diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,12034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 20000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "grad_norm": 0.7395622134208679, + "learning_rate": 9e-07, + "loss": 1.1397, + "step": 10 + }, + { + "grad_norm": 0.797982394695282, + "learning_rate": 1.9e-06, + "loss": 1.1333, + "step": 20 + }, + { + "grad_norm": 0.6013365983963013, + "learning_rate": 2.9e-06, + "loss": 1.1018, + "step": 30 + }, + { + "grad_norm": 0.3074468970298767, + "learning_rate": 3.9e-06, + "loss": 1.0844, + "step": 40 + }, + { + "grad_norm": 0.4337396025657654, + "learning_rate": 4.9000000000000005e-06, + "loss": 1.0541, + "step": 50 + }, + { + "grad_norm": 0.44008368253707886, + "learning_rate": 5.9e-06, + "loss": 1.0576, + "step": 60 + }, + { + "grad_norm": 0.46346142888069153, + "learning_rate": 6.900000000000001e-06, + "loss": 1.0477, + "step": 70 + }, + { + "grad_norm": 0.8132529854774475, + "learning_rate": 7.9e-06, + "loss": 1.0377, + "step": 80 + }, + { + "grad_norm": 0.589951753616333, + "learning_rate": 8.9e-06, + "loss": 1.0318, + "step": 90 + }, + { + "grad_norm": 0.34533777832984924, + "learning_rate": 9.900000000000002e-06, + "loss": 1.0304, + "step": 100 + }, + { + "grad_norm": 0.5165931582450867, + "learning_rate": 1.09e-05, + "loss": 1.0245, + "step": 110 + }, + { + "grad_norm": 1.5801116228103638, + "learning_rate": 1.19e-05, + "loss": 0.9693, + "step": 120 + }, + { + "grad_norm": 2.446295976638794, + "learning_rate": 1.29e-05, + "loss": 0.7924, + "step": 130 + }, + { + "grad_norm": 2.7408406734466553, + "learning_rate": 1.3900000000000002e-05, + "loss": 0.623, + "step": 140 + }, + { + "grad_norm": 2.2150840759277344, + "learning_rate": 1.49e-05, + "loss": 0.4795, + "step": 150 + }, + { + "grad_norm": 2.8643667697906494, + "learning_rate": 1.59e-05, + "loss": 0.3721, + "step": 160 + }, + { + "grad_norm": 2.9574365615844727, + "learning_rate": 1.69e-05, + "loss": 0.2909, + "step": 170 + }, + { + "grad_norm": 1.9344878196716309, + "learning_rate": 1.79e-05, + "loss": 0.2165, + "step": 180 + }, + { + "grad_norm": 2.583017587661743, + "learning_rate": 1.8900000000000002e-05, + "loss": 0.2015, + "step": 190 + }, + { + "grad_norm": 2.3045356273651123, + "learning_rate": 1.9900000000000003e-05, + "loss": 0.1684, + "step": 200 + }, + { + "grad_norm": 1.5953307151794434, + "learning_rate": 2.09e-05, + "loss": 0.1595, + "step": 210 + }, + { + "grad_norm": 2.159213066101074, + "learning_rate": 2.19e-05, + "loss": 0.1514, + "step": 220 + }, + { + "grad_norm": 2.5493297576904297, + "learning_rate": 2.29e-05, + "loss": 0.1417, + "step": 230 + }, + { + "grad_norm": 1.7852596044540405, + "learning_rate": 2.39e-05, + "loss": 0.1369, + "step": 240 + }, + { + "grad_norm": 1.8975121974945068, + "learning_rate": 2.4900000000000002e-05, + "loss": 0.1293, + "step": 250 + }, + { + "grad_norm": 1.9026151895523071, + "learning_rate": 2.5900000000000003e-05, + "loss": 0.1243, + "step": 260 + }, + { + "grad_norm": 1.4586244821548462, + "learning_rate": 2.6900000000000003e-05, + "loss": 0.1155, + "step": 270 + }, + { + "grad_norm": 1.688338279724121, + "learning_rate": 2.7900000000000004e-05, + "loss": 0.1066, + "step": 280 + }, + { + "grad_norm": 1.1545861959457397, + "learning_rate": 2.8899999999999998e-05, + "loss": 0.1108, + "step": 290 + }, + { + "grad_norm": 2.1053218841552734, + "learning_rate": 2.9900000000000002e-05, + "loss": 0.1145, + "step": 300 + }, + { + "grad_norm": 2.336277961730957, + "learning_rate": 3.09e-05, + "loss": 0.1093, + "step": 310 + }, + { + "grad_norm": 2.255950689315796, + "learning_rate": 3.19e-05, + "loss": 0.0986, + "step": 320 + }, + { + "grad_norm": 1.8794231414794922, + "learning_rate": 3.29e-05, + "loss": 0.1034, + "step": 330 + }, + { + "grad_norm": 1.811044454574585, + "learning_rate": 3.3900000000000004e-05, + "loss": 0.1063, + "step": 340 + }, + { + "grad_norm": 1.7343348264694214, + "learning_rate": 3.49e-05, + "loss": 0.0961, + "step": 350 + }, + { + "grad_norm": 1.1818451881408691, + "learning_rate": 3.59e-05, + "loss": 0.1069, + "step": 360 + }, + { + "grad_norm": 1.313843846321106, + "learning_rate": 3.69e-05, + "loss": 0.0889, + "step": 370 + }, + { + "grad_norm": 1.465733289718628, + "learning_rate": 3.79e-05, + "loss": 0.0934, + "step": 380 + }, + { + "grad_norm": 1.4527413845062256, + "learning_rate": 3.8900000000000004e-05, + "loss": 0.0912, + "step": 390 + }, + { + "grad_norm": 1.3359452486038208, + "learning_rate": 3.99e-05, + "loss": 0.0921, + "step": 400 + }, + { + "grad_norm": 1.16392183303833, + "learning_rate": 4.09e-05, + "loss": 0.0922, + "step": 410 + }, + { + "grad_norm": 1.2726706266403198, + "learning_rate": 4.19e-05, + "loss": 0.0907, + "step": 420 + }, + { + "grad_norm": 1.0239323377609253, + "learning_rate": 4.29e-05, + "loss": 0.0992, + "step": 430 + }, + { + "grad_norm": 1.0375834703445435, + "learning_rate": 4.39e-05, + "loss": 0.0829, + "step": 440 + }, + { + "grad_norm": 0.9939393401145935, + "learning_rate": 4.49e-05, + "loss": 0.0859, + "step": 450 + }, + { + "grad_norm": 1.231859803199768, + "learning_rate": 4.5900000000000004e-05, + "loss": 0.0869, + "step": 460 + }, + { + "grad_norm": 1.371180534362793, + "learning_rate": 4.69e-05, + "loss": 0.08, + "step": 470 + }, + { + "grad_norm": 1.1594104766845703, + "learning_rate": 4.79e-05, + "loss": 0.0866, + "step": 480 + }, + { + "grad_norm": 1.2831006050109863, + "learning_rate": 4.89e-05, + "loss": 0.0879, + "step": 490 + }, + { + "grad_norm": 0.9735462069511414, + "learning_rate": 4.99e-05, + "loss": 0.0831, + "step": 500 + }, + { + "grad_norm": 1.5769312381744385, + "learning_rate": 5.0900000000000004e-05, + "loss": 0.0819, + "step": 510 + }, + { + "grad_norm": 1.0352920293807983, + "learning_rate": 5.19e-05, + "loss": 0.0807, + "step": 520 + }, + { + "grad_norm": 0.8500271439552307, + "learning_rate": 5.2900000000000005e-05, + "loss": 0.0801, + "step": 530 + }, + { + "grad_norm": 1.1866955757141113, + "learning_rate": 5.390000000000001e-05, + "loss": 0.0798, + "step": 540 + }, + { + "grad_norm": 1.0223989486694336, + "learning_rate": 5.4900000000000006e-05, + "loss": 0.0731, + "step": 550 + }, + { + "grad_norm": 0.832564651966095, + "learning_rate": 5.590000000000001e-05, + "loss": 0.0788, + "step": 560 + }, + { + "grad_norm": 0.9835225939750671, + "learning_rate": 5.69e-05, + "loss": 0.0824, + "step": 570 + }, + { + "grad_norm": 0.7991761565208435, + "learning_rate": 5.79e-05, + "loss": 0.0772, + "step": 580 + }, + { + "grad_norm": 0.7952585816383362, + "learning_rate": 5.89e-05, + "loss": 0.0786, + "step": 590 + }, + { + "grad_norm": 0.9018053412437439, + "learning_rate": 5.99e-05, + "loss": 0.0784, + "step": 600 + }, + { + "grad_norm": 0.7779418230056763, + "learning_rate": 6.09e-05, + "loss": 0.0724, + "step": 610 + }, + { + "grad_norm": 0.8213123679161072, + "learning_rate": 6.19e-05, + "loss": 0.0739, + "step": 620 + }, + { + "grad_norm": 0.8696741461753845, + "learning_rate": 6.29e-05, + "loss": 0.0768, + "step": 630 + }, + { + "grad_norm": 0.7442203760147095, + "learning_rate": 6.390000000000001e-05, + "loss": 0.0822, + "step": 640 + }, + { + "grad_norm": 0.6419669985771179, + "learning_rate": 6.49e-05, + "loss": 0.0777, + "step": 650 + }, + { + "grad_norm": 0.7887861132621765, + "learning_rate": 6.59e-05, + "loss": 0.0726, + "step": 660 + }, + { + "grad_norm": 0.7971125245094299, + "learning_rate": 6.690000000000001e-05, + "loss": 0.0689, + "step": 670 + }, + { + "grad_norm": 0.9036775231361389, + "learning_rate": 6.790000000000001e-05, + "loss": 0.0636, + "step": 680 + }, + { + "grad_norm": 0.6422489285469055, + "learning_rate": 6.89e-05, + "loss": 0.0689, + "step": 690 + }, + { + "grad_norm": 0.977107048034668, + "learning_rate": 6.99e-05, + "loss": 0.0697, + "step": 700 + }, + { + "grad_norm": 0.8799305558204651, + "learning_rate": 7.09e-05, + "loss": 0.0661, + "step": 710 + }, + { + "grad_norm": 0.6521508693695068, + "learning_rate": 7.19e-05, + "loss": 0.0684, + "step": 720 + }, + { + "grad_norm": 0.7238072752952576, + "learning_rate": 7.29e-05, + "loss": 0.0706, + "step": 730 + }, + { + "grad_norm": 0.7217698097229004, + "learning_rate": 7.390000000000001e-05, + "loss": 0.0761, + "step": 740 + }, + { + "grad_norm": 0.85261070728302, + "learning_rate": 7.49e-05, + "loss": 0.0749, + "step": 750 + }, + { + "grad_norm": 0.784895122051239, + "learning_rate": 7.59e-05, + "loss": 0.0727, + "step": 760 + }, + { + "grad_norm": 0.8261733055114746, + "learning_rate": 7.69e-05, + "loss": 0.0774, + "step": 770 + }, + { + "grad_norm": 0.7314831614494324, + "learning_rate": 7.790000000000001e-05, + "loss": 0.0713, + "step": 780 + }, + { + "grad_norm": 0.641707718372345, + "learning_rate": 7.890000000000001e-05, + "loss": 0.0654, + "step": 790 + }, + { + "grad_norm": 0.7922617793083191, + "learning_rate": 7.99e-05, + "loss": 0.0629, + "step": 800 + }, + { + "grad_norm": 1.0089313983917236, + "learning_rate": 8.090000000000001e-05, + "loss": 0.0644, + "step": 810 + }, + { + "grad_norm": 0.4889024794101715, + "learning_rate": 8.19e-05, + "loss": 0.0697, + "step": 820 + }, + { + "grad_norm": 0.813235878944397, + "learning_rate": 8.29e-05, + "loss": 0.0701, + "step": 830 + }, + { + "grad_norm": 0.5666770935058594, + "learning_rate": 8.39e-05, + "loss": 0.0685, + "step": 840 + }, + { + "grad_norm": 0.6462077498435974, + "learning_rate": 8.49e-05, + "loss": 0.0701, + "step": 850 + }, + { + "grad_norm": 0.6827865839004517, + "learning_rate": 8.59e-05, + "loss": 0.065, + "step": 860 + }, + { + "grad_norm": 0.5797464847564697, + "learning_rate": 8.69e-05, + "loss": 0.0665, + "step": 870 + }, + { + "grad_norm": 0.590549647808075, + "learning_rate": 8.790000000000001e-05, + "loss": 0.0649, + "step": 880 + }, + { + "grad_norm": 0.5490359663963318, + "learning_rate": 8.89e-05, + "loss": 0.0645, + "step": 890 + }, + { + "grad_norm": 0.6479198932647705, + "learning_rate": 8.99e-05, + "loss": 0.0691, + "step": 900 + }, + { + "grad_norm": 0.39340826869010925, + "learning_rate": 9.090000000000001e-05, + "loss": 0.0586, + "step": 910 + }, + { + "grad_norm": 0.5507171750068665, + "learning_rate": 9.190000000000001e-05, + "loss": 0.0642, + "step": 920 + }, + { + "grad_norm": 0.8585442304611206, + "learning_rate": 9.290000000000001e-05, + "loss": 0.065, + "step": 930 + }, + { + "grad_norm": 0.49833944439888, + "learning_rate": 9.39e-05, + "loss": 0.0676, + "step": 940 + }, + { + "grad_norm": 0.8173732161521912, + "learning_rate": 9.49e-05, + "loss": 0.0673, + "step": 950 + }, + { + "grad_norm": 0.6390983462333679, + "learning_rate": 9.59e-05, + "loss": 0.0644, + "step": 960 + }, + { + "grad_norm": 0.5229816436767578, + "learning_rate": 9.69e-05, + "loss": 0.0664, + "step": 970 + }, + { + "grad_norm": 0.45658552646636963, + "learning_rate": 9.790000000000001e-05, + "loss": 0.0551, + "step": 980 + }, + { + "grad_norm": 0.5670000314712524, + "learning_rate": 9.89e-05, + "loss": 0.0685, + "step": 990 + }, + { + "grad_norm": 0.5052791833877563, + "learning_rate": 9.99e-05, + "loss": 0.0571, + "step": 1000 + }, + { + "grad_norm": 0.5222781896591187, + "learning_rate": 9.999994463727085e-05, + "loss": 0.0573, + "step": 1010 + }, + { + "grad_norm": 0.39941611886024475, + "learning_rate": 9.999975326009292e-05, + "loss": 0.0612, + "step": 1020 + }, + { + "grad_norm": 0.5795792937278748, + "learning_rate": 9.999942518549879e-05, + "loss": 0.0649, + "step": 1030 + }, + { + "grad_norm": 0.5003178715705872, + "learning_rate": 9.999896041438544e-05, + "loss": 0.059, + "step": 1040 + }, + { + "grad_norm": 0.7415033578872681, + "learning_rate": 9.999835894802353e-05, + "loss": 0.0562, + "step": 1050 + }, + { + "grad_norm": 0.533107340335846, + "learning_rate": 9.999762078805743e-05, + "loss": 0.0572, + "step": 1060 + }, + { + "grad_norm": 0.5728627443313599, + "learning_rate": 9.999674593650526e-05, + "loss": 0.0631, + "step": 1070 + }, + { + "grad_norm": 0.4351804852485657, + "learning_rate": 9.99957343957588e-05, + "loss": 0.0576, + "step": 1080 + }, + { + "grad_norm": 0.5238198041915894, + "learning_rate": 9.99945861685836e-05, + "loss": 0.0635, + "step": 1090 + }, + { + "grad_norm": 0.5429909825325012, + "learning_rate": 9.999330125811884e-05, + "loss": 0.0611, + "step": 1100 + }, + { + "grad_norm": 0.5178719162940979, + "learning_rate": 9.999187966787744e-05, + "loss": 0.0529, + "step": 1110 + }, + { + "grad_norm": 0.6097676157951355, + "learning_rate": 9.999032140174595e-05, + "loss": 0.0572, + "step": 1120 + }, + { + "grad_norm": 0.36114999651908875, + "learning_rate": 9.998862646398464e-05, + "loss": 0.0629, + "step": 1130 + }, + { + "grad_norm": 0.7032138109207153, + "learning_rate": 9.998679485922739e-05, + "loss": 0.0589, + "step": 1140 + }, + { + "grad_norm": 0.3891696631908417, + "learning_rate": 9.998482659248174e-05, + "loss": 0.053, + "step": 1150 + }, + { + "grad_norm": 0.5694180727005005, + "learning_rate": 9.998272166912883e-05, + "loss": 0.0575, + "step": 1160 + }, + { + "grad_norm": 0.5297912955284119, + "learning_rate": 9.998048009492347e-05, + "loss": 0.0584, + "step": 1170 + }, + { + "grad_norm": 0.4530673921108246, + "learning_rate": 9.997810187599403e-05, + "loss": 0.0559, + "step": 1180 + }, + { + "grad_norm": 0.5671072006225586, + "learning_rate": 9.997558701884249e-05, + "loss": 0.0548, + "step": 1190 + }, + { + "grad_norm": 0.6415340900421143, + "learning_rate": 9.997293553034433e-05, + "loss": 0.0588, + "step": 1200 + }, + { + "grad_norm": 0.5936589241027832, + "learning_rate": 9.997014741774866e-05, + "loss": 0.053, + "step": 1210 + }, + { + "grad_norm": 0.36478346586227417, + "learning_rate": 9.996722268867803e-05, + "loss": 0.051, + "step": 1220 + }, + { + "grad_norm": 0.5621365308761597, + "learning_rate": 9.996416135112858e-05, + "loss": 0.0533, + "step": 1230 + }, + { + "grad_norm": 0.5009233355522156, + "learning_rate": 9.996096341346988e-05, + "loss": 0.0544, + "step": 1240 + }, + { + "grad_norm": 0.43907880783081055, + "learning_rate": 9.995762888444495e-05, + "loss": 0.0477, + "step": 1250 + }, + { + "grad_norm": 0.47513559460639954, + "learning_rate": 9.995415777317027e-05, + "loss": 0.0502, + "step": 1260 + }, + { + "grad_norm": 0.3658427894115448, + "learning_rate": 9.995055008913574e-05, + "loss": 0.0503, + "step": 1270 + }, + { + "grad_norm": 0.30684512853622437, + "learning_rate": 9.994680584220463e-05, + "loss": 0.0537, + "step": 1280 + }, + { + "grad_norm": 0.43991294503211975, + "learning_rate": 9.994292504261355e-05, + "loss": 0.0502, + "step": 1290 + }, + { + "grad_norm": 0.405241459608078, + "learning_rate": 9.993890770097247e-05, + "loss": 0.0531, + "step": 1300 + }, + { + "grad_norm": 0.48382768034935, + "learning_rate": 9.993475382826467e-05, + "loss": 0.0523, + "step": 1310 + }, + { + "grad_norm": 0.5165296196937561, + "learning_rate": 9.993046343584664e-05, + "loss": 0.0483, + "step": 1320 + }, + { + "grad_norm": 0.3143220543861389, + "learning_rate": 9.992603653544816e-05, + "loss": 0.0553, + "step": 1330 + }, + { + "grad_norm": 0.39124631881713867, + "learning_rate": 9.992147313917222e-05, + "loss": 0.0519, + "step": 1340 + }, + { + "grad_norm": 0.46568813920021057, + "learning_rate": 9.991677325949497e-05, + "loss": 0.0535, + "step": 1350 + }, + { + "grad_norm": 0.445818692445755, + "learning_rate": 9.991193690926568e-05, + "loss": 0.0477, + "step": 1360 + }, + { + "grad_norm": 0.34483277797698975, + "learning_rate": 9.990696410170678e-05, + "loss": 0.0513, + "step": 1370 + }, + { + "grad_norm": 0.3180125653743744, + "learning_rate": 9.990185485041371e-05, + "loss": 0.0437, + "step": 1380 + }, + { + "grad_norm": 0.3598921597003937, + "learning_rate": 9.989660916935498e-05, + "loss": 0.0456, + "step": 1390 + }, + { + "grad_norm": 0.5585952401161194, + "learning_rate": 9.989122707287208e-05, + "loss": 0.0481, + "step": 1400 + }, + { + "grad_norm": 0.4544965624809265, + "learning_rate": 9.988570857567945e-05, + "loss": 0.0472, + "step": 1410 + }, + { + "grad_norm": 0.46985459327697754, + "learning_rate": 9.988005369286446e-05, + "loss": 0.0478, + "step": 1420 + }, + { + "grad_norm": 0.3741517663002014, + "learning_rate": 9.987426243988734e-05, + "loss": 0.048, + "step": 1430 + }, + { + "grad_norm": 0.4785807728767395, + "learning_rate": 9.986833483258114e-05, + "loss": 0.047, + "step": 1440 + }, + { + "grad_norm": 0.5361403226852417, + "learning_rate": 9.986227088715173e-05, + "loss": 0.0491, + "step": 1450 + }, + { + "grad_norm": 0.4387078583240509, + "learning_rate": 9.98560706201777e-05, + "loss": 0.0547, + "step": 1460 + }, + { + "grad_norm": 0.7411963939666748, + "learning_rate": 9.984973404861036e-05, + "loss": 0.0477, + "step": 1470 + }, + { + "grad_norm": 0.2766841650009155, + "learning_rate": 9.984326118977361e-05, + "loss": 0.0512, + "step": 1480 + }, + { + "grad_norm": 0.34676310420036316, + "learning_rate": 9.983665206136406e-05, + "loss": 0.0475, + "step": 1490 + }, + { + "grad_norm": 0.37037667632102966, + "learning_rate": 9.982990668145075e-05, + "loss": 0.0493, + "step": 1500 + }, + { + "grad_norm": 0.6023677587509155, + "learning_rate": 9.982302506847534e-05, + "loss": 0.0483, + "step": 1510 + }, + { + "grad_norm": 0.2910383343696594, + "learning_rate": 9.981600724125189e-05, + "loss": 0.0461, + "step": 1520 + }, + { + "grad_norm": 0.3197880685329437, + "learning_rate": 9.980885321896685e-05, + "loss": 0.0466, + "step": 1530 + }, + { + "grad_norm": 0.4805852174758911, + "learning_rate": 9.980156302117905e-05, + "loss": 0.0499, + "step": 1540 + }, + { + "grad_norm": 0.458696573972702, + "learning_rate": 9.979413666781963e-05, + "loss": 0.048, + "step": 1550 + }, + { + "grad_norm": 0.5970068573951721, + "learning_rate": 9.978657417919193e-05, + "loss": 0.0515, + "step": 1560 + }, + { + "grad_norm": 0.5699660778045654, + "learning_rate": 9.977887557597153e-05, + "loss": 0.0506, + "step": 1570 + }, + { + "grad_norm": 0.3206250071525574, + "learning_rate": 9.97710408792061e-05, + "loss": 0.0488, + "step": 1580 + }, + { + "grad_norm": 0.5578159093856812, + "learning_rate": 9.976307011031542e-05, + "loss": 0.0463, + "step": 1590 + }, + { + "grad_norm": 0.45685502886772156, + "learning_rate": 9.975496329109126e-05, + "loss": 0.0523, + "step": 1600 + }, + { + "grad_norm": 0.3745459318161011, + "learning_rate": 9.974672044369732e-05, + "loss": 0.0478, + "step": 1610 + }, + { + "grad_norm": 0.49243006110191345, + "learning_rate": 9.97383415906693e-05, + "loss": 0.0444, + "step": 1620 + }, + { + "grad_norm": 0.3611481189727783, + "learning_rate": 9.97298267549146e-05, + "loss": 0.048, + "step": 1630 + }, + { + "grad_norm": 0.35129472613334656, + "learning_rate": 9.972117595971249e-05, + "loss": 0.0439, + "step": 1640 + }, + { + "grad_norm": 0.5138766765594482, + "learning_rate": 9.971238922871391e-05, + "loss": 0.042, + "step": 1650 + }, + { + "grad_norm": 0.26045840978622437, + "learning_rate": 9.970346658594142e-05, + "loss": 0.0448, + "step": 1660 + }, + { + "grad_norm": 0.3174215853214264, + "learning_rate": 9.969440805578923e-05, + "loss": 0.0475, + "step": 1670 + }, + { + "grad_norm": 0.565101683139801, + "learning_rate": 9.968521366302298e-05, + "loss": 0.0498, + "step": 1680 + }, + { + "grad_norm": 0.4355429708957672, + "learning_rate": 9.967588343277981e-05, + "loss": 0.0569, + "step": 1690 + }, + { + "grad_norm": 0.5504177808761597, + "learning_rate": 9.966641739056818e-05, + "loss": 0.0521, + "step": 1700 + }, + { + "grad_norm": 0.4876561164855957, + "learning_rate": 9.965681556226793e-05, + "loss": 0.0488, + "step": 1710 + }, + { + "grad_norm": 0.45689451694488525, + "learning_rate": 9.964707797413006e-05, + "loss": 0.0488, + "step": 1720 + }, + { + "grad_norm": 0.4040215015411377, + "learning_rate": 9.963720465277679e-05, + "loss": 0.0455, + "step": 1730 + }, + { + "grad_norm": 0.44894132018089294, + "learning_rate": 9.96271956252014e-05, + "loss": 0.0493, + "step": 1740 + }, + { + "grad_norm": 0.4604091942310333, + "learning_rate": 9.961705091876816e-05, + "loss": 0.047, + "step": 1750 + }, + { + "grad_norm": 0.34120258688926697, + "learning_rate": 9.960677056121235e-05, + "loss": 0.0458, + "step": 1760 + }, + { + "grad_norm": 0.35755547881126404, + "learning_rate": 9.959635458064005e-05, + "loss": 0.05, + "step": 1770 + }, + { + "grad_norm": 0.4225219488143921, + "learning_rate": 9.958580300552815e-05, + "loss": 0.0453, + "step": 1780 + }, + { + "grad_norm": 0.4329666793346405, + "learning_rate": 9.957511586472426e-05, + "loss": 0.0499, + "step": 1790 + }, + { + "grad_norm": 0.5889803171157837, + "learning_rate": 9.956429318744662e-05, + "loss": 0.0461, + "step": 1800 + }, + { + "grad_norm": 0.29788509011268616, + "learning_rate": 9.955333500328404e-05, + "loss": 0.0438, + "step": 1810 + }, + { + "grad_norm": 0.5277188420295715, + "learning_rate": 9.95422413421957e-05, + "loss": 0.0455, + "step": 1820 + }, + { + "grad_norm": 0.4105511009693146, + "learning_rate": 9.953101223451133e-05, + "loss": 0.0411, + "step": 1830 + }, + { + "grad_norm": 0.39845457673072815, + "learning_rate": 9.951964771093085e-05, + "loss": 0.0462, + "step": 1840 + }, + { + "grad_norm": 0.4860050678253174, + "learning_rate": 9.950814780252442e-05, + "loss": 0.0492, + "step": 1850 + }, + { + "grad_norm": 0.3665149509906769, + "learning_rate": 9.949651254073236e-05, + "loss": 0.0482, + "step": 1860 + }, + { + "grad_norm": 0.3214748799800873, + "learning_rate": 9.948474195736504e-05, + "loss": 0.0461, + "step": 1870 + }, + { + "grad_norm": 0.26840588450431824, + "learning_rate": 9.947283608460277e-05, + "loss": 0.0449, + "step": 1880 + }, + { + "grad_norm": 0.4030681550502777, + "learning_rate": 9.946079495499577e-05, + "loss": 0.0438, + "step": 1890 + }, + { + "grad_norm": 0.4224938452243805, + "learning_rate": 9.944861860146401e-05, + "loss": 0.0414, + "step": 1900 + }, + { + "grad_norm": 0.3272496163845062, + "learning_rate": 9.943630705729719e-05, + "loss": 0.0417, + "step": 1910 + }, + { + "grad_norm": 0.44444161653518677, + "learning_rate": 9.942386035615459e-05, + "loss": 0.0453, + "step": 1920 + }, + { + "grad_norm": 0.4095990061759949, + "learning_rate": 9.941127853206503e-05, + "loss": 0.0462, + "step": 1930 + }, + { + "grad_norm": 0.34056487679481506, + "learning_rate": 9.939856161942673e-05, + "loss": 0.0444, + "step": 1940 + }, + { + "grad_norm": 0.245171457529068, + "learning_rate": 9.938570965300724e-05, + "loss": 0.0413, + "step": 1950 + }, + { + "grad_norm": 0.32479041814804077, + "learning_rate": 9.937272266794335e-05, + "loss": 0.043, + "step": 1960 + }, + { + "grad_norm": 0.31490927934646606, + "learning_rate": 9.935960069974096e-05, + "loss": 0.0447, + "step": 1970 + }, + { + "grad_norm": 0.2531657814979553, + "learning_rate": 9.934634378427506e-05, + "loss": 0.0418, + "step": 1980 + }, + { + "grad_norm": 0.4324510395526886, + "learning_rate": 9.933295195778954e-05, + "loss": 0.0445, + "step": 1990 + }, + { + "grad_norm": 0.31543952226638794, + "learning_rate": 9.931942525689715e-05, + "loss": 0.0435, + "step": 2000 + }, + { + "grad_norm": 0.3936459422111511, + "learning_rate": 9.930576371857936e-05, + "loss": 0.0454, + "step": 2010 + }, + { + "grad_norm": 0.4099113941192627, + "learning_rate": 9.929196738018629e-05, + "loss": 0.0456, + "step": 2020 + }, + { + "grad_norm": 0.2855170667171478, + "learning_rate": 9.927803627943662e-05, + "loss": 0.0461, + "step": 2030 + }, + { + "grad_norm": 0.25442031025886536, + "learning_rate": 9.926397045441744e-05, + "loss": 0.0401, + "step": 2040 + }, + { + "grad_norm": 0.3943769037723541, + "learning_rate": 9.924976994358417e-05, + "loss": 0.0435, + "step": 2050 + }, + { + "grad_norm": 0.4695734977722168, + "learning_rate": 9.923543478576048e-05, + "loss": 0.0426, + "step": 2060 + }, + { + "grad_norm": 0.545697033405304, + "learning_rate": 9.922096502013813e-05, + "loss": 0.0449, + "step": 2070 + }, + { + "grad_norm": 0.300885409116745, + "learning_rate": 9.92063606862769e-05, + "loss": 0.0411, + "step": 2080 + }, + { + "grad_norm": 0.322279691696167, + "learning_rate": 9.919162182410453e-05, + "loss": 0.0467, + "step": 2090 + }, + { + "grad_norm": 0.3021215498447418, + "learning_rate": 9.917674847391645e-05, + "loss": 0.0448, + "step": 2100 + }, + { + "grad_norm": 0.3706432282924652, + "learning_rate": 9.916174067637584e-05, + "loss": 0.0393, + "step": 2110 + }, + { + "grad_norm": 0.4282996356487274, + "learning_rate": 9.914659847251348e-05, + "loss": 0.0411, + "step": 2120 + }, + { + "grad_norm": 0.509666919708252, + "learning_rate": 9.913132190372753e-05, + "loss": 0.0462, + "step": 2130 + }, + { + "grad_norm": 0.3906203508377075, + "learning_rate": 9.911591101178359e-05, + "loss": 0.0405, + "step": 2140 + }, + { + "grad_norm": 0.4438856244087219, + "learning_rate": 9.910036583881443e-05, + "loss": 0.0442, + "step": 2150 + }, + { + "grad_norm": 0.2545924782752991, + "learning_rate": 9.908468642731995e-05, + "loss": 0.0442, + "step": 2160 + }, + { + "grad_norm": 0.28592199087142944, + "learning_rate": 9.906887282016707e-05, + "loss": 0.041, + "step": 2170 + }, + { + "grad_norm": 0.3645898699760437, + "learning_rate": 9.90529250605896e-05, + "loss": 0.0445, + "step": 2180 + }, + { + "grad_norm": 0.4444137215614319, + "learning_rate": 9.903684319218809e-05, + "loss": 0.0422, + "step": 2190 + }, + { + "grad_norm": 0.37915918231010437, + "learning_rate": 9.902062725892976e-05, + "loss": 0.0494, + "step": 2200 + }, + { + "grad_norm": 0.28666216135025024, + "learning_rate": 9.900427730514834e-05, + "loss": 0.0447, + "step": 2210 + }, + { + "grad_norm": 0.2940952777862549, + "learning_rate": 9.8987793375544e-05, + "loss": 0.0434, + "step": 2220 + }, + { + "grad_norm": 0.37347811460494995, + "learning_rate": 9.897117551518318e-05, + "loss": 0.0377, + "step": 2230 + }, + { + "grad_norm": 0.395631343126297, + "learning_rate": 9.895442376949844e-05, + "loss": 0.0397, + "step": 2240 + }, + { + "grad_norm": 0.33471351861953735, + "learning_rate": 9.893753818428845e-05, + "loss": 0.0438, + "step": 2250 + }, + { + "grad_norm": 0.2978215217590332, + "learning_rate": 9.892051880571773e-05, + "loss": 0.0485, + "step": 2260 + }, + { + "grad_norm": 0.37751349806785583, + "learning_rate": 9.890336568031663e-05, + "loss": 0.0459, + "step": 2270 + }, + { + "grad_norm": 0.3284722566604614, + "learning_rate": 9.888607885498113e-05, + "loss": 0.0397, + "step": 2280 + }, + { + "grad_norm": 0.3817616105079651, + "learning_rate": 9.886865837697275e-05, + "loss": 0.0406, + "step": 2290 + }, + { + "grad_norm": 0.2554313540458679, + "learning_rate": 9.88511042939184e-05, + "loss": 0.0379, + "step": 2300 + }, + { + "grad_norm": 0.24379108846187592, + "learning_rate": 9.883341665381028e-05, + "loss": 0.0398, + "step": 2310 + }, + { + "grad_norm": 0.41589999198913574, + "learning_rate": 9.881559550500575e-05, + "loss": 0.0401, + "step": 2320 + }, + { + "grad_norm": 0.30908241868019104, + "learning_rate": 9.879764089622712e-05, + "loss": 0.0485, + "step": 2330 + }, + { + "grad_norm": 0.3617725968360901, + "learning_rate": 9.87795528765616e-05, + "loss": 0.0383, + "step": 2340 + }, + { + "grad_norm": 0.3001377284526825, + "learning_rate": 9.876133149546118e-05, + "loss": 0.047, + "step": 2350 + }, + { + "grad_norm": 0.3865724802017212, + "learning_rate": 9.874297680274238e-05, + "loss": 0.0425, + "step": 2360 + }, + { + "grad_norm": 0.32195764780044556, + "learning_rate": 9.872448884858624e-05, + "loss": 0.0428, + "step": 2370 + }, + { + "grad_norm": 0.399752676486969, + "learning_rate": 9.870586768353815e-05, + "loss": 0.0427, + "step": 2380 + }, + { + "grad_norm": 0.43321412801742554, + "learning_rate": 9.868711335850764e-05, + "loss": 0.0395, + "step": 2390 + }, + { + "grad_norm": 0.43004199862480164, + "learning_rate": 9.866822592476833e-05, + "loss": 0.0404, + "step": 2400 + }, + { + "grad_norm": 0.24631984531879425, + "learning_rate": 9.86492054339577e-05, + "loss": 0.0364, + "step": 2410 + }, + { + "grad_norm": 0.3020949959754944, + "learning_rate": 9.863005193807711e-05, + "loss": 0.0466, + "step": 2420 + }, + { + "grad_norm": 0.3032028377056122, + "learning_rate": 9.861076548949143e-05, + "loss": 0.0377, + "step": 2430 + }, + { + "grad_norm": 0.32794851064682007, + "learning_rate": 9.859134614092912e-05, + "loss": 0.0401, + "step": 2440 + }, + { + "grad_norm": 0.2794005274772644, + "learning_rate": 9.857179394548191e-05, + "loss": 0.0377, + "step": 2450 + }, + { + "grad_norm": 0.2938261926174164, + "learning_rate": 9.855210895660477e-05, + "loss": 0.0375, + "step": 2460 + }, + { + "grad_norm": 0.26473769545555115, + "learning_rate": 9.853229122811568e-05, + "loss": 0.0423, + "step": 2470 + }, + { + "grad_norm": 0.31394049525260925, + "learning_rate": 9.851234081419559e-05, + "loss": 0.0356, + "step": 2480 + }, + { + "grad_norm": 0.3999021649360657, + "learning_rate": 9.849225776938814e-05, + "loss": 0.0476, + "step": 2490 + }, + { + "grad_norm": 0.28824809193611145, + "learning_rate": 9.847204214859964e-05, + "loss": 0.0402, + "step": 2500 + }, + { + "grad_norm": 0.29327067732810974, + "learning_rate": 9.845169400709879e-05, + "loss": 0.0455, + "step": 2510 + }, + { + "grad_norm": 0.3950151801109314, + "learning_rate": 9.843121340051664e-05, + "loss": 0.0436, + "step": 2520 + }, + { + "grad_norm": 0.3586881458759308, + "learning_rate": 9.841060038484641e-05, + "loss": 0.041, + "step": 2530 + }, + { + "grad_norm": 0.29140523076057434, + "learning_rate": 9.838985501644328e-05, + "loss": 0.0356, + "step": 2540 + }, + { + "grad_norm": 0.40005186200141907, + "learning_rate": 9.83689773520243e-05, + "loss": 0.0422, + "step": 2550 + }, + { + "grad_norm": 0.32647910714149475, + "learning_rate": 9.834796744866819e-05, + "loss": 0.0407, + "step": 2560 + }, + { + "grad_norm": 0.2351028472185135, + "learning_rate": 9.832682536381525e-05, + "loss": 0.0378, + "step": 2570 + }, + { + "grad_norm": 0.24619437754154205, + "learning_rate": 9.830555115526711e-05, + "loss": 0.0392, + "step": 2580 + }, + { + "grad_norm": 0.3009263575077057, + "learning_rate": 9.828414488118667e-05, + "loss": 0.0345, + "step": 2590 + }, + { + "grad_norm": 0.32350802421569824, + "learning_rate": 9.826260660009785e-05, + "loss": 0.0402, + "step": 2600 + }, + { + "grad_norm": 0.3015955984592438, + "learning_rate": 9.824093637088547e-05, + "loss": 0.0391, + "step": 2610 + }, + { + "grad_norm": 0.30525708198547363, + "learning_rate": 9.821913425279514e-05, + "loss": 0.0389, + "step": 2620 + }, + { + "grad_norm": 0.22946375608444214, + "learning_rate": 9.8197200305433e-05, + "loss": 0.0363, + "step": 2630 + }, + { + "grad_norm": 0.3912782669067383, + "learning_rate": 9.817513458876564e-05, + "loss": 0.0385, + "step": 2640 + }, + { + "grad_norm": 0.28584063053131104, + "learning_rate": 9.815293716311987e-05, + "loss": 0.0383, + "step": 2650 + }, + { + "grad_norm": 0.34168604016304016, + "learning_rate": 9.813060808918262e-05, + "loss": 0.04, + "step": 2660 + }, + { + "grad_norm": 0.2768874764442444, + "learning_rate": 9.810814742800069e-05, + "loss": 0.0388, + "step": 2670 + }, + { + "grad_norm": 0.31885719299316406, + "learning_rate": 9.808555524098074e-05, + "loss": 0.0364, + "step": 2680 + }, + { + "grad_norm": 0.320256769657135, + "learning_rate": 9.806283158988887e-05, + "loss": 0.0376, + "step": 2690 + }, + { + "grad_norm": 0.4063742756843567, + "learning_rate": 9.803997653685072e-05, + "loss": 0.0395, + "step": 2700 + }, + { + "grad_norm": 0.24014152586460114, + "learning_rate": 9.801699014435112e-05, + "loss": 0.0413, + "step": 2710 + }, + { + "grad_norm": 0.3252515196800232, + "learning_rate": 9.799387247523398e-05, + "loss": 0.0419, + "step": 2720 + }, + { + "grad_norm": 0.30467653274536133, + "learning_rate": 9.797062359270215e-05, + "loss": 0.0382, + "step": 2730 + }, + { + "grad_norm": 0.31427666544914246, + "learning_rate": 9.794724356031715e-05, + "loss": 0.0348, + "step": 2740 + }, + { + "grad_norm": 0.23562613129615784, + "learning_rate": 9.792373244199913e-05, + "loss": 0.0366, + "step": 2750 + }, + { + "grad_norm": 0.33076491951942444, + "learning_rate": 9.790009030202658e-05, + "loss": 0.0375, + "step": 2760 + }, + { + "grad_norm": 0.35463470220565796, + "learning_rate": 9.78763172050362e-05, + "loss": 0.0393, + "step": 2770 + }, + { + "grad_norm": 0.3029499053955078, + "learning_rate": 9.785241321602274e-05, + "loss": 0.0363, + "step": 2780 + }, + { + "grad_norm": 0.285557359457016, + "learning_rate": 9.782837840033879e-05, + "loss": 0.0342, + "step": 2790 + }, + { + "grad_norm": 0.45977655053138733, + "learning_rate": 9.780421282369461e-05, + "loss": 0.0367, + "step": 2800 + }, + { + "grad_norm": 0.2490273416042328, + "learning_rate": 9.777991655215797e-05, + "loss": 0.0416, + "step": 2810 + }, + { + "grad_norm": 0.28247377276420593, + "learning_rate": 9.775548965215394e-05, + "loss": 0.041, + "step": 2820 + }, + { + "grad_norm": 0.41089487075805664, + "learning_rate": 9.773093219046474e-05, + "loss": 0.0363, + "step": 2830 + }, + { + "grad_norm": 0.35828647017478943, + "learning_rate": 9.770624423422954e-05, + "loss": 0.0397, + "step": 2840 + }, + { + "grad_norm": 0.28381896018981934, + "learning_rate": 9.768142585094426e-05, + "loss": 0.0399, + "step": 2850 + }, + { + "grad_norm": 0.26408782601356506, + "learning_rate": 9.765647710846142e-05, + "loss": 0.0419, + "step": 2860 + }, + { + "grad_norm": 0.5101597905158997, + "learning_rate": 9.763139807498991e-05, + "loss": 0.0388, + "step": 2870 + }, + { + "grad_norm": 0.22416441142559052, + "learning_rate": 9.760618881909487e-05, + "loss": 0.0372, + "step": 2880 + }, + { + "grad_norm": 0.2925640642642975, + "learning_rate": 9.758084940969744e-05, + "loss": 0.0399, + "step": 2890 + }, + { + "grad_norm": 0.35014769434928894, + "learning_rate": 9.755537991607459e-05, + "loss": 0.0362, + "step": 2900 + }, + { + "grad_norm": 0.24538226425647736, + "learning_rate": 9.752978040785895e-05, + "loss": 0.0375, + "step": 2910 + }, + { + "grad_norm": 0.3659448027610779, + "learning_rate": 9.750405095503859e-05, + "loss": 0.04, + "step": 2920 + }, + { + "grad_norm": 0.3367757201194763, + "learning_rate": 9.747819162795686e-05, + "loss": 0.0402, + "step": 2930 + }, + { + "grad_norm": 0.2855086922645569, + "learning_rate": 9.745220249731217e-05, + "loss": 0.0417, + "step": 2940 + }, + { + "grad_norm": 0.4202945828437805, + "learning_rate": 9.742608363415781e-05, + "loss": 0.0402, + "step": 2950 + }, + { + "grad_norm": 0.414223849773407, + "learning_rate": 9.739983510990176e-05, + "loss": 0.0413, + "step": 2960 + }, + { + "grad_norm": 0.2509484887123108, + "learning_rate": 9.737345699630647e-05, + "loss": 0.0374, + "step": 2970 + }, + { + "grad_norm": 0.4225872755050659, + "learning_rate": 9.734694936548869e-05, + "loss": 0.0378, + "step": 2980 + }, + { + "grad_norm": 0.34358322620391846, + "learning_rate": 9.732031228991932e-05, + "loss": 0.0358, + "step": 2990 + }, + { + "grad_norm": 0.4824865758419037, + "learning_rate": 9.729354584242302e-05, + "loss": 0.0379, + "step": 3000 + }, + { + "grad_norm": 0.2874622642993927, + "learning_rate": 9.726665009617832e-05, + "loss": 0.0384, + "step": 3010 + }, + { + "grad_norm": 0.40659430623054504, + "learning_rate": 9.723962512471714e-05, + "loss": 0.04, + "step": 3020 + }, + { + "grad_norm": 0.32021278142929077, + "learning_rate": 9.72124710019247e-05, + "loss": 0.034, + "step": 3030 + }, + { + "grad_norm": 0.2295599728822708, + "learning_rate": 9.718518780203934e-05, + "loss": 0.0409, + "step": 3040 + }, + { + "grad_norm": 0.40697047114372253, + "learning_rate": 9.715777559965228e-05, + "loss": 0.0348, + "step": 3050 + }, + { + "grad_norm": 0.3346352279186249, + "learning_rate": 9.713023446970746e-05, + "loss": 0.036, + "step": 3060 + }, + { + "grad_norm": 0.2871566712856293, + "learning_rate": 9.710256448750126e-05, + "loss": 0.0347, + "step": 3070 + }, + { + "grad_norm": 0.31119096279144287, + "learning_rate": 9.707476572868235e-05, + "loss": 0.0344, + "step": 3080 + }, + { + "grad_norm": 0.3240164518356323, + "learning_rate": 9.704683826925149e-05, + "loss": 0.041, + "step": 3090 + }, + { + "grad_norm": 0.22162267565727234, + "learning_rate": 9.701878218556129e-05, + "loss": 0.0382, + "step": 3100 + }, + { + "grad_norm": 0.22689464688301086, + "learning_rate": 9.699059755431598e-05, + "loss": 0.0367, + "step": 3110 + }, + { + "grad_norm": 0.27198609709739685, + "learning_rate": 9.696228445257132e-05, + "loss": 0.035, + "step": 3120 + }, + { + "grad_norm": 0.23883281648159027, + "learning_rate": 9.693384295773419e-05, + "loss": 0.0363, + "step": 3130 + }, + { + "grad_norm": 0.20870637893676758, + "learning_rate": 9.690527314756259e-05, + "loss": 0.0348, + "step": 3140 + }, + { + "grad_norm": 0.39633700251579285, + "learning_rate": 9.687657510016527e-05, + "loss": 0.0384, + "step": 3150 + }, + { + "grad_norm": 0.2578440308570862, + "learning_rate": 9.684774889400161e-05, + "loss": 0.0345, + "step": 3160 + }, + { + "grad_norm": 0.23767991364002228, + "learning_rate": 9.681879460788135e-05, + "loss": 0.033, + "step": 3170 + }, + { + "grad_norm": 0.38834938406944275, + "learning_rate": 9.67897123209644e-05, + "loss": 0.0342, + "step": 3180 + }, + { + "grad_norm": 0.3847854733467102, + "learning_rate": 9.676050211276062e-05, + "loss": 0.0371, + "step": 3190 + }, + { + "grad_norm": 0.29791954159736633, + "learning_rate": 9.673116406312962e-05, + "loss": 0.0372, + "step": 3200 + }, + { + "grad_norm": 0.2858186960220337, + "learning_rate": 9.67016982522805e-05, + "loss": 0.0394, + "step": 3210 + }, + { + "grad_norm": 0.31906649470329285, + "learning_rate": 9.667210476077164e-05, + "loss": 0.0352, + "step": 3220 + }, + { + "grad_norm": 0.2839038670063019, + "learning_rate": 9.664238366951055e-05, + "loss": 0.031, + "step": 3230 + }, + { + "grad_norm": 0.206875279545784, + "learning_rate": 9.661253505975355e-05, + "loss": 0.0359, + "step": 3240 + }, + { + "grad_norm": 0.25959134101867676, + "learning_rate": 9.658255901310557e-05, + "loss": 0.0318, + "step": 3250 + }, + { + "grad_norm": 0.28207266330718994, + "learning_rate": 9.655245561152e-05, + "loss": 0.037, + "step": 3260 + }, + { + "grad_norm": 0.20452620089054108, + "learning_rate": 9.65222249372984e-05, + "loss": 0.0322, + "step": 3270 + }, + { + "grad_norm": 0.20434054732322693, + "learning_rate": 9.649186707309026e-05, + "loss": 0.0364, + "step": 3280 + }, + { + "grad_norm": 0.4000234305858612, + "learning_rate": 9.646138210189283e-05, + "loss": 0.039, + "step": 3290 + }, + { + "grad_norm": 0.4067232012748718, + "learning_rate": 9.643077010705087e-05, + "loss": 0.0384, + "step": 3300 + }, + { + "grad_norm": 0.30296677350997925, + "learning_rate": 9.640003117225637e-05, + "loss": 0.036, + "step": 3310 + }, + { + "grad_norm": 0.3965762257575989, + "learning_rate": 9.636916538154846e-05, + "loss": 0.0359, + "step": 3320 + }, + { + "grad_norm": 0.3527912199497223, + "learning_rate": 9.633817281931296e-05, + "loss": 0.0341, + "step": 3330 + }, + { + "grad_norm": 0.34234344959259033, + "learning_rate": 9.630705357028242e-05, + "loss": 0.0372, + "step": 3340 + }, + { + "grad_norm": 0.292255699634552, + "learning_rate": 9.627580771953563e-05, + "loss": 0.0346, + "step": 3350 + }, + { + "grad_norm": 0.21248452365398407, + "learning_rate": 9.624443535249759e-05, + "loss": 0.0338, + "step": 3360 + }, + { + "grad_norm": 0.3184260129928589, + "learning_rate": 9.621293655493913e-05, + "loss": 0.0358, + "step": 3370 + }, + { + "grad_norm": 0.4206657409667969, + "learning_rate": 9.618131141297675e-05, + "loss": 0.0341, + "step": 3380 + }, + { + "grad_norm": 0.20620489120483398, + "learning_rate": 9.614956001307242e-05, + "loss": 0.0335, + "step": 3390 + }, + { + "grad_norm": 0.29552599787712097, + "learning_rate": 9.611768244203321e-05, + "loss": 0.0348, + "step": 3400 + }, + { + "grad_norm": 0.41781607270240784, + "learning_rate": 9.60856787870112e-05, + "loss": 0.0356, + "step": 3410 + }, + { + "grad_norm": 0.3652491569519043, + "learning_rate": 9.605354913550318e-05, + "loss": 0.0391, + "step": 3420 + }, + { + "grad_norm": 0.5129753351211548, + "learning_rate": 9.602129357535037e-05, + "loss": 0.0394, + "step": 3430 + }, + { + "grad_norm": 0.3734123110771179, + "learning_rate": 9.598891219473825e-05, + "loss": 0.0441, + "step": 3440 + }, + { + "grad_norm": 0.36078277230262756, + "learning_rate": 9.595640508219625e-05, + "loss": 0.0403, + "step": 3450 + }, + { + "grad_norm": 0.26294559240341187, + "learning_rate": 9.592377232659761e-05, + "loss": 0.0431, + "step": 3460 + }, + { + "grad_norm": 0.2595246136188507, + "learning_rate": 9.589101401715904e-05, + "loss": 0.0337, + "step": 3470 + }, + { + "grad_norm": 0.34734177589416504, + "learning_rate": 9.585813024344045e-05, + "loss": 0.0387, + "step": 3480 + }, + { + "grad_norm": 0.2872370481491089, + "learning_rate": 9.58251210953449e-05, + "loss": 0.035, + "step": 3490 + }, + { + "grad_norm": 0.25488224625587463, + "learning_rate": 9.579198666311809e-05, + "loss": 0.0332, + "step": 3500 + }, + { + "grad_norm": 0.23626229166984558, + "learning_rate": 9.575872703734832e-05, + "loss": 0.0319, + "step": 3510 + }, + { + "grad_norm": 0.34818294644355774, + "learning_rate": 9.572534230896611e-05, + "loss": 0.0346, + "step": 3520 + }, + { + "grad_norm": 0.5608959794044495, + "learning_rate": 9.569183256924403e-05, + "loss": 0.0347, + "step": 3530 + }, + { + "grad_norm": 0.24106062948703766, + "learning_rate": 9.565819790979646e-05, + "loss": 0.0341, + "step": 3540 + }, + { + "grad_norm": 0.24844060838222504, + "learning_rate": 9.562443842257925e-05, + "loss": 0.0352, + "step": 3550 + }, + { + "grad_norm": 0.36223217844963074, + "learning_rate": 9.559055419988956e-05, + "loss": 0.0363, + "step": 3560 + }, + { + "grad_norm": 0.224070206284523, + "learning_rate": 9.555654533436557e-05, + "loss": 0.0358, + "step": 3570 + }, + { + "grad_norm": 0.2377307415008545, + "learning_rate": 9.552241191898621e-05, + "loss": 0.0381, + "step": 3580 + }, + { + "grad_norm": 0.24801868200302124, + "learning_rate": 9.548815404707092e-05, + "loss": 0.0378, + "step": 3590 + }, + { + "grad_norm": 0.3089374005794525, + "learning_rate": 9.545377181227942e-05, + "loss": 0.0389, + "step": 3600 + }, + { + "grad_norm": 0.3225456178188324, + "learning_rate": 9.541926530861145e-05, + "loss": 0.0353, + "step": 3610 + }, + { + "grad_norm": 0.286188006401062, + "learning_rate": 9.538463463040645e-05, + "loss": 0.0308, + "step": 3620 + }, + { + "grad_norm": 0.26595228910446167, + "learning_rate": 9.534987987234337e-05, + "loss": 0.0311, + "step": 3630 + }, + { + "grad_norm": 0.2935989499092102, + "learning_rate": 9.53150011294404e-05, + "loss": 0.0359, + "step": 3640 + }, + { + "grad_norm": 0.299089252948761, + "learning_rate": 9.527999849705471e-05, + "loss": 0.0389, + "step": 3650 + }, + { + "grad_norm": 0.3336467146873474, + "learning_rate": 9.524487207088213e-05, + "loss": 0.0373, + "step": 3660 + }, + { + "grad_norm": 0.254970908164978, + "learning_rate": 9.520962194695698e-05, + "loss": 0.0322, + "step": 3670 + }, + { + "grad_norm": 0.20083194971084595, + "learning_rate": 9.517424822165175e-05, + "loss": 0.0306, + "step": 3680 + }, + { + "grad_norm": 0.3437754809856415, + "learning_rate": 9.513875099167685e-05, + "loss": 0.035, + "step": 3690 + }, + { + "grad_norm": 0.1960253119468689, + "learning_rate": 9.510313035408035e-05, + "loss": 0.0344, + "step": 3700 + }, + { + "grad_norm": 0.26334255933761597, + "learning_rate": 9.506738640624775e-05, + "loss": 0.0322, + "step": 3710 + }, + { + "grad_norm": 0.34374117851257324, + "learning_rate": 9.50315192459016e-05, + "loss": 0.0359, + "step": 3720 + }, + { + "grad_norm": 0.2908255159854889, + "learning_rate": 9.499552897110136e-05, + "loss": 0.0353, + "step": 3730 + }, + { + "grad_norm": 0.2676212191581726, + "learning_rate": 9.495941568024304e-05, + "loss": 0.0339, + "step": 3740 + }, + { + "grad_norm": 0.26183634996414185, + "learning_rate": 9.492317947205904e-05, + "loss": 0.0357, + "step": 3750 + }, + { + "grad_norm": 0.21402321755886078, + "learning_rate": 9.488682044561775e-05, + "loss": 0.0324, + "step": 3760 + }, + { + "grad_norm": 0.2705502212047577, + "learning_rate": 9.485033870032335e-05, + "loss": 0.0353, + "step": 3770 + }, + { + "grad_norm": 0.21902285516262054, + "learning_rate": 9.481373433591556e-05, + "loss": 0.0347, + "step": 3780 + }, + { + "grad_norm": 0.29380369186401367, + "learning_rate": 9.47770074524693e-05, + "loss": 0.0346, + "step": 3790 + }, + { + "grad_norm": 0.27559059858322144, + "learning_rate": 9.474015815039446e-05, + "loss": 0.0325, + "step": 3800 + }, + { + "grad_norm": 0.3386359214782715, + "learning_rate": 9.470318653043565e-05, + "loss": 0.0298, + "step": 3810 + }, + { + "grad_norm": 0.29516127705574036, + "learning_rate": 9.466609269367185e-05, + "loss": 0.0307, + "step": 3820 + }, + { + "grad_norm": 0.17525973916053772, + "learning_rate": 9.46288767415162e-05, + "loss": 0.0338, + "step": 3830 + }, + { + "grad_norm": 0.21319764852523804, + "learning_rate": 9.459153877571567e-05, + "loss": 0.0335, + "step": 3840 + }, + { + "grad_norm": 0.3554450273513794, + "learning_rate": 9.455407889835087e-05, + "loss": 0.0347, + "step": 3850 + }, + { + "grad_norm": 0.2817513346672058, + "learning_rate": 9.451649721183564e-05, + "loss": 0.0306, + "step": 3860 + }, + { + "grad_norm": 0.36978018283843994, + "learning_rate": 9.447879381891692e-05, + "loss": 0.0349, + "step": 3870 + }, + { + "grad_norm": 0.20787593722343445, + "learning_rate": 9.444096882267428e-05, + "loss": 0.0336, + "step": 3880 + }, + { + "grad_norm": 0.2597019672393799, + "learning_rate": 9.440302232651988e-05, + "loss": 0.0323, + "step": 3890 + }, + { + "grad_norm": 0.23970812559127808, + "learning_rate": 9.436495443419795e-05, + "loss": 0.0344, + "step": 3900 + }, + { + "grad_norm": 0.2618948221206665, + "learning_rate": 9.432676524978466e-05, + "loss": 0.0342, + "step": 3910 + }, + { + "grad_norm": 0.27909228205680847, + "learning_rate": 9.42884548776878e-05, + "loss": 0.0379, + "step": 3920 + }, + { + "grad_norm": 0.4070633351802826, + "learning_rate": 9.425002342264646e-05, + "loss": 0.0326, + "step": 3930 + }, + { + "grad_norm": 0.23352625966072083, + "learning_rate": 9.421147098973077e-05, + "loss": 0.0334, + "step": 3940 + }, + { + "grad_norm": 0.2052670568227768, + "learning_rate": 9.41727976843416e-05, + "loss": 0.03, + "step": 3950 + }, + { + "grad_norm": 0.19285006821155548, + "learning_rate": 9.413400361221029e-05, + "loss": 0.0318, + "step": 3960 + }, + { + "grad_norm": 0.2674274444580078, + "learning_rate": 9.409508887939835e-05, + "loss": 0.0334, + "step": 3970 + }, + { + "grad_norm": 0.2135963886976242, + "learning_rate": 9.40560535922972e-05, + "loss": 0.0387, + "step": 3980 + }, + { + "grad_norm": 0.2925035059452057, + "learning_rate": 9.40168978576278e-05, + "loss": 0.0376, + "step": 3990 + }, + { + "grad_norm": 0.1648014485836029, + "learning_rate": 9.397762178244043e-05, + "loss": 0.0334, + "step": 4000 + }, + { + "grad_norm": 0.18689607083797455, + "learning_rate": 9.393822547411439e-05, + "loss": 0.0348, + "step": 4010 + }, + { + "grad_norm": 0.29493579268455505, + "learning_rate": 9.389870904035769e-05, + "loss": 0.0317, + "step": 4020 + }, + { + "grad_norm": 0.32554978132247925, + "learning_rate": 9.385907258920672e-05, + "loss": 0.0323, + "step": 4030 + }, + { + "grad_norm": 0.22445693612098694, + "learning_rate": 9.381931622902607e-05, + "loss": 0.0329, + "step": 4040 + }, + { + "grad_norm": 0.24071559309959412, + "learning_rate": 9.377944006850807e-05, + "loss": 0.03, + "step": 4050 + }, + { + "grad_norm": 0.2176637053489685, + "learning_rate": 9.373944421667265e-05, + "loss": 0.0327, + "step": 4060 + }, + { + "grad_norm": 0.35246598720550537, + "learning_rate": 9.369932878286691e-05, + "loss": 0.0322, + "step": 4070 + }, + { + "grad_norm": 0.24312129616737366, + "learning_rate": 9.365909387676494e-05, + "loss": 0.0341, + "step": 4080 + }, + { + "grad_norm": 0.21892638504505157, + "learning_rate": 9.361873960836744e-05, + "loss": 0.0329, + "step": 4090 + }, + { + "grad_norm": 0.23584529757499695, + "learning_rate": 9.357826608800142e-05, + "loss": 0.0318, + "step": 4100 + }, + { + "grad_norm": 0.29266729950904846, + "learning_rate": 9.353767342631994e-05, + "loss": 0.0348, + "step": 4110 + }, + { + "grad_norm": 0.32518109679222107, + "learning_rate": 9.34969617343018e-05, + "loss": 0.0352, + "step": 4120 + }, + { + "grad_norm": 0.27469193935394287, + "learning_rate": 9.345613112325122e-05, + "loss": 0.0309, + "step": 4130 + }, + { + "grad_norm": 0.3353689908981323, + "learning_rate": 9.34151817047975e-05, + "loss": 0.0358, + "step": 4140 + }, + { + "grad_norm": 0.30090591311454773, + "learning_rate": 9.33741135908948e-05, + "loss": 0.0334, + "step": 4150 + }, + { + "grad_norm": 0.31856924295425415, + "learning_rate": 9.33329268938218e-05, + "loss": 0.0347, + "step": 4160 + }, + { + "grad_norm": 0.35957279801368713, + "learning_rate": 9.329162172618132e-05, + "loss": 0.0369, + "step": 4170 + }, + { + "grad_norm": 0.30077677965164185, + "learning_rate": 9.325019820090013e-05, + "loss": 0.0362, + "step": 4180 + }, + { + "grad_norm": 0.2290671020746231, + "learning_rate": 9.320865643122855e-05, + "loss": 0.0304, + "step": 4190 + }, + { + "grad_norm": 0.18252457678318024, + "learning_rate": 9.316699653074023e-05, + "loss": 0.0319, + "step": 4200 + }, + { + "grad_norm": 0.19245073199272156, + "learning_rate": 9.312521861333172e-05, + "loss": 0.0283, + "step": 4210 + }, + { + "grad_norm": 0.25932082533836365, + "learning_rate": 9.308332279322224e-05, + "loss": 0.0298, + "step": 4220 + }, + { + "grad_norm": 0.2076045721769333, + "learning_rate": 9.304130918495338e-05, + "loss": 0.0291, + "step": 4230 + }, + { + "grad_norm": 0.39151453971862793, + "learning_rate": 9.299917790338874e-05, + "loss": 0.0331, + "step": 4240 + }, + { + "grad_norm": 0.3864510655403137, + "learning_rate": 9.295692906371363e-05, + "loss": 0.0325, + "step": 4250 + }, + { + "grad_norm": 0.27132487297058105, + "learning_rate": 9.291456278143476e-05, + "loss": 0.0347, + "step": 4260 + }, + { + "grad_norm": 0.19841168820858002, + "learning_rate": 9.287207917237994e-05, + "loss": 0.0287, + "step": 4270 + }, + { + "grad_norm": 0.22019748389720917, + "learning_rate": 9.282947835269773e-05, + "loss": 0.0309, + "step": 4280 + }, + { + "grad_norm": 0.15275084972381592, + "learning_rate": 9.278676043885715e-05, + "loss": 0.0328, + "step": 4290 + }, + { + "grad_norm": 0.2983507215976715, + "learning_rate": 9.274392554764733e-05, + "loss": 0.0343, + "step": 4300 + }, + { + "grad_norm": 0.23987573385238647, + "learning_rate": 9.270097379617723e-05, + "loss": 0.0328, + "step": 4310 + }, + { + "grad_norm": 0.27825701236724854, + "learning_rate": 9.26579053018753e-05, + "loss": 0.0307, + "step": 4320 + }, + { + "grad_norm": 0.21437597274780273, + "learning_rate": 9.261472018248918e-05, + "loss": 0.0322, + "step": 4330 + }, + { + "grad_norm": 0.26730191707611084, + "learning_rate": 9.25714185560853e-05, + "loss": 0.0327, + "step": 4340 + }, + { + "grad_norm": 0.2268911749124527, + "learning_rate": 9.252800054104868e-05, + "loss": 0.0329, + "step": 4350 + }, + { + "grad_norm": 0.16261471807956696, + "learning_rate": 9.248446625608252e-05, + "loss": 0.0319, + "step": 4360 + }, + { + "grad_norm": 0.3054570257663727, + "learning_rate": 9.244081582020789e-05, + "loss": 0.0347, + "step": 4370 + }, + { + "grad_norm": 0.25517579913139343, + "learning_rate": 9.239704935276339e-05, + "loss": 0.0325, + "step": 4380 + }, + { + "grad_norm": 0.3075014352798462, + "learning_rate": 9.235316697340489e-05, + "loss": 0.0303, + "step": 4390 + }, + { + "grad_norm": 0.2481297105550766, + "learning_rate": 9.230916880210512e-05, + "loss": 0.0342, + "step": 4400 + }, + { + "grad_norm": 0.32948264479637146, + "learning_rate": 9.226505495915342e-05, + "loss": 0.0332, + "step": 4410 + }, + { + "grad_norm": 0.25925540924072266, + "learning_rate": 9.222082556515536e-05, + "loss": 0.031, + "step": 4420 + }, + { + "grad_norm": 0.16929636895656586, + "learning_rate": 9.217648074103242e-05, + "loss": 0.033, + "step": 4430 + }, + { + "grad_norm": 0.21182797849178314, + "learning_rate": 9.213202060802161e-05, + "loss": 0.0286, + "step": 4440 + }, + { + "grad_norm": 0.197372704744339, + "learning_rate": 9.208744528767528e-05, + "loss": 0.0303, + "step": 4450 + }, + { + "grad_norm": 0.2988884449005127, + "learning_rate": 9.204275490186064e-05, + "loss": 0.0302, + "step": 4460 + }, + { + "grad_norm": 0.16220621764659882, + "learning_rate": 9.199794957275949e-05, + "loss": 0.0278, + "step": 4470 + }, + { + "grad_norm": 0.22180549800395966, + "learning_rate": 9.19530294228679e-05, + "loss": 0.0302, + "step": 4480 + }, + { + "grad_norm": 0.22499610483646393, + "learning_rate": 9.190799457499583e-05, + "loss": 0.0303, + "step": 4490 + }, + { + "grad_norm": 0.23255053162574768, + "learning_rate": 9.186284515226686e-05, + "loss": 0.0352, + "step": 4500 + }, + { + "grad_norm": 0.2958816587924957, + "learning_rate": 9.181758127811777e-05, + "loss": 0.0297, + "step": 4510 + }, + { + "grad_norm": 0.19010356068611145, + "learning_rate": 9.177220307629825e-05, + "loss": 0.0299, + "step": 4520 + }, + { + "grad_norm": 0.15465109050273895, + "learning_rate": 9.172671067087059e-05, + "loss": 0.0292, + "step": 4530 + }, + { + "grad_norm": 0.3397411108016968, + "learning_rate": 9.16811041862093e-05, + "loss": 0.0283, + "step": 4540 + }, + { + "grad_norm": 0.2775285542011261, + "learning_rate": 9.163538374700076e-05, + "loss": 0.0287, + "step": 4550 + }, + { + "grad_norm": 0.22243374586105347, + "learning_rate": 9.158954947824287e-05, + "loss": 0.032, + "step": 4560 + }, + { + "grad_norm": 0.28789588809013367, + "learning_rate": 9.154360150524482e-05, + "loss": 0.0303, + "step": 4570 + }, + { + "grad_norm": 0.28991883993148804, + "learning_rate": 9.14975399536266e-05, + "loss": 0.0329, + "step": 4580 + }, + { + "grad_norm": 0.22245454788208008, + "learning_rate": 9.14513649493187e-05, + "loss": 0.0333, + "step": 4590 + }, + { + "grad_norm": 0.22494885325431824, + "learning_rate": 9.140507661856187e-05, + "loss": 0.0309, + "step": 4600 + }, + { + "grad_norm": 0.24359627068042755, + "learning_rate": 9.135867508790661e-05, + "loss": 0.0329, + "step": 4610 + }, + { + "grad_norm": 0.285066694021225, + "learning_rate": 9.131216048421291e-05, + "loss": 0.0291, + "step": 4620 + }, + { + "grad_norm": 0.24179978668689728, + "learning_rate": 9.126553293464998e-05, + "loss": 0.0335, + "step": 4630 + }, + { + "grad_norm": 0.2951452136039734, + "learning_rate": 9.121879256669572e-05, + "loss": 0.0304, + "step": 4640 + }, + { + "grad_norm": 0.160237118601799, + "learning_rate": 9.117193950813652e-05, + "loss": 0.0308, + "step": 4650 + }, + { + "grad_norm": 0.1899791806936264, + "learning_rate": 9.112497388706685e-05, + "loss": 0.0307, + "step": 4660 + }, + { + "grad_norm": 0.24163338541984558, + "learning_rate": 9.10778958318889e-05, + "loss": 0.0305, + "step": 4670 + }, + { + "grad_norm": 0.3649671673774719, + "learning_rate": 9.103070547131232e-05, + "loss": 0.0316, + "step": 4680 + }, + { + "grad_norm": 0.3322865068912506, + "learning_rate": 9.098340293435375e-05, + "loss": 0.0338, + "step": 4690 + }, + { + "grad_norm": 0.27509939670562744, + "learning_rate": 9.093598835033649e-05, + "loss": 0.0321, + "step": 4700 + }, + { + "grad_norm": 0.2817743420600891, + "learning_rate": 9.088846184889021e-05, + "loss": 0.0324, + "step": 4710 + }, + { + "grad_norm": 0.22843751311302185, + "learning_rate": 9.084082355995057e-05, + "loss": 0.0343, + "step": 4720 + }, + { + "grad_norm": 0.3018876016139984, + "learning_rate": 9.079307361375882e-05, + "loss": 0.0309, + "step": 4730 + }, + { + "grad_norm": 0.2945747375488281, + "learning_rate": 9.074521214086149e-05, + "loss": 0.0315, + "step": 4740 + }, + { + "grad_norm": 0.2847113013267517, + "learning_rate": 9.069723927211001e-05, + "loss": 0.0327, + "step": 4750 + }, + { + "grad_norm": 0.23866991698741913, + "learning_rate": 9.064915513866037e-05, + "loss": 0.0302, + "step": 4760 + }, + { + "grad_norm": 0.24398328363895416, + "learning_rate": 9.060095987197279e-05, + "loss": 0.0314, + "step": 4770 + }, + { + "grad_norm": 0.26327499747276306, + "learning_rate": 9.055265360381126e-05, + "loss": 0.0307, + "step": 4780 + }, + { + "grad_norm": 0.2463245689868927, + "learning_rate": 9.050423646624326e-05, + "loss": 0.0314, + "step": 4790 + }, + { + "grad_norm": 0.24274016916751862, + "learning_rate": 9.045570859163943e-05, + "loss": 0.0273, + "step": 4800 + }, + { + "grad_norm": 0.2603799104690552, + "learning_rate": 9.04070701126731e-05, + "loss": 0.0323, + "step": 4810 + }, + { + "grad_norm": 0.2634144425392151, + "learning_rate": 9.035832116232001e-05, + "loss": 0.0319, + "step": 4820 + }, + { + "grad_norm": 0.28296008706092834, + "learning_rate": 9.030946187385796e-05, + "loss": 0.0306, + "step": 4830 + }, + { + "grad_norm": 0.21780265867710114, + "learning_rate": 9.026049238086635e-05, + "loss": 0.0299, + "step": 4840 + }, + { + "grad_norm": 0.293646901845932, + "learning_rate": 9.021141281722591e-05, + "loss": 0.0306, + "step": 4850 + }, + { + "grad_norm": 0.26014596223831177, + "learning_rate": 9.01622233171183e-05, + "loss": 0.0302, + "step": 4860 + }, + { + "grad_norm": 0.1846085637807846, + "learning_rate": 9.011292401502574e-05, + "loss": 0.0335, + "step": 4870 + }, + { + "grad_norm": 0.32345864176750183, + "learning_rate": 9.006351504573063e-05, + "loss": 0.0297, + "step": 4880 + }, + { + "grad_norm": 0.27427902817726135, + "learning_rate": 9.001399654431519e-05, + "loss": 0.0297, + "step": 4890 + }, + { + "grad_norm": 0.2501916289329529, + "learning_rate": 8.996436864616116e-05, + "loss": 0.0287, + "step": 4900 + }, + { + "grad_norm": 0.1802971214056015, + "learning_rate": 8.991463148694925e-05, + "loss": 0.0275, + "step": 4910 + }, + { + "grad_norm": 0.2993827760219574, + "learning_rate": 8.986478520265902e-05, + "loss": 0.0322, + "step": 4920 + }, + { + "grad_norm": 0.25177472829818726, + "learning_rate": 8.981482992956827e-05, + "loss": 0.0328, + "step": 4930 + }, + { + "grad_norm": 0.31737279891967773, + "learning_rate": 8.976476580425282e-05, + "loss": 0.0308, + "step": 4940 + }, + { + "grad_norm": 0.2443665713071823, + "learning_rate": 8.971459296358606e-05, + "loss": 0.0337, + "step": 4950 + }, + { + "grad_norm": 0.23701953887939453, + "learning_rate": 8.966431154473864e-05, + "loss": 0.0331, + "step": 4960 + }, + { + "grad_norm": 0.2281855344772339, + "learning_rate": 8.961392168517803e-05, + "loss": 0.0303, + "step": 4970 + }, + { + "grad_norm": 0.30207300186157227, + "learning_rate": 8.956342352266821e-05, + "loss": 0.0296, + "step": 4980 + }, + { + "grad_norm": 0.17806647717952728, + "learning_rate": 8.95128171952692e-05, + "loss": 0.0291, + "step": 4990 + }, + { + "grad_norm": 0.25516825914382935, + "learning_rate": 8.946210284133676e-05, + "loss": 0.0299, + "step": 5000 + }, + { + "grad_norm": 0.23243404924869537, + "learning_rate": 8.941128059952201e-05, + "loss": 0.0319, + "step": 5010 + }, + { + "grad_norm": 0.2368650734424591, + "learning_rate": 8.936035060877102e-05, + "loss": 0.0329, + "step": 5020 + }, + { + "grad_norm": 0.23695982992649078, + "learning_rate": 8.930931300832443e-05, + "loss": 0.0274, + "step": 5030 + }, + { + "grad_norm": 0.2600453794002533, + "learning_rate": 8.925816793771711e-05, + "loss": 0.0293, + "step": 5040 + }, + { + "grad_norm": 0.18863385915756226, + "learning_rate": 8.92069155367777e-05, + "loss": 0.0266, + "step": 5050 + }, + { + "grad_norm": 0.26943832635879517, + "learning_rate": 8.915555594562834e-05, + "loss": 0.0315, + "step": 5060 + }, + { + "grad_norm": 0.23639757931232452, + "learning_rate": 8.910408930468416e-05, + "loss": 0.028, + "step": 5070 + }, + { + "grad_norm": 0.181135356426239, + "learning_rate": 8.905251575465303e-05, + "loss": 0.0275, + "step": 5080 + }, + { + "grad_norm": 0.2265515774488449, + "learning_rate": 8.900083543653502e-05, + "loss": 0.0279, + "step": 5090 + }, + { + "grad_norm": 0.22376364469528198, + "learning_rate": 8.894904849162218e-05, + "loss": 0.0297, + "step": 5100 + }, + { + "grad_norm": 0.22845898568630219, + "learning_rate": 8.889715506149802e-05, + "loss": 0.0306, + "step": 5110 + }, + { + "grad_norm": 0.2636346220970154, + "learning_rate": 8.884515528803722e-05, + "loss": 0.0262, + "step": 5120 + }, + { + "grad_norm": 0.25118422508239746, + "learning_rate": 8.879304931340517e-05, + "loss": 0.0283, + "step": 5130 + }, + { + "grad_norm": 0.23598280549049377, + "learning_rate": 8.874083728005759e-05, + "loss": 0.0322, + "step": 5140 + }, + { + "grad_norm": 0.23262402415275574, + "learning_rate": 8.868851933074021e-05, + "loss": 0.0291, + "step": 5150 + }, + { + "grad_norm": 0.3212668299674988, + "learning_rate": 8.863609560848829e-05, + "loss": 0.0343, + "step": 5160 + }, + { + "grad_norm": 0.1705205887556076, + "learning_rate": 8.85835662566263e-05, + "loss": 0.0275, + "step": 5170 + }, + { + "grad_norm": 0.27302443981170654, + "learning_rate": 8.853093141876747e-05, + "loss": 0.0286, + "step": 5180 + }, + { + "grad_norm": 0.237864151597023, + "learning_rate": 8.847819123881343e-05, + "loss": 0.0307, + "step": 5190 + }, + { + "grad_norm": 0.2379344403743744, + "learning_rate": 8.842534586095383e-05, + "loss": 0.0305, + "step": 5200 + }, + { + "grad_norm": 0.20280928909778595, + "learning_rate": 8.837239542966593e-05, + "loss": 0.0285, + "step": 5210 + }, + { + "grad_norm": 0.1828172206878662, + "learning_rate": 8.831934008971417e-05, + "loss": 0.0277, + "step": 5220 + }, + { + "grad_norm": 0.2206842005252838, + "learning_rate": 8.826617998614982e-05, + "loss": 0.0299, + "step": 5230 + }, + { + "grad_norm": 0.16984917223453522, + "learning_rate": 8.821291526431056e-05, + "loss": 0.029, + "step": 5240 + }, + { + "grad_norm": 0.17552043497562408, + "learning_rate": 8.815954606982015e-05, + "loss": 0.0302, + "step": 5250 + }, + { + "grad_norm": 0.2597530782222748, + "learning_rate": 8.810607254858789e-05, + "loss": 0.033, + "step": 5260 + }, + { + "grad_norm": 0.25801724195480347, + "learning_rate": 8.805249484680838e-05, + "loss": 0.0271, + "step": 5270 + }, + { + "grad_norm": 0.24760890007019043, + "learning_rate": 8.799881311096096e-05, + "loss": 0.0282, + "step": 5280 + }, + { + "grad_norm": 0.2581833004951477, + "learning_rate": 8.794502748780949e-05, + "loss": 0.0281, + "step": 5290 + }, + { + "grad_norm": 0.24304334819316864, + "learning_rate": 8.78911381244018e-05, + "loss": 0.0284, + "step": 5300 + }, + { + "grad_norm": 0.16503192484378815, + "learning_rate": 8.783714516806933e-05, + "loss": 0.0261, + "step": 5310 + }, + { + "grad_norm": 0.19780272245407104, + "learning_rate": 8.77830487664268e-05, + "loss": 0.0273, + "step": 5320 + }, + { + "grad_norm": 0.19942815601825714, + "learning_rate": 8.772884906737167e-05, + "loss": 0.0281, + "step": 5330 + }, + { + "grad_norm": 0.24979932606220245, + "learning_rate": 8.767454621908387e-05, + "loss": 0.0276, + "step": 5340 + }, + { + "grad_norm": 0.16779963672161102, + "learning_rate": 8.76201403700253e-05, + "loss": 0.0279, + "step": 5350 + }, + { + "grad_norm": 0.18714742362499237, + "learning_rate": 8.756563166893949e-05, + "loss": 0.0284, + "step": 5360 + }, + { + "grad_norm": 0.2028001844882965, + "learning_rate": 8.751102026485113e-05, + "loss": 0.0263, + "step": 5370 + }, + { + "grad_norm": 0.2669692039489746, + "learning_rate": 8.745630630706571e-05, + "loss": 0.0276, + "step": 5380 + }, + { + "grad_norm": 0.19449053704738617, + "learning_rate": 8.740148994516912e-05, + "loss": 0.0263, + "step": 5390 + }, + { + "grad_norm": 0.3075883984565735, + "learning_rate": 8.73465713290272e-05, + "loss": 0.0276, + "step": 5400 + }, + { + "grad_norm": 0.19841042160987854, + "learning_rate": 8.729155060878533e-05, + "loss": 0.0297, + "step": 5410 + }, + { + "grad_norm": 0.2630940079689026, + "learning_rate": 8.723642793486809e-05, + "loss": 0.0292, + "step": 5420 + }, + { + "grad_norm": 0.24401918053627014, + "learning_rate": 8.718120345797873e-05, + "loss": 0.027, + "step": 5430 + }, + { + "grad_norm": 0.22380056977272034, + "learning_rate": 8.712587732909889e-05, + "loss": 0.026, + "step": 5440 + }, + { + "grad_norm": 0.18872298300266266, + "learning_rate": 8.707044969948806e-05, + "loss": 0.0276, + "step": 5450 + }, + { + "grad_norm": 0.2413512021303177, + "learning_rate": 8.701492072068329e-05, + "loss": 0.0252, + "step": 5460 + }, + { + "grad_norm": 0.26963767409324646, + "learning_rate": 8.695929054449869e-05, + "loss": 0.0288, + "step": 5470 + }, + { + "grad_norm": 0.21836189925670624, + "learning_rate": 8.690355932302501e-05, + "loss": 0.0267, + "step": 5480 + }, + { + "grad_norm": 0.2140367478132248, + "learning_rate": 8.684772720862931e-05, + "loss": 0.0278, + "step": 5490 + }, + { + "grad_norm": 0.2408180832862854, + "learning_rate": 8.679179435395446e-05, + "loss": 0.0286, + "step": 5500 + }, + { + "grad_norm": 0.23574891686439514, + "learning_rate": 8.673576091191874e-05, + "loss": 0.0263, + "step": 5510 + }, + { + "grad_norm": 0.17144161462783813, + "learning_rate": 8.667962703571541e-05, + "loss": 0.0279, + "step": 5520 + }, + { + "grad_norm": 0.22829054296016693, + "learning_rate": 8.662339287881238e-05, + "loss": 0.0251, + "step": 5530 + }, + { + "grad_norm": 0.21960297226905823, + "learning_rate": 8.656705859495169e-05, + "loss": 0.0327, + "step": 5540 + }, + { + "grad_norm": 0.2673276364803314, + "learning_rate": 8.651062433814912e-05, + "loss": 0.0282, + "step": 5550 + }, + { + "grad_norm": 0.25558650493621826, + "learning_rate": 8.645409026269375e-05, + "loss": 0.0292, + "step": 5560 + }, + { + "grad_norm": 0.26786279678344727, + "learning_rate": 8.639745652314759e-05, + "loss": 0.029, + "step": 5570 + }, + { + "grad_norm": 0.28067299723625183, + "learning_rate": 8.634072327434515e-05, + "loss": 0.0329, + "step": 5580 + }, + { + "grad_norm": 0.23317526280879974, + "learning_rate": 8.628389067139294e-05, + "loss": 0.026, + "step": 5590 + }, + { + "grad_norm": 0.2960446774959564, + "learning_rate": 8.622695886966911e-05, + "loss": 0.0346, + "step": 5600 + }, + { + "grad_norm": 0.24711379408836365, + "learning_rate": 8.616992802482308e-05, + "loss": 0.0301, + "step": 5610 + }, + { + "grad_norm": 0.24605536460876465, + "learning_rate": 8.611279829277496e-05, + "loss": 0.0279, + "step": 5620 + }, + { + "grad_norm": 0.15064962208271027, + "learning_rate": 8.605556982971528e-05, + "loss": 0.0267, + "step": 5630 + }, + { + "grad_norm": 0.22357802093029022, + "learning_rate": 8.599824279210447e-05, + "loss": 0.0258, + "step": 5640 + }, + { + "grad_norm": 0.2887904644012451, + "learning_rate": 8.594081733667243e-05, + "loss": 0.0307, + "step": 5650 + }, + { + "grad_norm": 0.26563066244125366, + "learning_rate": 8.58832936204182e-05, + "loss": 0.0282, + "step": 5660 + }, + { + "grad_norm": 0.2543182969093323, + "learning_rate": 8.582567180060942e-05, + "loss": 0.0311, + "step": 5670 + }, + { + "grad_norm": 0.22355759143829346, + "learning_rate": 8.576795203478194e-05, + "loss": 0.0277, + "step": 5680 + }, + { + "grad_norm": 0.24217355251312256, + "learning_rate": 8.571013448073939e-05, + "loss": 0.0286, + "step": 5690 + }, + { + "grad_norm": 0.258706271648407, + "learning_rate": 8.565221929655275e-05, + "loss": 0.027, + "step": 5700 + }, + { + "grad_norm": 0.1835828572511673, + "learning_rate": 8.559420664055992e-05, + "loss": 0.03, + "step": 5710 + }, + { + "grad_norm": 0.24048325419425964, + "learning_rate": 8.553609667136532e-05, + "loss": 0.0282, + "step": 5720 + }, + { + "grad_norm": 0.2555566132068634, + "learning_rate": 8.547788954783936e-05, + "loss": 0.0326, + "step": 5730 + }, + { + "grad_norm": 0.1845780909061432, + "learning_rate": 8.541958542911808e-05, + "loss": 0.0284, + "step": 5740 + }, + { + "grad_norm": 0.24985064566135406, + "learning_rate": 8.536118447460275e-05, + "loss": 0.027, + "step": 5750 + }, + { + "grad_norm": 0.24968400597572327, + "learning_rate": 8.530268684395932e-05, + "loss": 0.0283, + "step": 5760 + }, + { + "grad_norm": 0.21790499985218048, + "learning_rate": 8.524409269711807e-05, + "loss": 0.0278, + "step": 5770 + }, + { + "grad_norm": 0.19446101784706116, + "learning_rate": 8.51854021942732e-05, + "loss": 0.0273, + "step": 5780 + }, + { + "grad_norm": 0.23789428174495697, + "learning_rate": 8.512661549588227e-05, + "loss": 0.0246, + "step": 5790 + }, + { + "grad_norm": 0.16820183396339417, + "learning_rate": 8.506773276266588e-05, + "loss": 0.026, + "step": 5800 + }, + { + "grad_norm": 0.18751686811447144, + "learning_rate": 8.500875415560721e-05, + "loss": 0.0267, + "step": 5810 + }, + { + "grad_norm": 0.1670578569173813, + "learning_rate": 8.494967983595144e-05, + "loss": 0.0264, + "step": 5820 + }, + { + "grad_norm": 0.18704158067703247, + "learning_rate": 8.489050996520558e-05, + "loss": 0.0255, + "step": 5830 + }, + { + "grad_norm": 0.3310665488243103, + "learning_rate": 8.483124470513775e-05, + "loss": 0.026, + "step": 5840 + }, + { + "grad_norm": 0.22714407742023468, + "learning_rate": 8.477188421777692e-05, + "loss": 0.0279, + "step": 5850 + }, + { + "grad_norm": 0.1409146636724472, + "learning_rate": 8.47124286654124e-05, + "loss": 0.0259, + "step": 5860 + }, + { + "grad_norm": 0.19318713247776031, + "learning_rate": 8.465287821059341e-05, + "loss": 0.0263, + "step": 5870 + }, + { + "grad_norm": 0.19850215315818787, + "learning_rate": 8.45932330161286e-05, + "loss": 0.0268, + "step": 5880 + }, + { + "grad_norm": 0.18577902019023895, + "learning_rate": 8.453349324508567e-05, + "loss": 0.024, + "step": 5890 + }, + { + "grad_norm": 0.21152952313423157, + "learning_rate": 8.447365906079088e-05, + "loss": 0.0249, + "step": 5900 + }, + { + "grad_norm": 0.16427931189537048, + "learning_rate": 8.441373062682856e-05, + "loss": 0.0274, + "step": 5910 + }, + { + "grad_norm": 0.20545101165771484, + "learning_rate": 8.43537081070408e-05, + "loss": 0.0295, + "step": 5920 + }, + { + "grad_norm": 0.2445710301399231, + "learning_rate": 8.429359166552689e-05, + "loss": 0.0307, + "step": 5930 + }, + { + "grad_norm": 0.18652495741844177, + "learning_rate": 8.423338146664284e-05, + "loss": 0.0265, + "step": 5940 + }, + { + "grad_norm": 0.22506183385849, + "learning_rate": 8.417307767500107e-05, + "loss": 0.0279, + "step": 5950 + }, + { + "grad_norm": 0.30442968010902405, + "learning_rate": 8.411268045546983e-05, + "loss": 0.032, + "step": 5960 + }, + { + "grad_norm": 0.1586800068616867, + "learning_rate": 8.405218997317281e-05, + "loss": 0.0314, + "step": 5970 + }, + { + "grad_norm": 0.27731549739837646, + "learning_rate": 8.399160639348869e-05, + "loss": 0.0285, + "step": 5980 + }, + { + "grad_norm": 0.2733120620250702, + "learning_rate": 8.393092988205065e-05, + "loss": 0.0298, + "step": 5990 + }, + { + "grad_norm": 0.19086717069149017, + "learning_rate": 8.387016060474597e-05, + "loss": 0.0237, + "step": 6000 + }, + { + "grad_norm": 0.25084781646728516, + "learning_rate": 8.380929872771551e-05, + "loss": 0.0292, + "step": 6010 + }, + { + "grad_norm": 0.16830351948738098, + "learning_rate": 8.374834441735335e-05, + "loss": 0.0278, + "step": 6020 + }, + { + "grad_norm": 0.2337435632944107, + "learning_rate": 8.368729784030622e-05, + "loss": 0.0301, + "step": 6030 + }, + { + "grad_norm": 0.19408833980560303, + "learning_rate": 8.362615916347315e-05, + "loss": 0.0277, + "step": 6040 + }, + { + "grad_norm": 0.2253166288137436, + "learning_rate": 8.356492855400493e-05, + "loss": 0.0254, + "step": 6050 + }, + { + "grad_norm": 0.2161058932542801, + "learning_rate": 8.350360617930371e-05, + "loss": 0.0258, + "step": 6060 + }, + { + "grad_norm": 0.26727095246315, + "learning_rate": 8.344219220702255e-05, + "loss": 0.0282, + "step": 6070 + }, + { + "grad_norm": 0.17742595076560974, + "learning_rate": 8.338068680506485e-05, + "loss": 0.0306, + "step": 6080 + }, + { + "grad_norm": 0.2415505349636078, + "learning_rate": 8.33190901415841e-05, + "loss": 0.0287, + "step": 6090 + }, + { + "grad_norm": 0.21782615780830383, + "learning_rate": 8.325740238498317e-05, + "loss": 0.0262, + "step": 6100 + }, + { + "grad_norm": 0.27476173639297485, + "learning_rate": 8.319562370391406e-05, + "loss": 0.0271, + "step": 6110 + }, + { + "grad_norm": 0.2107502818107605, + "learning_rate": 8.31337542672773e-05, + "loss": 0.0299, + "step": 6120 + }, + { + "grad_norm": 0.17578455805778503, + "learning_rate": 8.307179424422158e-05, + "loss": 0.0274, + "step": 6130 + }, + { + "grad_norm": 0.2937401235103607, + "learning_rate": 8.300974380414327e-05, + "loss": 0.0284, + "step": 6140 + }, + { + "grad_norm": 0.2051600217819214, + "learning_rate": 8.294760311668586e-05, + "loss": 0.0269, + "step": 6150 + }, + { + "grad_norm": 0.15926064550876617, + "learning_rate": 8.288537235173961e-05, + "loss": 0.025, + "step": 6160 + }, + { + "grad_norm": 0.18960604071617126, + "learning_rate": 8.282305167944108e-05, + "loss": 0.0238, + "step": 6170 + }, + { + "grad_norm": 0.18251734972000122, + "learning_rate": 8.276064127017262e-05, + "loss": 0.0249, + "step": 6180 + }, + { + "grad_norm": 0.19407083094120026, + "learning_rate": 8.269814129456189e-05, + "loss": 0.0236, + "step": 6190 + }, + { + "grad_norm": 0.3516760766506195, + "learning_rate": 8.263555192348143e-05, + "loss": 0.0277, + "step": 6200 + }, + { + "grad_norm": 0.2286289930343628, + "learning_rate": 8.257287332804819e-05, + "loss": 0.0261, + "step": 6210 + }, + { + "grad_norm": 0.2606516182422638, + "learning_rate": 8.251010567962307e-05, + "loss": 0.0267, + "step": 6220 + }, + { + "grad_norm": 0.17926591634750366, + "learning_rate": 8.244724914981041e-05, + "loss": 0.0274, + "step": 6230 + }, + { + "grad_norm": 0.25607237219810486, + "learning_rate": 8.238430391045757e-05, + "loss": 0.0287, + "step": 6240 + }, + { + "grad_norm": 0.23498065769672394, + "learning_rate": 8.232127013365445e-05, + "loss": 0.0252, + "step": 6250 + }, + { + "grad_norm": 0.23208433389663696, + "learning_rate": 8.225814799173295e-05, + "loss": 0.0245, + "step": 6260 + }, + { + "grad_norm": 0.286391943693161, + "learning_rate": 8.219493765726663e-05, + "loss": 0.0242, + "step": 6270 + }, + { + "grad_norm": 0.15979978442192078, + "learning_rate": 8.21316393030701e-05, + "loss": 0.0251, + "step": 6280 + }, + { + "grad_norm": 0.20175860822200775, + "learning_rate": 8.206825310219865e-05, + "loss": 0.0291, + "step": 6290 + }, + { + "grad_norm": 0.29250213503837585, + "learning_rate": 8.200477922794776e-05, + "loss": 0.0261, + "step": 6300 + }, + { + "grad_norm": 0.14572352170944214, + "learning_rate": 8.194121785385256e-05, + "loss": 0.0256, + "step": 6310 + }, + { + "grad_norm": 0.23009555041790009, + "learning_rate": 8.187756915368741e-05, + "loss": 0.0265, + "step": 6320 + }, + { + "grad_norm": 0.22957858443260193, + "learning_rate": 8.181383330146544e-05, + "loss": 0.0271, + "step": 6330 + }, + { + "grad_norm": 0.1760542243719101, + "learning_rate": 8.175001047143804e-05, + "loss": 0.0232, + "step": 6340 + }, + { + "grad_norm": 0.1835332214832306, + "learning_rate": 8.168610083809438e-05, + "loss": 0.0279, + "step": 6350 + }, + { + "grad_norm": 0.18589144945144653, + "learning_rate": 8.162210457616095e-05, + "loss": 0.0252, + "step": 6360 + }, + { + "grad_norm": 0.16829219460487366, + "learning_rate": 8.155802186060109e-05, + "loss": 0.0239, + "step": 6370 + }, + { + "grad_norm": 0.21911458671092987, + "learning_rate": 8.149385286661453e-05, + "loss": 0.0252, + "step": 6380 + }, + { + "grad_norm": 0.20423123240470886, + "learning_rate": 8.14295977696368e-05, + "loss": 0.0242, + "step": 6390 + }, + { + "grad_norm": 0.2617279291152954, + "learning_rate": 8.13652567453389e-05, + "loss": 0.0307, + "step": 6400 + }, + { + "grad_norm": 0.235616996884346, + "learning_rate": 8.130082996962676e-05, + "loss": 0.027, + "step": 6410 + }, + { + "grad_norm": 0.2028554379940033, + "learning_rate": 8.123631761864068e-05, + "loss": 0.0281, + "step": 6420 + }, + { + "grad_norm": 0.23416590690612793, + "learning_rate": 8.1171719868755e-05, + "loss": 0.0286, + "step": 6430 + }, + { + "grad_norm": 0.18758827447891235, + "learning_rate": 8.110703689657748e-05, + "loss": 0.0261, + "step": 6440 + }, + { + "grad_norm": 0.17477427423000336, + "learning_rate": 8.104226887894892e-05, + "loss": 0.0239, + "step": 6450 + }, + { + "grad_norm": 0.1996232122182846, + "learning_rate": 8.097741599294257e-05, + "loss": 0.0265, + "step": 6460 + }, + { + "grad_norm": 0.26463672518730164, + "learning_rate": 8.091247841586378e-05, + "loss": 0.0239, + "step": 6470 + }, + { + "grad_norm": 0.19640767574310303, + "learning_rate": 8.084745632524939e-05, + "loss": 0.0269, + "step": 6480 + }, + { + "grad_norm": 0.17365293204784393, + "learning_rate": 8.07823498988673e-05, + "loss": 0.0244, + "step": 6490 + }, + { + "grad_norm": 0.20144912600517273, + "learning_rate": 8.071715931471602e-05, + "loss": 0.0285, + "step": 6500 + }, + { + "grad_norm": 0.2680608034133911, + "learning_rate": 8.06518847510241e-05, + "loss": 0.0279, + "step": 6510 + }, + { + "grad_norm": 0.2847196161746979, + "learning_rate": 8.058652638624971e-05, + "loss": 0.0247, + "step": 6520 + }, + { + "grad_norm": 0.2829527258872986, + "learning_rate": 8.052108439908013e-05, + "loss": 0.0286, + "step": 6530 + }, + { + "grad_norm": 0.24273407459259033, + "learning_rate": 8.045555896843125e-05, + "loss": 0.0222, + "step": 6540 + }, + { + "grad_norm": 0.2342250645160675, + "learning_rate": 8.03899502734471e-05, + "loss": 0.0271, + "step": 6550 + }, + { + "grad_norm": 0.17223294079303741, + "learning_rate": 8.032425849349931e-05, + "loss": 0.0263, + "step": 6560 + }, + { + "grad_norm": 0.1771479845046997, + "learning_rate": 8.025848380818674e-05, + "loss": 0.0253, + "step": 6570 + }, + { + "grad_norm": 0.15529488027095795, + "learning_rate": 8.019262639733487e-05, + "loss": 0.0212, + "step": 6580 + }, + { + "grad_norm": 0.1952378898859024, + "learning_rate": 8.012668644099531e-05, + "loss": 0.025, + "step": 6590 + }, + { + "grad_norm": 0.23330336809158325, + "learning_rate": 8.006066411944542e-05, + "loss": 0.0275, + "step": 6600 + }, + { + "grad_norm": 0.21989727020263672, + "learning_rate": 7.999455961318769e-05, + "loss": 0.0259, + "step": 6610 + }, + { + "grad_norm": 0.18205955624580383, + "learning_rate": 7.992837310294932e-05, + "loss": 0.0265, + "step": 6620 + }, + { + "grad_norm": 0.2020057886838913, + "learning_rate": 7.986210476968167e-05, + "loss": 0.0271, + "step": 6630 + }, + { + "grad_norm": 0.18366611003875732, + "learning_rate": 7.97957547945599e-05, + "loss": 0.0279, + "step": 6640 + }, + { + "grad_norm": 0.28971606492996216, + "learning_rate": 7.972932335898226e-05, + "loss": 0.0283, + "step": 6650 + }, + { + "grad_norm": 0.31382375955581665, + "learning_rate": 7.966281064456975e-05, + "loss": 0.0257, + "step": 6660 + }, + { + "grad_norm": 0.26659446954727173, + "learning_rate": 7.959621683316563e-05, + "loss": 0.0292, + "step": 6670 + }, + { + "grad_norm": 0.1979912519454956, + "learning_rate": 7.952954210683481e-05, + "loss": 0.023, + "step": 6680 + }, + { + "grad_norm": 0.2972131073474884, + "learning_rate": 7.946278664786345e-05, + "loss": 0.0249, + "step": 6690 + }, + { + "grad_norm": 0.23077769577503204, + "learning_rate": 7.939595063875842e-05, + "loss": 0.0277, + "step": 6700 + }, + { + "grad_norm": 0.23427477478981018, + "learning_rate": 7.932903426224683e-05, + "loss": 0.0269, + "step": 6710 + }, + { + "grad_norm": 0.2006346583366394, + "learning_rate": 7.926203770127552e-05, + "loss": 0.0281, + "step": 6720 + }, + { + "grad_norm": 0.15825718641281128, + "learning_rate": 7.919496113901046e-05, + "loss": 0.0273, + "step": 6730 + }, + { + "grad_norm": 0.20323115587234497, + "learning_rate": 7.912780475883649e-05, + "loss": 0.0267, + "step": 6740 + }, + { + "grad_norm": 0.2942650616168976, + "learning_rate": 7.906056874435652e-05, + "loss": 0.0296, + "step": 6750 + }, + { + "grad_norm": 0.229061096906662, + "learning_rate": 7.899325327939131e-05, + "loss": 0.0259, + "step": 6760 + }, + { + "grad_norm": 0.19965054094791412, + "learning_rate": 7.892585854797872e-05, + "loss": 0.0282, + "step": 6770 + }, + { + "grad_norm": 0.1717044860124588, + "learning_rate": 7.88583847343734e-05, + "loss": 0.0269, + "step": 6780 + }, + { + "grad_norm": 0.18811571598052979, + "learning_rate": 7.879083202304616e-05, + "loss": 0.025, + "step": 6790 + }, + { + "grad_norm": 0.2910924255847931, + "learning_rate": 7.872320059868355e-05, + "loss": 0.028, + "step": 6800 + }, + { + "grad_norm": 0.21640610694885254, + "learning_rate": 7.865549064618729e-05, + "loss": 0.0262, + "step": 6810 + }, + { + "grad_norm": 0.20532046258449554, + "learning_rate": 7.858770235067381e-05, + "loss": 0.0252, + "step": 6820 + }, + { + "grad_norm": 0.2995338439941406, + "learning_rate": 7.851983589747374e-05, + "loss": 0.0231, + "step": 6830 + }, + { + "grad_norm": 0.3033715486526489, + "learning_rate": 7.845189147213133e-05, + "loss": 0.026, + "step": 6840 + }, + { + "grad_norm": 0.18233098089694977, + "learning_rate": 7.838386926040407e-05, + "loss": 0.0251, + "step": 6850 + }, + { + "grad_norm": 0.16308864951133728, + "learning_rate": 7.83157694482621e-05, + "loss": 0.025, + "step": 6860 + }, + { + "grad_norm": 0.18623046576976776, + "learning_rate": 7.824759222188768e-05, + "loss": 0.0268, + "step": 6870 + }, + { + "grad_norm": 0.2004988193511963, + "learning_rate": 7.817933776767478e-05, + "loss": 0.025, + "step": 6880 + }, + { + "grad_norm": 0.17624622583389282, + "learning_rate": 7.811100627222842e-05, + "loss": 0.0238, + "step": 6890 + }, + { + "grad_norm": 0.2512792646884918, + "learning_rate": 7.804259792236435e-05, + "loss": 0.0244, + "step": 6900 + }, + { + "grad_norm": 0.1901821792125702, + "learning_rate": 7.797411290510835e-05, + "loss": 0.0268, + "step": 6910 + }, + { + "grad_norm": 0.1449953019618988, + "learning_rate": 7.790555140769586e-05, + "loss": 0.0246, + "step": 6920 + }, + { + "grad_norm": 0.21970738470554352, + "learning_rate": 7.78369136175714e-05, + "loss": 0.0262, + "step": 6930 + }, + { + "grad_norm": 0.17989793419837952, + "learning_rate": 7.776819972238806e-05, + "loss": 0.0227, + "step": 6940 + }, + { + "grad_norm": 0.20827268064022064, + "learning_rate": 7.7699409910007e-05, + "loss": 0.026, + "step": 6950 + }, + { + "grad_norm": 0.3076477646827698, + "learning_rate": 7.763054436849694e-05, + "loss": 0.0276, + "step": 6960 + }, + { + "grad_norm": 0.18687370419502258, + "learning_rate": 7.756160328613364e-05, + "loss": 0.0258, + "step": 6970 + }, + { + "grad_norm": 0.2509337067604065, + "learning_rate": 7.749258685139942e-05, + "loss": 0.0272, + "step": 6980 + }, + { + "grad_norm": 0.16400770843029022, + "learning_rate": 7.742349525298253e-05, + "loss": 0.0244, + "step": 6990 + }, + { + "grad_norm": 0.15971729159355164, + "learning_rate": 7.735432867977679e-05, + "loss": 0.025, + "step": 7000 + }, + { + "grad_norm": 0.1516847014427185, + "learning_rate": 7.728508732088096e-05, + "loss": 0.0257, + "step": 7010 + }, + { + "grad_norm": 0.20461945235729218, + "learning_rate": 7.721577136559825e-05, + "loss": 0.023, + "step": 7020 + }, + { + "grad_norm": 0.20047405362129211, + "learning_rate": 7.714638100343588e-05, + "loss": 0.0232, + "step": 7030 + }, + { + "grad_norm": 0.15968237817287445, + "learning_rate": 7.707691642410444e-05, + "loss": 0.0235, + "step": 7040 + }, + { + "grad_norm": 0.20452530682086945, + "learning_rate": 7.70073778175174e-05, + "loss": 0.0252, + "step": 7050 + }, + { + "grad_norm": 0.1929885298013687, + "learning_rate": 7.69377653737907e-05, + "loss": 0.0223, + "step": 7060 + }, + { + "grad_norm": 0.17922548949718475, + "learning_rate": 7.686807928324209e-05, + "loss": 0.0259, + "step": 7070 + }, + { + "grad_norm": 0.24653008580207825, + "learning_rate": 7.679831973639065e-05, + "loss": 0.0246, + "step": 7080 + }, + { + "grad_norm": 0.1668480634689331, + "learning_rate": 7.672848692395637e-05, + "loss": 0.0239, + "step": 7090 + }, + { + "grad_norm": 0.23397453129291534, + "learning_rate": 7.665858103685944e-05, + "loss": 0.0236, + "step": 7100 + }, + { + "grad_norm": 0.23826593160629272, + "learning_rate": 7.658860226621991e-05, + "loss": 0.0252, + "step": 7110 + }, + { + "grad_norm": 0.19321885704994202, + "learning_rate": 7.651855080335708e-05, + "loss": 0.0254, + "step": 7120 + }, + { + "grad_norm": 0.19148577749729156, + "learning_rate": 7.644842683978896e-05, + "loss": 0.0264, + "step": 7130 + }, + { + "grad_norm": 0.21372193098068237, + "learning_rate": 7.63782305672318e-05, + "loss": 0.0233, + "step": 7140 + }, + { + "grad_norm": 0.23095247149467468, + "learning_rate": 7.63079621775995e-05, + "loss": 0.0209, + "step": 7150 + }, + { + "grad_norm": 0.24704545736312866, + "learning_rate": 7.623762186300319e-05, + "loss": 0.0259, + "step": 7160 + }, + { + "grad_norm": 0.21102583408355713, + "learning_rate": 7.616720981575057e-05, + "loss": 0.0249, + "step": 7170 + }, + { + "grad_norm": 0.21966096758842468, + "learning_rate": 7.609672622834552e-05, + "loss": 0.0243, + "step": 7180 + }, + { + "grad_norm": 0.14871662855148315, + "learning_rate": 7.602617129348747e-05, + "loss": 0.0249, + "step": 7190 + }, + { + "grad_norm": 0.19188743829727173, + "learning_rate": 7.595554520407088e-05, + "loss": 0.0261, + "step": 7200 + }, + { + "grad_norm": 0.2074108123779297, + "learning_rate": 7.588484815318484e-05, + "loss": 0.0242, + "step": 7210 + }, + { + "grad_norm": 0.20805217325687408, + "learning_rate": 7.581408033411234e-05, + "loss": 0.0226, + "step": 7220 + }, + { + "grad_norm": 0.20965856313705444, + "learning_rate": 7.574324194032995e-05, + "loss": 0.0245, + "step": 7230 + }, + { + "grad_norm": 0.19314010441303253, + "learning_rate": 7.567233316550705e-05, + "loss": 0.027, + "step": 7240 + }, + { + "grad_norm": 0.1669759899377823, + "learning_rate": 7.560135420350562e-05, + "loss": 0.022, + "step": 7250 + }, + { + "grad_norm": 0.21310827136039734, + "learning_rate": 7.553030524837935e-05, + "loss": 0.0253, + "step": 7260 + }, + { + "grad_norm": 0.2044450342655182, + "learning_rate": 7.545918649437341e-05, + "loss": 0.0263, + "step": 7270 + }, + { + "grad_norm": 0.1826537698507309, + "learning_rate": 7.538799813592377e-05, + "loss": 0.0235, + "step": 7280 + }, + { + "grad_norm": 0.1640489250421524, + "learning_rate": 7.531674036765662e-05, + "loss": 0.0237, + "step": 7290 + }, + { + "grad_norm": 0.2104564756155014, + "learning_rate": 7.524541338438807e-05, + "loss": 0.021, + "step": 7300 + }, + { + "grad_norm": 0.22137579321861267, + "learning_rate": 7.517401738112328e-05, + "loss": 0.021, + "step": 7310 + }, + { + "grad_norm": 0.26817333698272705, + "learning_rate": 7.510255255305628e-05, + "loss": 0.0276, + "step": 7320 + }, + { + "grad_norm": 0.20775100588798523, + "learning_rate": 7.503101909556911e-05, + "loss": 0.0216, + "step": 7330 + }, + { + "grad_norm": 0.24248790740966797, + "learning_rate": 7.495941720423154e-05, + "loss": 0.0234, + "step": 7340 + }, + { + "grad_norm": 0.21298962831497192, + "learning_rate": 7.488774707480042e-05, + "loss": 0.0216, + "step": 7350 + }, + { + "grad_norm": 0.2223520576953888, + "learning_rate": 7.481600890321911e-05, + "loss": 0.0303, + "step": 7360 + }, + { + "grad_norm": 0.1866648942232132, + "learning_rate": 7.474420288561708e-05, + "loss": 0.0282, + "step": 7370 + }, + { + "grad_norm": 0.20475749671459198, + "learning_rate": 7.467232921830921e-05, + "loss": 0.0215, + "step": 7380 + }, + { + "grad_norm": 0.17132557928562164, + "learning_rate": 7.460038809779537e-05, + "loss": 0.0254, + "step": 7390 + }, + { + "grad_norm": 0.16491803526878357, + "learning_rate": 7.452837972075983e-05, + "loss": 0.0237, + "step": 7400 + }, + { + "grad_norm": 0.1937810331583023, + "learning_rate": 7.445630428407074e-05, + "loss": 0.0253, + "step": 7410 + }, + { + "grad_norm": 0.16583487391471863, + "learning_rate": 7.43841619847796e-05, + "loss": 0.0225, + "step": 7420 + }, + { + "grad_norm": 0.18042494356632233, + "learning_rate": 7.431195302012072e-05, + "loss": 0.0242, + "step": 7430 + }, + { + "grad_norm": 0.21080712974071503, + "learning_rate": 7.423967758751061e-05, + "loss": 0.0233, + "step": 7440 + }, + { + "grad_norm": 0.1872524917125702, + "learning_rate": 7.416733588454758e-05, + "loss": 0.0227, + "step": 7450 + }, + { + "grad_norm": 0.2792978286743164, + "learning_rate": 7.409492810901106e-05, + "loss": 0.0241, + "step": 7460 + }, + { + "grad_norm": 0.13574199378490448, + "learning_rate": 7.402245445886116e-05, + "loss": 0.0214, + "step": 7470 + }, + { + "grad_norm": 0.19132237136363983, + "learning_rate": 7.394991513223806e-05, + "loss": 0.023, + "step": 7480 + }, + { + "grad_norm": 0.16104218363761902, + "learning_rate": 7.38773103274615e-05, + "loss": 0.0223, + "step": 7490 + }, + { + "grad_norm": 0.12513796985149384, + "learning_rate": 7.380464024303028e-05, + "loss": 0.025, + "step": 7500 + }, + { + "grad_norm": 0.16853372752666473, + "learning_rate": 7.373190507762162e-05, + "loss": 0.0195, + "step": 7510 + }, + { + "grad_norm": 0.201432004570961, + "learning_rate": 7.365910503009066e-05, + "loss": 0.0247, + "step": 7520 + }, + { + "grad_norm": 0.1393406093120575, + "learning_rate": 7.358624029946996e-05, + "loss": 0.0218, + "step": 7530 + }, + { + "grad_norm": 0.2042279690504074, + "learning_rate": 7.351331108496893e-05, + "loss": 0.022, + "step": 7540 + }, + { + "grad_norm": 0.20944632589817047, + "learning_rate": 7.344031758597325e-05, + "loss": 0.0226, + "step": 7550 + }, + { + "grad_norm": 0.1793331354856491, + "learning_rate": 7.336726000204435e-05, + "loss": 0.0251, + "step": 7560 + }, + { + "grad_norm": 0.24198517203330994, + "learning_rate": 7.32941385329189e-05, + "loss": 0.0215, + "step": 7570 + }, + { + "grad_norm": 0.20446063578128815, + "learning_rate": 7.322095337850816e-05, + "loss": 0.0208, + "step": 7580 + }, + { + "grad_norm": 0.15520139038562775, + "learning_rate": 7.314770473889758e-05, + "loss": 0.0213, + "step": 7590 + }, + { + "grad_norm": 0.19187678396701813, + "learning_rate": 7.307439281434615e-05, + "loss": 0.0215, + "step": 7600 + }, + { + "grad_norm": 0.20064619183540344, + "learning_rate": 7.300101780528585e-05, + "loss": 0.0215, + "step": 7610 + }, + { + "grad_norm": 0.26868581771850586, + "learning_rate": 7.292757991232117e-05, + "loss": 0.0229, + "step": 7620 + }, + { + "grad_norm": 0.2599526643753052, + "learning_rate": 7.285407933622848e-05, + "loss": 0.0241, + "step": 7630 + }, + { + "grad_norm": 0.2054545283317566, + "learning_rate": 7.278051627795557e-05, + "loss": 0.0222, + "step": 7640 + }, + { + "grad_norm": 0.16483281552791595, + "learning_rate": 7.270689093862105e-05, + "loss": 0.0236, + "step": 7650 + }, + { + "grad_norm": 0.22294221818447113, + "learning_rate": 7.263320351951374e-05, + "loss": 0.0283, + "step": 7660 + }, + { + "grad_norm": 0.21725963056087494, + "learning_rate": 7.255945422209227e-05, + "loss": 0.0234, + "step": 7670 + }, + { + "grad_norm": 0.15655584633350372, + "learning_rate": 7.248564324798437e-05, + "loss": 0.0214, + "step": 7680 + }, + { + "grad_norm": 0.1951420158147812, + "learning_rate": 7.241177079898644e-05, + "loss": 0.0204, + "step": 7690 + }, + { + "grad_norm": 0.1373668611049652, + "learning_rate": 7.233783707706295e-05, + "loss": 0.0238, + "step": 7700 + }, + { + "grad_norm": 0.20325444638729095, + "learning_rate": 7.226384228434586e-05, + "loss": 0.0208, + "step": 7710 + }, + { + "grad_norm": 0.18210965394973755, + "learning_rate": 7.21897866231341e-05, + "loss": 0.0198, + "step": 7720 + }, + { + "grad_norm": 0.20808078348636627, + "learning_rate": 7.211567029589303e-05, + "loss": 0.023, + "step": 7730 + }, + { + "grad_norm": 0.20421084761619568, + "learning_rate": 7.204149350525387e-05, + "loss": 0.0198, + "step": 7740 + }, + { + "grad_norm": 0.19235551357269287, + "learning_rate": 7.196725645401309e-05, + "loss": 0.0228, + "step": 7750 + }, + { + "grad_norm": 0.17785964906215668, + "learning_rate": 7.1892959345132e-05, + "loss": 0.0209, + "step": 7760 + }, + { + "grad_norm": 0.1593187004327774, + "learning_rate": 7.181860238173605e-05, + "loss": 0.0249, + "step": 7770 + }, + { + "grad_norm": 0.13027040660381317, + "learning_rate": 7.174418576711432e-05, + "loss": 0.0218, + "step": 7780 + }, + { + "grad_norm": 0.19390296936035156, + "learning_rate": 7.1669709704719e-05, + "loss": 0.0221, + "step": 7790 + }, + { + "grad_norm": 0.22560721635818481, + "learning_rate": 7.159517439816481e-05, + "loss": 0.0208, + "step": 7800 + }, + { + "grad_norm": 0.3117069900035858, + "learning_rate": 7.152058005122842e-05, + "loss": 0.022, + "step": 7810 + }, + { + "grad_norm": 0.20474955439567566, + "learning_rate": 7.144592686784793e-05, + "loss": 0.0185, + "step": 7820 + }, + { + "grad_norm": 0.21816793084144592, + "learning_rate": 7.137121505212229e-05, + "loss": 0.0237, + "step": 7830 + }, + { + "grad_norm": 0.20002390444278717, + "learning_rate": 7.129644480831077e-05, + "loss": 0.0214, + "step": 7840 + }, + { + "grad_norm": 0.16056178510189056, + "learning_rate": 7.122161634083234e-05, + "loss": 0.0211, + "step": 7850 + }, + { + "grad_norm": 0.19346041977405548, + "learning_rate": 7.114672985426516e-05, + "loss": 0.0222, + "step": 7860 + }, + { + "grad_norm": 0.17714178562164307, + "learning_rate": 7.107178555334606e-05, + "loss": 0.0202, + "step": 7870 + }, + { + "grad_norm": 0.19057969748973846, + "learning_rate": 7.099678364296989e-05, + "loss": 0.0228, + "step": 7880 + }, + { + "grad_norm": 0.27934834361076355, + "learning_rate": 7.0921724328189e-05, + "loss": 0.0254, + "step": 7890 + }, + { + "grad_norm": 0.2795708477497101, + "learning_rate": 7.084660781421268e-05, + "loss": 0.0231, + "step": 7900 + }, + { + "grad_norm": 0.16426512598991394, + "learning_rate": 7.077143430640662e-05, + "loss": 0.0227, + "step": 7910 + }, + { + "grad_norm": 0.19890974462032318, + "learning_rate": 7.069620401029232e-05, + "loss": 0.0202, + "step": 7920 + }, + { + "grad_norm": 0.14402411878108978, + "learning_rate": 7.062091713154655e-05, + "loss": 0.0228, + "step": 7930 + }, + { + "grad_norm": 0.19238942861557007, + "learning_rate": 7.054557387600075e-05, + "loss": 0.0235, + "step": 7940 + }, + { + "grad_norm": 0.19047386944293976, + "learning_rate": 7.04701744496405e-05, + "loss": 0.0235, + "step": 7950 + }, + { + "grad_norm": 0.2381032407283783, + "learning_rate": 7.039471905860495e-05, + "loss": 0.0244, + "step": 7960 + }, + { + "grad_norm": 0.22217124700546265, + "learning_rate": 7.031920790918628e-05, + "loss": 0.0275, + "step": 7970 + }, + { + "grad_norm": 0.30119267106056213, + "learning_rate": 7.024364120782906e-05, + "loss": 0.0247, + "step": 7980 + }, + { + "grad_norm": 0.22664779424667358, + "learning_rate": 7.016801916112978e-05, + "loss": 0.0233, + "step": 7990 + }, + { + "grad_norm": 0.14178867638111115, + "learning_rate": 7.009234197583623e-05, + "loss": 0.0208, + "step": 8000 + }, + { + "grad_norm": 0.18423017859458923, + "learning_rate": 7.001660985884692e-05, + "loss": 0.0201, + "step": 8010 + }, + { + "grad_norm": 0.17040474712848663, + "learning_rate": 6.994082301721063e-05, + "loss": 0.0214, + "step": 8020 + }, + { + "grad_norm": 0.1411750316619873, + "learning_rate": 6.986498165812563e-05, + "loss": 0.0219, + "step": 8030 + }, + { + "grad_norm": 0.24464115500450134, + "learning_rate": 6.978908598893932e-05, + "loss": 0.0234, + "step": 8040 + }, + { + "grad_norm": 0.24485130608081818, + "learning_rate": 6.971313621714756e-05, + "loss": 0.0219, + "step": 8050 + }, + { + "grad_norm": 0.14656926691532135, + "learning_rate": 6.96371325503941e-05, + "loss": 0.0216, + "step": 8060 + }, + { + "grad_norm": 0.24350827932357788, + "learning_rate": 6.956107519647014e-05, + "loss": 0.0215, + "step": 8070 + }, + { + "grad_norm": 0.15665143728256226, + "learning_rate": 6.94849643633135e-05, + "loss": 0.0241, + "step": 8080 + }, + { + "grad_norm": 0.1463271826505661, + "learning_rate": 6.940880025900834e-05, + "loss": 0.0214, + "step": 8090 + }, + { + "grad_norm": 0.16267117857933044, + "learning_rate": 6.933258309178438e-05, + "loss": 0.0213, + "step": 8100 + }, + { + "grad_norm": 0.21984468400478363, + "learning_rate": 6.925631307001646e-05, + "loss": 0.0207, + "step": 8110 + }, + { + "grad_norm": 0.25808578729629517, + "learning_rate": 6.91799904022239e-05, + "loss": 0.0216, + "step": 8120 + }, + { + "grad_norm": 0.20477712154388428, + "learning_rate": 6.910361529706997e-05, + "loss": 0.0228, + "step": 8130 + }, + { + "grad_norm": 0.18936611711978912, + "learning_rate": 6.902718796336131e-05, + "loss": 0.0226, + "step": 8140 + }, + { + "grad_norm": 0.1834283322095871, + "learning_rate": 6.895070861004729e-05, + "loss": 0.022, + "step": 8150 + }, + { + "grad_norm": 0.20454129576683044, + "learning_rate": 6.887417744621956e-05, + "loss": 0.0226, + "step": 8160 + }, + { + "grad_norm": 0.16775032877922058, + "learning_rate": 6.87975946811114e-05, + "loss": 0.0227, + "step": 8170 + }, + { + "grad_norm": 0.21968281269073486, + "learning_rate": 6.872096052409718e-05, + "loss": 0.0209, + "step": 8180 + }, + { + "grad_norm": 0.17906726896762848, + "learning_rate": 6.864427518469174e-05, + "loss": 0.0234, + "step": 8190 + }, + { + "grad_norm": 0.16236279904842377, + "learning_rate": 6.856753887254986e-05, + "loss": 0.021, + "step": 8200 + }, + { + "grad_norm": 0.19435322284698486, + "learning_rate": 6.849075179746572e-05, + "loss": 0.0231, + "step": 8210 + }, + { + "grad_norm": 0.22213879227638245, + "learning_rate": 6.841391416937221e-05, + "loss": 0.0224, + "step": 8220 + }, + { + "grad_norm": 0.22625139355659485, + "learning_rate": 6.833702619834053e-05, + "loss": 0.0224, + "step": 8230 + }, + { + "grad_norm": 0.2535237669944763, + "learning_rate": 6.82600880945794e-05, + "loss": 0.024, + "step": 8240 + }, + { + "grad_norm": 0.16354459524154663, + "learning_rate": 6.818310006843468e-05, + "loss": 0.021, + "step": 8250 + }, + { + "grad_norm": 0.1509534865617752, + "learning_rate": 6.810606233038868e-05, + "loss": 0.0198, + "step": 8260 + }, + { + "grad_norm": 0.16952399909496307, + "learning_rate": 6.802897509105966e-05, + "loss": 0.0187, + "step": 8270 + }, + { + "grad_norm": 0.13748878240585327, + "learning_rate": 6.79518385612012e-05, + "loss": 0.0213, + "step": 8280 + }, + { + "grad_norm": 0.16778123378753662, + "learning_rate": 6.787465295170157e-05, + "loss": 0.0204, + "step": 8290 + }, + { + "grad_norm": 0.1332818567752838, + "learning_rate": 6.779741847358332e-05, + "loss": 0.0205, + "step": 8300 + }, + { + "grad_norm": 0.21745198965072632, + "learning_rate": 6.772013533800256e-05, + "loss": 0.0213, + "step": 8310 + }, + { + "grad_norm": 0.14549212157726288, + "learning_rate": 6.764280375624843e-05, + "loss": 0.0231, + "step": 8320 + }, + { + "grad_norm": 0.14919663965702057, + "learning_rate": 6.756542393974252e-05, + "loss": 0.0223, + "step": 8330 + }, + { + "grad_norm": 0.2004791796207428, + "learning_rate": 6.748799610003828e-05, + "loss": 0.0234, + "step": 8340 + }, + { + "grad_norm": 0.219511479139328, + "learning_rate": 6.741052044882048e-05, + "loss": 0.0225, + "step": 8350 + }, + { + "grad_norm": 0.16925059258937836, + "learning_rate": 6.73329971979046e-05, + "loss": 0.0209, + "step": 8360 + }, + { + "grad_norm": 0.14986781775951385, + "learning_rate": 6.725542655923625e-05, + "loss": 0.0187, + "step": 8370 + }, + { + "grad_norm": 0.21496498584747314, + "learning_rate": 6.717780874489057e-05, + "loss": 0.0191, + "step": 8380 + }, + { + "grad_norm": 0.19724883139133453, + "learning_rate": 6.710014396707172e-05, + "loss": 0.0216, + "step": 8390 + }, + { + "grad_norm": 0.12446457147598267, + "learning_rate": 6.702243243811221e-05, + "loss": 0.0215, + "step": 8400 + }, + { + "grad_norm": 0.17544637620449066, + "learning_rate": 6.694467437047244e-05, + "loss": 0.0231, + "step": 8410 + }, + { + "grad_norm": 0.17463184893131256, + "learning_rate": 6.686686997673997e-05, + "loss": 0.0232, + "step": 8420 + }, + { + "grad_norm": 0.1377185881137848, + "learning_rate": 6.678901946962903e-05, + "loss": 0.0198, + "step": 8430 + }, + { + "grad_norm": 0.15128378570079803, + "learning_rate": 6.671112306197996e-05, + "loss": 0.0243, + "step": 8440 + }, + { + "grad_norm": 0.22358955442905426, + "learning_rate": 6.663318096675854e-05, + "loss": 0.0272, + "step": 8450 + }, + { + "grad_norm": 0.19586408138275146, + "learning_rate": 6.655519339705552e-05, + "loss": 0.0199, + "step": 8460 + }, + { + "grad_norm": 0.1928173005580902, + "learning_rate": 6.647716056608588e-05, + "loss": 0.0214, + "step": 8470 + }, + { + "grad_norm": 0.1851319670677185, + "learning_rate": 6.639908268718843e-05, + "loss": 0.0201, + "step": 8480 + }, + { + "grad_norm": 0.14243553578853607, + "learning_rate": 6.632095997382514e-05, + "loss": 0.0222, + "step": 8490 + }, + { + "grad_norm": 0.20461443066596985, + "learning_rate": 6.624279263958047e-05, + "loss": 0.0239, + "step": 8500 + }, + { + "grad_norm": 0.1716454178094864, + "learning_rate": 6.616458089816097e-05, + "loss": 0.0192, + "step": 8510 + }, + { + "grad_norm": 0.15453378856182098, + "learning_rate": 6.608632496339454e-05, + "loss": 0.0207, + "step": 8520 + }, + { + "grad_norm": 0.11159835010766983, + "learning_rate": 6.600802504922988e-05, + "loss": 0.0207, + "step": 8530 + }, + { + "grad_norm": 0.1560823768377304, + "learning_rate": 6.592968136973604e-05, + "loss": 0.0208, + "step": 8540 + }, + { + "grad_norm": 0.23835600912570953, + "learning_rate": 6.585129413910159e-05, + "loss": 0.0209, + "step": 8550 + }, + { + "grad_norm": 0.25999143719673157, + "learning_rate": 6.577286357163424e-05, + "loss": 0.0225, + "step": 8560 + }, + { + "grad_norm": 0.17386066913604736, + "learning_rate": 6.569438988176018e-05, + "loss": 0.0201, + "step": 8570 + }, + { + "grad_norm": 0.21149365603923798, + "learning_rate": 6.561587328402347e-05, + "loss": 0.0229, + "step": 8580 + }, + { + "grad_norm": 0.1868230551481247, + "learning_rate": 6.553731399308549e-05, + "loss": 0.0214, + "step": 8590 + }, + { + "grad_norm": 0.18246783316135406, + "learning_rate": 6.545871222372436e-05, + "loss": 0.0218, + "step": 8600 + }, + { + "grad_norm": 0.14454898238182068, + "learning_rate": 6.538006819083426e-05, + "loss": 0.0239, + "step": 8610 + }, + { + "grad_norm": 0.13045312464237213, + "learning_rate": 6.530138210942505e-05, + "loss": 0.0239, + "step": 8620 + }, + { + "grad_norm": 0.156825453042984, + "learning_rate": 6.522265419462141e-05, + "loss": 0.0214, + "step": 8630 + }, + { + "grad_norm": 0.23999051749706268, + "learning_rate": 6.514388466166248e-05, + "loss": 0.0224, + "step": 8640 + }, + { + "grad_norm": 0.2515720725059509, + "learning_rate": 6.506507372590119e-05, + "loss": 0.0217, + "step": 8650 + }, + { + "grad_norm": 0.167621910572052, + "learning_rate": 6.498622160280355e-05, + "loss": 0.0192, + "step": 8660 + }, + { + "grad_norm": 0.1321040689945221, + "learning_rate": 6.490732850794832e-05, + "loss": 0.0236, + "step": 8670 + }, + { + "grad_norm": 0.1927858591079712, + "learning_rate": 6.482839465702616e-05, + "loss": 0.0201, + "step": 8680 + }, + { + "grad_norm": 0.14037522673606873, + "learning_rate": 6.474942026583923e-05, + "loss": 0.0185, + "step": 8690 + }, + { + "grad_norm": 0.1516280323266983, + "learning_rate": 6.467040555030052e-05, + "loss": 0.0181, + "step": 8700 + }, + { + "grad_norm": 0.18472100794315338, + "learning_rate": 6.459135072643321e-05, + "loss": 0.0229, + "step": 8710 + }, + { + "grad_norm": 0.17317551374435425, + "learning_rate": 6.451225601037019e-05, + "loss": 0.0203, + "step": 8720 + }, + { + "grad_norm": 0.24339652061462402, + "learning_rate": 6.443312161835338e-05, + "loss": 0.0187, + "step": 8730 + }, + { + "grad_norm": 0.16814760863780975, + "learning_rate": 6.43539477667332e-05, + "loss": 0.0206, + "step": 8740 + }, + { + "grad_norm": 0.18786193430423737, + "learning_rate": 6.427473467196793e-05, + "loss": 0.0182, + "step": 8750 + }, + { + "grad_norm": 0.21564754843711853, + "learning_rate": 6.419548255062315e-05, + "loss": 0.0186, + "step": 8760 + }, + { + "grad_norm": 0.27725014090538025, + "learning_rate": 6.411619161937112e-05, + "loss": 0.0254, + "step": 8770 + }, + { + "grad_norm": 0.21169643104076385, + "learning_rate": 6.403686209499022e-05, + "loss": 0.0226, + "step": 8780 + }, + { + "grad_norm": 0.16932587325572968, + "learning_rate": 6.395749419436437e-05, + "loss": 0.0201, + "step": 8790 + }, + { + "grad_norm": 0.17667268216609955, + "learning_rate": 6.387808813448234e-05, + "loss": 0.017, + "step": 8800 + }, + { + "grad_norm": 0.16228170692920685, + "learning_rate": 6.37986441324373e-05, + "loss": 0.0175, + "step": 8810 + }, + { + "grad_norm": 0.12692728638648987, + "learning_rate": 6.37191624054261e-05, + "loss": 0.0205, + "step": 8820 + }, + { + "grad_norm": 0.15658240020275116, + "learning_rate": 6.363964317074872e-05, + "loss": 0.0195, + "step": 8830 + }, + { + "grad_norm": 0.13041932880878448, + "learning_rate": 6.356008664580776e-05, + "loss": 0.0205, + "step": 8840 + }, + { + "grad_norm": 0.14340585470199585, + "learning_rate": 6.348049304810771e-05, + "loss": 0.0212, + "step": 8850 + }, + { + "grad_norm": 0.1328858584165573, + "learning_rate": 6.340086259525442e-05, + "loss": 0.0198, + "step": 8860 + }, + { + "grad_norm": 0.1854487955570221, + "learning_rate": 6.332119550495448e-05, + "loss": 0.0231, + "step": 8870 + }, + { + "grad_norm": 0.14328166842460632, + "learning_rate": 6.324149199501473e-05, + "loss": 0.0192, + "step": 8880 + }, + { + "grad_norm": 0.13808801770210266, + "learning_rate": 6.316175228334146e-05, + "loss": 0.0226, + "step": 8890 + }, + { + "grad_norm": 0.19612519443035126, + "learning_rate": 6.308197658794003e-05, + "loss": 0.0235, + "step": 8900 + }, + { + "grad_norm": 0.15023820102214813, + "learning_rate": 6.300216512691417e-05, + "loss": 0.0186, + "step": 8910 + }, + { + "grad_norm": 0.1928299516439438, + "learning_rate": 6.292231811846532e-05, + "loss": 0.0208, + "step": 8920 + }, + { + "grad_norm": 0.20941108465194702, + "learning_rate": 6.284243578089217e-05, + "loss": 0.0241, + "step": 8930 + }, + { + "grad_norm": 0.28566795587539673, + "learning_rate": 6.276251833258999e-05, + "loss": 0.0219, + "step": 8940 + }, + { + "grad_norm": 0.18755096197128296, + "learning_rate": 6.268256599205003e-05, + "loss": 0.0217, + "step": 8950 + }, + { + "grad_norm": 0.1262677013874054, + "learning_rate": 6.260257897785892e-05, + "loss": 0.0205, + "step": 8960 + }, + { + "grad_norm": 0.1488276720046997, + "learning_rate": 6.252255750869811e-05, + "loss": 0.0228, + "step": 8970 + }, + { + "grad_norm": 0.2091875672340393, + "learning_rate": 6.244250180334325e-05, + "loss": 0.02, + "step": 8980 + }, + { + "grad_norm": 0.19262540340423584, + "learning_rate": 6.236241208066356e-05, + "loss": 0.0217, + "step": 8990 + }, + { + "grad_norm": 0.16485196352005005, + "learning_rate": 6.228228855962133e-05, + "loss": 0.0216, + "step": 9000 + }, + { + "grad_norm": 0.18421217799186707, + "learning_rate": 6.220213145927115e-05, + "loss": 0.0224, + "step": 9010 + }, + { + "grad_norm": 0.24218332767486572, + "learning_rate": 6.212194099875951e-05, + "loss": 0.0226, + "step": 9020 + }, + { + "grad_norm": 0.1489834040403366, + "learning_rate": 6.204171739732405e-05, + "loss": 0.0223, + "step": 9030 + }, + { + "grad_norm": 0.17691490054130554, + "learning_rate": 6.196146087429303e-05, + "loss": 0.0195, + "step": 9040 + }, + { + "grad_norm": 0.24188517034053802, + "learning_rate": 6.188117164908474e-05, + "loss": 0.0184, + "step": 9050 + }, + { + "grad_norm": 0.2162814885377884, + "learning_rate": 6.180084994120684e-05, + "loss": 0.0199, + "step": 9060 + }, + { + "grad_norm": 0.21242709457874298, + "learning_rate": 6.17204959702558e-05, + "loss": 0.0206, + "step": 9070 + }, + { + "grad_norm": 0.17017142474651337, + "learning_rate": 6.164010995591635e-05, + "loss": 0.0194, + "step": 9080 + }, + { + "grad_norm": 0.17777535319328308, + "learning_rate": 6.155969211796076e-05, + "loss": 0.0189, + "step": 9090 + }, + { + "grad_norm": 0.12886588275432587, + "learning_rate": 6.147924267624829e-05, + "loss": 0.0188, + "step": 9100 + }, + { + "grad_norm": 0.2046012282371521, + "learning_rate": 6.13987618507247e-05, + "loss": 0.0174, + "step": 9110 + }, + { + "grad_norm": 0.15767820179462433, + "learning_rate": 6.131824986142147e-05, + "loss": 0.0217, + "step": 9120 + }, + { + "grad_norm": 0.1215910017490387, + "learning_rate": 6.123770692845529e-05, + "loss": 0.0189, + "step": 9130 + }, + { + "grad_norm": 0.14569929242134094, + "learning_rate": 6.11571332720275e-05, + "loss": 0.0197, + "step": 9140 + }, + { + "grad_norm": 0.2217453122138977, + "learning_rate": 6.107652911242336e-05, + "loss": 0.02, + "step": 9150 + }, + { + "grad_norm": 0.13841713964939117, + "learning_rate": 6.0995894670011586e-05, + "loss": 0.019, + "step": 9160 + }, + { + "grad_norm": 0.1597103476524353, + "learning_rate": 6.091523016524368e-05, + "loss": 0.0193, + "step": 9170 + }, + { + "grad_norm": 0.14008677005767822, + "learning_rate": 6.083453581865328e-05, + "loss": 0.0194, + "step": 9180 + }, + { + "grad_norm": 0.15293079614639282, + "learning_rate": 6.075381185085568e-05, + "loss": 0.0195, + "step": 9190 + }, + { + "grad_norm": 0.12000234425067902, + "learning_rate": 6.067305848254709e-05, + "loss": 0.0191, + "step": 9200 + }, + { + "grad_norm": 0.1507752388715744, + "learning_rate": 6.059227593450418e-05, + "loss": 0.0205, + "step": 9210 + }, + { + "grad_norm": 0.1724635362625122, + "learning_rate": 6.051146442758333e-05, + "loss": 0.0197, + "step": 9220 + }, + { + "grad_norm": 0.14416766166687012, + "learning_rate": 6.043062418272012e-05, + "loss": 0.019, + "step": 9230 + }, + { + "grad_norm": 0.1530875414609909, + "learning_rate": 6.0349755420928666e-05, + "loss": 0.0179, + "step": 9240 + }, + { + "grad_norm": 0.12776997685432434, + "learning_rate": 6.0268858363301105e-05, + "loss": 0.0191, + "step": 9250 + }, + { + "grad_norm": 0.14690600335597992, + "learning_rate": 6.018793323100689e-05, + "loss": 0.0206, + "step": 9260 + }, + { + "grad_norm": 0.1346130520105362, + "learning_rate": 6.0106980245292255e-05, + "loss": 0.0219, + "step": 9270 + }, + { + "grad_norm": 0.18886178731918335, + "learning_rate": 6.002599962747957e-05, + "loss": 0.019, + "step": 9280 + }, + { + "grad_norm": 0.2060827910900116, + "learning_rate": 5.994499159896673e-05, + "loss": 0.0203, + "step": 9290 + }, + { + "grad_norm": 0.17232877016067505, + "learning_rate": 5.9863956381226607e-05, + "loss": 0.0189, + "step": 9300 + }, + { + "grad_norm": 0.13772423565387726, + "learning_rate": 5.9782894195806394e-05, + "loss": 0.017, + "step": 9310 + }, + { + "grad_norm": 0.19502772390842438, + "learning_rate": 5.9701805264327004e-05, + "loss": 0.0191, + "step": 9320 + }, + { + "grad_norm": 0.16178490221500397, + "learning_rate": 5.96206898084825e-05, + "loss": 0.0192, + "step": 9330 + }, + { + "grad_norm": 0.1435016691684723, + "learning_rate": 5.953954805003942e-05, + "loss": 0.0227, + "step": 9340 + }, + { + "grad_norm": 0.23129305243492126, + "learning_rate": 5.945838021083623e-05, + "loss": 0.0202, + "step": 9350 + }, + { + "grad_norm": 0.20442074537277222, + "learning_rate": 5.9377186512782714e-05, + "loss": 0.0182, + "step": 9360 + }, + { + "grad_norm": 0.19649991393089294, + "learning_rate": 5.929596717785935e-05, + "loss": 0.0187, + "step": 9370 + }, + { + "grad_norm": 0.1303817629814148, + "learning_rate": 5.921472242811668e-05, + "loss": 0.0174, + "step": 9380 + }, + { + "grad_norm": 0.16999246180057526, + "learning_rate": 5.913345248567475e-05, + "loss": 0.0185, + "step": 9390 + }, + { + "grad_norm": 0.17592598497867584, + "learning_rate": 5.905215757272248e-05, + "loss": 0.0201, + "step": 9400 + }, + { + "grad_norm": 0.19038279354572296, + "learning_rate": 5.897083791151706e-05, + "loss": 0.02, + "step": 9410 + }, + { + "grad_norm": 0.146666020154953, + "learning_rate": 5.888949372438336e-05, + "loss": 0.018, + "step": 9420 + }, + { + "grad_norm": 0.1196974664926529, + "learning_rate": 5.8808125233713255e-05, + "loss": 0.0174, + "step": 9430 + }, + { + "grad_norm": 0.1861487179994583, + "learning_rate": 5.872673266196509e-05, + "loss": 0.0204, + "step": 9440 + }, + { + "grad_norm": 0.16816885769367218, + "learning_rate": 5.864531623166305e-05, + "loss": 0.0175, + "step": 9450 + }, + { + "grad_norm": 0.19386325776576996, + "learning_rate": 5.856387616539656e-05, + "loss": 0.0207, + "step": 9460 + }, + { + "grad_norm": 0.16816996037960052, + "learning_rate": 5.848241268581967e-05, + "loss": 0.0195, + "step": 9470 + }, + { + "grad_norm": 0.16529600322246552, + "learning_rate": 5.840092601565037e-05, + "loss": 0.0184, + "step": 9480 + }, + { + "grad_norm": 0.20107638835906982, + "learning_rate": 5.8319416377670144e-05, + "loss": 0.017, + "step": 9490 + }, + { + "grad_norm": 0.18497569859027863, + "learning_rate": 5.82378839947232e-05, + "loss": 0.0166, + "step": 9500 + }, + { + "grad_norm": 0.1693001389503479, + "learning_rate": 5.815632908971599e-05, + "loss": 0.0163, + "step": 9510 + }, + { + "grad_norm": 0.17620375752449036, + "learning_rate": 5.80747518856165e-05, + "loss": 0.0195, + "step": 9520 + }, + { + "grad_norm": 0.14993426203727722, + "learning_rate": 5.799315260545367e-05, + "loss": 0.018, + "step": 9530 + }, + { + "grad_norm": 0.10744142532348633, + "learning_rate": 5.791153147231686e-05, + "loss": 0.019, + "step": 9540 + }, + { + "grad_norm": 0.16672174632549286, + "learning_rate": 5.782988870935509e-05, + "loss": 0.0198, + "step": 9550 + }, + { + "grad_norm": 0.23908977210521698, + "learning_rate": 5.774822453977657e-05, + "loss": 0.02, + "step": 9560 + }, + { + "grad_norm": 0.19433216750621796, + "learning_rate": 5.7666539186848036e-05, + "loss": 0.0207, + "step": 9570 + }, + { + "grad_norm": 0.16219832003116608, + "learning_rate": 5.758483287389411e-05, + "loss": 0.0203, + "step": 9580 + }, + { + "grad_norm": 0.11743449419736862, + "learning_rate": 5.7503105824296735e-05, + "loss": 0.0205, + "step": 9590 + }, + { + "grad_norm": 0.1957157403230667, + "learning_rate": 5.742135826149453e-05, + "loss": 0.0185, + "step": 9600 + }, + { + "grad_norm": 0.1528787463903427, + "learning_rate": 5.7339590408982223e-05, + "loss": 0.0169, + "step": 9610 + }, + { + "grad_norm": 0.12242112308740616, + "learning_rate": 5.725780249031e-05, + "loss": 0.0186, + "step": 9620 + }, + { + "grad_norm": 0.16867554187774658, + "learning_rate": 5.717599472908292e-05, + "loss": 0.0179, + "step": 9630 + }, + { + "grad_norm": 0.14338399469852448, + "learning_rate": 5.7094167348960237e-05, + "loss": 0.0188, + "step": 9640 + }, + { + "grad_norm": 0.1741122454404831, + "learning_rate": 5.7012320573654945e-05, + "loss": 0.0174, + "step": 9650 + }, + { + "grad_norm": 0.1467796117067337, + "learning_rate": 5.693045462693295e-05, + "loss": 0.0172, + "step": 9660 + }, + { + "grad_norm": 0.15410570800304413, + "learning_rate": 5.684856973261266e-05, + "loss": 0.0192, + "step": 9670 + }, + { + "grad_norm": 0.13427449762821198, + "learning_rate": 5.6766666114564215e-05, + "loss": 0.0186, + "step": 9680 + }, + { + "grad_norm": 0.14436767995357513, + "learning_rate": 5.668474399670899e-05, + "loss": 0.0176, + "step": 9690 + }, + { + "grad_norm": 0.15177932381629944, + "learning_rate": 5.660280360301896e-05, + "loss": 0.0178, + "step": 9700 + }, + { + "grad_norm": 0.1563161462545395, + "learning_rate": 5.652084515751599e-05, + "loss": 0.0192, + "step": 9710 + }, + { + "grad_norm": 0.19394555687904358, + "learning_rate": 5.643886888427137e-05, + "loss": 0.0176, + "step": 9720 + }, + { + "grad_norm": 0.1550029218196869, + "learning_rate": 5.6356875007405074e-05, + "loss": 0.0179, + "step": 9730 + }, + { + "grad_norm": 0.13084563612937927, + "learning_rate": 5.627486375108525e-05, + "loss": 0.02, + "step": 9740 + }, + { + "grad_norm": 0.16264773905277252, + "learning_rate": 5.619283533952754e-05, + "loss": 0.0207, + "step": 9750 + }, + { + "grad_norm": 0.1014072597026825, + "learning_rate": 5.6110789996994474e-05, + "loss": 0.0168, + "step": 9760 + }, + { + "grad_norm": 0.14121492207050323, + "learning_rate": 5.602872794779491e-05, + "loss": 0.0188, + "step": 9770 + }, + { + "grad_norm": 0.14768218994140625, + "learning_rate": 5.594664941628334e-05, + "loss": 0.0194, + "step": 9780 + }, + { + "grad_norm": 0.20915895700454712, + "learning_rate": 5.5864554626859324e-05, + "loss": 0.0199, + "step": 9790 + }, + { + "grad_norm": 0.17958924174308777, + "learning_rate": 5.578244380396691e-05, + "loss": 0.0162, + "step": 9800 + }, + { + "grad_norm": 0.19344228506088257, + "learning_rate": 5.570031717209394e-05, + "loss": 0.0172, + "step": 9810 + }, + { + "grad_norm": 0.1290435642004013, + "learning_rate": 5.561817495577147e-05, + "loss": 0.0187, + "step": 9820 + }, + { + "grad_norm": 0.16577839851379395, + "learning_rate": 5.5536017379573215e-05, + "loss": 0.0178, + "step": 9830 + }, + { + "grad_norm": 0.14159929752349854, + "learning_rate": 5.545384466811483e-05, + "loss": 0.0188, + "step": 9840 + }, + { + "grad_norm": 0.20798265933990479, + "learning_rate": 5.5371657046053384e-05, + "loss": 0.0192, + "step": 9850 + }, + { + "grad_norm": 0.12706337869167328, + "learning_rate": 5.528945473808669e-05, + "loss": 0.0169, + "step": 9860 + }, + { + "grad_norm": 0.15028949081897736, + "learning_rate": 5.520723796895272e-05, + "loss": 0.0176, + "step": 9870 + }, + { + "grad_norm": 0.10984886437654495, + "learning_rate": 5.512500696342897e-05, + "loss": 0.0182, + "step": 9880 + }, + { + "grad_norm": 0.1143980547785759, + "learning_rate": 5.504276194633188e-05, + "loss": 0.02, + "step": 9890 + }, + { + "grad_norm": 0.1573047935962677, + "learning_rate": 5.49605031425162e-05, + "loss": 0.0166, + "step": 9900 + }, + { + "grad_norm": 0.11620697379112244, + "learning_rate": 5.487823077687434e-05, + "loss": 0.0165, + "step": 9910 + }, + { + "grad_norm": 0.14061188697814941, + "learning_rate": 5.4795945074335806e-05, + "loss": 0.0192, + "step": 9920 + }, + { + "grad_norm": 0.19839583337306976, + "learning_rate": 5.471364625986657e-05, + "loss": 0.0209, + "step": 9930 + }, + { + "grad_norm": 0.1596444845199585, + "learning_rate": 5.463133455846845e-05, + "loss": 0.0161, + "step": 9940 + }, + { + "grad_norm": 0.17750969529151917, + "learning_rate": 5.4549010195178505e-05, + "loss": 0.0195, + "step": 9950 + }, + { + "grad_norm": 0.1519780308008194, + "learning_rate": 5.446667339506838e-05, + "loss": 0.0173, + "step": 9960 + }, + { + "grad_norm": 0.17411664128303528, + "learning_rate": 5.4384324383243756e-05, + "loss": 0.0171, + "step": 9970 + }, + { + "grad_norm": 0.176849827170372, + "learning_rate": 5.430196338484368e-05, + "loss": 0.0181, + "step": 9980 + }, + { + "grad_norm": 0.16095474362373352, + "learning_rate": 5.4219590625039975e-05, + "loss": 0.0161, + "step": 9990 + }, + { + "grad_norm": 0.15348845720291138, + "learning_rate": 5.413720632903664e-05, + "loss": 0.0179, + "step": 10000 + }, + { + "grad_norm": 0.1386663168668747, + "learning_rate": 5.405481072206917e-05, + "loss": 0.0161, + "step": 10010 + }, + { + "grad_norm": 0.12349103391170502, + "learning_rate": 5.397240402940402e-05, + "loss": 0.0183, + "step": 10020 + }, + { + "grad_norm": 0.16608865559101105, + "learning_rate": 5.388998647633794e-05, + "loss": 0.0185, + "step": 10030 + }, + { + "grad_norm": 0.16761597990989685, + "learning_rate": 5.380755828819737e-05, + "loss": 0.0165, + "step": 10040 + }, + { + "grad_norm": 0.10259400308132172, + "learning_rate": 5.3725119690337846e-05, + "loss": 0.0177, + "step": 10050 + }, + { + "grad_norm": 0.11251728236675262, + "learning_rate": 5.3642670908143324e-05, + "loss": 0.0153, + "step": 10060 + }, + { + "grad_norm": 0.16619464755058289, + "learning_rate": 5.356021216702562e-05, + "loss": 0.0186, + "step": 10070 + }, + { + "grad_norm": 0.1266852170228958, + "learning_rate": 5.347774369242381e-05, + "loss": 0.0186, + "step": 10080 + }, + { + "grad_norm": 0.09114649146795273, + "learning_rate": 5.3395265709803545e-05, + "loss": 0.0167, + "step": 10090 + }, + { + "grad_norm": 0.20421481132507324, + "learning_rate": 5.331277844465647e-05, + "loss": 0.0206, + "step": 10100 + }, + { + "grad_norm": 0.20459061861038208, + "learning_rate": 5.323028212249963e-05, + "loss": 0.0177, + "step": 10110 + }, + { + "grad_norm": 0.21775595843791962, + "learning_rate": 5.314777696887481e-05, + "loss": 0.0172, + "step": 10120 + }, + { + "grad_norm": 0.17398659884929657, + "learning_rate": 5.306526320934796e-05, + "loss": 0.0171, + "step": 10130 + }, + { + "grad_norm": 0.10391732305288315, + "learning_rate": 5.298274106950854e-05, + "loss": 0.017, + "step": 10140 + }, + { + "grad_norm": 0.17731687426567078, + "learning_rate": 5.290021077496893e-05, + "loss": 0.0158, + "step": 10150 + }, + { + "grad_norm": 0.12182687222957611, + "learning_rate": 5.2817672551363816e-05, + "loss": 0.0167, + "step": 10160 + }, + { + "grad_norm": 0.194319486618042, + "learning_rate": 5.273512662434952e-05, + "loss": 0.0189, + "step": 10170 + }, + { + "grad_norm": 0.17227765917778015, + "learning_rate": 5.265257321960349e-05, + "loss": 0.0184, + "step": 10180 + }, + { + "grad_norm": 0.10741866379976273, + "learning_rate": 5.257001256282357e-05, + "loss": 0.0166, + "step": 10190 + }, + { + "grad_norm": 0.18176622688770294, + "learning_rate": 5.248744487972742e-05, + "loss": 0.0159, + "step": 10200 + }, + { + "grad_norm": 0.21171052753925323, + "learning_rate": 5.240487039605196e-05, + "loss": 0.0161, + "step": 10210 + }, + { + "grad_norm": 0.18893016874790192, + "learning_rate": 5.232228933755267e-05, + "loss": 0.0186, + "step": 10220 + }, + { + "grad_norm": 0.14247570931911469, + "learning_rate": 5.2239701930003006e-05, + "loss": 0.0165, + "step": 10230 + }, + { + "grad_norm": 0.1444028615951538, + "learning_rate": 5.215710839919379e-05, + "loss": 0.017, + "step": 10240 + }, + { + "grad_norm": 0.12994956970214844, + "learning_rate": 5.207450897093257e-05, + "loss": 0.0154, + "step": 10250 + }, + { + "grad_norm": 0.15160958468914032, + "learning_rate": 5.1991903871043046e-05, + "loss": 0.0198, + "step": 10260 + }, + { + "grad_norm": 0.13079297542572021, + "learning_rate": 5.190929332536439e-05, + "loss": 0.0194, + "step": 10270 + }, + { + "grad_norm": 0.13523326814174652, + "learning_rate": 5.182667755975071e-05, + "loss": 0.0181, + "step": 10280 + }, + { + "grad_norm": 0.11969475448131561, + "learning_rate": 5.1744056800070315e-05, + "loss": 0.017, + "step": 10290 + }, + { + "grad_norm": 0.15299108624458313, + "learning_rate": 5.166143127220524e-05, + "loss": 0.0181, + "step": 10300 + }, + { + "grad_norm": 0.15768375992774963, + "learning_rate": 5.1578801202050485e-05, + "loss": 0.0169, + "step": 10310 + }, + { + "grad_norm": 0.15589089691638947, + "learning_rate": 5.149616681551355e-05, + "loss": 0.0144, + "step": 10320 + }, + { + "grad_norm": 0.18822535872459412, + "learning_rate": 5.141352833851367e-05, + "loss": 0.0197, + "step": 10330 + }, + { + "grad_norm": 0.14233779907226562, + "learning_rate": 5.1330885996981285e-05, + "loss": 0.0187, + "step": 10340 + }, + { + "grad_norm": 0.12203872948884964, + "learning_rate": 5.124824001685741e-05, + "loss": 0.0185, + "step": 10350 + }, + { + "grad_norm": 0.187170147895813, + "learning_rate": 5.116559062409298e-05, + "loss": 0.018, + "step": 10360 + }, + { + "grad_norm": 0.12957622110843658, + "learning_rate": 5.10829380446483e-05, + "loss": 0.0174, + "step": 10370 + }, + { + "grad_norm": 0.10922467708587646, + "learning_rate": 5.100028250449235e-05, + "loss": 0.0148, + "step": 10380 + }, + { + "grad_norm": 0.1294066160917282, + "learning_rate": 5.0917624229602234e-05, + "loss": 0.0149, + "step": 10390 + }, + { + "grad_norm": 0.08882381021976471, + "learning_rate": 5.0834963445962524e-05, + "loss": 0.0198, + "step": 10400 + }, + { + "grad_norm": 0.1957225650548935, + "learning_rate": 5.075230037956461e-05, + "loss": 0.0172, + "step": 10410 + }, + { + "grad_norm": 0.16142338514328003, + "learning_rate": 5.0669635256406213e-05, + "loss": 0.0158, + "step": 10420 + }, + { + "grad_norm": 0.13807645440101624, + "learning_rate": 5.058696830249058e-05, + "loss": 0.0151, + "step": 10430 + }, + { + "grad_norm": 0.1055712029337883, + "learning_rate": 5.050429974382602e-05, + "loss": 0.0149, + "step": 10440 + }, + { + "grad_norm": 0.11258240789175034, + "learning_rate": 5.042162980642523e-05, + "loss": 0.0134, + "step": 10450 + }, + { + "grad_norm": 0.13824427127838135, + "learning_rate": 5.033895871630462e-05, + "loss": 0.0174, + "step": 10460 + }, + { + "grad_norm": 0.1465044766664505, + "learning_rate": 5.025628669948386e-05, + "loss": 0.0154, + "step": 10470 + }, + { + "grad_norm": 0.1220289096236229, + "learning_rate": 5.017361398198502e-05, + "loss": 0.0174, + "step": 10480 + }, + { + "grad_norm": 0.17498435080051422, + "learning_rate": 5.009094078983221e-05, + "loss": 0.0205, + "step": 10490 + }, + { + "grad_norm": 0.09604810923337936, + "learning_rate": 5.000826734905073e-05, + "loss": 0.0202, + "step": 10500 + }, + { + "grad_norm": 0.12474366277456284, + "learning_rate": 4.9925593885666645e-05, + "loss": 0.0137, + "step": 10510 + }, + { + "grad_norm": 0.1423722207546234, + "learning_rate": 4.984292062570602e-05, + "loss": 0.0141, + "step": 10520 + }, + { + "grad_norm": 0.12184154987335205, + "learning_rate": 4.976024779519442e-05, + "loss": 0.0162, + "step": 10530 + }, + { + "grad_norm": 0.15307633578777313, + "learning_rate": 4.9677575620156194e-05, + "loss": 0.018, + "step": 10540 + }, + { + "grad_norm": 0.13412360846996307, + "learning_rate": 4.959490432661391e-05, + "loss": 0.0175, + "step": 10550 + }, + { + "grad_norm": 0.10879155248403549, + "learning_rate": 4.9512234140587726e-05, + "loss": 0.0163, + "step": 10560 + }, + { + "grad_norm": 0.13189014792442322, + "learning_rate": 4.942956528809477e-05, + "loss": 0.0158, + "step": 10570 + }, + { + "grad_norm": 0.15158042311668396, + "learning_rate": 4.934689799514854e-05, + "loss": 0.0156, + "step": 10580 + }, + { + "grad_norm": 0.11107157170772552, + "learning_rate": 4.926423248775827e-05, + "loss": 0.0145, + "step": 10590 + }, + { + "grad_norm": 0.12662722170352936, + "learning_rate": 4.918156899192826e-05, + "loss": 0.0171, + "step": 10600 + }, + { + "grad_norm": 0.18004392087459564, + "learning_rate": 4.909890773365738e-05, + "loss": 0.016, + "step": 10610 + }, + { + "grad_norm": 0.15429238975048065, + "learning_rate": 4.9016248938938344e-05, + "loss": 0.0177, + "step": 10620 + }, + { + "grad_norm": 0.15285173058509827, + "learning_rate": 4.8933592833757156e-05, + "loss": 0.0173, + "step": 10630 + }, + { + "grad_norm": 0.1551215648651123, + "learning_rate": 4.8850939644092435e-05, + "loss": 0.0191, + "step": 10640 + }, + { + "grad_norm": 0.17197580635547638, + "learning_rate": 4.876828959591485e-05, + "loss": 0.0178, + "step": 10650 + }, + { + "grad_norm": 0.169466033577919, + "learning_rate": 4.8685642915186474e-05, + "loss": 0.016, + "step": 10660 + }, + { + "grad_norm": 0.1530524343252182, + "learning_rate": 4.860299982786018e-05, + "loss": 0.0178, + "step": 10670 + }, + { + "grad_norm": 0.1665390282869339, + "learning_rate": 4.852036055987901e-05, + "loss": 0.0189, + "step": 10680 + }, + { + "grad_norm": 0.15927554666996002, + "learning_rate": 4.843772533717558e-05, + "loss": 0.0167, + "step": 10690 + }, + { + "grad_norm": 0.1281796097755432, + "learning_rate": 4.835509438567142e-05, + "loss": 0.0154, + "step": 10700 + }, + { + "grad_norm": 0.12752528488636017, + "learning_rate": 4.827246793127639e-05, + "loss": 0.0162, + "step": 10710 + }, + { + "grad_norm": 0.15971392393112183, + "learning_rate": 4.818984619988807e-05, + "loss": 0.0161, + "step": 10720 + }, + { + "grad_norm": 0.15452678501605988, + "learning_rate": 4.810722941739115e-05, + "loss": 0.0169, + "step": 10730 + }, + { + "grad_norm": 0.1358589082956314, + "learning_rate": 4.8024617809656684e-05, + "loss": 0.0189, + "step": 10740 + }, + { + "grad_norm": 0.14593958854675293, + "learning_rate": 4.794201160254171e-05, + "loss": 0.0171, + "step": 10750 + }, + { + "grad_norm": 0.1561196744441986, + "learning_rate": 4.785941102188844e-05, + "loss": 0.0157, + "step": 10760 + }, + { + "grad_norm": 0.15398050844669342, + "learning_rate": 4.7776816293523686e-05, + "loss": 0.0147, + "step": 10770 + }, + { + "grad_norm": 0.1320110708475113, + "learning_rate": 4.769422764325832e-05, + "loss": 0.014, + "step": 10780 + }, + { + "grad_norm": 0.14990536868572235, + "learning_rate": 4.76116452968865e-05, + "loss": 0.0163, + "step": 10790 + }, + { + "grad_norm": 0.15348578989505768, + "learning_rate": 4.752906948018525e-05, + "loss": 0.0151, + "step": 10800 + }, + { + "grad_norm": 0.1372896283864975, + "learning_rate": 4.7446500418913684e-05, + "loss": 0.0178, + "step": 10810 + }, + { + "grad_norm": 0.13223505020141602, + "learning_rate": 4.736393833881247e-05, + "loss": 0.0169, + "step": 10820 + }, + { + "grad_norm": 0.16496150195598602, + "learning_rate": 4.7281383465603194e-05, + "loss": 0.0143, + "step": 10830 + }, + { + "grad_norm": 0.18933743238449097, + "learning_rate": 4.71988360249877e-05, + "loss": 0.0177, + "step": 10840 + }, + { + "grad_norm": 0.15667422115802765, + "learning_rate": 4.7116296242647554e-05, + "loss": 0.0179, + "step": 10850 + }, + { + "grad_norm": 0.18266332149505615, + "learning_rate": 4.703376434424336e-05, + "loss": 0.0152, + "step": 10860 + }, + { + "grad_norm": 0.12549003958702087, + "learning_rate": 4.695124055541421e-05, + "loss": 0.0174, + "step": 10870 + }, + { + "grad_norm": 0.16347110271453857, + "learning_rate": 4.6868725101776934e-05, + "loss": 0.017, + "step": 10880 + }, + { + "grad_norm": 0.16633811593055725, + "learning_rate": 4.678621820892567e-05, + "loss": 0.0155, + "step": 10890 + }, + { + "grad_norm": 0.12812109291553497, + "learning_rate": 4.670372010243111e-05, + "loss": 0.0156, + "step": 10900 + }, + { + "grad_norm": 0.15393684804439545, + "learning_rate": 4.662123100783992e-05, + "loss": 0.0146, + "step": 10910 + }, + { + "grad_norm": 0.1706119030714035, + "learning_rate": 4.653875115067415e-05, + "loss": 0.0157, + "step": 10920 + }, + { + "grad_norm": 0.1704203486442566, + "learning_rate": 4.6456280756430545e-05, + "loss": 0.0162, + "step": 10930 + }, + { + "grad_norm": 0.1871538758277893, + "learning_rate": 4.637382005058004e-05, + "loss": 0.0167, + "step": 10940 + }, + { + "grad_norm": 0.1782759130001068, + "learning_rate": 4.629136925856705e-05, + "loss": 0.0142, + "step": 10950 + }, + { + "grad_norm": 0.15308071672916412, + "learning_rate": 4.6208928605808895e-05, + "loss": 0.0156, + "step": 10960 + }, + { + "grad_norm": 0.15167579054832458, + "learning_rate": 4.612649831769519e-05, + "loss": 0.0166, + "step": 10970 + }, + { + "grad_norm": 0.11804986745119095, + "learning_rate": 4.604407861958715e-05, + "loss": 0.0153, + "step": 10980 + }, + { + "grad_norm": 0.1385330855846405, + "learning_rate": 4.5961669736817114e-05, + "loss": 0.0162, + "step": 10990 + }, + { + "grad_norm": 0.12075856328010559, + "learning_rate": 4.5879271894687814e-05, + "loss": 0.015, + "step": 11000 + }, + { + "grad_norm": 0.12524457275867462, + "learning_rate": 4.5796885318471826e-05, + "loss": 0.0177, + "step": 11010 + }, + { + "grad_norm": 0.19302158057689667, + "learning_rate": 4.571451023341086e-05, + "loss": 0.0171, + "step": 11020 + }, + { + "grad_norm": 0.16245462000370026, + "learning_rate": 4.563214686471527e-05, + "loss": 0.0149, + "step": 11030 + }, + { + "grad_norm": 0.14350807666778564, + "learning_rate": 4.5549795437563365e-05, + "loss": 0.0175, + "step": 11040 + }, + { + "grad_norm": 0.1560516655445099, + "learning_rate": 4.546745617710081e-05, + "loss": 0.0165, + "step": 11050 + }, + { + "grad_norm": 0.1624559462070465, + "learning_rate": 4.5385129308440014e-05, + "loss": 0.0163, + "step": 11060 + }, + { + "grad_norm": 0.13163962960243225, + "learning_rate": 4.530281505665944e-05, + "loss": 0.0142, + "step": 11070 + }, + { + "grad_norm": 0.1744009107351303, + "learning_rate": 4.5220513646803134e-05, + "loss": 0.0168, + "step": 11080 + }, + { + "grad_norm": 0.22604048252105713, + "learning_rate": 4.513822530388003e-05, + "loss": 0.0162, + "step": 11090 + }, + { + "grad_norm": 0.14584003388881683, + "learning_rate": 4.5055950252863296e-05, + "loss": 0.0162, + "step": 11100 + }, + { + "grad_norm": 0.1038496196269989, + "learning_rate": 4.4973688718689803e-05, + "loss": 0.015, + "step": 11110 + }, + { + "grad_norm": 0.17408005893230438, + "learning_rate": 4.4891440926259406e-05, + "loss": 0.015, + "step": 11120 + }, + { + "grad_norm": 0.13025762140750885, + "learning_rate": 4.480920710043443e-05, + "loss": 0.0163, + "step": 11130 + }, + { + "grad_norm": 0.1203264519572258, + "learning_rate": 4.4726987466039044e-05, + "loss": 0.0167, + "step": 11140 + }, + { + "grad_norm": 0.13094156980514526, + "learning_rate": 4.46447822478586e-05, + "loss": 0.0141, + "step": 11150 + }, + { + "grad_norm": 0.14033634960651398, + "learning_rate": 4.4562591670638974e-05, + "loss": 0.0165, + "step": 11160 + }, + { + "grad_norm": 0.10585950314998627, + "learning_rate": 4.4480415959086105e-05, + "loss": 0.0157, + "step": 11170 + }, + { + "grad_norm": 0.10921136289834976, + "learning_rate": 4.439825533786522e-05, + "loss": 0.0149, + "step": 11180 + }, + { + "grad_norm": 0.16036371886730194, + "learning_rate": 4.431611003160035e-05, + "loss": 0.0174, + "step": 11190 + }, + { + "grad_norm": 0.1530076116323471, + "learning_rate": 4.4233980264873636e-05, + "loss": 0.0165, + "step": 11200 + }, + { + "grad_norm": 0.13739179074764252, + "learning_rate": 4.4151866262224684e-05, + "loss": 0.0179, + "step": 11210 + }, + { + "grad_norm": 0.0982503667473793, + "learning_rate": 4.406976824815006e-05, + "loss": 0.0127, + "step": 11220 + }, + { + "grad_norm": 0.15988130867481232, + "learning_rate": 4.3987686447102595e-05, + "loss": 0.0185, + "step": 11230 + }, + { + "grad_norm": 0.1203751266002655, + "learning_rate": 4.3905621083490804e-05, + "loss": 0.0152, + "step": 11240 + }, + { + "grad_norm": 0.13053373992443085, + "learning_rate": 4.3823572381678286e-05, + "loss": 0.0149, + "step": 11250 + }, + { + "grad_norm": 0.13859491050243378, + "learning_rate": 4.374154056598301e-05, + "loss": 0.0144, + "step": 11260 + }, + { + "grad_norm": 0.1259634792804718, + "learning_rate": 4.3659525860676845e-05, + "loss": 0.0148, + "step": 11270 + }, + { + "grad_norm": 0.13633209466934204, + "learning_rate": 4.3577528489984854e-05, + "loss": 0.0143, + "step": 11280 + }, + { + "grad_norm": 0.10781408846378326, + "learning_rate": 4.349554867808476e-05, + "loss": 0.0131, + "step": 11290 + }, + { + "grad_norm": 0.12550900876522064, + "learning_rate": 4.34135866491062e-05, + "loss": 0.0117, + "step": 11300 + }, + { + "grad_norm": 0.17351140081882477, + "learning_rate": 4.333164262713022e-05, + "loss": 0.0157, + "step": 11310 + }, + { + "grad_norm": 0.11698879301548004, + "learning_rate": 4.324971683618868e-05, + "loss": 0.0154, + "step": 11320 + }, + { + "grad_norm": 0.14138099551200867, + "learning_rate": 4.316780950026354e-05, + "loss": 0.014, + "step": 11330 + }, + { + "grad_norm": 0.12799644470214844, + "learning_rate": 4.308592084328637e-05, + "loss": 0.0157, + "step": 11340 + }, + { + "grad_norm": 0.13187606632709503, + "learning_rate": 4.3004051089137576e-05, + "loss": 0.0182, + "step": 11350 + }, + { + "grad_norm": 0.08409903198480606, + "learning_rate": 4.292220046164597e-05, + "loss": 0.0158, + "step": 11360 + }, + { + "grad_norm": 0.13147087395191193, + "learning_rate": 4.2840369184588035e-05, + "loss": 0.0179, + "step": 11370 + }, + { + "grad_norm": 0.1346842497587204, + "learning_rate": 4.2758557481687345e-05, + "loss": 0.0151, + "step": 11380 + }, + { + "grad_norm": 0.14093443751335144, + "learning_rate": 4.267676557661403e-05, + "loss": 0.0146, + "step": 11390 + }, + { + "grad_norm": 0.20171605050563812, + "learning_rate": 4.2594993692983955e-05, + "loss": 0.0166, + "step": 11400 + }, + { + "grad_norm": 0.12062186002731323, + "learning_rate": 4.251324205435837e-05, + "loss": 0.0166, + "step": 11410 + }, + { + "grad_norm": 0.13168655335903168, + "learning_rate": 4.243151088424312e-05, + "loss": 0.0149, + "step": 11420 + }, + { + "grad_norm": 0.1192360669374466, + "learning_rate": 4.234980040608813e-05, + "loss": 0.0166, + "step": 11430 + }, + { + "grad_norm": 0.12471235543489456, + "learning_rate": 4.22681108432867e-05, + "loss": 0.0143, + "step": 11440 + }, + { + "grad_norm": 0.12549076974391937, + "learning_rate": 4.2186442419174984e-05, + "loss": 0.015, + "step": 11450 + }, + { + "grad_norm": 0.17939548194408417, + "learning_rate": 4.210479535703133e-05, + "loss": 0.0152, + "step": 11460 + }, + { + "grad_norm": 0.09044060111045837, + "learning_rate": 4.202316988007567e-05, + "loss": 0.0146, + "step": 11470 + }, + { + "grad_norm": 0.11873579770326614, + "learning_rate": 4.194156621146901e-05, + "loss": 0.0129, + "step": 11480 + }, + { + "grad_norm": 0.1776263564825058, + "learning_rate": 4.1859984574312596e-05, + "loss": 0.015, + "step": 11490 + }, + { + "grad_norm": 0.14686188101768494, + "learning_rate": 4.177842519164752e-05, + "loss": 0.014, + "step": 11500 + }, + { + "grad_norm": 0.16359327733516693, + "learning_rate": 4.169688828645404e-05, + "loss": 0.0148, + "step": 11510 + }, + { + "grad_norm": 0.20391836762428284, + "learning_rate": 4.161537408165092e-05, + "loss": 0.0156, + "step": 11520 + }, + { + "grad_norm": 0.14375638961791992, + "learning_rate": 4.1533882800094924e-05, + "loss": 0.0177, + "step": 11530 + }, + { + "grad_norm": 0.0956331118941307, + "learning_rate": 4.145241466458005e-05, + "loss": 0.0173, + "step": 11540 + }, + { + "grad_norm": 0.14115546643733978, + "learning_rate": 4.13709698978371e-05, + "loss": 0.0157, + "step": 11550 + }, + { + "grad_norm": 0.13794995844364166, + "learning_rate": 4.1289548722532944e-05, + "loss": 0.016, + "step": 11560 + }, + { + "grad_norm": 0.1263679713010788, + "learning_rate": 4.120815136126999e-05, + "loss": 0.0149, + "step": 11570 + }, + { + "grad_norm": 0.112674281001091, + "learning_rate": 4.112677803658548e-05, + "loss": 0.0181, + "step": 11580 + }, + { + "grad_norm": 0.16525673866271973, + "learning_rate": 4.1045428970951e-05, + "loss": 0.0158, + "step": 11590 + }, + { + "grad_norm": 0.1573369950056076, + "learning_rate": 4.0964104386771785e-05, + "loss": 0.0148, + "step": 11600 + }, + { + "grad_norm": 0.14529259502887726, + "learning_rate": 4.0882804506386144e-05, + "loss": 0.0175, + "step": 11610 + }, + { + "grad_norm": 0.1014423593878746, + "learning_rate": 4.080152955206485e-05, + "loss": 0.0194, + "step": 11620 + }, + { + "grad_norm": 0.20214314758777618, + "learning_rate": 4.0720279746010505e-05, + "loss": 0.0188, + "step": 11630 + }, + { + "grad_norm": 0.13978305459022522, + "learning_rate": 4.063905531035699e-05, + "loss": 0.0155, + "step": 11640 + }, + { + "grad_norm": 0.14400368928909302, + "learning_rate": 4.055785646716882e-05, + "loss": 0.0146, + "step": 11650 + }, + { + "grad_norm": 0.1247081607580185, + "learning_rate": 4.047668343844051e-05, + "loss": 0.0144, + "step": 11660 + }, + { + "grad_norm": 0.145439013838768, + "learning_rate": 4.039553644609604e-05, + "loss": 0.0157, + "step": 11670 + }, + { + "grad_norm": 0.1491650491952896, + "learning_rate": 4.0314415711988176e-05, + "loss": 0.0143, + "step": 11680 + }, + { + "grad_norm": 0.15634377300739288, + "learning_rate": 4.023332145789792e-05, + "loss": 0.0149, + "step": 11690 + }, + { + "grad_norm": 0.15609179437160492, + "learning_rate": 4.015225390553385e-05, + "loss": 0.0173, + "step": 11700 + }, + { + "grad_norm": 0.1583641618490219, + "learning_rate": 4.007121327653158e-05, + "loss": 0.0155, + "step": 11710 + }, + { + "grad_norm": 0.1330731362104416, + "learning_rate": 3.9990199792453064e-05, + "loss": 0.0133, + "step": 11720 + }, + { + "grad_norm": 0.11838053166866302, + "learning_rate": 3.9909213674786103e-05, + "loss": 0.0142, + "step": 11730 + }, + { + "grad_norm": 0.09256942570209503, + "learning_rate": 3.982825514494363e-05, + "loss": 0.0151, + "step": 11740 + }, + { + "grad_norm": 0.1232718825340271, + "learning_rate": 3.974732442426319e-05, + "loss": 0.0154, + "step": 11750 + }, + { + "grad_norm": 0.13511840999126434, + "learning_rate": 3.966642173400629e-05, + "loss": 0.0147, + "step": 11760 + }, + { + "grad_norm": 0.15866443514823914, + "learning_rate": 3.9585547295357764e-05, + "loss": 0.014, + "step": 11770 + }, + { + "grad_norm": 0.12842710316181183, + "learning_rate": 3.950470132942526e-05, + "loss": 0.0142, + "step": 11780 + }, + { + "grad_norm": 0.13316993415355682, + "learning_rate": 3.942388405723856e-05, + "loss": 0.0152, + "step": 11790 + }, + { + "grad_norm": 0.12535324692726135, + "learning_rate": 3.9343095699749e-05, + "loss": 0.0147, + "step": 11800 + }, + { + "grad_norm": 0.1451651006937027, + "learning_rate": 3.9262336477828874e-05, + "loss": 0.0152, + "step": 11810 + }, + { + "grad_norm": 0.1585710495710373, + "learning_rate": 3.9181606612270794e-05, + "loss": 0.0162, + "step": 11820 + }, + { + "grad_norm": 0.1250775009393692, + "learning_rate": 3.910090632378713e-05, + "loss": 0.017, + "step": 11830 + }, + { + "grad_norm": 0.16008895635604858, + "learning_rate": 3.90202358330094e-05, + "loss": 0.0156, + "step": 11840 + }, + { + "grad_norm": 0.14432847499847412, + "learning_rate": 3.8939595360487656e-05, + "loss": 0.0159, + "step": 11850 + }, + { + "grad_norm": 0.14685307443141937, + "learning_rate": 3.885898512668984e-05, + "loss": 0.0154, + "step": 11860 + }, + { + "grad_norm": 0.12076714634895325, + "learning_rate": 3.877840535200127e-05, + "loss": 0.0163, + "step": 11870 + }, + { + "grad_norm": 0.13044147193431854, + "learning_rate": 3.869785625672397e-05, + "loss": 0.0155, + "step": 11880 + }, + { + "grad_norm": 0.14798730611801147, + "learning_rate": 3.8617338061076094e-05, + "loss": 0.015, + "step": 11890 + }, + { + "grad_norm": 0.15533815324306488, + "learning_rate": 3.853685098519132e-05, + "loss": 0.0145, + "step": 11900 + }, + { + "grad_norm": 0.1446336954832077, + "learning_rate": 3.845639524911823e-05, + "loss": 0.0148, + "step": 11910 + }, + { + "grad_norm": 0.1601383090019226, + "learning_rate": 3.837597107281974e-05, + "loss": 0.0148, + "step": 11920 + }, + { + "grad_norm": 0.11422371864318848, + "learning_rate": 3.829557867617247e-05, + "loss": 0.0179, + "step": 11930 + }, + { + "grad_norm": 0.20501482486724854, + "learning_rate": 3.821521827896618e-05, + "loss": 0.0162, + "step": 11940 + }, + { + "grad_norm": 0.110312320291996, + "learning_rate": 3.81348901009031e-05, + "loss": 0.0128, + "step": 11950 + }, + { + "grad_norm": 0.14152300357818604, + "learning_rate": 3.805459436159741e-05, + "loss": 0.0159, + "step": 11960 + }, + { + "grad_norm": 0.14452366530895233, + "learning_rate": 3.797433128057461e-05, + "loss": 0.0146, + "step": 11970 + }, + { + "grad_norm": 0.12984851002693176, + "learning_rate": 3.789410107727089e-05, + "loss": 0.015, + "step": 11980 + }, + { + "grad_norm": 0.1130700632929802, + "learning_rate": 3.781390397103257e-05, + "loss": 0.0139, + "step": 11990 + }, + { + "grad_norm": 0.1647651493549347, + "learning_rate": 3.7733740181115455e-05, + "loss": 0.0166, + "step": 12000 + }, + { + "grad_norm": 0.18699532747268677, + "learning_rate": 3.7653609926684306e-05, + "loss": 0.016, + "step": 12010 + }, + { + "grad_norm": 0.15569140017032623, + "learning_rate": 3.757351342681217e-05, + "loss": 0.0142, + "step": 12020 + }, + { + "grad_norm": 0.1413174867630005, + "learning_rate": 3.749345090047982e-05, + "loss": 0.0136, + "step": 12030 + }, + { + "grad_norm": 0.13615532219409943, + "learning_rate": 3.741342256657515e-05, + "loss": 0.0146, + "step": 12040 + }, + { + "grad_norm": 0.13739195466041565, + "learning_rate": 3.7333428643892567e-05, + "loss": 0.0156, + "step": 12050 + }, + { + "grad_norm": 0.1359894871711731, + "learning_rate": 3.725346935113239e-05, + "loss": 0.0153, + "step": 12060 + }, + { + "grad_norm": 0.10953269153833389, + "learning_rate": 3.717354490690029e-05, + "loss": 0.0135, + "step": 12070 + }, + { + "grad_norm": 0.15777398645877838, + "learning_rate": 3.709365552970664e-05, + "loss": 0.014, + "step": 12080 + }, + { + "grad_norm": 0.11898507922887802, + "learning_rate": 3.7013801437965945e-05, + "loss": 0.0136, + "step": 12090 + }, + { + "grad_norm": 0.13760054111480713, + "learning_rate": 3.693398284999623e-05, + "loss": 0.0135, + "step": 12100 + }, + { + "grad_norm": 0.15929579734802246, + "learning_rate": 3.6854199984018484e-05, + "loss": 0.0159, + "step": 12110 + }, + { + "grad_norm": 0.16446372866630554, + "learning_rate": 3.677445305815601e-05, + "loss": 0.0146, + "step": 12120 + }, + { + "grad_norm": 0.1320229470729828, + "learning_rate": 3.669474229043387e-05, + "loss": 0.0139, + "step": 12130 + }, + { + "grad_norm": 0.10756336897611618, + "learning_rate": 3.6615067898778235e-05, + "loss": 0.0125, + "step": 12140 + }, + { + "grad_norm": 0.15356606245040894, + "learning_rate": 3.6535430101015866e-05, + "loss": 0.0147, + "step": 12150 + }, + { + "grad_norm": 0.10317644476890564, + "learning_rate": 3.645582911487345e-05, + "loss": 0.0143, + "step": 12160 + }, + { + "grad_norm": 0.12801851332187653, + "learning_rate": 3.637626515797706e-05, + "loss": 0.0127, + "step": 12170 + }, + { + "grad_norm": 0.17475886642932892, + "learning_rate": 3.629673844785152e-05, + "loss": 0.0142, + "step": 12180 + }, + { + "grad_norm": 0.1472134292125702, + "learning_rate": 3.621724920191979e-05, + "loss": 0.0124, + "step": 12190 + }, + { + "grad_norm": 0.1240682303905487, + "learning_rate": 3.6137797637502444e-05, + "loss": 0.0158, + "step": 12200 + }, + { + "grad_norm": 0.1859951913356781, + "learning_rate": 3.6058383971817035e-05, + "loss": 0.0167, + "step": 12210 + }, + { + "grad_norm": 0.13284876942634583, + "learning_rate": 3.59790084219775e-05, + "loss": 0.0128, + "step": 12220 + }, + { + "grad_norm": 0.13101252913475037, + "learning_rate": 3.589967120499353e-05, + "loss": 0.0166, + "step": 12230 + }, + { + "grad_norm": 0.14012062549591064, + "learning_rate": 3.5820372537770075e-05, + "loss": 0.0166, + "step": 12240 + }, + { + "grad_norm": 0.08424890786409378, + "learning_rate": 3.5741112637106655e-05, + "loss": 0.0123, + "step": 12250 + }, + { + "grad_norm": 0.18260999023914337, + "learning_rate": 3.5661891719696804e-05, + "loss": 0.0139, + "step": 12260 + }, + { + "grad_norm": 0.13156132400035858, + "learning_rate": 3.5582710002127504e-05, + "loss": 0.0155, + "step": 12270 + }, + { + "grad_norm": 0.13513632118701935, + "learning_rate": 3.550356770087853e-05, + "loss": 0.0138, + "step": 12280 + }, + { + "grad_norm": 0.11364325135946274, + "learning_rate": 3.5424465032321914e-05, + "loss": 0.0137, + "step": 12290 + }, + { + "grad_norm": 0.13653968274593353, + "learning_rate": 3.5345402212721335e-05, + "loss": 0.0161, + "step": 12300 + }, + { + "grad_norm": 0.13233284652233124, + "learning_rate": 3.526637945823152e-05, + "loss": 0.0137, + "step": 12310 + }, + { + "grad_norm": 0.1129714846611023, + "learning_rate": 3.518739698489767e-05, + "loss": 0.0137, + "step": 12320 + }, + { + "grad_norm": 0.1178167536854744, + "learning_rate": 3.510845500865485e-05, + "loss": 0.0115, + "step": 12330 + }, + { + "grad_norm": 0.13436441123485565, + "learning_rate": 3.502955374532739e-05, + "loss": 0.0135, + "step": 12340 + }, + { + "grad_norm": 0.1632746160030365, + "learning_rate": 3.495069341062836e-05, + "loss": 0.0155, + "step": 12350 + }, + { + "grad_norm": 0.15889474749565125, + "learning_rate": 3.4871874220158896e-05, + "loss": 0.0138, + "step": 12360 + }, + { + "grad_norm": 0.14875760674476624, + "learning_rate": 3.479309638940762e-05, + "loss": 0.0128, + "step": 12370 + }, + { + "grad_norm": 0.11081046611070633, + "learning_rate": 3.4714360133750146e-05, + "loss": 0.0135, + "step": 12380 + }, + { + "grad_norm": 0.10351568460464478, + "learning_rate": 3.463566566844839e-05, + "loss": 0.0122, + "step": 12390 + }, + { + "grad_norm": 0.13017849624156952, + "learning_rate": 3.4557013208650016e-05, + "loss": 0.0137, + "step": 12400 + }, + { + "grad_norm": 0.13252219557762146, + "learning_rate": 3.4478402969387857e-05, + "loss": 0.0115, + "step": 12410 + }, + { + "grad_norm": 0.2398897409439087, + "learning_rate": 3.4399835165579266e-05, + "loss": 0.0127, + "step": 12420 + }, + { + "grad_norm": 0.08913063257932663, + "learning_rate": 3.4321310012025645e-05, + "loss": 0.0156, + "step": 12430 + }, + { + "grad_norm": 0.12528473138809204, + "learning_rate": 3.424282772341176e-05, + "loss": 0.0133, + "step": 12440 + }, + { + "grad_norm": 0.2074984759092331, + "learning_rate": 3.416438851430519e-05, + "loss": 0.0128, + "step": 12450 + }, + { + "grad_norm": 0.09442967176437378, + "learning_rate": 3.408599259915577e-05, + "loss": 0.0122, + "step": 12460 + }, + { + "grad_norm": 0.10211984813213348, + "learning_rate": 3.400764019229487e-05, + "loss": 0.0161, + "step": 12470 + }, + { + "grad_norm": 0.1321975290775299, + "learning_rate": 3.3929331507935035e-05, + "loss": 0.0129, + "step": 12480 + }, + { + "grad_norm": 0.10855752229690552, + "learning_rate": 3.3851066760169196e-05, + "loss": 0.0174, + "step": 12490 + }, + { + "grad_norm": 0.1284664422273636, + "learning_rate": 3.377284616297021e-05, + "loss": 0.0131, + "step": 12500 + }, + { + "grad_norm": 0.11752691119909286, + "learning_rate": 3.3694669930190166e-05, + "loss": 0.0138, + "step": 12510 + }, + { + "grad_norm": 0.10745704919099808, + "learning_rate": 3.36165382755599e-05, + "loss": 0.0111, + "step": 12520 + }, + { + "grad_norm": 0.126668319106102, + "learning_rate": 3.35384514126884e-05, + "loss": 0.0135, + "step": 12530 + }, + { + "grad_norm": 0.13161689043045044, + "learning_rate": 3.3460409555062154e-05, + "loss": 0.0116, + "step": 12540 + }, + { + "grad_norm": 0.1370457261800766, + "learning_rate": 3.3382412916044645e-05, + "loss": 0.013, + "step": 12550 + }, + { + "grad_norm": 0.11534154415130615, + "learning_rate": 3.330446170887566e-05, + "loss": 0.0166, + "step": 12560 + }, + { + "grad_norm": 0.12002326548099518, + "learning_rate": 3.3226556146670834e-05, + "loss": 0.0145, + "step": 12570 + }, + { + "grad_norm": 0.12132298946380615, + "learning_rate": 3.314869644242102e-05, + "loss": 0.0162, + "step": 12580 + }, + { + "grad_norm": 0.11905062198638916, + "learning_rate": 3.3070882808991674e-05, + "loss": 0.0125, + "step": 12590 + }, + { + "grad_norm": 0.11214898526668549, + "learning_rate": 3.2993115459122305e-05, + "loss": 0.0113, + "step": 12600 + }, + { + "grad_norm": 0.09516862779855728, + "learning_rate": 3.2915394605425835e-05, + "loss": 0.011, + "step": 12610 + }, + { + "grad_norm": 0.10201261937618256, + "learning_rate": 3.283772046038816e-05, + "loss": 0.0134, + "step": 12620 + }, + { + "grad_norm": 0.12799382209777832, + "learning_rate": 3.276009323636739e-05, + "loss": 0.0128, + "step": 12630 + }, + { + "grad_norm": 0.10710633546113968, + "learning_rate": 3.268251314559344e-05, + "loss": 0.0135, + "step": 12640 + }, + { + "grad_norm": 0.14563435316085815, + "learning_rate": 3.2604980400167254e-05, + "loss": 0.0136, + "step": 12650 + }, + { + "grad_norm": 0.14591513574123383, + "learning_rate": 3.252749521206042e-05, + "loss": 0.0134, + "step": 12660 + }, + { + "grad_norm": 0.14750060439109802, + "learning_rate": 3.2450057793114494e-05, + "loss": 0.0135, + "step": 12670 + }, + { + "grad_norm": 0.1272416114807129, + "learning_rate": 3.2372668355040435e-05, + "loss": 0.0137, + "step": 12680 + }, + { + "grad_norm": 0.1310454159975052, + "learning_rate": 3.2295327109418005e-05, + "loss": 0.0151, + "step": 12690 + }, + { + "grad_norm": 0.13678844273090363, + "learning_rate": 3.221803426769518e-05, + "loss": 0.0124, + "step": 12700 + }, + { + "grad_norm": 0.11800270527601242, + "learning_rate": 3.214079004118768e-05, + "loss": 0.0129, + "step": 12710 + }, + { + "grad_norm": 0.10299576073884964, + "learning_rate": 3.2063594641078234e-05, + "loss": 0.0138, + "step": 12720 + }, + { + "grad_norm": 0.1434839814901352, + "learning_rate": 3.198644827841616e-05, + "loss": 0.0162, + "step": 12730 + }, + { + "grad_norm": 0.13421641290187836, + "learning_rate": 3.1909351164116654e-05, + "loss": 0.0122, + "step": 12740 + }, + { + "grad_norm": 0.12477222830057144, + "learning_rate": 3.183230350896026e-05, + "loss": 0.0137, + "step": 12750 + }, + { + "grad_norm": 0.10647811740636826, + "learning_rate": 3.1755305523592337e-05, + "loss": 0.016, + "step": 12760 + }, + { + "grad_norm": 0.1225629672408104, + "learning_rate": 3.167835741852245e-05, + "loss": 0.0157, + "step": 12770 + }, + { + "grad_norm": 0.13227476179599762, + "learning_rate": 3.160145940412378e-05, + "loss": 0.015, + "step": 12780 + }, + { + "grad_norm": 0.13701468706130981, + "learning_rate": 3.1524611690632545e-05, + "loss": 0.0146, + "step": 12790 + }, + { + "grad_norm": 0.11905475705862045, + "learning_rate": 3.144781448814746e-05, + "loss": 0.014, + "step": 12800 + }, + { + "grad_norm": 0.12421087920665741, + "learning_rate": 3.1371068006629145e-05, + "loss": 0.0118, + "step": 12810 + }, + { + "grad_norm": 0.09958793222904205, + "learning_rate": 3.129437245589956e-05, + "loss": 0.0148, + "step": 12820 + }, + { + "grad_norm": 0.11220111697912216, + "learning_rate": 3.121772804564143e-05, + "loss": 0.0118, + "step": 12830 + }, + { + "grad_norm": 0.13333453238010406, + "learning_rate": 3.11411349853976e-05, + "loss": 0.0128, + "step": 12840 + }, + { + "grad_norm": 0.113494373857975, + "learning_rate": 3.10645934845706e-05, + "loss": 0.0139, + "step": 12850 + }, + { + "grad_norm": 0.15671712160110474, + "learning_rate": 3.098810375242196e-05, + "loss": 0.0122, + "step": 12860 + }, + { + "grad_norm": 0.1252211630344391, + "learning_rate": 3.0911665998071704e-05, + "loss": 0.0136, + "step": 12870 + }, + { + "grad_norm": 0.09996208548545837, + "learning_rate": 3.083528043049774e-05, + "loss": 0.0131, + "step": 12880 + }, + { + "grad_norm": 0.183076411485672, + "learning_rate": 3.0758947258535255e-05, + "loss": 0.0137, + "step": 12890 + }, + { + "grad_norm": 0.126839280128479, + "learning_rate": 3.068266669087625e-05, + "loss": 0.013, + "step": 12900 + }, + { + "grad_norm": 0.1417466253042221, + "learning_rate": 3.060643893606887e-05, + "loss": 0.0111, + "step": 12910 + }, + { + "grad_norm": 0.168529212474823, + "learning_rate": 3.053026420251693e-05, + "loss": 0.014, + "step": 12920 + }, + { + "grad_norm": 0.16259074211120605, + "learning_rate": 3.0454142698479183e-05, + "loss": 0.0147, + "step": 12930 + }, + { + "grad_norm": 0.09722009301185608, + "learning_rate": 3.0378074632068954e-05, + "loss": 0.0125, + "step": 12940 + }, + { + "grad_norm": 0.12659865617752075, + "learning_rate": 3.0302060211253408e-05, + "loss": 0.0139, + "step": 12950 + }, + { + "grad_norm": 0.11219172924757004, + "learning_rate": 3.0226099643853073e-05, + "loss": 0.0134, + "step": 12960 + }, + { + "grad_norm": 0.11740846186876297, + "learning_rate": 3.0150193137541283e-05, + "loss": 0.013, + "step": 12970 + }, + { + "grad_norm": 0.16941086947917938, + "learning_rate": 3.0074340899843467e-05, + "loss": 0.0128, + "step": 12980 + }, + { + "grad_norm": 0.15864995121955872, + "learning_rate": 2.999854313813677e-05, + "loss": 0.0151, + "step": 12990 + }, + { + "grad_norm": 0.12871518731117249, + "learning_rate": 2.9922800059649382e-05, + "loss": 0.0139, + "step": 13000 + }, + { + "grad_norm": 0.11101030558347702, + "learning_rate": 2.9847111871459976e-05, + "loss": 0.0132, + "step": 13010 + }, + { + "grad_norm": 0.1117987111210823, + "learning_rate": 2.977147878049721e-05, + "loss": 0.0144, + "step": 13020 + }, + { + "grad_norm": 0.16980081796646118, + "learning_rate": 2.9695900993539006e-05, + "loss": 0.0149, + "step": 13030 + }, + { + "grad_norm": 0.10659262537956238, + "learning_rate": 2.9620378717212183e-05, + "loss": 0.0119, + "step": 13040 + }, + { + "grad_norm": 0.13019990921020508, + "learning_rate": 2.9544912157991745e-05, + "loss": 0.0138, + "step": 13050 + }, + { + "grad_norm": 0.14628802239894867, + "learning_rate": 2.9469501522200405e-05, + "loss": 0.0137, + "step": 13060 + }, + { + "grad_norm": 0.10786167532205582, + "learning_rate": 2.9394147016007946e-05, + "loss": 0.0108, + "step": 13070 + }, + { + "grad_norm": 0.14825493097305298, + "learning_rate": 2.9318848845430702e-05, + "loss": 0.0135, + "step": 13080 + }, + { + "grad_norm": 0.09231767803430557, + "learning_rate": 2.9243607216331013e-05, + "loss": 0.013, + "step": 13090 + }, + { + "grad_norm": 0.147687628865242, + "learning_rate": 2.916842233441661e-05, + "loss": 0.0116, + "step": 13100 + }, + { + "grad_norm": 0.11064019799232483, + "learning_rate": 2.90932944052401e-05, + "loss": 0.0123, + "step": 13110 + }, + { + "grad_norm": 0.11891503632068634, + "learning_rate": 2.9018223634198354e-05, + "loss": 0.0129, + "step": 13120 + }, + { + "grad_norm": 0.10179593414068222, + "learning_rate": 2.8943210226532025e-05, + "loss": 0.0132, + "step": 13130 + }, + { + "grad_norm": 0.16178156435489655, + "learning_rate": 2.8868254387324857e-05, + "loss": 0.0139, + "step": 13140 + }, + { + "grad_norm": 0.1412755846977234, + "learning_rate": 2.8793356321503306e-05, + "loss": 0.0137, + "step": 13150 + }, + { + "grad_norm": 0.09551704674959183, + "learning_rate": 2.87185162338358e-05, + "loss": 0.0117, + "step": 13160 + }, + { + "grad_norm": 0.10173319280147552, + "learning_rate": 2.8643734328932253e-05, + "loss": 0.0141, + "step": 13170 + }, + { + "grad_norm": 0.12520283460617065, + "learning_rate": 2.856901081124359e-05, + "loss": 0.0131, + "step": 13180 + }, + { + "grad_norm": 0.12247993052005768, + "learning_rate": 2.8494345885061002e-05, + "loss": 0.0138, + "step": 13190 + }, + { + "grad_norm": 0.13983102142810822, + "learning_rate": 2.8419739754515616e-05, + "loss": 0.0107, + "step": 13200 + }, + { + "grad_norm": 0.09981770813465118, + "learning_rate": 2.8345192623577666e-05, + "loss": 0.0124, + "step": 13210 + }, + { + "grad_norm": 0.14096727967262268, + "learning_rate": 2.8270704696056193e-05, + "loss": 0.0119, + "step": 13220 + }, + { + "grad_norm": 0.09955500811338425, + "learning_rate": 2.8196276175598367e-05, + "loss": 0.0146, + "step": 13230 + }, + { + "grad_norm": 0.12892650067806244, + "learning_rate": 2.8121907265688884e-05, + "loss": 0.0118, + "step": 13240 + }, + { + "grad_norm": 0.12986475229263306, + "learning_rate": 2.804759816964957e-05, + "loss": 0.0122, + "step": 13250 + }, + { + "grad_norm": 0.13239455223083496, + "learning_rate": 2.797334909063857e-05, + "loss": 0.0143, + "step": 13260 + }, + { + "grad_norm": 0.10744354128837585, + "learning_rate": 2.7899160231650056e-05, + "loss": 0.0129, + "step": 13270 + }, + { + "grad_norm": 0.10660792887210846, + "learning_rate": 2.7825031795513585e-05, + "loss": 0.0134, + "step": 13280 + }, + { + "grad_norm": 0.12968522310256958, + "learning_rate": 2.775096398489341e-05, + "loss": 0.0152, + "step": 13290 + }, + { + "grad_norm": 0.12975488603115082, + "learning_rate": 2.7676957002288163e-05, + "loss": 0.0136, + "step": 13300 + }, + { + "grad_norm": 0.12956364452838898, + "learning_rate": 2.760301105003003e-05, + "loss": 0.0122, + "step": 13310 + }, + { + "grad_norm": 0.14073896408081055, + "learning_rate": 2.752912633028446e-05, + "loss": 0.0144, + "step": 13320 + }, + { + "grad_norm": 0.12888725101947784, + "learning_rate": 2.7455303045049474e-05, + "loss": 0.014, + "step": 13330 + }, + { + "grad_norm": 0.14723517000675201, + "learning_rate": 2.7381541396155098e-05, + "loss": 0.0138, + "step": 13340 + }, + { + "grad_norm": 0.12811152637004852, + "learning_rate": 2.730784158526286e-05, + "loss": 0.0121, + "step": 13350 + }, + { + "grad_norm": 0.1611941009759903, + "learning_rate": 2.723420381386521e-05, + "loss": 0.0161, + "step": 13360 + }, + { + "grad_norm": 0.13735568523406982, + "learning_rate": 2.7160628283285018e-05, + "loss": 0.0109, + "step": 13370 + }, + { + "grad_norm": 0.13074660301208496, + "learning_rate": 2.7087115194675007e-05, + "loss": 0.0133, + "step": 13380 + }, + { + "grad_norm": 0.12460041046142578, + "learning_rate": 2.701366474901712e-05, + "loss": 0.0117, + "step": 13390 + }, + { + "grad_norm": 0.11751213669776917, + "learning_rate": 2.6940277147122085e-05, + "loss": 0.0135, + "step": 13400 + }, + { + "grad_norm": 0.13527220487594604, + "learning_rate": 2.686695258962878e-05, + "loss": 0.012, + "step": 13410 + }, + { + "grad_norm": 0.13173392415046692, + "learning_rate": 2.679369127700375e-05, + "loss": 0.0144, + "step": 13420 + }, + { + "grad_norm": 0.17878493666648865, + "learning_rate": 2.672049340954067e-05, + "loss": 0.0169, + "step": 13430 + }, + { + "grad_norm": 0.10489585995674133, + "learning_rate": 2.6647359187359676e-05, + "loss": 0.0125, + "step": 13440 + }, + { + "grad_norm": 0.11608193069696426, + "learning_rate": 2.6574288810406946e-05, + "loss": 0.0117, + "step": 13450 + }, + { + "grad_norm": 0.0915096178650856, + "learning_rate": 2.6501282478454083e-05, + "loss": 0.0114, + "step": 13460 + }, + { + "grad_norm": 0.16352947056293488, + "learning_rate": 2.6428340391097618e-05, + "loss": 0.0148, + "step": 13470 + }, + { + "grad_norm": 0.10337302833795547, + "learning_rate": 2.6355462747758485e-05, + "loss": 0.0119, + "step": 13480 + }, + { + "grad_norm": 0.10317914187908173, + "learning_rate": 2.6282649747681304e-05, + "loss": 0.0109, + "step": 13490 + }, + { + "grad_norm": 0.11913973838090897, + "learning_rate": 2.620990158993406e-05, + "loss": 0.0123, + "step": 13500 + }, + { + "grad_norm": 0.1330854743719101, + "learning_rate": 2.6137218473407477e-05, + "loss": 0.0132, + "step": 13510 + }, + { + "grad_norm": 0.12897777557373047, + "learning_rate": 2.606460059681436e-05, + "loss": 0.0113, + "step": 13520 + }, + { + "grad_norm": 0.0806979313492775, + "learning_rate": 2.599204815868928e-05, + "loss": 0.0113, + "step": 13530 + }, + { + "grad_norm": 0.09123429656028748, + "learning_rate": 2.5919561357387756e-05, + "loss": 0.014, + "step": 13540 + }, + { + "grad_norm": 0.06981328874826431, + "learning_rate": 2.5847140391085972e-05, + "loss": 0.0136, + "step": 13550 + }, + { + "grad_norm": 0.14494577050209045, + "learning_rate": 2.5774785457780103e-05, + "loss": 0.013, + "step": 13560 + }, + { + "grad_norm": 0.13974572718143463, + "learning_rate": 2.5702496755285753e-05, + "loss": 0.0127, + "step": 13570 + }, + { + "grad_norm": 0.12633803486824036, + "learning_rate": 2.5630274481237483e-05, + "loss": 0.013, + "step": 13580 + }, + { + "grad_norm": 0.14200319349765778, + "learning_rate": 2.5558118833088197e-05, + "loss": 0.0135, + "step": 13590 + }, + { + "grad_norm": 0.10983054339885712, + "learning_rate": 2.548603000810872e-05, + "loss": 0.013, + "step": 13600 + }, + { + "grad_norm": 0.12070131301879883, + "learning_rate": 2.5414008203387152e-05, + "loss": 0.011, + "step": 13610 + }, + { + "grad_norm": 0.09498189389705658, + "learning_rate": 2.534205361582834e-05, + "loss": 0.0111, + "step": 13620 + }, + { + "grad_norm": 0.10117040574550629, + "learning_rate": 2.527016644215338e-05, + "loss": 0.0129, + "step": 13630 + }, + { + "grad_norm": 0.08696826547384262, + "learning_rate": 2.519834687889905e-05, + "loss": 0.0118, + "step": 13640 + }, + { + "grad_norm": 0.10143747180700302, + "learning_rate": 2.5126595122417295e-05, + "loss": 0.0142, + "step": 13650 + }, + { + "grad_norm": 0.1503414660692215, + "learning_rate": 2.5054911368874713e-05, + "loss": 0.0122, + "step": 13660 + }, + { + "grad_norm": 0.1367895007133484, + "learning_rate": 2.4983295814251916e-05, + "loss": 0.0128, + "step": 13670 + }, + { + "grad_norm": 0.11799922585487366, + "learning_rate": 2.4911748654343105e-05, + "loss": 0.0142, + "step": 13680 + }, + { + "grad_norm": 0.11096301674842834, + "learning_rate": 2.4840270084755463e-05, + "loss": 0.0125, + "step": 13690 + }, + { + "grad_norm": 0.1219366043806076, + "learning_rate": 2.4768860300908685e-05, + "loss": 0.0139, + "step": 13700 + }, + { + "grad_norm": 0.10047262161970139, + "learning_rate": 2.469751949803443e-05, + "loss": 0.0129, + "step": 13710 + }, + { + "grad_norm": 0.10014031827449799, + "learning_rate": 2.4626247871175666e-05, + "loss": 0.0103, + "step": 13720 + }, + { + "grad_norm": 0.09014005959033966, + "learning_rate": 2.4555045615186346e-05, + "loss": 0.0161, + "step": 13730 + }, + { + "grad_norm": 0.12338422983884811, + "learning_rate": 2.4483912924730677e-05, + "loss": 0.0123, + "step": 13740 + }, + { + "grad_norm": 0.1309572458267212, + "learning_rate": 2.4412849994282742e-05, + "loss": 0.0147, + "step": 13750 + }, + { + "grad_norm": 0.1429263800382614, + "learning_rate": 2.434185701812592e-05, + "loss": 0.013, + "step": 13760 + }, + { + "grad_norm": 0.11043151468038559, + "learning_rate": 2.4270934190352218e-05, + "loss": 0.0111, + "step": 13770 + }, + { + "grad_norm": 0.09178663045167923, + "learning_rate": 2.4200081704861998e-05, + "loss": 0.0113, + "step": 13780 + }, + { + "grad_norm": 0.09059850126504898, + "learning_rate": 2.412929975536321e-05, + "loss": 0.0137, + "step": 13790 + }, + { + "grad_norm": 0.10262386500835419, + "learning_rate": 2.4058588535371017e-05, + "loss": 0.0114, + "step": 13800 + }, + { + "grad_norm": 0.14737792313098907, + "learning_rate": 2.3987948238207243e-05, + "loss": 0.0126, + "step": 13810 + }, + { + "grad_norm": 0.11754263937473297, + "learning_rate": 2.3917379056999678e-05, + "loss": 0.0119, + "step": 13820 + }, + { + "grad_norm": 0.13891632854938507, + "learning_rate": 2.3846881184681824e-05, + "loss": 0.0125, + "step": 13830 + }, + { + "grad_norm": 0.1205991879105568, + "learning_rate": 2.377645481399214e-05, + "loss": 0.0123, + "step": 13840 + }, + { + "grad_norm": 0.11768194288015366, + "learning_rate": 2.3706100137473667e-05, + "loss": 0.0112, + "step": 13850 + }, + { + "grad_norm": 0.1152067556977272, + "learning_rate": 2.3635817347473394e-05, + "loss": 0.0102, + "step": 13860 + }, + { + "grad_norm": 0.11635255813598633, + "learning_rate": 2.3565606636141757e-05, + "loss": 0.0142, + "step": 13870 + }, + { + "grad_norm": 0.10487168282270432, + "learning_rate": 2.3495468195432203e-05, + "loss": 0.0114, + "step": 13880 + }, + { + "grad_norm": 0.11733433604240417, + "learning_rate": 2.3425402217100507e-05, + "loss": 0.011, + "step": 13890 + }, + { + "grad_norm": 0.11520569771528244, + "learning_rate": 2.3355408892704424e-05, + "loss": 0.0114, + "step": 13900 + }, + { + "grad_norm": 0.09005911648273468, + "learning_rate": 2.3285488413603003e-05, + "loss": 0.0114, + "step": 13910 + }, + { + "grad_norm": 0.12757213413715363, + "learning_rate": 2.321564097095615e-05, + "loss": 0.0117, + "step": 13920 + }, + { + "grad_norm": 0.10761798173189163, + "learning_rate": 2.3145866755724142e-05, + "loss": 0.0111, + "step": 13930 + }, + { + "grad_norm": 0.09664459526538849, + "learning_rate": 2.307616595866699e-05, + "loss": 0.0108, + "step": 13940 + }, + { + "grad_norm": 0.13742592930793762, + "learning_rate": 2.3006538770344032e-05, + "loss": 0.0114, + "step": 13950 + }, + { + "grad_norm": 0.11331397294998169, + "learning_rate": 2.293698538111334e-05, + "loss": 0.0135, + "step": 13960 + }, + { + "grad_norm": 0.10556375235319138, + "learning_rate": 2.28675059811312e-05, + "loss": 0.0137, + "step": 13970 + }, + { + "grad_norm": 0.09589388966560364, + "learning_rate": 2.279810076035167e-05, + "loss": 0.0103, + "step": 13980 + }, + { + "grad_norm": 0.11377747356891632, + "learning_rate": 2.272876990852596e-05, + "loss": 0.0127, + "step": 13990 + }, + { + "grad_norm": 0.12210900336503983, + "learning_rate": 2.265951361520195e-05, + "loss": 0.0094, + "step": 14000 + }, + { + "grad_norm": 0.1865001916885376, + "learning_rate": 2.2590332069723748e-05, + "loss": 0.0109, + "step": 14010 + }, + { + "grad_norm": 0.12490807473659515, + "learning_rate": 2.2521225461231004e-05, + "loss": 0.0122, + "step": 14020 + }, + { + "grad_norm": 0.10145914554595947, + "learning_rate": 2.2452193978658597e-05, + "loss": 0.0104, + "step": 14030 + }, + { + "grad_norm": 0.09514082968235016, + "learning_rate": 2.238323781073594e-05, + "loss": 0.0099, + "step": 14040 + }, + { + "grad_norm": 0.12236886471509933, + "learning_rate": 2.2314357145986552e-05, + "loss": 0.0131, + "step": 14050 + }, + { + "grad_norm": 0.09739136695861816, + "learning_rate": 2.224555217272757e-05, + "loss": 0.0116, + "step": 14060 + }, + { + "grad_norm": 0.09764561802148819, + "learning_rate": 2.2176823079069127e-05, + "loss": 0.0088, + "step": 14070 + }, + { + "grad_norm": 0.12355641275644302, + "learning_rate": 2.210817005291398e-05, + "loss": 0.0098, + "step": 14080 + }, + { + "grad_norm": 0.11599385738372803, + "learning_rate": 2.203959328195686e-05, + "loss": 0.0112, + "step": 14090 + }, + { + "grad_norm": 0.15587230026721954, + "learning_rate": 2.1971092953684026e-05, + "loss": 0.0105, + "step": 14100 + }, + { + "grad_norm": 0.13454626500606537, + "learning_rate": 2.1902669255372788e-05, + "loss": 0.0101, + "step": 14110 + }, + { + "grad_norm": 0.11352057754993439, + "learning_rate": 2.1834322374090897e-05, + "loss": 0.0121, + "step": 14120 + }, + { + "grad_norm": 0.0980420708656311, + "learning_rate": 2.1766052496696153e-05, + "loss": 0.0112, + "step": 14130 + }, + { + "grad_norm": 0.10784881561994553, + "learning_rate": 2.169785980983577e-05, + "loss": 0.0111, + "step": 14140 + }, + { + "grad_norm": 0.11844468116760254, + "learning_rate": 2.162974449994593e-05, + "loss": 0.0099, + "step": 14150 + }, + { + "grad_norm": 0.10883970558643341, + "learning_rate": 2.1561706753251337e-05, + "loss": 0.0098, + "step": 14160 + }, + { + "grad_norm": 0.11267892271280289, + "learning_rate": 2.1493746755764544e-05, + "loss": 0.0121, + "step": 14170 + }, + { + "grad_norm": 0.14904311299324036, + "learning_rate": 2.1425864693285635e-05, + "loss": 0.0116, + "step": 14180 + }, + { + "grad_norm": 0.1703404039144516, + "learning_rate": 2.1358060751401547e-05, + "loss": 0.0141, + "step": 14190 + }, + { + "grad_norm": 0.1179344430565834, + "learning_rate": 2.129033511548566e-05, + "loss": 0.0104, + "step": 14200 + }, + { + "grad_norm": 0.113071508705616, + "learning_rate": 2.1222687970697315e-05, + "loss": 0.0124, + "step": 14210 + }, + { + "grad_norm": 0.10459605604410172, + "learning_rate": 2.1155119501981173e-05, + "loss": 0.0103, + "step": 14220 + }, + { + "grad_norm": 0.14708572626113892, + "learning_rate": 2.1087629894066895e-05, + "loss": 0.0135, + "step": 14230 + }, + { + "grad_norm": 0.1336475908756256, + "learning_rate": 2.1020219331468473e-05, + "loss": 0.0123, + "step": 14240 + }, + { + "grad_norm": 0.10219430923461914, + "learning_rate": 2.095288799848379e-05, + "loss": 0.0122, + "step": 14250 + }, + { + "grad_norm": 0.11198897659778595, + "learning_rate": 2.088563607919417e-05, + "loss": 0.0114, + "step": 14260 + }, + { + "grad_norm": 0.10178937017917633, + "learning_rate": 2.0818463757463786e-05, + "loss": 0.0115, + "step": 14270 + }, + { + "grad_norm": 0.09164676815271378, + "learning_rate": 2.0751371216939175e-05, + "loss": 0.0097, + "step": 14280 + }, + { + "grad_norm": 0.09583322703838348, + "learning_rate": 2.068435864104882e-05, + "loss": 0.0103, + "step": 14290 + }, + { + "grad_norm": 0.09914228320121765, + "learning_rate": 2.0617426213002506e-05, + "loss": 0.0108, + "step": 14300 + }, + { + "grad_norm": 0.07630550116300583, + "learning_rate": 2.055057411579097e-05, + "loss": 0.013, + "step": 14310 + }, + { + "grad_norm": 0.15564921498298645, + "learning_rate": 2.0483802532185286e-05, + "loss": 0.0103, + "step": 14320 + }, + { + "grad_norm": 0.10511241853237152, + "learning_rate": 2.041711164473638e-05, + "loss": 0.0099, + "step": 14330 + }, + { + "grad_norm": 0.09792450070381165, + "learning_rate": 2.0350501635774637e-05, + "loss": 0.0127, + "step": 14340 + }, + { + "grad_norm": 0.08692257851362228, + "learning_rate": 2.0283972687409247e-05, + "loss": 0.01, + "step": 14350 + }, + { + "grad_norm": 0.0911308154463768, + "learning_rate": 2.021752498152784e-05, + "loss": 0.0122, + "step": 14360 + }, + { + "grad_norm": 0.0754559263586998, + "learning_rate": 2.015115869979589e-05, + "loss": 0.01, + "step": 14370 + }, + { + "grad_norm": 0.09537773579359055, + "learning_rate": 2.0084874023656265e-05, + "loss": 0.0113, + "step": 14380 + }, + { + "grad_norm": 0.11594177037477493, + "learning_rate": 2.001867113432877e-05, + "loss": 0.0105, + "step": 14390 + }, + { + "grad_norm": 0.13202723860740662, + "learning_rate": 1.995255021280954e-05, + "loss": 0.0105, + "step": 14400 + }, + { + "grad_norm": 0.0932583436369896, + "learning_rate": 1.9886511439870688e-05, + "loss": 0.0106, + "step": 14410 + }, + { + "grad_norm": 0.0999087318778038, + "learning_rate": 1.9820554996059675e-05, + "loss": 0.0108, + "step": 14420 + }, + { + "grad_norm": 0.17903649806976318, + "learning_rate": 1.9754681061698893e-05, + "loss": 0.0112, + "step": 14430 + }, + { + "grad_norm": 0.10335102677345276, + "learning_rate": 1.9688889816885185e-05, + "loss": 0.0122, + "step": 14440 + }, + { + "grad_norm": 0.09123598039150238, + "learning_rate": 1.962318144148928e-05, + "loss": 0.0101, + "step": 14450 + }, + { + "grad_norm": 0.09797730296850204, + "learning_rate": 1.955755611515539e-05, + "loss": 0.0109, + "step": 14460 + }, + { + "grad_norm": 0.15314173698425293, + "learning_rate": 1.9492014017300642e-05, + "loss": 0.0103, + "step": 14470 + }, + { + "grad_norm": 0.12908782064914703, + "learning_rate": 1.942655532711461e-05, + "loss": 0.0097, + "step": 14480 + }, + { + "grad_norm": 0.1173621341586113, + "learning_rate": 1.9361180223558882e-05, + "loss": 0.0097, + "step": 14490 + }, + { + "grad_norm": 0.10326097905635834, + "learning_rate": 1.929588888536647e-05, + "loss": 0.0119, + "step": 14500 + }, + { + "grad_norm": 0.12936003506183624, + "learning_rate": 1.9230681491041425e-05, + "loss": 0.0116, + "step": 14510 + }, + { + "grad_norm": 0.12333337962627411, + "learning_rate": 1.9165558218858264e-05, + "loss": 0.011, + "step": 14520 + }, + { + "grad_norm": 0.13529351353645325, + "learning_rate": 1.9100519246861505e-05, + "loss": 0.0124, + "step": 14530 + }, + { + "grad_norm": 0.12396562844514847, + "learning_rate": 1.9035564752865248e-05, + "loss": 0.0094, + "step": 14540 + }, + { + "grad_norm": 0.0972055122256279, + "learning_rate": 1.897069491445258e-05, + "loss": 0.0121, + "step": 14550 + }, + { + "grad_norm": 0.11542759835720062, + "learning_rate": 1.890590990897515e-05, + "loss": 0.0126, + "step": 14560 + }, + { + "grad_norm": 0.10494445264339447, + "learning_rate": 1.884120991355272e-05, + "loss": 0.01, + "step": 14570 + }, + { + "grad_norm": 0.09562406688928604, + "learning_rate": 1.8776595105072576e-05, + "loss": 0.0096, + "step": 14580 + }, + { + "grad_norm": 0.09210069477558136, + "learning_rate": 1.8712065660189166e-05, + "loss": 0.0107, + "step": 14590 + }, + { + "grad_norm": 0.09391437470912933, + "learning_rate": 1.8647621755323513e-05, + "loss": 0.0122, + "step": 14600 + }, + { + "grad_norm": 0.12200205028057098, + "learning_rate": 1.858326356666278e-05, + "loss": 0.011, + "step": 14610 + }, + { + "grad_norm": 0.08863028138875961, + "learning_rate": 1.851899127015983e-05, + "loss": 0.0106, + "step": 14620 + }, + { + "grad_norm": 0.10602623969316483, + "learning_rate": 1.8454805041532626e-05, + "loss": 0.0102, + "step": 14630 + }, + { + "grad_norm": 0.10410308837890625, + "learning_rate": 1.8390705056263906e-05, + "loss": 0.0113, + "step": 14640 + }, + { + "grad_norm": 0.07248975336551666, + "learning_rate": 1.832669148960057e-05, + "loss": 0.01, + "step": 14650 + }, + { + "grad_norm": 0.10344856977462769, + "learning_rate": 1.8262764516553233e-05, + "loss": 0.0107, + "step": 14660 + }, + { + "grad_norm": 0.08993128687143326, + "learning_rate": 1.8198924311895843e-05, + "loss": 0.0115, + "step": 14670 + }, + { + "grad_norm": 0.09466415643692017, + "learning_rate": 1.813517105016505e-05, + "loss": 0.0109, + "step": 14680 + }, + { + "grad_norm": 0.1439206302165985, + "learning_rate": 1.8071504905659888e-05, + "loss": 0.0107, + "step": 14690 + }, + { + "grad_norm": 0.11155963689088821, + "learning_rate": 1.800792605244109e-05, + "loss": 0.012, + "step": 14700 + }, + { + "grad_norm": 0.09676221013069153, + "learning_rate": 1.7944434664330844e-05, + "loss": 0.0106, + "step": 14710 + }, + { + "grad_norm": 0.0695369765162468, + "learning_rate": 1.7881030914912212e-05, + "loss": 0.0098, + "step": 14720 + }, + { + "grad_norm": 0.11870352923870087, + "learning_rate": 1.7817714977528577e-05, + "loss": 0.0105, + "step": 14730 + }, + { + "grad_norm": 0.09013182669878006, + "learning_rate": 1.7754487025283332e-05, + "loss": 0.0104, + "step": 14740 + }, + { + "grad_norm": 0.12589795887470245, + "learning_rate": 1.7691347231039275e-05, + "loss": 0.0086, + "step": 14750 + }, + { + "grad_norm": 0.16133029758930206, + "learning_rate": 1.7628295767418164e-05, + "loss": 0.0133, + "step": 14760 + }, + { + "grad_norm": 0.13052208721637726, + "learning_rate": 1.7565332806800333e-05, + "loss": 0.0116, + "step": 14770 + }, + { + "grad_norm": 0.06853306293487549, + "learning_rate": 1.750245852132408e-05, + "loss": 0.0106, + "step": 14780 + }, + { + "grad_norm": 0.09216471016407013, + "learning_rate": 1.7439673082885323e-05, + "loss": 0.0104, + "step": 14790 + }, + { + "grad_norm": 0.07402797788381577, + "learning_rate": 1.7376976663137047e-05, + "loss": 0.013, + "step": 14800 + }, + { + "grad_norm": 0.09811637550592422, + "learning_rate": 1.7314369433488853e-05, + "loss": 0.0104, + "step": 14810 + }, + { + "grad_norm": 0.08107998222112656, + "learning_rate": 1.7251851565106548e-05, + "loss": 0.01, + "step": 14820 + }, + { + "grad_norm": 0.09519882500171661, + "learning_rate": 1.7189423228911574e-05, + "loss": 0.0131, + "step": 14830 + }, + { + "grad_norm": 0.10114264488220215, + "learning_rate": 1.7127084595580606e-05, + "loss": 0.0117, + "step": 14840 + }, + { + "grad_norm": 0.0987163633108139, + "learning_rate": 1.706483583554513e-05, + "loss": 0.0106, + "step": 14850 + }, + { + "grad_norm": 0.09987714141607285, + "learning_rate": 1.700267711899083e-05, + "loss": 0.0112, + "step": 14860 + }, + { + "grad_norm": 0.10829785466194153, + "learning_rate": 1.69406086158573e-05, + "loss": 0.0114, + "step": 14870 + }, + { + "grad_norm": 0.09733504056930542, + "learning_rate": 1.6878630495837455e-05, + "loss": 0.0101, + "step": 14880 + }, + { + "grad_norm": 0.10879287868738174, + "learning_rate": 1.681674292837707e-05, + "loss": 0.0102, + "step": 14890 + }, + { + "grad_norm": 0.15284156799316406, + "learning_rate": 1.6754946082674444e-05, + "loss": 0.0124, + "step": 14900 + }, + { + "grad_norm": 0.11000285297632217, + "learning_rate": 1.6693240127679748e-05, + "loss": 0.0111, + "step": 14910 + }, + { + "grad_norm": 0.07903582602739334, + "learning_rate": 1.663162523209475e-05, + "loss": 0.0112, + "step": 14920 + }, + { + "grad_norm": 0.11662151664495468, + "learning_rate": 1.6570101564372193e-05, + "loss": 0.0135, + "step": 14930 + }, + { + "grad_norm": 0.11682089418172836, + "learning_rate": 1.650866929271543e-05, + "loss": 0.0102, + "step": 14940 + }, + { + "grad_norm": 0.09364151954650879, + "learning_rate": 1.644732858507797e-05, + "loss": 0.0096, + "step": 14950 + }, + { + "grad_norm": 0.10374051332473755, + "learning_rate": 1.6386079609162943e-05, + "loss": 0.0096, + "step": 14960 + }, + { + "grad_norm": 0.09023815393447876, + "learning_rate": 1.6324922532422742e-05, + "loss": 0.011, + "step": 14970 + }, + { + "grad_norm": 0.11377856135368347, + "learning_rate": 1.6263857522058434e-05, + "loss": 0.0115, + "step": 14980 + }, + { + "grad_norm": 0.12053507566452026, + "learning_rate": 1.6202884745019443e-05, + "loss": 0.0096, + "step": 14990 + }, + { + "grad_norm": 0.08721872419118881, + "learning_rate": 1.614200436800304e-05, + "loss": 0.0109, + "step": 15000 + }, + { + "grad_norm": 0.1320895254611969, + "learning_rate": 1.6081216557453814e-05, + "loss": 0.0139, + "step": 15010 + }, + { + "grad_norm": 0.08385386317968369, + "learning_rate": 1.6020521479563367e-05, + "loss": 0.0104, + "step": 15020 + }, + { + "grad_norm": 0.13200055062770844, + "learning_rate": 1.5959919300269654e-05, + "loss": 0.0091, + "step": 15030 + }, + { + "grad_norm": 0.10987170040607452, + "learning_rate": 1.5899410185256764e-05, + "loss": 0.0097, + "step": 15040 + }, + { + "grad_norm": 0.10612881928682327, + "learning_rate": 1.583899429995431e-05, + "loss": 0.0112, + "step": 15050 + }, + { + "grad_norm": 0.09509719163179398, + "learning_rate": 1.5778671809536993e-05, + "loss": 0.0093, + "step": 15060 + }, + { + "grad_norm": 0.15343886613845825, + "learning_rate": 1.5718442878924246e-05, + "loss": 0.0121, + "step": 15070 + }, + { + "grad_norm": 0.12362321466207504, + "learning_rate": 1.5658307672779593e-05, + "loss": 0.0111, + "step": 15080 + }, + { + "grad_norm": 0.1134907528758049, + "learning_rate": 1.5598266355510427e-05, + "loss": 0.0119, + "step": 15090 + }, + { + "grad_norm": 0.07345488667488098, + "learning_rate": 1.553831909126744e-05, + "loss": 0.0112, + "step": 15100 + }, + { + "grad_norm": 0.0879063680768013, + "learning_rate": 1.5478466043944135e-05, + "loss": 0.0115, + "step": 15110 + }, + { + "grad_norm": 0.11450450122356415, + "learning_rate": 1.5418707377176468e-05, + "loss": 0.0107, + "step": 15120 + }, + { + "grad_norm": 0.08029681444168091, + "learning_rate": 1.535904325434233e-05, + "loss": 0.01, + "step": 15130 + }, + { + "grad_norm": 0.11782268434762955, + "learning_rate": 1.529947383856118e-05, + "loss": 0.0113, + "step": 15140 + }, + { + "grad_norm": 0.09880916774272919, + "learning_rate": 1.5239999292693524e-05, + "loss": 0.0086, + "step": 15150 + }, + { + "grad_norm": 0.12976066768169403, + "learning_rate": 1.5180619779340505e-05, + "loss": 0.0091, + "step": 15160 + }, + { + "grad_norm": 0.11458215862512589, + "learning_rate": 1.5121335460843428e-05, + "loss": 0.011, + "step": 15170 + }, + { + "grad_norm": 0.10735385864973068, + "learning_rate": 1.5062146499283347e-05, + "loss": 0.0106, + "step": 15180 + }, + { + "grad_norm": 0.10807356983423233, + "learning_rate": 1.5003053056480643e-05, + "loss": 0.0105, + "step": 15190 + }, + { + "grad_norm": 0.1063939779996872, + "learning_rate": 1.4944055293994551e-05, + "loss": 0.0116, + "step": 15200 + }, + { + "grad_norm": 0.0911322683095932, + "learning_rate": 1.4885153373122656e-05, + "loss": 0.0085, + "step": 15210 + }, + { + "grad_norm": 0.08222576975822449, + "learning_rate": 1.482634745490059e-05, + "loss": 0.01, + "step": 15220 + }, + { + "grad_norm": 0.10335645824670792, + "learning_rate": 1.4767637700101466e-05, + "loss": 0.0107, + "step": 15230 + }, + { + "grad_norm": 0.11196567863225937, + "learning_rate": 1.4709024269235528e-05, + "loss": 0.0109, + "step": 15240 + }, + { + "grad_norm": 0.10949188470840454, + "learning_rate": 1.4650507322549684e-05, + "loss": 0.011, + "step": 15250 + }, + { + "grad_norm": 0.07919946312904358, + "learning_rate": 1.4592087020026972e-05, + "loss": 0.01, + "step": 15260 + }, + { + "grad_norm": 0.09153652936220169, + "learning_rate": 1.4533763521386318e-05, + "loss": 0.0108, + "step": 15270 + }, + { + "grad_norm": 0.08978869020938873, + "learning_rate": 1.44755369860819e-05, + "loss": 0.0124, + "step": 15280 + }, + { + "grad_norm": 0.08744938671588898, + "learning_rate": 1.441740757330287e-05, + "loss": 0.0102, + "step": 15290 + }, + { + "grad_norm": 0.10775104910135269, + "learning_rate": 1.4359375441972844e-05, + "loss": 0.0112, + "step": 15300 + }, + { + "grad_norm": 0.07762865722179413, + "learning_rate": 1.4301440750749395e-05, + "loss": 0.0092, + "step": 15310 + }, + { + "grad_norm": 0.0667906403541565, + "learning_rate": 1.4243603658023808e-05, + "loss": 0.0075, + "step": 15320 + }, + { + "grad_norm": 0.08997557312250137, + "learning_rate": 1.4185864321920444e-05, + "loss": 0.0091, + "step": 15330 + }, + { + "grad_norm": 0.12192272394895554, + "learning_rate": 1.4128222900296485e-05, + "loss": 0.0112, + "step": 15340 + }, + { + "grad_norm": 0.10414731502532959, + "learning_rate": 1.407067955074135e-05, + "loss": 0.0094, + "step": 15350 + }, + { + "grad_norm": 0.11222764849662781, + "learning_rate": 1.4013234430576356e-05, + "loss": 0.0105, + "step": 15360 + }, + { + "grad_norm": 0.12625601887702942, + "learning_rate": 1.3955887696854286e-05, + "loss": 0.0117, + "step": 15370 + }, + { + "grad_norm": 0.10820724815130234, + "learning_rate": 1.38986395063589e-05, + "loss": 0.0088, + "step": 15380 + }, + { + "grad_norm": 0.06910783797502518, + "learning_rate": 1.3841490015604597e-05, + "loss": 0.0111, + "step": 15390 + }, + { + "grad_norm": 0.07153230160474777, + "learning_rate": 1.3784439380835879e-05, + "loss": 0.0105, + "step": 15400 + }, + { + "grad_norm": 0.12468352168798447, + "learning_rate": 1.3727487758026986e-05, + "loss": 0.0091, + "step": 15410 + }, + { + "grad_norm": 0.10580624639987946, + "learning_rate": 1.3670635302881525e-05, + "loss": 0.012, + "step": 15420 + }, + { + "grad_norm": 0.08015213161706924, + "learning_rate": 1.3613882170831888e-05, + "loss": 0.0099, + "step": 15430 + }, + { + "grad_norm": 0.10252169519662857, + "learning_rate": 1.355722851703901e-05, + "loss": 0.0116, + "step": 15440 + }, + { + "grad_norm": 0.12380716949701309, + "learning_rate": 1.3500674496391814e-05, + "loss": 0.0097, + "step": 15450 + }, + { + "grad_norm": 0.09673014283180237, + "learning_rate": 1.3444220263506795e-05, + "loss": 0.0083, + "step": 15460 + }, + { + "grad_norm": 0.09148132801055908, + "learning_rate": 1.3387865972727714e-05, + "loss": 0.0081, + "step": 15470 + }, + { + "grad_norm": 0.10601256787776947, + "learning_rate": 1.3331611778125036e-05, + "loss": 0.0117, + "step": 15480 + }, + { + "grad_norm": 0.1094682440161705, + "learning_rate": 1.3275457833495564e-05, + "loss": 0.0114, + "step": 15490 + }, + { + "grad_norm": 0.10962852090597153, + "learning_rate": 1.3219404292362065e-05, + "loss": 0.0126, + "step": 15500 + }, + { + "grad_norm": 0.10417924076318741, + "learning_rate": 1.3163451307972751e-05, + "loss": 0.0103, + "step": 15510 + }, + { + "grad_norm": 0.11326111108064651, + "learning_rate": 1.3107599033300977e-05, + "loss": 0.011, + "step": 15520 + }, + { + "grad_norm": 0.14387348294258118, + "learning_rate": 1.305184762104471e-05, + "loss": 0.0104, + "step": 15530 + }, + { + "grad_norm": 0.07867037504911423, + "learning_rate": 1.2996197223626178e-05, + "loss": 0.0098, + "step": 15540 + }, + { + "grad_norm": 0.09793301671743393, + "learning_rate": 1.2940647993191457e-05, + "loss": 0.0112, + "step": 15550 + }, + { + "grad_norm": 0.084642194211483, + "learning_rate": 1.2885200081610005e-05, + "loss": 0.0107, + "step": 15560 + }, + { + "grad_norm": 0.09194007515907288, + "learning_rate": 1.2829853640474316e-05, + "loss": 0.0121, + "step": 15570 + }, + { + "grad_norm": 0.07772580534219742, + "learning_rate": 1.2774608821099438e-05, + "loss": 0.0116, + "step": 15580 + }, + { + "grad_norm": 0.08480673283338547, + "learning_rate": 1.2719465774522577e-05, + "loss": 0.0094, + "step": 15590 + }, + { + "grad_norm": 0.08304253220558167, + "learning_rate": 1.2664424651502755e-05, + "loss": 0.0095, + "step": 15600 + }, + { + "grad_norm": 0.09991034865379333, + "learning_rate": 1.260948560252026e-05, + "loss": 0.0097, + "step": 15610 + }, + { + "grad_norm": 0.10061522573232651, + "learning_rate": 1.2554648777776396e-05, + "loss": 0.0087, + "step": 15620 + }, + { + "grad_norm": 0.07117509841918945, + "learning_rate": 1.2499914327192919e-05, + "loss": 0.0087, + "step": 15630 + }, + { + "grad_norm": 0.07055433094501495, + "learning_rate": 1.2445282400411722e-05, + "loss": 0.0101, + "step": 15640 + }, + { + "grad_norm": 0.09905356913805008, + "learning_rate": 1.2390753146794437e-05, + "loss": 0.0093, + "step": 15650 + }, + { + "grad_norm": 0.10582245141267776, + "learning_rate": 1.2336326715421925e-05, + "loss": 0.0096, + "step": 15660 + }, + { + "grad_norm": 0.13323351740837097, + "learning_rate": 1.2282003255094005e-05, + "loss": 0.0098, + "step": 15670 + }, + { + "grad_norm": 0.11265382915735245, + "learning_rate": 1.2227782914328928e-05, + "loss": 0.009, + "step": 15680 + }, + { + "grad_norm": 0.09909597784280777, + "learning_rate": 1.2173665841363018e-05, + "loss": 0.0092, + "step": 15690 + }, + { + "grad_norm": 0.08489767462015152, + "learning_rate": 1.211965218415032e-05, + "loss": 0.0095, + "step": 15700 + }, + { + "grad_norm": 0.09848932176828384, + "learning_rate": 1.2065742090362082e-05, + "loss": 0.0104, + "step": 15710 + }, + { + "grad_norm": 0.08533177524805069, + "learning_rate": 1.2011935707386457e-05, + "loss": 0.0092, + "step": 15720 + }, + { + "grad_norm": 0.057840920984745026, + "learning_rate": 1.1958233182328044e-05, + "loss": 0.0096, + "step": 15730 + }, + { + "grad_norm": 0.08860820531845093, + "learning_rate": 1.1904634662007474e-05, + "loss": 0.0103, + "step": 15740 + }, + { + "grad_norm": 0.16067075729370117, + "learning_rate": 1.1851140292961088e-05, + "loss": 0.0113, + "step": 15750 + }, + { + "grad_norm": 0.11732985079288483, + "learning_rate": 1.1797750221440424e-05, + "loss": 0.01, + "step": 15760 + }, + { + "grad_norm": 0.0913938358426094, + "learning_rate": 1.1744464593411897e-05, + "loss": 0.0106, + "step": 15770 + }, + { + "grad_norm": 0.12666913866996765, + "learning_rate": 1.1691283554556399e-05, + "loss": 0.0106, + "step": 15780 + }, + { + "grad_norm": 0.0712522566318512, + "learning_rate": 1.1638207250268834e-05, + "loss": 0.0108, + "step": 15790 + }, + { + "grad_norm": 0.10091933608055115, + "learning_rate": 1.158523582565782e-05, + "loss": 0.0098, + "step": 15800 + }, + { + "grad_norm": 0.07214750349521637, + "learning_rate": 1.1532369425545192e-05, + "loss": 0.0095, + "step": 15810 + }, + { + "grad_norm": 0.15435372292995453, + "learning_rate": 1.1479608194465662e-05, + "loss": 0.011, + "step": 15820 + }, + { + "grad_norm": 0.13618561625480652, + "learning_rate": 1.1426952276666442e-05, + "loss": 0.0113, + "step": 15830 + }, + { + "grad_norm": 0.1350582242012024, + "learning_rate": 1.1374401816106778e-05, + "loss": 0.0097, + "step": 15840 + }, + { + "grad_norm": 0.07411269098520279, + "learning_rate": 1.1321956956457646e-05, + "loss": 0.0109, + "step": 15850 + }, + { + "grad_norm": 0.09586109220981598, + "learning_rate": 1.1269617841101277e-05, + "loss": 0.0104, + "step": 15860 + }, + { + "grad_norm": 0.10644227266311646, + "learning_rate": 1.1217384613130804e-05, + "loss": 0.0093, + "step": 15870 + }, + { + "grad_norm": 0.09538836032152176, + "learning_rate": 1.11652574153499e-05, + "loss": 0.0095, + "step": 15880 + }, + { + "grad_norm": 0.1250450760126114, + "learning_rate": 1.1113236390272303e-05, + "loss": 0.0119, + "step": 15890 + }, + { + "grad_norm": 0.10659480839967728, + "learning_rate": 1.106132168012155e-05, + "loss": 0.0086, + "step": 15900 + }, + { + "grad_norm": 0.10112997144460678, + "learning_rate": 1.1009513426830448e-05, + "loss": 0.0097, + "step": 15910 + }, + { + "grad_norm": 0.1233617290854454, + "learning_rate": 1.0957811772040777e-05, + "loss": 0.0089, + "step": 15920 + }, + { + "grad_norm": 0.09904201328754425, + "learning_rate": 1.0906216857102913e-05, + "loss": 0.0099, + "step": 15930 + }, + { + "grad_norm": 0.10568784922361374, + "learning_rate": 1.0854728823075355e-05, + "loss": 0.0092, + "step": 15940 + }, + { + "grad_norm": 0.09177477657794952, + "learning_rate": 1.0803347810724452e-05, + "loss": 0.0109, + "step": 15950 + }, + { + "grad_norm": 0.09217534214258194, + "learning_rate": 1.0752073960523911e-05, + "loss": 0.0081, + "step": 15960 + }, + { + "grad_norm": 0.12068036198616028, + "learning_rate": 1.070090741265447e-05, + "loss": 0.0108, + "step": 15970 + }, + { + "grad_norm": 0.13883650302886963, + "learning_rate": 1.0649848307003547e-05, + "loss": 0.0113, + "step": 15980 + }, + { + "grad_norm": 0.10264614224433899, + "learning_rate": 1.0598896783164757e-05, + "loss": 0.0095, + "step": 15990 + }, + { + "grad_norm": 0.08448878675699234, + "learning_rate": 1.0548052980437645e-05, + "loss": 0.0092, + "step": 16000 + }, + { + "grad_norm": 0.09325902909040451, + "learning_rate": 1.049731703782722e-05, + "loss": 0.008, + "step": 16010 + }, + { + "grad_norm": 0.11123044043779373, + "learning_rate": 1.0446689094043587e-05, + "loss": 0.0104, + "step": 16020 + }, + { + "grad_norm": 0.08915429562330246, + "learning_rate": 1.039616928750165e-05, + "loss": 0.008, + "step": 16030 + }, + { + "grad_norm": 0.11341921985149384, + "learning_rate": 1.0345757756320612e-05, + "loss": 0.01, + "step": 16040 + }, + { + "grad_norm": 0.11879666149616241, + "learning_rate": 1.0295454638323666e-05, + "loss": 0.01, + "step": 16050 + }, + { + "grad_norm": 0.10207688808441162, + "learning_rate": 1.0245260071037632e-05, + "loss": 0.011, + "step": 16060 + }, + { + "grad_norm": 0.11344747245311737, + "learning_rate": 1.0195174191692518e-05, + "loss": 0.0084, + "step": 16070 + }, + { + "grad_norm": 0.05890040099620819, + "learning_rate": 1.014519713722124e-05, + "loss": 0.0102, + "step": 16080 + }, + { + "grad_norm": 0.10120263695716858, + "learning_rate": 1.0095329044259132e-05, + "loss": 0.0109, + "step": 16090 + }, + { + "grad_norm": 0.10087117552757263, + "learning_rate": 1.004557004914365e-05, + "loss": 0.0078, + "step": 16100 + }, + { + "grad_norm": 0.12072423845529556, + "learning_rate": 9.995920287914007e-06, + "loss": 0.0098, + "step": 16110 + }, + { + "grad_norm": 0.10275515168905258, + "learning_rate": 9.946379896310737e-06, + "loss": 0.0088, + "step": 16120 + }, + { + "grad_norm": 0.10975579172372818, + "learning_rate": 9.896949009775396e-06, + "loss": 0.0102, + "step": 16130 + }, + { + "grad_norm": 0.09533862769603729, + "learning_rate": 9.847627763450134e-06, + "loss": 0.009, + "step": 16140 + }, + { + "grad_norm": 0.08571784198284149, + "learning_rate": 9.798416292177337e-06, + "loss": 0.0075, + "step": 16150 + }, + { + "grad_norm": 0.08448456972837448, + "learning_rate": 9.74931473049932e-06, + "loss": 0.0089, + "step": 16160 + }, + { + "grad_norm": 0.06716963648796082, + "learning_rate": 9.700323212657847e-06, + "loss": 0.0107, + "step": 16170 + }, + { + "grad_norm": 0.11258897185325623, + "learning_rate": 9.65144187259388e-06, + "loss": 0.0092, + "step": 16180 + }, + { + "grad_norm": 0.15475143492221832, + "learning_rate": 9.602670843947132e-06, + "loss": 0.0095, + "step": 16190 + }, + { + "grad_norm": 0.11258062720298767, + "learning_rate": 9.554010260055713e-06, + "loss": 0.0115, + "step": 16200 + }, + { + "grad_norm": 0.10187021642923355, + "learning_rate": 9.505460253955834e-06, + "loss": 0.0089, + "step": 16210 + }, + { + "grad_norm": 0.10581238567829132, + "learning_rate": 9.457020958381324e-06, + "loss": 0.0102, + "step": 16220 + }, + { + "grad_norm": 0.0811745673418045, + "learning_rate": 9.408692505763395e-06, + "loss": 0.0078, + "step": 16230 + }, + { + "grad_norm": 0.07950105518102646, + "learning_rate": 9.360475028230181e-06, + "loss": 0.0087, + "step": 16240 + }, + { + "grad_norm": 0.09894289821386337, + "learning_rate": 9.312368657606412e-06, + "loss": 0.0109, + "step": 16250 + }, + { + "grad_norm": 0.10434219241142273, + "learning_rate": 9.264373525413096e-06, + "loss": 0.0095, + "step": 16260 + }, + { + "grad_norm": 0.11856606602668762, + "learning_rate": 9.216489762867058e-06, + "loss": 0.0112, + "step": 16270 + }, + { + "grad_norm": 0.09584639221429825, + "learning_rate": 9.168717500880708e-06, + "loss": 0.0091, + "step": 16280 + }, + { + "grad_norm": 0.0687025934457779, + "learning_rate": 9.121056870061574e-06, + "loss": 0.0112, + "step": 16290 + }, + { + "grad_norm": 0.08826236426830292, + "learning_rate": 9.073508000711983e-06, + "loss": 0.0107, + "step": 16300 + }, + { + "grad_norm": 0.1256347894668579, + "learning_rate": 9.026071022828758e-06, + "loss": 0.0098, + "step": 16310 + }, + { + "grad_norm": 0.09770868718624115, + "learning_rate": 8.978746066102771e-06, + "loss": 0.0088, + "step": 16320 + }, + { + "grad_norm": 0.09318437427282333, + "learning_rate": 8.931533259918634e-06, + "loss": 0.009, + "step": 16330 + }, + { + "grad_norm": 0.06524844467639923, + "learning_rate": 8.884432733354382e-06, + "loss": 0.0109, + "step": 16340 + }, + { + "grad_norm": 0.10596481710672379, + "learning_rate": 8.837444615181029e-06, + "loss": 0.0085, + "step": 16350 + }, + { + "grad_norm": 0.09532684087753296, + "learning_rate": 8.790569033862323e-06, + "loss": 0.0102, + "step": 16360 + }, + { + "grad_norm": 0.13039983808994293, + "learning_rate": 8.7438061175543e-06, + "loss": 0.0081, + "step": 16370 + }, + { + "grad_norm": 0.06694814562797546, + "learning_rate": 8.697155994104978e-06, + "loss": 0.0088, + "step": 16380 + }, + { + "grad_norm": 0.11142653226852417, + "learning_rate": 8.650618791054033e-06, + "loss": 0.0109, + "step": 16390 + }, + { + "grad_norm": 0.08565206080675125, + "learning_rate": 8.604194635632373e-06, + "loss": 0.0094, + "step": 16400 + }, + { + "grad_norm": 0.10072365403175354, + "learning_rate": 8.557883654761906e-06, + "loss": 0.0081, + "step": 16410 + }, + { + "grad_norm": 0.09272979199886322, + "learning_rate": 8.511685975055061e-06, + "loss": 0.0082, + "step": 16420 + }, + { + "grad_norm": 0.10406430810689926, + "learning_rate": 8.46560172281452e-06, + "loss": 0.0094, + "step": 16430 + }, + { + "grad_norm": 0.09129655361175537, + "learning_rate": 8.419631024032893e-06, + "loss": 0.0077, + "step": 16440 + }, + { + "grad_norm": 0.08670595288276672, + "learning_rate": 8.373774004392293e-06, + "loss": 0.0116, + "step": 16450 + }, + { + "grad_norm": 0.0978437066078186, + "learning_rate": 8.32803078926409e-06, + "loss": 0.0079, + "step": 16460 + }, + { + "grad_norm": 0.064898282289505, + "learning_rate": 8.282401503708454e-06, + "loss": 0.0076, + "step": 16470 + }, + { + "grad_norm": 0.14275825023651123, + "learning_rate": 8.23688627247412e-06, + "loss": 0.0106, + "step": 16480 + }, + { + "grad_norm": 0.12254878133535385, + "learning_rate": 8.191485219998007e-06, + "loss": 0.0114, + "step": 16490 + }, + { + "grad_norm": 0.11008410155773163, + "learning_rate": 8.146198470404843e-06, + "loss": 0.0079, + "step": 16500 + }, + { + "grad_norm": 0.11701522767543793, + "learning_rate": 8.101026147506897e-06, + "loss": 0.0101, + "step": 16510 + }, + { + "grad_norm": 0.09101291000843048, + "learning_rate": 8.05596837480353e-06, + "loss": 0.0079, + "step": 16520 + }, + { + "grad_norm": 0.08418968319892883, + "learning_rate": 8.011025275480998e-06, + "loss": 0.0084, + "step": 16530 + }, + { + "grad_norm": 0.09053727984428406, + "learning_rate": 7.966196972412027e-06, + "loss": 0.0079, + "step": 16540 + }, + { + "grad_norm": 0.10216031968593597, + "learning_rate": 7.92148358815547e-06, + "loss": 0.009, + "step": 16550 + }, + { + "grad_norm": 0.1010800153017044, + "learning_rate": 7.87688524495604e-06, + "loss": 0.0085, + "step": 16560 + }, + { + "grad_norm": 0.09330249577760696, + "learning_rate": 7.83240206474386e-06, + "loss": 0.0087, + "step": 16570 + }, + { + "grad_norm": 0.0894162654876709, + "learning_rate": 7.788034169134272e-06, + "loss": 0.0089, + "step": 16580 + }, + { + "grad_norm": 0.12389563769102097, + "learning_rate": 7.743781679427414e-06, + "loss": 0.0099, + "step": 16590 + }, + { + "grad_norm": 0.08647796511650085, + "learning_rate": 7.699644716607895e-06, + "loss": 0.0107, + "step": 16600 + }, + { + "grad_norm": 0.058433812111616135, + "learning_rate": 7.655623401344486e-06, + "loss": 0.0087, + "step": 16610 + }, + { + "grad_norm": 0.0737694576382637, + "learning_rate": 7.611717853989775e-06, + "loss": 0.0087, + "step": 16620 + }, + { + "grad_norm": 0.12006530910730362, + "learning_rate": 7.567928194579854e-06, + "loss": 0.0092, + "step": 16630 + }, + { + "grad_norm": 0.12729163467884064, + "learning_rate": 7.524254542833997e-06, + "loss": 0.0093, + "step": 16640 + }, + { + "grad_norm": 0.11596929281949997, + "learning_rate": 7.480697018154286e-06, + "loss": 0.0086, + "step": 16650 + }, + { + "grad_norm": 0.09019852429628372, + "learning_rate": 7.437255739625332e-06, + "loss": 0.0083, + "step": 16660 + }, + { + "grad_norm": 0.11755993962287903, + "learning_rate": 7.393930826013923e-06, + "loss": 0.0109, + "step": 16670 + }, + { + "grad_norm": 0.12654249370098114, + "learning_rate": 7.350722395768722e-06, + "loss": 0.0106, + "step": 16680 + }, + { + "grad_norm": 0.1370524764060974, + "learning_rate": 7.307630567019963e-06, + "loss": 0.011, + "step": 16690 + }, + { + "grad_norm": 0.11299826204776764, + "learning_rate": 7.264655457579e-06, + "loss": 0.0079, + "step": 16700 + }, + { + "grad_norm": 0.08943434804677963, + "learning_rate": 7.221797184938184e-06, + "loss": 0.0089, + "step": 16710 + }, + { + "grad_norm": 0.08322244882583618, + "learning_rate": 7.179055866270373e-06, + "loss": 0.0089, + "step": 16720 + }, + { + "grad_norm": 0.08710954338312149, + "learning_rate": 7.136431618428707e-06, + "loss": 0.0077, + "step": 16730 + }, + { + "grad_norm": 0.07177197188138962, + "learning_rate": 7.09392455794628e-06, + "loss": 0.0073, + "step": 16740 + }, + { + "grad_norm": 0.12535834312438965, + "learning_rate": 7.051534801035725e-06, + "loss": 0.0108, + "step": 16750 + }, + { + "grad_norm": 0.08635842055082321, + "learning_rate": 7.00926246358905e-06, + "loss": 0.0091, + "step": 16760 + }, + { + "grad_norm": 0.07984359562397003, + "learning_rate": 6.967107661177191e-06, + "loss": 0.0086, + "step": 16770 + }, + { + "grad_norm": 0.08598160743713379, + "learning_rate": 6.925070509049786e-06, + "loss": 0.0085, + "step": 16780 + }, + { + "grad_norm": 0.06658700853586197, + "learning_rate": 6.883151122134812e-06, + "loss": 0.0081, + "step": 16790 + }, + { + "grad_norm": 0.12147262692451477, + "learning_rate": 6.8413496150382394e-06, + "loss": 0.0093, + "step": 16800 + }, + { + "grad_norm": 0.14221180975437164, + "learning_rate": 6.7996661020438165e-06, + "loss": 0.0107, + "step": 16810 + }, + { + "grad_norm": 0.12473465502262115, + "learning_rate": 6.758100697112662e-06, + "loss": 0.0106, + "step": 16820 + }, + { + "grad_norm": 0.08993563055992126, + "learning_rate": 6.716653513883026e-06, + "loss": 0.0103, + "step": 16830 + }, + { + "grad_norm": 0.1235961988568306, + "learning_rate": 6.675324665669913e-06, + "loss": 0.0083, + "step": 16840 + }, + { + "grad_norm": 0.0859520435333252, + "learning_rate": 6.634114265464803e-06, + "loss": 0.0092, + "step": 16850 + }, + { + "grad_norm": 0.07648740708827972, + "learning_rate": 6.59302242593538e-06, + "loss": 0.0075, + "step": 16860 + }, + { + "grad_norm": 0.14654943346977234, + "learning_rate": 6.552049259425141e-06, + "loss": 0.0098, + "step": 16870 + }, + { + "grad_norm": 0.09150251001119614, + "learning_rate": 6.511194877953181e-06, + "loss": 0.0079, + "step": 16880 + }, + { + "grad_norm": 0.08237707614898682, + "learning_rate": 6.470459393213813e-06, + "loss": 0.0112, + "step": 16890 + }, + { + "grad_norm": 0.13840873539447784, + "learning_rate": 6.429842916576279e-06, + "loss": 0.0098, + "step": 16900 + }, + { + "grad_norm": 0.13403399288654327, + "learning_rate": 6.389345559084503e-06, + "loss": 0.0088, + "step": 16910 + }, + { + "grad_norm": 0.11110247671604156, + "learning_rate": 6.348967431456682e-06, + "loss": 0.0094, + "step": 16920 + }, + { + "grad_norm": 0.07588594406843185, + "learning_rate": 6.30870864408511e-06, + "loss": 0.0083, + "step": 16930 + }, + { + "grad_norm": 0.08848042041063309, + "learning_rate": 6.268569307035754e-06, + "loss": 0.0081, + "step": 16940 + }, + { + "grad_norm": 0.08812838047742844, + "learning_rate": 6.228549530048022e-06, + "loss": 0.008, + "step": 16950 + }, + { + "grad_norm": 0.08061771839857101, + "learning_rate": 6.1886494225344814e-06, + "loss": 0.0079, + "step": 16960 + }, + { + "grad_norm": 0.0747426301240921, + "learning_rate": 6.148869093580479e-06, + "loss": 0.0077, + "step": 16970 + }, + { + "grad_norm": 0.10628330707550049, + "learning_rate": 6.109208651943921e-06, + "loss": 0.0095, + "step": 16980 + }, + { + "grad_norm": 0.09798549115657806, + "learning_rate": 6.069668206054946e-06, + "loss": 0.0099, + "step": 16990 + }, + { + "grad_norm": 0.13766677677631378, + "learning_rate": 6.0302478640156145e-06, + "loss": 0.0103, + "step": 17000 + }, + { + "grad_norm": 0.10698185861110687, + "learning_rate": 5.990947733599644e-06, + "loss": 0.011, + "step": 17010 + }, + { + "grad_norm": 0.1182551309466362, + "learning_rate": 5.951767922252105e-06, + "loss": 0.0094, + "step": 17020 + }, + { + "grad_norm": 0.10787802934646606, + "learning_rate": 5.912708537089068e-06, + "loss": 0.0083, + "step": 17030 + }, + { + "grad_norm": 0.1129554957151413, + "learning_rate": 5.873769684897434e-06, + "loss": 0.0096, + "step": 17040 + }, + { + "grad_norm": 0.08732520788908005, + "learning_rate": 5.834951472134514e-06, + "loss": 0.0079, + "step": 17050 + }, + { + "grad_norm": 0.08840211480855942, + "learning_rate": 5.796254004927832e-06, + "loss": 0.0079, + "step": 17060 + }, + { + "grad_norm": 0.13193069398403168, + "learning_rate": 5.757677389074806e-06, + "loss": 0.0083, + "step": 17070 + }, + { + "grad_norm": 0.10598991066217422, + "learning_rate": 5.719221730042385e-06, + "loss": 0.0088, + "step": 17080 + }, + { + "grad_norm": 0.08788826316595078, + "learning_rate": 5.680887132966911e-06, + "loss": 0.0095, + "step": 17090 + }, + { + "grad_norm": 0.08623786270618439, + "learning_rate": 5.642673702653683e-06, + "loss": 0.0103, + "step": 17100 + }, + { + "grad_norm": 0.09824397414922714, + "learning_rate": 5.604581543576781e-06, + "loss": 0.0098, + "step": 17110 + }, + { + "grad_norm": 0.07001456618309021, + "learning_rate": 5.566610759878704e-06, + "loss": 0.0068, + "step": 17120 + }, + { + "grad_norm": 0.16719165444374084, + "learning_rate": 5.528761455370119e-06, + "loss": 0.008, + "step": 17130 + }, + { + "grad_norm": 0.12537311017513275, + "learning_rate": 5.491033733529594e-06, + "loss": 0.0088, + "step": 17140 + }, + { + "grad_norm": 0.08659331500530243, + "learning_rate": 5.453427697503255e-06, + "loss": 0.0083, + "step": 17150 + }, + { + "grad_norm": 0.09163559228181839, + "learning_rate": 5.415943450104599e-06, + "loss": 0.0096, + "step": 17160 + }, + { + "grad_norm": 0.09231021255254745, + "learning_rate": 5.378581093814111e-06, + "loss": 0.0094, + "step": 17170 + }, + { + "grad_norm": 0.10288187861442566, + "learning_rate": 5.3413407307790375e-06, + "loss": 0.0089, + "step": 17180 + }, + { + "grad_norm": 0.09350934624671936, + "learning_rate": 5.30422246281313e-06, + "loss": 0.0094, + "step": 17190 + }, + { + "grad_norm": 0.07568471878767014, + "learning_rate": 5.267226391396296e-06, + "loss": 0.0093, + "step": 17200 + }, + { + "grad_norm": 0.08966884762048721, + "learning_rate": 5.2303526176744e-06, + "loss": 0.0088, + "step": 17210 + }, + { + "grad_norm": 0.10326549410820007, + "learning_rate": 5.193601242458929e-06, + "loss": 0.012, + "step": 17220 + }, + { + "grad_norm": 0.1096484512090683, + "learning_rate": 5.156972366226714e-06, + "loss": 0.0088, + "step": 17230 + }, + { + "grad_norm": 0.10975171625614166, + "learning_rate": 5.120466089119735e-06, + "loss": 0.0085, + "step": 17240 + }, + { + "grad_norm": 0.09211853891611099, + "learning_rate": 5.084082510944749e-06, + "loss": 0.0083, + "step": 17250 + }, + { + "grad_norm": 0.08604861050844193, + "learning_rate": 5.047821731173058e-06, + "loss": 0.0084, + "step": 17260 + }, + { + "grad_norm": 0.05837802216410637, + "learning_rate": 5.011683848940274e-06, + "loss": 0.007, + "step": 17270 + }, + { + "grad_norm": 0.09806860983371735, + "learning_rate": 4.975668963045954e-06, + "loss": 0.0091, + "step": 17280 + }, + { + "grad_norm": 0.0689844861626625, + "learning_rate": 4.9397771719534525e-06, + "loss": 0.0072, + "step": 17290 + }, + { + "grad_norm": 0.07109055668115616, + "learning_rate": 4.904008573789548e-06, + "loss": 0.0096, + "step": 17300 + }, + { + "grad_norm": 0.11152077466249466, + "learning_rate": 4.8683632663442005e-06, + "loss": 0.0104, + "step": 17310 + }, + { + "grad_norm": 0.07792787998914719, + "learning_rate": 4.832841347070343e-06, + "loss": 0.0084, + "step": 17320 + }, + { + "grad_norm": 0.08081085979938507, + "learning_rate": 4.797442913083539e-06, + "loss": 0.0087, + "step": 17330 + }, + { + "grad_norm": 0.08944491297006607, + "learning_rate": 4.7621680611617596e-06, + "loss": 0.0065, + "step": 17340 + }, + { + "grad_norm": 0.11780090630054474, + "learning_rate": 4.727016887745095e-06, + "loss": 0.009, + "step": 17350 + }, + { + "grad_norm": 0.12969744205474854, + "learning_rate": 4.691989488935511e-06, + "loss": 0.0091, + "step": 17360 + }, + { + "grad_norm": 0.10791114717721939, + "learning_rate": 4.657085960496588e-06, + "loss": 0.0087, + "step": 17370 + }, + { + "grad_norm": 0.09812215715646744, + "learning_rate": 4.6223063978532265e-06, + "loss": 0.0085, + "step": 17380 + }, + { + "grad_norm": 0.15001660585403442, + "learning_rate": 4.587650896091439e-06, + "loss": 0.0107, + "step": 17390 + }, + { + "grad_norm": 0.11266276985406876, + "learning_rate": 4.553119549958035e-06, + "loss": 0.0115, + "step": 17400 + }, + { + "grad_norm": 0.10915940254926682, + "learning_rate": 4.518712453860385e-06, + "loss": 0.0084, + "step": 17410 + }, + { + "grad_norm": 0.07811658084392548, + "learning_rate": 4.484429701866205e-06, + "loss": 0.0074, + "step": 17420 + }, + { + "grad_norm": 0.10885101556777954, + "learning_rate": 4.4502713877031975e-06, + "loss": 0.0084, + "step": 17430 + }, + { + "grad_norm": 0.07058259844779968, + "learning_rate": 4.416237604758911e-06, + "loss": 0.0072, + "step": 17440 + }, + { + "grad_norm": 0.0840296521782875, + "learning_rate": 4.3823284460804025e-06, + "loss": 0.007, + "step": 17450 + }, + { + "grad_norm": 0.10486243665218353, + "learning_rate": 4.348544004374011e-06, + "loss": 0.0099, + "step": 17460 + }, + { + "grad_norm": 0.07803904265165329, + "learning_rate": 4.314884372005123e-06, + "loss": 0.0085, + "step": 17470 + }, + { + "grad_norm": 0.07243572175502777, + "learning_rate": 4.281349640997867e-06, + "loss": 0.0092, + "step": 17480 + }, + { + "grad_norm": 0.12921324372291565, + "learning_rate": 4.247939903034942e-06, + "loss": 0.0093, + "step": 17490 + }, + { + "grad_norm": 0.07304882258176804, + "learning_rate": 4.214655249457284e-06, + "loss": 0.0078, + "step": 17500 + }, + { + "grad_norm": 0.11342873424291611, + "learning_rate": 4.181495771263855e-06, + "loss": 0.0073, + "step": 17510 + }, + { + "grad_norm": 0.07445833832025528, + "learning_rate": 4.148461559111427e-06, + "loss": 0.008, + "step": 17520 + }, + { + "grad_norm": 0.047203417867422104, + "learning_rate": 4.115552703314252e-06, + "loss": 0.0087, + "step": 17530 + }, + { + "grad_norm": 0.1024160087108612, + "learning_rate": 4.082769293843886e-06, + "loss": 0.0089, + "step": 17540 + }, + { + "grad_norm": 0.11823720484972, + "learning_rate": 4.050111420328939e-06, + "loss": 0.0084, + "step": 17550 + }, + { + "grad_norm": 0.1030588150024414, + "learning_rate": 4.017579172054764e-06, + "loss": 0.0091, + "step": 17560 + }, + { + "grad_norm": 0.08110244572162628, + "learning_rate": 3.985172637963308e-06, + "loss": 0.0102, + "step": 17570 + }, + { + "grad_norm": 0.08997591584920883, + "learning_rate": 3.952891906652784e-06, + "loss": 0.009, + "step": 17580 + }, + { + "grad_norm": 0.09126926213502884, + "learning_rate": 3.920737066377478e-06, + "loss": 0.0083, + "step": 17590 + }, + { + "grad_norm": 0.07523714751005173, + "learning_rate": 3.888708205047509e-06, + "loss": 0.0098, + "step": 17600 + }, + { + "grad_norm": 0.07901276648044586, + "learning_rate": 3.856805410228542e-06, + "loss": 0.0076, + "step": 17610 + }, + { + "grad_norm": 0.09152697771787643, + "learning_rate": 3.82502876914162e-06, + "loss": 0.0068, + "step": 17620 + }, + { + "grad_norm": 0.13128548860549927, + "learning_rate": 3.7933783686628586e-06, + "loss": 0.0106, + "step": 17630 + }, + { + "grad_norm": 0.09366084635257721, + "learning_rate": 3.7618542953232306e-06, + "loss": 0.0084, + "step": 17640 + }, + { + "grad_norm": 0.07082492113113403, + "learning_rate": 3.7304566353083658e-06, + "loss": 0.008, + "step": 17650 + }, + { + "grad_norm": 0.09562533348798752, + "learning_rate": 3.6991854744582555e-06, + "loss": 0.0084, + "step": 17660 + }, + { + "grad_norm": 0.07046911865472794, + "learning_rate": 3.6680408982670777e-06, + "loss": 0.007, + "step": 17670 + }, + { + "grad_norm": 0.10544536262750626, + "learning_rate": 3.637022991882899e-06, + "loss": 0.0079, + "step": 17680 + }, + { + "grad_norm": 0.09704677015542984, + "learning_rate": 3.606131840107485e-06, + "loss": 0.0112, + "step": 17690 + }, + { + "grad_norm": 0.09045834094285965, + "learning_rate": 3.575367527396084e-06, + "loss": 0.0088, + "step": 17700 + }, + { + "grad_norm": 0.08584032952785492, + "learning_rate": 3.5447301378571386e-06, + "loss": 0.0078, + "step": 17710 + }, + { + "grad_norm": 0.1277010142803192, + "learning_rate": 3.514219755252113e-06, + "loss": 0.0084, + "step": 17720 + }, + { + "grad_norm": 0.07208079099655151, + "learning_rate": 3.4838364629952213e-06, + "loss": 0.008, + "step": 17730 + }, + { + "grad_norm": 0.0757041722536087, + "learning_rate": 3.4535803441532123e-06, + "loss": 0.0082, + "step": 17740 + }, + { + "grad_norm": 0.0726761743426323, + "learning_rate": 3.4234514814451836e-06, + "loss": 0.0073, + "step": 17750 + }, + { + "grad_norm": 0.08730766922235489, + "learning_rate": 3.393449957242273e-06, + "loss": 0.009, + "step": 17760 + }, + { + "grad_norm": 0.0633537620306015, + "learning_rate": 3.363575853567524e-06, + "loss": 0.0068, + "step": 17770 + }, + { + "grad_norm": 0.15596207976341248, + "learning_rate": 3.3338292520955826e-06, + "loss": 0.0075, + "step": 17780 + }, + { + "grad_norm": 0.0953638106584549, + "learning_rate": 3.304210234152516e-06, + "loss": 0.0082, + "step": 17790 + }, + { + "grad_norm": 0.09487409144639969, + "learning_rate": 3.2747188807155993e-06, + "loss": 0.0078, + "step": 17800 + }, + { + "grad_norm": 0.062127385288476944, + "learning_rate": 3.2453552724130643e-06, + "loss": 0.0082, + "step": 17810 + }, + { + "grad_norm": 0.09691774100065231, + "learning_rate": 3.216119489523889e-06, + "loss": 0.0095, + "step": 17820 + }, + { + "grad_norm": 0.09746340662240982, + "learning_rate": 3.1870116119775917e-06, + "loss": 0.0087, + "step": 17830 + }, + { + "grad_norm": 0.08026193827390671, + "learning_rate": 3.158031719353999e-06, + "loss": 0.0083, + "step": 17840 + }, + { + "grad_norm": 0.09739769250154495, + "learning_rate": 3.1291798908830273e-06, + "loss": 0.0095, + "step": 17850 + }, + { + "grad_norm": 0.08513207733631134, + "learning_rate": 3.1004562054444853e-06, + "loss": 0.0081, + "step": 17860 + }, + { + "grad_norm": 0.10273633897304535, + "learning_rate": 3.071860741567806e-06, + "loss": 0.0082, + "step": 17870 + }, + { + "grad_norm": 0.08400917798280716, + "learning_rate": 3.04339357743193e-06, + "loss": 0.0077, + "step": 17880 + }, + { + "grad_norm": 0.06768491864204407, + "learning_rate": 3.0150547908649628e-06, + "loss": 0.0086, + "step": 17890 + }, + { + "grad_norm": 0.09743641316890717, + "learning_rate": 2.9868444593440957e-06, + "loss": 0.0071, + "step": 17900 + }, + { + "grad_norm": 0.05410058796405792, + "learning_rate": 2.9587626599952846e-06, + "loss": 0.0072, + "step": 17910 + }, + { + "grad_norm": 0.1437024623155594, + "learning_rate": 2.930809469593082e-06, + "loss": 0.0102, + "step": 17920 + }, + { + "grad_norm": 0.10102169960737228, + "learning_rate": 2.9029849645604733e-06, + "loss": 0.0073, + "step": 17930 + }, + { + "grad_norm": 0.08796143531799316, + "learning_rate": 2.8752892209685632e-06, + "loss": 0.0083, + "step": 17940 + }, + { + "grad_norm": 0.10380808264017105, + "learning_rate": 2.847722314536483e-06, + "loss": 0.0076, + "step": 17950 + }, + { + "grad_norm": 0.07070038467645645, + "learning_rate": 2.820284320631078e-06, + "loss": 0.0071, + "step": 17960 + }, + { + "grad_norm": 0.07691735029220581, + "learning_rate": 2.792975314266788e-06, + "loss": 0.0087, + "step": 17970 + }, + { + "grad_norm": 0.07807575911283493, + "learning_rate": 2.7657953701054007e-06, + "loss": 0.0067, + "step": 17980 + }, + { + "grad_norm": 0.07985598593950272, + "learning_rate": 2.7387445624558306e-06, + "loss": 0.0067, + "step": 17990 + }, + { + "grad_norm": 0.09975605458021164, + "learning_rate": 2.7118229652739747e-06, + "loss": 0.0075, + "step": 18000 + }, + { + "grad_norm": 0.09057799726724625, + "learning_rate": 2.6850306521624236e-06, + "loss": 0.0078, + "step": 18010 + }, + { + "grad_norm": 0.0745016485452652, + "learning_rate": 2.6583676963703507e-06, + "loss": 0.0085, + "step": 18020 + }, + { + "grad_norm": 0.11493494361639023, + "learning_rate": 2.631834170793268e-06, + "loss": 0.0089, + "step": 18030 + }, + { + "grad_norm": 0.08134123682975769, + "learning_rate": 2.6054301479728036e-06, + "loss": 0.01, + "step": 18040 + }, + { + "grad_norm": 0.08473201841115952, + "learning_rate": 2.579155700096575e-06, + "loss": 0.009, + "step": 18050 + }, + { + "grad_norm": 0.06887421011924744, + "learning_rate": 2.5530108989978873e-06, + "loss": 0.0077, + "step": 18060 + }, + { + "grad_norm": 0.09340028464794159, + "learning_rate": 2.5269958161556416e-06, + "loss": 0.008, + "step": 18070 + }, + { + "grad_norm": 0.10058394819498062, + "learning_rate": 2.5011105226940888e-06, + "loss": 0.0075, + "step": 18080 + }, + { + "grad_norm": 0.11359404772520065, + "learning_rate": 2.4753550893826248e-06, + "loss": 0.0084, + "step": 18090 + }, + { + "grad_norm": 0.0884319394826889, + "learning_rate": 2.4497295866356296e-06, + "loss": 0.007, + "step": 18100 + }, + { + "grad_norm": 0.09588965028524399, + "learning_rate": 2.424234084512228e-06, + "loss": 0.0078, + "step": 18110 + }, + { + "grad_norm": 0.12072261422872543, + "learning_rate": 2.3988686527161687e-06, + "loss": 0.0087, + "step": 18120 + }, + { + "grad_norm": 0.09171867370605469, + "learning_rate": 2.373633360595573e-06, + "loss": 0.0079, + "step": 18130 + }, + { + "grad_norm": 0.06788451969623566, + "learning_rate": 2.3485282771427585e-06, + "loss": 0.0108, + "step": 18140 + }, + { + "grad_norm": 0.07304449379444122, + "learning_rate": 2.3235534709940665e-06, + "loss": 0.0078, + "step": 18150 + }, + { + "grad_norm": 0.05841727554798126, + "learning_rate": 2.2987090104296617e-06, + "loss": 0.0082, + "step": 18160 + }, + { + "grad_norm": 0.05940954387187958, + "learning_rate": 2.273994963373355e-06, + "loss": 0.0109, + "step": 18170 + }, + { + "grad_norm": 0.0758310854434967, + "learning_rate": 2.249411397392409e-06, + "loss": 0.0089, + "step": 18180 + }, + { + "grad_norm": 0.11268885433673859, + "learning_rate": 2.2249583796973506e-06, + "loss": 0.0087, + "step": 18190 + }, + { + "grad_norm": 0.06662435084581375, + "learning_rate": 2.200635977141796e-06, + "loss": 0.0065, + "step": 18200 + }, + { + "grad_norm": 0.09335529804229736, + "learning_rate": 2.17644425622226e-06, + "loss": 0.0072, + "step": 18210 + }, + { + "grad_norm": 0.09296659380197525, + "learning_rate": 2.152383283077991e-06, + "loss": 0.0089, + "step": 18220 + }, + { + "grad_norm": 0.07471774518489838, + "learning_rate": 2.128453123490781e-06, + "loss": 0.0073, + "step": 18230 + }, + { + "grad_norm": 0.13518674671649933, + "learning_rate": 2.1046538428847462e-06, + "loss": 0.0087, + "step": 18240 + }, + { + "grad_norm": 0.0745004266500473, + "learning_rate": 2.0809855063262273e-06, + "loss": 0.0088, + "step": 18250 + }, + { + "grad_norm": 0.09193255007266998, + "learning_rate": 2.057448178523558e-06, + "loss": 0.0074, + "step": 18260 + }, + { + "grad_norm": 0.07844199985265732, + "learning_rate": 2.034041923826885e-06, + "loss": 0.0082, + "step": 18270 + }, + { + "grad_norm": 0.09172441810369492, + "learning_rate": 2.0107668062280204e-06, + "loss": 0.009, + "step": 18280 + }, + { + "grad_norm": 0.07932326197624207, + "learning_rate": 1.9876228893602357e-06, + "loss": 0.0081, + "step": 18290 + }, + { + "grad_norm": 0.08357483893632889, + "learning_rate": 1.9646102364981266e-06, + "loss": 0.0086, + "step": 18300 + }, + { + "grad_norm": 0.0705217719078064, + "learning_rate": 1.9417289105574053e-06, + "loss": 0.0091, + "step": 18310 + }, + { + "grad_norm": 0.06657382845878601, + "learning_rate": 1.9189789740947427e-06, + "loss": 0.0082, + "step": 18320 + }, + { + "grad_norm": 0.11797548830509186, + "learning_rate": 1.896360489307597e-06, + "loss": 0.0092, + "step": 18330 + }, + { + "grad_norm": 0.07438908517360687, + "learning_rate": 1.8738735180340362e-06, + "loss": 0.0081, + "step": 18340 + }, + { + "grad_norm": 0.06076909229159355, + "learning_rate": 1.8515181217525824e-06, + "loss": 0.007, + "step": 18350 + }, + { + "grad_norm": 0.062138210982084274, + "learning_rate": 1.8292943615820457e-06, + "loss": 0.0084, + "step": 18360 + }, + { + "grad_norm": 0.07920871675014496, + "learning_rate": 1.8072022982813296e-06, + "loss": 0.009, + "step": 18370 + }, + { + "grad_norm": 0.08604487031698227, + "learning_rate": 1.7852419922492925e-06, + "loss": 0.0088, + "step": 18380 + }, + { + "grad_norm": 0.08762957155704498, + "learning_rate": 1.763413503524569e-06, + "loss": 0.0075, + "step": 18390 + }, + { + "grad_norm": 0.0878095030784607, + "learning_rate": 1.7417168917854165e-06, + "loss": 0.0086, + "step": 18400 + }, + { + "grad_norm": 0.07590161263942719, + "learning_rate": 1.720152216349552e-06, + "loss": 0.0104, + "step": 18410 + }, + { + "grad_norm": 0.07685063779354095, + "learning_rate": 1.6987195361739595e-06, + "loss": 0.0078, + "step": 18420 + }, + { + "grad_norm": 0.11251430958509445, + "learning_rate": 1.6774189098547832e-06, + "loss": 0.0096, + "step": 18430 + }, + { + "grad_norm": 0.08727696537971497, + "learning_rate": 1.6562503956271069e-06, + "loss": 0.0065, + "step": 18440 + }, + { + "grad_norm": 0.12324109673500061, + "learning_rate": 1.6352140513648417e-06, + "loss": 0.0083, + "step": 18450 + }, + { + "grad_norm": 0.09162585437297821, + "learning_rate": 1.6143099345805712e-06, + "loss": 0.0084, + "step": 18460 + }, + { + "grad_norm": 0.07423223555088043, + "learning_rate": 1.5935381024253293e-06, + "loss": 0.0074, + "step": 18470 + }, + { + "grad_norm": 0.09696771949529648, + "learning_rate": 1.572898611688517e-06, + "loss": 0.0085, + "step": 18480 + }, + { + "grad_norm": 0.05576842650771141, + "learning_rate": 1.5523915187977133e-06, + "loss": 0.0066, + "step": 18490 + }, + { + "grad_norm": 0.08985871821641922, + "learning_rate": 1.532016879818532e-06, + "loss": 0.0073, + "step": 18500 + }, + { + "grad_norm": 0.09337108582258224, + "learning_rate": 1.51177475045447e-06, + "loss": 0.0086, + "step": 18510 + }, + { + "grad_norm": 0.09257137030363083, + "learning_rate": 1.4916651860467035e-06, + "loss": 0.0098, + "step": 18520 + }, + { + "grad_norm": 0.08665672689676285, + "learning_rate": 1.471688241574043e-06, + "loss": 0.0071, + "step": 18530 + }, + { + "grad_norm": 0.11311785876750946, + "learning_rate": 1.451843971652672e-06, + "loss": 0.0085, + "step": 18540 + }, + { + "grad_norm": 0.09265606850385666, + "learning_rate": 1.432132430536076e-06, + "loss": 0.0082, + "step": 18550 + }, + { + "grad_norm": 0.07042396813631058, + "learning_rate": 1.412553672114869e-06, + "loss": 0.0073, + "step": 18560 + }, + { + "grad_norm": 0.06209905445575714, + "learning_rate": 1.3931077499166056e-06, + "loss": 0.0085, + "step": 18570 + }, + { + "grad_norm": 0.03796195238828659, + "learning_rate": 1.3737947171057085e-06, + "loss": 0.0066, + "step": 18580 + }, + { + "grad_norm": 0.06222688406705856, + "learning_rate": 1.3546146264832582e-06, + "loss": 0.0086, + "step": 18590 + }, + { + "grad_norm": 0.09720593690872192, + "learning_rate": 1.3355675304869086e-06, + "loss": 0.008, + "step": 18600 + }, + { + "grad_norm": 0.07502838224172592, + "learning_rate": 1.3166534811906827e-06, + "loss": 0.0097, + "step": 18610 + }, + { + "grad_norm": 0.10019935667514801, + "learning_rate": 1.2978725303048666e-06, + "loss": 0.0074, + "step": 18620 + }, + { + "grad_norm": 0.08895155042409897, + "learning_rate": 1.2792247291758762e-06, + "loss": 0.0077, + "step": 18630 + }, + { + "grad_norm": 0.07851336896419525, + "learning_rate": 1.2607101287860635e-06, + "loss": 0.0079, + "step": 18640 + }, + { + "grad_norm": 0.10427725315093994, + "learning_rate": 1.2423287797536654e-06, + "loss": 0.0073, + "step": 18650 + }, + { + "grad_norm": 0.07401717454195023, + "learning_rate": 1.2240807323325776e-06, + "loss": 0.0073, + "step": 18660 + }, + { + "grad_norm": 0.08853729814291, + "learning_rate": 1.205966036412254e-06, + "loss": 0.0059, + "step": 18670 + }, + { + "grad_norm": 0.09655343741178513, + "learning_rate": 1.1879847415175949e-06, + "loss": 0.0087, + "step": 18680 + }, + { + "grad_norm": 0.07455024123191833, + "learning_rate": 1.1701368968087712e-06, + "loss": 0.0085, + "step": 18690 + }, + { + "grad_norm": 0.08170092105865479, + "learning_rate": 1.1524225510811116e-06, + "loss": 0.0082, + "step": 18700 + }, + { + "grad_norm": 0.09190782159566879, + "learning_rate": 1.1348417527649535e-06, + "loss": 0.0075, + "step": 18710 + }, + { + "grad_norm": 0.10425201058387756, + "learning_rate": 1.1173945499255268e-06, + "loss": 0.0084, + "step": 18720 + }, + { + "grad_norm": 0.0772399827837944, + "learning_rate": 1.1000809902628307e-06, + "loss": 0.007, + "step": 18730 + }, + { + "grad_norm": 0.07126295566558838, + "learning_rate": 1.082901121111468e-06, + "loss": 0.0101, + "step": 18740 + }, + { + "grad_norm": 0.08075746148824692, + "learning_rate": 1.0658549894405456e-06, + "loss": 0.0081, + "step": 18750 + }, + { + "grad_norm": 0.07402067631483078, + "learning_rate": 1.0489426418535342e-06, + "loss": 0.0072, + "step": 18760 + }, + { + "grad_norm": 0.06745612621307373, + "learning_rate": 1.0321641245881474e-06, + "loss": 0.0067, + "step": 18770 + }, + { + "grad_norm": 0.07484298944473267, + "learning_rate": 1.015519483516214e-06, + "loss": 0.0085, + "step": 18780 + }, + { + "grad_norm": 0.08844636380672455, + "learning_rate": 9.990087641435443e-07, + "loss": 0.0094, + "step": 18790 + }, + { + "grad_norm": 0.08883892744779587, + "learning_rate": 9.826320116098132e-07, + "loss": 0.009, + "step": 18800 + }, + { + "grad_norm": 0.11147583276033401, + "learning_rate": 9.663892706884447e-07, + "loss": 0.0077, + "step": 18810 + }, + { + "grad_norm": 0.08384449779987335, + "learning_rate": 9.502805857864616e-07, + "loss": 0.0079, + "step": 18820 + }, + { + "grad_norm": 0.1048063188791275, + "learning_rate": 9.34306000944396e-07, + "loss": 0.0068, + "step": 18830 + }, + { + "grad_norm": 0.09604956954717636, + "learning_rate": 9.184655598361624e-07, + "loss": 0.0082, + "step": 18840 + }, + { + "grad_norm": 0.11281401664018631, + "learning_rate": 9.027593057689076e-07, + "loss": 0.0083, + "step": 18850 + }, + { + "grad_norm": 0.04877932369709015, + "learning_rate": 8.871872816829441e-07, + "loss": 0.0078, + "step": 18860 + }, + { + "grad_norm": 0.07863446325063705, + "learning_rate": 8.717495301515777e-07, + "loss": 0.0072, + "step": 18870 + }, + { + "grad_norm": 0.07426071912050247, + "learning_rate": 8.564460933810415e-07, + "loss": 0.0077, + "step": 18880 + }, + { + "grad_norm": 0.09083437919616699, + "learning_rate": 8.412770132103453e-07, + "loss": 0.0078, + "step": 18890 + }, + { + "grad_norm": 0.08448003232479095, + "learning_rate": 8.262423311111711e-07, + "loss": 0.0076, + "step": 18900 + }, + { + "grad_norm": 0.2685222625732422, + "learning_rate": 8.113420881877665e-07, + "loss": 0.0083, + "step": 18910 + }, + { + "grad_norm": 0.09670383483171463, + "learning_rate": 7.965763251768288e-07, + "loss": 0.0094, + "step": 18920 + }, + { + "grad_norm": 0.07262090593576431, + "learning_rate": 7.819450824473995e-07, + "loss": 0.0078, + "step": 18930 + }, + { + "grad_norm": 0.07297039031982422, + "learning_rate": 7.674484000007198e-07, + "loss": 0.0065, + "step": 18940 + }, + { + "grad_norm": 0.10263441503047943, + "learning_rate": 7.530863174701752e-07, + "loss": 0.0102, + "step": 18950 + }, + { + "grad_norm": 0.09818470478057861, + "learning_rate": 7.38858874121151e-07, + "loss": 0.0077, + "step": 18960 + }, + { + "grad_norm": 0.053600430488586426, + "learning_rate": 7.247661088509328e-07, + "loss": 0.0082, + "step": 18970 + }, + { + "grad_norm": 0.09788445383310318, + "learning_rate": 7.108080601886002e-07, + "loss": 0.0094, + "step": 18980 + }, + { + "grad_norm": 0.13908198475837708, + "learning_rate": 6.969847662949336e-07, + "loss": 0.0082, + "step": 18990 + }, + { + "grad_norm": 0.08889535069465637, + "learning_rate": 6.832962649622798e-07, + "loss": 0.0078, + "step": 19000 + }, + { + "grad_norm": 0.06614526361227036, + "learning_rate": 6.697425936144863e-07, + "loss": 0.0075, + "step": 19010 + }, + { + "grad_norm": 0.09962449967861176, + "learning_rate": 6.563237893067731e-07, + "loss": 0.0064, + "step": 19020 + }, + { + "grad_norm": 0.07745740562677383, + "learning_rate": 6.430398887256328e-07, + "loss": 0.0095, + "step": 19030 + }, + { + "grad_norm": 0.1160113513469696, + "learning_rate": 6.298909281887478e-07, + "loss": 0.0089, + "step": 19040 + }, + { + "grad_norm": 0.08460378646850586, + "learning_rate": 6.168769436448673e-07, + "loss": 0.0089, + "step": 19050 + }, + { + "grad_norm": 0.11441825330257416, + "learning_rate": 6.03997970673742e-07, + "loss": 0.0089, + "step": 19060 + }, + { + "grad_norm": 0.09307875484228134, + "learning_rate": 5.912540444859782e-07, + "loss": 0.0089, + "step": 19070 + }, + { + "grad_norm": 0.055086757987737656, + "learning_rate": 5.786451999229837e-07, + "loss": 0.0086, + "step": 19080 + }, + { + "grad_norm": 0.12894394993782043, + "learning_rate": 5.661714714568722e-07, + "loss": 0.0081, + "step": 19090 + }, + { + "grad_norm": 0.07145680487155914, + "learning_rate": 5.538328931903259e-07, + "loss": 0.0096, + "step": 19100 + }, + { + "grad_norm": 0.06126881018280983, + "learning_rate": 5.416294988565551e-07, + "loss": 0.0111, + "step": 19110 + }, + { + "grad_norm": 0.08967404067516327, + "learning_rate": 5.29561321819172e-07, + "loss": 0.0076, + "step": 19120 + }, + { + "grad_norm": 0.11542192101478577, + "learning_rate": 5.176283950721061e-07, + "loss": 0.009, + "step": 19130 + }, + { + "grad_norm": 0.1454596221446991, + "learning_rate": 5.058307512395332e-07, + "loss": 0.0105, + "step": 19140 + }, + { + "grad_norm": 0.0667036771774292, + "learning_rate": 4.941684225757526e-07, + "loss": 0.0081, + "step": 19150 + }, + { + "grad_norm": 0.0938059389591217, + "learning_rate": 4.826414409651314e-07, + "loss": 0.008, + "step": 19160 + }, + { + "grad_norm": 0.07873714715242386, + "learning_rate": 4.712498379219943e-07, + "loss": 0.0064, + "step": 19170 + }, + { + "grad_norm": 0.1003437414765358, + "learning_rate": 4.599936445905506e-07, + "loss": 0.0076, + "step": 19180 + }, + { + "grad_norm": 0.1182265654206276, + "learning_rate": 4.4887289174480594e-07, + "loss": 0.0102, + "step": 19190 + }, + { + "grad_norm": 0.06163545325398445, + "learning_rate": 4.378876097884621e-07, + "loss": 0.0104, + "step": 19200 + }, + { + "grad_norm": 0.07294057309627533, + "learning_rate": 4.2703782875487264e-07, + "loss": 0.007, + "step": 19210 + }, + { + "grad_norm": 0.08125778287649155, + "learning_rate": 4.163235783069208e-07, + "loss": 0.008, + "step": 19220 + }, + { + "grad_norm": 0.0974658951163292, + "learning_rate": 4.057448877369585e-07, + "loss": 0.0096, + "step": 19230 + }, + { + "grad_norm": 0.08408542722463608, + "learning_rate": 3.9530178596672295e-07, + "loss": 0.0073, + "step": 19240 + }, + { + "grad_norm": 0.08791803568601608, + "learning_rate": 3.849943015472479e-07, + "loss": 0.0089, + "step": 19250 + }, + { + "grad_norm": 0.06575709581375122, + "learning_rate": 3.748224626588137e-07, + "loss": 0.0087, + "step": 19260 + }, + { + "grad_norm": 0.0887456014752388, + "learning_rate": 3.647862971108307e-07, + "loss": 0.0064, + "step": 19270 + }, + { + "grad_norm": 0.13375157117843628, + "learning_rate": 3.5488583234179473e-07, + "loss": 0.0094, + "step": 19280 + }, + { + "grad_norm": 0.08476066589355469, + "learning_rate": 3.4512109541920413e-07, + "loss": 0.0076, + "step": 19290 + }, + { + "grad_norm": 0.12561450898647308, + "learning_rate": 3.354921130394706e-07, + "loss": 0.0086, + "step": 19300 + }, + { + "grad_norm": 0.06015969067811966, + "learning_rate": 3.259989115278639e-07, + "loss": 0.0085, + "step": 19310 + }, + { + "grad_norm": 0.081431083381176, + "learning_rate": 3.1664151683843403e-07, + "loss": 0.0092, + "step": 19320 + }, + { + "grad_norm": 0.08083204180002213, + "learning_rate": 3.074199545539447e-07, + "loss": 0.0076, + "step": 19330 + }, + { + "grad_norm": 0.10698910802602768, + "learning_rate": 2.983342498857955e-07, + "loss": 0.0085, + "step": 19340 + }, + { + "grad_norm": 0.0598200224339962, + "learning_rate": 2.893844276739499e-07, + "loss": 0.0079, + "step": 19350 + }, + { + "grad_norm": 0.0912698283791542, + "learning_rate": 2.8057051238688514e-07, + "loss": 0.0076, + "step": 19360 + }, + { + "grad_norm": 0.11089014261960983, + "learning_rate": 2.71892528121509e-07, + "loss": 0.0097, + "step": 19370 + }, + { + "grad_norm": 0.06108818203210831, + "learning_rate": 2.633504986030988e-07, + "loss": 0.0069, + "step": 19380 + }, + { + "grad_norm": 0.17292183637619019, + "learning_rate": 2.549444471852347e-07, + "loss": 0.0093, + "step": 19390 + }, + { + "grad_norm": 0.0797327309846878, + "learning_rate": 2.4667439684974423e-07, + "loss": 0.0075, + "step": 19400 + }, + { + "grad_norm": 0.09049943089485168, + "learning_rate": 2.3854037020662467e-07, + "loss": 0.0099, + "step": 19410 + }, + { + "grad_norm": 0.07103713601827621, + "learning_rate": 2.3054238949399288e-07, + "loss": 0.0087, + "step": 19420 + }, + { + "grad_norm": 0.0830058604478836, + "learning_rate": 2.2268047657802993e-07, + "loss": 0.0065, + "step": 19430 + }, + { + "grad_norm": 0.09744752198457718, + "learning_rate": 2.149546529529034e-07, + "loss": 0.0093, + "step": 19440 + }, + { + "grad_norm": 0.08256909251213074, + "learning_rate": 2.0736493974071736e-07, + "loss": 0.0063, + "step": 19450 + }, + { + "grad_norm": 0.06695324182510376, + "learning_rate": 1.9991135769145686e-07, + "loss": 0.0071, + "step": 19460 + }, + { + "grad_norm": 0.111125148832798, + "learning_rate": 1.9259392718293245e-07, + "loss": 0.008, + "step": 19470 + }, + { + "grad_norm": 0.09279434382915497, + "learning_rate": 1.8541266822072467e-07, + "loss": 0.0078, + "step": 19480 + }, + { + "grad_norm": 0.07287320494651794, + "learning_rate": 1.7836760043811184e-07, + "loss": 0.0072, + "step": 19490 + }, + { + "grad_norm": 0.06699864566326141, + "learning_rate": 1.7145874309604792e-07, + "loss": 0.0069, + "step": 19500 + }, + { + "grad_norm": 0.08563697338104248, + "learning_rate": 1.6468611508308474e-07, + "loss": 0.0106, + "step": 19510 + }, + { + "grad_norm": 0.10029300302267075, + "learning_rate": 1.5804973491532204e-07, + "loss": 0.0068, + "step": 19520 + }, + { + "grad_norm": 0.08745329082012177, + "learning_rate": 1.5154962073637424e-07, + "loss": 0.0077, + "step": 19530 + }, + { + "grad_norm": 0.0749882236123085, + "learning_rate": 1.4518579031730372e-07, + "loss": 0.0068, + "step": 19540 + }, + { + "grad_norm": 0.11291056126356125, + "learning_rate": 1.389582610565876e-07, + "loss": 0.0079, + "step": 19550 + }, + { + "grad_norm": 0.06588900834321976, + "learning_rate": 1.3286704998003995e-07, + "loss": 0.0089, + "step": 19560 + }, + { + "grad_norm": 0.06938900053501129, + "learning_rate": 1.2691217374080632e-07, + "loss": 0.0086, + "step": 19570 + }, + { + "grad_norm": 0.10236141085624695, + "learning_rate": 1.2109364861929705e-07, + "loss": 0.0076, + "step": 19580 + }, + { + "grad_norm": 0.05483406037092209, + "learning_rate": 1.1541149052312628e-07, + "loss": 0.0057, + "step": 19590 + }, + { + "grad_norm": 0.07891669869422913, + "learning_rate": 1.0986571498710074e-07, + "loss": 0.0078, + "step": 19600 + }, + { + "grad_norm": 0.046337854117155075, + "learning_rate": 1.0445633717316438e-07, + "loss": 0.0079, + "step": 19610 + }, + { + "grad_norm": 0.06285147368907928, + "learning_rate": 9.918337187034277e-08, + "loss": 0.0073, + "step": 19620 + }, + { + "grad_norm": 0.07573024928569794, + "learning_rate": 9.404683349472643e-08, + "loss": 0.0091, + "step": 19630 + }, + { + "grad_norm": 0.08258235454559326, + "learning_rate": 8.904673608940983e-08, + "loss": 0.0067, + "step": 19640 + }, + { + "grad_norm": 0.09584877640008926, + "learning_rate": 8.418309332447471e-08, + "loss": 0.0074, + "step": 19650 + }, + { + "grad_norm": 0.0916801169514656, + "learning_rate": 7.945591849692902e-08, + "loss": 0.0101, + "step": 19660 + }, + { + "grad_norm": 0.10481783002614975, + "learning_rate": 7.486522453069578e-08, + "loss": 0.0106, + "step": 19670 + }, + { + "grad_norm": 0.09190323203802109, + "learning_rate": 7.041102397655208e-08, + "loss": 0.0074, + "step": 19680 + }, + { + "grad_norm": 0.12194039672613144, + "learning_rate": 6.609332901210685e-08, + "loss": 0.0079, + "step": 19690 + }, + { + "grad_norm": 0.08534438908100128, + "learning_rate": 6.191215144178419e-08, + "loss": 0.0074, + "step": 19700 + }, + { + "grad_norm": 0.030948689207434654, + "learning_rate": 5.786750269675678e-08, + "loss": 0.0073, + "step": 19710 + }, + { + "grad_norm": 0.08273988217115402, + "learning_rate": 5.395939383494031e-08, + "loss": 0.0073, + "step": 19720 + }, + { + "grad_norm": 0.0580388680100441, + "learning_rate": 5.018783554095463e-08, + "loss": 0.0072, + "step": 19730 + }, + { + "grad_norm": 0.08530529588460922, + "learning_rate": 4.655283812610156e-08, + "loss": 0.0081, + "step": 19740 + }, + { + "grad_norm": 0.0832894966006279, + "learning_rate": 4.305441152831491e-08, + "loss": 0.009, + "step": 19750 + }, + { + "grad_norm": 0.07896947860717773, + "learning_rate": 3.9692565312171584e-08, + "loss": 0.0091, + "step": 19760 + }, + { + "grad_norm": 0.08613897860050201, + "learning_rate": 3.6467308668824975e-08, + "loss": 0.0074, + "step": 19770 + }, + { + "grad_norm": 0.08991219848394394, + "learning_rate": 3.3378650416004964e-08, + "loss": 0.008, + "step": 19780 + }, + { + "grad_norm": 0.05563666298985481, + "learning_rate": 3.042659899797906e-08, + "loss": 0.0082, + "step": 19790 + }, + { + "grad_norm": 0.11001113802194595, + "learning_rate": 2.76111624855524e-08, + "loss": 0.0089, + "step": 19800 + }, + { + "grad_norm": 0.07874969393014908, + "learning_rate": 2.4932348576017784e-08, + "loss": 0.0088, + "step": 19810 + }, + { + "grad_norm": 0.09826279431581497, + "learning_rate": 2.239016459314458e-08, + "loss": 0.0079, + "step": 19820 + }, + { + "grad_norm": 0.09135234355926514, + "learning_rate": 1.9984617487173174e-08, + "loss": 0.01, + "step": 19830 + }, + { + "grad_norm": 0.06302394717931747, + "learning_rate": 1.7715713834776105e-08, + "loss": 0.0079, + "step": 19840 + }, + { + "grad_norm": 0.07780909538269043, + "learning_rate": 1.5583459839046964e-08, + "loss": 0.0076, + "step": 19850 + }, + { + "grad_norm": 0.09754228591918945, + "learning_rate": 1.3587861329489304e-08, + "loss": 0.008, + "step": 19860 + }, + { + "grad_norm": 0.07660409808158875, + "learning_rate": 1.1728923761994415e-08, + "loss": 0.0082, + "step": 19870 + }, + { + "grad_norm": 0.09123298525810242, + "learning_rate": 1.0006652218819135e-08, + "loss": 0.0084, + "step": 19880 + }, + { + "grad_norm": 0.11355910450220108, + "learning_rate": 8.421051408596947e-09, + "loss": 0.0085, + "step": 19890 + }, + { + "grad_norm": 0.06773153692483902, + "learning_rate": 6.972125666299123e-09, + "loss": 0.0068, + "step": 19900 + }, + { + "grad_norm": 0.08153941482305527, + "learning_rate": 5.659878953229169e-09, + "loss": 0.0082, + "step": 19910 + }, + { + "grad_norm": 0.078322634100914, + "learning_rate": 4.48431485701728e-09, + "loss": 0.0085, + "step": 19920 + }, + { + "grad_norm": 0.10496822744607925, + "learning_rate": 3.4454365916203322e-09, + "loss": 0.0115, + "step": 19930 + }, + { + "grad_norm": 0.05381660535931587, + "learning_rate": 2.5432469972830332e-09, + "loss": 0.0084, + "step": 19940 + }, + { + "grad_norm": 0.047633007168769836, + "learning_rate": 1.7777485405601203e-09, + "loss": 0.0076, + "step": 19950 + }, + { + "grad_norm": 0.11699728667736053, + "learning_rate": 1.1489433142941597e-09, + "loss": 0.0105, + "step": 19960 + }, + { + "grad_norm": 0.09142055362462997, + "learning_rate": 6.568330376210963e-10, + "loss": 0.0067, + "step": 19970 + }, + { + "grad_norm": 0.05864863842725754, + "learning_rate": 3.0141905594249787e-10, + "loss": 0.0075, + "step": 19980 + }, + { + "grad_norm": 0.08082906901836395, + "learning_rate": 8.270234094776008e-11, + "loss": 0.0072, + "step": 19990 + }, + { + "grad_norm": 0.06887168437242508, + "learning_rate": 6.834906085551041e-13, + "loss": 0.0081, + "step": 20000 + } + ], + "logging_steps": 10, + "max_steps": 20000, + "num_input_tokens_seen": 0, + "num_train_epochs": 9223372036854775807, + "save_steps": 5000, + "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": 64, + "trial_name": null, + "trial_params": null +}