868506e-05, "loss": 0.0289, "step": 3260 }, { "epoch": 7.975609756097561, "grad_norm": 0.2521304786205292, "learning_rate": 8.046211526134888e-05, "loss": 0.0254, "step": 3270 }, { "epoch": 8.0, "grad_norm": 0.26578664779663086, "learning_rate": 8.033083140520065e-05, "loss": 0.0272, "step": 3280 }, { "epoch": 8.024390243902438, "grad_norm": 0.3784792721271515, "learning_rate": 8.019921585593962e-05, "loss": 0.0248, "step": 3290 }, { "epoch": 8.048780487804878, "grad_norm": 0.28874078392982483, "learning_rate": 8.006727005289232e-05, "loss": 0.0213, "step": 3300 }, { "epoch": 8.073170731707316, "grad_norm": 0.449009507894516, "learning_rate": 7.993499543899692e-05, "loss": 0.0206, "step": 3310 }, { "epoch": 8.097560975609756, "grad_norm": 0.5118389129638672, "learning_rate": 7.980239346078742e-05, "loss": 0.026, "step": 3320 }, { "epoch": 8.121951219512194, "grad_norm": 0.29989659786224365, "learning_rate": 7.966946556837778e-05, "loss": 0.0219, "step": 3330 }, { "epoch": 8.146341463414634, "grad_norm": 0.27204805612564087, "learning_rate": 7.953621321544616e-05, "loss": 0.0259, "step": 3340 }, { "epoch": 8.170731707317072, "grad_norm": 0.2746070921421051, "learning_rate": 7.940263785921896e-05, "loss": 0.0217, "step": 3350 }, { "epoch": 8.195121951219512, "grad_norm": 0.21347470581531525, "learning_rate": 7.926874096045482e-05, "loss": 0.0248, "step": 3360 }, { "epoch": 8.21951219512195, "grad_norm": 0.2145281881093979, "learning_rate": 7.913452398342881e-05, "loss": 0.0271, "step": 3370 }, { "epoch": 8.24390243902439, "grad_norm": 0.3854407072067261, "learning_rate": 7.89999883959163e-05, "loss": 0.0269, "step": 3380 }, { "epoch": 8.268292682926829, "grad_norm": 0.3943921625614166, "learning_rate": 7.886513566917687e-05, "loss": 0.0288, "step": 3390 }, { "epoch": 8.292682926829269, "grad_norm": 0.3694443106651306, "learning_rate": 7.872996727793838e-05, "loss": 0.0242, "step": 3400 }, { "epoch": 8.317073170731707, "grad_norm": 0.3698720335960388, "learning_rate": 7.859448470038069e-05, "loss": 0.0271, "step": 3410 }, { "epoch": 8.341463414634147, "grad_norm": 0.4029603600502014, "learning_rate": 7.845868941811956e-05, "loss": 0.0246, "step": 3420 }, { "epoch": 8.365853658536585, "grad_norm": 0.3008832633495331, "learning_rate": 7.832258291619043e-05, "loss": 0.0217, "step": 3430 }, { "epoch": 8.390243902439025, "grad_norm": 0.31910908222198486, "learning_rate": 7.81861666830322e-05, "loss": 0.0202, "step": 3440 }, { "epoch": 8.414634146341463, "grad_norm": 0.356024831533432, "learning_rate": 7.804944221047097e-05, "loss": 0.0291, "step": 3450 }, { "epoch": 8.439024390243903, "grad_norm": 0.220293790102005, "learning_rate": 7.791241099370364e-05, "loss": 0.0241, "step": 3460 }, { "epoch": 8.463414634146341, "grad_norm": 0.22763630747795105, "learning_rate": 7.777507453128163e-05, "loss": 0.0231, "step": 3470 }, { "epoch": 8.487804878048781, "grad_norm": 0.3023100197315216, "learning_rate": 7.763743432509451e-05, "loss": 0.0262, "step": 3480 }, { "epoch": 8.512195121951219, "grad_norm": 0.26969295740127563, "learning_rate": 7.749949188035353e-05, "loss": 0.0257, "step": 3490 }, { "epoch": 8.536585365853659, "grad_norm": 0.40289902687072754, "learning_rate": 7.736124870557516e-05, "loss": 0.0291, "step": 3500 }, { "epoch": 8.560975609756097, "grad_norm": 0.3947031795978546, "learning_rate": 7.722270631256459e-05, "loss": 0.027, "step": 3510 }, { "epoch": 8.585365853658537, "grad_norm": 0.3091443181037903, "learning_rate": 7.708386621639925e-05, "loss": 0.0238, "step": 3520 }, { "epoch": 8.609756097560975, "grad_norm": 0.2447778433561325, "learning_rate": 7.694472993541219e-05, "loss": 0.0239, "step": 3530 }, { "epoch": 8.634146341463415, "grad_norm": 0.2967393100261688, "learning_rate": 7.680529899117547e-05, "loss": 0.0238, "step": 3540 }, { "epoch": 8.658536585365853, "grad_norm": 0.2527478039264679, "learning_rate": 7.666557490848358e-05, "loss": 0.0233, "step": 3550 }, { "epoch": 8.682926829268293, "grad_norm": 0.38460373878479004, "learning_rate": 7.65255592153367e-05, "loss": 0.025, "step": 3560 }, { "epoch": 8.707317073170731, "grad_norm": 0.4034731388092041, "learning_rate": 7.638525344292402e-05, "loss": 0.0284, "step": 3570 }, { "epoch": 8.731707317073171, "grad_norm": 0.43001022934913635, "learning_rate": 7.624465912560697e-05, "loss": 0.0238, "step": 3580 }, { "epoch": 8.75609756097561, "grad_norm": 0.31898584961891174, "learning_rate": 7.610377780090249e-05, "loss": 0.0272, "step": 3590 }, { "epoch": 8.78048780487805, "grad_norm": 0.4841764271259308, "learning_rate": 7.596261100946618e-05, "loss": 0.0223, "step": 3600 }, { "epoch": 8.804878048780488, "grad_norm": 0.29002511501312256, "learning_rate": 7.582116029507542e-05, "loss": 0.0255, "step": 3610 }, { "epoch": 8.829268292682928, "grad_norm": 0.4290894865989685, "learning_rate": 7.56794272046126e-05, "loss": 0.0232, "step": 3620 }, { "epoch": 8.853658536585366, "grad_norm": 0.3998239040374756, "learning_rate": 7.55374132880481e-05, "loss": 0.0263, "step": 3630 }, { "epoch": 8.878048780487806, "grad_norm": 0.3078030049800873, "learning_rate": 7.539512009842333e-05, "loss": 0.0236, "step": 3640 }, { "epoch": 8.902439024390244, "grad_norm": 0.4083644449710846, "learning_rate": 7.525254919183382e-05, "loss": 0.0268, "step": 3650 }, { "epoch": 8.926829268292684, "grad_norm": 0.3749852478504181, "learning_rate": 7.510970212741215e-05, "loss": 0.0206, "step": 3660 }, { "epoch": 8.951219512195122, "grad_norm": 0.2569369673728943, "learning_rate": 7.496658046731096e-05, "loss": 0.0206, "step": 3670 }, { "epoch": 8.975609756097562, "grad_norm": 0.35169756412506104, "learning_rate": 7.482318577668578e-05, "loss": 0.0257, "step": 3680 }, { "epoch": 9.0, "grad_norm": 0.31908467411994934, "learning_rate": 7.467951962367796e-05, "loss": 0.0253, "step": 3690 }, { "epoch": 9.024390243902438, "grad_norm": 0.3758999705314636, "learning_rate": 7.453558357939755e-05, "loss": 0.025, "step": 3700 }, { "epoch": 9.048780487804878, "grad_norm": 0.24222075939178467, "learning_rate": 7.439137921790606e-05, "loss": 0.0243, "step": 3710 }, { "epoch": 9.073170731707316, "grad_norm": 0.3466871380805969, "learning_rate": 7.42469081161993e-05, "loss": 0.0286, "step": 3720 }, { "epoch": 9.097560975609756, "grad_norm": 0.41073858737945557, "learning_rate": 7.410217185419006e-05, "loss": 0.0228, "step": 3730 }, { "epoch": 9.121951219512194, "grad_norm": 0.2821403443813324, "learning_rate": 7.395717201469095e-05, "loss": 0.0202, "step": 3740 }, { "epoch": 9.146341463414634, "grad_norm": 0.2303513139486313, "learning_rate": 7.381191018339696e-05, "loss": 0.0268, "step": 3750 }, { "epoch": 9.170731707317072, "grad_norm": 0.2297833412885666, "learning_rate": 7.36663879488682e-05, "loss": 0.024, "step": 3760 }, { "epoch": 9.195121951219512, "grad_norm": 0.19966505467891693, "learning_rate": 7.352060690251254e-05, "loss": 0.0277, "step": 3770 }, { "epoch": 9.21951219512195, "grad_norm": 0.24993550777435303, "learning_rate": 7.337456863856811e-05, "loss": 0.0246, "step": 3780 }, { "epoch": 9.24390243902439, "grad_norm": 0.29471948742866516, "learning_rate": 7.3228274754086e-05, "loss": 0.0161, "step": 3790 }, { "epoch": 9.268292682926829, "grad_norm": 0.2377634197473526, "learning_rate": 7.308172684891267e-05, "loss": 0.0232, "step": 3800 }, { "epoch": 9.292682926829269, "grad_norm": 0.3629654049873352, "learning_rate": 7.293492652567255e-05, "loss": 0.0229, "step": 3810 }, { "epoch": 9.317073170731707, "grad_norm": 0.25956496596336365, "learning_rate": 7.278787538975043e-05, "loss": 0.0245, "step": 3820 }, { "epoch": 9.341463414634147, "grad_norm": 0.467792272567749, "learning_rate": 7.2640575049274e-05, "loss": 0.0231, "step": 3830 }, { "epoch": 9.365853658536585, "grad_norm": 0.45132941007614136, "learning_rate": 7.249302711509616e-05, "loss": 0.0238, "step": 3840 }, { "epoch": 9.390243902439025, "grad_norm": 0.23344075679779053, "learning_rate": 7.23452332007775e-05, "loss": 0.0263, "step": 3850 }, { "epoch": 9.414634146341463, "grad_norm": 0.3689783215522766, "learning_rate": 7.219719492256858e-05, "loss": 0.0197, "step": 3860 }, { "epoch": 9.439024390243903, "grad_norm": 0.29409611225128174, "learning_rate": 7.20489138993923e-05, "loss": 0.0257, "step": 3870 }, { "epoch": 9.463414634146341, "grad_norm": 0.3899075388908386, "learning_rate": 7.190039175282614e-05, "loss": 0.0222, "step": 3880 }, { "epoch": 9.487804878048781, "grad_norm": 0.22943243384361267, "learning_rate": 7.175163010708455e-05, "loss": 0.0176, "step": 3890 }, { "epoch": 9.512195121951219, "grad_norm": 0.24564087390899658, "learning_rate": 7.1602630589001e-05, "loss": 0.0224, "step": 3900 }, { "epoch": 9.536585365853659, "grad_norm": 0.45181769132614136, "learning_rate": 7.14533948280104e-05, "loss": 0.0205, "step": 3910 }, { "epoch": 9.560975609756097, "grad_norm": 0.463695764541626, "learning_rate": 7.130392445613109e-05, "loss": 0.0245, "step": 3920 }, { "epoch": 9.585365853658537, "grad_norm": 0.3602858781814575, "learning_rate": 7.115422110794711e-05, "loss": 0.0227, "step": 3930 }, { "epoch": 9.609756097560975, "grad_norm": 0.23871228098869324, "learning_rate": 7.100428642059033e-05, "loss": 0.0243, "step": 3940 }, { "epoch": 9.634146341463415, "grad_norm": 0.22365263104438782, "learning_rate": 7.08541220337224e-05, "loss": 0.0253, "step": 3950 }, { "epoch": 9.658536585365853, "grad_norm": 0.22427475452423096, "learning_rate": 7.070372958951706e-05, "loss": 0.0254, "step": 3960 }, { "epoch": 9.682926829268293, "grad_norm": 0.198167622089386, "learning_rate": 7.055311073264194e-05, "loss": 0.0212, "step": 3970 }, { "epoch": 9.707317073170731, "grad_norm": 0.2180306762456894, "learning_rate": 7.040226711024077e-05, "loss": 0.0223, "step": 3980 }, { "epoch": 9.731707317073171, "grad_norm": 0.29318174719810486, "learning_rate": 7.02512003719152e-05, "loss": 0.0222, "step": 3990 }, { "epoch": 9.75609756097561, "grad_norm": 0.14693811535835266, "learning_rate": 7.00999121697069e-05, "loss": 0.0217, "step": 4000 }, { "epoch": 9.78048780487805, "grad_norm": 0.328864187002182, "learning_rate": 6.99484041580794e-05, "loss": 0.0246, "step": 4010 }, { "epoch": 9.804878048780488, "grad_norm": 0.33691155910491943, "learning_rate": 6.979667799390004e-05, "loss": 0.0216, "step": 4020 }, { "epoch": 9.829268292682928, "grad_norm": 0.2644006907939911, "learning_rate": 6.964473533642185e-05, "loss": 0.0205, "step": 4030 }, { "epoch": 9.853658536585366, "grad_norm": 0.2517907917499542, "learning_rate": 6.949257784726539e-05, "loss": 0.0231, "step": 4040 }, { "epoch": 9.878048780487806, "grad_norm": 0.1978023499250412, "learning_rate": 6.934020719040056e-05, "loss": 0.0233, "step": 4050 }, { "epoch": 9.902439024390244, "grad_norm": 0.21526019275188446, "learning_rate": 6.918762503212848e-05, "loss": 0.0243, "step": 4060 }, { "epoch": 9.926829268292684, "grad_norm": 0.22828538715839386, "learning_rate": 6.903483304106319e-05, "loss": 0.0264, "step": 4070 }, { "epoch": 9.951219512195122, "grad_norm": 0.27751001715660095, "learning_rate": 6.888183288811341e-05, "loss": 0.02, "step": 4080 }, { "epoch": 9.975609756097562, "grad_norm": 0.2941833436489105, "learning_rate": 6.87286262464643e-05, "loss": 0.0235, "step": 4090 }, { "epoch": 10.0, "grad_norm": 0.2639453709125519, "learning_rate": 6.857521479155915e-05, "loss": 0.0215, "step": 4100 }, { "epoch": 10.024390243902438, "grad_norm": 0.2582087218761444, "learning_rate": 6.842160020108104e-05, "loss": 0.0213, "step": 4110 }, { "epoch": 10.048780487804878, "grad_norm": 0.1619865447282791, "learning_rate": 6.826778415493455e-05, "loss": 0.0198, "step": 4120 }, { "epoch": 10.073170731707316, "grad_norm": 0.3544345200061798, "learning_rate": 6.811376833522729e-05, "loss": 0.0228, "step": 4130 }, { "epoch": 10.097560975609756, "grad_norm": 0.3791655898094177, "learning_rate": 6.795955442625159e-05, "loss": 0.0231, "step": 4140 }, { "epoch": 10.121951219512194, "grad_norm": 0.31393322348594666, "learning_rate": 6.780514411446608e-05, "loss": 0.0273, "step": 4150 }, { "epoch": 10.146341463414634, "grad_norm": 0.4197497069835663, "learning_rate": 6.765053908847716e-05, "loss": 0.0191, "step": 4160 }, { "epoch": 10.170731707317072, "grad_norm": 0.4402354657649994, "learning_rate": 6.749574103902064e-05, "loss": 0.0239, "step": 4170 }, { "epoch": 10.195121951219512, "grad_norm": 0.2804039716720581, "learning_rate": 6.734075165894317e-05, "loss": 0.0234, "step": 4180 }, { "epoch": 10.21951219512195, "grad_norm": 0.3387048840522766, "learning_rate": 6.71855726431838e-05, "loss": 0.0261, "step": 4190 }, { "epoch": 10.24390243902439, "grad_norm": 0.2285296469926834, "learning_rate": 6.703020568875538e-05, "loss": 0.0213, "step": 4200 }, { "epoch": 10.268292682926829, "grad_norm": 0.1758410930633545, "learning_rate": 6.687465249472603e-05, "loss": 0.0205, "step": 4210 }, { "epoch": 10.292682926829269, "grad_norm": 0.2568712830543518, "learning_rate": 6.671891476220055e-05, "loss": 0.0191, "step": 4220 }, { "epoch": 10.317073170731707, "grad_norm": 0.23384162783622742, "learning_rate": 6.656299419430183e-05, "loss": 0.0189, "step": 4230 }, { "epoch": 10.341463414634147, "grad_norm": 0.2457735687494278, "learning_rate": 6.640689249615223e-05, "loss": 0.022, "step": 4240 }, { "epoch": 10.365853658536585, "grad_norm": 0.3043486475944519, "learning_rate": 6.625061137485491e-05, "loss": 0.0231, "step": 4250 }, { "epoch": 10.390243902439025, "grad_norm": 0.2883990705013275, "learning_rate": 6.609415253947517e-05, "loss": 0.0274, "step": 4260 }, { "epoch": 10.414634146341463, "grad_norm": 0.34007298946380615, "learning_rate": 6.593751770102178e-05, "loss": 0.0209, "step": 4270 }, { "epoch": 10.439024390243903, "grad_norm": 0.3055611848831177, "learning_rate": 6.578070857242823e-05, "loss": 0.0225, "step": 4280 }, { "epoch": 10.463414634146341, "grad_norm": 0.2238772064447403, "learning_rate": 6.562372686853402e-05, "loss": 0.0189, "step": 4290 }, { "epoch": 10.487804878048781, "grad_norm": 0.20618516206741333, "learning_rate": 6.546657430606593e-05, "loss": 0.0234, "step": 4300 }, { "epoch": 10.512195121951219, "grad_norm": 0.28194308280944824, "learning_rate": 6.530925260361918e-05, "loss": 0.0206, "step": 4310 }, { "epoch": 10.536585365853659, "grad_norm": 0.22390757501125336, "learning_rate": 6.515176348163871e-05, "loss": 0.0195, "step": 4320 }, { "epoch": 10.560975609756097, "grad_norm": 0.22681041061878204, "learning_rate": 6.499410866240032e-05, "loss": 0.0187, "step": 4330 }, { "epoch": 10.585365853658537, "grad_norm": 0.3045816421508789, "learning_rate": 6.48362898699919e-05, "loss": 0.0152, "step": 4340 }, { "epoch": 10.609756097560975, "grad_norm": 0.3179272413253784, "learning_rate": 6.467830883029443e-05, "loss": 0.0214, "step": 4350 }, { "epoch": 10.634146341463415, "grad_norm": 0.2916434109210968, "learning_rate": 6.452016727096326e-05, "loss": 0.0202, "step": 4360 }, { "epoch": 10.658536585365853, "grad_norm": 0.24403797090053558, "learning_rate": 6.436186692140916e-05, "loss": 0.0221, "step": 4370 }, { "epoch": 10.682926829268293, "grad_norm": 0.22020524740219116, "learning_rate": 6.420340951277938e-05, "loss": 0.0174, "step": 4380 }, { "epoch": 10.707317073170731, "grad_norm": 0.3301219642162323, "learning_rate": 6.404479677793874e-05, "loss": 0.0178, "step": 4390 }, { "epoch": 10.731707317073171, "grad_norm": 0.3293200135231018, "learning_rate": 6.388603045145075e-05, "loss": 0.025, "step": 4400 }, { "epoch": 10.75609756097561, "grad_norm": 0.27123138308525085, "learning_rate": 6.372711226955843e-05, "loss": 0.0265, "step": 4410 }, { "epoch": 10.78048780487805, "grad_norm": 0.32547059655189514, "learning_rate": 6.356804397016564e-05, "loss": 0.0191, "step": 4420 }, { "epoch": 10.804878048780488, "grad_norm": 0.298409104347229, "learning_rate": 6.340882729281779e-05, "loss": 0.0287, "step": 4430 }, { "epoch": 10.829268292682928, "grad_norm": 0.24305979907512665, "learning_rate": 6.324946397868294e-05, "loss": 0.0231, "step": 4440 }, { "epoch": 10.853658536585366, "grad_norm": 0.3580092191696167, "learning_rate": 6.308995577053276e-05, "loss": 0.0244, "step": 4450 }, { "epoch": 10.878048780487806, "grad_norm": 0.17259186506271362, "learning_rate": 6.293030441272347e-05, "loss": 0.0178, "step": 4460 }, { "epoch": 10.902439024390244, "grad_norm": 0.22783319652080536, "learning_rate": 6.277051165117677e-05, "loss": 0.0205, "step": 4470 }, { "epoch": 10.926829268292684, "grad_norm": 0.2750856578350067, "learning_rate": 6.261057923336064e-05, "loss": 0.0201, "step": 4480 }, { "epoch": 10.951219512195122, "grad_norm": 0.3185718059539795, "learning_rate": 6.245050890827042e-05, "loss": 0.0221, "step": 4490 }, { "epoch": 10.975609756097562, "grad_norm": 0.2536439597606659, "learning_rate": 6.229030242640952e-05, "loss": 0.022, "step": 4500 }, { "epoch": 11.0, "grad_norm": 0.35280293226242065, "learning_rate": 6.212996153977037e-05, "loss": 0.0198, "step": 4510 }, { "epoch": 11.024390243902438, "grad_norm": 0.34211188554763794, "learning_rate": 6.196948800181523e-05, "loss": 0.0223, "step": 4520 }, { "epoch": 11.048780487804878, "grad_norm": 0.3623008131980896, "learning_rate": 6.180888356745695e-05, "loss": 0.023, "step": 4530 }, { "epoch": 11.073170731707316, "grad_norm": 0.2666378915309906, "learning_rate": 6.164814999303995e-05, "loss": 0.0193, "step": 4540 }, { "epoch": 11.097560975609756, "grad_norm": 0.2366725504398346, "learning_rate": 6.148728903632081e-05, "loss": 0.0181, "step": 4550 }, { "epoch": 11.121951219512194, "grad_norm": 0.17401465773582458, "learning_rate": 6.132630245644921e-05, "loss": 0.0171, "step": 4560 }, { "epoch": 11.146341463414634, "grad_norm": 0.19249685108661652, "learning_rate": 6.116519201394857e-05, "loss": 0.0212, "step": 4570 }, { "epoch": 11.170731707317072, "grad_norm": 0.19817626476287842, "learning_rate": 6.10039594706969e-05, "loss": 0.0199, "step": 4580 }, { "epoch": 11.195121951219512, "grad_norm": 0.37751418352127075, "learning_rate": 6.084260658990744e-05, "loss": 0.0214, "step": 4590 }, { "epoch": 11.21951219512195, "grad_norm": 0.24243248999118805, "learning_rate": 6.068113513610943e-05, "loss": 0.0206, "step": 4600 }, { "epoch": 11.24390243902439, "grad_norm": 0.21889227628707886, "learning_rate": 6.0519546875128876e-05, "loss": 0.0222, "step": 4610 }, { "epoch": 11.268292682926829, "grad_norm": 0.29077640175819397, "learning_rate": 6.035784357406906e-05, "loss": 0.0211, "step": 4620 }, { "epoch": 11.292682926829269, "grad_norm": 0.1955326795578003, "learning_rate": 6.01960270012914e-05, "loss": 0.0211, "step": 4630 }, { "epoch": 11.317073170731707, "grad_norm": 0.25588569045066833, "learning_rate": 6.003409892639599e-05, "loss": 0.0201, "step": 4640 }, { "epoch": 11.341463414634147, "grad_norm": 0.23723214864730835, "learning_rate": 5.9872061120202336e-05, "loss": 0.0185, "step": 4650 }, { "epoch": 11.365853658536585, "grad_norm": 0.32810404896736145, "learning_rate": 5.9709915354729914e-05, "loss": 0.0206, "step": 4660 }, { "epoch": 11.390243902439025, "grad_norm": 0.36235862970352173, "learning_rate": 5.9547663403178824e-05, "loss": 0.0283, "step": 4670 }, { "epoch": 11.414634146341463, "grad_norm": 0.241432324051857, "learning_rate": 5.9385307039910445e-05, "loss": 0.0176, "step": 4680 }, { "epoch": 11.439024390243903, "grad_norm": 0.30528154969215393, "learning_rate": 5.922284804042792e-05, "loss": 0.0233, "step": 4690 }, { "epoch": 11.463414634146341, "grad_norm": 0.2939111590385437, "learning_rate": 5.906028818135687e-05, "loss": 0.0219, "step": 4700 }, { "epoch": 11.487804878048781, "grad_norm": 0.26651132106781006, "learning_rate": 5.889762924042585e-05, "loss": 0.0228, "step": 4710 }, { "epoch": 11.512195121951219, "grad_norm": 0.36398816108703613, "learning_rate": 5.873487299644699e-05, "loss": 0.0216, "step": 4720 }, { "epoch": 11.536585365853659, "grad_norm": 0.2878786623477936, "learning_rate": 5.857202122929649e-05, "loss": 0.0214, "step": 4730 }, { "epoch": 11.560975609756097, "grad_norm": 0.20646262168884277, "learning_rate": 5.840907571989518e-05, "loss": 0.018, "step": 4740 }, { "epoch": 11.585365853658537, "grad_norm": 0.2746175229549408, "learning_rate": 5.824603825018904e-05, "loss": 0.0184, "step": 4750 }, { "epoch": 11.609756097560975, "grad_norm": 0.2174173891544342, "learning_rate": 5.808291060312975e-05, "loss": 0.0208, "step": 4760 }, { "epoch": 11.634146341463415, "grad_norm": 0.1724894940853119, "learning_rate": 5.7919694562655083e-05, "loss": 0.0193, "step": 4770 }, { "epoch": 11.658536585365853, "grad_norm": 0.3776164650917053, "learning_rate": 5.775639191366954e-05, "loss": 0.0231, "step": 4780 }, { "epoch": 11.682926829268293, "grad_norm": 0.2553853690624237, "learning_rate": 5.75930044420247e-05, "loss": 0.0199, "step": 4790 }, { "epoch": 11.707317073170731, "grad_norm": 0.3227883577346802, "learning_rate": 5.74295339344998e-05, "loss": 0.018, "step": 4800 }, { "epoch": 11.731707317073171, "grad_norm": 0.21700100600719452, "learning_rate": 5.726598217878211e-05, "loss": 0.0235, "step": 4810 }, { "epoch": 11.75609756097561, "grad_norm": 0.28760379552841187, "learning_rate": 5.71023509634474e-05, "loss": 0.0194, "step": 4820 }, { "epoch": 11.78048780487805, "grad_norm": 0.2390815019607544, "learning_rate": 5.693864207794049e-05, "loss": 0.0165, "step": 4830 }, { "epoch": 11.804878048780488, "grad_norm": 0.26883596181869507, "learning_rate": 5.677485731255545e-05, "loss": 0.0234, "step": 4840 }, { "epoch": 11.829268292682928, "grad_norm": 0.2687983512878418, "learning_rate": 5.6610998458416296e-05, "loss": 0.0225, "step": 4850 }, { "epoch": 11.853658536585366, "grad_norm": 0.252943754196167, "learning_rate": 5.644706730745716e-05, "loss": 0.0205, "step": 4860 }, { "epoch": 11.878048780487806, "grad_norm": 0.2529122531414032, "learning_rate": 5.628306565240287e-05, "loss": 0.0174, "step": 4870 }, { "epoch": 11.902439024390244, "grad_norm": 0.23580294847488403, "learning_rate": 5.611899528674923e-05, "loss": 0.0203, "step": 4880 }, { "epoch": 11.926829268292684, "grad_norm": 0.26065850257873535, "learning_rate": 5.595485800474349e-05, "loss": 0.0187, "step": 4890 }, { "epoch": 11.951219512195122, "grad_norm": 0.2584094703197479, "learning_rate": 5.579065560136467e-05, "loss": 0.0217, "step": 4900 }, { "epoch": 11.975609756097562, "grad_norm": 0.22429655492305756, "learning_rate": 5.562638987230392e-05, "loss": 0.0201, "step": 4910 }, { "epoch": 12.0, "grad_norm": 0.48434099555015564, "learning_rate": 5.546206261394498e-05, "loss": 0.0174, "step": 4920 }, { "epoch": 12.024390243902438, "grad_norm": 0.38327252864837646, "learning_rate": 5.529767562334437e-05, "loss": 0.0208, "step": 4930 }, { "epoch": 12.048780487804878, "grad_norm": 0.26438337564468384, "learning_rate": 5.5133230698211926e-05, "loss": 0.021, "step": 4940 }, { "epoch": 12.073170731707316, "grad_norm": 0.4111882150173187, "learning_rate": 5.496872963689096e-05, "loss": 0.0197, "step": 4950 }, { "epoch": 12.097560975609756, "grad_norm": 0.21874411404132843, "learning_rate": 5.4804174238338756e-05, "loss": 0.0211, "step": 4960 }, { "epoch": 12.121951219512194, "grad_norm": 0.3088352382183075, "learning_rate": 5.463956630210678e-05, "loss": 0.021, "step": 4970 }, { "epoch": 12.146341463414634, "grad_norm": 0.24208848178386688, "learning_rate": 5.4474907628321046e-05, "loss": 0.0222, "step": 4980 }, { "epoch": 12.170731707317072, "grad_norm": 0.18094708025455475, "learning_rate": 5.431020001766244e-05, "loss": 0.021, "step": 4990 }, { "epoch": 12.195121951219512, "grad_norm": 0.20863591134548187, "learning_rate": 5.4145445271346986e-05, "loss": 0.0176, "step": 5000 }, { "epoch": 12.21951219512195, "grad_norm": 0.24203473329544067, "learning_rate": 5.398064519110622e-05, "loss": 0.0186, "step": 5010 }, { "epoch": 12.24390243902439, "grad_norm": 0.21313101053237915, "learning_rate": 5.3815801579167394e-05, "loss": 0.0209, "step": 5020 }, { "epoch": 12.268292682926829, "grad_norm": 0.20778658986091614, "learning_rate": 5.365091623823382e-05, "loss": 0.018, "step": 5030 }, { "epoch": 12.292682926829269, "grad_norm": 0.1912977546453476, "learning_rate": 5.348599097146521e-05, "loss": 0.0179, "step": 5040 }, { "epoch": 12.317073170731707, "grad_norm": 0.2596966624259949, "learning_rate": 5.3321027582457836e-05, "loss": 0.0174, "step": 5050 }, { "epoch": 12.341463414634147, "grad_norm": 0.22290438413619995, "learning_rate": 5.315602787522491e-05, "loss": 0.0191, "step": 5060 }, { "epoch": 12.365853658536585, "grad_norm": 0.18755024671554565, "learning_rate": 5.299099365417678e-05, "loss": 0.0172, "step": 5070 }, { "epoch": 12.390243902439025, "grad_norm": 0.18188652396202087, "learning_rate": 5.2825926724101236e-05, "loss": 0.0169, "step": 5080 }, { "epoch": 12.414634146341463, "grad_norm": 0.24289856851100922, "learning_rate": 5.26608288901438e-05, "loss": 0.0197, "step": 5090 }, { "epoch": 12.439024390243903, "grad_norm": 0.2156391441822052, "learning_rate": 5.24957019577879e-05, "loss": 0.0214, "step": 5100 }, { "epoch": 12.463414634146341, "grad_norm": 0.3585338890552521, "learning_rate": 5.2330547732835266e-05, "loss": 0.0215, "step": 5110 }, { "epoch": 12.487804878048781, "grad_norm": 0.23013228178024292, "learning_rate": 5.2165368021385996e-05, "loss": 0.0208, "step": 5120 }, { "epoch": 12.512195121951219, "grad_norm": 0.25682762265205383, "learning_rate": 5.200016462981897e-05, "loss": 0.0213, "step": 5130 }, { "epoch": 12.536585365853659, "grad_norm": 0.27759674191474915, "learning_rate": 5.1834939364772015e-05, "loss": 0.0169, "step": 5140 }, { "epoch": 12.560975609756097, "grad_norm": 0.25582432746887207, "learning_rate": 5.166969403312214e-05, "loss": 0.0158, "step": 5150 }, { "epoch": 12.585365853658537, "grad_norm": 0.26869234442710876, "learning_rate": 5.1504430441965844e-05, "loss": 0.0205, "step": 5160 }, { "epoch": 12.609756097560975, "grad_norm": 0.19383768737316132, "learning_rate": 5.133915039859923e-05, "loss": 0.0194, "step": 5170 }, { "epoch": 12.634146341463415, "grad_norm": 0.2720479667186737, "learning_rate": 5.1173855710498444e-05, "loss": 0.0148, "step": 5180 }, { "epoch": 12.658536585365853, "grad_norm": 0.21669845283031464, "learning_rate": 5.100854818529967e-05, "loss": 0.0195, "step": 5190 }, { "epoch": 12.682926829268293, "grad_norm": 0.2963488698005676, "learning_rate": 5.084322963077951e-05, "loss": 0.0195, "step": 5200 }, { "epoch": 12.707317073170731, "grad_norm": 0.24955260753631592, "learning_rate": 5.067790185483522e-05, "loss": 0.0172, "step": 5210 }, { "epoch": 12.731707317073171, "grad_norm": 0.29309767484664917, "learning_rate": 5.0512566665464844e-05, "loss": 0.0184, "step": 5220 }, { "epoch": 12.75609756097561, "grad_norm": 0.21202300488948822, "learning_rate": 5.034722587074755e-05, "loss": 0.0186, "step": 5230 }, { "epoch": 12.78048780487805, "grad_norm": 0.24876736104488373, "learning_rate": 5.018188127882375e-05, "loss": 0.0163, "step": 5240 }, { "epoch": 12.804878048780488, "grad_norm": 0.22734910249710083, "learning_rate": 5.0016534697875417e-05, "loss": 0.0166, "step": 5250 }, { "epoch": 12.829268292682928, "grad_norm": 0.1786053329706192, "learning_rate": 4.9851187936106294e-05, "loss": 0.0173, "step": 5260 }, { "epoch": 12.853658536585366, "grad_norm": 0.2518689036369324, "learning_rate": 4.968584280172206e-05, "loss": 0.0189, "step": 5270 }, { "epoch": 12.878048780487806, "grad_norm": 0.2367246001958847, "learning_rate": 4.95205011029106e-05, "loss": 0.0189, "step": 5280 }, { "epoch": 12.902439024390244, "grad_norm": 0.23083904385566711, "learning_rate": 4.935516464782227e-05, "loss": 0.0207, "step": 5290 }, { "epoch": 12.926829268292684, "grad_norm": 0.23726493120193481, "learning_rate": 4.918983524455003e-05, "loss": 0.018, "step": 5300 }, { "epoch": 12.951219512195122, "grad_norm": 0.18990957736968994, "learning_rate": 4.9024514701109766e-05, "loss": 0.0186, "step": 5310 }, { "epoch": 12.975609756097562, "grad_norm": 0.2555699348449707, "learning_rate": 4.885920482542043e-05, "loss": 0.0186, "step": 5320 }, { "epoch": 13.0, "grad_norm": 0.3495965003967285, "learning_rate": 4.869390742528438e-05, "loss": 0.0158, "step": 5330 }, { "epoch": 13.024390243902438, "grad_norm": 0.240470752120018, "learning_rate": 4.852862430836744e-05, "loss": 0.0183, "step": 5340 }, { "epoch": 13.048780487804878, "grad_norm": 0.2017982006072998, "learning_rate": 4.836335728217933e-05, "loss": 0.0165, "step": 5350 }, { "epoch": 13.073170731707316, "grad_norm": 0.2704800069332123, "learning_rate": 4.819810815405379e-05, "loss": 0.0195, "step": 5360 }, { "epoch": 13.097560975609756, "grad_norm": 0.25249332189559937, "learning_rate": 4.803287873112877e-05, "loss": 0.0193, "step": 5370 }, { "epoch": 13.121951219512194, "grad_norm": 0.2804461717605591, "learning_rate": 4.786767082032681e-05, "loss": 0.02, "step": 5380 }, { "epoch": 13.146341463414634, "grad_norm": 0.1664564460515976, "learning_rate": 4.77024862283351e-05, "loss": 0.0148, "step": 5390 }, { "epoch": 13.170731707317072, "grad_norm": 0.19321493804454803, "learning_rate": 4.753732676158593e-05, "loss": 0.0154, "step": 5400 }, { "epoch": 13.195121951219512, "grad_norm": 0.1905934363603592, "learning_rate": 4.737219422623672e-05, "loss": 0.0162, "step": 5410 }, { "epoch": 13.21951219512195, "grad_norm": 0.21867506206035614, "learning_rate": 4.720709042815044e-05, "loss": 0.0157, "step": 5420 }, { "epoch": 13.24390243902439, "grad_norm": 0.1973230242729187, "learning_rate": 4.704201717287578e-05, "loss": 0.0163, "step": 5430 }, { "epoch": 13.268292682926829, "grad_norm": 0.21163010597229004, "learning_rate": 4.6876976265627404e-05, "loss": 0.0144, "step": 5440 }, { "epoch": 13.292682926829269, "grad_norm": 0.2771659791469574, "learning_rate": 4.671196951126626e-05, "loss": 0.0209, "step": 5450 }, { "epoch": 13.317073170731707, "grad_norm": 0.28584933280944824, "learning_rate": 4.654699871427971e-05, "loss": 0.0178, "step": 5460 }, { "epoch": 13.341463414634147, "grad_norm": 0.28675249218940735, "learning_rate": 4.6382065678762034e-05, "loss": 0.0179, "step": 5470 }, { "epoch": 13.365853658536585, "grad_norm": 0.25331366062164307, "learning_rate": 4.6217172208394424e-05, "loss": 0.0176, "step": 5480 }, { "epoch": 13.390243902439025, "grad_norm": 0.38044095039367676, "learning_rate": 4.605232010642549e-05, "loss": 0.0181, "step": 5490 }, { "epoch": 13.414634146341463, "grad_norm": 0.12201747298240662, "learning_rate": 4.588751117565142e-05, "loss": 0.0161, "step": 5500 }, { "epoch": 13.439024390243903, "grad_norm": 0.23347687721252441, "learning_rate": 4.5722747218396214e-05, "loss": 0.0132, "step": 5510 }, { "epoch": 13.463414634146341, "grad_norm": 0.15709710121154785, "learning_rate": 4.5558030036492194e-05, "loss": 0.0167, "step": 5520 }, { "epoch": 13.487804878048781, "grad_norm": 0.13767477869987488, "learning_rate": 4.539336143125999e-05, "loss": 0.0154, "step": 5530 }, { "epoch": 13.512195121951219, "grad_norm": 0.2261703461408615, "learning_rate": 4.522874320348916e-05, "loss": 0.0202, "step": 5540 }, { "epoch": 13.536585365853659, "grad_norm": 0.3044319748878479, "learning_rate": 4.506417715341821e-05, "loss": 0.0142, "step": 5550 }, { "epoch": 13.560975609756097, "grad_norm": 0.19989942014217377, "learning_rate": 4.489966508071511e-05, "loss": 0.0168, "step": 5560 }, { "epoch": 13.585365853658537, "grad_norm": 0.23851622641086578, "learning_rate": 4.4735208784457575e-05, "loss": 0.0151, "step": 5570 }, { "epoch": 13.609756097560975, "grad_norm": 0.2146359533071518, "learning_rate": 4.457081006311325e-05, "loss": 0.0143, "step": 5580 }, { "epoch": 13.634146341463415, "grad_norm": 0.2587135136127472, "learning_rate": 4.440647071452027e-05, "loss": 0.016, "step": 5590 }, { "epoch": 13.658536585365853, "grad_norm": 0.32874974608421326, "learning_rate": 4.424219253586737e-05, "loss": 0.0209, "step": 5600 }, { "epoch": 13.682926829268293, "grad_norm": 0.2050141990184784, "learning_rate": 4.407797732367443e-05, "loss": 0.0185, "step": 5610 }, { "epoch": 13.707317073170731, "grad_norm": 0.32777833938598633, "learning_rate": 4.391382687377268e-05, "loss": 0.015, "step": 5620 }, { "epoch": 13.731707317073171, "grad_norm": 0.2871174216270447, "learning_rate": 4.374974298128512e-05, "loss": 0.0169, "step": 5630 }, { "epoch": 13.75609756097561, "grad_norm": 0.2602051794528961, "learning_rate": 4.358572744060699e-05, "loss": 0.017, "step": 5640 }, { "epoch": 13.78048780487805, "grad_norm": 0.19220633804798126, "learning_rate": 4.342178204538588e-05, "loss": 0.0194, "step": 5650 }, { "epoch": 13.804878048780488, "grad_norm": 0.2587337791919708, "learning_rate": 4.325790858850241e-05, "loss": 0.0139, "step": 5660 }, { "epoch": 13.829268292682928, "grad_norm": 0.18754558265209198, "learning_rate": 4.309410886205043e-05, "loss": 0.0189, "step": 5670 }, { "epoch": 13.853658536585366, "grad_norm": 0.214426651597023, "learning_rate": 4.293038465731752e-05, "loss": 0.0159, "step": 5680 }, { "epoch": 13.878048780487806, "grad_norm": 0.21553276479244232, "learning_rate": 4.276673776476533e-05, "loss": 0.0175, "step": 5690 }, { "epoch": 13.902439024390244, "grad_norm": 0.21011918783187866, "learning_rate": 4.260316997401007e-05, "loss": 0.022, "step": 5700 }, { "epoch": 13.926829268292684, "grad_norm": 0.2373148798942566, "learning_rate": 4.243968307380293e-05, "loss": 0.0153, "step": 5710 }, { "epoch": 13.951219512195122, "grad_norm": 0.29248884320259094, "learning_rate": 4.22762788520104e-05, "loss": 0.0228, "step": 5720 }, { "epoch": 13.975609756097562, "grad_norm": 0.3254092335700989, "learning_rate": 4.211295909559491e-05, "loss": 0.0155, "step": 5730 }, { "epoch": 14.0, "grad_norm": 0.357650488615036, "learning_rate": 4.194972559059511e-05, "loss": 0.0207, "step": 5740 }, { "epoch": 14.024390243902438, "grad_norm": 0.2461772859096527, "learning_rate": 4.178658012210651e-05, "loss": 0.0162, "step": 5750 }, { "epoch": 14.048780487804878, "grad_norm": 0.2035420835018158, "learning_rate": 4.162352447426177e-05, "loss": 0.0132, "step": 5760 }, { "epoch": 14.073170731707316, "grad_norm": 0.19651339948177338, "learning_rate": 4.146056043021135e-05, "loss": 0.0176, "step": 5770 }, { "epoch": 14.097560975609756, "grad_norm": 0.24692706763744354, "learning_rate": 4.1297689772103944e-05, "loss": 0.0143, "step": 5780 }, { "epoch": 14.121951219512194, "grad_norm": 0.20717059075832367, "learning_rate": 4.113491428106694e-05, "loss": 0.0172, "step": 5790 }, { "epoch": 14.146341463414634, "grad_norm": 0.24303783476352692, "learning_rate": 4.0972235737187055e-05, "loss": 0.0155, "step": 5800 }, { "epoch": 14.170731707317072, "grad_norm": 0.18483713269233704, "learning_rate": 4.080965591949076e-05, "loss": 0.0149, "step": 5810 }, { "epoch": 14.195121951219512, "grad_norm": 0.16167880594730377, "learning_rate": 4.0647176605924924e-05, "loss": 0.0164, "step": 5820 }, { "epoch": 14.21951219512195, "grad_norm": 0.2552972137928009, "learning_rate": 4.0484799573337255e-05, "loss": 0.016, "step": 5830 }, { "epoch": 14.24390243902439, "grad_norm": 0.21522434055805206, "learning_rate": 4.032252659745699e-05, "loss": 0.0161, "step": 5840 }, { "epoch": 14.268292682926829, "grad_norm": 0.35629358887672424, "learning_rate": 4.016035945287539e-05, "loss": 0.0175, "step": 5850 }, { "epoch": 14.292682926829269, "grad_norm": 0.2886747419834137, "learning_rate": 3.999829991302635e-05, "loss": 0.0255, "step": 5860 }, { "epoch": 14.317073170731707, "grad_norm": 0.23283573985099792, "learning_rate": 3.983634975016707e-05, "loss": 0.0157, "step": 5870 }, { "epoch": 14.341463414634147, "grad_norm": 0.2919616401195526, "learning_rate": 3.967451073535854e-05, "loss": 0.0145, "step": 5880 }, { "epoch": 14.365853658536585, "grad_norm": 0.22208772599697113, "learning_rate": 3.951278463844633e-05, "loss": 0.0232, "step": 5890 }, { "epoch": 14.390243902439025, "grad_norm": 0.16758930683135986, "learning_rate": 3.935117322804111e-05, "loss": 0.0152, "step": 5900 }, { "epoch": 14.414634146341463, "grad_norm": 0.3052211105823517, "learning_rate": 3.918967827149938e-05, "loss": 0.0182, "step": 5910 }, { "epoch": 14.439024390243903, "grad_norm": 0.2488154023885727, "learning_rate": 3.9028301534904094e-05, "loss": 0.0169, "step": 5920 }, { "epoch": 14.463414634146341, "grad_norm": 0.1735934168100357, "learning_rate": 3.88670447830454e-05, "loss": 0.0157, "step": 5930 }, { "epoch": 14.487804878048781, "grad_norm": 0.1864284723997116, "learning_rate": 3.870590977940132e-05, "loss": 0.0145, "step": 5940 }, { "epoch": 14.512195121951219, "grad_norm": 0.31341204047203064, "learning_rate": 3.8544898286118404e-05, "loss": 0.0152, "step": 5950 }, { "epoch": 14.536585365853659, "grad_norm": 0.3601032495498657, "learning_rate": 3.838401206399257e-05, "loss": 0.0145, "step": 5960 }, { "epoch": 14.560975609756097, "grad_norm": 0.42088907957077026, "learning_rate": 3.822325287244975e-05, "loss": 0.0145, "step": 5970 }, { "epoch": 14.585365853658537, "grad_norm": 0.14860713481903076, "learning_rate": 3.8062622469526725e-05, "loss": 0.0157, "step": 5980 }, { "epoch": 14.609756097560975, "grad_norm": 0.3302834928035736, "learning_rate": 3.790212261185183e-05, "loss": 0.0186, "step": 5990 }, { "epoch": 14.634146341463415, "grad_norm": 0.30182769894599915, "learning_rate": 3.7741755054625794e-05, "loss": 0.0145, "step": 6000 }, { "epoch": 14.658536585365853, "grad_norm": 0.21025842428207397, "learning_rate": 3.758152155160255e-05, "loss": 0.0147, "step": 6010 }, { "epoch": 14.682926829268293, "grad_norm": 0.22440677881240845, "learning_rate": 3.742142385506999e-05, "loss": 0.0172, "step": 6020 }, { "epoch": 14.707317073170731, "grad_norm": 0.22785454988479614, "learning_rate": 3.72614637158309e-05, "loss": 0.0163, "step": 6030 }, { "epoch": 14.731707317073171, "grad_norm": 0.29240235686302185, "learning_rate": 3.710164288318371e-05, "loss": 0.0149, "step": 6040 }, { "epoch": 14.75609756097561, "grad_norm": 0.24598218500614166, "learning_rate": 3.694196310490345e-05, "loss": 0.0157, "step": 6050 }, { "epoch": 14.78048780487805, "grad_norm": 0.20531202852725983, "learning_rate": 3.678242612722259e-05, "loss": 0.0136, "step": 6060 }, { "epoch": 14.804878048780488, "grad_norm": 0.2538522481918335, "learning_rate": 3.6623033694811953e-05, "loss": 0.0135, "step": 6070 }, { "epoch": 14.829268292682928, "grad_norm": 0.354500412940979, "learning_rate": 3.6463787550761665e-05, "loss": 0.0168, "step": 6080 }, { "epoch": 14.853658536585366, "grad_norm": 0.18036897480487823, "learning_rate": 3.630468943656202e-05, "loss": 0.018, "step": 6090 }, { "epoch": 14.878048780487806, "grad_norm": 0.23048214614391327, "learning_rate": 3.6145741092084523e-05, "loss": 0.0182, "step": 6100 }, { "epoch": 14.902439024390244, "grad_norm": 0.25603458285331726, "learning_rate": 3.598694425556278e-05, "loss": 0.0136, "step": 6110 }, { "epoch": 14.926829268292684, "grad_norm": 0.23654544353485107, "learning_rate": 3.58283006635736e-05, "loss": 0.013, "step": 6120 }, { "epoch": 14.951219512195122, "grad_norm": 0.1820974200963974, "learning_rate": 3.566981205101781e-05, "loss": 0.0161, "step": 6130 }, { "epoch": 14.975609756097562, "grad_norm": 0.239983469247818, "learning_rate": 3.5511480151101556e-05, "loss": 0.0151, "step": 6140 }, { "epoch": 15.0, "grad_norm": 0.18575075268745422, "learning_rate": 3.5353306695317104e-05, "loss": 0.0154, "step": 6150 }, { "epoch": 15.024390243902438, "grad_norm": 0.20420274138450623, "learning_rate": 3.519529341342402e-05, "loss": 0.0149, "step": 6160 }, { "epoch": 15.048780487804878, "grad_norm": 0.1670485883951187, "learning_rate": 3.503744203343026e-05, "loss": 0.0145, "step": 6170 }, { "epoch": 15.073170731707316, "grad_norm": 0.1589074581861496, "learning_rate": 3.487975428157318e-05, "loss": 0.0165, "step": 6180 }, { "epoch": 15.097560975609756, "grad_norm": 0.23744134604930878, "learning_rate": 3.472223188230083e-05, "loss": 0.0177, "step": 6190 }, { "epoch": 15.121951219512194, "grad_norm": 0.19014030694961548, "learning_rate": 3.4564876558252866e-05, "loss": 0.0164, "step": 6200 }, { "epoch": 15.146341463414634, "grad_norm": 0.19612963497638702, "learning_rate": 3.440769003024195e-05, "loss": 0.0127, "step": 6210 }, { "epoch": 15.170731707317072, "grad_norm": 0.23854143917560577, "learning_rate": 3.425067401723477e-05, "loss": 0.0156, "step": 6220 }, { "epoch": 15.195121951219512, "grad_norm": 0.20270146429538727, "learning_rate": 3.409383023633325e-05, "loss": 0.014, "step": 6230 }, { "epoch": 15.21951219512195, "grad_norm": 0.17673933506011963, "learning_rate": 3.3937160402755894e-05, "loss": 0.0146, "step": 6240 }, { "epoch": 15.24390243902439, "grad_norm": 0.17498810589313507, "learning_rate": 3.378066622981885e-05, "loss": 0.0121, "step": 6250 }, { "epoch": 15.268292682926829, "grad_norm": 0.1387321501970291, "learning_rate": 3.362434942891738e-05, "loss": 0.0192, "step": 6260 }, { "epoch": 15.292682926829269, "grad_norm": 0.1843465268611908, "learning_rate": 3.346821170950693e-05, "loss": 0.0158, "step": 6270 }, { "epoch": 15.317073170731707, "grad_norm": 0.19278454780578613, "learning_rate": 3.3312254779084585e-05, "loss": 0.0135, "step": 6280 }, { "epoch": 15.341463414634147, "grad_norm": 0.2112695872783661, "learning_rate": 3.315648034317039e-05, "loss": 0.0187, "step": 6290 }, { "epoch": 15.365853658536585, "grad_norm": 0.3191860020160675, "learning_rate": 3.3000890105288564e-05, "loss": 0.016, "step": 6300 }, { "epoch": 15.390243902439025, "grad_norm": 0.2431771457195282, "learning_rate": 3.284548576694908e-05, "loss": 0.0154, "step": 6310 }, { "epoch": 15.414634146341463, "grad_norm": 0.0985310897231102, "learning_rate": 3.2690269027628815e-05, "loss": 0.0149, "step": 6320 }, { "epoch": 15.439024390243903, "grad_norm": 0.29152584075927734, "learning_rate": 3.253524158475324e-05, "loss": 0.0134, "step": 6330 }, { "epoch": 15.463414634146341, "grad_norm": 0.287227600812912, "learning_rate": 3.238040513367757e-05, "loss": 0.0169, "step": 6340 }, { "epoch": 15.487804878048781, "grad_norm": 0.23408304154872894, "learning_rate": 3.222576136766843e-05, "loss": 0.0114, "step": 6350 }, { "epoch": 15.512195121951219, "grad_norm": 0.24744589626789093, "learning_rate": 3.2071311977885324e-05, "loss": 0.0128, "step": 6360 }, { "epoch": 15.536585365853659, "grad_norm": 0.28945887088775635, "learning_rate": 3.191705865336197e-05, "loss": 0.018, "step": 6370 }, { "epoch": 15.560975609756097, "grad_norm": 0.24864605069160461, "learning_rate": 3.1763003080988075e-05, "loss": 0.016, "step": 6380 }, { "epoch": 15.585365853658537, "grad_norm": 0.17159052193164825, "learning_rate": 3.160914694549063e-05, "loss": 0.0134, "step": 6390 }, { "epoch": 15.609756097560975, "grad_norm": 0.22714361548423767, "learning_rate": 3.145549192941573e-05, "loss": 0.0143, "step": 6400 }, { "epoch": 15.634146341463415, "grad_norm": 0.26719456911087036, "learning_rate": 3.130203971310999e-05, "loss": 0.0142, "step": 6410 }, { "epoch": 15.658536585365853, "grad_norm": 0.20024561882019043, "learning_rate": 3.114879197470225e-05, "loss": 0.0165, "step": 6420 }, { "epoch": 15.682926829268293, "grad_norm": 0.26084479689598083, "learning_rate": 3.0995750390085285e-05, "loss": 0.0151, "step": 6430 }, { "epoch": 15.707317073170731, "grad_norm": 0.20982839167118073, "learning_rate": 3.084291663289728e-05, "loss": 0.0139, "step": 6440 }, { "epoch": 15.731707317073171, "grad_norm": 0.1374421864748001, "learning_rate": 3.069029237450375e-05, "loss": 0.0128, "step": 6450 }, { "epoch": 15.75609756097561, "grad_norm": 0.17894846200942993, "learning_rate": 3.053787928397911e-05, "loss": 0.0115, "step": 6460 }, { "epoch": 15.78048780487805, "grad_norm": 0.24954095482826233, "learning_rate": 3.0385679028088526e-05, "loss": 0.0145, "step": 6470 }, { "epoch": 15.804878048780488, "grad_norm": 0.17149020731449127, "learning_rate": 3.023369327126959e-05, "loss": 0.0187, "step": 6480 }, { "epoch": 15.829268292682928, "grad_norm": 0.22541311383247375, "learning_rate": 3.0081923675614198e-05, "loss": 0.0102, "step": 6490 }, { "epoch": 15.853658536585366, "grad_norm": 0.20929937064647675, "learning_rate": 2.993037190085034e-05, "loss": 0.013, "step": 6500 }, { "epoch": 15.878048780487806, "grad_norm": 0.13958941400051117, "learning_rate": 2.977903960432392e-05, "loss": 0.0109, "step": 6510 }, { "epoch": 15.902439024390244, "grad_norm": 0.2834900915622711, "learning_rate": 2.9627928440980722e-05, "loss": 0.0159, "step": 6520 }, { "epoch": 15.926829268292684, "grad_norm": 0.19795532524585724, "learning_rate": 2.9477040063348183e-05, "loss": 0.0177, "step": 6530 }, { "epoch": 15.951219512195122, "grad_norm": 0.26454272866249084, "learning_rate": 2.9326376121517456e-05, "loss": 0.0147, "step": 6540 }, { "epoch": 15.975609756097562, "grad_norm": 0.1460547000169754, "learning_rate": 2.9175938263125236e-05, "loss": 0.0168, "step": 6550 }, { "epoch": 16.0, "grad_norm": 0.18856273591518402, "learning_rate": 2.9025728133335873e-05, "loss": 0.0134, "step": 6560 }, { "epoch": 16.024390243902438, "grad_norm": 0.31814947724342346, "learning_rate": 2.8875747374823288e-05, "loss": 0.0136, "step": 6570 }, { "epoch": 16.048780487804876, "grad_norm": 0.22137674689292908, "learning_rate": 2.872599762775298e-05, "loss": 0.0175, "step": 6580 }, { "epoch": 16.073170731707318, "grad_norm": 0.20111139118671417, "learning_rate": 2.857648052976425e-05, "loss": 0.0217, "step": 6590 }, { "epoch": 16.097560975609756, "grad_norm": 0.28660285472869873, "learning_rate": 2.8427197715952047e-05, "loss": 0.0123, "step": 6600 }, { "epoch": 16.121951219512194, "grad_norm": 0.27037322521209717, "learning_rate": 2.8278150818849393e-05, "loss": 0.0131, "step": 6610 }, { "epoch": 16.146341463414632, "grad_norm": 0.20074114203453064, "learning_rate": 2.812934146840922e-05, "loss": 0.0141, "step": 6620 }, { "epoch": 16.170731707317074, "grad_norm": 0.23057211935520172, "learning_rate": 2.7980771291986764e-05, "loss": 0.0116, "step": 6630 }, { "epoch": 16.195121951219512, "grad_norm": 0.2560122013092041, "learning_rate": 2.783244191432167e-05, "loss": 0.0161, "step": 6640 }, { "epoch": 16.21951219512195, "grad_norm": 0.19868651032447815, "learning_rate": 2.768435495752022e-05, "loss": 0.0169, "step": 6650 }, { "epoch": 16.24390243902439, "grad_norm": 0.2207840532064438, "learning_rate": 2.753651204103771e-05, "loss": 0.0112, "step": 6660 }, { "epoch": 16.26829268292683, "grad_norm": 0.15919135510921478, "learning_rate": 2.7388914781660523e-05, "loss": 0.0139, "step": 6670 }, { "epoch": 16.29268292682927, "grad_norm": 0.19716724753379822, "learning_rate": 2.7241564793488693e-05, "loss": 0.0139, "step": 6680 }, { "epoch": 16.317073170731707, "grad_norm": 0.20820879936218262, "learning_rate": 2.7094463687918037e-05, "loss": 0.0147, "step": 6690 }, { "epoch": 16.341463414634145, "grad_norm": 0.1732071489095688, "learning_rate": 2.694761307362268e-05, "loss": 0.0128, "step": 6700 }, { "epoch": 16.365853658536587, "grad_norm": 0.14973878860473633, "learning_rate": 2.6801014556537467e-05, "loss": 0.0117, "step": 6710 }, { "epoch": 16.390243902439025, "grad_norm": 0.16591908037662506, "learning_rate": 2.6654669739840243e-05, "loss": 0.0138, "step": 6720 }, { "epoch": 16.414634146341463, "grad_norm": 0.18221981823444366, "learning_rate": 2.650858022393451e-05, "loss": 0.0155, "step": 6730 }, { "epoch": 16.4390243902439, "grad_norm": 0.2837972640991211, "learning_rate": 2.6362747606431747e-05, "loss": 0.0149, "step": 6740 }, { "epoch": 16.463414634146343, "grad_norm": 0.17040933668613434, "learning_rate": 2.6217173482134172e-05, "loss": 0.0152, "step": 6750 }, { "epoch": 16.48780487804878, "grad_norm": 0.19737546145915985, "learning_rate": 2.6071859443017044e-05, "loss": 0.0117, "step": 6760 }, { "epoch": 16.51219512195122, "grad_norm": 0.23664474487304688, "learning_rate": 2.5926807078211414e-05, "loss": 0.0137, "step": 6770 }, { "epoch": 16.536585365853657, "grad_norm": 0.16656813025474548, "learning_rate": 2.5782017973986728e-05, "loss": 0.0113, "step": 6780 }, { "epoch": 16.5609756097561, "grad_norm": 0.1537283957004547, "learning_rate": 2.5637493713733374e-05, "loss": 0.0116, "step": 6790 }, { "epoch": 16.585365853658537, "grad_norm": 0.1819530874490738, "learning_rate": 2.549323587794559e-05, "loss": 0.0137, "step": 6800 }, { "epoch": 16.609756097560975, "grad_norm": 0.18603681027889252, "learning_rate": 2.5349246044203895e-05, "loss": 0.0166, "step": 6810 }, { "epoch": 16.634146341463413, "grad_norm": 0.2403397560119629, "learning_rate": 2.520552578715808e-05, "loss": 0.0122, "step": 6820 }, { "epoch": 16.658536585365855, "grad_norm": 0.19402626156806946, "learning_rate": 2.506207667850981e-05, "loss": 0.0099, "step": 6830 }, { "epoch": 16.682926829268293, "grad_norm": 0.2495782971382141, "learning_rate": 2.4918900286995555e-05, "loss": 0.0126, "step": 6840 }, { "epoch": 16.70731707317073, "grad_norm": 0.1913372129201889, "learning_rate": 2.4775998178369458e-05, "loss": 0.011, "step": 6850 }, { "epoch": 16.73170731707317, "grad_norm": 0.2134193629026413, "learning_rate": 2.4633371915386017e-05, "loss": 0.0127, "step": 6860 }, { "epoch": 16.75609756097561, "grad_norm": 0.20887625217437744, "learning_rate": 2.4491023057783235e-05, "loss": 0.0158, "step": 6870 }, { "epoch": 16.78048780487805, "grad_norm": 0.2741149961948395, "learning_rate": 2.4348953162265375e-05, "loss": 0.0142, "step": 6880 }, { "epoch": 16.804878048780488, "grad_norm": 0.22255656123161316, "learning_rate": 2.420716378248607e-05, "loss": 0.0177, "step": 6890 }, { "epoch": 16.829268292682926, "grad_norm": 0.23807673156261444, "learning_rate": 2.4065656469031266e-05, "loss": 0.0121, "step": 6900 }, { "epoch": 16.853658536585368, "grad_norm": 0.3277212977409363, "learning_rate": 2.3924432769402268e-05, "loss": 0.0141, "step": 6910 }, { "epoch": 16.878048780487806, "grad_norm": 0.3649060130119324, "learning_rate": 2.3783494227998844e-05, "loss": 0.0124, "step": 6920 }, { "epoch": 16.902439024390244, "grad_norm": 0.17507562041282654, "learning_rate": 2.3642842386102264e-05, "loss": 0.0171, "step": 6930 }, { "epoch": 16.926829268292682, "grad_norm": 0.26026996970176697, "learning_rate": 2.3502478781858567e-05, "loss": 0.012, "step": 6940 }, { "epoch": 16.951219512195124, "grad_norm": 0.24546192586421967, "learning_rate": 2.3362404950261628e-05, "loss": 0.012, "step": 6950 }, { "epoch": 16.975609756097562, "grad_norm": 0.20729102194309235, "learning_rate": 2.3222622423136458e-05, "loss": 0.0134, "step": 6960 }, { "epoch": 17.0, "grad_norm": 0.18805541098117828, "learning_rate": 2.3083132729122332e-05, "loss": 0.0128, "step": 6970 }, { "epoch": 17.024390243902438, "grad_norm": 0.2050151228904724, "learning_rate": 2.294393739365621e-05, "loss": 0.0131, "step": 6980 }, { "epoch": 17.048780487804876, "grad_norm": 0.1572803109884262, "learning_rate": 2.2805037938956e-05, "loss": 0.0113, "step": 6990 }, { "epoch": 17.073170731707318, "grad_norm": 0.13220667839050293, "learning_rate": 2.266643588400386e-05, "loss": 0.0135, "step": 7000 }, { "epoch": 17.097560975609756, "grad_norm": 0.2161470651626587, "learning_rate": 2.252813274452969e-05, "loss": 0.0116, "step": 7010 }, { "epoch": 17.121951219512194, "grad_norm": 0.24905510246753693, "learning_rate": 2.2390130032994427e-05, "loss": 0.0121, "step": 7020 }, { "epoch": 17.146341463414632, "grad_norm": 0.16867220401763916, "learning_rate": 2.2252429258573633e-05, "loss": 0.0098, "step": 7030 }, { "epoch": 17.170731707317074, "grad_norm": 0.22317847609519958, "learning_rate": 2.2115031927140904e-05, "loss": 0.0137, "step": 7040 }, { "epoch": 17.195121951219512, "grad_norm": 0.15547192096710205, "learning_rate": 2.1977939541251463e-05, "loss": 0.0131, "step": 7050 }, { "epoch": 17.21951219512195, "grad_norm": 0.20826084911823273, "learning_rate": 2.1841153600125684e-05, "loss": 0.0176, "step": 7060 }, { "epoch": 17.24390243902439, "grad_norm": 0.22580330073833466, "learning_rate": 2.170467559963267e-05, "loss": 0.0132, "step": 7070 }, { "epoch": 17.26829268292683, "grad_norm": 0.1746339648962021, "learning_rate": 2.1568507032273982e-05, "loss": 0.0136, "step": 7080 }, { "epoch": 17.29268292682927, "grad_norm": 0.16227588057518005, "learning_rate": 2.1432649387167264e-05, "loss": 0.0091, "step": 7090 }, { "epoch": 17.317073170731707, "grad_norm": 0.2125353217124939, "learning_rate": 2.1297104150029973e-05, "loss": 0.0119, "step": 7100 }, { "epoch": 17.341463414634145, "grad_norm": 0.12792453169822693, "learning_rate": 2.116187280316307e-05, "loss": 0.0119, "step": 7110 }, { "epoch": 17.365853658536587, "grad_norm": 0.22305063903331757, "learning_rate": 2.1026956825434908e-05, "loss": 0.0135, "step": 7120 }, { "epoch": 17.390243902439025, "grad_norm": 0.2930857241153717, "learning_rate": 2.0892357692265017e-05, "loss": 0.0135, "step": 7130 }, { "epoch": 17.414634146341463, "grad_norm": 0.24668855965137482, "learning_rate": 2.0758076875607947e-05, "loss": 0.0127, "step": 7140 }, { "epoch": 17.4390243902439, "grad_norm": 0.3125517964363098, "learning_rate": 2.0624115843937207e-05, "loss": 0.0135, "step": 7150 }, { "epoch": 17.463414634146343, "grad_norm": 0.19358424842357635, "learning_rate": 2.0490476062229157e-05, "loss": 0.0146, "step": 7160 }, { "epoch": 17.48780487804878, "grad_norm": 0.167482390999794, "learning_rate": 2.035715899194704e-05, "loss": 0.0128, "step": 7170 }, { "epoch": 17.51219512195122, "grad_norm": 0.2172875553369522, "learning_rate": 2.022416609102499e-05, "loss": 0.0123, "step": 7180 }, { "epoch": 17.536585365853657, "grad_norm": 0.17573276162147522, "learning_rate": 2.009149881385205e-05, "loss": 0.013, "step": 7190 }, { "epoch": 17.5609756097561, "grad_norm": 0.12452840059995651, "learning_rate": 1.995915861125634e-05, "loss": 0.0169, "step": 7200 }, { "epoch": 17.585365853658537, "grad_norm": 0.17774614691734314, "learning_rate": 1.9827146930489065e-05, "loss": 0.0108, "step": 7210 }, { "epoch": 17.609756097560975, "grad_norm": 0.191396564245224, "learning_rate": 1.9695465215208848e-05, "loss": 0.0085, "step": 7220 }, { "epoch": 17.634146341463413, "grad_norm": 0.13030576705932617, "learning_rate": 1.9564114905465813e-05, "loss": 0.0154, "step": 7230 }, { "epoch": 17.658536585365855, "grad_norm": 0.2462076097726822, "learning_rate": 1.9433097437685936e-05, "loss": 0.01, "step": 7240 }, { "epoch": 17.682926829268293, "grad_norm": 0.19551673531532288, "learning_rate": 1.930241424465521e-05, "loss": 0.013, "step": 7250 }, { "epoch": 17.70731707317073, "grad_norm": 0.15283554792404175, "learning_rate": 1.9172066755504115e-05, "loss": 0.0109, "step": 7260 }, { "epoch": 17.73170731707317, "grad_norm": 0.22944970428943634, "learning_rate": 1.9042056395691914e-05, "loss": 0.0128, "step": 7270 }, { "epoch": 17.75609756097561, "grad_norm": 0.2392381727695465, "learning_rate": 1.8912384586991066e-05, "loss": 0.0127, "step": 7280 }, { "epoch": 17.78048780487805, "grad_norm": 0.3608357906341553, "learning_rate": 1.8783052747471717e-05, "loss": 0.0115, "step": 7290 }, { "epoch": 17.804878048780488, "grad_norm": 0.18236228823661804, "learning_rate": 1.865406229148611e-05, "loss": 0.0123, "step": 7300 }, { "epoch": 17.829268292682926, "grad_norm": 0.1398136168718338, "learning_rate": 1.8525414629653233e-05, "loss": 0.0144, "step": 7310 }, { "epoch": 17.853658536585368, "grad_norm": 0.13034047186374664, "learning_rate": 1.8397111168843255e-05, "loss": 0.0148, "step": 7320 }, { "epoch": 17.878048780487806, "grad_norm": 0.25905653834342957, "learning_rate": 1.8269153312162323e-05, "loss": 0.0114, "step": 7330 }, { "epoch": 17.902439024390244, "grad_norm": 0.13741643726825714, "learning_rate": 1.8141542458937054e-05, "loss": 0.0125, "step": 7340 }, { "epoch": 17.926829268292682, "grad_norm": 0.16979466378688812, "learning_rate": 1.8014280004699268e-05, "loss": 0.0123, "step": 7350 }, { "epoch": 17.951219512195124, "grad_norm": 0.1729811280965805, "learning_rate": 1.788736734117078e-05, "loss": 0.0104, "step": 7360 }, { "epoch": 17.975609756097562, "grad_norm": 0.1397097110748291, "learning_rate": 1.7760805856248152e-05, "loss": 0.0115, "step": 7370 }, { "epoch": 18.0, "grad_norm": 0.3280600905418396, "learning_rate": 1.7634596933987518e-05, "loss": 0.0106, "step": 7380 }, { "epoch": 18.024390243902438, "grad_norm": 0.2011936902999878, "learning_rate": 1.7508741954589404e-05, "loss": 0.0132, "step": 7390 }, { "epoch": 18.048780487804876, "grad_norm": 0.18528711795806885, "learning_rate": 1.7383242294383717e-05, "loss": 0.0137, "step": 7400 }, { "epoch": 18.073170731707318, "grad_norm": 0.23939597606658936, "learning_rate": 1.7258099325814632e-05, "loss": 0.0123, "step": 7410 }, { "epoch": 18.097560975609756, "grad_norm": 0.17862501740455627, "learning_rate": 1.7133314417425594e-05, "loss": 0.0102, "step": 7420 }, { "epoch": 18.121951219512194, "grad_norm": 0.2118883728981018, "learning_rate": 1.7008888933844408e-05, "loss": 0.013, "step": 7430 }, { "epoch": 18.146341463414632, "grad_norm": 0.1714843213558197, "learning_rate": 1.6884824235768172e-05, "loss": 0.0095, "step": 7440 }, { "epoch": 18.170731707317074, "grad_norm": 0.19325602054595947, "learning_rate": 1.6761121679948592e-05, "loss": 0.0147, "step": 7450 }, { "epoch": 18.195121951219512, "grad_norm": 0.22706739604473114, "learning_rate": 1.663778261917695e-05, "loss": 0.0098, "step": 7460 }, { "epoch": 18.21951219512195, "grad_norm": 0.1989172101020813, "learning_rate": 1.651480840226952e-05, "loss": 0.0134, "step": 7470 }, { "epoch": 18.24390243902439, "grad_norm": 0.2651884853839874, "learning_rate": 1.639220037405258e-05, "loss": 0.0111, "step": 7480 }, { "epoch": 18.26829268292683, "grad_norm": 0.1672966331243515, "learning_rate": 1.6269959875347906e-05, "loss": 0.0122, "step": 7490 }, { "epoch": 18.29268292682927, "grad_norm": 0.1557215452194214, "learning_rate": 1.614808824295802e-05, "loss": 0.0129, "step": 7500 }, { "epoch": 18.317073170731707, "grad_norm": 0.10942487418651581, "learning_rate": 1.602658680965152e-05, "loss": 0.0127, "step": 7510 }, { "epoch": 18.341463414634145, "grad_norm": 0.18013854324817657, "learning_rate": 1.5905456904148686e-05, "loss": 0.009, "step": 7520 }, { "epoch": 18.365853658536587, "grad_norm": 0.14081823825836182, "learning_rate": 1.57846998511067e-05, "loss": 0.0159, "step": 7530 }, { "epoch": 18.390243902439025, "grad_norm": 0.21240636706352234, "learning_rate": 1.566431697110538e-05, "loss": 0.0136, "step": 7540 }, { "epoch": 18.414634146341463, "grad_norm": 0.11294695734977722, "learning_rate": 1.554430958063259e-05, "loss": 0.0104, "step": 7550 }, { "epoch": 18.4390243902439, "grad_norm": 0.18327893316745758, "learning_rate": 1.5424678992069912e-05, "loss": 0.0104, "step": 7560 }, { "epoch": 18.463414634146343, "grad_norm": 0.2680053114891052, "learning_rate": 1.5305426513678362e-05, "loss": 0.011, "step": 7570 }, { "epoch": 18.48780487804878, "grad_norm": 0.17958179116249084, "learning_rate": 1.518655344958388e-05, "loss": 0.0138, "step": 7580 }, { "epoch": 18.51219512195122, "grad_norm": 0.20701059699058533, "learning_rate": 1.5068061099763275e-05, "loss": 0.0147, "step": 7590 }, { "epoch": 18.536585365853657, "grad_norm": 0.24321326613426208, "learning_rate": 1.494995076002988e-05, "loss": 0.0128, "step": 7600 }, { "epoch": 18.5609756097561, "grad_norm": 0.19996492564678192, "learning_rate": 1.4832223722019456e-05, "loss": 0.0125, "step": 7610 }, { "epoch": 18.585365853658537, "grad_norm": 0.21097694337368011, "learning_rate": 1.4714881273176035e-05, "loss": 0.0101, "step": 7620 }, { "epoch": 18.609756097560975, "grad_norm": 0.14146944880485535, "learning_rate": 1.4597924696737835e-05, "loss": 0.0114, "step": 7630 }, { "epoch": 18.634146341463413, "grad_norm": 0.13050012290477753, "learning_rate": 1.4481355271723252e-05, "loss": 0.0113, "step": 7640 }, { "epoch": 18.658536585365855, "grad_norm": 0.18829785287380219, "learning_rate": 1.4365174272916809e-05, "loss": 0.0107, "step": 7650 }, { "epoch": 18.682926829268293, "grad_norm": 0.23382699489593506, "learning_rate": 1.4249382970855319e-05, "loss": 0.0135, "step": 7660 }, { "epoch": 18.70731707317073, "grad_norm": 0.14902684092521667, "learning_rate": 1.4133982631813903e-05, "loss": 0.0103, "step": 7670 }, { "epoch": 18.73170731707317, "grad_norm": 0.20700950920581818, "learning_rate": 1.4018974517792194e-05, "loss": 0.0138, "step": 7680 }, { "epoch": 18.75609756097561, "grad_norm": 0.11210900545120239, "learning_rate": 1.390435988650048e-05, "loss": 0.0088, "step": 7690 }, { "epoch": 18.78048780487805, "grad_norm": 0.22468966245651245, "learning_rate": 1.3790139991346006e-05, "loss": 0.0096, "step": 7700 }, { "epoch": 18.804878048780488, "grad_norm": 0.2505734860897064, "learning_rate": 1.367631608141926e-05, "loss": 0.0112, "step": 7710 }, { "epoch": 18.829268292682926, "grad_norm": 0.21094031631946564, "learning_rate": 1.3562889401480278e-05, "loss": 0.0107, "step": 7720 }, { "epoch": 18.853658536585368, "grad_norm": 0.16165322065353394, "learning_rate": 1.3449861191945074e-05, "loss": 0.0106, "step": 7730 }, { "epoch": 18.878048780487806, "grad_norm": 0.18205440044403076, "learning_rate": 1.3337232688872009e-05, "loss": 0.0128, "step": 7740 }, { "epoch": 18.902439024390244, "grad_norm": 0.1979338526725769, "learning_rate": 1.3225005123948364e-05, "loss": 0.0096, "step": 7750 }, { "epoch": 18.926829268292682, "grad_norm": 0.1613461822271347, "learning_rate": 1.311317972447681e-05, "loss": 0.0102, "step": 7760 }, { "epoch": 18.951219512195124, "grad_norm": 0.26150792837142944, "learning_rate": 1.3001757713361996e-05, "loss": 0.0105, "step": 7770 }, { "epoch": 18.975609756097562, "grad_norm": 0.23715780675411224, "learning_rate": 1.2890740309097204e-05, "loss": 0.0085, "step": 7780 }, { "epoch": 19.0, "grad_norm": 0.11350850015878677, "learning_rate": 1.2780128725750944e-05, "loss": 0.0113, "step": 7790 }, { "epoch": 19.024390243902438, "grad_norm": 0.16082502901554108, "learning_rate": 1.266992417295379e-05, "loss": 0.0092, "step": 7800 }, { "epoch": 19.048780487804876, "grad_norm": 0.18362541496753693, "learning_rate": 1.2560127855885073e-05, "loss": 0.0125, "step": 7810 }, { "epoch": 19.073170731707318, "grad_norm": 0.17274896800518036, "learning_rate": 1.2450740975259745e-05, "loss": 0.0106, "step": 7820 }, { "epoch": 19.097560975609756, "grad_norm": 0.2610225975513458, "learning_rate": 1.234176472731517e-05, "loss": 0.0137, "step": 7830 }, { "epoch": 19.121951219512194, "grad_norm": 0.13999223709106445, "learning_rate": 1.2233200303798158e-05, "loss": 0.0156, "step": 7840 }, { "epoch": 19.146341463414632, "grad_norm": 0.15542002022266388, "learning_rate": 1.2125048891951846e-05, "loss": 0.0099, "step": 7850 }, { "epoch": 19.170731707317074, "grad_norm": 0.13463877141475677, "learning_rate": 1.2017311674502745e-05, "loss": 0.0106, "step": 7860 }, { "epoch": 19.195121951219512, "grad_norm": 0.16056817770004272, "learning_rate": 1.1909989829647822e-05, "loss": 0.0088, "step": 7870 }, { "epoch": 19.21951219512195, "grad_norm": 0.1865416169166565, "learning_rate": 1.1803084531041553e-05, "loss": 0.0094, "step": 7880 }, { "epoch": 19.24390243902439, "grad_norm": 0.18612182140350342, "learning_rate": 1.1696596947783162e-05, "loss": 0.0114, "step": 7890 }, { "epoch": 19.26829268292683, "grad_norm": 0.22550585865974426, "learning_rate": 1.1590528244403803e-05, "loss": 0.0104, "step": 7900 }, { "epoch": 19.29268292682927, "grad_norm": 0.10861923545598984, "learning_rate": 1.148487958085382e-05, "loss": 0.0096, "step": 7910 }, { "epoch": 19.317073170731707, "grad_norm": 0.17277413606643677, "learning_rate": 1.1379652112490086e-05, "loss": 0.0113, "step": 7920 }, { "epoch": 19.341463414634145, "grad_norm": 0.15615837275981903, "learning_rate": 1.1274846990063315e-05, "loss": 0.0113, "step": 7930 }, { "epoch": 19.365853658536587, "grad_norm": 0.1589335948228836, "learning_rate": 1.117046535970554e-05, "loss": 0.0124, "step": 7940 }, { "epoch": 19.390243902439025, "grad_norm": 0.2003336250782013, "learning_rate": 1.106650836291755e-05, "loss": 0.0128, "step": 7950 }, { "epoch": 19.414634146341463, "grad_norm": 0.12606054544448853, "learning_rate": 1.0962977136556418e-05, "loss": 0.015, "step": 7960 }, { "epoch": 19.4390243902439, "grad_norm": 0.1572599858045578, "learning_rate": 1.0859872812823024e-05, "loss": 0.0148, "step": 7970 }, { "epoch": 19.463414634146343, "grad_norm": 0.11029129475355148, "learning_rate": 1.0757196519249747e-05, "loss": 0.0107, "step": 7980 }, { "epoch": 19.48780487804878, "grad_norm": 0.2257085144519806, "learning_rate": 1.0654949378688077e-05, "loss": 0.0191, "step": 7990 }, { "epoch": 19.51219512195122, "grad_norm": 0.17479510605335236, "learning_rate": 1.0553132509296376e-05, "loss": 0.012, "step": 8000 }, { "epoch": 19.536585365853657, "grad_norm": 0.1896139234304428, "learning_rate": 1.0451747024527613e-05, "loss": 0.0111, "step": 8010 }, { "epoch": 19.5609756097561, "grad_norm": 0.30604201555252075, "learning_rate": 1.0350794033117189e-05, "loss": 0.0124, "step": 8020 }, { "epoch": 19.585365853658537, "grad_norm": 0.12198792397975922, "learning_rate": 1.0250274639070856e-05, "loss": 0.0107, "step": 8030 }, { "epoch": 19.609756097560975, "grad_norm": 0.16436730325222015, "learning_rate": 1.0150189941652599e-05, "loss": 0.0115, "step": 8040 }, { "epoch": 19.634146341463413, "grad_norm": 0.12333788722753525, "learning_rate": 1.0050541035372635e-05, "loss": 0.0088, "step": 8050 }, { "epoch": 19.658536585365855, "grad_norm": 0.25077614188194275, "learning_rate": 9.951329009975458e-06, "loss": 0.0087, "step": 8060 }, { "epoch": 19.682926829268293, "grad_norm": 0.14324992895126343, "learning_rate": 9.852554950427845e-06, "loss": 0.0108, "step": 8070 }, { "epoch": 19.70731707317073, "grad_norm": 0.1492728441953659, "learning_rate": 9.754219936907105e-06, "loss": 0.0091, "step": 8080 }, { "epoch": 19.73170731707317, "grad_norm": 0.0740441381931305, "learning_rate": 9.656325044789194e-06, "loss": 0.009, "step": 8090 }, { "epoch": 19.75609756097561, "grad_norm": 0.17115770280361176, "learning_rate": 9.55887134463697e-06, "loss": 0.0081, "step": 8100 }, { "epoch": 19.78048780487805, "grad_norm": 0.1268915832042694, "learning_rate": 9.461859902188475e-06, "loss": 0.0071, "step": 8110 }, { "epoch": 19.804878048780488, "grad_norm": 0.13579000532627106, "learning_rate": 9.365291778345303e-06, "loss": 0.0101, "step": 8120 }, { "epoch": 19.829268292682926, "grad_norm": 0.1489654928445816, "learning_rate": 9.269168029160991e-06, "loss": 0.0111, "step": 8130 }, { "epoch": 19.853658536585368, "grad_norm": 0.17810888588428497, "learning_rate": 9.173489705829447e-06, "loss": 0.0084, "step": 8140 }, { "epoch": 19.878048780487806, "grad_norm": 0.19767481088638306, "learning_rate": 9.078257854673516e-06, "loss": 0.0125, "step": 8150 }, { "epoch": 19.902439024390244, "grad_norm": 0.2359165996313095, "learning_rate": 8.983473517133429e-06, "loss": 0.0119, "step": 8160 }, { "epoch": 19.926829268292682, "grad_norm": 0.09948579221963882, "learning_rate": 8.889137729755537e-06, "loss": 0.0109, "step": 8170 }, { "epoch": 19.951219512195124, "grad_norm": 0.21918344497680664, "learning_rate": 8.79525152418087e-06, "loss": 0.0095, "step": 8180 }, { "epoch": 19.975609756097562, "grad_norm": 0.136754110455513, "learning_rate": 8.701815927133961e-06, "loss": 0.0117, "step": 8190 }, { "epoch": 20.0, "grad_norm": 0.17661526799201965, "learning_rate": 8.608831960411534e-06, "loss": 0.0108, "step": 8200 }, { "epoch": 20.024390243902438, "grad_norm": 0.15653300285339355, "learning_rate": 8.516300640871321e-06, "loss": 0.0089, "step": 8210 }, { "epoch": 20.048780487804876, "grad_norm": 0.11913203448057175, "learning_rate": 8.424222980421038e-06, "loss": 0.0094, "step": 8220 }, { "epoch": 20.073170731707318, "grad_norm": 0.17709387838840485, "learning_rate": 8.332599986007184e-06, "loss": 0.0116, "step": 8230 }, { "epoch": 20.097560975609756, "grad_norm": 0.14333659410476685, "learning_rate": 8.241432659604203e-06, "loss": 0.0101, "step": 8240 }, { "epoch": 20.121951219512194, "grad_norm": 0.17693360149860382, "learning_rate": 8.150721998203331e-06, "loss": 0.0084, "step": 8250 }, { "epoch": 20.146341463414632, "grad_norm": 0.14758910238742828, "learning_rate": 8.06046899380184e-06, "loss": 0.0081, "step": 8260 }, { "epoch": 20.170731707317074, "grad_norm": 0.25577157735824585, "learning_rate": 7.970674633392133e-06, "loss": 0.0098, "step": 8270 }, { "epoch": 20.195121951219512, "grad_norm": 0.1601947695016861, "learning_rate": 7.881339898950924e-06, "loss": 0.0115, "step": 8280 }, { "epoch": 20.21951219512195, "grad_norm": 0.20418229699134827, "learning_rate": 7.792465767428597e-06, "loss": 0.0126, "step": 8290 }, { "epoch": 20.24390243902439, "grad_norm": 0.18942424654960632, "learning_rate": 7.704053210738376e-06, "loss": 0.0099, "step": 8300 }, { "epoch": 20.26829268292683, "grad_norm": 0.10163628309965134, "learning_rate": 7.6161031957458494e-06, "loss": 0.0087, "step": 8310 }, { "epoch": 20.29268292682927, "grad_norm": 0.0965646430850029, "learning_rate": 7.5286166842582605e-06, "loss": 0.011, "step": 8320 }, { "epoch": 20.317073170731707, "grad_norm": 0.25612872838974, "learning_rate": 7.4415946330140814e-06, "loss": 0.0103, "step": 8330 }, { "epoch": 20.341463414634145, "grad_norm": 0.09951556473970413, "learning_rate": 7.3550379936725644e-06, "loss": 0.0093, "step": 8340 }, { "epoch": 20.365853658536587, "grad_norm": 0.14025788009166718, "learning_rate": 7.2689477128032035e-06, "loss": 0.009, "step": 8350 }, { "epoch": 20.390243902439025, "grad_norm": 0.08568873256444931, "learning_rate": 7.183324731875551e-06, "loss": 0.0122, "step": 8360 }, { "epoch": 20.414634146341463, "grad_norm": 0.22724655270576477, "learning_rate": 7.098169987248782e-06, "loss": 0.0083, "step": 8370 }, { "epoch": 20.4390243902439, "grad_norm": 0.16363540291786194, "learning_rate": 7.013484410161553e-06, "loss": 0.0081, "step": 8380 }, { "epoch": 20.463414634146343, "grad_norm": 0.13467468321323395, "learning_rate": 6.92926892672176e-06, "loss": 0.0104, "step": 8390 }, { "epoch": 20.48780487804878, "grad_norm": 0.13649263978004456, "learning_rate": 6.845524457896446e-06, "loss": 0.0097, "step": 8400 }, { "epoch": 20.51219512195122, "grad_norm": 0.17734812200069427, "learning_rate": 6.7622519195017165e-06, "loss": 0.009, "step": 8410 }, { "epoch": 20.536585365853657, "grad_norm": 0.13624143600463867, "learning_rate": 6.679452222192684e-06, "loss": 0.0085, "step": 8420 }, { "epoch": 20.5609756097561, "grad_norm": 0.2331664115190506, "learning_rate": 6.597126271453579e-06, "loss": 0.0101, "step": 8430 }, { "epoch": 20.585365853658537, "grad_norm": 0.1888420134782791, "learning_rate": 6.51527496758782e-06, "loss": 0.0087, "step": 8440 }, { "epoch": 20.609756097560975, "grad_norm": 0.1492602378129959, "learning_rate": 6.433899205708155e-06, "loss": 0.0087, "step": 8450 }, { "epoch": 20.634146341463413, "grad_norm": 0.22079570591449738, "learning_rate": 6.352999875726856e-06, "loss": 0.0099, "step": 8460 }, { "epoch": 20.658536585365855, "grad_norm": 0.15431448817253113, "learning_rate": 6.272577862346052e-06, "loss": 0.0065, "step": 8470 }, { "epoch": 20.682926829268293, "grad_norm": 0.13980098068714142, "learning_rate": 6.192634045047996e-06, "loss": 0.0076, "step": 8480 }, { "epoch": 20.70731707317073, "grad_norm": 0.1788775771856308, "learning_rate": 6.113169298085458e-06, "loss": 0.0109, "step": 8490 }, { "epoch": 20.73170731707317, "grad_norm": 0.1268845647573471, "learning_rate": 6.034184490472195e-06, "loss": 0.0123, "step": 8500 }, { "epoch": 20.75609756097561, "grad_norm": 0.1849120706319809, "learning_rate": 5.955680485973386e-06, "loss": 0.0057, "step": 8510 }, { "epoch": 20.78048780487805, "grad_norm": 0.22011688351631165, "learning_rate": 5.877658143096265e-06, "loss": 0.0096, "step": 8520 }, { "epoch": 20.804878048780488, "grad_norm": 0.12771470844745636, "learning_rate": 5.800118315080661e-06, "loss": 0.0096, "step": 8530 }, { "epoch": 20.829268292682926, "grad_norm": 0.10316302627325058, "learning_rate": 5.723061849889716e-06, "loss": 0.0073, "step": 8540 }, { "epoch": 20.853658536585368, "grad_norm": 0.14179106056690216, "learning_rate": 5.646489590200604e-06, "loss": 0.0076, "step": 8550 }, { "epoch": 20.878048780487806, "grad_norm": 0.16923163831233978, "learning_rate": 5.570402373395256e-06, "loss": 0.0101, "step": 8560 }, { "epoch": 20.902439024390244, "grad_norm": 0.21445897221565247, "learning_rate": 5.494801031551305e-06, "loss": 0.0079, "step": 8570 }, { "epoch": 20.926829268292682, "grad_norm": 0.17787350714206696, "learning_rate": 5.41968639143291e-06, "loss": 0.0118, "step": 8580 }, { "epoch": 20.951219512195124, "grad_norm": 0.21737384796142578, "learning_rate": 5.345059274481751e-06, "loss": 0.0082, "step": 8590 }, { "epoch": 20.975609756097562, "grad_norm": 0.13147112727165222, "learning_rate": 5.270920496808002e-06, "loss": 0.0077, "step": 8600 }, { "epoch": 21.0, "grad_norm": 0.13363362848758698, "learning_rate": 5.1972708691814695e-06, "loss": 0.0095, "step": 8610 }, { "epoch": 21.024390243902438, "grad_norm": 0.08537745475769043, "learning_rate": 5.124111197022674e-06, "loss": 0.01, "step": 8620 }, { "epoch": 21.048780487804876, "grad_norm": 0.21417123079299927, "learning_rate": 5.051442280394081e-06, "loss": 0.0131, "step": 8630 }, { "epoch": 21.073170731707318, "grad_norm": 0.09984221309423447, "learning_rate": 4.979264913991322e-06, "loss": 0.007, "step": 8640 }, { "epoch": 21.097560975609756, "grad_norm": 0.1674463450908661, "learning_rate": 4.907579887134489e-06, "loss": 0.009, "step": 8650 }, { "epoch": 21.121951219512194, "grad_norm": 0.17915357649326324, "learning_rate": 4.836387983759572e-06, "loss": 0.0105, "step": 8660 }, { "epoch": 21.146341463414632, "grad_norm": 0.1776980608701706, "learning_rate": 4.765689982409816e-06, "loss": 0.0083, "step": 8670 }, { "epoch": 21.170731707317074, "grad_norm": 0.1178482174873352, "learning_rate": 4.695486656227233e-06, "loss": 0.0073, "step": 8680 }, { "epoch": 21.195121951219512, "grad_norm": 0.12703382968902588, "learning_rate": 4.625778772944156e-06, "loss": 0.0126, "step": 8690 }, { "epoch": 21.21951219512195, "grad_norm": 0.16913475096225739, "learning_rate": 4.556567094874825e-06, "loss": 0.0103, "step": 8700 }, { "epoch": 21.24390243902439, "grad_norm": 0.14067861437797546, "learning_rate": 4.487852378907059e-06, "loss": 0.0087, "step": 8710 }, { "epoch": 21.26829268292683, "grad_norm": 0.12038375437259674, "learning_rate": 4.419635376493986e-06, "loss": 0.0065, "step": 8720 }, { "epoch": 21.29268292682927, "grad_norm": 0.13224199414253235, "learning_rate": 4.351916833645825e-06, "loss": 0.0096, "step": 8730 }, { "epoch": 21.317073170731707, "grad_norm": 0.0995870977640152, "learning_rate": 4.284697490921691e-06, "loss": 0.0078, "step": 8740 }, { "epoch": 21.341463414634145, "grad_norm": 0.23361696302890778, "learning_rate": 4.2179780834215585e-06, "loss": 0.0087, "step": 8750 }, { "epoch": 21.365853658536587, "grad_norm": 0.13679705560207367, "learning_rate": 4.151759340778178e-06, "loss": 0.0121, "step": 8760 }, { "epoch": 21.390243902439025, "grad_norm": 0.12412291765213013, "learning_rate": 4.086041987149109e-06, "loss": 0.008, "step": 8770 }, { "epoch": 21.414634146341463, "grad_norm": 0.19725491106510162, "learning_rate": 4.020826741208811e-06, "loss": 0.0089, "step": 8780 }, { "epoch": 21.4390243902439, "grad_norm": 0.1982433944940567, "learning_rate": 3.956114316140746e-06, "loss": 0.0117, "step": 8790 }, { "epoch": 21.463414634146343, "grad_norm": 0.19211578369140625, "learning_rate": 3.891905419629643e-06, "loss": 0.0129, "step": 8800 }, { "epoch": 21.48780487804878, "grad_norm": 0.09827245771884918, "learning_rate": 3.8282007538536946e-06, "loss": 0.0077, "step": 8810 }, { "epoch": 21.51219512195122, "grad_norm": 0.11677441000938416, "learning_rate": 3.7650010154769265e-06, "loss": 0.0118, "step": 8820 }, { "epoch": 21.536585365853657, "grad_norm": 0.09769511222839355, "learning_rate": 3.7023068956415608e-06, "loss": 0.0073, "step": 8830 }, { "epoch": 21.5609756097561, "grad_norm": 0.21006107330322266, "learning_rate": 3.6401190799604303e-06, "loss": 0.0115, "step": 8840 }, { "epoch": 21.585365853658537, "grad_norm": 0.13288800418376923, "learning_rate": 3.578438248509536e-06, "loss": 0.0084, "step": 8850 }, { "epoch": 21.609756097560975, "grad_norm": 0.10836083441972733, "learning_rate": 3.5172650758205583e-06, "loss": 0.0076, "step": 8860 }, { "epoch": 21.634146341463413, "grad_norm": 0.15194399654865265, "learning_rate": 3.45660023087353e-06, "loss": 0.0106, "step": 8870 }, { "epoch": 21.658536585365855, "grad_norm": 0.13392139971256256, "learning_rate": 3.3964443770894528e-06, "loss": 0.0076, "step": 8880 }, { "epoch": 21.682926829268293, "grad_norm": 0.13574053347110748, "learning_rate": 3.3367981723231245e-06, "loss": 0.0089, "step": 8890 }, { "epoch": 21.70731707317073, "grad_norm": 0.14829783141613007, "learning_rate": 3.2776622688558746e-06, "loss": 0.0085, "step": 8900 }, { "epoch": 21.73170731707317, "grad_norm": 0.21344609558582306, "learning_rate": 3.2190373133884677e-06, "loss": 0.0084, "step": 8910 }, { "epoch": 21.75609756097561, "grad_norm": 0.09232096374034882, "learning_rate": 3.1609239470340446e-06, "loss": 0.0091, "step": 8920 }, { "epoch": 21.78048780487805, "grad_norm": 0.11385582387447357, "learning_rate": 3.1033228053110373e-06, "loss": 0.0084, "step": 8930 }, { "epoch": 21.804878048780488, "grad_norm": 0.07174244523048401, "learning_rate": 3.0462345181363314e-06, "loss": 0.0071, "step": 8940 }, { "epoch": 21.829268292682926, "grad_norm": 0.1953529566526413, "learning_rate": 2.9896597098182654e-06, "loss": 0.0086, "step": 8950 }, { "epoch": 21.853658536585368, "grad_norm": 0.13533097505569458, "learning_rate": 2.933598999049891e-06, "loss": 0.0093, "step": 8960 }, { "epoch": 21.878048780487806, "grad_norm": 0.12223199754953384, "learning_rate": 2.8780529989021697e-06, "loss": 0.0085, "step": 8970 }, { "epoch": 21.902439024390244, "grad_norm": 0.12146934121847153, "learning_rate": 2.823022316817242e-06, "loss": 0.0076, "step": 8980 }, { "epoch": 21.926829268292682, "grad_norm": 0.19187134504318237, "learning_rate": 2.7685075546018456e-06, "loss": 0.013, "step": 8990 }, { "epoch": 21.951219512195124, "grad_norm": 0.12114346027374268, "learning_rate": 2.7145093084206598e-06, "loss": 0.0097, "step": 9000 }, { "epoch": 21.975609756097562, "grad_norm": 0.17190483212471008, "learning_rate": 2.661028168789892e-06, "loss": 0.009, "step": 9010 }, { "epoch": 22.0, "grad_norm": 0.11163271963596344, "learning_rate": 2.6080647205706855e-06, "loss": 0.0101, "step": 9020 }, { "epoch": 22.024390243902438, "grad_norm": 0.10256139189004898, "learning_rate": 2.555619542962834e-06, "loss": 0.0074, "step": 9030 }, { "epoch": 22.048780487804876, "grad_norm": 0.11299065500497818, "learning_rate": 2.503693209498409e-06, "loss": 0.0091, "step": 9040 }, { "epoch": 22.073170731707318, "grad_norm": 0.12215982377529144, "learning_rate": 2.452286288035449e-06, "loss": 0.006, "step": 9050 }, { "epoch": 22.097560975609756, "grad_norm": 0.13071924448013306, "learning_rate": 2.4013993407518363e-06, "loss": 0.0089, "step": 9060 }, { "epoch": 22.121951219512194, "grad_norm": 0.10491999983787537, "learning_rate": 2.351032924139063e-06, "loss": 0.0098, "step": 9070 }, { "epoch": 22.146341463414632, "grad_norm": 0.12704725563526154, "learning_rate": 2.30118758899619e-06, "loss": 0.0064, "step": 9080 }, { "epoch": 22.170731707317074, "grad_norm": 0.1342615783214569, "learning_rate": 2.2518638804238157e-06, "loss": 0.0112, "step": 9090 }, { "epoch": 22.195121951219512, "grad_norm": 0.15303771197795868, "learning_rate": 2.203062337818118e-06, "loss": 0.0065, "step": 9100 }, { "epoch": 22.21951219512195, "grad_norm": 0.12761712074279785, "learning_rate": 2.1547834948649483e-06, "loss": 0.0082, "step": 9110 }, { "epoch": 22.24390243902439, "grad_norm": 0.15862001478672028, "learning_rate": 2.1070278795340017e-06, "loss": 0.0105, "step": 9120 }, { "epoch": 22.26829268292683, "grad_norm": 0.11628398299217224, "learning_rate": 2.059796014073029e-06, "loss": 0.0091, "step": 9130 }, { "epoch": 22.29268292682927, "grad_norm": 0.1255343109369278, "learning_rate": 2.01308841500214e-06, "loss": 0.0073, "step": 9140 }, { "epoch": 22.317073170731707, "grad_norm": 0.1320139318704605, "learning_rate": 1.9669055931081704e-06, "loss": 0.0082, "step": 9150 }, { "epoch": 22.341463414634145, "grad_norm": 0.20194648206233978, "learning_rate": 1.9212480534390507e-06, "loss": 0.0113, "step": 9160 }, { "epoch": 22.365853658536587, "grad_norm": 0.12820903956890106, "learning_rate": 1.8761162952983246e-06, "loss": 0.0076, "step": 9170 }, { "epoch": 22.390243902439025, "grad_norm": 0.14364054799079895, "learning_rate": 1.8315108122396618e-06, "loss": 0.0082, "step": 9180 }, { "epoch": 22.414634146341463, "grad_norm": 0.14670322835445404, "learning_rate": 1.787432092061475e-06, "loss": 0.0102, "step": 9190 }, { "epoch": 22.4390243902439, "grad_norm": 0.14936046302318573, "learning_rate": 1.743880616801602e-06, "loss": 0.0086, "step": 9200 }, { "epoch": 22.463414634146343, "grad_norm": 0.15991879999637604, "learning_rate": 1.7008568627319865e-06, "loss": 0.0101, "step": 9210 }, { "epoch": 22.48780487804878, "grad_norm": 0.09684566408395767, "learning_rate": 1.6583613003535226e-06, "loss": 0.0089, "step": 9220 }, { "epoch": 22.51219512195122, "grad_norm": 0.16733086109161377, "learning_rate": 1.6163943943908522e-06, "loss": 0.01, "step": 9230 }, { "epoch": 22.536585365853657, "grad_norm": 0.13386715948581696, "learning_rate": 1.5749566037873476e-06, "loss": 0.0114, "step": 9240 }, { "epoch": 22.5609756097561, "grad_norm": 0.08313728868961334, "learning_rate": 1.5340483817000428e-06, "loss": 0.0081, "step": 9250 }, { "epoch": 22.585365853658537, "grad_norm": 0.11027350276708603, "learning_rate": 1.4936701754947101e-06, "loss": 0.011, "step": 9260 }, { "epoch": 22.609756097560975, "grad_norm": 0.10279306024312973, "learning_rate": 1.4538224267409361e-06, "loss": 0.0081, "step": 9270 }, { "epoch": 22.634146341463413, "grad_norm": 0.13070428371429443, "learning_rate": 1.414505571207314e-06, "loss": 0.0089, "step": 9280 }, { "epoch": 22.658536585365855, "grad_norm": 0.10510098189115524, "learning_rate": 1.3757200388566816e-06, "loss": 0.0078, "step": 9290 }, { "epoch": 22.682926829268293, "grad_norm": 0.1142260804772377, "learning_rate": 1.3374662538414074e-06, "loss": 0.0076, "step": 9300 }, { "epoch": 22.70731707317073, "grad_norm": 0.08060184121131897, "learning_rate": 1.2997446344987617e-06, "loss": 0.01, "step": 9310 }, { "epoch": 22.73170731707317, "grad_norm": 0.07056501507759094, "learning_rate": 1.262555593346315e-06, "loss": 0.0089, "step": 9320 }, { "epoch": 22.75609756097561, "grad_norm": 0.08530158549547195, "learning_rate": 1.2258995370774685e-06, "loss": 0.0095, "step": 9330 }, { "epoch": 22.78048780487805, "grad_norm": 0.20514100790023804, "learning_rate": 1.1897768665569798e-06, "loss": 0.011, "step": 9340 }, { "epoch": 22.804878048780488, "grad_norm": 0.14554913341999054, "learning_rate": 1.1541879768165954e-06, "loss": 0.009, "step": 9350 }, { "epoch": 22.829268292682926, "grad_norm": 0.10517197847366333, "learning_rate": 1.1191332570507085e-06, "loss": 0.0068, "step": 9360 }, { "epoch": 22.853658536585368, "grad_norm": 0.08250486850738525, "learning_rate": 1.0846130906121132e-06, "loss": 0.0071, "step": 9370 }, { "epoch": 22.878048780487806, "grad_norm": 0.1687975823879242, "learning_rate": 1.0506278550078131e-06, "loss": 0.0149, "step": 9380 }, { "epoch": 22.902439024390244, "grad_norm": 0.17175151407718658, "learning_rate": 1.0171779218949185e-06, "loss": 0.0094, "step": 9390 }, { "epoch": 22.926829268292682, "grad_norm": 0.16291427612304688, "learning_rate": 9.842636570765174e-07, "loss": 0.0095, "step": 9400 }, { "epoch": 22.951219512195124, "grad_norm": 0.22573085129261017, "learning_rate": 9.518854204977612e-07, "loss": 0.0082, "step": 9410 }, { "epoch": 22.975609756097562, "grad_norm": 0.19846147298812866, "learning_rate": 9.200435662418349e-07, "loss": 0.0072, "step": 9420 }, { "epoch": 23.0, "grad_norm": 0.17329593002796173, "learning_rate": 8.887384425261658e-07, "loss": 0.0076, "step": 9430 }, { "epoch": 23.024390243902438, "grad_norm": 0.07042143493890762, "learning_rate": 8.579703916985648e-07, "loss": 0.0074, "step": 9440 }, { "epoch": 23.048780487804876, "grad_norm": 0.26697543263435364, "learning_rate": 8.277397502335194e-07, "loss": 0.0089, "step": 9450 }, { "epoch": 23.073170731707318, "grad_norm": 0.16999194025993347, "learning_rate": 7.980468487284675e-07, "loss": 0.0094, "step": 9460 }, { "epoch": 23.097560975609756, "grad_norm": 0.17188598215579987, "learning_rate": 7.688920119002297e-07, "loss": 0.0077, "step": 9470 }, { "epoch": 23.121951219512194, "grad_norm": 0.0863710269331932, "learning_rate": 7.402755585814269e-07, "loss": 0.0087, "step": 9480 }, { "epoch": 23.146341463414632, "grad_norm": 0.07592741400003433, "learning_rate": 7.121978017170073e-07, "loss": 0.0092, "step": 9490 }, { "epoch": 23.170731707317074, "grad_norm": 0.23436231911182404, "learning_rate": 6.846590483608306e-07, "loss": 0.0088, "step": 9500 }, { "epoch": 23.195121951219512, "grad_norm": 0.08250889927148819, "learning_rate": 6.576595996722834e-07, "loss": 0.0073, "step": 9510 }, { "epoch": 23.21951219512195, "grad_norm": 0.10812386870384216, "learning_rate": 6.311997509130141e-07, "loss": 0.0135, "step": 9520 }, { "epoch": 23.24390243902439, "grad_norm": 0.08859504014253616, "learning_rate": 6.052797914436803e-07, "loss": 0.0092, "step": 9530 }, { "epoch": 23.26829268292683, "grad_norm": 0.12982113659381866, "learning_rate": 5.799000047208181e-07, "loss": 0.008, "step": 9540 }, { "epoch": 23.29268292682927, "grad_norm": 0.11518954485654831, "learning_rate": 5.550606682937054e-07, "loss": 0.0109, "step": 9550 }, { "epoch": 23.317073170731707, "grad_norm": 0.16649867594242096, "learning_rate": 5.307620538013481e-07, "loss": 0.0117, "step": 9560 }, { "epoch": 23.341463414634145, "grad_norm": 0.2819383144378662, "learning_rate": 5.070044269694874e-07, "loss": 0.012, "step": 9570 }, { "epoch": 23.365853658536587, "grad_norm": 0.14769881963729858, "learning_rate": 4.837880476077417e-07, "loss": 0.0081, "step": 9580 }, { "epoch": 23.390243902439025, "grad_norm": 0.2314777970314026, "learning_rate": 4.6111316960670835e-07, "loss": 0.0084, "step": 9590 }, { "epoch": 23.414634146341463, "grad_norm": 0.10781470686197281, "learning_rate": 4.389800409352218e-07, "loss": 0.0098, "step": 9600 }, { "epoch": 23.4390243902439, "grad_norm": 0.15466098487377167, "learning_rate": 4.173889036376277e-07, "loss": 0.0071, "step": 9610 }, { "epoch": 23.463414634146343, "grad_norm": 0.15470187366008759, "learning_rate": 3.963399938311463e-07, "loss": 0.0113, "step": 9620 }, { "epoch": 23.48780487804878, "grad_norm": 0.17244833707809448, "learning_rate": 3.7583354170328545e-07, "loss": 0.011, "step": 9630 }, { "epoch": 23.51219512195122, "grad_norm": 0.12076525390148163, "learning_rate": 3.558697715093207e-07, "loss": 0.0105, "step": 9640 }, { "epoch": 23.536585365853657, "grad_norm": 0.12389826029539108, "learning_rate": 3.3644890156983576e-07, "loss": 0.0083, "step": 9650 }, { "epoch": 23.5609756097561, "grad_norm": 0.13878820836544037, "learning_rate": 3.175711442683638e-07, "loss": 0.0092, "step": 9660 }, { "epoch": 23.585365853658537, "grad_norm": 0.13089226186275482, "learning_rate": 2.9923670604902197e-07, "loss": 0.0116, "step": 9670 }, { "epoch": 23.609756097560975, "grad_norm": 0.0701451227068901, "learning_rate": 2.814457874143028e-07, "loss": 0.0079, "step": 9680 }, { "epoch": 23.634146341463413, "grad_norm": 0.061592455953359604, "learning_rate": 2.641985829228366e-07, "loss": 0.0095, "step": 9690 }, { "epoch": 23.658536585365855, "grad_norm": 0.15562207996845245, "learning_rate": 2.474952811872877e-07, "loss": 0.0083, "step": 9700 }, { "epoch": 23.682926829268293, "grad_norm": 0.07680628448724747, "learning_rate": 2.3133606487228397e-07, "loss": 0.008, "step": 9710 }, { "epoch": 23.70731707317073, "grad_norm": 0.07263928651809692, "learning_rate": 2.157211106924295e-07, "loss": 0.0077, "step": 9720 }, { "epoch": 23.73170731707317, "grad_norm": 0.12193568050861359, "learning_rate": 2.006505894103672e-07, "loss": 0.0109, "step": 9730 }, { "epoch": 23.75609756097561, "grad_norm": 0.20887912809848785, "learning_rate": 1.8612466583489696e-07, "loss": 0.0089, "step": 9740 }, { "epoch": 23.78048780487805, "grad_norm": 0.09030177444219589, "learning_rate": 1.7214349881918834e-07, "loss": 0.0102, "step": 9750 }, { "epoch": 23.804878048780488, "grad_norm": 0.19849129021167755, "learning_rate": 1.5870724125904845e-07, "loss": 0.0086, "step": 9760 }, { "epoch": 23.829268292682926, "grad_norm": 0.16899527609348297, "learning_rate": 1.4581604009124006e-07, "loss": 0.0082, "step": 9770 }, { "epoch": 23.853658536585368, "grad_norm": 0.10104808211326599, "learning_rate": 1.334700362918717e-07, "loss": 0.0079, "step": 9780 }, { "epoch": 23.878048780487806, "grad_norm": 0.16594862937927246, "learning_rate": 1.2166936487486015e-07, "loss": 0.0075, "step": 9790 }, { "epoch": 23.902439024390244, "grad_norm": 0.2572060525417328, "learning_rate": 1.1041415489045914e-07, "loss": 0.009, "step": 9800 }, { "epoch": 23.926829268292682, "grad_norm": 0.16250790655612946, "learning_rate": 9.970452942384412e-08, "loss": 0.0062, "step": 9810 }, { "epoch": 23.951219512195124, "grad_norm": 0.15023371577262878, "learning_rate": 8.954060559375754e-08, "loss": 0.0063, "step": 9820 }, { "epoch": 23.975609756097562, "grad_norm": 0.15030191838741302, "learning_rate": 7.99224945512489e-08, "loss": 0.008, "step": 9830 }, { "epoch": 24.0, "grad_norm": 0.13905693590641022, "learning_rate": 7.085030147843675e-08, "loss": 0.0115, "step": 9840 }, { "epoch": 24.024390243902438, "grad_norm": 0.23413218557834625, "learning_rate": 6.232412558736523e-08, "loss": 0.0102, "step": 9850 }, { "epoch": 24.048780487804876, "grad_norm": 0.23484253883361816, "learning_rate": 5.434406011893822e-08, "loss": 0.0087, "step": 9860 }, { "epoch": 24.073170731707318, "grad_norm": 0.11507958918809891, "learning_rate": 4.6910192341864664e-08, "loss": 0.0115, "step": 9870 }, { "epoch": 24.097560975609756, "grad_norm": 0.1549917459487915, "learning_rate": 4.0022603551737035e-08, "loss": 0.0078, "step": 9880 }, { "epoch": 24.121951219512194, "grad_norm": 0.10509636998176575, "learning_rate": 3.3681369070120985e-08, "loss": 0.01, "step": 9890 }, { "epoch": 24.146341463414632, "grad_norm": 0.22248610854148865, "learning_rate": 2.7886558243744866e-08, "loss": 0.0091, "step": 9900 }, { "epoch": 24.170731707317074, "grad_norm": 0.20585030317306519, "learning_rate": 2.2638234443722596e-08, "loss": 0.0124, "step": 9910 }, { "epoch": 24.195121951219512, "grad_norm": 0.22081707417964935, "learning_rate": 1.7936455064887504e-08, "loss": 0.0088, "step": 9920 }, { "epoch": 24.21951219512195, "grad_norm": 0.18969430029392242, "learning_rate": 1.378127152514841e-08, "loss": 0.0142, "step": 9930 }, { "epoch": 24.24390243902439, "grad_norm": 0.1392698734998703, "learning_rate": 1.0172729264917857e-08, "loss": 0.0068, "step": 9940 }, { "epoch": 24.26829268292683, "grad_norm": 0.1857910007238388, "learning_rate": 7.1108677466458215e-09, "loss": 0.0103, "step": 9950 }, { "epoch": 24.29268292682927, "grad_norm": 0.18373022973537445, "learning_rate": 4.595720454353414e-09, "loss": 0.0095, "step": 9960 }, { "epoch": 24.317073170731707, "grad_norm": 0.12904109060764313, "learning_rate": 2.627314893294264e-09, "loss": 0.0102, "step": 9970 }, { "epoch": 24.341463414634145, "grad_norm": 0.06782866269350052, "learning_rate": 1.2056725896270048e-09, "loss": 0.007, "step": 9980 }, { "epoch": 24.365853658536587, "grad_norm": 0.22785444557666779, "learning_rate": 3.308090902098826e-10, "loss": 0.0084, "step": 9990 }, { "epoch": 24.390243902439025, "grad_norm": 0.1324882060289383, "learning_rate": 2.7339624120159555e-12, "loss": 0.0097, "step": 10000 }, { "epoch": 24.390243902439025, "step": 10000, "total_flos": 0.0, "train_loss": 0.025912553722411392, "train_runtime": 10249.0493, "train_samples_per_second": 31.222, "train_steps_per_second": 0.976 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 25, "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": 32, "trial_name": null, "trial_params": null }