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": 4.4084038734436035, + "learning_rate": 9e-07, + "loss": 1.3469, + "step": 10 + }, + { + "grad_norm": 2.854828119277954, + "learning_rate": 1.9e-06, + "loss": 1.3141, + "step": 20 + }, + { + "grad_norm": 1.38284432888031, + "learning_rate": 2.9e-06, + "loss": 1.2086, + "step": 30 + }, + { + "grad_norm": 0.6830054521560669, + "learning_rate": 3.9e-06, + "loss": 1.1273, + "step": 40 + }, + { + "grad_norm": 0.3296288847923279, + "learning_rate": 4.9000000000000005e-06, + "loss": 1.0875, + "step": 50 + }, + { + "grad_norm": 0.3340240716934204, + "learning_rate": 5.9e-06, + "loss": 1.075, + "step": 60 + }, + { + "grad_norm": 0.2849108874797821, + "learning_rate": 6.900000000000001e-06, + "loss": 1.0703, + "step": 70 + }, + { + "grad_norm": 0.31928372383117676, + "learning_rate": 7.9e-06, + "loss": 1.0582, + "step": 80 + }, + { + "grad_norm": 0.3880952298641205, + "learning_rate": 8.9e-06, + "loss": 1.0555, + "step": 90 + }, + { + "grad_norm": 0.30215585231781006, + "learning_rate": 9.900000000000002e-06, + "loss": 1.048, + "step": 100 + }, + { + "grad_norm": 0.43695512413978577, + "learning_rate": 1.09e-05, + "loss": 1.043, + "step": 110 + }, + { + "grad_norm": 0.4662553668022156, + "learning_rate": 1.19e-05, + "loss": 1.0406, + "step": 120 + }, + { + "grad_norm": 0.7485678195953369, + "learning_rate": 1.29e-05, + "loss": 1.0172, + "step": 130 + }, + { + "grad_norm": 0.7978800535202026, + "learning_rate": 1.3900000000000002e-05, + "loss": 1.0072, + "step": 140 + }, + { + "grad_norm": 0.7703062295913696, + "learning_rate": 1.49e-05, + "loss": 0.9766, + "step": 150 + }, + { + "grad_norm": 1.0931109189987183, + "learning_rate": 1.59e-05, + "loss": 0.9051, + "step": 160 + }, + { + "grad_norm": 1.3214606046676636, + "learning_rate": 1.69e-05, + "loss": 0.8164, + "step": 170 + }, + { + "grad_norm": 1.6509478092193604, + "learning_rate": 1.79e-05, + "loss": 0.7283, + "step": 180 + }, + { + "grad_norm": 1.649338722229004, + "learning_rate": 1.8900000000000002e-05, + "loss": 0.6459, + "step": 190 + }, + { + "grad_norm": 1.5137853622436523, + "learning_rate": 1.9900000000000003e-05, + "loss": 0.5609, + "step": 200 + }, + { + "grad_norm": 1.676445722579956, + "learning_rate": 2.09e-05, + "loss": 0.5043, + "step": 210 + }, + { + "grad_norm": 1.5472042560577393, + "learning_rate": 2.19e-05, + "loss": 0.4341, + "step": 220 + }, + { + "grad_norm": 2.0228428840637207, + "learning_rate": 2.29e-05, + "loss": 0.38, + "step": 230 + }, + { + "grad_norm": 2.0979020595550537, + "learning_rate": 2.39e-05, + "loss": 0.3448, + "step": 240 + }, + { + "grad_norm": 1.9566534757614136, + "learning_rate": 2.4900000000000002e-05, + "loss": 0.3183, + "step": 250 + }, + { + "grad_norm": 1.620675802230835, + "learning_rate": 2.5900000000000003e-05, + "loss": 0.2805, + "step": 260 + }, + { + "grad_norm": 2.3286774158477783, + "learning_rate": 2.6900000000000003e-05, + "loss": 0.2455, + "step": 270 + }, + { + "grad_norm": 1.9605939388275146, + "learning_rate": 2.7900000000000004e-05, + "loss": 0.2306, + "step": 280 + }, + { + "grad_norm": 2.058300733566284, + "learning_rate": 2.8899999999999998e-05, + "loss": 0.2136, + "step": 290 + }, + { + "grad_norm": 2.3196494579315186, + "learning_rate": 2.9900000000000002e-05, + "loss": 0.1968, + "step": 300 + }, + { + "grad_norm": 2.2974038124084473, + "learning_rate": 3.09e-05, + "loss": 0.1902, + "step": 310 + }, + { + "grad_norm": 2.038285732269287, + "learning_rate": 3.19e-05, + "loss": 0.161, + "step": 320 + }, + { + "grad_norm": 1.532688856124878, + "learning_rate": 3.29e-05, + "loss": 0.1528, + "step": 330 + }, + { + "grad_norm": 2.092379331588745, + "learning_rate": 3.3900000000000004e-05, + "loss": 0.143, + "step": 340 + }, + { + "grad_norm": 2.0367798805236816, + "learning_rate": 3.49e-05, + "loss": 0.1339, + "step": 350 + }, + { + "grad_norm": 2.3606696128845215, + "learning_rate": 3.59e-05, + "loss": 0.1321, + "step": 360 + }, + { + "grad_norm": 1.742854118347168, + "learning_rate": 3.69e-05, + "loss": 0.1226, + "step": 370 + }, + { + "grad_norm": 2.3252644538879395, + "learning_rate": 3.79e-05, + "loss": 0.1165, + "step": 380 + }, + { + "grad_norm": 2.4858038425445557, + "learning_rate": 3.8900000000000004e-05, + "loss": 0.1215, + "step": 390 + }, + { + "grad_norm": 2.094243049621582, + "learning_rate": 3.99e-05, + "loss": 0.1252, + "step": 400 + }, + { + "grad_norm": 1.9110815525054932, + "learning_rate": 4.09e-05, + "loss": 0.1219, + "step": 410 + }, + { + "grad_norm": 2.131819725036621, + "learning_rate": 4.19e-05, + "loss": 0.1087, + "step": 420 + }, + { + "grad_norm": 2.0112192630767822, + "learning_rate": 4.29e-05, + "loss": 0.1132, + "step": 430 + }, + { + "grad_norm": 2.2110557556152344, + "learning_rate": 4.39e-05, + "loss": 0.1113, + "step": 440 + }, + { + "grad_norm": 2.171856164932251, + "learning_rate": 4.49e-05, + "loss": 0.1198, + "step": 450 + }, + { + "grad_norm": 1.8911032676696777, + "learning_rate": 4.5900000000000004e-05, + "loss": 0.1119, + "step": 460 + }, + { + "grad_norm": 2.165710926055908, + "learning_rate": 4.69e-05, + "loss": 0.1167, + "step": 470 + }, + { + "grad_norm": 2.065707206726074, + "learning_rate": 4.79e-05, + "loss": 0.1074, + "step": 480 + }, + { + "grad_norm": 1.9315952062606812, + "learning_rate": 4.89e-05, + "loss": 0.1016, + "step": 490 + }, + { + "grad_norm": 2.0874738693237305, + "learning_rate": 4.99e-05, + "loss": 0.0999, + "step": 500 + }, + { + "grad_norm": 1.8715001344680786, + "learning_rate": 5.0900000000000004e-05, + "loss": 0.0998, + "step": 510 + }, + { + "grad_norm": 1.6038424968719482, + "learning_rate": 5.19e-05, + "loss": 0.0966, + "step": 520 + }, + { + "grad_norm": 1.9660725593566895, + "learning_rate": 5.2900000000000005e-05, + "loss": 0.0977, + "step": 530 + }, + { + "grad_norm": 1.8537325859069824, + "learning_rate": 5.390000000000001e-05, + "loss": 0.1056, + "step": 540 + }, + { + "grad_norm": 1.6274349689483643, + "learning_rate": 5.4900000000000006e-05, + "loss": 0.094, + "step": 550 + }, + { + "grad_norm": 1.6068252325057983, + "learning_rate": 5.590000000000001e-05, + "loss": 0.1028, + "step": 560 + }, + { + "grad_norm": 1.5526419878005981, + "learning_rate": 5.69e-05, + "loss": 0.1021, + "step": 570 + }, + { + "grad_norm": 1.9871398210525513, + "learning_rate": 5.79e-05, + "loss": 0.0982, + "step": 580 + }, + { + "grad_norm": 1.8751955032348633, + "learning_rate": 5.89e-05, + "loss": 0.0894, + "step": 590 + }, + { + "grad_norm": 1.8565336465835571, + "learning_rate": 5.99e-05, + "loss": 0.0947, + "step": 600 + }, + { + "grad_norm": 1.7967112064361572, + "learning_rate": 6.09e-05, + "loss": 0.0876, + "step": 610 + }, + { + "grad_norm": 1.6690943241119385, + "learning_rate": 6.19e-05, + "loss": 0.0907, + "step": 620 + }, + { + "grad_norm": 1.8858799934387207, + "learning_rate": 6.29e-05, + "loss": 0.0909, + "step": 630 + }, + { + "grad_norm": 1.6637696027755737, + "learning_rate": 6.390000000000001e-05, + "loss": 0.0874, + "step": 640 + }, + { + "grad_norm": 1.7460479736328125, + "learning_rate": 6.49e-05, + "loss": 0.09, + "step": 650 + }, + { + "grad_norm": 1.7727495431900024, + "learning_rate": 6.59e-05, + "loss": 0.092, + "step": 660 + }, + { + "grad_norm": 1.4950904846191406, + "learning_rate": 6.690000000000001e-05, + "loss": 0.0923, + "step": 670 + }, + { + "grad_norm": 1.44473397731781, + "learning_rate": 6.790000000000001e-05, + "loss": 0.0891, + "step": 680 + }, + { + "grad_norm": 1.5126936435699463, + "learning_rate": 6.89e-05, + "loss": 0.0874, + "step": 690 + }, + { + "grad_norm": 1.394876480102539, + "learning_rate": 6.99e-05, + "loss": 0.0849, + "step": 700 + }, + { + "grad_norm": 1.681939721107483, + "learning_rate": 7.09e-05, + "loss": 0.0966, + "step": 710 + }, + { + "grad_norm": 1.6944371461868286, + "learning_rate": 7.19e-05, + "loss": 0.0864, + "step": 720 + }, + { + "grad_norm": 1.6403831243515015, + "learning_rate": 7.29e-05, + "loss": 0.0764, + "step": 730 + }, + { + "grad_norm": 1.59284245967865, + "learning_rate": 7.390000000000001e-05, + "loss": 0.0874, + "step": 740 + }, + { + "grad_norm": 1.707362174987793, + "learning_rate": 7.49e-05, + "loss": 0.0837, + "step": 750 + }, + { + "grad_norm": 1.5193947553634644, + "learning_rate": 7.59e-05, + "loss": 0.0939, + "step": 760 + }, + { + "grad_norm": 1.5924936532974243, + "learning_rate": 7.69e-05, + "loss": 0.0903, + "step": 770 + }, + { + "grad_norm": 1.5755540132522583, + "learning_rate": 7.790000000000001e-05, + "loss": 0.0854, + "step": 780 + }, + { + "grad_norm": 1.28225576877594, + "learning_rate": 7.890000000000001e-05, + "loss": 0.0902, + "step": 790 + }, + { + "grad_norm": 1.3005540370941162, + "learning_rate": 7.99e-05, + "loss": 0.0885, + "step": 800 + }, + { + "grad_norm": 1.4294734001159668, + "learning_rate": 8.090000000000001e-05, + "loss": 0.084, + "step": 810 + }, + { + "grad_norm": 1.44032883644104, + "learning_rate": 8.19e-05, + "loss": 0.0816, + "step": 820 + }, + { + "grad_norm": 1.2144129276275635, + "learning_rate": 8.29e-05, + "loss": 0.0845, + "step": 830 + }, + { + "grad_norm": 1.4295352697372437, + "learning_rate": 8.39e-05, + "loss": 0.0838, + "step": 840 + }, + { + "grad_norm": 1.4729937314987183, + "learning_rate": 8.49e-05, + "loss": 0.0849, + "step": 850 + }, + { + "grad_norm": 1.2805378437042236, + "learning_rate": 8.59e-05, + "loss": 0.0899, + "step": 860 + }, + { + "grad_norm": 1.2634493112564087, + "learning_rate": 8.69e-05, + "loss": 0.0815, + "step": 870 + }, + { + "grad_norm": 1.3772571086883545, + "learning_rate": 8.790000000000001e-05, + "loss": 0.0818, + "step": 880 + }, + { + "grad_norm": 1.173386573791504, + "learning_rate": 8.89e-05, + "loss": 0.0792, + "step": 890 + }, + { + "grad_norm": 1.140397548675537, + "learning_rate": 8.99e-05, + "loss": 0.0738, + "step": 900 + }, + { + "grad_norm": 1.2523930072784424, + "learning_rate": 9.090000000000001e-05, + "loss": 0.0728, + "step": 910 + }, + { + "grad_norm": 1.2446190118789673, + "learning_rate": 9.190000000000001e-05, + "loss": 0.0733, + "step": 920 + }, + { + "grad_norm": 1.123772144317627, + "learning_rate": 9.290000000000001e-05, + "loss": 0.0752, + "step": 930 + }, + { + "grad_norm": 1.2798773050308228, + "learning_rate": 9.39e-05, + "loss": 0.0718, + "step": 940 + }, + { + "grad_norm": 1.1665390729904175, + "learning_rate": 9.49e-05, + "loss": 0.08, + "step": 950 + }, + { + "grad_norm": 1.1190357208251953, + "learning_rate": 9.59e-05, + "loss": 0.0757, + "step": 960 + }, + { + "grad_norm": 1.172494888305664, + "learning_rate": 9.69e-05, + "loss": 0.0768, + "step": 970 + }, + { + "grad_norm": 1.2636840343475342, + "learning_rate": 9.790000000000001e-05, + "loss": 0.0777, + "step": 980 + }, + { + "grad_norm": 1.0782413482666016, + "learning_rate": 9.89e-05, + "loss": 0.0781, + "step": 990 + }, + { + "grad_norm": 1.1838284730911255, + "learning_rate": 9.99e-05, + "loss": 0.0818, + "step": 1000 + }, + { + "grad_norm": 1.178709626197815, + "learning_rate": 9.999994463727085e-05, + "loss": 0.0745, + "step": 1010 + }, + { + "grad_norm": 1.0647282600402832, + "learning_rate": 9.999975326009292e-05, + "loss": 0.0764, + "step": 1020 + }, + { + "grad_norm": 1.127692461013794, + "learning_rate": 9.999942518549879e-05, + "loss": 0.0814, + "step": 1030 + }, + { + "grad_norm": 1.0108965635299683, + "learning_rate": 9.999896041438544e-05, + "loss": 0.0792, + "step": 1040 + }, + { + "grad_norm": 0.9025714993476868, + "learning_rate": 9.999835894802353e-05, + "loss": 0.0722, + "step": 1050 + }, + { + "grad_norm": 1.064939022064209, + "learning_rate": 9.999762078805743e-05, + "loss": 0.0717, + "step": 1060 + }, + { + "grad_norm": 1.1429728269577026, + "learning_rate": 9.999674593650526e-05, + "loss": 0.069, + "step": 1070 + }, + { + "grad_norm": 1.0148142576217651, + "learning_rate": 9.99957343957588e-05, + "loss": 0.0643, + "step": 1080 + }, + { + "grad_norm": 0.9264249205589294, + "learning_rate": 9.99945861685836e-05, + "loss": 0.0679, + "step": 1090 + }, + { + "grad_norm": 1.1930052042007446, + "learning_rate": 9.999330125811884e-05, + "loss": 0.0737, + "step": 1100 + }, + { + "grad_norm": 1.004508137702942, + "learning_rate": 9.999187966787744e-05, + "loss": 0.0643, + "step": 1110 + }, + { + "grad_norm": 0.9657332897186279, + "learning_rate": 9.999032140174595e-05, + "loss": 0.0749, + "step": 1120 + }, + { + "grad_norm": 0.8985906839370728, + "learning_rate": 9.998862646398464e-05, + "loss": 0.0746, + "step": 1130 + }, + { + "grad_norm": 1.174508810043335, + "learning_rate": 9.998679485922739e-05, + "loss": 0.068, + "step": 1140 + }, + { + "grad_norm": 1.1320881843566895, + "learning_rate": 9.998482659248174e-05, + "loss": 0.0749, + "step": 1150 + }, + { + "grad_norm": 0.8892739415168762, + "learning_rate": 9.998272166912883e-05, + "loss": 0.065, + "step": 1160 + }, + { + "grad_norm": 1.0125502347946167, + "learning_rate": 9.998048009492347e-05, + "loss": 0.0676, + "step": 1170 + }, + { + "grad_norm": 0.9409289956092834, + "learning_rate": 9.997810187599403e-05, + "loss": 0.0691, + "step": 1180 + }, + { + "grad_norm": 1.0810723304748535, + "learning_rate": 9.997558701884249e-05, + "loss": 0.0672, + "step": 1190 + }, + { + "grad_norm": 0.9851813316345215, + "learning_rate": 9.997293553034433e-05, + "loss": 0.0711, + "step": 1200 + }, + { + "grad_norm": 0.9922870993614197, + "learning_rate": 9.997014741774866e-05, + "loss": 0.0646, + "step": 1210 + }, + { + "grad_norm": 0.8688313364982605, + "learning_rate": 9.996722268867803e-05, + "loss": 0.0598, + "step": 1220 + }, + { + "grad_norm": 0.7204305529594421, + "learning_rate": 9.996416135112858e-05, + "loss": 0.061, + "step": 1230 + }, + { + "grad_norm": 0.9210626482963562, + "learning_rate": 9.996096341346988e-05, + "loss": 0.0651, + "step": 1240 + }, + { + "grad_norm": 0.894849419593811, + "learning_rate": 9.995762888444495e-05, + "loss": 0.0677, + "step": 1250 + }, + { + "grad_norm": 0.8269813656806946, + "learning_rate": 9.995415777317027e-05, + "loss": 0.0586, + "step": 1260 + }, + { + "grad_norm": 0.705187976360321, + "learning_rate": 9.995055008913574e-05, + "loss": 0.0716, + "step": 1270 + }, + { + "grad_norm": 0.9853078722953796, + "learning_rate": 9.994680584220463e-05, + "loss": 0.0643, + "step": 1280 + }, + { + "grad_norm": 0.8162582516670227, + "learning_rate": 9.994292504261355e-05, + "loss": 0.0662, + "step": 1290 + }, + { + "grad_norm": 0.975835919380188, + "learning_rate": 9.993890770097247e-05, + "loss": 0.0635, + "step": 1300 + }, + { + "grad_norm": 1.0302761793136597, + "learning_rate": 9.993475382826467e-05, + "loss": 0.0572, + "step": 1310 + }, + { + "grad_norm": 0.7187492251396179, + "learning_rate": 9.993046343584664e-05, + "loss": 0.0697, + "step": 1320 + }, + { + "grad_norm": 0.7942295670509338, + "learning_rate": 9.992603653544816e-05, + "loss": 0.0577, + "step": 1330 + }, + { + "grad_norm": 0.8199202418327332, + "learning_rate": 9.992147313917222e-05, + "loss": 0.0529, + "step": 1340 + }, + { + "grad_norm": 0.8115194439888, + "learning_rate": 9.991677325949497e-05, + "loss": 0.0624, + "step": 1350 + }, + { + "grad_norm": 0.6720649600028992, + "learning_rate": 9.991193690926568e-05, + "loss": 0.0646, + "step": 1360 + }, + { + "grad_norm": 0.9936930537223816, + "learning_rate": 9.990696410170678e-05, + "loss": 0.0564, + "step": 1370 + }, + { + "grad_norm": 0.7325035929679871, + "learning_rate": 9.990185485041371e-05, + "loss": 0.0564, + "step": 1380 + }, + { + "grad_norm": 0.7739620804786682, + "learning_rate": 9.989660916935498e-05, + "loss": 0.0522, + "step": 1390 + }, + { + "grad_norm": 0.6889415979385376, + "learning_rate": 9.989122707287208e-05, + "loss": 0.0494, + "step": 1400 + }, + { + "grad_norm": 0.7498676776885986, + "learning_rate": 9.988570857567945e-05, + "loss": 0.0511, + "step": 1410 + }, + { + "grad_norm": 0.7862585783004761, + "learning_rate": 9.988005369286446e-05, + "loss": 0.0514, + "step": 1420 + }, + { + "grad_norm": 0.86005699634552, + "learning_rate": 9.987426243988734e-05, + "loss": 0.0594, + "step": 1430 + }, + { + "grad_norm": 0.7341459393501282, + "learning_rate": 9.986833483258114e-05, + "loss": 0.0524, + "step": 1440 + }, + { + "grad_norm": 0.6387414336204529, + "learning_rate": 9.986227088715173e-05, + "loss": 0.0502, + "step": 1450 + }, + { + "grad_norm": 0.6452542543411255, + "learning_rate": 9.98560706201777e-05, + "loss": 0.0552, + "step": 1460 + }, + { + "grad_norm": 0.6964207291603088, + "learning_rate": 9.984973404861036e-05, + "loss": 0.053, + "step": 1470 + }, + { + "grad_norm": 0.9187918305397034, + "learning_rate": 9.984326118977361e-05, + "loss": 0.0523, + "step": 1480 + }, + { + "grad_norm": 0.5924795269966125, + "learning_rate": 9.983665206136406e-05, + "loss": 0.0572, + "step": 1490 + }, + { + "grad_norm": 0.7473819851875305, + "learning_rate": 9.982990668145075e-05, + "loss": 0.0536, + "step": 1500 + }, + { + "grad_norm": 0.7378273606300354, + "learning_rate": 9.982302506847534e-05, + "loss": 0.0563, + "step": 1510 + }, + { + "grad_norm": 0.7275499105453491, + "learning_rate": 9.981600724125189e-05, + "loss": 0.054, + "step": 1520 + }, + { + "grad_norm": 0.8484076261520386, + "learning_rate": 9.980885321896685e-05, + "loss": 0.0517, + "step": 1530 + }, + { + "grad_norm": 0.7533107995986938, + "learning_rate": 9.980156302117905e-05, + "loss": 0.0555, + "step": 1540 + }, + { + "grad_norm": 0.7630091905593872, + "learning_rate": 9.979413666781963e-05, + "loss": 0.0537, + "step": 1550 + }, + { + "grad_norm": 0.6897077560424805, + "learning_rate": 9.978657417919193e-05, + "loss": 0.0524, + "step": 1560 + }, + { + "grad_norm": 0.8400923013687134, + "learning_rate": 9.977887557597153e-05, + "loss": 0.0539, + "step": 1570 + }, + { + "grad_norm": 0.9766692519187927, + "learning_rate": 9.97710408792061e-05, + "loss": 0.0511, + "step": 1580 + }, + { + "grad_norm": 0.8205817341804504, + "learning_rate": 9.976307011031542e-05, + "loss": 0.0521, + "step": 1590 + }, + { + "grad_norm": 0.8941313624382019, + "learning_rate": 9.975496329109126e-05, + "loss": 0.0572, + "step": 1600 + }, + { + "grad_norm": 0.68316650390625, + "learning_rate": 9.974672044369732e-05, + "loss": 0.0495, + "step": 1610 + }, + { + "grad_norm": 0.7722634077072144, + "learning_rate": 9.97383415906693e-05, + "loss": 0.048, + "step": 1620 + }, + { + "grad_norm": 0.7082585692405701, + "learning_rate": 9.97298267549146e-05, + "loss": 0.0501, + "step": 1630 + }, + { + "grad_norm": 0.9164435863494873, + "learning_rate": 9.972117595971249e-05, + "loss": 0.0517, + "step": 1640 + }, + { + "grad_norm": 0.6809993386268616, + "learning_rate": 9.971238922871391e-05, + "loss": 0.0471, + "step": 1650 + }, + { + "grad_norm": 0.5702314376831055, + "learning_rate": 9.970346658594142e-05, + "loss": 0.0449, + "step": 1660 + }, + { + "grad_norm": 0.773526668548584, + "learning_rate": 9.969440805578923e-05, + "loss": 0.0498, + "step": 1670 + }, + { + "grad_norm": 0.7197460532188416, + "learning_rate": 9.968521366302298e-05, + "loss": 0.0467, + "step": 1680 + }, + { + "grad_norm": 0.6625840067863464, + "learning_rate": 9.967588343277981e-05, + "loss": 0.045, + "step": 1690 + }, + { + "grad_norm": 0.8736388087272644, + "learning_rate": 9.966641739056818e-05, + "loss": 0.0545, + "step": 1700 + }, + { + "grad_norm": 0.750792384147644, + "learning_rate": 9.965681556226793e-05, + "loss": 0.0473, + "step": 1710 + }, + { + "grad_norm": 0.786637008190155, + "learning_rate": 9.964707797413006e-05, + "loss": 0.0507, + "step": 1720 + }, + { + "grad_norm": 0.6412456631660461, + "learning_rate": 9.963720465277679e-05, + "loss": 0.0487, + "step": 1730 + }, + { + "grad_norm": 0.7418937087059021, + "learning_rate": 9.96271956252014e-05, + "loss": 0.0466, + "step": 1740 + }, + { + "grad_norm": 0.654487669467926, + "learning_rate": 9.961705091876816e-05, + "loss": 0.0437, + "step": 1750 + }, + { + "grad_norm": 0.5268855690956116, + "learning_rate": 9.960677056121235e-05, + "loss": 0.0425, + "step": 1760 + }, + { + "grad_norm": 0.6897957921028137, + "learning_rate": 9.959635458064005e-05, + "loss": 0.0524, + "step": 1770 + }, + { + "grad_norm": 0.7622124552726746, + "learning_rate": 9.958580300552815e-05, + "loss": 0.0486, + "step": 1780 + }, + { + "grad_norm": 0.6903027296066284, + "learning_rate": 9.957511586472426e-05, + "loss": 0.0484, + "step": 1790 + }, + { + "grad_norm": 0.6610746383666992, + "learning_rate": 9.956429318744662e-05, + "loss": 0.0501, + "step": 1800 + }, + { + "grad_norm": 0.6257076263427734, + "learning_rate": 9.955333500328404e-05, + "loss": 0.0461, + "step": 1810 + }, + { + "grad_norm": 0.6198020577430725, + "learning_rate": 9.95422413421957e-05, + "loss": 0.0471, + "step": 1820 + }, + { + "grad_norm": 0.5837280750274658, + "learning_rate": 9.953101223451133e-05, + "loss": 0.0458, + "step": 1830 + }, + { + "grad_norm": 0.5286070704460144, + "learning_rate": 9.951964771093085e-05, + "loss": 0.0439, + "step": 1840 + }, + { + "grad_norm": 0.700628399848938, + "learning_rate": 9.950814780252442e-05, + "loss": 0.0449, + "step": 1850 + }, + { + "grad_norm": 0.659744918346405, + "learning_rate": 9.949651254073236e-05, + "loss": 0.0422, + "step": 1860 + }, + { + "grad_norm": 0.6888746619224548, + "learning_rate": 9.948474195736504e-05, + "loss": 0.0437, + "step": 1870 + }, + { + "grad_norm": 0.7431710958480835, + "learning_rate": 9.947283608460277e-05, + "loss": 0.0444, + "step": 1880 + }, + { + "grad_norm": 0.5975768566131592, + "learning_rate": 9.946079495499577e-05, + "loss": 0.0428, + "step": 1890 + }, + { + "grad_norm": 0.7167385816574097, + "learning_rate": 9.944861860146401e-05, + "loss": 0.0478, + "step": 1900 + }, + { + "grad_norm": 0.8195943236351013, + "learning_rate": 9.943630705729719e-05, + "loss": 0.0462, + "step": 1910 + }, + { + "grad_norm": 0.6716237664222717, + "learning_rate": 9.942386035615459e-05, + "loss": 0.0403, + "step": 1920 + }, + { + "grad_norm": 0.609624981880188, + "learning_rate": 9.941127853206503e-05, + "loss": 0.0417, + "step": 1930 + }, + { + "grad_norm": 0.5453461408615112, + "learning_rate": 9.939856161942673e-05, + "loss": 0.0513, + "step": 1940 + }, + { + "grad_norm": 0.6557628512382507, + "learning_rate": 9.938570965300724e-05, + "loss": 0.0426, + "step": 1950 + }, + { + "grad_norm": 0.7052960991859436, + "learning_rate": 9.937272266794335e-05, + "loss": 0.0427, + "step": 1960 + }, + { + "grad_norm": 0.7427675724029541, + "learning_rate": 9.935960069974096e-05, + "loss": 0.0437, + "step": 1970 + }, + { + "grad_norm": 0.7904492616653442, + "learning_rate": 9.934634378427506e-05, + "loss": 0.0474, + "step": 1980 + }, + { + "grad_norm": 0.6603449583053589, + "learning_rate": 9.933295195778954e-05, + "loss": 0.041, + "step": 1990 + }, + { + "grad_norm": 0.6774188876152039, + "learning_rate": 9.931942525689715e-05, + "loss": 0.0419, + "step": 2000 + }, + { + "grad_norm": 0.8625200390815735, + "learning_rate": 9.930576371857936e-05, + "loss": 0.0433, + "step": 2010 + }, + { + "grad_norm": 0.6208633184432983, + "learning_rate": 9.929196738018629e-05, + "loss": 0.0434, + "step": 2020 + }, + { + "grad_norm": 0.7620671391487122, + "learning_rate": 9.927803627943662e-05, + "loss": 0.0396, + "step": 2030 + }, + { + "grad_norm": 0.7727156281471252, + "learning_rate": 9.926397045441744e-05, + "loss": 0.0438, + "step": 2040 + }, + { + "grad_norm": 0.7673093676567078, + "learning_rate": 9.924976994358417e-05, + "loss": 0.0463, + "step": 2050 + }, + { + "grad_norm": 0.6097151041030884, + "learning_rate": 9.923543478576048e-05, + "loss": 0.0424, + "step": 2060 + }, + { + "grad_norm": 0.8347402215003967, + "learning_rate": 9.922096502013813e-05, + "loss": 0.0433, + "step": 2070 + }, + { + "grad_norm": 0.7491519451141357, + "learning_rate": 9.92063606862769e-05, + "loss": 0.0463, + "step": 2080 + }, + { + "grad_norm": 0.7769274711608887, + "learning_rate": 9.919162182410453e-05, + "loss": 0.0468, + "step": 2090 + }, + { + "grad_norm": 0.5936280488967896, + "learning_rate": 9.917674847391645e-05, + "loss": 0.0429, + "step": 2100 + }, + { + "grad_norm": 0.6929172873497009, + "learning_rate": 9.916174067637584e-05, + "loss": 0.0431, + "step": 2110 + }, + { + "grad_norm": 0.6833594441413879, + "learning_rate": 9.914659847251348e-05, + "loss": 0.0417, + "step": 2120 + }, + { + "grad_norm": 0.7112385630607605, + "learning_rate": 9.913132190372753e-05, + "loss": 0.044, + "step": 2130 + }, + { + "grad_norm": 0.48244747519493103, + "learning_rate": 9.911591101178359e-05, + "loss": 0.0437, + "step": 2140 + }, + { + "grad_norm": 0.7037522792816162, + "learning_rate": 9.910036583881443e-05, + "loss": 0.0414, + "step": 2150 + }, + { + "grad_norm": 0.6704358458518982, + "learning_rate": 9.908468642731995e-05, + "loss": 0.0366, + "step": 2160 + }, + { + "grad_norm": 0.5901470184326172, + "learning_rate": 9.906887282016707e-05, + "loss": 0.0446, + "step": 2170 + }, + { + "grad_norm": 0.6559775471687317, + "learning_rate": 9.90529250605896e-05, + "loss": 0.0422, + "step": 2180 + }, + { + "grad_norm": 0.8128564357757568, + "learning_rate": 9.903684319218809e-05, + "loss": 0.0444, + "step": 2190 + }, + { + "grad_norm": 0.5437960624694824, + "learning_rate": 9.902062725892976e-05, + "loss": 0.044, + "step": 2200 + }, + { + "grad_norm": 0.8856530785560608, + "learning_rate": 9.900427730514834e-05, + "loss": 0.0466, + "step": 2210 + }, + { + "grad_norm": 0.5385216474533081, + "learning_rate": 9.8987793375544e-05, + "loss": 0.0451, + "step": 2220 + }, + { + "grad_norm": 0.6559721231460571, + "learning_rate": 9.897117551518318e-05, + "loss": 0.0437, + "step": 2230 + }, + { + "grad_norm": 0.8500446081161499, + "learning_rate": 9.895442376949844e-05, + "loss": 0.0429, + "step": 2240 + }, + { + "grad_norm": 0.538585364818573, + "learning_rate": 9.893753818428845e-05, + "loss": 0.0475, + "step": 2250 + }, + { + "grad_norm": 0.608948826789856, + "learning_rate": 9.892051880571773e-05, + "loss": 0.0468, + "step": 2260 + }, + { + "grad_norm": 0.6505542993545532, + "learning_rate": 9.890336568031663e-05, + "loss": 0.0467, + "step": 2270 + }, + { + "grad_norm": 0.5538124442100525, + "learning_rate": 9.888607885498113e-05, + "loss": 0.0409, + "step": 2280 + }, + { + "grad_norm": 0.6162983775138855, + "learning_rate": 9.886865837697275e-05, + "loss": 0.0385, + "step": 2290 + }, + { + "grad_norm": 0.5613241195678711, + "learning_rate": 9.88511042939184e-05, + "loss": 0.0403, + "step": 2300 + }, + { + "grad_norm": 0.6898803114891052, + "learning_rate": 9.883341665381028e-05, + "loss": 0.0406, + "step": 2310 + }, + { + "grad_norm": 0.7443089485168457, + "learning_rate": 9.881559550500575e-05, + "loss": 0.04, + "step": 2320 + }, + { + "grad_norm": 0.6974295377731323, + "learning_rate": 9.879764089622712e-05, + "loss": 0.0449, + "step": 2330 + }, + { + "grad_norm": 0.6060269474983215, + "learning_rate": 9.87795528765616e-05, + "loss": 0.0416, + "step": 2340 + }, + { + "grad_norm": 0.6247705221176147, + "learning_rate": 9.876133149546118e-05, + "loss": 0.0404, + "step": 2350 + }, + { + "grad_norm": 0.5524375438690186, + "learning_rate": 9.874297680274238e-05, + "loss": 0.0429, + "step": 2360 + }, + { + "grad_norm": 0.6544674634933472, + "learning_rate": 9.872448884858624e-05, + "loss": 0.0419, + "step": 2370 + }, + { + "grad_norm": 0.7922749519348145, + "learning_rate": 9.870586768353815e-05, + "loss": 0.0403, + "step": 2380 + }, + { + "grad_norm": 0.5878010392189026, + "learning_rate": 9.868711335850764e-05, + "loss": 0.0413, + "step": 2390 + }, + { + "grad_norm": 0.577335774898529, + "learning_rate": 9.866822592476833e-05, + "loss": 0.0403, + "step": 2400 + }, + { + "grad_norm": 0.6214291453361511, + "learning_rate": 9.86492054339577e-05, + "loss": 0.036, + "step": 2410 + }, + { + "grad_norm": 0.6292899250984192, + "learning_rate": 9.863005193807711e-05, + "loss": 0.0387, + "step": 2420 + }, + { + "grad_norm": 0.6217591166496277, + "learning_rate": 9.861076548949143e-05, + "loss": 0.0386, + "step": 2430 + }, + { + "grad_norm": 0.642356812953949, + "learning_rate": 9.859134614092912e-05, + "loss": 0.0381, + "step": 2440 + }, + { + "grad_norm": 0.615687906742096, + "learning_rate": 9.857179394548191e-05, + "loss": 0.0387, + "step": 2450 + }, + { + "grad_norm": 0.6285825967788696, + "learning_rate": 9.855210895660477e-05, + "loss": 0.0422, + "step": 2460 + }, + { + "grad_norm": 0.5457389950752258, + "learning_rate": 9.853229122811568e-05, + "loss": 0.036, + "step": 2470 + }, + { + "grad_norm": 0.7624576091766357, + "learning_rate": 9.851234081419559e-05, + "loss": 0.0456, + "step": 2480 + }, + { + "grad_norm": 0.6762944459915161, + "learning_rate": 9.849225776938814e-05, + "loss": 0.039, + "step": 2490 + }, + { + "grad_norm": 0.607412576675415, + "learning_rate": 9.847204214859964e-05, + "loss": 0.0407, + "step": 2500 + }, + { + "grad_norm": 0.5367524027824402, + "learning_rate": 9.845169400709879e-05, + "loss": 0.0409, + "step": 2510 + }, + { + "grad_norm": 0.6632624864578247, + "learning_rate": 9.843121340051664e-05, + "loss": 0.0377, + "step": 2520 + }, + { + "grad_norm": 0.6622194647789001, + "learning_rate": 9.841060038484641e-05, + "loss": 0.0369, + "step": 2530 + }, + { + "grad_norm": 0.4699934124946594, + "learning_rate": 9.838985501644328e-05, + "loss": 0.0346, + "step": 2540 + }, + { + "grad_norm": 0.7941675186157227, + "learning_rate": 9.83689773520243e-05, + "loss": 0.0417, + "step": 2550 + }, + { + "grad_norm": 0.5687143802642822, + "learning_rate": 9.834796744866819e-05, + "loss": 0.0358, + "step": 2560 + }, + { + "grad_norm": 0.5689958333969116, + "learning_rate": 9.832682536381525e-05, + "loss": 0.0353, + "step": 2570 + }, + { + "grad_norm": 0.7229174971580505, + "learning_rate": 9.830555115526711e-05, + "loss": 0.0398, + "step": 2580 + }, + { + "grad_norm": 0.6610782742500305, + "learning_rate": 9.828414488118667e-05, + "loss": 0.0373, + "step": 2590 + }, + { + "grad_norm": 0.6783902049064636, + "learning_rate": 9.826260660009785e-05, + "loss": 0.0378, + "step": 2600 + }, + { + "grad_norm": 0.5838411450386047, + "learning_rate": 9.824093637088547e-05, + "loss": 0.0374, + "step": 2610 + }, + { + "grad_norm": 0.5158525109291077, + "learning_rate": 9.821913425279514e-05, + "loss": 0.0346, + "step": 2620 + }, + { + "grad_norm": 0.5010576248168945, + "learning_rate": 9.8197200305433e-05, + "loss": 0.0367, + "step": 2630 + }, + { + "grad_norm": 0.5873668193817139, + "learning_rate": 9.817513458876564e-05, + "loss": 0.0389, + "step": 2640 + }, + { + "grad_norm": 0.669737696647644, + "learning_rate": 9.815293716311987e-05, + "loss": 0.0391, + "step": 2650 + }, + { + "grad_norm": 0.6051730513572693, + "learning_rate": 9.813060808918262e-05, + "loss": 0.0358, + "step": 2660 + }, + { + "grad_norm": 0.6769428253173828, + "learning_rate": 9.810814742800069e-05, + "loss": 0.0376, + "step": 2670 + }, + { + "grad_norm": 0.5549777150154114, + "learning_rate": 9.808555524098074e-05, + "loss": 0.035, + "step": 2680 + }, + { + "grad_norm": 0.6324145197868347, + "learning_rate": 9.806283158988887e-05, + "loss": 0.037, + "step": 2690 + }, + { + "grad_norm": 0.522504448890686, + "learning_rate": 9.803997653685072e-05, + "loss": 0.0353, + "step": 2700 + }, + { + "grad_norm": 0.6257413029670715, + "learning_rate": 9.801699014435112e-05, + "loss": 0.0398, + "step": 2710 + }, + { + "grad_norm": 0.6243777871131897, + "learning_rate": 9.799387247523398e-05, + "loss": 0.0329, + "step": 2720 + }, + { + "grad_norm": 0.5824418663978577, + "learning_rate": 9.797062359270215e-05, + "loss": 0.0368, + "step": 2730 + }, + { + "grad_norm": 0.5653295516967773, + "learning_rate": 9.794724356031715e-05, + "loss": 0.038, + "step": 2740 + }, + { + "grad_norm": 0.5860871076583862, + "learning_rate": 9.792373244199913e-05, + "loss": 0.0366, + "step": 2750 + }, + { + "grad_norm": 0.6080830693244934, + "learning_rate": 9.790009030202658e-05, + "loss": 0.0391, + "step": 2760 + }, + { + "grad_norm": 0.5480026602745056, + "learning_rate": 9.78763172050362e-05, + "loss": 0.0342, + "step": 2770 + }, + { + "grad_norm": 0.4915301203727722, + "learning_rate": 9.785241321602274e-05, + "loss": 0.0382, + "step": 2780 + }, + { + "grad_norm": 0.5721039772033691, + "learning_rate": 9.782837840033879e-05, + "loss": 0.0387, + "step": 2790 + }, + { + "grad_norm": 0.6181113123893738, + "learning_rate": 9.780421282369461e-05, + "loss": 0.0343, + "step": 2800 + }, + { + "grad_norm": 0.5936803817749023, + "learning_rate": 9.777991655215797e-05, + "loss": 0.0325, + "step": 2810 + }, + { + "grad_norm": 0.6910586357116699, + "learning_rate": 9.775548965215394e-05, + "loss": 0.0314, + "step": 2820 + }, + { + "grad_norm": 0.5371761918067932, + "learning_rate": 9.773093219046474e-05, + "loss": 0.036, + "step": 2830 + }, + { + "grad_norm": 0.4973508417606354, + "learning_rate": 9.770624423422954e-05, + "loss": 0.0376, + "step": 2840 + }, + { + "grad_norm": 0.6510023474693298, + "learning_rate": 9.768142585094426e-05, + "loss": 0.0353, + "step": 2850 + }, + { + "grad_norm": 0.6040503978729248, + "learning_rate": 9.765647710846142e-05, + "loss": 0.0347, + "step": 2860 + }, + { + "grad_norm": 0.5976170301437378, + "learning_rate": 9.763139807498991e-05, + "loss": 0.034, + "step": 2870 + }, + { + "grad_norm": 0.6797834038734436, + "learning_rate": 9.760618881909487e-05, + "loss": 0.0339, + "step": 2880 + }, + { + "grad_norm": 0.5329861044883728, + "learning_rate": 9.758084940969744e-05, + "loss": 0.0317, + "step": 2890 + }, + { + "grad_norm": 0.6726340055465698, + "learning_rate": 9.755537991607459e-05, + "loss": 0.0319, + "step": 2900 + }, + { + "grad_norm": 0.608772873878479, + "learning_rate": 9.752978040785895e-05, + "loss": 0.0313, + "step": 2910 + }, + { + "grad_norm": 0.7140541076660156, + "learning_rate": 9.750405095503859e-05, + "loss": 0.0311, + "step": 2920 + }, + { + "grad_norm": 0.5499521493911743, + "learning_rate": 9.747819162795686e-05, + "loss": 0.0367, + "step": 2930 + }, + { + "grad_norm": 0.6744058728218079, + "learning_rate": 9.745220249731217e-05, + "loss": 0.0407, + "step": 2940 + }, + { + "grad_norm": 0.7178553342819214, + "learning_rate": 9.742608363415781e-05, + "loss": 0.041, + "step": 2950 + }, + { + "grad_norm": 0.6183580160140991, + "learning_rate": 9.739983510990176e-05, + "loss": 0.0394, + "step": 2960 + }, + { + "grad_norm": 0.553803026676178, + "learning_rate": 9.737345699630647e-05, + "loss": 0.0329, + "step": 2970 + }, + { + "grad_norm": 0.4599069654941559, + "learning_rate": 9.734694936548869e-05, + "loss": 0.0313, + "step": 2980 + }, + { + "grad_norm": 0.5753762125968933, + "learning_rate": 9.732031228991932e-05, + "loss": 0.0304, + "step": 2990 + }, + { + "grad_norm": 0.6017200946807861, + "learning_rate": 9.729354584242302e-05, + "loss": 0.0343, + "step": 3000 + }, + { + "grad_norm": 0.6461474299430847, + "learning_rate": 9.726665009617832e-05, + "loss": 0.0379, + "step": 3010 + }, + { + "grad_norm": 0.5760048627853394, + "learning_rate": 9.723962512471714e-05, + "loss": 0.0342, + "step": 3020 + }, + { + "grad_norm": 0.621891438961029, + "learning_rate": 9.72124710019247e-05, + "loss": 0.0316, + "step": 3030 + }, + { + "grad_norm": 0.524480402469635, + "learning_rate": 9.718518780203934e-05, + "loss": 0.0292, + "step": 3040 + }, + { + "grad_norm": 0.6318420171737671, + "learning_rate": 9.715777559965228e-05, + "loss": 0.0316, + "step": 3050 + }, + { + "grad_norm": 0.566210150718689, + "learning_rate": 9.713023446970746e-05, + "loss": 0.0309, + "step": 3060 + }, + { + "grad_norm": 0.704804003238678, + "learning_rate": 9.710256448750126e-05, + "loss": 0.0332, + "step": 3070 + }, + { + "grad_norm": 0.7099645733833313, + "learning_rate": 9.707476572868235e-05, + "loss": 0.0333, + "step": 3080 + }, + { + "grad_norm": 0.6047917604446411, + "learning_rate": 9.704683826925149e-05, + "loss": 0.0314, + "step": 3090 + }, + { + "grad_norm": 0.7446325421333313, + "learning_rate": 9.701878218556129e-05, + "loss": 0.035, + "step": 3100 + }, + { + "grad_norm": 0.6377901434898376, + "learning_rate": 9.699059755431598e-05, + "loss": 0.0387, + "step": 3110 + }, + { + "grad_norm": 0.5415878295898438, + "learning_rate": 9.696228445257132e-05, + "loss": 0.0299, + "step": 3120 + }, + { + "grad_norm": 0.6148945093154907, + "learning_rate": 9.693384295773419e-05, + "loss": 0.0391, + "step": 3130 + }, + { + "grad_norm": 0.578273594379425, + "learning_rate": 9.690527314756259e-05, + "loss": 0.0305, + "step": 3140 + }, + { + "grad_norm": 0.5767846703529358, + "learning_rate": 9.687657510016527e-05, + "loss": 0.0327, + "step": 3150 + }, + { + "grad_norm": 0.532577633857727, + "learning_rate": 9.684774889400161e-05, + "loss": 0.0276, + "step": 3160 + }, + { + "grad_norm": 0.5682797431945801, + "learning_rate": 9.681879460788135e-05, + "loss": 0.0295, + "step": 3170 + }, + { + "grad_norm": 0.6185994744300842, + "learning_rate": 9.67897123209644e-05, + "loss": 0.0306, + "step": 3180 + }, + { + "grad_norm": 0.6122663617134094, + "learning_rate": 9.676050211276062e-05, + "loss": 0.0323, + "step": 3190 + }, + { + "grad_norm": 0.6145076751708984, + "learning_rate": 9.673116406312962e-05, + "loss": 0.0289, + "step": 3200 + }, + { + "grad_norm": 0.7365098595619202, + "learning_rate": 9.67016982522805e-05, + "loss": 0.0264, + "step": 3210 + }, + { + "grad_norm": 0.6660006642341614, + "learning_rate": 9.667210476077164e-05, + "loss": 0.0307, + "step": 3220 + }, + { + "grad_norm": 0.43615856766700745, + "learning_rate": 9.664238366951055e-05, + "loss": 0.0347, + "step": 3230 + }, + { + "grad_norm": 0.5423561334609985, + "learning_rate": 9.661253505975355e-05, + "loss": 0.0286, + "step": 3240 + }, + { + "grad_norm": 0.6123466491699219, + "learning_rate": 9.658255901310557e-05, + "loss": 0.0297, + "step": 3250 + }, + { + "grad_norm": 0.5747650861740112, + "learning_rate": 9.655245561152e-05, + "loss": 0.0311, + "step": 3260 + }, + { + "grad_norm": 0.552060604095459, + "learning_rate": 9.65222249372984e-05, + "loss": 0.0287, + "step": 3270 + }, + { + "grad_norm": 0.5556790828704834, + "learning_rate": 9.649186707309026e-05, + "loss": 0.0274, + "step": 3280 + }, + { + "grad_norm": 0.5429115295410156, + "learning_rate": 9.646138210189283e-05, + "loss": 0.0297, + "step": 3290 + }, + { + "grad_norm": 0.6687198281288147, + "learning_rate": 9.643077010705087e-05, + "loss": 0.029, + "step": 3300 + }, + { + "grad_norm": 0.5589491724967957, + "learning_rate": 9.640003117225637e-05, + "loss": 0.0296, + "step": 3310 + }, + { + "grad_norm": 0.5250815749168396, + "learning_rate": 9.636916538154846e-05, + "loss": 0.0313, + "step": 3320 + }, + { + "grad_norm": 0.5884456038475037, + "learning_rate": 9.633817281931296e-05, + "loss": 0.0279, + "step": 3330 + }, + { + "grad_norm": 0.5253452658653259, + "learning_rate": 9.630705357028242e-05, + "loss": 0.0297, + "step": 3340 + }, + { + "grad_norm": 0.44990262389183044, + "learning_rate": 9.627580771953563e-05, + "loss": 0.0293, + "step": 3350 + }, + { + "grad_norm": 0.5788272023200989, + "learning_rate": 9.624443535249759e-05, + "loss": 0.028, + "step": 3360 + }, + { + "grad_norm": 0.4931959807872772, + "learning_rate": 9.621293655493913e-05, + "loss": 0.0282, + "step": 3370 + }, + { + "grad_norm": 0.49001696705818176, + "learning_rate": 9.618131141297675e-05, + "loss": 0.025, + "step": 3380 + }, + { + "grad_norm": 0.6005767583847046, + "learning_rate": 9.614956001307242e-05, + "loss": 0.0282, + "step": 3390 + }, + { + "grad_norm": 0.5819389224052429, + "learning_rate": 9.611768244203321e-05, + "loss": 0.0287, + "step": 3400 + }, + { + "grad_norm": 0.5611231923103333, + "learning_rate": 9.60856787870112e-05, + "loss": 0.0315, + "step": 3410 + }, + { + "grad_norm": 0.49898526072502136, + "learning_rate": 9.605354913550318e-05, + "loss": 0.0267, + "step": 3420 + }, + { + "grad_norm": 0.5152934789657593, + "learning_rate": 9.602129357535037e-05, + "loss": 0.0248, + "step": 3430 + }, + { + "grad_norm": 0.4927721917629242, + "learning_rate": 9.598891219473825e-05, + "loss": 0.0272, + "step": 3440 + }, + { + "grad_norm": 0.5922147035598755, + "learning_rate": 9.595640508219625e-05, + "loss": 0.0225, + "step": 3450 + }, + { + "grad_norm": 0.6173586845397949, + "learning_rate": 9.592377232659761e-05, + "loss": 0.0269, + "step": 3460 + }, + { + "grad_norm": 0.6940552592277527, + "learning_rate": 9.589101401715904e-05, + "loss": 0.0314, + "step": 3470 + }, + { + "grad_norm": 0.5696406364440918, + "learning_rate": 9.585813024344045e-05, + "loss": 0.0271, + "step": 3480 + }, + { + "grad_norm": 0.5202314853668213, + "learning_rate": 9.58251210953449e-05, + "loss": 0.0247, + "step": 3490 + }, + { + "grad_norm": 0.5222875475883484, + "learning_rate": 9.579198666311809e-05, + "loss": 0.0238, + "step": 3500 + }, + { + "grad_norm": 0.6505278944969177, + "learning_rate": 9.575872703734832e-05, + "loss": 0.0295, + "step": 3510 + }, + { + "grad_norm": 0.619454026222229, + "learning_rate": 9.572534230896611e-05, + "loss": 0.0268, + "step": 3520 + }, + { + "grad_norm": 0.48343542218208313, + "learning_rate": 9.569183256924403e-05, + "loss": 0.0269, + "step": 3530 + }, + { + "grad_norm": 0.4461105763912201, + "learning_rate": 9.565819790979646e-05, + "loss": 0.0255, + "step": 3540 + }, + { + "grad_norm": 0.5143640637397766, + "learning_rate": 9.562443842257925e-05, + "loss": 0.0257, + "step": 3550 + }, + { + "grad_norm": 0.5106943249702454, + "learning_rate": 9.559055419988956e-05, + "loss": 0.0265, + "step": 3560 + }, + { + "grad_norm": 0.6030021905899048, + "learning_rate": 9.555654533436557e-05, + "loss": 0.0269, + "step": 3570 + }, + { + "grad_norm": 0.5566527843475342, + "learning_rate": 9.552241191898621e-05, + "loss": 0.0279, + "step": 3580 + }, + { + "grad_norm": 0.4313097894191742, + "learning_rate": 9.548815404707092e-05, + "loss": 0.0246, + "step": 3590 + }, + { + "grad_norm": 0.6710801720619202, + "learning_rate": 9.545377181227942e-05, + "loss": 0.0242, + "step": 3600 + }, + { + "grad_norm": 0.4369684159755707, + "learning_rate": 9.541926530861145e-05, + "loss": 0.0238, + "step": 3610 + }, + { + "grad_norm": 0.48853105306625366, + "learning_rate": 9.538463463040645e-05, + "loss": 0.0226, + "step": 3620 + }, + { + "grad_norm": 0.5221312642097473, + "learning_rate": 9.534987987234337e-05, + "loss": 0.0267, + "step": 3630 + }, + { + "grad_norm": 0.5919408798217773, + "learning_rate": 9.53150011294404e-05, + "loss": 0.0234, + "step": 3640 + }, + { + "grad_norm": 0.5544537305831909, + "learning_rate": 9.527999849705471e-05, + "loss": 0.0229, + "step": 3650 + }, + { + "grad_norm": 0.5028519034385681, + "learning_rate": 9.524487207088213e-05, + "loss": 0.0229, + "step": 3660 + }, + { + "grad_norm": 0.5428268313407898, + "learning_rate": 9.520962194695698e-05, + "loss": 0.0234, + "step": 3670 + }, + { + "grad_norm": 0.4685244858264923, + "learning_rate": 9.517424822165175e-05, + "loss": 0.023, + "step": 3680 + }, + { + "grad_norm": 0.6033672094345093, + "learning_rate": 9.513875099167685e-05, + "loss": 0.028, + "step": 3690 + }, + { + "grad_norm": 0.5230802297592163, + "learning_rate": 9.510313035408035e-05, + "loss": 0.0211, + "step": 3700 + }, + { + "grad_norm": 0.5911836624145508, + "learning_rate": 9.506738640624775e-05, + "loss": 0.0226, + "step": 3710 + }, + { + "grad_norm": 0.4973941445350647, + "learning_rate": 9.50315192459016e-05, + "loss": 0.0233, + "step": 3720 + }, + { + "grad_norm": 0.5748140811920166, + "learning_rate": 9.499552897110136e-05, + "loss": 0.0258, + "step": 3730 + }, + { + "grad_norm": 0.5241221785545349, + "learning_rate": 9.495941568024304e-05, + "loss": 0.025, + "step": 3740 + }, + { + "grad_norm": 0.42789241671562195, + "learning_rate": 9.492317947205904e-05, + "loss": 0.0231, + "step": 3750 + }, + { + "grad_norm": 0.6186679601669312, + "learning_rate": 9.488682044561775e-05, + "loss": 0.0259, + "step": 3760 + }, + { + "grad_norm": 0.6295249462127686, + "learning_rate": 9.485033870032335e-05, + "loss": 0.0204, + "step": 3770 + }, + { + "grad_norm": 0.5628687739372253, + "learning_rate": 9.481373433591556e-05, + "loss": 0.0252, + "step": 3780 + }, + { + "grad_norm": 0.4300912916660309, + "learning_rate": 9.47770074524693e-05, + "loss": 0.0224, + "step": 3790 + }, + { + "grad_norm": 0.5620883703231812, + "learning_rate": 9.474015815039446e-05, + "loss": 0.0224, + "step": 3800 + }, + { + "grad_norm": 0.4041639268398285, + "learning_rate": 9.470318653043565e-05, + "loss": 0.0275, + "step": 3810 + }, + { + "grad_norm": 0.6825946569442749, + "learning_rate": 9.466609269367185e-05, + "loss": 0.0288, + "step": 3820 + }, + { + "grad_norm": 0.528630256652832, + "learning_rate": 9.46288767415162e-05, + "loss": 0.0222, + "step": 3830 + }, + { + "grad_norm": 0.5332393646240234, + "learning_rate": 9.459153877571567e-05, + "loss": 0.0231, + "step": 3840 + }, + { + "grad_norm": 0.41178372502326965, + "learning_rate": 9.455407889835087e-05, + "loss": 0.0234, + "step": 3850 + }, + { + "grad_norm": 0.4013999402523041, + "learning_rate": 9.451649721183564e-05, + "loss": 0.026, + "step": 3860 + }, + { + "grad_norm": 0.5963751673698425, + "learning_rate": 9.447879381891692e-05, + "loss": 0.0207, + "step": 3870 + }, + { + "grad_norm": 0.46209827065467834, + "learning_rate": 9.444096882267428e-05, + "loss": 0.0217, + "step": 3880 + }, + { + "grad_norm": 0.5197754502296448, + "learning_rate": 9.440302232651988e-05, + "loss": 0.0265, + "step": 3890 + }, + { + "grad_norm": 0.6915350556373596, + "learning_rate": 9.436495443419795e-05, + "loss": 0.0297, + "step": 3900 + }, + { + "grad_norm": 0.48685312271118164, + "learning_rate": 9.432676524978466e-05, + "loss": 0.0233, + "step": 3910 + }, + { + "grad_norm": 0.47423553466796875, + "learning_rate": 9.42884548776878e-05, + "loss": 0.0288, + "step": 3920 + }, + { + "grad_norm": 0.4048287868499756, + "learning_rate": 9.425002342264646e-05, + "loss": 0.0211, + "step": 3930 + }, + { + "grad_norm": 0.44486644864082336, + "learning_rate": 9.421147098973077e-05, + "loss": 0.0238, + "step": 3940 + }, + { + "grad_norm": 0.5044620633125305, + "learning_rate": 9.41727976843416e-05, + "loss": 0.023, + "step": 3950 + }, + { + "grad_norm": 0.6554492712020874, + "learning_rate": 9.413400361221029e-05, + "loss": 0.0224, + "step": 3960 + }, + { + "grad_norm": 0.5063493251800537, + "learning_rate": 9.409508887939835e-05, + "loss": 0.02, + "step": 3970 + }, + { + "grad_norm": 0.5956079959869385, + "learning_rate": 9.40560535922972e-05, + "loss": 0.02, + "step": 3980 + }, + { + "grad_norm": 0.6585922837257385, + "learning_rate": 9.40168978576278e-05, + "loss": 0.0224, + "step": 3990 + }, + { + "grad_norm": 0.4550802707672119, + "learning_rate": 9.397762178244043e-05, + "loss": 0.0235, + "step": 4000 + }, + { + "grad_norm": 0.4042211174964905, + "learning_rate": 9.393822547411439e-05, + "loss": 0.0208, + "step": 4010 + }, + { + "grad_norm": 0.4356859028339386, + "learning_rate": 9.389870904035769e-05, + "loss": 0.0199, + "step": 4020 + }, + { + "grad_norm": 0.45227131247520447, + "learning_rate": 9.385907258920672e-05, + "loss": 0.0227, + "step": 4030 + }, + { + "grad_norm": 0.4879400134086609, + "learning_rate": 9.381931622902607e-05, + "loss": 0.0192, + "step": 4040 + }, + { + "grad_norm": 0.48629337549209595, + "learning_rate": 9.377944006850807e-05, + "loss": 0.0225, + "step": 4050 + }, + { + "grad_norm": 0.4865284562110901, + "learning_rate": 9.373944421667265e-05, + "loss": 0.0212, + "step": 4060 + }, + { + "grad_norm": 0.6001793742179871, + "learning_rate": 9.369932878286691e-05, + "loss": 0.0184, + "step": 4070 + }, + { + "grad_norm": 0.5513994097709656, + "learning_rate": 9.365909387676494e-05, + "loss": 0.021, + "step": 4080 + }, + { + "grad_norm": 0.6248122453689575, + "learning_rate": 9.361873960836744e-05, + "loss": 0.0208, + "step": 4090 + }, + { + "grad_norm": 0.4238356947898865, + "learning_rate": 9.357826608800142e-05, + "loss": 0.0182, + "step": 4100 + }, + { + "grad_norm": 0.4596106708049774, + "learning_rate": 9.353767342631994e-05, + "loss": 0.023, + "step": 4110 + }, + { + "grad_norm": 0.39670488238334656, + "learning_rate": 9.34969617343018e-05, + "loss": 0.023, + "step": 4120 + }, + { + "grad_norm": 0.5592398047447205, + "learning_rate": 9.345613112325122e-05, + "loss": 0.0237, + "step": 4130 + }, + { + "grad_norm": 0.5025811195373535, + "learning_rate": 9.34151817047975e-05, + "loss": 0.0207, + "step": 4140 + }, + { + "grad_norm": 0.5580839514732361, + "learning_rate": 9.33741135908948e-05, + "loss": 0.0224, + "step": 4150 + }, + { + "grad_norm": 0.5201109051704407, + "learning_rate": 9.33329268938218e-05, + "loss": 0.0206, + "step": 4160 + }, + { + "grad_norm": 0.513161838054657, + "learning_rate": 9.329162172618132e-05, + "loss": 0.0231, + "step": 4170 + }, + { + "grad_norm": 0.4309871792793274, + "learning_rate": 9.325019820090013e-05, + "loss": 0.022, + "step": 4180 + }, + { + "grad_norm": 0.43937987089157104, + "learning_rate": 9.320865643122855e-05, + "loss": 0.0233, + "step": 4190 + }, + { + "grad_norm": 0.4912075400352478, + "learning_rate": 9.316699653074023e-05, + "loss": 0.0182, + "step": 4200 + }, + { + "grad_norm": 0.4246988892555237, + "learning_rate": 9.312521861333172e-05, + "loss": 0.0192, + "step": 4210 + }, + { + "grad_norm": 0.39220497012138367, + "learning_rate": 9.308332279322224e-05, + "loss": 0.0194, + "step": 4220 + }, + { + "grad_norm": 0.5391877293586731, + "learning_rate": 9.304130918495338e-05, + "loss": 0.02, + "step": 4230 + }, + { + "grad_norm": 0.4661605954170227, + "learning_rate": 9.299917790338874e-05, + "loss": 0.0201, + "step": 4240 + }, + { + "grad_norm": 0.3973395824432373, + "learning_rate": 9.295692906371363e-05, + "loss": 0.0213, + "step": 4250 + }, + { + "grad_norm": 0.4798072278499603, + "learning_rate": 9.291456278143476e-05, + "loss": 0.02, + "step": 4260 + }, + { + "grad_norm": 0.6149327158927917, + "learning_rate": 9.287207917237994e-05, + "loss": 0.0203, + "step": 4270 + }, + { + "grad_norm": 0.5336242318153381, + "learning_rate": 9.282947835269773e-05, + "loss": 0.0215, + "step": 4280 + }, + { + "grad_norm": 0.5178649425506592, + "learning_rate": 9.278676043885715e-05, + "loss": 0.0194, + "step": 4290 + }, + { + "grad_norm": 0.4814247190952301, + "learning_rate": 9.274392554764733e-05, + "loss": 0.0171, + "step": 4300 + }, + { + "grad_norm": 0.32441624999046326, + "learning_rate": 9.270097379617723e-05, + "loss": 0.0184, + "step": 4310 + }, + { + "grad_norm": 0.5904157757759094, + "learning_rate": 9.26579053018753e-05, + "loss": 0.0204, + "step": 4320 + }, + { + "grad_norm": 0.45750078558921814, + "learning_rate": 9.261472018248918e-05, + "loss": 0.0191, + "step": 4330 + }, + { + "grad_norm": 0.5417255759239197, + "learning_rate": 9.25714185560853e-05, + "loss": 0.0208, + "step": 4340 + }, + { + "grad_norm": 0.41097697615623474, + "learning_rate": 9.252800054104868e-05, + "loss": 0.0238, + "step": 4350 + }, + { + "grad_norm": 0.42212045192718506, + "learning_rate": 9.248446625608252e-05, + "loss": 0.0196, + "step": 4360 + }, + { + "grad_norm": 0.5629220008850098, + "learning_rate": 9.244081582020789e-05, + "loss": 0.0208, + "step": 4370 + }, + { + "grad_norm": 0.4684229791164398, + "learning_rate": 9.239704935276339e-05, + "loss": 0.0226, + "step": 4380 + }, + { + "grad_norm": 0.43924668431282043, + "learning_rate": 9.235316697340489e-05, + "loss": 0.0247, + "step": 4390 + }, + { + "grad_norm": 0.5363799929618835, + "learning_rate": 9.230916880210512e-05, + "loss": 0.0193, + "step": 4400 + }, + { + "grad_norm": 0.4902143180370331, + "learning_rate": 9.226505495915342e-05, + "loss": 0.0204, + "step": 4410 + }, + { + "grad_norm": 0.4934154450893402, + "learning_rate": 9.222082556515536e-05, + "loss": 0.0204, + "step": 4420 + }, + { + "grad_norm": 0.5217421650886536, + "learning_rate": 9.217648074103242e-05, + "loss": 0.0187, + "step": 4430 + }, + { + "grad_norm": 0.5157290101051331, + "learning_rate": 9.213202060802161e-05, + "loss": 0.0193, + "step": 4440 + }, + { + "grad_norm": 0.48323342204093933, + "learning_rate": 9.208744528767528e-05, + "loss": 0.0222, + "step": 4450 + }, + { + "grad_norm": 0.46333587169647217, + "learning_rate": 9.204275490186064e-05, + "loss": 0.0188, + "step": 4460 + }, + { + "grad_norm": 0.48506951332092285, + "learning_rate": 9.199794957275949e-05, + "loss": 0.0181, + "step": 4470 + }, + { + "grad_norm": 0.547903835773468, + "learning_rate": 9.19530294228679e-05, + "loss": 0.0201, + "step": 4480 + }, + { + "grad_norm": 0.5428500771522522, + "learning_rate": 9.190799457499583e-05, + "loss": 0.023, + "step": 4490 + }, + { + "grad_norm": 0.4775592088699341, + "learning_rate": 9.186284515226686e-05, + "loss": 0.0187, + "step": 4500 + }, + { + "grad_norm": 0.515190839767456, + "learning_rate": 9.181758127811777e-05, + "loss": 0.0174, + "step": 4510 + }, + { + "grad_norm": 0.523912250995636, + "learning_rate": 9.177220307629825e-05, + "loss": 0.0183, + "step": 4520 + }, + { + "grad_norm": 0.46641191840171814, + "learning_rate": 9.172671067087059e-05, + "loss": 0.0218, + "step": 4530 + }, + { + "grad_norm": 0.4439499080181122, + "learning_rate": 9.16811041862093e-05, + "loss": 0.0194, + "step": 4540 + }, + { + "grad_norm": 0.39287662506103516, + "learning_rate": 9.163538374700076e-05, + "loss": 0.0174, + "step": 4550 + }, + { + "grad_norm": 0.49717986583709717, + "learning_rate": 9.158954947824287e-05, + "loss": 0.0192, + "step": 4560 + }, + { + "grad_norm": 0.4734252691268921, + "learning_rate": 9.154360150524482e-05, + "loss": 0.0177, + "step": 4570 + }, + { + "grad_norm": 0.37044256925582886, + "learning_rate": 9.14975399536266e-05, + "loss": 0.0216, + "step": 4580 + }, + { + "grad_norm": 0.36238211393356323, + "learning_rate": 9.14513649493187e-05, + "loss": 0.0241, + "step": 4590 + }, + { + "grad_norm": 0.3750474750995636, + "learning_rate": 9.140507661856187e-05, + "loss": 0.0188, + "step": 4600 + }, + { + "grad_norm": 0.6232521533966064, + "learning_rate": 9.135867508790661e-05, + "loss": 0.0232, + "step": 4610 + }, + { + "grad_norm": 0.40693479776382446, + "learning_rate": 9.131216048421291e-05, + "loss": 0.0191, + "step": 4620 + }, + { + "grad_norm": 0.46917590498924255, + "learning_rate": 9.126553293464998e-05, + "loss": 0.0192, + "step": 4630 + }, + { + "grad_norm": 0.6374298930168152, + "learning_rate": 9.121879256669572e-05, + "loss": 0.0209, + "step": 4640 + }, + { + "grad_norm": 0.49459344148635864, + "learning_rate": 9.117193950813652e-05, + "loss": 0.0191, + "step": 4650 + }, + { + "grad_norm": 0.3625936508178711, + "learning_rate": 9.112497388706685e-05, + "loss": 0.02, + "step": 4660 + }, + { + "grad_norm": 0.5517452955245972, + "learning_rate": 9.10778958318889e-05, + "loss": 0.0178, + "step": 4670 + }, + { + "grad_norm": 0.5170832276344299, + "learning_rate": 9.103070547131232e-05, + "loss": 0.0172, + "step": 4680 + }, + { + "grad_norm": 0.5125618577003479, + "learning_rate": 9.098340293435375e-05, + "loss": 0.0172, + "step": 4690 + }, + { + "grad_norm": 0.6093714833259583, + "learning_rate": 9.093598835033649e-05, + "loss": 0.0253, + "step": 4700 + }, + { + "grad_norm": 0.48880383372306824, + "learning_rate": 9.088846184889021e-05, + "loss": 0.02, + "step": 4710 + }, + { + "grad_norm": 0.3739651143550873, + "learning_rate": 9.084082355995057e-05, + "loss": 0.0172, + "step": 4720 + }, + { + "grad_norm": 0.41044220328330994, + "learning_rate": 9.079307361375882e-05, + "loss": 0.0178, + "step": 4730 + }, + { + "grad_norm": 0.42307567596435547, + "learning_rate": 9.074521214086149e-05, + "loss": 0.0184, + "step": 4740 + }, + { + "grad_norm": 0.39302605390548706, + "learning_rate": 9.069723927211001e-05, + "loss": 0.0175, + "step": 4750 + }, + { + "grad_norm": 0.4822540581226349, + "learning_rate": 9.064915513866037e-05, + "loss": 0.0205, + "step": 4760 + }, + { + "grad_norm": 0.4448596239089966, + "learning_rate": 9.060095987197279e-05, + "loss": 0.0246, + "step": 4770 + }, + { + "grad_norm": 0.3715053200721741, + "learning_rate": 9.055265360381126e-05, + "loss": 0.0182, + "step": 4780 + }, + { + "grad_norm": 0.3781905770301819, + "learning_rate": 9.050423646624326e-05, + "loss": 0.0218, + "step": 4790 + }, + { + "grad_norm": 0.4780808091163635, + "learning_rate": 9.045570859163943e-05, + "loss": 0.0203, + "step": 4800 + }, + { + "grad_norm": 0.45449933409690857, + "learning_rate": 9.04070701126731e-05, + "loss": 0.0168, + "step": 4810 + }, + { + "grad_norm": 0.455399751663208, + "learning_rate": 9.035832116232001e-05, + "loss": 0.0192, + "step": 4820 + }, + { + "grad_norm": 0.569148600101471, + "learning_rate": 9.030946187385796e-05, + "loss": 0.0191, + "step": 4830 + }, + { + "grad_norm": 0.39798951148986816, + "learning_rate": 9.026049238086635e-05, + "loss": 0.0183, + "step": 4840 + }, + { + "grad_norm": 0.4741237759590149, + "learning_rate": 9.021141281722591e-05, + "loss": 0.0185, + "step": 4850 + }, + { + "grad_norm": 0.44035014510154724, + "learning_rate": 9.01622233171183e-05, + "loss": 0.0169, + "step": 4860 + }, + { + "grad_norm": 0.4502478241920471, + "learning_rate": 9.011292401502574e-05, + "loss": 0.0207, + "step": 4870 + }, + { + "grad_norm": 0.38435491919517517, + "learning_rate": 9.006351504573063e-05, + "loss": 0.0183, + "step": 4880 + }, + { + "grad_norm": 0.36374086141586304, + "learning_rate": 9.001399654431519e-05, + "loss": 0.0195, + "step": 4890 + }, + { + "grad_norm": 0.4774911105632782, + "learning_rate": 8.996436864616116e-05, + "loss": 0.0177, + "step": 4900 + }, + { + "grad_norm": 0.3256385624408722, + "learning_rate": 8.991463148694925e-05, + "loss": 0.0166, + "step": 4910 + }, + { + "grad_norm": 0.5121222734451294, + "learning_rate": 8.986478520265902e-05, + "loss": 0.0191, + "step": 4920 + }, + { + "grad_norm": 0.5058743953704834, + "learning_rate": 8.981482992956827e-05, + "loss": 0.0202, + "step": 4930 + }, + { + "grad_norm": 0.436591774225235, + "learning_rate": 8.976476580425282e-05, + "loss": 0.0165, + "step": 4940 + }, + { + "grad_norm": 0.4705229699611664, + "learning_rate": 8.971459296358606e-05, + "loss": 0.0195, + "step": 4950 + }, + { + "grad_norm": 0.423905611038208, + "learning_rate": 8.966431154473864e-05, + "loss": 0.0201, + "step": 4960 + }, + { + "grad_norm": 0.4129910469055176, + "learning_rate": 8.961392168517803e-05, + "loss": 0.0149, + "step": 4970 + }, + { + "grad_norm": 0.45044130086898804, + "learning_rate": 8.956342352266821e-05, + "loss": 0.0209, + "step": 4980 + }, + { + "grad_norm": 0.43892666697502136, + "learning_rate": 8.95128171952692e-05, + "loss": 0.0204, + "step": 4990 + }, + { + "grad_norm": 0.4579017758369446, + "learning_rate": 8.946210284133676e-05, + "loss": 0.0172, + "step": 5000 + }, + { + "grad_norm": 0.38334211707115173, + "learning_rate": 8.941128059952201e-05, + "loss": 0.0176, + "step": 5010 + }, + { + "grad_norm": 0.4932248592376709, + "learning_rate": 8.936035060877102e-05, + "loss": 0.0197, + "step": 5020 + }, + { + "grad_norm": 0.4217004179954529, + "learning_rate": 8.930931300832443e-05, + "loss": 0.0159, + "step": 5030 + }, + { + "grad_norm": 0.5817981958389282, + "learning_rate": 8.925816793771711e-05, + "loss": 0.0162, + "step": 5040 + }, + { + "grad_norm": 0.43254992365837097, + "learning_rate": 8.92069155367777e-05, + "loss": 0.0193, + "step": 5050 + }, + { + "grad_norm": 0.5035724639892578, + "learning_rate": 8.915555594562834e-05, + "loss": 0.0208, + "step": 5060 + }, + { + "grad_norm": 0.4568447172641754, + "learning_rate": 8.910408930468416e-05, + "loss": 0.0165, + "step": 5070 + }, + { + "grad_norm": 0.43123510479927063, + "learning_rate": 8.905251575465303e-05, + "loss": 0.0158, + "step": 5080 + }, + { + "grad_norm": 0.4527429938316345, + "learning_rate": 8.900083543653502e-05, + "loss": 0.0169, + "step": 5090 + }, + { + "grad_norm": 0.41953304409980774, + "learning_rate": 8.894904849162218e-05, + "loss": 0.0184, + "step": 5100 + }, + { + "grad_norm": 0.415479838848114, + "learning_rate": 8.889715506149802e-05, + "loss": 0.0184, + "step": 5110 + }, + { + "grad_norm": 0.3731771409511566, + "learning_rate": 8.884515528803722e-05, + "loss": 0.0154, + "step": 5120 + }, + { + "grad_norm": 0.41724300384521484, + "learning_rate": 8.879304931340517e-05, + "loss": 0.0163, + "step": 5130 + }, + { + "grad_norm": 0.47969841957092285, + "learning_rate": 8.874083728005759e-05, + "loss": 0.0177, + "step": 5140 + }, + { + "grad_norm": 0.34264758229255676, + "learning_rate": 8.868851933074021e-05, + "loss": 0.0182, + "step": 5150 + }, + { + "grad_norm": 0.4063286781311035, + "learning_rate": 8.863609560848829e-05, + "loss": 0.0182, + "step": 5160 + }, + { + "grad_norm": 0.42877396941185, + "learning_rate": 8.85835662566263e-05, + "loss": 0.0165, + "step": 5170 + }, + { + "grad_norm": 0.4081937372684479, + "learning_rate": 8.853093141876747e-05, + "loss": 0.0148, + "step": 5180 + }, + { + "grad_norm": 0.40108522772789, + "learning_rate": 8.847819123881343e-05, + "loss": 0.0174, + "step": 5190 + }, + { + "grad_norm": 0.4415458142757416, + "learning_rate": 8.842534586095383e-05, + "loss": 0.0165, + "step": 5200 + }, + { + "grad_norm": 0.48253053426742554, + "learning_rate": 8.837239542966593e-05, + "loss": 0.0212, + "step": 5210 + }, + { + "grad_norm": 0.4505639970302582, + "learning_rate": 8.831934008971417e-05, + "loss": 0.0211, + "step": 5220 + }, + { + "grad_norm": 0.4596928656101227, + "learning_rate": 8.826617998614982e-05, + "loss": 0.0182, + "step": 5230 + }, + { + "grad_norm": 0.4037300646305084, + "learning_rate": 8.821291526431056e-05, + "loss": 0.0146, + "step": 5240 + }, + { + "grad_norm": 0.3879428505897522, + "learning_rate": 8.815954606982015e-05, + "loss": 0.0173, + "step": 5250 + }, + { + "grad_norm": 0.4592052102088928, + "learning_rate": 8.810607254858789e-05, + "loss": 0.0158, + "step": 5260 + }, + { + "grad_norm": 0.43258506059646606, + "learning_rate": 8.805249484680838e-05, + "loss": 0.0176, + "step": 5270 + }, + { + "grad_norm": 0.5278488993644714, + "learning_rate": 8.799881311096096e-05, + "loss": 0.0175, + "step": 5280 + }, + { + "grad_norm": 0.536786675453186, + "learning_rate": 8.794502748780949e-05, + "loss": 0.0191, + "step": 5290 + }, + { + "grad_norm": 0.39231422543525696, + "learning_rate": 8.78911381244018e-05, + "loss": 0.0213, + "step": 5300 + }, + { + "grad_norm": 0.3669500946998596, + "learning_rate": 8.783714516806933e-05, + "loss": 0.0227, + "step": 5310 + }, + { + "grad_norm": 0.4094989001750946, + "learning_rate": 8.77830487664268e-05, + "loss": 0.0239, + "step": 5320 + }, + { + "grad_norm": 0.3868635296821594, + "learning_rate": 8.772884906737167e-05, + "loss": 0.0174, + "step": 5330 + }, + { + "grad_norm": 0.4408529996871948, + "learning_rate": 8.767454621908387e-05, + "loss": 0.02, + "step": 5340 + }, + { + "grad_norm": 0.4143405258655548, + "learning_rate": 8.76201403700253e-05, + "loss": 0.0188, + "step": 5350 + }, + { + "grad_norm": 0.4451104402542114, + "learning_rate": 8.756563166893949e-05, + "loss": 0.0157, + "step": 5360 + }, + { + "grad_norm": 0.45531290769577026, + "learning_rate": 8.751102026485113e-05, + "loss": 0.0164, + "step": 5370 + }, + { + "grad_norm": 0.4830329418182373, + "learning_rate": 8.745630630706571e-05, + "loss": 0.0164, + "step": 5380 + }, + { + "grad_norm": 0.390863835811615, + "learning_rate": 8.740148994516912e-05, + "loss": 0.0157, + "step": 5390 + }, + { + "grad_norm": 0.32199040055274963, + "learning_rate": 8.73465713290272e-05, + "loss": 0.0147, + "step": 5400 + }, + { + "grad_norm": 0.4165448844432831, + "learning_rate": 8.729155060878533e-05, + "loss": 0.0162, + "step": 5410 + }, + { + "grad_norm": 0.3976683020591736, + "learning_rate": 8.723642793486809e-05, + "loss": 0.0155, + "step": 5420 + }, + { + "grad_norm": 0.38279062509536743, + "learning_rate": 8.718120345797873e-05, + "loss": 0.0187, + "step": 5430 + }, + { + "grad_norm": 0.3204955458641052, + "learning_rate": 8.712587732909889e-05, + "loss": 0.0173, + "step": 5440 + }, + { + "grad_norm": 0.4341621398925781, + "learning_rate": 8.707044969948806e-05, + "loss": 0.0164, + "step": 5450 + }, + { + "grad_norm": 0.38356325030326843, + "learning_rate": 8.701492072068329e-05, + "loss": 0.0171, + "step": 5460 + }, + { + "grad_norm": 0.45178765058517456, + "learning_rate": 8.695929054449869e-05, + "loss": 0.0159, + "step": 5470 + }, + { + "grad_norm": 0.34176012873649597, + "learning_rate": 8.690355932302501e-05, + "loss": 0.0191, + "step": 5480 + }, + { + "grad_norm": 0.4089793562889099, + "learning_rate": 8.684772720862931e-05, + "loss": 0.0171, + "step": 5490 + }, + { + "grad_norm": 0.48740607500076294, + "learning_rate": 8.679179435395446e-05, + "loss": 0.0156, + "step": 5500 + }, + { + "grad_norm": 0.45742067694664, + "learning_rate": 8.673576091191874e-05, + "loss": 0.018, + "step": 5510 + }, + { + "grad_norm": 0.334735244512558, + "learning_rate": 8.667962703571541e-05, + "loss": 0.0194, + "step": 5520 + }, + { + "grad_norm": 0.40391767024993896, + "learning_rate": 8.662339287881238e-05, + "loss": 0.0168, + "step": 5530 + }, + { + "grad_norm": 0.46624624729156494, + "learning_rate": 8.656705859495169e-05, + "loss": 0.0178, + "step": 5540 + }, + { + "grad_norm": 0.33784207701683044, + "learning_rate": 8.651062433814912e-05, + "loss": 0.0173, + "step": 5550 + }, + { + "grad_norm": 0.4161977469921112, + "learning_rate": 8.645409026269375e-05, + "loss": 0.0162, + "step": 5560 + }, + { + "grad_norm": 0.4053378701210022, + "learning_rate": 8.639745652314759e-05, + "loss": 0.0168, + "step": 5570 + }, + { + "grad_norm": 0.3393418490886688, + "learning_rate": 8.634072327434515e-05, + "loss": 0.0149, + "step": 5580 + }, + { + "grad_norm": 0.40152478218078613, + "learning_rate": 8.628389067139294e-05, + "loss": 0.0163, + "step": 5590 + }, + { + "grad_norm": 0.3247140347957611, + "learning_rate": 8.622695886966911e-05, + "loss": 0.0154, + "step": 5600 + }, + { + "grad_norm": 0.48909687995910645, + "learning_rate": 8.616992802482308e-05, + "loss": 0.0143, + "step": 5610 + }, + { + "grad_norm": 0.4496266841888428, + "learning_rate": 8.611279829277496e-05, + "loss": 0.0148, + "step": 5620 + }, + { + "grad_norm": 0.4135376513004303, + "learning_rate": 8.605556982971528e-05, + "loss": 0.0142, + "step": 5630 + }, + { + "grad_norm": 0.47809404134750366, + "learning_rate": 8.599824279210447e-05, + "loss": 0.0142, + "step": 5640 + }, + { + "grad_norm": 0.5461645126342773, + "learning_rate": 8.594081733667243e-05, + "loss": 0.0149, + "step": 5650 + }, + { + "grad_norm": 0.358334481716156, + "learning_rate": 8.58832936204182e-05, + "loss": 0.0227, + "step": 5660 + }, + { + "grad_norm": 0.5356438755989075, + "learning_rate": 8.582567180060942e-05, + "loss": 0.0153, + "step": 5670 + }, + { + "grad_norm": 0.5985286831855774, + "learning_rate": 8.576795203478194e-05, + "loss": 0.0173, + "step": 5680 + }, + { + "grad_norm": 0.44569990038871765, + "learning_rate": 8.571013448073939e-05, + "loss": 0.0193, + "step": 5690 + }, + { + "grad_norm": 0.4444361627101898, + "learning_rate": 8.565221929655275e-05, + "loss": 0.0216, + "step": 5700 + }, + { + "grad_norm": 0.3684118986129761, + "learning_rate": 8.559420664055992e-05, + "loss": 0.0207, + "step": 5710 + }, + { + "grad_norm": 0.45600274205207825, + "learning_rate": 8.553609667136532e-05, + "loss": 0.0202, + "step": 5720 + }, + { + "grad_norm": 0.45347586274147034, + "learning_rate": 8.547788954783936e-05, + "loss": 0.02, + "step": 5730 + }, + { + "grad_norm": 0.477226585149765, + "learning_rate": 8.541958542911808e-05, + "loss": 0.0232, + "step": 5740 + }, + { + "grad_norm": 0.4014797508716583, + "learning_rate": 8.536118447460275e-05, + "loss": 0.0206, + "step": 5750 + }, + { + "grad_norm": 0.3849544823169708, + "learning_rate": 8.530268684395932e-05, + "loss": 0.0188, + "step": 5760 + }, + { + "grad_norm": 0.43162935972213745, + "learning_rate": 8.524409269711807e-05, + "loss": 0.0216, + "step": 5770 + }, + { + "grad_norm": 0.3942539691925049, + "learning_rate": 8.51854021942732e-05, + "loss": 0.0196, + "step": 5780 + }, + { + "grad_norm": 0.4294084906578064, + "learning_rate": 8.512661549588227e-05, + "loss": 0.0192, + "step": 5790 + }, + { + "grad_norm": 0.45174744725227356, + "learning_rate": 8.506773276266588e-05, + "loss": 0.0182, + "step": 5800 + }, + { + "grad_norm": 0.39539965987205505, + "learning_rate": 8.500875415560721e-05, + "loss": 0.016, + "step": 5810 + }, + { + "grad_norm": 0.46286967396736145, + "learning_rate": 8.494967983595144e-05, + "loss": 0.0175, + "step": 5820 + }, + { + "grad_norm": 0.46430596709251404, + "learning_rate": 8.489050996520558e-05, + "loss": 0.0182, + "step": 5830 + }, + { + "grad_norm": 0.45418673753738403, + "learning_rate": 8.483124470513775e-05, + "loss": 0.0173, + "step": 5840 + }, + { + "grad_norm": 0.34853413701057434, + "learning_rate": 8.477188421777692e-05, + "loss": 0.0172, + "step": 5850 + }, + { + "grad_norm": 0.39042121171951294, + "learning_rate": 8.47124286654124e-05, + "loss": 0.018, + "step": 5860 + }, + { + "grad_norm": 0.40436068177223206, + "learning_rate": 8.465287821059341e-05, + "loss": 0.0186, + "step": 5870 + }, + { + "grad_norm": 0.46700358390808105, + "learning_rate": 8.45932330161286e-05, + "loss": 0.017, + "step": 5880 + }, + { + "grad_norm": 0.42868727445602417, + "learning_rate": 8.453349324508567e-05, + "loss": 0.015, + "step": 5890 + }, + { + "grad_norm": 0.49199414253234863, + "learning_rate": 8.447365906079088e-05, + "loss": 0.0168, + "step": 5900 + }, + { + "grad_norm": 0.35933196544647217, + "learning_rate": 8.441373062682856e-05, + "loss": 0.017, + "step": 5910 + }, + { + "grad_norm": 0.34801632165908813, + "learning_rate": 8.43537081070408e-05, + "loss": 0.0151, + "step": 5920 + }, + { + "grad_norm": 0.3106619715690613, + "learning_rate": 8.429359166552689e-05, + "loss": 0.0131, + "step": 5930 + }, + { + "grad_norm": 0.3804507255554199, + "learning_rate": 8.423338146664284e-05, + "loss": 0.0173, + "step": 5940 + }, + { + "grad_norm": 0.3459800183773041, + "learning_rate": 8.417307767500107e-05, + "loss": 0.0133, + "step": 5950 + }, + { + "grad_norm": 0.40529242157936096, + "learning_rate": 8.411268045546983e-05, + "loss": 0.0134, + "step": 5960 + }, + { + "grad_norm": 0.40981313586235046, + "learning_rate": 8.405218997317281e-05, + "loss": 0.0172, + "step": 5970 + }, + { + "grad_norm": 0.36945050954818726, + "learning_rate": 8.399160639348869e-05, + "loss": 0.0148, + "step": 5980 + }, + { + "grad_norm": 0.48703187704086304, + "learning_rate": 8.393092988205065e-05, + "loss": 0.0172, + "step": 5990 + }, + { + "grad_norm": 0.45176467299461365, + "learning_rate": 8.387016060474597e-05, + "loss": 0.0182, + "step": 6000 + }, + { + "grad_norm": 0.3710270822048187, + "learning_rate": 8.380929872771551e-05, + "loss": 0.0139, + "step": 6010 + }, + { + "grad_norm": 0.423194020986557, + "learning_rate": 8.374834441735335e-05, + "loss": 0.0141, + "step": 6020 + }, + { + "grad_norm": 0.40545573830604553, + "learning_rate": 8.368729784030622e-05, + "loss": 0.0136, + "step": 6030 + }, + { + "grad_norm": 0.2971954345703125, + "learning_rate": 8.362615916347315e-05, + "loss": 0.0158, + "step": 6040 + }, + { + "grad_norm": 0.3459785580635071, + "learning_rate": 8.356492855400493e-05, + "loss": 0.0151, + "step": 6050 + }, + { + "grad_norm": 0.3283204734325409, + "learning_rate": 8.350360617930371e-05, + "loss": 0.0135, + "step": 6060 + }, + { + "grad_norm": 0.4347304701805115, + "learning_rate": 8.344219220702255e-05, + "loss": 0.0155, + "step": 6070 + }, + { + "grad_norm": 0.4429846405982971, + "learning_rate": 8.338068680506485e-05, + "loss": 0.0165, + "step": 6080 + }, + { + "grad_norm": 0.45707541704177856, + "learning_rate": 8.33190901415841e-05, + "loss": 0.0132, + "step": 6090 + }, + { + "grad_norm": 0.4395197927951813, + "learning_rate": 8.325740238498317e-05, + "loss": 0.0168, + "step": 6100 + }, + { + "grad_norm": 0.4128543436527252, + "learning_rate": 8.319562370391406e-05, + "loss": 0.0157, + "step": 6110 + }, + { + "grad_norm": 0.4431177079677582, + "learning_rate": 8.31337542672773e-05, + "loss": 0.0195, + "step": 6120 + }, + { + "grad_norm": 0.4641300141811371, + "learning_rate": 8.307179424422158e-05, + "loss": 0.0164, + "step": 6130 + }, + { + "grad_norm": 0.4123786687850952, + "learning_rate": 8.300974380414327e-05, + "loss": 0.0147, + "step": 6140 + }, + { + "grad_norm": 0.42310845851898193, + "learning_rate": 8.294760311668586e-05, + "loss": 0.0184, + "step": 6150 + }, + { + "grad_norm": 0.4000084102153778, + "learning_rate": 8.288537235173961e-05, + "loss": 0.0149, + "step": 6160 + }, + { + "grad_norm": 0.4384673237800598, + "learning_rate": 8.282305167944108e-05, + "loss": 0.0174, + "step": 6170 + }, + { + "grad_norm": 0.2722078263759613, + "learning_rate": 8.276064127017262e-05, + "loss": 0.0154, + "step": 6180 + }, + { + "grad_norm": 0.31681233644485474, + "learning_rate": 8.269814129456189e-05, + "loss": 0.0139, + "step": 6190 + }, + { + "grad_norm": 0.3676801919937134, + "learning_rate": 8.263555192348143e-05, + "loss": 0.0145, + "step": 6200 + }, + { + "grad_norm": 0.38583555817604065, + "learning_rate": 8.257287332804819e-05, + "loss": 0.0148, + "step": 6210 + }, + { + "grad_norm": 0.38549646735191345, + "learning_rate": 8.251010567962307e-05, + "loss": 0.0157, + "step": 6220 + }, + { + "grad_norm": 0.38993483781814575, + "learning_rate": 8.244724914981041e-05, + "loss": 0.0151, + "step": 6230 + }, + { + "grad_norm": 0.4005882441997528, + "learning_rate": 8.238430391045757e-05, + "loss": 0.0148, + "step": 6240 + }, + { + "grad_norm": 0.34108078479766846, + "learning_rate": 8.232127013365445e-05, + "loss": 0.0176, + "step": 6250 + }, + { + "grad_norm": 0.4081633985042572, + "learning_rate": 8.225814799173295e-05, + "loss": 0.0181, + "step": 6260 + }, + { + "grad_norm": 0.24599479138851166, + "learning_rate": 8.219493765726663e-05, + "loss": 0.0136, + "step": 6270 + }, + { + "grad_norm": 0.4071696698665619, + "learning_rate": 8.21316393030701e-05, + "loss": 0.0159, + "step": 6280 + }, + { + "grad_norm": 0.41659772396087646, + "learning_rate": 8.206825310219865e-05, + "loss": 0.0133, + "step": 6290 + }, + { + "grad_norm": 0.4144860804080963, + "learning_rate": 8.200477922794776e-05, + "loss": 0.0178, + "step": 6300 + }, + { + "grad_norm": 0.3404088318347931, + "learning_rate": 8.194121785385256e-05, + "loss": 0.016, + "step": 6310 + }, + { + "grad_norm": 0.4269247055053711, + "learning_rate": 8.187756915368741e-05, + "loss": 0.0144, + "step": 6320 + }, + { + "grad_norm": 0.34299951791763306, + "learning_rate": 8.181383330146544e-05, + "loss": 0.0161, + "step": 6330 + }, + { + "grad_norm": 0.27657780051231384, + "learning_rate": 8.175001047143804e-05, + "loss": 0.0172, + "step": 6340 + }, + { + "grad_norm": 0.4395776391029358, + "learning_rate": 8.168610083809438e-05, + "loss": 0.0173, + "step": 6350 + }, + { + "grad_norm": 0.41414085030555725, + "learning_rate": 8.162210457616095e-05, + "loss": 0.0182, + "step": 6360 + }, + { + "grad_norm": 0.3814326822757721, + "learning_rate": 8.155802186060109e-05, + "loss": 0.0171, + "step": 6370 + }, + { + "grad_norm": 0.3817397654056549, + "learning_rate": 8.149385286661453e-05, + "loss": 0.0174, + "step": 6380 + }, + { + "grad_norm": 0.40936893224716187, + "learning_rate": 8.14295977696368e-05, + "loss": 0.0171, + "step": 6390 + }, + { + "grad_norm": 0.36244097352027893, + "learning_rate": 8.13652567453389e-05, + "loss": 0.0134, + "step": 6400 + }, + { + "grad_norm": 0.4146624803543091, + "learning_rate": 8.130082996962676e-05, + "loss": 0.0175, + "step": 6410 + }, + { + "grad_norm": 0.4154786169528961, + "learning_rate": 8.123631761864068e-05, + "loss": 0.0178, + "step": 6420 + }, + { + "grad_norm": 0.4006327986717224, + "learning_rate": 8.1171719868755e-05, + "loss": 0.0142, + "step": 6430 + }, + { + "grad_norm": 0.45199131965637207, + "learning_rate": 8.110703689657748e-05, + "loss": 0.0164, + "step": 6440 + }, + { + "grad_norm": 0.35770922899246216, + "learning_rate": 8.104226887894892e-05, + "loss": 0.0147, + "step": 6450 + }, + { + "grad_norm": 0.30330339074134827, + "learning_rate": 8.097741599294257e-05, + "loss": 0.0143, + "step": 6460 + }, + { + "grad_norm": 0.32697442173957825, + "learning_rate": 8.091247841586378e-05, + "loss": 0.0143, + "step": 6470 + }, + { + "grad_norm": 0.3217572867870331, + "learning_rate": 8.084745632524939e-05, + "loss": 0.016, + "step": 6480 + }, + { + "grad_norm": 0.42020222544670105, + "learning_rate": 8.07823498988673e-05, + "loss": 0.017, + "step": 6490 + }, + { + "grad_norm": 0.3531644642353058, + "learning_rate": 8.071715931471602e-05, + "loss": 0.0144, + "step": 6500 + }, + { + "grad_norm": 0.33470600843429565, + "learning_rate": 8.06518847510241e-05, + "loss": 0.0135, + "step": 6510 + }, + { + "grad_norm": 0.34689798951148987, + "learning_rate": 8.058652638624971e-05, + "loss": 0.0139, + "step": 6520 + }, + { + "grad_norm": 0.32701587677001953, + "learning_rate": 8.052108439908013e-05, + "loss": 0.0158, + "step": 6530 + }, + { + "grad_norm": 0.38156113028526306, + "learning_rate": 8.045555896843125e-05, + "loss": 0.0155, + "step": 6540 + }, + { + "grad_norm": 0.448223352432251, + "learning_rate": 8.03899502734471e-05, + "loss": 0.0157, + "step": 6550 + }, + { + "grad_norm": 0.5395865440368652, + "learning_rate": 8.032425849349931e-05, + "loss": 0.0158, + "step": 6560 + }, + { + "grad_norm": 0.5085408091545105, + "learning_rate": 8.025848380818674e-05, + "loss": 0.0194, + "step": 6570 + }, + { + "grad_norm": 0.3302910625934601, + "learning_rate": 8.019262639733487e-05, + "loss": 0.0141, + "step": 6580 + }, + { + "grad_norm": 0.5127670764923096, + "learning_rate": 8.012668644099531e-05, + "loss": 0.0159, + "step": 6590 + }, + { + "grad_norm": 0.36805611848831177, + "learning_rate": 8.006066411944542e-05, + "loss": 0.0183, + "step": 6600 + }, + { + "grad_norm": 0.35506105422973633, + "learning_rate": 7.999455961318769e-05, + "loss": 0.0196, + "step": 6610 + }, + { + "grad_norm": 0.3373585343360901, + "learning_rate": 7.992837310294932e-05, + "loss": 0.0125, + "step": 6620 + }, + { + "grad_norm": 0.4130808413028717, + "learning_rate": 7.986210476968167e-05, + "loss": 0.0174, + "step": 6630 + }, + { + "grad_norm": 0.455818772315979, + "learning_rate": 7.97957547945599e-05, + "loss": 0.0161, + "step": 6640 + }, + { + "grad_norm": 0.35132884979248047, + "learning_rate": 7.972932335898226e-05, + "loss": 0.0138, + "step": 6650 + }, + { + "grad_norm": 0.39049220085144043, + "learning_rate": 7.966281064456975e-05, + "loss": 0.0164, + "step": 6660 + }, + { + "grad_norm": 0.46879708766937256, + "learning_rate": 7.959621683316563e-05, + "loss": 0.0172, + "step": 6670 + }, + { + "grad_norm": 0.4622355103492737, + "learning_rate": 7.952954210683481e-05, + "loss": 0.0158, + "step": 6680 + }, + { + "grad_norm": 0.35040467977523804, + "learning_rate": 7.946278664786345e-05, + "loss": 0.015, + "step": 6690 + }, + { + "grad_norm": 0.36279842257499695, + "learning_rate": 7.939595063875842e-05, + "loss": 0.0144, + "step": 6700 + }, + { + "grad_norm": 0.3879755735397339, + "learning_rate": 7.932903426224683e-05, + "loss": 0.015, + "step": 6710 + }, + { + "grad_norm": 0.33389654755592346, + "learning_rate": 7.926203770127552e-05, + "loss": 0.0155, + "step": 6720 + }, + { + "grad_norm": 0.36863645911216736, + "learning_rate": 7.919496113901046e-05, + "loss": 0.0125, + "step": 6730 + }, + { + "grad_norm": 0.33035746216773987, + "learning_rate": 7.912780475883649e-05, + "loss": 0.0143, + "step": 6740 + }, + { + "grad_norm": 0.377729594707489, + "learning_rate": 7.906056874435652e-05, + "loss": 0.0125, + "step": 6750 + }, + { + "grad_norm": 0.311613529920578, + "learning_rate": 7.899325327939131e-05, + "loss": 0.0121, + "step": 6760 + }, + { + "grad_norm": 0.3309403955936432, + "learning_rate": 7.892585854797872e-05, + "loss": 0.0123, + "step": 6770 + }, + { + "grad_norm": 0.3694356083869934, + "learning_rate": 7.88583847343734e-05, + "loss": 0.0141, + "step": 6780 + }, + { + "grad_norm": 0.46677365899086, + "learning_rate": 7.879083202304616e-05, + "loss": 0.0134, + "step": 6790 + }, + { + "grad_norm": 0.3483867347240448, + "learning_rate": 7.872320059868355e-05, + "loss": 0.0143, + "step": 6800 + }, + { + "grad_norm": 0.30982255935668945, + "learning_rate": 7.865549064618729e-05, + "loss": 0.0135, + "step": 6810 + }, + { + "grad_norm": 0.3763626217842102, + "learning_rate": 7.858770235067381e-05, + "loss": 0.0155, + "step": 6820 + }, + { + "grad_norm": 0.43991050124168396, + "learning_rate": 7.851983589747374e-05, + "loss": 0.0126, + "step": 6830 + }, + { + "grad_norm": 0.4573172330856323, + "learning_rate": 7.845189147213133e-05, + "loss": 0.0154, + "step": 6840 + }, + { + "grad_norm": 0.5023182034492493, + "learning_rate": 7.838386926040407e-05, + "loss": 0.0163, + "step": 6850 + }, + { + "grad_norm": 0.4336654841899872, + "learning_rate": 7.83157694482621e-05, + "loss": 0.0173, + "step": 6860 + }, + { + "grad_norm": 0.2743140459060669, + "learning_rate": 7.824759222188768e-05, + "loss": 0.0125, + "step": 6870 + }, + { + "grad_norm": 0.3396323621273041, + "learning_rate": 7.817933776767478e-05, + "loss": 0.0165, + "step": 6880 + }, + { + "grad_norm": 0.35954833030700684, + "learning_rate": 7.811100627222842e-05, + "loss": 0.012, + "step": 6890 + }, + { + "grad_norm": 0.3257628381252289, + "learning_rate": 7.804259792236435e-05, + "loss": 0.0135, + "step": 6900 + }, + { + "grad_norm": 0.49320679903030396, + "learning_rate": 7.797411290510835e-05, + "loss": 0.0128, + "step": 6910 + }, + { + "grad_norm": 0.36475473642349243, + "learning_rate": 7.790555140769586e-05, + "loss": 0.0148, + "step": 6920 + }, + { + "grad_norm": 0.30804768204689026, + "learning_rate": 7.78369136175714e-05, + "loss": 0.0158, + "step": 6930 + }, + { + "grad_norm": 0.2891436219215393, + "learning_rate": 7.776819972238806e-05, + "loss": 0.0127, + "step": 6940 + }, + { + "grad_norm": 0.3640880584716797, + "learning_rate": 7.7699409910007e-05, + "loss": 0.0135, + "step": 6950 + }, + { + "grad_norm": 0.29762402176856995, + "learning_rate": 7.763054436849694e-05, + "loss": 0.0148, + "step": 6960 + }, + { + "grad_norm": 0.3800078332424164, + "learning_rate": 7.756160328613364e-05, + "loss": 0.0124, + "step": 6970 + }, + { + "grad_norm": 0.3265116810798645, + "learning_rate": 7.749258685139942e-05, + "loss": 0.0148, + "step": 6980 + }, + { + "grad_norm": 0.37737640738487244, + "learning_rate": 7.742349525298253e-05, + "loss": 0.0132, + "step": 6990 + }, + { + "grad_norm": 0.21071003377437592, + "learning_rate": 7.735432867977679e-05, + "loss": 0.0136, + "step": 7000 + }, + { + "grad_norm": 0.30626237392425537, + "learning_rate": 7.728508732088096e-05, + "loss": 0.0168, + "step": 7010 + }, + { + "grad_norm": 0.3753643035888672, + "learning_rate": 7.721577136559825e-05, + "loss": 0.0154, + "step": 7020 + }, + { + "grad_norm": 0.3369854986667633, + "learning_rate": 7.714638100343588e-05, + "loss": 0.0126, + "step": 7030 + }, + { + "grad_norm": 0.36830559372901917, + "learning_rate": 7.707691642410444e-05, + "loss": 0.014, + "step": 7040 + }, + { + "grad_norm": 0.4115355908870697, + "learning_rate": 7.70073778175174e-05, + "loss": 0.0148, + "step": 7050 + }, + { + "grad_norm": 0.2858520448207855, + "learning_rate": 7.69377653737907e-05, + "loss": 0.0139, + "step": 7060 + }, + { + "grad_norm": 0.38406574726104736, + "learning_rate": 7.686807928324209e-05, + "loss": 0.013, + "step": 7070 + }, + { + "grad_norm": 0.3661177456378937, + "learning_rate": 7.679831973639065e-05, + "loss": 0.0128, + "step": 7080 + }, + { + "grad_norm": 0.2673167586326599, + "learning_rate": 7.672848692395637e-05, + "loss": 0.0135, + "step": 7090 + }, + { + "grad_norm": 0.3895997703075409, + "learning_rate": 7.665858103685944e-05, + "loss": 0.0116, + "step": 7100 + }, + { + "grad_norm": 0.3720957934856415, + "learning_rate": 7.658860226621991e-05, + "loss": 0.0144, + "step": 7110 + }, + { + "grad_norm": 0.3137901723384857, + "learning_rate": 7.651855080335708e-05, + "loss": 0.0146, + "step": 7120 + }, + { + "grad_norm": 0.29563164710998535, + "learning_rate": 7.644842683978896e-05, + "loss": 0.0123, + "step": 7130 + }, + { + "grad_norm": 0.39437034726142883, + "learning_rate": 7.63782305672318e-05, + "loss": 0.0155, + "step": 7140 + }, + { + "grad_norm": 0.3885636627674103, + "learning_rate": 7.63079621775995e-05, + "loss": 0.0124, + "step": 7150 + }, + { + "grad_norm": 0.35217997431755066, + "learning_rate": 7.623762186300319e-05, + "loss": 0.0138, + "step": 7160 + }, + { + "grad_norm": 0.3057458698749542, + "learning_rate": 7.616720981575057e-05, + "loss": 0.0116, + "step": 7170 + }, + { + "grad_norm": 0.3529524505138397, + "learning_rate": 7.609672622834552e-05, + "loss": 0.0134, + "step": 7180 + }, + { + "grad_norm": 0.3752933442592621, + "learning_rate": 7.602617129348747e-05, + "loss": 0.0118, + "step": 7190 + }, + { + "grad_norm": 0.34898602962493896, + "learning_rate": 7.595554520407088e-05, + "loss": 0.0145, + "step": 7200 + }, + { + "grad_norm": 0.39017370343208313, + "learning_rate": 7.588484815318484e-05, + "loss": 0.0135, + "step": 7210 + }, + { + "grad_norm": 0.37441468238830566, + "learning_rate": 7.581408033411234e-05, + "loss": 0.0137, + "step": 7220 + }, + { + "grad_norm": 0.40523502230644226, + "learning_rate": 7.574324194032995e-05, + "loss": 0.0147, + "step": 7230 + }, + { + "grad_norm": 0.2746640145778656, + "learning_rate": 7.567233316550705e-05, + "loss": 0.0118, + "step": 7240 + }, + { + "grad_norm": 0.3495345115661621, + "learning_rate": 7.560135420350562e-05, + "loss": 0.0117, + "step": 7250 + }, + { + "grad_norm": 0.4313659965991974, + "learning_rate": 7.553030524837935e-05, + "loss": 0.0128, + "step": 7260 + }, + { + "grad_norm": 0.4218427538871765, + "learning_rate": 7.545918649437341e-05, + "loss": 0.0121, + "step": 7270 + }, + { + "grad_norm": 0.37825003266334534, + "learning_rate": 7.538799813592377e-05, + "loss": 0.0107, + "step": 7280 + }, + { + "grad_norm": 0.36705559492111206, + "learning_rate": 7.531674036765662e-05, + "loss": 0.0134, + "step": 7290 + }, + { + "grad_norm": 0.3276546001434326, + "learning_rate": 7.524541338438807e-05, + "loss": 0.0145, + "step": 7300 + }, + { + "grad_norm": 0.45257559418678284, + "learning_rate": 7.517401738112328e-05, + "loss": 0.0138, + "step": 7310 + }, + { + "grad_norm": 0.36200499534606934, + "learning_rate": 7.510255255305628e-05, + "loss": 0.0143, + "step": 7320 + }, + { + "grad_norm": 0.37191298604011536, + "learning_rate": 7.503101909556911e-05, + "loss": 0.0166, + "step": 7330 + }, + { + "grad_norm": 0.3438247740268707, + "learning_rate": 7.495941720423154e-05, + "loss": 0.0142, + "step": 7340 + }, + { + "grad_norm": 0.4581180810928345, + "learning_rate": 7.488774707480042e-05, + "loss": 0.0153, + "step": 7350 + }, + { + "grad_norm": 0.359670490026474, + "learning_rate": 7.481600890321911e-05, + "loss": 0.0122, + "step": 7360 + }, + { + "grad_norm": 0.46836525201797485, + "learning_rate": 7.474420288561708e-05, + "loss": 0.0135, + "step": 7370 + }, + { + "grad_norm": 0.33389872312545776, + "learning_rate": 7.467232921830921e-05, + "loss": 0.0134, + "step": 7380 + }, + { + "grad_norm": 0.33688122034072876, + "learning_rate": 7.460038809779537e-05, + "loss": 0.0136, + "step": 7390 + }, + { + "grad_norm": 0.41096070408821106, + "learning_rate": 7.452837972075983e-05, + "loss": 0.0127, + "step": 7400 + }, + { + "grad_norm": 0.41126012802124023, + "learning_rate": 7.445630428407074e-05, + "loss": 0.015, + "step": 7410 + }, + { + "grad_norm": 0.3702179789543152, + "learning_rate": 7.43841619847796e-05, + "loss": 0.0126, + "step": 7420 + }, + { + "grad_norm": 0.35743647813796997, + "learning_rate": 7.431195302012072e-05, + "loss": 0.0197, + "step": 7430 + }, + { + "grad_norm": 0.2985547184944153, + "learning_rate": 7.423967758751061e-05, + "loss": 0.0159, + "step": 7440 + }, + { + "grad_norm": 0.39072185754776, + "learning_rate": 7.416733588454758e-05, + "loss": 0.0154, + "step": 7450 + }, + { + "grad_norm": 0.4116431474685669, + "learning_rate": 7.409492810901106e-05, + "loss": 0.0163, + "step": 7460 + }, + { + "grad_norm": 0.4209757149219513, + "learning_rate": 7.402245445886116e-05, + "loss": 0.0197, + "step": 7470 + }, + { + "grad_norm": 0.348718523979187, + "learning_rate": 7.394991513223806e-05, + "loss": 0.0142, + "step": 7480 + }, + { + "grad_norm": 0.3451424837112427, + "learning_rate": 7.38773103274615e-05, + "loss": 0.0147, + "step": 7490 + }, + { + "grad_norm": 0.32023975253105164, + "learning_rate": 7.380464024303028e-05, + "loss": 0.0159, + "step": 7500 + }, + { + "grad_norm": 0.4120250344276428, + "learning_rate": 7.373190507762162e-05, + "loss": 0.0123, + "step": 7510 + }, + { + "grad_norm": 0.3678809106349945, + "learning_rate": 7.365910503009066e-05, + "loss": 0.0117, + "step": 7520 + }, + { + "grad_norm": 0.36750277876853943, + "learning_rate": 7.358624029946996e-05, + "loss": 0.0136, + "step": 7530 + }, + { + "grad_norm": 0.3163510262966156, + "learning_rate": 7.351331108496893e-05, + "loss": 0.014, + "step": 7540 + }, + { + "grad_norm": 0.4134257435798645, + "learning_rate": 7.344031758597325e-05, + "loss": 0.0133, + "step": 7550 + }, + { + "grad_norm": 0.392423540353775, + "learning_rate": 7.336726000204435e-05, + "loss": 0.0141, + "step": 7560 + }, + { + "grad_norm": 0.3356896638870239, + "learning_rate": 7.32941385329189e-05, + "loss": 0.0125, + "step": 7570 + }, + { + "grad_norm": 0.31061065196990967, + "learning_rate": 7.322095337850816e-05, + "loss": 0.0153, + "step": 7580 + }, + { + "grad_norm": 0.33278927206993103, + "learning_rate": 7.314770473889758e-05, + "loss": 0.0104, + "step": 7590 + }, + { + "grad_norm": 0.35223352909088135, + "learning_rate": 7.307439281434615e-05, + "loss": 0.0117, + "step": 7600 + }, + { + "grad_norm": 0.37879201769828796, + "learning_rate": 7.300101780528585e-05, + "loss": 0.0132, + "step": 7610 + }, + { + "grad_norm": 0.33135613799095154, + "learning_rate": 7.292757991232117e-05, + "loss": 0.0142, + "step": 7620 + }, + { + "grad_norm": 0.3964645564556122, + "learning_rate": 7.285407933622848e-05, + "loss": 0.012, + "step": 7630 + }, + { + "grad_norm": 0.3188134431838989, + "learning_rate": 7.278051627795557e-05, + "loss": 0.0182, + "step": 7640 + }, + { + "grad_norm": 0.36549797654151917, + "learning_rate": 7.270689093862105e-05, + "loss": 0.0146, + "step": 7650 + }, + { + "grad_norm": 0.3195458948612213, + "learning_rate": 7.263320351951374e-05, + "loss": 0.0105, + "step": 7660 + }, + { + "grad_norm": 0.3117104768753052, + "learning_rate": 7.255945422209227e-05, + "loss": 0.0127, + "step": 7670 + }, + { + "grad_norm": 0.31559836864471436, + "learning_rate": 7.248564324798437e-05, + "loss": 0.0139, + "step": 7680 + }, + { + "grad_norm": 0.2703145742416382, + "learning_rate": 7.241177079898644e-05, + "loss": 0.014, + "step": 7690 + }, + { + "grad_norm": 0.30897757411003113, + "learning_rate": 7.233783707706295e-05, + "loss": 0.0127, + "step": 7700 + }, + { + "grad_norm": 0.3173775374889374, + "learning_rate": 7.226384228434586e-05, + "loss": 0.0133, + "step": 7710 + }, + { + "grad_norm": 0.36799511313438416, + "learning_rate": 7.21897866231341e-05, + "loss": 0.0105, + "step": 7720 + }, + { + "grad_norm": 0.2746727764606476, + "learning_rate": 7.211567029589303e-05, + "loss": 0.0138, + "step": 7730 + }, + { + "grad_norm": 0.3362380266189575, + "learning_rate": 7.204149350525387e-05, + "loss": 0.0123, + "step": 7740 + }, + { + "grad_norm": 0.33995118737220764, + "learning_rate": 7.196725645401309e-05, + "loss": 0.0127, + "step": 7750 + }, + { + "grad_norm": 0.332607626914978, + "learning_rate": 7.1892959345132e-05, + "loss": 0.0139, + "step": 7760 + }, + { + "grad_norm": 0.2877747416496277, + "learning_rate": 7.181860238173605e-05, + "loss": 0.0111, + "step": 7770 + }, + { + "grad_norm": 0.30644649267196655, + "learning_rate": 7.174418576711432e-05, + "loss": 0.012, + "step": 7780 + }, + { + "grad_norm": 0.29927390813827515, + "learning_rate": 7.1669709704719e-05, + "loss": 0.0109, + "step": 7790 + }, + { + "grad_norm": 0.3413328528404236, + "learning_rate": 7.159517439816481e-05, + "loss": 0.0143, + "step": 7800 + }, + { + "grad_norm": 0.24794118106365204, + "learning_rate": 7.152058005122842e-05, + "loss": 0.0148, + "step": 7810 + }, + { + "grad_norm": 0.30149683356285095, + "learning_rate": 7.144592686784793e-05, + "loss": 0.0118, + "step": 7820 + }, + { + "grad_norm": 0.33201098442077637, + "learning_rate": 7.137121505212229e-05, + "loss": 0.0111, + "step": 7830 + }, + { + "grad_norm": 0.40392014384269714, + "learning_rate": 7.129644480831077e-05, + "loss": 0.0126, + "step": 7840 + }, + { + "grad_norm": 0.32815903425216675, + "learning_rate": 7.122161634083234e-05, + "loss": 0.0104, + "step": 7850 + }, + { + "grad_norm": 0.31929904222488403, + "learning_rate": 7.114672985426516e-05, + "loss": 0.0121, + "step": 7860 + }, + { + "grad_norm": 0.3497408926486969, + "learning_rate": 7.107178555334606e-05, + "loss": 0.0113, + "step": 7870 + }, + { + "grad_norm": 0.3377891778945923, + "learning_rate": 7.099678364296989e-05, + "loss": 0.0135, + "step": 7880 + }, + { + "grad_norm": 0.27795737981796265, + "learning_rate": 7.0921724328189e-05, + "loss": 0.0115, + "step": 7890 + }, + { + "grad_norm": 0.3064858615398407, + "learning_rate": 7.084660781421268e-05, + "loss": 0.0142, + "step": 7900 + }, + { + "grad_norm": 0.32354864478111267, + "learning_rate": 7.077143430640662e-05, + "loss": 0.0154, + "step": 7910 + }, + { + "grad_norm": 0.38219061493873596, + "learning_rate": 7.069620401029232e-05, + "loss": 0.0116, + "step": 7920 + }, + { + "grad_norm": 0.3809864819049835, + "learning_rate": 7.062091713154655e-05, + "loss": 0.0132, + "step": 7930 + }, + { + "grad_norm": 0.3125777542591095, + "learning_rate": 7.054557387600075e-05, + "loss": 0.0149, + "step": 7940 + }, + { + "grad_norm": 0.3452930152416229, + "learning_rate": 7.04701744496405e-05, + "loss": 0.0106, + "step": 7950 + }, + { + "grad_norm": 0.37026354670524597, + "learning_rate": 7.039471905860495e-05, + "loss": 0.0122, + "step": 7960 + }, + { + "grad_norm": 0.32580164074897766, + "learning_rate": 7.031920790918628e-05, + "loss": 0.013, + "step": 7970 + }, + { + "grad_norm": 0.288991242647171, + "learning_rate": 7.024364120782906e-05, + "loss": 0.0132, + "step": 7980 + }, + { + "grad_norm": 0.325033575296402, + "learning_rate": 7.016801916112978e-05, + "loss": 0.0102, + "step": 7990 + }, + { + "grad_norm": 0.27642202377319336, + "learning_rate": 7.009234197583623e-05, + "loss": 0.0111, + "step": 8000 + }, + { + "grad_norm": 0.3329485058784485, + "learning_rate": 7.001660985884692e-05, + "loss": 0.0123, + "step": 8010 + }, + { + "grad_norm": 0.3785879611968994, + "learning_rate": 6.994082301721063e-05, + "loss": 0.0151, + "step": 8020 + }, + { + "grad_norm": 0.34150007367134094, + "learning_rate": 6.986498165812563e-05, + "loss": 0.0117, + "step": 8030 + }, + { + "grad_norm": 0.439264178276062, + "learning_rate": 6.978908598893932e-05, + "loss": 0.0139, + "step": 8040 + }, + { + "grad_norm": 0.3500205874443054, + "learning_rate": 6.971313621714756e-05, + "loss": 0.0112, + "step": 8050 + }, + { + "grad_norm": 0.2777945399284363, + "learning_rate": 6.96371325503941e-05, + "loss": 0.011, + "step": 8060 + }, + { + "grad_norm": 0.382794052362442, + "learning_rate": 6.956107519647014e-05, + "loss": 0.0125, + "step": 8070 + }, + { + "grad_norm": 0.30284425616264343, + "learning_rate": 6.94849643633135e-05, + "loss": 0.0133, + "step": 8080 + }, + { + "grad_norm": 0.24802178144454956, + "learning_rate": 6.940880025900834e-05, + "loss": 0.0107, + "step": 8090 + }, + { + "grad_norm": 0.2981727719306946, + "learning_rate": 6.933258309178438e-05, + "loss": 0.0122, + "step": 8100 + }, + { + "grad_norm": 0.21943475306034088, + "learning_rate": 6.925631307001646e-05, + "loss": 0.0165, + "step": 8110 + }, + { + "grad_norm": 0.3769550621509552, + "learning_rate": 6.91799904022239e-05, + "loss": 0.0143, + "step": 8120 + }, + { + "grad_norm": 0.2876106798648834, + "learning_rate": 6.910361529706997e-05, + "loss": 0.0131, + "step": 8130 + }, + { + "grad_norm": 0.33470556139945984, + "learning_rate": 6.902718796336131e-05, + "loss": 0.0137, + "step": 8140 + }, + { + "grad_norm": 0.4283612370491028, + "learning_rate": 6.895070861004729e-05, + "loss": 0.0129, + "step": 8150 + }, + { + "grad_norm": 0.2765333652496338, + "learning_rate": 6.887417744621956e-05, + "loss": 0.0137, + "step": 8160 + }, + { + "grad_norm": 0.32283779978752136, + "learning_rate": 6.87975946811114e-05, + "loss": 0.0125, + "step": 8170 + }, + { + "grad_norm": 0.3863534927368164, + "learning_rate": 6.872096052409718e-05, + "loss": 0.0139, + "step": 8180 + }, + { + "grad_norm": 0.29665520787239075, + "learning_rate": 6.864427518469174e-05, + "loss": 0.0129, + "step": 8190 + }, + { + "grad_norm": 0.3554975986480713, + "learning_rate": 6.856753887254986e-05, + "loss": 0.0107, + "step": 8200 + }, + { + "grad_norm": 0.3163433074951172, + "learning_rate": 6.849075179746572e-05, + "loss": 0.0115, + "step": 8210 + }, + { + "grad_norm": 0.3252999782562256, + "learning_rate": 6.841391416937221e-05, + "loss": 0.01, + "step": 8220 + }, + { + "grad_norm": 0.38887593150138855, + "learning_rate": 6.833702619834053e-05, + "loss": 0.0107, + "step": 8230 + }, + { + "grad_norm": 0.3632328510284424, + "learning_rate": 6.82600880945794e-05, + "loss": 0.0106, + "step": 8240 + }, + { + "grad_norm": 0.2744136154651642, + "learning_rate": 6.818310006843468e-05, + "loss": 0.0108, + "step": 8250 + }, + { + "grad_norm": 0.3894250988960266, + "learning_rate": 6.810606233038868e-05, + "loss": 0.0117, + "step": 8260 + }, + { + "grad_norm": 0.3185424506664276, + "learning_rate": 6.802897509105966e-05, + "loss": 0.0136, + "step": 8270 + }, + { + "grad_norm": 0.28028440475463867, + "learning_rate": 6.79518385612012e-05, + "loss": 0.0129, + "step": 8280 + }, + { + "grad_norm": 0.25754812359809875, + "learning_rate": 6.787465295170157e-05, + "loss": 0.0121, + "step": 8290 + }, + { + "grad_norm": 0.2997465431690216, + "learning_rate": 6.779741847358332e-05, + "loss": 0.0108, + "step": 8300 + }, + { + "grad_norm": 0.3573647141456604, + "learning_rate": 6.772013533800256e-05, + "loss": 0.0133, + "step": 8310 + }, + { + "grad_norm": 0.3089340627193451, + "learning_rate": 6.764280375624843e-05, + "loss": 0.0116, + "step": 8320 + }, + { + "grad_norm": 0.31227874755859375, + "learning_rate": 6.756542393974252e-05, + "loss": 0.0108, + "step": 8330 + }, + { + "grad_norm": 0.381428986787796, + "learning_rate": 6.748799610003828e-05, + "loss": 0.0112, + "step": 8340 + }, + { + "grad_norm": 0.278055340051651, + "learning_rate": 6.741052044882048e-05, + "loss": 0.0117, + "step": 8350 + }, + { + "grad_norm": 0.2949462831020355, + "learning_rate": 6.73329971979046e-05, + "loss": 0.0108, + "step": 8360 + }, + { + "grad_norm": 0.3816727101802826, + "learning_rate": 6.725542655923625e-05, + "loss": 0.0116, + "step": 8370 + }, + { + "grad_norm": 0.2636021673679352, + "learning_rate": 6.717780874489057e-05, + "loss": 0.0128, + "step": 8380 + }, + { + "grad_norm": 0.38993480801582336, + "learning_rate": 6.710014396707172e-05, + "loss": 0.0119, + "step": 8390 + }, + { + "grad_norm": 0.3024043142795563, + "learning_rate": 6.702243243811221e-05, + "loss": 0.0109, + "step": 8400 + }, + { + "grad_norm": 0.36139100790023804, + "learning_rate": 6.694467437047244e-05, + "loss": 0.0118, + "step": 8410 + }, + { + "grad_norm": 0.34032484889030457, + "learning_rate": 6.686686997673997e-05, + "loss": 0.0122, + "step": 8420 + }, + { + "grad_norm": 0.3963174819946289, + "learning_rate": 6.678901946962903e-05, + "loss": 0.0104, + "step": 8430 + }, + { + "grad_norm": 0.2162044644355774, + "learning_rate": 6.671112306197996e-05, + "loss": 0.0124, + "step": 8440 + }, + { + "grad_norm": 0.2702636122703552, + "learning_rate": 6.663318096675854e-05, + "loss": 0.0125, + "step": 8450 + }, + { + "grad_norm": 0.42705151438713074, + "learning_rate": 6.655519339705552e-05, + "loss": 0.0118, + "step": 8460 + }, + { + "grad_norm": 0.29198169708251953, + "learning_rate": 6.647716056608588e-05, + "loss": 0.0132, + "step": 8470 + }, + { + "grad_norm": 0.24624188244342804, + "learning_rate": 6.639908268718843e-05, + "loss": 0.011, + "step": 8480 + }, + { + "grad_norm": 0.29472094774246216, + "learning_rate": 6.632095997382514e-05, + "loss": 0.0103, + "step": 8490 + }, + { + "grad_norm": 0.28369250893592834, + "learning_rate": 6.624279263958047e-05, + "loss": 0.0121, + "step": 8500 + }, + { + "grad_norm": 0.3822849690914154, + "learning_rate": 6.616458089816097e-05, + "loss": 0.0121, + "step": 8510 + }, + { + "grad_norm": 0.3733149766921997, + "learning_rate": 6.608632496339454e-05, + "loss": 0.0127, + "step": 8520 + }, + { + "grad_norm": 0.25586628913879395, + "learning_rate": 6.600802504922988e-05, + "loss": 0.0149, + "step": 8530 + }, + { + "grad_norm": 0.36027905344963074, + "learning_rate": 6.592968136973604e-05, + "loss": 0.0132, + "step": 8540 + }, + { + "grad_norm": 0.30015572905540466, + "learning_rate": 6.585129413910159e-05, + "loss": 0.0119, + "step": 8550 + }, + { + "grad_norm": 0.3715488016605377, + "learning_rate": 6.577286357163424e-05, + "loss": 0.0123, + "step": 8560 + }, + { + "grad_norm": 0.39481574296951294, + "learning_rate": 6.569438988176018e-05, + "loss": 0.0103, + "step": 8570 + }, + { + "grad_norm": 0.35289764404296875, + "learning_rate": 6.561587328402347e-05, + "loss": 0.0127, + "step": 8580 + }, + { + "grad_norm": 0.34876325726509094, + "learning_rate": 6.553731399308549e-05, + "loss": 0.0127, + "step": 8590 + }, + { + "grad_norm": 0.26030921936035156, + "learning_rate": 6.545871222372436e-05, + "loss": 0.011, + "step": 8600 + }, + { + "grad_norm": 0.3516961336135864, + "learning_rate": 6.538006819083426e-05, + "loss": 0.0115, + "step": 8610 + }, + { + "grad_norm": 0.3771360218524933, + "learning_rate": 6.530138210942505e-05, + "loss": 0.011, + "step": 8620 + }, + { + "grad_norm": 0.2810113728046417, + "learning_rate": 6.522265419462141e-05, + "loss": 0.0089, + "step": 8630 + }, + { + "grad_norm": 0.2637802064418793, + "learning_rate": 6.514388466166248e-05, + "loss": 0.0105, + "step": 8640 + }, + { + "grad_norm": 0.2962155342102051, + "learning_rate": 6.506507372590119e-05, + "loss": 0.0126, + "step": 8650 + }, + { + "grad_norm": 0.2789287269115448, + "learning_rate": 6.498622160280355e-05, + "loss": 0.0128, + "step": 8660 + }, + { + "grad_norm": 0.2849268317222595, + "learning_rate": 6.490732850794832e-05, + "loss": 0.0133, + "step": 8670 + }, + { + "grad_norm": 0.26055100560188293, + "learning_rate": 6.482839465702616e-05, + "loss": 0.0131, + "step": 8680 + }, + { + "grad_norm": 0.32526013255119324, + "learning_rate": 6.474942026583923e-05, + "loss": 0.0127, + "step": 8690 + }, + { + "grad_norm": 0.33485621213912964, + "learning_rate": 6.467040555030052e-05, + "loss": 0.0108, + "step": 8700 + }, + { + "grad_norm": 0.2516840994358063, + "learning_rate": 6.459135072643321e-05, + "loss": 0.0115, + "step": 8710 + }, + { + "grad_norm": 0.31756162643432617, + "learning_rate": 6.451225601037019e-05, + "loss": 0.0091, + "step": 8720 + }, + { + "grad_norm": 0.2860639691352844, + "learning_rate": 6.443312161835338e-05, + "loss": 0.0128, + "step": 8730 + }, + { + "grad_norm": 0.2940915822982788, + "learning_rate": 6.43539477667332e-05, + "loss": 0.0104, + "step": 8740 + }, + { + "grad_norm": 0.24016410112380981, + "learning_rate": 6.427473467196793e-05, + "loss": 0.0143, + "step": 8750 + }, + { + "grad_norm": 0.29369375109672546, + "learning_rate": 6.419548255062315e-05, + "loss": 0.0103, + "step": 8760 + }, + { + "grad_norm": 0.32075247168540955, + "learning_rate": 6.411619161937112e-05, + "loss": 0.0097, + "step": 8770 + }, + { + "grad_norm": 0.3262137174606323, + "learning_rate": 6.403686209499022e-05, + "loss": 0.0104, + "step": 8780 + }, + { + "grad_norm": 0.3471458852291107, + "learning_rate": 6.395749419436437e-05, + "loss": 0.0099, + "step": 8790 + }, + { + "grad_norm": 0.30499032139778137, + "learning_rate": 6.387808813448234e-05, + "loss": 0.0133, + "step": 8800 + }, + { + "grad_norm": 0.34709998965263367, + "learning_rate": 6.37986441324373e-05, + "loss": 0.0118, + "step": 8810 + }, + { + "grad_norm": 0.2934912443161011, + "learning_rate": 6.37191624054261e-05, + "loss": 0.0125, + "step": 8820 + }, + { + "grad_norm": 0.324332058429718, + "learning_rate": 6.363964317074872e-05, + "loss": 0.0154, + "step": 8830 + }, + { + "grad_norm": 0.2829475402832031, + "learning_rate": 6.356008664580776e-05, + "loss": 0.0134, + "step": 8840 + }, + { + "grad_norm": 0.2962392270565033, + "learning_rate": 6.348049304810771e-05, + "loss": 0.0142, + "step": 8850 + }, + { + "grad_norm": 0.3759327530860901, + "learning_rate": 6.340086259525442e-05, + "loss": 0.0129, + "step": 8860 + }, + { + "grad_norm": 0.36553898453712463, + "learning_rate": 6.332119550495448e-05, + "loss": 0.0131, + "step": 8870 + }, + { + "grad_norm": 0.33665499091148376, + "learning_rate": 6.324149199501473e-05, + "loss": 0.0113, + "step": 8880 + }, + { + "grad_norm": 0.3106566071510315, + "learning_rate": 6.316175228334146e-05, + "loss": 0.0128, + "step": 8890 + }, + { + "grad_norm": 0.3394717276096344, + "learning_rate": 6.308197658794003e-05, + "loss": 0.0089, + "step": 8900 + }, + { + "grad_norm": 0.30129188299179077, + "learning_rate": 6.300216512691417e-05, + "loss": 0.0114, + "step": 8910 + }, + { + "grad_norm": 0.3877800405025482, + "learning_rate": 6.292231811846532e-05, + "loss": 0.0112, + "step": 8920 + }, + { + "grad_norm": 0.31548699736595154, + "learning_rate": 6.284243578089217e-05, + "loss": 0.0124, + "step": 8930 + }, + { + "grad_norm": 0.286324143409729, + "learning_rate": 6.276251833258999e-05, + "loss": 0.0104, + "step": 8940 + }, + { + "grad_norm": 0.31045469641685486, + "learning_rate": 6.268256599205003e-05, + "loss": 0.0124, + "step": 8950 + }, + { + "grad_norm": 0.31697219610214233, + "learning_rate": 6.260257897785892e-05, + "loss": 0.0137, + "step": 8960 + }, + { + "grad_norm": 0.37782853841781616, + "learning_rate": 6.252255750869811e-05, + "loss": 0.0118, + "step": 8970 + }, + { + "grad_norm": 0.30432772636413574, + "learning_rate": 6.244250180334325e-05, + "loss": 0.0128, + "step": 8980 + }, + { + "grad_norm": 0.3652268052101135, + "learning_rate": 6.236241208066356e-05, + "loss": 0.0127, + "step": 8990 + }, + { + "grad_norm": 0.2324247658252716, + "learning_rate": 6.228228855962133e-05, + "loss": 0.0162, + "step": 9000 + }, + { + "grad_norm": 0.31128183007240295, + "learning_rate": 6.220213145927115e-05, + "loss": 0.0135, + "step": 9010 + }, + { + "grad_norm": 0.32616373896598816, + "learning_rate": 6.212194099875951e-05, + "loss": 0.0101, + "step": 9020 + }, + { + "grad_norm": 0.25326231122016907, + "learning_rate": 6.204171739732405e-05, + "loss": 0.0133, + "step": 9030 + }, + { + "grad_norm": 0.3350052237510681, + "learning_rate": 6.196146087429303e-05, + "loss": 0.0134, + "step": 9040 + }, + { + "grad_norm": 0.40826472640037537, + "learning_rate": 6.188117164908474e-05, + "loss": 0.0112, + "step": 9050 + }, + { + "grad_norm": 0.2690710127353668, + "learning_rate": 6.180084994120684e-05, + "loss": 0.0116, + "step": 9060 + }, + { + "grad_norm": 0.20586135983467102, + "learning_rate": 6.17204959702558e-05, + "loss": 0.012, + "step": 9070 + }, + { + "grad_norm": 0.30860739946365356, + "learning_rate": 6.164010995591635e-05, + "loss": 0.0128, + "step": 9080 + }, + { + "grad_norm": 0.36940038204193115, + "learning_rate": 6.155969211796076e-05, + "loss": 0.0118, + "step": 9090 + }, + { + "grad_norm": 0.32097283005714417, + "learning_rate": 6.147924267624829e-05, + "loss": 0.0122, + "step": 9100 + }, + { + "grad_norm": 0.3452341556549072, + "learning_rate": 6.13987618507247e-05, + "loss": 0.0124, + "step": 9110 + }, + { + "grad_norm": 0.25516965985298157, + "learning_rate": 6.131824986142147e-05, + "loss": 0.0123, + "step": 9120 + }, + { + "grad_norm": 0.3512190878391266, + "learning_rate": 6.123770692845529e-05, + "loss": 0.0108, + "step": 9130 + }, + { + "grad_norm": 0.3161102831363678, + "learning_rate": 6.11571332720275e-05, + "loss": 0.0121, + "step": 9140 + }, + { + "grad_norm": 0.2909663915634155, + "learning_rate": 6.107652911242336e-05, + "loss": 0.0099, + "step": 9150 + }, + { + "grad_norm": 0.3107944130897522, + "learning_rate": 6.0995894670011586e-05, + "loss": 0.0108, + "step": 9160 + }, + { + "grad_norm": 0.2862789034843445, + "learning_rate": 6.091523016524368e-05, + "loss": 0.0121, + "step": 9170 + }, + { + "grad_norm": 0.3161313235759735, + "learning_rate": 6.083453581865328e-05, + "loss": 0.0105, + "step": 9180 + }, + { + "grad_norm": 0.2773940861225128, + "learning_rate": 6.075381185085568e-05, + "loss": 0.0108, + "step": 9190 + }, + { + "grad_norm": 0.2752029299736023, + "learning_rate": 6.067305848254709e-05, + "loss": 0.0115, + "step": 9200 + }, + { + "grad_norm": 0.35372427105903625, + "learning_rate": 6.059227593450418e-05, + "loss": 0.0119, + "step": 9210 + }, + { + "grad_norm": 0.3142034411430359, + "learning_rate": 6.051146442758333e-05, + "loss": 0.0125, + "step": 9220 + }, + { + "grad_norm": 0.2529461979866028, + "learning_rate": 6.043062418272012e-05, + "loss": 0.0107, + "step": 9230 + }, + { + "grad_norm": 0.3262726962566376, + "learning_rate": 6.0349755420928666e-05, + "loss": 0.0127, + "step": 9240 + }, + { + "grad_norm": 0.26650935411453247, + "learning_rate": 6.0268858363301105e-05, + "loss": 0.0106, + "step": 9250 + }, + { + "grad_norm": 0.3812367916107178, + "learning_rate": 6.018793323100689e-05, + "loss": 0.0104, + "step": 9260 + }, + { + "grad_norm": 0.2846297025680542, + "learning_rate": 6.0106980245292255e-05, + "loss": 0.0116, + "step": 9270 + }, + { + "grad_norm": 0.3379952907562256, + "learning_rate": 6.002599962747957e-05, + "loss": 0.0131, + "step": 9280 + }, + { + "grad_norm": 0.28340259194374084, + "learning_rate": 5.994499159896673e-05, + "loss": 0.0109, + "step": 9290 + }, + { + "grad_norm": 0.32905933260917664, + "learning_rate": 5.9863956381226607e-05, + "loss": 0.0127, + "step": 9300 + }, + { + "grad_norm": 0.34781473875045776, + "learning_rate": 5.9782894195806394e-05, + "loss": 0.0097, + "step": 9310 + }, + { + "grad_norm": 0.3538030683994293, + "learning_rate": 5.9701805264327004e-05, + "loss": 0.0101, + "step": 9320 + }, + { + "grad_norm": 0.3605461120605469, + "learning_rate": 5.96206898084825e-05, + "loss": 0.0102, + "step": 9330 + }, + { + "grad_norm": 0.29366081953048706, + "learning_rate": 5.953954805003942e-05, + "loss": 0.0105, + "step": 9340 + }, + { + "grad_norm": 0.26942163705825806, + "learning_rate": 5.945838021083623e-05, + "loss": 0.0119, + "step": 9350 + }, + { + "grad_norm": 0.3048942983150482, + "learning_rate": 5.9377186512782714e-05, + "loss": 0.0104, + "step": 9360 + }, + { + "grad_norm": 0.3276144862174988, + "learning_rate": 5.929596717785935e-05, + "loss": 0.0092, + "step": 9370 + }, + { + "grad_norm": 0.327134370803833, + "learning_rate": 5.921472242811668e-05, + "loss": 0.0106, + "step": 9380 + }, + { + "grad_norm": 0.2601287066936493, + "learning_rate": 5.913345248567475e-05, + "loss": 0.0095, + "step": 9390 + }, + { + "grad_norm": 0.3127315044403076, + "learning_rate": 5.905215757272248e-05, + "loss": 0.0113, + "step": 9400 + }, + { + "grad_norm": 0.2669365406036377, + "learning_rate": 5.897083791151706e-05, + "loss": 0.0082, + "step": 9410 + }, + { + "grad_norm": 0.3623141348361969, + "learning_rate": 5.888949372438336e-05, + "loss": 0.0114, + "step": 9420 + }, + { + "grad_norm": 0.28753525018692017, + "learning_rate": 5.8808125233713255e-05, + "loss": 0.0123, + "step": 9430 + }, + { + "grad_norm": 0.2539946436882019, + "learning_rate": 5.872673266196509e-05, + "loss": 0.0113, + "step": 9440 + }, + { + "grad_norm": 0.30448299646377563, + "learning_rate": 5.864531623166305e-05, + "loss": 0.0131, + "step": 9450 + }, + { + "grad_norm": 0.35516154766082764, + "learning_rate": 5.856387616539656e-05, + "loss": 0.0103, + "step": 9460 + }, + { + "grad_norm": 0.2838006615638733, + "learning_rate": 5.848241268581967e-05, + "loss": 0.0099, + "step": 9470 + }, + { + "grad_norm": 0.2405579686164856, + "learning_rate": 5.840092601565037e-05, + "loss": 0.0117, + "step": 9480 + }, + { + "grad_norm": 0.2804074287414551, + "learning_rate": 5.8319416377670144e-05, + "loss": 0.0104, + "step": 9490 + }, + { + "grad_norm": 0.3132031559944153, + "learning_rate": 5.82378839947232e-05, + "loss": 0.0102, + "step": 9500 + }, + { + "grad_norm": 0.2594463527202606, + "learning_rate": 5.815632908971599e-05, + "loss": 0.0123, + "step": 9510 + }, + { + "grad_norm": 0.3764408230781555, + "learning_rate": 5.80747518856165e-05, + "loss": 0.0088, + "step": 9520 + }, + { + "grad_norm": 0.3109181821346283, + "learning_rate": 5.799315260545367e-05, + "loss": 0.0078, + "step": 9530 + }, + { + "grad_norm": 0.26161524653434753, + "learning_rate": 5.791153147231686e-05, + "loss": 0.01, + "step": 9540 + }, + { + "grad_norm": 0.38785234093666077, + "learning_rate": 5.782988870935509e-05, + "loss": 0.0135, + "step": 9550 + }, + { + "grad_norm": 0.3314606547355652, + "learning_rate": 5.774822453977657e-05, + "loss": 0.0114, + "step": 9560 + }, + { + "grad_norm": 0.2779804766178131, + "learning_rate": 5.7666539186848036e-05, + "loss": 0.011, + "step": 9570 + }, + { + "grad_norm": 0.2708481252193451, + "learning_rate": 5.758483287389411e-05, + "loss": 0.0128, + "step": 9580 + }, + { + "grad_norm": 0.27679356932640076, + "learning_rate": 5.7503105824296735e-05, + "loss": 0.0128, + "step": 9590 + }, + { + "grad_norm": 0.2817783057689667, + "learning_rate": 5.742135826149453e-05, + "loss": 0.0091, + "step": 9600 + }, + { + "grad_norm": 0.2382742166519165, + "learning_rate": 5.7339590408982223e-05, + "loss": 0.0115, + "step": 9610 + }, + { + "grad_norm": 0.23625558614730835, + "learning_rate": 5.725780249031e-05, + "loss": 0.0086, + "step": 9620 + }, + { + "grad_norm": 0.30190321803092957, + "learning_rate": 5.717599472908292e-05, + "loss": 0.0097, + "step": 9630 + }, + { + "grad_norm": 0.31747740507125854, + "learning_rate": 5.7094167348960237e-05, + "loss": 0.0129, + "step": 9640 + }, + { + "grad_norm": 0.27004364132881165, + "learning_rate": 5.7012320573654945e-05, + "loss": 0.0126, + "step": 9650 + }, + { + "grad_norm": 0.2586357593536377, + "learning_rate": 5.693045462693295e-05, + "loss": 0.0124, + "step": 9660 + }, + { + "grad_norm": 0.2702503502368927, + "learning_rate": 5.684856973261266e-05, + "loss": 0.009, + "step": 9670 + }, + { + "grad_norm": 0.33263716101646423, + "learning_rate": 5.6766666114564215e-05, + "loss": 0.0087, + "step": 9680 + }, + { + "grad_norm": 0.3644093871116638, + "learning_rate": 5.668474399670899e-05, + "loss": 0.0119, + "step": 9690 + }, + { + "grad_norm": 0.29676637053489685, + "learning_rate": 5.660280360301896e-05, + "loss": 0.0115, + "step": 9700 + }, + { + "grad_norm": 0.2950310707092285, + "learning_rate": 5.652084515751599e-05, + "loss": 0.0098, + "step": 9710 + }, + { + "grad_norm": 0.30780714750289917, + "learning_rate": 5.643886888427137e-05, + "loss": 0.0105, + "step": 9720 + }, + { + "grad_norm": 0.31481608748435974, + "learning_rate": 5.6356875007405074e-05, + "loss": 0.0101, + "step": 9730 + }, + { + "grad_norm": 0.2832598090171814, + "learning_rate": 5.627486375108525e-05, + "loss": 0.0113, + "step": 9740 + }, + { + "grad_norm": 0.3724987208843231, + "learning_rate": 5.619283533952754e-05, + "loss": 0.0122, + "step": 9750 + }, + { + "grad_norm": 0.47322213649749756, + "learning_rate": 5.6110789996994474e-05, + "loss": 0.0114, + "step": 9760 + }, + { + "grad_norm": 0.31387490034103394, + "learning_rate": 5.602872794779491e-05, + "loss": 0.0097, + "step": 9770 + }, + { + "grad_norm": 0.33728551864624023, + "learning_rate": 5.594664941628334e-05, + "loss": 0.0084, + "step": 9780 + }, + { + "grad_norm": 0.2974473237991333, + "learning_rate": 5.5864554626859324e-05, + "loss": 0.0096, + "step": 9790 + }, + { + "grad_norm": 0.2743162214756012, + "learning_rate": 5.578244380396691e-05, + "loss": 0.0102, + "step": 9800 + }, + { + "grad_norm": 0.2396203577518463, + "learning_rate": 5.570031717209394e-05, + "loss": 0.0172, + "step": 9810 + }, + { + "grad_norm": 0.3160618245601654, + "learning_rate": 5.561817495577147e-05, + "loss": 0.0117, + "step": 9820 + }, + { + "grad_norm": 0.31184670329093933, + "learning_rate": 5.5536017379573215e-05, + "loss": 0.0131, + "step": 9830 + }, + { + "grad_norm": 0.2700205445289612, + "learning_rate": 5.545384466811483e-05, + "loss": 0.0131, + "step": 9840 + }, + { + "grad_norm": 0.29386454820632935, + "learning_rate": 5.5371657046053384e-05, + "loss": 0.013, + "step": 9850 + }, + { + "grad_norm": 0.30857086181640625, + "learning_rate": 5.528945473808669e-05, + "loss": 0.0112, + "step": 9860 + }, + { + "grad_norm": 0.31214991211891174, + "learning_rate": 5.520723796895272e-05, + "loss": 0.0082, + "step": 9870 + }, + { + "grad_norm": 0.3384791910648346, + "learning_rate": 5.512500696342897e-05, + "loss": 0.0087, + "step": 9880 + }, + { + "grad_norm": 0.26187098026275635, + "learning_rate": 5.504276194633188e-05, + "loss": 0.0093, + "step": 9890 + }, + { + "grad_norm": 0.3037872314453125, + "learning_rate": 5.49605031425162e-05, + "loss": 0.0092, + "step": 9900 + }, + { + "grad_norm": 0.28506723046302795, + "learning_rate": 5.487823077687434e-05, + "loss": 0.0099, + "step": 9910 + }, + { + "grad_norm": 0.3275642693042755, + "learning_rate": 5.4795945074335806e-05, + "loss": 0.009, + "step": 9920 + }, + { + "grad_norm": 0.29687318205833435, + "learning_rate": 5.471364625986657e-05, + "loss": 0.0112, + "step": 9930 + }, + { + "grad_norm": 0.24183917045593262, + "learning_rate": 5.463133455846845e-05, + "loss": 0.0088, + "step": 9940 + }, + { + "grad_norm": 0.32168254256248474, + "learning_rate": 5.4549010195178505e-05, + "loss": 0.0141, + "step": 9950 + }, + { + "grad_norm": 0.2662130892276764, + "learning_rate": 5.446667339506838e-05, + "loss": 0.0104, + "step": 9960 + }, + { + "grad_norm": 0.32166045904159546, + "learning_rate": 5.4384324383243756e-05, + "loss": 0.0098, + "step": 9970 + }, + { + "grad_norm": 0.22883684933185577, + "learning_rate": 5.430196338484368e-05, + "loss": 0.0078, + "step": 9980 + }, + { + "grad_norm": 0.2235536277294159, + "learning_rate": 5.4219590625039975e-05, + "loss": 0.0097, + "step": 9990 + }, + { + "grad_norm": 0.2653367221355438, + "learning_rate": 5.413720632903664e-05, + "loss": 0.0103, + "step": 10000 + }, + { + "grad_norm": 0.25344082713127136, + "learning_rate": 5.405481072206917e-05, + "loss": 0.0089, + "step": 10010 + }, + { + "grad_norm": 0.335708886384964, + "learning_rate": 5.397240402940402e-05, + "loss": 0.0102, + "step": 10020 + }, + { + "grad_norm": 0.2804220914840698, + "learning_rate": 5.388998647633794e-05, + "loss": 0.0119, + "step": 10030 + }, + { + "grad_norm": 0.3300700783729553, + "learning_rate": 5.380755828819737e-05, + "loss": 0.009, + "step": 10040 + }, + { + "grad_norm": 0.3184926211833954, + "learning_rate": 5.3725119690337846e-05, + "loss": 0.01, + "step": 10050 + }, + { + "grad_norm": 0.2888861894607544, + "learning_rate": 5.3642670908143324e-05, + "loss": 0.0085, + "step": 10060 + }, + { + "grad_norm": 0.2378275841474533, + "learning_rate": 5.356021216702562e-05, + "loss": 0.0083, + "step": 10070 + }, + { + "grad_norm": 0.30194157361984253, + "learning_rate": 5.347774369242381e-05, + "loss": 0.0144, + "step": 10080 + }, + { + "grad_norm": 0.30574026703834534, + "learning_rate": 5.3395265709803545e-05, + "loss": 0.0106, + "step": 10090 + }, + { + "grad_norm": 0.2707749307155609, + "learning_rate": 5.331277844465647e-05, + "loss": 0.009, + "step": 10100 + }, + { + "grad_norm": 0.2563587427139282, + "learning_rate": 5.323028212249963e-05, + "loss": 0.0079, + "step": 10110 + }, + { + "grad_norm": 0.23816479742527008, + "learning_rate": 5.314777696887481e-05, + "loss": 0.0092, + "step": 10120 + }, + { + "grad_norm": 0.2609695494174957, + "learning_rate": 5.306526320934796e-05, + "loss": 0.0115, + "step": 10130 + }, + { + "grad_norm": 0.32091790437698364, + "learning_rate": 5.298274106950854e-05, + "loss": 0.009, + "step": 10140 + }, + { + "grad_norm": 0.352911114692688, + "learning_rate": 5.290021077496893e-05, + "loss": 0.0088, + "step": 10150 + }, + { + "grad_norm": 0.2584521174430847, + "learning_rate": 5.2817672551363816e-05, + "loss": 0.0098, + "step": 10160 + }, + { + "grad_norm": 0.25784599781036377, + "learning_rate": 5.273512662434952e-05, + "loss": 0.0107, + "step": 10170 + }, + { + "grad_norm": 0.3719595670700073, + "learning_rate": 5.265257321960349e-05, + "loss": 0.011, + "step": 10180 + }, + { + "grad_norm": 0.30774036049842834, + "learning_rate": 5.257001256282357e-05, + "loss": 0.0096, + "step": 10190 + }, + { + "grad_norm": 0.3356902599334717, + "learning_rate": 5.248744487972742e-05, + "loss": 0.0094, + "step": 10200 + }, + { + "grad_norm": 0.3237636685371399, + "learning_rate": 5.240487039605196e-05, + "loss": 0.0113, + "step": 10210 + }, + { + "grad_norm": 0.3064771890640259, + "learning_rate": 5.232228933755267e-05, + "loss": 0.0128, + "step": 10220 + }, + { + "grad_norm": 0.3693506419658661, + "learning_rate": 5.2239701930003006e-05, + "loss": 0.0099, + "step": 10230 + }, + { + "grad_norm": 0.2279125452041626, + "learning_rate": 5.215710839919379e-05, + "loss": 0.0092, + "step": 10240 + }, + { + "grad_norm": 0.2967457175254822, + "learning_rate": 5.207450897093257e-05, + "loss": 0.0082, + "step": 10250 + }, + { + "grad_norm": 0.2727642059326172, + "learning_rate": 5.1991903871043046e-05, + "loss": 0.0125, + "step": 10260 + }, + { + "grad_norm": 0.3166670501232147, + "learning_rate": 5.190929332536439e-05, + "loss": 0.0109, + "step": 10270 + }, + { + "grad_norm": 0.3417263329029083, + "learning_rate": 5.182667755975071e-05, + "loss": 0.0121, + "step": 10280 + }, + { + "grad_norm": 0.24792048335075378, + "learning_rate": 5.1744056800070315e-05, + "loss": 0.0089, + "step": 10290 + }, + { + "grad_norm": 0.28874701261520386, + "learning_rate": 5.166143127220524e-05, + "loss": 0.0099, + "step": 10300 + }, + { + "grad_norm": 0.33919915556907654, + "learning_rate": 5.1578801202050485e-05, + "loss": 0.0092, + "step": 10310 + }, + { + "grad_norm": 0.24734066426753998, + "learning_rate": 5.149616681551355e-05, + "loss": 0.0084, + "step": 10320 + }, + { + "grad_norm": 0.29600366950035095, + "learning_rate": 5.141352833851367e-05, + "loss": 0.011, + "step": 10330 + }, + { + "grad_norm": 0.26344043016433716, + "learning_rate": 5.1330885996981285e-05, + "loss": 0.0084, + "step": 10340 + }, + { + "grad_norm": 0.29486021399497986, + "learning_rate": 5.124824001685741e-05, + "loss": 0.0128, + "step": 10350 + }, + { + "grad_norm": 0.3321808874607086, + "learning_rate": 5.116559062409298e-05, + "loss": 0.0108, + "step": 10360 + }, + { + "grad_norm": 0.3323744535446167, + "learning_rate": 5.10829380446483e-05, + "loss": 0.0102, + "step": 10370 + }, + { + "grad_norm": 0.2918391227722168, + "learning_rate": 5.100028250449235e-05, + "loss": 0.0086, + "step": 10380 + }, + { + "grad_norm": 0.2917460501194, + "learning_rate": 5.0917624229602234e-05, + "loss": 0.0092, + "step": 10390 + }, + { + "grad_norm": 0.24862314760684967, + "learning_rate": 5.0834963445962524e-05, + "loss": 0.0092, + "step": 10400 + }, + { + "grad_norm": 0.3336447775363922, + "learning_rate": 5.075230037956461e-05, + "loss": 0.0089, + "step": 10410 + }, + { + "grad_norm": 0.2436840683221817, + "learning_rate": 5.0669635256406213e-05, + "loss": 0.0074, + "step": 10420 + }, + { + "grad_norm": 0.23710206151008606, + "learning_rate": 5.058696830249058e-05, + "loss": 0.0109, + "step": 10430 + }, + { + "grad_norm": 0.2947542071342468, + "learning_rate": 5.050429974382602e-05, + "loss": 0.0099, + "step": 10440 + }, + { + "grad_norm": 0.28282612562179565, + "learning_rate": 5.042162980642523e-05, + "loss": 0.0083, + "step": 10450 + }, + { + "grad_norm": 0.2697611451148987, + "learning_rate": 5.033895871630462e-05, + "loss": 0.0096, + "step": 10460 + }, + { + "grad_norm": 0.28840044140815735, + "learning_rate": 5.025628669948386e-05, + "loss": 0.0099, + "step": 10470 + }, + { + "grad_norm": 0.26551976799964905, + "learning_rate": 5.017361398198502e-05, + "loss": 0.0095, + "step": 10480 + }, + { + "grad_norm": 0.344992458820343, + "learning_rate": 5.009094078983221e-05, + "loss": 0.009, + "step": 10490 + }, + { + "grad_norm": 0.33870258927345276, + "learning_rate": 5.000826734905073e-05, + "loss": 0.0086, + "step": 10500 + }, + { + "grad_norm": 0.26343870162963867, + "learning_rate": 4.9925593885666645e-05, + "loss": 0.0095, + "step": 10510 + }, + { + "grad_norm": 0.30816709995269775, + "learning_rate": 4.984292062570602e-05, + "loss": 0.0095, + "step": 10520 + }, + { + "grad_norm": 0.24630755186080933, + "learning_rate": 4.976024779519442e-05, + "loss": 0.009, + "step": 10530 + }, + { + "grad_norm": 0.2802034616470337, + "learning_rate": 4.9677575620156194e-05, + "loss": 0.0094, + "step": 10540 + }, + { + "grad_norm": 0.41168075799942017, + "learning_rate": 4.959490432661391e-05, + "loss": 0.012, + "step": 10550 + }, + { + "grad_norm": 0.2359355241060257, + "learning_rate": 4.9512234140587726e-05, + "loss": 0.0089, + "step": 10560 + }, + { + "grad_norm": 0.26179102063179016, + "learning_rate": 4.942956528809477e-05, + "loss": 0.011, + "step": 10570 + }, + { + "grad_norm": 0.21987201273441315, + "learning_rate": 4.934689799514854e-05, + "loss": 0.008, + "step": 10580 + }, + { + "grad_norm": 0.2534409165382385, + "learning_rate": 4.926423248775827e-05, + "loss": 0.0107, + "step": 10590 + }, + { + "grad_norm": 0.32994967699050903, + "learning_rate": 4.918156899192826e-05, + "loss": 0.0106, + "step": 10600 + }, + { + "grad_norm": 0.19284388422966003, + "learning_rate": 4.909890773365738e-05, + "loss": 0.0117, + "step": 10610 + }, + { + "grad_norm": 0.2593536078929901, + "learning_rate": 4.9016248938938344e-05, + "loss": 0.0125, + "step": 10620 + }, + { + "grad_norm": 0.34668925404548645, + "learning_rate": 4.8933592833757156e-05, + "loss": 0.0087, + "step": 10630 + }, + { + "grad_norm": 0.30534157156944275, + "learning_rate": 4.8850939644092435e-05, + "loss": 0.0119, + "step": 10640 + }, + { + "grad_norm": 0.2502377927303314, + "learning_rate": 4.876828959591485e-05, + "loss": 0.0087, + "step": 10650 + }, + { + "grad_norm": 0.2559858560562134, + "learning_rate": 4.8685642915186474e-05, + "loss": 0.0098, + "step": 10660 + }, + { + "grad_norm": 0.17750589549541473, + "learning_rate": 4.860299982786018e-05, + "loss": 0.0084, + "step": 10670 + }, + { + "grad_norm": 0.29540231823921204, + "learning_rate": 4.852036055987901e-05, + "loss": 0.0069, + "step": 10680 + }, + { + "grad_norm": 0.20562411844730377, + "learning_rate": 4.843772533717558e-05, + "loss": 0.0083, + "step": 10690 + }, + { + "grad_norm": 0.25039413571357727, + "learning_rate": 4.835509438567142e-05, + "loss": 0.0092, + "step": 10700 + }, + { + "grad_norm": 0.2514692544937134, + "learning_rate": 4.827246793127639e-05, + "loss": 0.0077, + "step": 10710 + }, + { + "grad_norm": 0.30662453174591064, + "learning_rate": 4.818984619988807e-05, + "loss": 0.0107, + "step": 10720 + }, + { + "grad_norm": 0.26225242018699646, + "learning_rate": 4.810722941739115e-05, + "loss": 0.0104, + "step": 10730 + }, + { + "grad_norm": 0.3066365122795105, + "learning_rate": 4.8024617809656684e-05, + "loss": 0.0085, + "step": 10740 + }, + { + "grad_norm": 0.27893194556236267, + "learning_rate": 4.794201160254171e-05, + "loss": 0.0077, + "step": 10750 + }, + { + "grad_norm": 0.2807646095752716, + "learning_rate": 4.785941102188844e-05, + "loss": 0.0096, + "step": 10760 + }, + { + "grad_norm": 0.299655944108963, + "learning_rate": 4.7776816293523686e-05, + "loss": 0.0108, + "step": 10770 + }, + { + "grad_norm": 0.2901594638824463, + "learning_rate": 4.769422764325832e-05, + "loss": 0.0086, + "step": 10780 + }, + { + "grad_norm": 0.3350774943828583, + "learning_rate": 4.76116452968865e-05, + "loss": 0.0124, + "step": 10790 + }, + { + "grad_norm": 0.29051363468170166, + "learning_rate": 4.752906948018525e-05, + "loss": 0.0097, + "step": 10800 + }, + { + "grad_norm": 0.26909443736076355, + "learning_rate": 4.7446500418913684e-05, + "loss": 0.0076, + "step": 10810 + }, + { + "grad_norm": 0.28262028098106384, + "learning_rate": 4.736393833881247e-05, + "loss": 0.0097, + "step": 10820 + }, + { + "grad_norm": 0.22907830774784088, + "learning_rate": 4.7281383465603194e-05, + "loss": 0.0094, + "step": 10830 + }, + { + "grad_norm": 0.38901764154434204, + "learning_rate": 4.71988360249877e-05, + "loss": 0.0085, + "step": 10840 + }, + { + "grad_norm": 0.2833888828754425, + "learning_rate": 4.7116296242647554e-05, + "loss": 0.0076, + "step": 10850 + }, + { + "grad_norm": 0.2758328318595886, + "learning_rate": 4.703376434424336e-05, + "loss": 0.0106, + "step": 10860 + }, + { + "grad_norm": 0.26278847455978394, + "learning_rate": 4.695124055541421e-05, + "loss": 0.01, + "step": 10870 + }, + { + "grad_norm": 0.3171476125717163, + "learning_rate": 4.6868725101776934e-05, + "loss": 0.0086, + "step": 10880 + }, + { + "grad_norm": 0.3477509915828705, + "learning_rate": 4.678621820892567e-05, + "loss": 0.0101, + "step": 10890 + }, + { + "grad_norm": 0.30033016204833984, + "learning_rate": 4.670372010243111e-05, + "loss": 0.0088, + "step": 10900 + }, + { + "grad_norm": 0.3467949628829956, + "learning_rate": 4.662123100783992e-05, + "loss": 0.0095, + "step": 10910 + }, + { + "grad_norm": 0.24811244010925293, + "learning_rate": 4.653875115067415e-05, + "loss": 0.0102, + "step": 10920 + }, + { + "grad_norm": 0.33560365438461304, + "learning_rate": 4.6456280756430545e-05, + "loss": 0.0096, + "step": 10930 + }, + { + "grad_norm": 0.26739415526390076, + "learning_rate": 4.637382005058004e-05, + "loss": 0.01, + "step": 10940 + }, + { + "grad_norm": 0.2914515733718872, + "learning_rate": 4.629136925856705e-05, + "loss": 0.0068, + "step": 10950 + }, + { + "grad_norm": 0.256402850151062, + "learning_rate": 4.6208928605808895e-05, + "loss": 0.0079, + "step": 10960 + }, + { + "grad_norm": 0.19222067296504974, + "learning_rate": 4.612649831769519e-05, + "loss": 0.0082, + "step": 10970 + }, + { + "grad_norm": 0.33148807287216187, + "learning_rate": 4.604407861958715e-05, + "loss": 0.0094, + "step": 10980 + }, + { + "grad_norm": 0.35632526874542236, + "learning_rate": 4.5961669736817114e-05, + "loss": 0.0102, + "step": 10990 + }, + { + "grad_norm": 0.3180091381072998, + "learning_rate": 4.5879271894687814e-05, + "loss": 0.0096, + "step": 11000 + }, + { + "grad_norm": 0.3015517592430115, + "learning_rate": 4.5796885318471826e-05, + "loss": 0.0075, + "step": 11010 + }, + { + "grad_norm": 0.24222354590892792, + "learning_rate": 4.571451023341086e-05, + "loss": 0.0083, + "step": 11020 + }, + { + "grad_norm": 0.3168736696243286, + "learning_rate": 4.563214686471527e-05, + "loss": 0.0091, + "step": 11030 + }, + { + "grad_norm": 0.20201337337493896, + "learning_rate": 4.5549795437563365e-05, + "loss": 0.0082, + "step": 11040 + }, + { + "grad_norm": 0.3413758873939514, + "learning_rate": 4.546745617710081e-05, + "loss": 0.0091, + "step": 11050 + }, + { + "grad_norm": 0.19693852961063385, + "learning_rate": 4.5385129308440014e-05, + "loss": 0.0089, + "step": 11060 + }, + { + "grad_norm": 0.2525441646575928, + "learning_rate": 4.530281505665944e-05, + "loss": 0.0088, + "step": 11070 + }, + { + "grad_norm": 0.2551596462726593, + "learning_rate": 4.5220513646803134e-05, + "loss": 0.0105, + "step": 11080 + }, + { + "grad_norm": 0.1865827739238739, + "learning_rate": 4.513822530388003e-05, + "loss": 0.0077, + "step": 11090 + }, + { + "grad_norm": 0.25236818194389343, + "learning_rate": 4.5055950252863296e-05, + "loss": 0.0078, + "step": 11100 + }, + { + "grad_norm": 0.2237762063741684, + "learning_rate": 4.4973688718689803e-05, + "loss": 0.0081, + "step": 11110 + }, + { + "grad_norm": 0.22772948443889618, + "learning_rate": 4.4891440926259406e-05, + "loss": 0.0076, + "step": 11120 + }, + { + "grad_norm": 0.23078931868076324, + "learning_rate": 4.480920710043443e-05, + "loss": 0.0088, + "step": 11130 + }, + { + "grad_norm": 0.19781197607517242, + "learning_rate": 4.4726987466039044e-05, + "loss": 0.0098, + "step": 11140 + }, + { + "grad_norm": 0.24051713943481445, + "learning_rate": 4.46447822478586e-05, + "loss": 0.0118, + "step": 11150 + }, + { + "grad_norm": 0.23845115303993225, + "learning_rate": 4.4562591670638974e-05, + "loss": 0.0066, + "step": 11160 + }, + { + "grad_norm": 0.3238190710544586, + "learning_rate": 4.4480415959086105e-05, + "loss": 0.0073, + "step": 11170 + }, + { + "grad_norm": 0.2884441912174225, + "learning_rate": 4.439825533786522e-05, + "loss": 0.0105, + "step": 11180 + }, + { + "grad_norm": 0.3803757429122925, + "learning_rate": 4.431611003160035e-05, + "loss": 0.0086, + "step": 11190 + }, + { + "grad_norm": 0.17523354291915894, + "learning_rate": 4.4233980264873636e-05, + "loss": 0.0093, + "step": 11200 + }, + { + "grad_norm": 0.2824532687664032, + "learning_rate": 4.4151866262224684e-05, + "loss": 0.0108, + "step": 11210 + }, + { + "grad_norm": 0.3265339732170105, + "learning_rate": 4.406976824815006e-05, + "loss": 0.0079, + "step": 11220 + }, + { + "grad_norm": 0.24177534878253937, + "learning_rate": 4.3987686447102595e-05, + "loss": 0.0085, + "step": 11230 + }, + { + "grad_norm": 0.21874769032001495, + "learning_rate": 4.3905621083490804e-05, + "loss": 0.0076, + "step": 11240 + }, + { + "grad_norm": 0.2425031214952469, + "learning_rate": 4.3823572381678286e-05, + "loss": 0.0089, + "step": 11250 + }, + { + "grad_norm": 0.2610780894756317, + "learning_rate": 4.374154056598301e-05, + "loss": 0.0118, + "step": 11260 + }, + { + "grad_norm": 0.36238113045692444, + "learning_rate": 4.3659525860676845e-05, + "loss": 0.0131, + "step": 11270 + }, + { + "grad_norm": 0.25657838582992554, + "learning_rate": 4.3577528489984854e-05, + "loss": 0.0081, + "step": 11280 + }, + { + "grad_norm": 0.27643704414367676, + "learning_rate": 4.349554867808476e-05, + "loss": 0.008, + "step": 11290 + }, + { + "grad_norm": 0.2851879596710205, + "learning_rate": 4.34135866491062e-05, + "loss": 0.0084, + "step": 11300 + }, + { + "grad_norm": 0.3266191780567169, + "learning_rate": 4.333164262713022e-05, + "loss": 0.0119, + "step": 11310 + }, + { + "grad_norm": 0.3193758428096771, + "learning_rate": 4.324971683618868e-05, + "loss": 0.0075, + "step": 11320 + }, + { + "grad_norm": 0.32918694615364075, + "learning_rate": 4.316780950026354e-05, + "loss": 0.0098, + "step": 11330 + }, + { + "grad_norm": 0.2517104148864746, + "learning_rate": 4.308592084328637e-05, + "loss": 0.0098, + "step": 11340 + }, + { + "grad_norm": 0.31300199031829834, + "learning_rate": 4.3004051089137576e-05, + "loss": 0.0095, + "step": 11350 + }, + { + "grad_norm": 0.2529290020465851, + "learning_rate": 4.292220046164597e-05, + "loss": 0.0073, + "step": 11360 + }, + { + "grad_norm": 0.23040375113487244, + "learning_rate": 4.2840369184588035e-05, + "loss": 0.0077, + "step": 11370 + }, + { + "grad_norm": 0.30496788024902344, + "learning_rate": 4.2758557481687345e-05, + "loss": 0.0096, + "step": 11380 + }, + { + "grad_norm": 0.2186974138021469, + "learning_rate": 4.267676557661403e-05, + "loss": 0.0084, + "step": 11390 + }, + { + "grad_norm": 0.251447856426239, + "learning_rate": 4.2594993692983955e-05, + "loss": 0.0082, + "step": 11400 + }, + { + "grad_norm": 0.29543429613113403, + "learning_rate": 4.251324205435837e-05, + "loss": 0.0093, + "step": 11410 + }, + { + "grad_norm": 0.3692440688610077, + "learning_rate": 4.243151088424312e-05, + "loss": 0.0108, + "step": 11420 + }, + { + "grad_norm": 0.27703699469566345, + "learning_rate": 4.234980040608813e-05, + "loss": 0.0118, + "step": 11430 + }, + { + "grad_norm": 0.2971097528934479, + "learning_rate": 4.22681108432867e-05, + "loss": 0.0097, + "step": 11440 + }, + { + "grad_norm": 0.2624056041240692, + "learning_rate": 4.2186442419174984e-05, + "loss": 0.0085, + "step": 11450 + }, + { + "grad_norm": 0.2369023561477661, + "learning_rate": 4.210479535703133e-05, + "loss": 0.0081, + "step": 11460 + }, + { + "grad_norm": 0.18464338779449463, + "learning_rate": 4.202316988007567e-05, + "loss": 0.0095, + "step": 11470 + }, + { + "grad_norm": 0.21752451360225677, + "learning_rate": 4.194156621146901e-05, + "loss": 0.008, + "step": 11480 + }, + { + "grad_norm": 0.20453378558158875, + "learning_rate": 4.1859984574312596e-05, + "loss": 0.0069, + "step": 11490 + }, + { + "grad_norm": 0.21684104204177856, + "learning_rate": 4.177842519164752e-05, + "loss": 0.0117, + "step": 11500 + }, + { + "grad_norm": 0.22559356689453125, + "learning_rate": 4.169688828645404e-05, + "loss": 0.0082, + "step": 11510 + }, + { + "grad_norm": 0.31731340289115906, + "learning_rate": 4.161537408165092e-05, + "loss": 0.0098, + "step": 11520 + }, + { + "grad_norm": 0.2910335063934326, + "learning_rate": 4.1533882800094924e-05, + "loss": 0.0092, + "step": 11530 + }, + { + "grad_norm": 0.2638309597969055, + "learning_rate": 4.145241466458005e-05, + "loss": 0.0097, + "step": 11540 + }, + { + "grad_norm": 0.2704751193523407, + "learning_rate": 4.13709698978371e-05, + "loss": 0.0066, + "step": 11550 + }, + { + "grad_norm": 0.24327205121517181, + "learning_rate": 4.1289548722532944e-05, + "loss": 0.0101, + "step": 11560 + }, + { + "grad_norm": 0.2812352180480957, + "learning_rate": 4.120815136126999e-05, + "loss": 0.0096, + "step": 11570 + }, + { + "grad_norm": 0.1935969889163971, + "learning_rate": 4.112677803658548e-05, + "loss": 0.0073, + "step": 11580 + }, + { + "grad_norm": 0.2166028618812561, + "learning_rate": 4.1045428970951e-05, + "loss": 0.0073, + "step": 11590 + }, + { + "grad_norm": 0.20732693374156952, + "learning_rate": 4.0964104386771785e-05, + "loss": 0.0068, + "step": 11600 + }, + { + "grad_norm": 0.36443421244621277, + "learning_rate": 4.0882804506386144e-05, + "loss": 0.009, + "step": 11610 + }, + { + "grad_norm": 0.2749943733215332, + "learning_rate": 4.080152955206485e-05, + "loss": 0.011, + "step": 11620 + }, + { + "grad_norm": 0.34515616297721863, + "learning_rate": 4.0720279746010505e-05, + "loss": 0.0071, + "step": 11630 + }, + { + "grad_norm": 0.35089951753616333, + "learning_rate": 4.063905531035699e-05, + "loss": 0.0085, + "step": 11640 + }, + { + "grad_norm": 0.271034300327301, + "learning_rate": 4.055785646716882e-05, + "loss": 0.0083, + "step": 11650 + }, + { + "grad_norm": 0.3269774913787842, + "learning_rate": 4.047668343844051e-05, + "loss": 0.0079, + "step": 11660 + }, + { + "grad_norm": 0.2104874551296234, + "learning_rate": 4.039553644609604e-05, + "loss": 0.0066, + "step": 11670 + }, + { + "grad_norm": 0.23121942579746246, + "learning_rate": 4.0314415711988176e-05, + "loss": 0.0067, + "step": 11680 + }, + { + "grad_norm": 0.3726987838745117, + "learning_rate": 4.023332145789792e-05, + "loss": 0.0108, + "step": 11690 + }, + { + "grad_norm": 0.28051552176475525, + "learning_rate": 4.015225390553385e-05, + "loss": 0.0099, + "step": 11700 + }, + { + "grad_norm": 0.30161479115486145, + "learning_rate": 4.007121327653158e-05, + "loss": 0.0085, + "step": 11710 + }, + { + "grad_norm": 0.2895340919494629, + "learning_rate": 3.9990199792453064e-05, + "loss": 0.0101, + "step": 11720 + }, + { + "grad_norm": 0.20414268970489502, + "learning_rate": 3.9909213674786103e-05, + "loss": 0.0067, + "step": 11730 + }, + { + "grad_norm": 0.17530548572540283, + "learning_rate": 3.982825514494363e-05, + "loss": 0.0077, + "step": 11740 + }, + { + "grad_norm": 0.2952740490436554, + "learning_rate": 3.974732442426319e-05, + "loss": 0.0076, + "step": 11750 + }, + { + "grad_norm": 0.2763718068599701, + "learning_rate": 3.966642173400629e-05, + "loss": 0.0081, + "step": 11760 + }, + { + "grad_norm": 0.24036982655525208, + "learning_rate": 3.9585547295357764e-05, + "loss": 0.0078, + "step": 11770 + }, + { + "grad_norm": 0.1980796903371811, + "learning_rate": 3.950470132942526e-05, + "loss": 0.0107, + "step": 11780 + }, + { + "grad_norm": 0.22392122447490692, + "learning_rate": 3.942388405723856e-05, + "loss": 0.0072, + "step": 11790 + }, + { + "grad_norm": 0.3153569996356964, + "learning_rate": 3.9343095699749e-05, + "loss": 0.0118, + "step": 11800 + }, + { + "grad_norm": 0.2480199635028839, + "learning_rate": 3.9262336477828874e-05, + "loss": 0.008, + "step": 11810 + }, + { + "grad_norm": 0.33051490783691406, + "learning_rate": 3.9181606612270794e-05, + "loss": 0.0103, + "step": 11820 + }, + { + "grad_norm": 0.2385939210653305, + "learning_rate": 3.910090632378713e-05, + "loss": 0.0092, + "step": 11830 + }, + { + "grad_norm": 0.23588930070400238, + "learning_rate": 3.90202358330094e-05, + "loss": 0.0078, + "step": 11840 + }, + { + "grad_norm": 0.24540403485298157, + "learning_rate": 3.8939595360487656e-05, + "loss": 0.0098, + "step": 11850 + }, + { + "grad_norm": 0.2659599184989929, + "learning_rate": 3.885898512668984e-05, + "loss": 0.0099, + "step": 11860 + }, + { + "grad_norm": 0.2260056883096695, + "learning_rate": 3.877840535200127e-05, + "loss": 0.0111, + "step": 11870 + }, + { + "grad_norm": 0.2496788501739502, + "learning_rate": 3.869785625672397e-05, + "loss": 0.0078, + "step": 11880 + }, + { + "grad_norm": 0.24494805932044983, + "learning_rate": 3.8617338061076094e-05, + "loss": 0.0092, + "step": 11890 + }, + { + "grad_norm": 0.27489352226257324, + "learning_rate": 3.853685098519132e-05, + "loss": 0.0143, + "step": 11900 + }, + { + "grad_norm": 0.1963207870721817, + "learning_rate": 3.845639524911823e-05, + "loss": 0.0076, + "step": 11910 + }, + { + "grad_norm": 0.26988306641578674, + "learning_rate": 3.837597107281974e-05, + "loss": 0.0084, + "step": 11920 + }, + { + "grad_norm": 0.2814962565898895, + "learning_rate": 3.829557867617247e-05, + "loss": 0.0087, + "step": 11930 + }, + { + "grad_norm": 0.2576271891593933, + "learning_rate": 3.821521827896618e-05, + "loss": 0.0074, + "step": 11940 + }, + { + "grad_norm": 0.35358384251594543, + "learning_rate": 3.81348901009031e-05, + "loss": 0.0116, + "step": 11950 + }, + { + "grad_norm": 0.23834174871444702, + "learning_rate": 3.805459436159741e-05, + "loss": 0.0092, + "step": 11960 + }, + { + "grad_norm": 0.198640838265419, + "learning_rate": 3.797433128057461e-05, + "loss": 0.0081, + "step": 11970 + }, + { + "grad_norm": 0.2608760595321655, + "learning_rate": 3.789410107727089e-05, + "loss": 0.0118, + "step": 11980 + }, + { + "grad_norm": 0.3514273762702942, + "learning_rate": 3.781390397103257e-05, + "loss": 0.0086, + "step": 11990 + }, + { + "grad_norm": 0.2969302833080292, + "learning_rate": 3.7733740181115455e-05, + "loss": 0.0094, + "step": 12000 + }, + { + "grad_norm": 0.2919679880142212, + "learning_rate": 3.7653609926684306e-05, + "loss": 0.0109, + "step": 12010 + }, + { + "grad_norm": 0.25600647926330566, + "learning_rate": 3.757351342681217e-05, + "loss": 0.0099, + "step": 12020 + }, + { + "grad_norm": 0.3260495066642761, + "learning_rate": 3.749345090047982e-05, + "loss": 0.01, + "step": 12030 + }, + { + "grad_norm": 0.28559738397598267, + "learning_rate": 3.741342256657515e-05, + "loss": 0.0103, + "step": 12040 + }, + { + "grad_norm": 0.2582564055919647, + "learning_rate": 3.7333428643892567e-05, + "loss": 0.0075, + "step": 12050 + }, + { + "grad_norm": 0.21928203105926514, + "learning_rate": 3.725346935113239e-05, + "loss": 0.0077, + "step": 12060 + }, + { + "grad_norm": 0.2669205963611603, + "learning_rate": 3.717354490690029e-05, + "loss": 0.0084, + "step": 12070 + }, + { + "grad_norm": 0.21352745592594147, + "learning_rate": 3.709365552970664e-05, + "loss": 0.0086, + "step": 12080 + }, + { + "grad_norm": 0.25914400815963745, + "learning_rate": 3.7013801437965945e-05, + "loss": 0.0072, + "step": 12090 + }, + { + "grad_norm": 0.27507898211479187, + "learning_rate": 3.693398284999623e-05, + "loss": 0.0085, + "step": 12100 + }, + { + "grad_norm": 0.2519773840904236, + "learning_rate": 3.6854199984018484e-05, + "loss": 0.0083, + "step": 12110 + }, + { + "grad_norm": 0.2433474063873291, + "learning_rate": 3.677445305815601e-05, + "loss": 0.0098, + "step": 12120 + }, + { + "grad_norm": 0.22368106245994568, + "learning_rate": 3.669474229043387e-05, + "loss": 0.0079, + "step": 12130 + }, + { + "grad_norm": 0.22219599783420563, + "learning_rate": 3.6615067898778235e-05, + "loss": 0.0061, + "step": 12140 + }, + { + "grad_norm": 0.2246261090040207, + "learning_rate": 3.6535430101015866e-05, + "loss": 0.0076, + "step": 12150 + }, + { + "grad_norm": 0.24473929405212402, + "learning_rate": 3.645582911487345e-05, + "loss": 0.0088, + "step": 12160 + }, + { + "grad_norm": 0.28539374470710754, + "learning_rate": 3.637626515797706e-05, + "loss": 0.0088, + "step": 12170 + }, + { + "grad_norm": 0.25773289799690247, + "learning_rate": 3.629673844785152e-05, + "loss": 0.0067, + "step": 12180 + }, + { + "grad_norm": 0.21043069660663605, + "learning_rate": 3.621724920191979e-05, + "loss": 0.0074, + "step": 12190 + }, + { + "grad_norm": 0.2802955210208893, + "learning_rate": 3.6137797637502444e-05, + "loss": 0.0099, + "step": 12200 + }, + { + "grad_norm": 0.22656969726085663, + "learning_rate": 3.6058383971817035e-05, + "loss": 0.0068, + "step": 12210 + }, + { + "grad_norm": 0.33790233731269836, + "learning_rate": 3.59790084219775e-05, + "loss": 0.0084, + "step": 12220 + }, + { + "grad_norm": 0.29881054162979126, + "learning_rate": 3.589967120499353e-05, + "loss": 0.0069, + "step": 12230 + }, + { + "grad_norm": 0.3114658296108246, + "learning_rate": 3.5820372537770075e-05, + "loss": 0.0073, + "step": 12240 + }, + { + "grad_norm": 0.25627395510673523, + "learning_rate": 3.5741112637106655e-05, + "loss": 0.0089, + "step": 12250 + }, + { + "grad_norm": 0.20619019865989685, + "learning_rate": 3.5661891719696804e-05, + "loss": 0.008, + "step": 12260 + }, + { + "grad_norm": 0.24497458338737488, + "learning_rate": 3.5582710002127504e-05, + "loss": 0.011, + "step": 12270 + }, + { + "grad_norm": 0.23106102645397186, + "learning_rate": 3.550356770087853e-05, + "loss": 0.0086, + "step": 12280 + }, + { + "grad_norm": 0.2801083028316498, + "learning_rate": 3.5424465032321914e-05, + "loss": 0.0081, + "step": 12290 + }, + { + "grad_norm": 0.27025705575942993, + "learning_rate": 3.5345402212721335e-05, + "loss": 0.0082, + "step": 12300 + }, + { + "grad_norm": 0.2505914270877838, + "learning_rate": 3.526637945823152e-05, + "loss": 0.009, + "step": 12310 + }, + { + "grad_norm": 0.2746097445487976, + "learning_rate": 3.518739698489767e-05, + "loss": 0.0064, + "step": 12320 + }, + { + "grad_norm": 0.2840472459793091, + "learning_rate": 3.510845500865485e-05, + "loss": 0.0073, + "step": 12330 + }, + { + "grad_norm": 0.27252042293548584, + "learning_rate": 3.502955374532739e-05, + "loss": 0.011, + "step": 12340 + }, + { + "grad_norm": 0.21478882431983948, + "learning_rate": 3.495069341062836e-05, + "loss": 0.0082, + "step": 12350 + }, + { + "grad_norm": 0.30031028389930725, + "learning_rate": 3.4871874220158896e-05, + "loss": 0.0115, + "step": 12360 + }, + { + "grad_norm": 0.29065871238708496, + "learning_rate": 3.479309638940762e-05, + "loss": 0.009, + "step": 12370 + }, + { + "grad_norm": 0.3161035478115082, + "learning_rate": 3.4714360133750146e-05, + "loss": 0.0085, + "step": 12380 + }, + { + "grad_norm": 0.29907113313674927, + "learning_rate": 3.463566566844839e-05, + "loss": 0.0078, + "step": 12390 + }, + { + "grad_norm": 0.2610228359699249, + "learning_rate": 3.4557013208650016e-05, + "loss": 0.0093, + "step": 12400 + }, + { + "grad_norm": 0.19939985871315002, + "learning_rate": 3.4478402969387857e-05, + "loss": 0.0075, + "step": 12410 + }, + { + "grad_norm": 0.22081556916236877, + "learning_rate": 3.4399835165579266e-05, + "loss": 0.0075, + "step": 12420 + }, + { + "grad_norm": 0.20169289410114288, + "learning_rate": 3.4321310012025645e-05, + "loss": 0.0063, + "step": 12430 + }, + { + "grad_norm": 0.2489660531282425, + "learning_rate": 3.424282772341176e-05, + "loss": 0.0074, + "step": 12440 + }, + { + "grad_norm": 0.28628960251808167, + "learning_rate": 3.416438851430519e-05, + "loss": 0.013, + "step": 12450 + }, + { + "grad_norm": 0.29382604360580444, + "learning_rate": 3.408599259915577e-05, + "loss": 0.0077, + "step": 12460 + }, + { + "grad_norm": 0.3110322952270508, + "learning_rate": 3.400764019229487e-05, + "loss": 0.0108, + "step": 12470 + }, + { + "grad_norm": 0.23523154854774475, + "learning_rate": 3.3929331507935035e-05, + "loss": 0.0069, + "step": 12480 + }, + { + "grad_norm": 0.273687481880188, + "learning_rate": 3.3851066760169196e-05, + "loss": 0.0075, + "step": 12490 + }, + { + "grad_norm": 0.19932150840759277, + "learning_rate": 3.377284616297021e-05, + "loss": 0.007, + "step": 12500 + }, + { + "grad_norm": 0.24185536801815033, + "learning_rate": 3.3694669930190166e-05, + "loss": 0.0068, + "step": 12510 + }, + { + "grad_norm": 0.2853364646434784, + "learning_rate": 3.36165382755599e-05, + "loss": 0.0075, + "step": 12520 + }, + { + "grad_norm": 0.23251241445541382, + "learning_rate": 3.35384514126884e-05, + "loss": 0.009, + "step": 12530 + }, + { + "grad_norm": 0.2282250076532364, + "learning_rate": 3.3460409555062154e-05, + "loss": 0.007, + "step": 12540 + }, + { + "grad_norm": 0.2507518231868744, + "learning_rate": 3.3382412916044645e-05, + "loss": 0.0073, + "step": 12550 + }, + { + "grad_norm": 0.17678532004356384, + "learning_rate": 3.330446170887566e-05, + "loss": 0.006, + "step": 12560 + }, + { + "grad_norm": 0.22255797684192657, + "learning_rate": 3.3226556146670834e-05, + "loss": 0.0069, + "step": 12570 + }, + { + "grad_norm": 0.22277423739433289, + "learning_rate": 3.314869644242102e-05, + "loss": 0.0067, + "step": 12580 + }, + { + "grad_norm": 0.24665354192256927, + "learning_rate": 3.3070882808991674e-05, + "loss": 0.0077, + "step": 12590 + }, + { + "grad_norm": 0.2221645563840866, + "learning_rate": 3.2993115459122305e-05, + "loss": 0.008, + "step": 12600 + }, + { + "grad_norm": 0.25327399373054504, + "learning_rate": 3.2915394605425835e-05, + "loss": 0.007, + "step": 12610 + }, + { + "grad_norm": 0.2379884123802185, + "learning_rate": 3.283772046038816e-05, + "loss": 0.0084, + "step": 12620 + }, + { + "grad_norm": 0.2786004841327667, + "learning_rate": 3.276009323636739e-05, + "loss": 0.007, + "step": 12630 + }, + { + "grad_norm": 0.23235200345516205, + "learning_rate": 3.268251314559344e-05, + "loss": 0.0071, + "step": 12640 + }, + { + "grad_norm": 0.2734534740447998, + "learning_rate": 3.2604980400167254e-05, + "loss": 0.0065, + "step": 12650 + }, + { + "grad_norm": 0.20516322553157806, + "learning_rate": 3.252749521206042e-05, + "loss": 0.0057, + "step": 12660 + }, + { + "grad_norm": 0.2804936170578003, + "learning_rate": 3.2450057793114494e-05, + "loss": 0.0058, + "step": 12670 + }, + { + "grad_norm": 0.23371683061122894, + "learning_rate": 3.2372668355040435e-05, + "loss": 0.0072, + "step": 12680 + }, + { + "grad_norm": 0.24273928999900818, + "learning_rate": 3.2295327109418005e-05, + "loss": 0.0076, + "step": 12690 + }, + { + "grad_norm": 0.2837400734424591, + "learning_rate": 3.221803426769518e-05, + "loss": 0.0059, + "step": 12700 + }, + { + "grad_norm": 0.1821887344121933, + "learning_rate": 3.214079004118768e-05, + "loss": 0.0062, + "step": 12710 + }, + { + "grad_norm": 0.27825450897216797, + "learning_rate": 3.2063594641078234e-05, + "loss": 0.0075, + "step": 12720 + }, + { + "grad_norm": 0.30339550971984863, + "learning_rate": 3.198644827841616e-05, + "loss": 0.0066, + "step": 12730 + }, + { + "grad_norm": 0.23971441388130188, + "learning_rate": 3.1909351164116654e-05, + "loss": 0.0066, + "step": 12740 + }, + { + "grad_norm": 0.18594853579998016, + "learning_rate": 3.183230350896026e-05, + "loss": 0.0085, + "step": 12750 + }, + { + "grad_norm": 0.2139957696199417, + "learning_rate": 3.1755305523592337e-05, + "loss": 0.0084, + "step": 12760 + }, + { + "grad_norm": 0.282361775636673, + "learning_rate": 3.167835741852245e-05, + "loss": 0.0078, + "step": 12770 + }, + { + "grad_norm": 0.23739394545555115, + "learning_rate": 3.160145940412378e-05, + "loss": 0.0095, + "step": 12780 + }, + { + "grad_norm": 0.22370891273021698, + "learning_rate": 3.1524611690632545e-05, + "loss": 0.0059, + "step": 12790 + }, + { + "grad_norm": 0.2226971536874771, + "learning_rate": 3.144781448814746e-05, + "loss": 0.0074, + "step": 12800 + }, + { + "grad_norm": 0.27098506689071655, + "learning_rate": 3.1371068006629145e-05, + "loss": 0.006, + "step": 12810 + }, + { + "grad_norm": 0.23244546353816986, + "learning_rate": 3.129437245589956e-05, + "loss": 0.0071, + "step": 12820 + }, + { + "grad_norm": 0.23745259642601013, + "learning_rate": 3.121772804564143e-05, + "loss": 0.0066, + "step": 12830 + }, + { + "grad_norm": 0.25362375378608704, + "learning_rate": 3.11411349853976e-05, + "loss": 0.0064, + "step": 12840 + }, + { + "grad_norm": 0.2746460437774658, + "learning_rate": 3.10645934845706e-05, + "loss": 0.0064, + "step": 12850 + }, + { + "grad_norm": 0.2910458445549011, + "learning_rate": 3.098810375242196e-05, + "loss": 0.008, + "step": 12860 + }, + { + "grad_norm": 0.1930484175682068, + "learning_rate": 3.0911665998071704e-05, + "loss": 0.0058, + "step": 12870 + }, + { + "grad_norm": 0.26174429059028625, + "learning_rate": 3.083528043049774e-05, + "loss": 0.0061, + "step": 12880 + }, + { + "grad_norm": 0.237471342086792, + "learning_rate": 3.0758947258535255e-05, + "loss": 0.0082, + "step": 12890 + }, + { + "grad_norm": 0.2530456483364105, + "learning_rate": 3.068266669087625e-05, + "loss": 0.0078, + "step": 12900 + }, + { + "grad_norm": 0.21137702465057373, + "learning_rate": 3.060643893606887e-05, + "loss": 0.006, + "step": 12910 + }, + { + "grad_norm": 0.1681884080171585, + "learning_rate": 3.053026420251693e-05, + "loss": 0.0068, + "step": 12920 + }, + { + "grad_norm": 0.25592848658561707, + "learning_rate": 3.0454142698479183e-05, + "loss": 0.0069, + "step": 12930 + }, + { + "grad_norm": 0.2088608294725418, + "learning_rate": 3.0378074632068954e-05, + "loss": 0.0087, + "step": 12940 + }, + { + "grad_norm": 0.22707495093345642, + "learning_rate": 3.0302060211253408e-05, + "loss": 0.0066, + "step": 12950 + }, + { + "grad_norm": 0.2611274719238281, + "learning_rate": 3.0226099643853073e-05, + "loss": 0.0075, + "step": 12960 + }, + { + "grad_norm": 0.20526066422462463, + "learning_rate": 3.0150193137541283e-05, + "loss": 0.006, + "step": 12970 + }, + { + "grad_norm": 0.26528921723365784, + "learning_rate": 3.0074340899843467e-05, + "loss": 0.0074, + "step": 12980 + }, + { + "grad_norm": 0.294217973947525, + "learning_rate": 2.999854313813677e-05, + "loss": 0.0112, + "step": 12990 + }, + { + "grad_norm": 0.2798357307910919, + "learning_rate": 2.9922800059649382e-05, + "loss": 0.0074, + "step": 13000 + }, + { + "grad_norm": 0.25348061323165894, + "learning_rate": 2.9847111871459976e-05, + "loss": 0.008, + "step": 13010 + }, + { + "grad_norm": 0.25376802682876587, + "learning_rate": 2.977147878049721e-05, + "loss": 0.0072, + "step": 13020 + }, + { + "grad_norm": 0.23868699371814728, + "learning_rate": 2.9695900993539006e-05, + "loss": 0.0069, + "step": 13030 + }, + { + "grad_norm": 0.24584409594535828, + "learning_rate": 2.9620378717212183e-05, + "loss": 0.0113, + "step": 13040 + }, + { + "grad_norm": 0.2317948341369629, + "learning_rate": 2.9544912157991745e-05, + "loss": 0.0087, + "step": 13050 + }, + { + "grad_norm": 0.3443603217601776, + "learning_rate": 2.9469501522200405e-05, + "loss": 0.006, + "step": 13060 + }, + { + "grad_norm": 0.1847532093524933, + "learning_rate": 2.9394147016007946e-05, + "loss": 0.0064, + "step": 13070 + }, + { + "grad_norm": 0.2874523401260376, + "learning_rate": 2.9318848845430702e-05, + "loss": 0.0058, + "step": 13080 + }, + { + "grad_norm": 0.20815210044384003, + "learning_rate": 2.9243607216331013e-05, + "loss": 0.0061, + "step": 13090 + }, + { + "grad_norm": 0.1735462099313736, + "learning_rate": 2.916842233441661e-05, + "loss": 0.0049, + "step": 13100 + }, + { + "grad_norm": 0.2947513163089752, + "learning_rate": 2.90932944052401e-05, + "loss": 0.0078, + "step": 13110 + }, + { + "grad_norm": 0.185662180185318, + "learning_rate": 2.9018223634198354e-05, + "loss": 0.006, + "step": 13120 + }, + { + "grad_norm": 0.159962996840477, + "learning_rate": 2.8943210226532025e-05, + "loss": 0.0085, + "step": 13130 + }, + { + "grad_norm": 0.23721252381801605, + "learning_rate": 2.8868254387324857e-05, + "loss": 0.0082, + "step": 13140 + }, + { + "grad_norm": 0.2839058041572571, + "learning_rate": 2.8793356321503306e-05, + "loss": 0.0084, + "step": 13150 + }, + { + "grad_norm": 0.3259228765964508, + "learning_rate": 2.87185162338358e-05, + "loss": 0.0067, + "step": 13160 + }, + { + "grad_norm": 0.2640034258365631, + "learning_rate": 2.8643734328932253e-05, + "loss": 0.0087, + "step": 13170 + }, + { + "grad_norm": 0.2673685848712921, + "learning_rate": 2.856901081124359e-05, + "loss": 0.0068, + "step": 13180 + }, + { + "grad_norm": 0.2533598840236664, + "learning_rate": 2.8494345885061002e-05, + "loss": 0.0084, + "step": 13190 + }, + { + "grad_norm": 0.27771884202957153, + "learning_rate": 2.8419739754515616e-05, + "loss": 0.008, + "step": 13200 + }, + { + "grad_norm": 0.18158374726772308, + "learning_rate": 2.8345192623577666e-05, + "loss": 0.0066, + "step": 13210 + }, + { + "grad_norm": 0.20402629673480988, + "learning_rate": 2.8270704696056193e-05, + "loss": 0.0085, + "step": 13220 + }, + { + "grad_norm": 0.22918453812599182, + "learning_rate": 2.8196276175598367e-05, + "loss": 0.0089, + "step": 13230 + }, + { + "grad_norm": 0.26338842511177063, + "learning_rate": 2.8121907265688884e-05, + "loss": 0.0097, + "step": 13240 + }, + { + "grad_norm": 0.2552448511123657, + "learning_rate": 2.804759816964957e-05, + "loss": 0.007, + "step": 13250 + }, + { + "grad_norm": 0.2544209957122803, + "learning_rate": 2.797334909063857e-05, + "loss": 0.0068, + "step": 13260 + }, + { + "grad_norm": 0.20611265301704407, + "learning_rate": 2.7899160231650056e-05, + "loss": 0.0057, + "step": 13270 + }, + { + "grad_norm": 0.2455790638923645, + "learning_rate": 2.7825031795513585e-05, + "loss": 0.0064, + "step": 13280 + }, + { + "grad_norm": 0.16692101955413818, + "learning_rate": 2.775096398489341e-05, + "loss": 0.0054, + "step": 13290 + }, + { + "grad_norm": 0.19417795538902283, + "learning_rate": 2.7676957002288163e-05, + "loss": 0.0069, + "step": 13300 + }, + { + "grad_norm": 0.246934175491333, + "learning_rate": 2.760301105003003e-05, + "loss": 0.0067, + "step": 13310 + }, + { + "grad_norm": 0.18580402433872223, + "learning_rate": 2.752912633028446e-05, + "loss": 0.0065, + "step": 13320 + }, + { + "grad_norm": 0.18205900490283966, + "learning_rate": 2.7455303045049474e-05, + "loss": 0.0065, + "step": 13330 + }, + { + "grad_norm": 0.24550791084766388, + "learning_rate": 2.7381541396155098e-05, + "loss": 0.0069, + "step": 13340 + }, + { + "grad_norm": 0.2719643712043762, + "learning_rate": 2.730784158526286e-05, + "loss": 0.0059, + "step": 13350 + }, + { + "grad_norm": 0.19565212726593018, + "learning_rate": 2.723420381386521e-05, + "loss": 0.0072, + "step": 13360 + }, + { + "grad_norm": 0.2644125521183014, + "learning_rate": 2.7160628283285018e-05, + "loss": 0.0074, + "step": 13370 + }, + { + "grad_norm": 0.20331737399101257, + "learning_rate": 2.7087115194675007e-05, + "loss": 0.0059, + "step": 13380 + }, + { + "grad_norm": 0.18478238582611084, + "learning_rate": 2.701366474901712e-05, + "loss": 0.005, + "step": 13390 + }, + { + "grad_norm": 0.2070227712392807, + "learning_rate": 2.6940277147122085e-05, + "loss": 0.0063, + "step": 13400 + }, + { + "grad_norm": 0.24267740547657013, + "learning_rate": 2.686695258962878e-05, + "loss": 0.0061, + "step": 13410 + }, + { + "grad_norm": 0.1909894049167633, + "learning_rate": 2.679369127700375e-05, + "loss": 0.0056, + "step": 13420 + }, + { + "grad_norm": 0.19610315561294556, + "learning_rate": 2.672049340954067e-05, + "loss": 0.007, + "step": 13430 + }, + { + "grad_norm": 0.2082044929265976, + "learning_rate": 2.6647359187359676e-05, + "loss": 0.0086, + "step": 13440 + }, + { + "grad_norm": 0.22661364078521729, + "learning_rate": 2.6574288810406946e-05, + "loss": 0.0072, + "step": 13450 + }, + { + "grad_norm": 0.2202974408864975, + "learning_rate": 2.6501282478454083e-05, + "loss": 0.0077, + "step": 13460 + }, + { + "grad_norm": 0.2299632430076599, + "learning_rate": 2.6428340391097618e-05, + "loss": 0.0109, + "step": 13470 + }, + { + "grad_norm": 0.3141759932041168, + "learning_rate": 2.6355462747758485e-05, + "loss": 0.0065, + "step": 13480 + }, + { + "grad_norm": 0.23412247002124786, + "learning_rate": 2.6282649747681304e-05, + "loss": 0.0059, + "step": 13490 + }, + { + "grad_norm": 0.2105311006307602, + "learning_rate": 2.620990158993406e-05, + "loss": 0.0054, + "step": 13500 + }, + { + "grad_norm": 0.21451091766357422, + "learning_rate": 2.6137218473407477e-05, + "loss": 0.007, + "step": 13510 + }, + { + "grad_norm": 0.18160901963710785, + "learning_rate": 2.606460059681436e-05, + "loss": 0.0051, + "step": 13520 + }, + { + "grad_norm": 0.2187230885028839, + "learning_rate": 2.599204815868928e-05, + "loss": 0.0053, + "step": 13530 + }, + { + "grad_norm": 0.20273251831531525, + "learning_rate": 2.5919561357387756e-05, + "loss": 0.0054, + "step": 13540 + }, + { + "grad_norm": 0.23314417898654938, + "learning_rate": 2.5847140391085972e-05, + "loss": 0.006, + "step": 13550 + }, + { + "grad_norm": 0.2788357436656952, + "learning_rate": 2.5774785457780103e-05, + "loss": 0.0093, + "step": 13560 + }, + { + "grad_norm": 0.20386862754821777, + "learning_rate": 2.5702496755285753e-05, + "loss": 0.0064, + "step": 13570 + }, + { + "grad_norm": 0.21573026478290558, + "learning_rate": 2.5630274481237483e-05, + "loss": 0.0057, + "step": 13580 + }, + { + "grad_norm": 0.1928456872701645, + "learning_rate": 2.5558118833088197e-05, + "loss": 0.0065, + "step": 13590 + }, + { + "grad_norm": 0.23650315403938293, + "learning_rate": 2.548603000810872e-05, + "loss": 0.0049, + "step": 13600 + }, + { + "grad_norm": 0.19839148223400116, + "learning_rate": 2.5414008203387152e-05, + "loss": 0.0068, + "step": 13610 + }, + { + "grad_norm": 0.24050778150558472, + "learning_rate": 2.534205361582834e-05, + "loss": 0.006, + "step": 13620 + }, + { + "grad_norm": 0.18059442937374115, + "learning_rate": 2.527016644215338e-05, + "loss": 0.0067, + "step": 13630 + }, + { + "grad_norm": 0.18381164968013763, + "learning_rate": 2.519834687889905e-05, + "loss": 0.007, + "step": 13640 + }, + { + "grad_norm": 0.2277163565158844, + "learning_rate": 2.5126595122417295e-05, + "loss": 0.0074, + "step": 13650 + }, + { + "grad_norm": 0.1679578423500061, + "learning_rate": 2.5054911368874713e-05, + "loss": 0.0052, + "step": 13660 + }, + { + "grad_norm": 0.19412367045879364, + "learning_rate": 2.4983295814251916e-05, + "loss": 0.0077, + "step": 13670 + }, + { + "grad_norm": 0.22263267636299133, + "learning_rate": 2.4911748654343105e-05, + "loss": 0.006, + "step": 13680 + }, + { + "grad_norm": 0.23625227808952332, + "learning_rate": 2.4840270084755463e-05, + "loss": 0.0062, + "step": 13690 + }, + { + "grad_norm": 0.22303491830825806, + "learning_rate": 2.4768860300908685e-05, + "loss": 0.0059, + "step": 13700 + }, + { + "grad_norm": 0.20615343749523163, + "learning_rate": 2.469751949803443e-05, + "loss": 0.0069, + "step": 13710 + }, + { + "grad_norm": 0.22624443471431732, + "learning_rate": 2.4626247871175666e-05, + "loss": 0.008, + "step": 13720 + }, + { + "grad_norm": 0.25985249876976013, + "learning_rate": 2.4555045615186346e-05, + "loss": 0.0065, + "step": 13730 + }, + { + "grad_norm": 0.2458672672510147, + "learning_rate": 2.4483912924730677e-05, + "loss": 0.0072, + "step": 13740 + }, + { + "grad_norm": 0.21012453734874725, + "learning_rate": 2.4412849994282742e-05, + "loss": 0.0048, + "step": 13750 + }, + { + "grad_norm": 0.14065012335777283, + "learning_rate": 2.434185701812592e-05, + "loss": 0.0052, + "step": 13760 + }, + { + "grad_norm": 0.19618520140647888, + "learning_rate": 2.4270934190352218e-05, + "loss": 0.0071, + "step": 13770 + }, + { + "grad_norm": 0.1482352614402771, + "learning_rate": 2.4200081704861998e-05, + "loss": 0.0074, + "step": 13780 + }, + { + "grad_norm": 0.20872756838798523, + "learning_rate": 2.412929975536321e-05, + "loss": 0.0069, + "step": 13790 + }, + { + "grad_norm": 0.20830373466014862, + "learning_rate": 2.4058588535371017e-05, + "loss": 0.0062, + "step": 13800 + }, + { + "grad_norm": 0.2667636275291443, + "learning_rate": 2.3987948238207243e-05, + "loss": 0.0057, + "step": 13810 + }, + { + "grad_norm": 0.24201737344264984, + "learning_rate": 2.3917379056999678e-05, + "loss": 0.0067, + "step": 13820 + }, + { + "grad_norm": 0.17350678145885468, + "learning_rate": 2.3846881184681824e-05, + "loss": 0.0046, + "step": 13830 + }, + { + "grad_norm": 0.17855527997016907, + "learning_rate": 2.377645481399214e-05, + "loss": 0.0052, + "step": 13840 + }, + { + "grad_norm": 0.24410289525985718, + "learning_rate": 2.3706100137473667e-05, + "loss": 0.0069, + "step": 13850 + }, + { + "grad_norm": 0.17104095220565796, + "learning_rate": 2.3635817347473394e-05, + "loss": 0.0078, + "step": 13860 + }, + { + "grad_norm": 0.20315547287464142, + "learning_rate": 2.3565606636141757e-05, + "loss": 0.007, + "step": 13870 + }, + { + "grad_norm": 0.24569718539714813, + "learning_rate": 2.3495468195432203e-05, + "loss": 0.0071, + "step": 13880 + }, + { + "grad_norm": 0.1952163428068161, + "learning_rate": 2.3425402217100507e-05, + "loss": 0.0052, + "step": 13890 + }, + { + "grad_norm": 0.18253405392169952, + "learning_rate": 2.3355408892704424e-05, + "loss": 0.0068, + "step": 13900 + }, + { + "grad_norm": 0.16422688961029053, + "learning_rate": 2.3285488413603003e-05, + "loss": 0.0046, + "step": 13910 + }, + { + "grad_norm": 0.2229805588722229, + "learning_rate": 2.321564097095615e-05, + "loss": 0.0048, + "step": 13920 + }, + { + "grad_norm": 0.3146781921386719, + "learning_rate": 2.3145866755724142e-05, + "loss": 0.0062, + "step": 13930 + }, + { + "grad_norm": 0.24816946685314178, + "learning_rate": 2.307616595866699e-05, + "loss": 0.0063, + "step": 13940 + }, + { + "grad_norm": 0.20742735266685486, + "learning_rate": 2.3006538770344032e-05, + "loss": 0.0067, + "step": 13950 + }, + { + "grad_norm": 0.17962752282619476, + "learning_rate": 2.293698538111334e-05, + "loss": 0.0056, + "step": 13960 + }, + { + "grad_norm": 0.2705751359462738, + "learning_rate": 2.28675059811312e-05, + "loss": 0.0065, + "step": 13970 + }, + { + "grad_norm": 0.23835845291614532, + "learning_rate": 2.279810076035167e-05, + "loss": 0.0071, + "step": 13980 + }, + { + "grad_norm": 0.24078679084777832, + "learning_rate": 2.272876990852596e-05, + "loss": 0.0076, + "step": 13990 + }, + { + "grad_norm": 0.22596397995948792, + "learning_rate": 2.265951361520195e-05, + "loss": 0.0101, + "step": 14000 + }, + { + "grad_norm": 0.19356085360050201, + "learning_rate": 2.2590332069723748e-05, + "loss": 0.0048, + "step": 14010 + }, + { + "grad_norm": 0.16043022274971008, + "learning_rate": 2.2521225461231004e-05, + "loss": 0.0067, + "step": 14020 + }, + { + "grad_norm": 0.2451379895210266, + "learning_rate": 2.2452193978658597e-05, + "loss": 0.0049, + "step": 14030 + }, + { + "grad_norm": 0.19134803116321564, + "learning_rate": 2.238323781073594e-05, + "loss": 0.0071, + "step": 14040 + }, + { + "grad_norm": 0.1770501583814621, + "learning_rate": 2.2314357145986552e-05, + "loss": 0.007, + "step": 14050 + }, + { + "grad_norm": 0.19176732003688812, + "learning_rate": 2.224555217272757e-05, + "loss": 0.007, + "step": 14060 + }, + { + "grad_norm": 0.2289755940437317, + "learning_rate": 2.2176823079069127e-05, + "loss": 0.0065, + "step": 14070 + }, + { + "grad_norm": 0.27378588914871216, + "learning_rate": 2.210817005291398e-05, + "loss": 0.0058, + "step": 14080 + }, + { + "grad_norm": 0.2435673326253891, + "learning_rate": 2.203959328195686e-05, + "loss": 0.0069, + "step": 14090 + }, + { + "grad_norm": 0.22978395223617554, + "learning_rate": 2.1971092953684026e-05, + "loss": 0.006, + "step": 14100 + }, + { + "grad_norm": 0.19038143754005432, + "learning_rate": 2.1902669255372788e-05, + "loss": 0.0054, + "step": 14110 + }, + { + "grad_norm": 0.16993741691112518, + "learning_rate": 2.1834322374090897e-05, + "loss": 0.0064, + "step": 14120 + }, + { + "grad_norm": 0.22536249458789825, + "learning_rate": 2.1766052496696153e-05, + "loss": 0.0061, + "step": 14130 + }, + { + "grad_norm": 0.1695421189069748, + "learning_rate": 2.169785980983577e-05, + "loss": 0.0053, + "step": 14140 + }, + { + "grad_norm": 0.20765681564807892, + "learning_rate": 2.162974449994593e-05, + "loss": 0.0055, + "step": 14150 + }, + { + "grad_norm": 0.20967355370521545, + "learning_rate": 2.1561706753251337e-05, + "loss": 0.0071, + "step": 14160 + }, + { + "grad_norm": 0.24757470190525055, + "learning_rate": 2.1493746755764544e-05, + "loss": 0.006, + "step": 14170 + }, + { + "grad_norm": 0.17080722749233246, + "learning_rate": 2.1425864693285635e-05, + "loss": 0.0087, + "step": 14180 + }, + { + "grad_norm": 0.1654953807592392, + "learning_rate": 2.1358060751401547e-05, + "loss": 0.0062, + "step": 14190 + }, + { + "grad_norm": 0.23503398895263672, + "learning_rate": 2.129033511548566e-05, + "loss": 0.0078, + "step": 14200 + }, + { + "grad_norm": 0.19658979773521423, + "learning_rate": 2.1222687970697315e-05, + "loss": 0.0043, + "step": 14210 + }, + { + "grad_norm": 0.24506372213363647, + "learning_rate": 2.1155119501981173e-05, + "loss": 0.0062, + "step": 14220 + }, + { + "grad_norm": 0.23663446307182312, + "learning_rate": 2.1087629894066895e-05, + "loss": 0.0053, + "step": 14230 + }, + { + "grad_norm": 0.20866233110427856, + "learning_rate": 2.1020219331468473e-05, + "loss": 0.0076, + "step": 14240 + }, + { + "grad_norm": 0.2166251242160797, + "learning_rate": 2.095288799848379e-05, + "loss": 0.0069, + "step": 14250 + }, + { + "grad_norm": 0.17026379704475403, + "learning_rate": 2.088563607919417e-05, + "loss": 0.0052, + "step": 14260 + }, + { + "grad_norm": 0.26690876483917236, + "learning_rate": 2.0818463757463786e-05, + "loss": 0.0055, + "step": 14270 + }, + { + "grad_norm": 0.21364854276180267, + "learning_rate": 2.0751371216939175e-05, + "loss": 0.0062, + "step": 14280 + }, + { + "grad_norm": 0.25960573554039, + "learning_rate": 2.068435864104882e-05, + "loss": 0.0057, + "step": 14290 + }, + { + "grad_norm": 0.2138037234544754, + "learning_rate": 2.0617426213002506e-05, + "loss": 0.0089, + "step": 14300 + }, + { + "grad_norm": 0.20547112822532654, + "learning_rate": 2.055057411579097e-05, + "loss": 0.0059, + "step": 14310 + }, + { + "grad_norm": 0.32553014159202576, + "learning_rate": 2.0483802532185286e-05, + "loss": 0.0066, + "step": 14320 + }, + { + "grad_norm": 0.22871890664100647, + "learning_rate": 2.041711164473638e-05, + "loss": 0.0072, + "step": 14330 + }, + { + "grad_norm": 0.20716866850852966, + "learning_rate": 2.0350501635774637e-05, + "loss": 0.0055, + "step": 14340 + }, + { + "grad_norm": 0.17182990908622742, + "learning_rate": 2.0283972687409247e-05, + "loss": 0.0063, + "step": 14350 + }, + { + "grad_norm": 0.16304871439933777, + "learning_rate": 2.021752498152784e-05, + "loss": 0.0061, + "step": 14360 + }, + { + "grad_norm": 0.22496426105499268, + "learning_rate": 2.015115869979589e-05, + "loss": 0.0046, + "step": 14370 + }, + { + "grad_norm": 0.1372428685426712, + "learning_rate": 2.0084874023656265e-05, + "loss": 0.0049, + "step": 14380 + }, + { + "grad_norm": 0.22879809141159058, + "learning_rate": 2.001867113432877e-05, + "loss": 0.0075, + "step": 14390 + }, + { + "grad_norm": 0.18383388221263885, + "learning_rate": 1.995255021280954e-05, + "loss": 0.0059, + "step": 14400 + }, + { + "grad_norm": 0.1956668198108673, + "learning_rate": 1.9886511439870688e-05, + "loss": 0.007, + "step": 14410 + }, + { + "grad_norm": 0.2605588436126709, + "learning_rate": 1.9820554996059675e-05, + "loss": 0.0064, + "step": 14420 + }, + { + "grad_norm": 0.21275591850280762, + "learning_rate": 1.9754681061698893e-05, + "loss": 0.0047, + "step": 14430 + }, + { + "grad_norm": 0.15767215192317963, + "learning_rate": 1.9688889816885185e-05, + "loss": 0.0051, + "step": 14440 + }, + { + "grad_norm": 0.206788569688797, + "learning_rate": 1.962318144148928e-05, + "loss": 0.0053, + "step": 14450 + }, + { + "grad_norm": 0.21391050517559052, + "learning_rate": 1.955755611515539e-05, + "loss": 0.0054, + "step": 14460 + }, + { + "grad_norm": 0.15545658767223358, + "learning_rate": 1.9492014017300642e-05, + "loss": 0.0068, + "step": 14470 + }, + { + "grad_norm": 0.1992073655128479, + "learning_rate": 1.942655532711461e-05, + "loss": 0.005, + "step": 14480 + }, + { + "grad_norm": 0.17155469954013824, + "learning_rate": 1.9361180223558882e-05, + "loss": 0.0068, + "step": 14490 + }, + { + "grad_norm": 0.1756027191877365, + "learning_rate": 1.929588888536647e-05, + "loss": 0.0063, + "step": 14500 + }, + { + "grad_norm": 0.21194574236869812, + "learning_rate": 1.9230681491041425e-05, + "loss": 0.0057, + "step": 14510 + }, + { + "grad_norm": 0.18223026394844055, + "learning_rate": 1.9165558218858264e-05, + "loss": 0.0052, + "step": 14520 + }, + { + "grad_norm": 0.250331848859787, + "learning_rate": 1.9100519246861505e-05, + "loss": 0.0053, + "step": 14530 + }, + { + "grad_norm": 0.2652829885482788, + "learning_rate": 1.9035564752865248e-05, + "loss": 0.0076, + "step": 14540 + }, + { + "grad_norm": 0.23563922941684723, + "learning_rate": 1.897069491445258e-05, + "loss": 0.0057, + "step": 14550 + }, + { + "grad_norm": 0.21190163493156433, + "learning_rate": 1.890590990897515e-05, + "loss": 0.0057, + "step": 14560 + }, + { + "grad_norm": 0.19106173515319824, + "learning_rate": 1.884120991355272e-05, + "loss": 0.0059, + "step": 14570 + }, + { + "grad_norm": 0.19955740869045258, + "learning_rate": 1.8776595105072576e-05, + "loss": 0.0054, + "step": 14580 + }, + { + "grad_norm": 0.1958319991827011, + "learning_rate": 1.8712065660189166e-05, + "loss": 0.0054, + "step": 14590 + }, + { + "grad_norm": 0.20281006395816803, + "learning_rate": 1.8647621755323513e-05, + "loss": 0.0054, + "step": 14600 + }, + { + "grad_norm": 0.2089354693889618, + "learning_rate": 1.858326356666278e-05, + "loss": 0.0055, + "step": 14610 + }, + { + "grad_norm": 0.3210635781288147, + "learning_rate": 1.851899127015983e-05, + "loss": 0.0061, + "step": 14620 + }, + { + "grad_norm": 0.2990008592605591, + "learning_rate": 1.8454805041532626e-05, + "loss": 0.0067, + "step": 14630 + }, + { + "grad_norm": 0.2500159740447998, + "learning_rate": 1.8390705056263906e-05, + "loss": 0.0056, + "step": 14640 + }, + { + "grad_norm": 0.2781897485256195, + "learning_rate": 1.832669148960057e-05, + "loss": 0.0049, + "step": 14650 + }, + { + "grad_norm": 0.17028845846652985, + "learning_rate": 1.8262764516553233e-05, + "loss": 0.0085, + "step": 14660 + }, + { + "grad_norm": 0.2419242262840271, + "learning_rate": 1.8198924311895843e-05, + "loss": 0.0053, + "step": 14670 + }, + { + "grad_norm": 0.13097338378429413, + "learning_rate": 1.813517105016505e-05, + "loss": 0.0059, + "step": 14680 + }, + { + "grad_norm": 0.20473894476890564, + "learning_rate": 1.8071504905659888e-05, + "loss": 0.0055, + "step": 14690 + }, + { + "grad_norm": 0.1552768498659134, + "learning_rate": 1.800792605244109e-05, + "loss": 0.0053, + "step": 14700 + }, + { + "grad_norm": 0.14139609038829803, + "learning_rate": 1.7944434664330844e-05, + "loss": 0.0042, + "step": 14710 + }, + { + "grad_norm": 0.14989487826824188, + "learning_rate": 1.7881030914912212e-05, + "loss": 0.0051, + "step": 14720 + }, + { + "grad_norm": 0.2136131227016449, + "learning_rate": 1.7817714977528577e-05, + "loss": 0.0056, + "step": 14730 + }, + { + "grad_norm": 0.22079499065876007, + "learning_rate": 1.7754487025283332e-05, + "loss": 0.0078, + "step": 14740 + }, + { + "grad_norm": 0.18154866993427277, + "learning_rate": 1.7691347231039275e-05, + "loss": 0.0057, + "step": 14750 + }, + { + "grad_norm": 0.17814518511295319, + "learning_rate": 1.7628295767418164e-05, + "loss": 0.0047, + "step": 14760 + }, + { + "grad_norm": 0.17428627610206604, + "learning_rate": 1.7565332806800333e-05, + "loss": 0.0055, + "step": 14770 + }, + { + "grad_norm": 0.2982683777809143, + "learning_rate": 1.750245852132408e-05, + "loss": 0.0052, + "step": 14780 + }, + { + "grad_norm": 0.2404620200395584, + "learning_rate": 1.7439673082885323e-05, + "loss": 0.0057, + "step": 14790 + }, + { + "grad_norm": 0.14670974016189575, + "learning_rate": 1.7376976663137047e-05, + "loss": 0.0056, + "step": 14800 + }, + { + "grad_norm": 0.2664196193218231, + "learning_rate": 1.7314369433488853e-05, + "loss": 0.0051, + "step": 14810 + }, + { + "grad_norm": 0.19232824444770813, + "learning_rate": 1.7251851565106548e-05, + "loss": 0.0066, + "step": 14820 + }, + { + "grad_norm": 0.1771378368139267, + "learning_rate": 1.7189423228911574e-05, + "loss": 0.0058, + "step": 14830 + }, + { + "grad_norm": 0.17615680396556854, + "learning_rate": 1.7127084595580606e-05, + "loss": 0.0061, + "step": 14840 + }, + { + "grad_norm": 0.20852655172348022, + "learning_rate": 1.706483583554513e-05, + "loss": 0.0051, + "step": 14850 + }, + { + "grad_norm": 0.16545358300209045, + "learning_rate": 1.700267711899083e-05, + "loss": 0.0042, + "step": 14860 + }, + { + "grad_norm": 0.20421037077903748, + "learning_rate": 1.69406086158573e-05, + "loss": 0.0044, + "step": 14870 + }, + { + "grad_norm": 0.19449642300605774, + "learning_rate": 1.6878630495837455e-05, + "loss": 0.0061, + "step": 14880 + }, + { + "grad_norm": 0.23133471608161926, + "learning_rate": 1.681674292837707e-05, + "loss": 0.0064, + "step": 14890 + }, + { + "grad_norm": 0.14292384684085846, + "learning_rate": 1.6754946082674444e-05, + "loss": 0.0058, + "step": 14900 + }, + { + "grad_norm": 0.12971968948841095, + "learning_rate": 1.6693240127679748e-05, + "loss": 0.0044, + "step": 14910 + }, + { + "grad_norm": 0.1865580976009369, + "learning_rate": 1.663162523209475e-05, + "loss": 0.0071, + "step": 14920 + }, + { + "grad_norm": 0.18029199540615082, + "learning_rate": 1.6570101564372193e-05, + "loss": 0.007, + "step": 14930 + }, + { + "grad_norm": 0.24925662577152252, + "learning_rate": 1.650866929271543e-05, + "loss": 0.0062, + "step": 14940 + }, + { + "grad_norm": 0.17494648694992065, + "learning_rate": 1.644732858507797e-05, + "loss": 0.0055, + "step": 14950 + }, + { + "grad_norm": 0.15730054676532745, + "learning_rate": 1.6386079609162943e-05, + "loss": 0.0047, + "step": 14960 + }, + { + "grad_norm": 0.2508449852466583, + "learning_rate": 1.6324922532422742e-05, + "loss": 0.0098, + "step": 14970 + }, + { + "grad_norm": 0.190720796585083, + "learning_rate": 1.6263857522058434e-05, + "loss": 0.0054, + "step": 14980 + }, + { + "grad_norm": 0.1414359211921692, + "learning_rate": 1.6202884745019443e-05, + "loss": 0.0046, + "step": 14990 + }, + { + "grad_norm": 0.14060048758983612, + "learning_rate": 1.614200436800304e-05, + "loss": 0.0061, + "step": 15000 + }, + { + "grad_norm": 0.19313286244869232, + "learning_rate": 1.6081216557453814e-05, + "loss": 0.0049, + "step": 15010 + }, + { + "grad_norm": 0.17659884691238403, + "learning_rate": 1.6020521479563367e-05, + "loss": 0.0054, + "step": 15020 + }, + { + "grad_norm": 0.14806364476680756, + "learning_rate": 1.5959919300269654e-05, + "loss": 0.0041, + "step": 15030 + }, + { + "grad_norm": 0.15692979097366333, + "learning_rate": 1.5899410185256764e-05, + "loss": 0.0037, + "step": 15040 + }, + { + "grad_norm": 0.14825497567653656, + "learning_rate": 1.583899429995431e-05, + "loss": 0.0046, + "step": 15050 + }, + { + "grad_norm": 0.1588299572467804, + "learning_rate": 1.5778671809536993e-05, + "loss": 0.0043, + "step": 15060 + }, + { + "grad_norm": 0.13866159319877625, + "learning_rate": 1.5718442878924246e-05, + "loss": 0.0044, + "step": 15070 + }, + { + "grad_norm": 0.2178276777267456, + "learning_rate": 1.5658307672779593e-05, + "loss": 0.0055, + "step": 15080 + }, + { + "grad_norm": 0.1646864116191864, + "learning_rate": 1.5598266355510427e-05, + "loss": 0.0045, + "step": 15090 + }, + { + "grad_norm": 0.21160311996936798, + "learning_rate": 1.553831909126744e-05, + "loss": 0.0049, + "step": 15100 + }, + { + "grad_norm": 0.23382562398910522, + "learning_rate": 1.5478466043944135e-05, + "loss": 0.0066, + "step": 15110 + }, + { + "grad_norm": 0.18930543959140778, + "learning_rate": 1.5418707377176468e-05, + "loss": 0.0053, + "step": 15120 + }, + { + "grad_norm": 0.1883104145526886, + "learning_rate": 1.535904325434233e-05, + "loss": 0.0051, + "step": 15130 + }, + { + "grad_norm": 0.21340157091617584, + "learning_rate": 1.529947383856118e-05, + "loss": 0.0073, + "step": 15140 + }, + { + "grad_norm": 0.23821984231472015, + "learning_rate": 1.5239999292693524e-05, + "loss": 0.0067, + "step": 15150 + }, + { + "grad_norm": 0.14447997510433197, + "learning_rate": 1.5180619779340505e-05, + "loss": 0.0067, + "step": 15160 + }, + { + "grad_norm": 0.14597798883914948, + "learning_rate": 1.5121335460843428e-05, + "loss": 0.0051, + "step": 15170 + }, + { + "grad_norm": 0.19506734609603882, + "learning_rate": 1.5062146499283347e-05, + "loss": 0.0062, + "step": 15180 + }, + { + "grad_norm": 0.18002669513225555, + "learning_rate": 1.5003053056480643e-05, + "loss": 0.007, + "step": 15190 + }, + { + "grad_norm": 0.2782810926437378, + "learning_rate": 1.4944055293994551e-05, + "loss": 0.0051, + "step": 15200 + }, + { + "grad_norm": 0.2039763480424881, + "learning_rate": 1.4885153373122656e-05, + "loss": 0.006, + "step": 15210 + }, + { + "grad_norm": 0.1816200613975525, + "learning_rate": 1.482634745490059e-05, + "loss": 0.005, + "step": 15220 + }, + { + "grad_norm": 0.15879617631435394, + "learning_rate": 1.4767637700101466e-05, + "loss": 0.0047, + "step": 15230 + }, + { + "grad_norm": 0.19043923914432526, + "learning_rate": 1.4709024269235528e-05, + "loss": 0.0051, + "step": 15240 + }, + { + "grad_norm": 0.2104177474975586, + "learning_rate": 1.4650507322549684e-05, + "loss": 0.0047, + "step": 15250 + }, + { + "grad_norm": 0.22558549046516418, + "learning_rate": 1.4592087020026972e-05, + "loss": 0.0055, + "step": 15260 + }, + { + "grad_norm": 0.1903344690799713, + "learning_rate": 1.4533763521386318e-05, + "loss": 0.0051, + "step": 15270 + }, + { + "grad_norm": 0.21042105555534363, + "learning_rate": 1.44755369860819e-05, + "loss": 0.0059, + "step": 15280 + }, + { + "grad_norm": 0.2088128626346588, + "learning_rate": 1.441740757330287e-05, + "loss": 0.0048, + "step": 15290 + }, + { + "grad_norm": 0.2233678549528122, + "learning_rate": 1.4359375441972844e-05, + "loss": 0.0043, + "step": 15300 + }, + { + "grad_norm": 0.13520531356334686, + "learning_rate": 1.4301440750749395e-05, + "loss": 0.0038, + "step": 15310 + }, + { + "grad_norm": 0.21610109508037567, + "learning_rate": 1.4243603658023808e-05, + "loss": 0.0056, + "step": 15320 + }, + { + "grad_norm": 0.16219377517700195, + "learning_rate": 1.4185864321920444e-05, + "loss": 0.0043, + "step": 15330 + }, + { + "grad_norm": 0.2393598109483719, + "learning_rate": 1.4128222900296485e-05, + "loss": 0.0057, + "step": 15340 + }, + { + "grad_norm": 0.16220656037330627, + "learning_rate": 1.407067955074135e-05, + "loss": 0.0056, + "step": 15350 + }, + { + "grad_norm": 0.22726091742515564, + "learning_rate": 1.4013234430576356e-05, + "loss": 0.0054, + "step": 15360 + }, + { + "grad_norm": 0.1691230684518814, + "learning_rate": 1.3955887696854286e-05, + "loss": 0.0066, + "step": 15370 + }, + { + "grad_norm": 0.24777837097644806, + "learning_rate": 1.38986395063589e-05, + "loss": 0.0073, + "step": 15380 + }, + { + "grad_norm": 0.2199578583240509, + "learning_rate": 1.3841490015604597e-05, + "loss": 0.0059, + "step": 15390 + }, + { + "grad_norm": 0.22531665861606598, + "learning_rate": 1.3784439380835879e-05, + "loss": 0.0068, + "step": 15400 + }, + { + "grad_norm": 0.16674892604351044, + "learning_rate": 1.3727487758026986e-05, + "loss": 0.0055, + "step": 15410 + }, + { + "grad_norm": 0.19342897832393646, + "learning_rate": 1.3670635302881525e-05, + "loss": 0.0056, + "step": 15420 + }, + { + "grad_norm": 0.12783855199813843, + "learning_rate": 1.3613882170831888e-05, + "loss": 0.0066, + "step": 15430 + }, + { + "grad_norm": 0.17184005677700043, + "learning_rate": 1.355722851703901e-05, + "loss": 0.0069, + "step": 15440 + }, + { + "grad_norm": 0.20452845096588135, + "learning_rate": 1.3500674496391814e-05, + "loss": 0.0047, + "step": 15450 + }, + { + "grad_norm": 0.13389019668102264, + "learning_rate": 1.3444220263506795e-05, + "loss": 0.0066, + "step": 15460 + }, + { + "grad_norm": 0.24266892671585083, + "learning_rate": 1.3387865972727714e-05, + "loss": 0.0067, + "step": 15470 + }, + { + "grad_norm": 0.1648254692554474, + "learning_rate": 1.3331611778125036e-05, + "loss": 0.0065, + "step": 15480 + }, + { + "grad_norm": 0.20072923600673676, + "learning_rate": 1.3275457833495564e-05, + "loss": 0.0102, + "step": 15490 + }, + { + "grad_norm": 0.18705320358276367, + "learning_rate": 1.3219404292362065e-05, + "loss": 0.0077, + "step": 15500 + }, + { + "grad_norm": 0.1857132464647293, + "learning_rate": 1.3163451307972751e-05, + "loss": 0.0067, + "step": 15510 + }, + { + "grad_norm": 0.2010364979505539, + "learning_rate": 1.3107599033300977e-05, + "loss": 0.006, + "step": 15520 + }, + { + "grad_norm": 0.1669953614473343, + "learning_rate": 1.305184762104471e-05, + "loss": 0.006, + "step": 15530 + }, + { + "grad_norm": 0.16274145245552063, + "learning_rate": 1.2996197223626178e-05, + "loss": 0.0041, + "step": 15540 + }, + { + "grad_norm": 0.18755671381950378, + "learning_rate": 1.2940647993191457e-05, + "loss": 0.0053, + "step": 15550 + }, + { + "grad_norm": 0.20959579944610596, + "learning_rate": 1.2885200081610005e-05, + "loss": 0.005, + "step": 15560 + }, + { + "grad_norm": 0.14389164745807648, + "learning_rate": 1.2829853640474316e-05, + "loss": 0.0045, + "step": 15570 + }, + { + "grad_norm": 0.15106278657913208, + "learning_rate": 1.2774608821099438e-05, + "loss": 0.0044, + "step": 15580 + }, + { + "grad_norm": 0.1607772409915924, + "learning_rate": 1.2719465774522577e-05, + "loss": 0.0055, + "step": 15590 + }, + { + "grad_norm": 0.1950947344303131, + "learning_rate": 1.2664424651502755e-05, + "loss": 0.0064, + "step": 15600 + }, + { + "grad_norm": 0.20571626722812653, + "learning_rate": 1.260948560252026e-05, + "loss": 0.0054, + "step": 15610 + }, + { + "grad_norm": 0.1459115594625473, + "learning_rate": 1.2554648777776396e-05, + "loss": 0.0048, + "step": 15620 + }, + { + "grad_norm": 0.1424199789762497, + "learning_rate": 1.2499914327192919e-05, + "loss": 0.0056, + "step": 15630 + }, + { + "grad_norm": 0.13897298276424408, + "learning_rate": 1.2445282400411722e-05, + "loss": 0.0055, + "step": 15640 + }, + { + "grad_norm": 0.13337059319019318, + "learning_rate": 1.2390753146794437e-05, + "loss": 0.0052, + "step": 15650 + }, + { + "grad_norm": 0.22809644043445587, + "learning_rate": 1.2336326715421925e-05, + "loss": 0.0055, + "step": 15660 + }, + { + "grad_norm": 0.13422560691833496, + "learning_rate": 1.2282003255094005e-05, + "loss": 0.0051, + "step": 15670 + }, + { + "grad_norm": 0.18627096712589264, + "learning_rate": 1.2227782914328928e-05, + "loss": 0.0052, + "step": 15680 + }, + { + "grad_norm": 0.18976987898349762, + "learning_rate": 1.2173665841363018e-05, + "loss": 0.0044, + "step": 15690 + }, + { + "grad_norm": 0.16836141049861908, + "learning_rate": 1.211965218415032e-05, + "loss": 0.0041, + "step": 15700 + }, + { + "grad_norm": 0.14203710854053497, + "learning_rate": 1.2065742090362082e-05, + "loss": 0.0055, + "step": 15710 + }, + { + "grad_norm": 0.16482926905155182, + "learning_rate": 1.2011935707386457e-05, + "loss": 0.0055, + "step": 15720 + }, + { + "grad_norm": 0.23353098332881927, + "learning_rate": 1.1958233182328044e-05, + "loss": 0.0052, + "step": 15730 + }, + { + "grad_norm": 0.2196880280971527, + "learning_rate": 1.1904634662007474e-05, + "loss": 0.0052, + "step": 15740 + }, + { + "grad_norm": 0.20127354562282562, + "learning_rate": 1.1851140292961088e-05, + "loss": 0.0052, + "step": 15750 + }, + { + "grad_norm": 0.19632349908351898, + "learning_rate": 1.1797750221440424e-05, + "loss": 0.006, + "step": 15760 + }, + { + "grad_norm": 0.22671037912368774, + "learning_rate": 1.1744464593411897e-05, + "loss": 0.0061, + "step": 15770 + }, + { + "grad_norm": 0.1899847388267517, + "learning_rate": 1.1691283554556399e-05, + "loss": 0.005, + "step": 15780 + }, + { + "grad_norm": 0.2320375293493271, + "learning_rate": 1.1638207250268834e-05, + "loss": 0.0048, + "step": 15790 + }, + { + "grad_norm": 0.1803273856639862, + "learning_rate": 1.158523582565782e-05, + "loss": 0.0052, + "step": 15800 + }, + { + "grad_norm": 0.12426243722438812, + "learning_rate": 1.1532369425545192e-05, + "loss": 0.0037, + "step": 15810 + }, + { + "grad_norm": 0.15763874351978302, + "learning_rate": 1.1479608194465662e-05, + "loss": 0.0092, + "step": 15820 + }, + { + "grad_norm": 0.16793958842754364, + "learning_rate": 1.1426952276666442e-05, + "loss": 0.0044, + "step": 15830 + }, + { + "grad_norm": 0.18338541686534882, + "learning_rate": 1.1374401816106778e-05, + "loss": 0.0073, + "step": 15840 + }, + { + "grad_norm": 0.16605933010578156, + "learning_rate": 1.1321956956457646e-05, + "loss": 0.005, + "step": 15850 + }, + { + "grad_norm": 0.23071573674678802, + "learning_rate": 1.1269617841101277e-05, + "loss": 0.0058, + "step": 15860 + }, + { + "grad_norm": 0.16531772911548615, + "learning_rate": 1.1217384613130804e-05, + "loss": 0.0046, + "step": 15870 + }, + { + "grad_norm": 0.17097795009613037, + "learning_rate": 1.11652574153499e-05, + "loss": 0.0056, + "step": 15880 + }, + { + "grad_norm": 0.14396753907203674, + "learning_rate": 1.1113236390272303e-05, + "loss": 0.0052, + "step": 15890 + }, + { + "grad_norm": 0.19578398764133453, + "learning_rate": 1.106132168012155e-05, + "loss": 0.0074, + "step": 15900 + }, + { + "grad_norm": 0.15604498982429504, + "learning_rate": 1.1009513426830448e-05, + "loss": 0.0057, + "step": 15910 + }, + { + "grad_norm": 0.1655254364013672, + "learning_rate": 1.0957811772040777e-05, + "loss": 0.0081, + "step": 15920 + }, + { + "grad_norm": 0.20122294127941132, + "learning_rate": 1.0906216857102913e-05, + "loss": 0.005, + "step": 15930 + }, + { + "grad_norm": 0.1938413679599762, + "learning_rate": 1.0854728823075355e-05, + "loss": 0.0055, + "step": 15940 + }, + { + "grad_norm": 0.14376655220985413, + "learning_rate": 1.0803347810724452e-05, + "loss": 0.0068, + "step": 15950 + }, + { + "grad_norm": 0.14233747124671936, + "learning_rate": 1.0752073960523911e-05, + "loss": 0.0051, + "step": 15960 + }, + { + "grad_norm": 0.14280758798122406, + "learning_rate": 1.070090741265447e-05, + "loss": 0.0047, + "step": 15970 + }, + { + "grad_norm": 0.14656981825828552, + "learning_rate": 1.0649848307003547e-05, + "loss": 0.0051, + "step": 15980 + }, + { + "grad_norm": 0.13522081077098846, + "learning_rate": 1.0598896783164757e-05, + "loss": 0.0049, + "step": 15990 + }, + { + "grad_norm": 0.12241239100694656, + "learning_rate": 1.0548052980437645e-05, + "loss": 0.0073, + "step": 16000 + }, + { + "grad_norm": 0.1648959517478943, + "learning_rate": 1.049731703782722e-05, + "loss": 0.0055, + "step": 16010 + }, + { + "grad_norm": 0.13855500519275665, + "learning_rate": 1.0446689094043587e-05, + "loss": 0.0073, + "step": 16020 + }, + { + "grad_norm": 0.1769500970840454, + "learning_rate": 1.039616928750165e-05, + "loss": 0.0049, + "step": 16030 + }, + { + "grad_norm": 0.15365423262119293, + "learning_rate": 1.0345757756320612e-05, + "loss": 0.0043, + "step": 16040 + }, + { + "grad_norm": 0.1871398538351059, + "learning_rate": 1.0295454638323666e-05, + "loss": 0.0055, + "step": 16050 + }, + { + "grad_norm": 0.22193288803100586, + "learning_rate": 1.0245260071037632e-05, + "loss": 0.0053, + "step": 16060 + }, + { + "grad_norm": 0.18220891058444977, + "learning_rate": 1.0195174191692518e-05, + "loss": 0.0043, + "step": 16070 + }, + { + "grad_norm": 0.15648461878299713, + "learning_rate": 1.014519713722124e-05, + "loss": 0.0052, + "step": 16080 + }, + { + "grad_norm": 0.18102756142616272, + "learning_rate": 1.0095329044259132e-05, + "loss": 0.0043, + "step": 16090 + }, + { + "grad_norm": 0.14919497072696686, + "learning_rate": 1.004557004914365e-05, + "loss": 0.005, + "step": 16100 + }, + { + "grad_norm": 0.17376764118671417, + "learning_rate": 9.995920287914007e-06, + "loss": 0.0042, + "step": 16110 + }, + { + "grad_norm": 0.13198114931583405, + "learning_rate": 9.946379896310737e-06, + "loss": 0.004, + "step": 16120 + }, + { + "grad_norm": 0.16041940450668335, + "learning_rate": 9.896949009775396e-06, + "loss": 0.0048, + "step": 16130 + }, + { + "grad_norm": 0.20756278932094574, + "learning_rate": 9.847627763450134e-06, + "loss": 0.0049, + "step": 16140 + }, + { + "grad_norm": 0.22414100170135498, + "learning_rate": 9.798416292177337e-06, + "loss": 0.0045, + "step": 16150 + }, + { + "grad_norm": 0.22440825402736664, + "learning_rate": 9.74931473049932e-06, + "loss": 0.0054, + "step": 16160 + }, + { + "grad_norm": 0.1703796684741974, + "learning_rate": 9.700323212657847e-06, + "loss": 0.0042, + "step": 16170 + }, + { + "grad_norm": 0.15423569083213806, + "learning_rate": 9.65144187259388e-06, + "loss": 0.0053, + "step": 16180 + }, + { + "grad_norm": 0.16278226673603058, + "learning_rate": 9.602670843947132e-06, + "loss": 0.0049, + "step": 16190 + }, + { + "grad_norm": 0.1522931307554245, + "learning_rate": 9.554010260055713e-06, + "loss": 0.0046, + "step": 16200 + }, + { + "grad_norm": 0.1324155330657959, + "learning_rate": 9.505460253955834e-06, + "loss": 0.0044, + "step": 16210 + }, + { + "grad_norm": 0.12835343182086945, + "learning_rate": 9.457020958381324e-06, + "loss": 0.0055, + "step": 16220 + }, + { + "grad_norm": 0.13485650718212128, + "learning_rate": 9.408692505763395e-06, + "loss": 0.0048, + "step": 16230 + }, + { + "grad_norm": 0.17823177576065063, + "learning_rate": 9.360475028230181e-06, + "loss": 0.0051, + "step": 16240 + }, + { + "grad_norm": 0.13854046165943146, + "learning_rate": 9.312368657606412e-06, + "loss": 0.0038, + "step": 16250 + }, + { + "grad_norm": 0.20396800339221954, + "learning_rate": 9.264373525413096e-06, + "loss": 0.0072, + "step": 16260 + }, + { + "grad_norm": 0.13611064851284027, + "learning_rate": 9.216489762867058e-06, + "loss": 0.0042, + "step": 16270 + }, + { + "grad_norm": 0.23162850737571716, + "learning_rate": 9.168717500880708e-06, + "loss": 0.0068, + "step": 16280 + }, + { + "grad_norm": 0.14248082041740417, + "learning_rate": 9.121056870061574e-06, + "loss": 0.0035, + "step": 16290 + }, + { + "grad_norm": 0.20380237698554993, + "learning_rate": 9.073508000711983e-06, + "loss": 0.0048, + "step": 16300 + }, + { + "grad_norm": 0.18490472435951233, + "learning_rate": 9.026071022828758e-06, + "loss": 0.0062, + "step": 16310 + }, + { + "grad_norm": 0.22372393310070038, + "learning_rate": 8.978746066102771e-06, + "loss": 0.0049, + "step": 16320 + }, + { + "grad_norm": 0.15033039450645447, + "learning_rate": 8.931533259918634e-06, + "loss": 0.0037, + "step": 16330 + }, + { + "grad_norm": 0.13687333464622498, + "learning_rate": 8.884432733354382e-06, + "loss": 0.004, + "step": 16340 + }, + { + "grad_norm": 0.15738479793071747, + "learning_rate": 8.837444615181029e-06, + "loss": 0.005, + "step": 16350 + }, + { + "grad_norm": 0.2072679102420807, + "learning_rate": 8.790569033862323e-06, + "loss": 0.0062, + "step": 16360 + }, + { + "grad_norm": 0.19508348405361176, + "learning_rate": 8.7438061175543e-06, + "loss": 0.0039, + "step": 16370 + }, + { + "grad_norm": 0.18191103637218475, + "learning_rate": 8.697155994104978e-06, + "loss": 0.0037, + "step": 16380 + }, + { + "grad_norm": 0.16201141476631165, + "learning_rate": 8.650618791054033e-06, + "loss": 0.0046, + "step": 16390 + }, + { + "grad_norm": 0.1699906885623932, + "learning_rate": 8.604194635632373e-06, + "loss": 0.0056, + "step": 16400 + }, + { + "grad_norm": 0.14311489462852478, + "learning_rate": 8.557883654761906e-06, + "loss": 0.0051, + "step": 16410 + }, + { + "grad_norm": 0.17263031005859375, + "learning_rate": 8.511685975055061e-06, + "loss": 0.0052, + "step": 16420 + }, + { + "grad_norm": 0.15772181749343872, + "learning_rate": 8.46560172281452e-06, + "loss": 0.0038, + "step": 16430 + }, + { + "grad_norm": 0.1426163911819458, + "learning_rate": 8.419631024032893e-06, + "loss": 0.004, + "step": 16440 + }, + { + "grad_norm": 0.15554185211658478, + "learning_rate": 8.373774004392293e-06, + "loss": 0.0048, + "step": 16450 + }, + { + "grad_norm": 0.16801729798316956, + "learning_rate": 8.32803078926409e-06, + "loss": 0.0058, + "step": 16460 + }, + { + "grad_norm": 0.15965138375759125, + "learning_rate": 8.282401503708454e-06, + "loss": 0.0044, + "step": 16470 + }, + { + "grad_norm": 0.1287917047739029, + "learning_rate": 8.23688627247412e-06, + "loss": 0.0051, + "step": 16480 + }, + { + "grad_norm": 0.16028323769569397, + "learning_rate": 8.191485219998007e-06, + "loss": 0.0042, + "step": 16490 + }, + { + "grad_norm": 0.11672808229923248, + "learning_rate": 8.146198470404843e-06, + "loss": 0.0046, + "step": 16500 + }, + { + "grad_norm": 0.13132800161838531, + "learning_rate": 8.101026147506897e-06, + "loss": 0.003, + "step": 16510 + }, + { + "grad_norm": 0.1448666900396347, + "learning_rate": 8.05596837480353e-06, + "loss": 0.0053, + "step": 16520 + }, + { + "grad_norm": 0.13154615461826324, + "learning_rate": 8.011025275480998e-06, + "loss": 0.0051, + "step": 16530 + }, + { + "grad_norm": 0.1266883760690689, + "learning_rate": 7.966196972412027e-06, + "loss": 0.0037, + "step": 16540 + }, + { + "grad_norm": 0.1132773905992508, + "learning_rate": 7.92148358815547e-06, + "loss": 0.0034, + "step": 16550 + }, + { + "grad_norm": 0.16018636524677277, + "learning_rate": 7.87688524495604e-06, + "loss": 0.0049, + "step": 16560 + }, + { + "grad_norm": 0.1342909187078476, + "learning_rate": 7.83240206474386e-06, + "loss": 0.0054, + "step": 16570 + }, + { + "grad_norm": 0.12503987550735474, + "learning_rate": 7.788034169134272e-06, + "loss": 0.0043, + "step": 16580 + }, + { + "grad_norm": 0.11737470328807831, + "learning_rate": 7.743781679427414e-06, + "loss": 0.0043, + "step": 16590 + }, + { + "grad_norm": 0.11028105020523071, + "learning_rate": 7.699644716607895e-06, + "loss": 0.0038, + "step": 16600 + }, + { + "grad_norm": 0.09312699735164642, + "learning_rate": 7.655623401344486e-06, + "loss": 0.0036, + "step": 16610 + }, + { + "grad_norm": 0.1259354054927826, + "learning_rate": 7.611717853989775e-06, + "loss": 0.0045, + "step": 16620 + }, + { + "grad_norm": 0.14075908064842224, + "learning_rate": 7.567928194579854e-06, + "loss": 0.0058, + "step": 16630 + }, + { + "grad_norm": 0.16171559691429138, + "learning_rate": 7.524254542833997e-06, + "loss": 0.0049, + "step": 16640 + }, + { + "grad_norm": 0.1568921059370041, + "learning_rate": 7.480697018154286e-06, + "loss": 0.009, + "step": 16650 + }, + { + "grad_norm": 0.14268898963928223, + "learning_rate": 7.437255739625332e-06, + "loss": 0.0036, + "step": 16660 + }, + { + "grad_norm": 0.2208922803401947, + "learning_rate": 7.393930826013923e-06, + "loss": 0.0064, + "step": 16670 + }, + { + "grad_norm": 0.1548379510641098, + "learning_rate": 7.350722395768722e-06, + "loss": 0.0043, + "step": 16680 + }, + { + "grad_norm": 0.1384071558713913, + "learning_rate": 7.307630567019963e-06, + "loss": 0.0036, + "step": 16690 + }, + { + "grad_norm": 0.10652858018875122, + "learning_rate": 7.264655457579e-06, + "loss": 0.0041, + "step": 16700 + }, + { + "grad_norm": 0.10340376943349838, + "learning_rate": 7.221797184938184e-06, + "loss": 0.0059, + "step": 16710 + }, + { + "grad_norm": 0.11764318495988846, + "learning_rate": 7.179055866270373e-06, + "loss": 0.0036, + "step": 16720 + }, + { + "grad_norm": 0.21147853136062622, + "learning_rate": 7.136431618428707e-06, + "loss": 0.0045, + "step": 16730 + }, + { + "grad_norm": 0.19434772431850433, + "learning_rate": 7.09392455794628e-06, + "loss": 0.0062, + "step": 16740 + }, + { + "grad_norm": 0.14892810583114624, + "learning_rate": 7.051534801035725e-06, + "loss": 0.0053, + "step": 16750 + }, + { + "grad_norm": 0.17711889743804932, + "learning_rate": 7.00926246358905e-06, + "loss": 0.0055, + "step": 16760 + }, + { + "grad_norm": 0.13662099838256836, + "learning_rate": 6.967107661177191e-06, + "loss": 0.0056, + "step": 16770 + }, + { + "grad_norm": 0.13581331074237823, + "learning_rate": 6.925070509049786e-06, + "loss": 0.0053, + "step": 16780 + }, + { + "grad_norm": 0.09662957489490509, + "learning_rate": 6.883151122134812e-06, + "loss": 0.0037, + "step": 16790 + }, + { + "grad_norm": 0.14342249929904938, + "learning_rate": 6.8413496150382394e-06, + "loss": 0.0047, + "step": 16800 + }, + { + "grad_norm": 0.07608351111412048, + "learning_rate": 6.7996661020438165e-06, + "loss": 0.0029, + "step": 16810 + }, + { + "grad_norm": 0.14846809208393097, + "learning_rate": 6.758100697112662e-06, + "loss": 0.0038, + "step": 16820 + }, + { + "grad_norm": 0.18177205324172974, + "learning_rate": 6.716653513883026e-06, + "loss": 0.005, + "step": 16830 + }, + { + "grad_norm": 0.15042972564697266, + "learning_rate": 6.675324665669913e-06, + "loss": 0.0039, + "step": 16840 + }, + { + "grad_norm": 0.17908424139022827, + "learning_rate": 6.634114265464803e-06, + "loss": 0.0056, + "step": 16850 + }, + { + "grad_norm": 0.1749325394630432, + "learning_rate": 6.59302242593538e-06, + "loss": 0.0041, + "step": 16860 + }, + { + "grad_norm": 0.11305996030569077, + "learning_rate": 6.552049259425141e-06, + "loss": 0.0042, + "step": 16870 + }, + { + "grad_norm": 0.11477645486593246, + "learning_rate": 6.511194877953181e-06, + "loss": 0.0046, + "step": 16880 + }, + { + "grad_norm": 0.1085256040096283, + "learning_rate": 6.470459393213813e-06, + "loss": 0.0046, + "step": 16890 + }, + { + "grad_norm": 0.16111595928668976, + "learning_rate": 6.429842916576279e-06, + "loss": 0.0049, + "step": 16900 + }, + { + "grad_norm": 0.1468374878168106, + "learning_rate": 6.389345559084503e-06, + "loss": 0.0036, + "step": 16910 + }, + { + "grad_norm": 0.17139410972595215, + "learning_rate": 6.348967431456682e-06, + "loss": 0.0064, + "step": 16920 + }, + { + "grad_norm": 0.17251119017601013, + "learning_rate": 6.30870864408511e-06, + "loss": 0.0036, + "step": 16930 + }, + { + "grad_norm": 0.14736191928386688, + "learning_rate": 6.268569307035754e-06, + "loss": 0.0041, + "step": 16940 + }, + { + "grad_norm": 0.1500120311975479, + "learning_rate": 6.228549530048022e-06, + "loss": 0.0058, + "step": 16950 + }, + { + "grad_norm": 0.14884720742702484, + "learning_rate": 6.1886494225344814e-06, + "loss": 0.0064, + "step": 16960 + }, + { + "grad_norm": 0.15160712599754333, + "learning_rate": 6.148869093580479e-06, + "loss": 0.0056, + "step": 16970 + }, + { + "grad_norm": 0.1517210155725479, + "learning_rate": 6.109208651943921e-06, + "loss": 0.0047, + "step": 16980 + }, + { + "grad_norm": 0.155134379863739, + "learning_rate": 6.069668206054946e-06, + "loss": 0.0038, + "step": 16990 + }, + { + "grad_norm": 0.14380772411823273, + "learning_rate": 6.0302478640156145e-06, + "loss": 0.0053, + "step": 17000 + }, + { + "grad_norm": 0.16657549142837524, + "learning_rate": 5.990947733599644e-06, + "loss": 0.0068, + "step": 17010 + }, + { + "grad_norm": 0.16626840829849243, + "learning_rate": 5.951767922252105e-06, + "loss": 0.0059, + "step": 17020 + }, + { + "grad_norm": 0.14325283467769623, + "learning_rate": 5.912708537089068e-06, + "loss": 0.0039, + "step": 17030 + }, + { + "grad_norm": 0.15190692245960236, + "learning_rate": 5.873769684897434e-06, + "loss": 0.005, + "step": 17040 + }, + { + "grad_norm": 0.13617077469825745, + "learning_rate": 5.834951472134514e-06, + "loss": 0.0059, + "step": 17050 + }, + { + "grad_norm": 0.16800183057785034, + "learning_rate": 5.796254004927832e-06, + "loss": 0.0052, + "step": 17060 + }, + { + "grad_norm": 0.1636773645877838, + "learning_rate": 5.757677389074806e-06, + "loss": 0.0047, + "step": 17070 + }, + { + "grad_norm": 0.11355911195278168, + "learning_rate": 5.719221730042385e-06, + "loss": 0.0033, + "step": 17080 + }, + { + "grad_norm": 0.1289564073085785, + "learning_rate": 5.680887132966911e-06, + "loss": 0.0057, + "step": 17090 + }, + { + "grad_norm": 0.14525209367275238, + "learning_rate": 5.642673702653683e-06, + "loss": 0.004, + "step": 17100 + }, + { + "grad_norm": 0.18314318358898163, + "learning_rate": 5.604581543576781e-06, + "loss": 0.0041, + "step": 17110 + }, + { + "grad_norm": 0.17528216540813446, + "learning_rate": 5.566610759878704e-06, + "loss": 0.0052, + "step": 17120 + }, + { + "grad_norm": 0.1443951427936554, + "learning_rate": 5.528761455370119e-06, + "loss": 0.0041, + "step": 17130 + }, + { + "grad_norm": 0.20061659812927246, + "learning_rate": 5.491033733529594e-06, + "loss": 0.0051, + "step": 17140 + }, + { + "grad_norm": 0.12190601229667664, + "learning_rate": 5.453427697503255e-06, + "loss": 0.0049, + "step": 17150 + }, + { + "grad_norm": 0.13548479974269867, + "learning_rate": 5.415943450104599e-06, + "loss": 0.0047, + "step": 17160 + }, + { + "grad_norm": 0.15832222998142242, + "learning_rate": 5.378581093814111e-06, + "loss": 0.0037, + "step": 17170 + }, + { + "grad_norm": 0.10824442654848099, + "learning_rate": 5.3413407307790375e-06, + "loss": 0.0041, + "step": 17180 + }, + { + "grad_norm": 0.18595543503761292, + "learning_rate": 5.30422246281313e-06, + "loss": 0.0063, + "step": 17190 + }, + { + "grad_norm": 0.13622011244297028, + "learning_rate": 5.267226391396296e-06, + "loss": 0.0032, + "step": 17200 + }, + { + "grad_norm": 0.10188230872154236, + "learning_rate": 5.2303526176744e-06, + "loss": 0.003, + "step": 17210 + }, + { + "grad_norm": 0.13925378024578094, + "learning_rate": 5.193601242458929e-06, + "loss": 0.0043, + "step": 17220 + }, + { + "grad_norm": 0.11982742697000504, + "learning_rate": 5.156972366226714e-06, + "loss": 0.0058, + "step": 17230 + }, + { + "grad_norm": 0.1453980952501297, + "learning_rate": 5.120466089119735e-06, + "loss": 0.0052, + "step": 17240 + }, + { + "grad_norm": 0.12326829135417938, + "learning_rate": 5.084082510944749e-06, + "loss": 0.0039, + "step": 17250 + }, + { + "grad_norm": 0.17726564407348633, + "learning_rate": 5.047821731173058e-06, + "loss": 0.0047, + "step": 17260 + }, + { + "grad_norm": 0.15553253889083862, + "learning_rate": 5.011683848940274e-06, + "loss": 0.0042, + "step": 17270 + }, + { + "grad_norm": 0.11963417381048203, + "learning_rate": 4.975668963045954e-06, + "loss": 0.004, + "step": 17280 + }, + { + "grad_norm": 0.12937189638614655, + "learning_rate": 4.9397771719534525e-06, + "loss": 0.0032, + "step": 17290 + }, + { + "grad_norm": 0.12897104024887085, + "learning_rate": 4.904008573789548e-06, + "loss": 0.0045, + "step": 17300 + }, + { + "grad_norm": 0.09982036799192429, + "learning_rate": 4.8683632663442005e-06, + "loss": 0.0035, + "step": 17310 + }, + { + "grad_norm": 0.15100431442260742, + "learning_rate": 4.832841347070343e-06, + "loss": 0.0047, + "step": 17320 + }, + { + "grad_norm": 0.175477996468544, + "learning_rate": 4.797442913083539e-06, + "loss": 0.0059, + "step": 17330 + }, + { + "grad_norm": 0.14300963282585144, + "learning_rate": 4.7621680611617596e-06, + "loss": 0.0038, + "step": 17340 + }, + { + "grad_norm": 0.2541049122810364, + "learning_rate": 4.727016887745095e-06, + "loss": 0.0053, + "step": 17350 + }, + { + "grad_norm": 0.12595532834529877, + "learning_rate": 4.691989488935511e-06, + "loss": 0.0041, + "step": 17360 + }, + { + "grad_norm": 0.11065539717674255, + "learning_rate": 4.657085960496588e-06, + "loss": 0.0036, + "step": 17370 + }, + { + "grad_norm": 0.09352098405361176, + "learning_rate": 4.6223063978532265e-06, + "loss": 0.0052, + "step": 17380 + }, + { + "grad_norm": 0.09857888519763947, + "learning_rate": 4.587650896091439e-06, + "loss": 0.0046, + "step": 17390 + }, + { + "grad_norm": 0.1510113924741745, + "learning_rate": 4.553119549958035e-06, + "loss": 0.0059, + "step": 17400 + }, + { + "grad_norm": 0.14610646665096283, + "learning_rate": 4.518712453860385e-06, + "loss": 0.0047, + "step": 17410 + }, + { + "grad_norm": 0.10412902384996414, + "learning_rate": 4.484429701866205e-06, + "loss": 0.0038, + "step": 17420 + }, + { + "grad_norm": 0.1347024291753769, + "learning_rate": 4.4502713877031975e-06, + "loss": 0.004, + "step": 17430 + }, + { + "grad_norm": 0.11108066141605377, + "learning_rate": 4.416237604758911e-06, + "loss": 0.0044, + "step": 17440 + }, + { + "grad_norm": 0.17151008546352386, + "learning_rate": 4.3823284460804025e-06, + "loss": 0.0054, + "step": 17450 + }, + { + "grad_norm": 0.1379305124282837, + "learning_rate": 4.348544004374011e-06, + "loss": 0.0033, + "step": 17460 + }, + { + "grad_norm": 0.1281946748495102, + "learning_rate": 4.314884372005123e-06, + "loss": 0.0049, + "step": 17470 + }, + { + "grad_norm": 0.15030136704444885, + "learning_rate": 4.281349640997867e-06, + "loss": 0.0044, + "step": 17480 + }, + { + "grad_norm": 0.11368031054735184, + "learning_rate": 4.247939903034942e-06, + "loss": 0.0045, + "step": 17490 + }, + { + "grad_norm": 0.10988582670688629, + "learning_rate": 4.214655249457284e-06, + "loss": 0.0043, + "step": 17500 + }, + { + "grad_norm": 0.13686129450798035, + "learning_rate": 4.181495771263855e-06, + "loss": 0.0047, + "step": 17510 + }, + { + "grad_norm": 0.08308565616607666, + "learning_rate": 4.148461559111427e-06, + "loss": 0.003, + "step": 17520 + }, + { + "grad_norm": 0.10827736556529999, + "learning_rate": 4.115552703314252e-06, + "loss": 0.0044, + "step": 17530 + }, + { + "grad_norm": 0.10387372225522995, + "learning_rate": 4.082769293843886e-06, + "loss": 0.0036, + "step": 17540 + }, + { + "grad_norm": 0.06729380041360855, + "learning_rate": 4.050111420328939e-06, + "loss": 0.0034, + "step": 17550 + }, + { + "grad_norm": 0.1021871566772461, + "learning_rate": 4.017579172054764e-06, + "loss": 0.0039, + "step": 17560 + }, + { + "grad_norm": 0.1491686850786209, + "learning_rate": 3.985172637963308e-06, + "loss": 0.0043, + "step": 17570 + }, + { + "grad_norm": 0.14715075492858887, + "learning_rate": 3.952891906652784e-06, + "loss": 0.0036, + "step": 17580 + }, + { + "grad_norm": 0.10194284468889236, + "learning_rate": 3.920737066377478e-06, + "loss": 0.0051, + "step": 17590 + }, + { + "grad_norm": 0.09443909674882889, + "learning_rate": 3.888708205047509e-06, + "loss": 0.003, + "step": 17600 + }, + { + "grad_norm": 0.13160809874534607, + "learning_rate": 3.856805410228542e-06, + "loss": 0.0045, + "step": 17610 + }, + { + "grad_norm": 0.10030721873044968, + "learning_rate": 3.82502876914162e-06, + "loss": 0.0038, + "step": 17620 + }, + { + "grad_norm": 0.11953095346689224, + "learning_rate": 3.7933783686628586e-06, + "loss": 0.005, + "step": 17630 + }, + { + "grad_norm": 0.09695965051651001, + "learning_rate": 3.7618542953232306e-06, + "loss": 0.0046, + "step": 17640 + }, + { + "grad_norm": 0.07293650507926941, + "learning_rate": 3.7304566353083658e-06, + "loss": 0.0029, + "step": 17650 + }, + { + "grad_norm": 0.09447503089904785, + "learning_rate": 3.6991854744582555e-06, + "loss": 0.003, + "step": 17660 + }, + { + "grad_norm": 0.127981036901474, + "learning_rate": 3.6680408982670777e-06, + "loss": 0.0039, + "step": 17670 + }, + { + "grad_norm": 0.10943655669689178, + "learning_rate": 3.637022991882899e-06, + "loss": 0.0054, + "step": 17680 + }, + { + "grad_norm": 0.1618594229221344, + "learning_rate": 3.606131840107485e-06, + "loss": 0.0046, + "step": 17690 + }, + { + "grad_norm": 0.15760213136672974, + "learning_rate": 3.575367527396084e-06, + "loss": 0.0045, + "step": 17700 + }, + { + "grad_norm": 0.11885190010070801, + "learning_rate": 3.5447301378571386e-06, + "loss": 0.0034, + "step": 17710 + }, + { + "grad_norm": 0.10916793346405029, + "learning_rate": 3.514219755252113e-06, + "loss": 0.0054, + "step": 17720 + }, + { + "grad_norm": 0.11321078985929489, + "learning_rate": 3.4838364629952213e-06, + "loss": 0.0045, + "step": 17730 + }, + { + "grad_norm": 0.08594832569360733, + "learning_rate": 3.4535803441532123e-06, + "loss": 0.0066, + "step": 17740 + }, + { + "grad_norm": 0.13212276995182037, + "learning_rate": 3.4234514814451836e-06, + "loss": 0.0037, + "step": 17750 + }, + { + "grad_norm": 0.1763709932565689, + "learning_rate": 3.393449957242273e-06, + "loss": 0.0076, + "step": 17760 + }, + { + "grad_norm": 0.13847731053829193, + "learning_rate": 3.363575853567524e-06, + "loss": 0.0043, + "step": 17770 + }, + { + "grad_norm": 0.1049138754606247, + "learning_rate": 3.3338292520955826e-06, + "loss": 0.0049, + "step": 17780 + }, + { + "grad_norm": 0.11970741301774979, + "learning_rate": 3.304210234152516e-06, + "loss": 0.0046, + "step": 17790 + }, + { + "grad_norm": 0.1296226680278778, + "learning_rate": 3.2747188807155993e-06, + "loss": 0.0038, + "step": 17800 + }, + { + "grad_norm": 0.09861905127763748, + "learning_rate": 3.2453552724130643e-06, + "loss": 0.0053, + "step": 17810 + }, + { + "grad_norm": 0.12149041891098022, + "learning_rate": 3.216119489523889e-06, + "loss": 0.0047, + "step": 17820 + }, + { + "grad_norm": 0.12130892276763916, + "learning_rate": 3.1870116119775917e-06, + "loss": 0.0036, + "step": 17830 + }, + { + "grad_norm": 0.08184216916561127, + "learning_rate": 3.158031719353999e-06, + "loss": 0.0047, + "step": 17840 + }, + { + "grad_norm": 0.12849783897399902, + "learning_rate": 3.1291798908830273e-06, + "loss": 0.0037, + "step": 17850 + }, + { + "grad_norm": 0.0973939523100853, + "learning_rate": 3.1004562054444853e-06, + "loss": 0.004, + "step": 17860 + }, + { + "grad_norm": 0.10297653824090958, + "learning_rate": 3.071860741567806e-06, + "loss": 0.0056, + "step": 17870 + }, + { + "grad_norm": 0.1703691929578781, + "learning_rate": 3.04339357743193e-06, + "loss": 0.0036, + "step": 17880 + }, + { + "grad_norm": 0.09887855499982834, + "learning_rate": 3.0150547908649628e-06, + "loss": 0.0048, + "step": 17890 + }, + { + "grad_norm": 0.17256423830986023, + "learning_rate": 2.9868444593440957e-06, + "loss": 0.0044, + "step": 17900 + }, + { + "grad_norm": 0.14721833169460297, + "learning_rate": 2.9587626599952846e-06, + "loss": 0.0052, + "step": 17910 + }, + { + "grad_norm": 0.10446737706661224, + "learning_rate": 2.930809469593082e-06, + "loss": 0.0058, + "step": 17920 + }, + { + "grad_norm": 0.1434236317873001, + "learning_rate": 2.9029849645604733e-06, + "loss": 0.0038, + "step": 17930 + }, + { + "grad_norm": 0.1383984535932541, + "learning_rate": 2.8752892209685632e-06, + "loss": 0.0054, + "step": 17940 + }, + { + "grad_norm": 0.10035323351621628, + "learning_rate": 2.847722314536483e-06, + "loss": 0.0033, + "step": 17950 + }, + { + "grad_norm": 0.11949007958173752, + "learning_rate": 2.820284320631078e-06, + "loss": 0.0063, + "step": 17960 + }, + { + "grad_norm": 0.1266811639070511, + "learning_rate": 2.792975314266788e-06, + "loss": 0.0041, + "step": 17970 + }, + { + "grad_norm": 0.12894940376281738, + "learning_rate": 2.7657953701054007e-06, + "loss": 0.0037, + "step": 17980 + }, + { + "grad_norm": 0.08531206846237183, + "learning_rate": 2.7387445624558306e-06, + "loss": 0.0056, + "step": 17990 + }, + { + "grad_norm": 0.09096170216798782, + "learning_rate": 2.7118229652739747e-06, + "loss": 0.0036, + "step": 18000 + }, + { + "grad_norm": 0.13481144607067108, + "learning_rate": 2.6850306521624236e-06, + "loss": 0.0047, + "step": 18010 + }, + { + "grad_norm": 0.0847296267747879, + "learning_rate": 2.6583676963703507e-06, + "loss": 0.0034, + "step": 18020 + }, + { + "grad_norm": 0.07553955912590027, + "learning_rate": 2.631834170793268e-06, + "loss": 0.0039, + "step": 18030 + }, + { + "grad_norm": 0.1235010102391243, + "learning_rate": 2.6054301479728036e-06, + "loss": 0.0031, + "step": 18040 + }, + { + "grad_norm": 0.07297364622354507, + "learning_rate": 2.579155700096575e-06, + "loss": 0.0046, + "step": 18050 + }, + { + "grad_norm": 0.10736757516860962, + "learning_rate": 2.5530108989978873e-06, + "loss": 0.006, + "step": 18060 + }, + { + "grad_norm": 0.11740373820066452, + "learning_rate": 2.5269958161556416e-06, + "loss": 0.0064, + "step": 18070 + }, + { + "grad_norm": 0.0979749783873558, + "learning_rate": 2.5011105226940888e-06, + "loss": 0.0034, + "step": 18080 + }, + { + "grad_norm": 0.13769516348838806, + "learning_rate": 2.4753550893826248e-06, + "loss": 0.0048, + "step": 18090 + }, + { + "grad_norm": 0.08744233101606369, + "learning_rate": 2.4497295866356296e-06, + "loss": 0.004, + "step": 18100 + }, + { + "grad_norm": 0.09951108694076538, + "learning_rate": 2.424234084512228e-06, + "loss": 0.0036, + "step": 18110 + }, + { + "grad_norm": 0.1635640263557434, + "learning_rate": 2.3988686527161687e-06, + "loss": 0.0074, + "step": 18120 + }, + { + "grad_norm": 0.14396807551383972, + "learning_rate": 2.373633360595573e-06, + "loss": 0.005, + "step": 18130 + }, + { + "grad_norm": 0.14187557995319366, + "learning_rate": 2.3485282771427585e-06, + "loss": 0.0028, + "step": 18140 + }, + { + "grad_norm": 0.08442353457212448, + "learning_rate": 2.3235534709940665e-06, + "loss": 0.0031, + "step": 18150 + }, + { + "grad_norm": 0.1230187937617302, + "learning_rate": 2.2987090104296617e-06, + "loss": 0.004, + "step": 18160 + }, + { + "grad_norm": 0.10204652696847916, + "learning_rate": 2.273994963373355e-06, + "loss": 0.0038, + "step": 18170 + }, + { + "grad_norm": 0.07479498535394669, + "learning_rate": 2.249411397392409e-06, + "loss": 0.0045, + "step": 18180 + }, + { + "grad_norm": 0.10246019065380096, + "learning_rate": 2.2249583796973506e-06, + "loss": 0.0048, + "step": 18190 + }, + { + "grad_norm": 0.16304340958595276, + "learning_rate": 2.200635977141796e-06, + "loss": 0.0047, + "step": 18200 + }, + { + "grad_norm": 0.1268751174211502, + "learning_rate": 2.17644425622226e-06, + "loss": 0.0042, + "step": 18210 + }, + { + "grad_norm": 0.07579465210437775, + "learning_rate": 2.152383283077991e-06, + "loss": 0.0042, + "step": 18220 + }, + { + "grad_norm": 0.11345085501670837, + "learning_rate": 2.128453123490781e-06, + "loss": 0.0049, + "step": 18230 + }, + { + "grad_norm": 0.1875070482492447, + "learning_rate": 2.1046538428847462e-06, + "loss": 0.0053, + "step": 18240 + }, + { + "grad_norm": 0.09485717862844467, + "learning_rate": 2.0809855063262273e-06, + "loss": 0.0042, + "step": 18250 + }, + { + "grad_norm": 0.07932358235120773, + "learning_rate": 2.057448178523558e-06, + "loss": 0.0043, + "step": 18260 + }, + { + "grad_norm": 0.11418958753347397, + "learning_rate": 2.034041923826885e-06, + "loss": 0.0051, + "step": 18270 + }, + { + "grad_norm": 0.08102735131978989, + "learning_rate": 2.0107668062280204e-06, + "loss": 0.0048, + "step": 18280 + }, + { + "grad_norm": 0.08639351278543472, + "learning_rate": 1.9876228893602357e-06, + "loss": 0.0037, + "step": 18290 + }, + { + "grad_norm": 0.09763824194669724, + "learning_rate": 1.9646102364981266e-06, + "loss": 0.004, + "step": 18300 + }, + { + "grad_norm": 0.08055881410837173, + "learning_rate": 1.9417289105574053e-06, + "loss": 0.0072, + "step": 18310 + }, + { + "grad_norm": 0.09894809126853943, + "learning_rate": 1.9189789740947427e-06, + "loss": 0.0036, + "step": 18320 + }, + { + "grad_norm": 0.10058688372373581, + "learning_rate": 1.896360489307597e-06, + "loss": 0.0063, + "step": 18330 + }, + { + "grad_norm": 0.1028638482093811, + "learning_rate": 1.8738735180340362e-06, + "loss": 0.004, + "step": 18340 + }, + { + "grad_norm": 0.0817807987332344, + "learning_rate": 1.8515181217525824e-06, + "loss": 0.0038, + "step": 18350 + }, + { + "grad_norm": 0.07613148540258408, + "learning_rate": 1.8292943615820457e-06, + "loss": 0.0036, + "step": 18360 + }, + { + "grad_norm": 0.05594291910529137, + "learning_rate": 1.8072022982813296e-06, + "loss": 0.0034, + "step": 18370 + }, + { + "grad_norm": 0.10103534907102585, + "learning_rate": 1.7852419922492925e-06, + "loss": 0.005, + "step": 18380 + }, + { + "grad_norm": 0.06014038249850273, + "learning_rate": 1.763413503524569e-06, + "loss": 0.0031, + "step": 18390 + }, + { + "grad_norm": 0.11110295355319977, + "learning_rate": 1.7417168917854165e-06, + "loss": 0.0032, + "step": 18400 + }, + { + "grad_norm": 0.08939255774021149, + "learning_rate": 1.720152216349552e-06, + "loss": 0.0034, + "step": 18410 + }, + { + "grad_norm": 0.08942482620477676, + "learning_rate": 1.6987195361739595e-06, + "loss": 0.0036, + "step": 18420 + }, + { + "grad_norm": 0.1863851249217987, + "learning_rate": 1.6774189098547832e-06, + "loss": 0.0051, + "step": 18430 + }, + { + "grad_norm": 0.12386361509561539, + "learning_rate": 1.6562503956271069e-06, + "loss": 0.0047, + "step": 18440 + }, + { + "grad_norm": 0.12683850526809692, + "learning_rate": 1.6352140513648417e-06, + "loss": 0.0043, + "step": 18450 + }, + { + "grad_norm": 0.09401161968708038, + "learning_rate": 1.6143099345805712e-06, + "loss": 0.0029, + "step": 18460 + }, + { + "grad_norm": 0.08010539412498474, + "learning_rate": 1.5935381024253293e-06, + "loss": 0.0035, + "step": 18470 + }, + { + "grad_norm": 0.05355304852128029, + "learning_rate": 1.572898611688517e-06, + "loss": 0.0034, + "step": 18480 + }, + { + "grad_norm": 0.12306501716375351, + "learning_rate": 1.5523915187977133e-06, + "loss": 0.0056, + "step": 18490 + }, + { + "grad_norm": 0.10398924350738525, + "learning_rate": 1.532016879818532e-06, + "loss": 0.0039, + "step": 18500 + }, + { + "grad_norm": 0.07574418187141418, + "learning_rate": 1.51177475045447e-06, + "loss": 0.0037, + "step": 18510 + }, + { + "grad_norm": 0.0989949107170105, + "learning_rate": 1.4916651860467035e-06, + "loss": 0.0038, + "step": 18520 + }, + { + "grad_norm": 0.06689750403165817, + "learning_rate": 1.471688241574043e-06, + "loss": 0.0031, + "step": 18530 + }, + { + "grad_norm": 0.09178074449300766, + "learning_rate": 1.451843971652672e-06, + "loss": 0.0041, + "step": 18540 + }, + { + "grad_norm": 0.12429051101207733, + "learning_rate": 1.432132430536076e-06, + "loss": 0.0042, + "step": 18550 + }, + { + "grad_norm": 0.08976944535970688, + "learning_rate": 1.412553672114869e-06, + "loss": 0.0035, + "step": 18560 + }, + { + "grad_norm": 0.07835699617862701, + "learning_rate": 1.3931077499166056e-06, + "loss": 0.0036, + "step": 18570 + }, + { + "grad_norm": 0.07831160724163055, + "learning_rate": 1.3737947171057085e-06, + "loss": 0.0035, + "step": 18580 + }, + { + "grad_norm": 0.08674316853284836, + "learning_rate": 1.3546146264832582e-06, + "loss": 0.0038, + "step": 18590 + }, + { + "grad_norm": 0.06477545201778412, + "learning_rate": 1.3355675304869086e-06, + "loss": 0.0037, + "step": 18600 + }, + { + "grad_norm": 0.11489281058311462, + "learning_rate": 1.3166534811906827e-06, + "loss": 0.005, + "step": 18610 + }, + { + "grad_norm": 0.1530413031578064, + "learning_rate": 1.2978725303048666e-06, + "loss": 0.0049, + "step": 18620 + }, + { + "grad_norm": 0.12522374093532562, + "learning_rate": 1.2792247291758762e-06, + "loss": 0.0045, + "step": 18630 + }, + { + "grad_norm": 0.06969575583934784, + "learning_rate": 1.2607101287860635e-06, + "loss": 0.0036, + "step": 18640 + }, + { + "grad_norm": 0.08598778396844864, + "learning_rate": 1.2423287797536654e-06, + "loss": 0.0033, + "step": 18650 + }, + { + "grad_norm": 0.06391648203134537, + "learning_rate": 1.2240807323325776e-06, + "loss": 0.0037, + "step": 18660 + }, + { + "grad_norm": 0.10801202058792114, + "learning_rate": 1.205966036412254e-06, + "loss": 0.0033, + "step": 18670 + }, + { + "grad_norm": 0.11867705732584, + "learning_rate": 1.1879847415175949e-06, + "loss": 0.0051, + "step": 18680 + }, + { + "grad_norm": 0.11671072244644165, + "learning_rate": 1.1701368968087712e-06, + "loss": 0.0046, + "step": 18690 + }, + { + "grad_norm": 0.05049500986933708, + "learning_rate": 1.1524225510811116e-06, + "loss": 0.0036, + "step": 18700 + }, + { + "grad_norm": 0.07130085676908493, + "learning_rate": 1.1348417527649535e-06, + "loss": 0.0034, + "step": 18710 + }, + { + "grad_norm": 0.054281286895275116, + "learning_rate": 1.1173945499255268e-06, + "loss": 0.0031, + "step": 18720 + }, + { + "grad_norm": 0.07449881732463837, + "learning_rate": 1.1000809902628307e-06, + "loss": 0.0052, + "step": 18730 + }, + { + "grad_norm": 0.09705372154712677, + "learning_rate": 1.082901121111468e-06, + "loss": 0.0034, + "step": 18740 + }, + { + "grad_norm": 0.10779692232608795, + "learning_rate": 1.0658549894405456e-06, + "loss": 0.0032, + "step": 18750 + }, + { + "grad_norm": 0.050938431173563004, + "learning_rate": 1.0489426418535342e-06, + "loss": 0.0026, + "step": 18760 + }, + { + "grad_norm": 0.07344609498977661, + "learning_rate": 1.0321641245881474e-06, + "loss": 0.003, + "step": 18770 + }, + { + "grad_norm": 0.0742596834897995, + "learning_rate": 1.015519483516214e-06, + "loss": 0.0037, + "step": 18780 + }, + { + "grad_norm": 0.06837273389101028, + "learning_rate": 9.990087641435443e-07, + "loss": 0.003, + "step": 18790 + }, + { + "grad_norm": 0.08643893897533417, + "learning_rate": 9.826320116098132e-07, + "loss": 0.0044, + "step": 18800 + }, + { + "grad_norm": 0.04812360182404518, + "learning_rate": 9.663892706884447e-07, + "loss": 0.0031, + "step": 18810 + }, + { + "grad_norm": 0.08176058530807495, + "learning_rate": 9.502805857864616e-07, + "loss": 0.0061, + "step": 18820 + }, + { + "grad_norm": 0.08198662847280502, + "learning_rate": 9.34306000944396e-07, + "loss": 0.0039, + "step": 18830 + }, + { + "grad_norm": 0.07465307414531708, + "learning_rate": 9.184655598361624e-07, + "loss": 0.0034, + "step": 18840 + }, + { + "grad_norm": 0.07182762771844864, + "learning_rate": 9.027593057689076e-07, + "loss": 0.0038, + "step": 18850 + }, + { + "grad_norm": 0.0734962522983551, + "learning_rate": 8.871872816829441e-07, + "loss": 0.0031, + "step": 18860 + }, + { + "grad_norm": 0.06601841002702713, + "learning_rate": 8.717495301515777e-07, + "loss": 0.0028, + "step": 18870 + }, + { + "grad_norm": 0.10713382810354233, + "learning_rate": 8.564460933810415e-07, + "loss": 0.0044, + "step": 18880 + }, + { + "grad_norm": 0.09342646598815918, + "learning_rate": 8.412770132103453e-07, + "loss": 0.006, + "step": 18890 + }, + { + "grad_norm": 0.0634833425283432, + "learning_rate": 8.262423311111711e-07, + "loss": 0.0034, + "step": 18900 + }, + { + "grad_norm": 0.07692574709653854, + "learning_rate": 8.113420881877665e-07, + "loss": 0.0037, + "step": 18910 + }, + { + "grad_norm": 0.07447660714387894, + "learning_rate": 7.965763251768288e-07, + "loss": 0.004, + "step": 18920 + }, + { + "grad_norm": 0.1114739403128624, + "learning_rate": 7.819450824473995e-07, + "loss": 0.0046, + "step": 18930 + }, + { + "grad_norm": 0.07135684788227081, + "learning_rate": 7.674484000007198e-07, + "loss": 0.004, + "step": 18940 + }, + { + "grad_norm": 0.08017963916063309, + "learning_rate": 7.530863174701752e-07, + "loss": 0.0042, + "step": 18950 + }, + { + "grad_norm": 0.08928892016410828, + "learning_rate": 7.38858874121151e-07, + "loss": 0.0052, + "step": 18960 + }, + { + "grad_norm": 0.1515638679265976, + "learning_rate": 7.247661088509328e-07, + "loss": 0.0039, + "step": 18970 + }, + { + "grad_norm": 0.09839137643575668, + "learning_rate": 7.108080601886002e-07, + "loss": 0.0042, + "step": 18980 + }, + { + "grad_norm": 0.060578133910894394, + "learning_rate": 6.969847662949336e-07, + "loss": 0.0043, + "step": 18990 + }, + { + "grad_norm": 0.07045385986566544, + "learning_rate": 6.832962649622798e-07, + "loss": 0.0043, + "step": 19000 + }, + { + "grad_norm": 0.11689253896474838, + "learning_rate": 6.697425936144863e-07, + "loss": 0.0033, + "step": 19010 + }, + { + "grad_norm": 0.051963455975055695, + "learning_rate": 6.563237893067731e-07, + "loss": 0.0035, + "step": 19020 + }, + { + "grad_norm": 0.08776947855949402, + "learning_rate": 6.430398887256328e-07, + "loss": 0.0041, + "step": 19030 + }, + { + "grad_norm": 0.053857170045375824, + "learning_rate": 6.298909281887478e-07, + "loss": 0.0029, + "step": 19040 + }, + { + "grad_norm": 0.1286555677652359, + "learning_rate": 6.168769436448673e-07, + "loss": 0.0033, + "step": 19050 + }, + { + "grad_norm": 0.08176587522029877, + "learning_rate": 6.03997970673742e-07, + "loss": 0.003, + "step": 19060 + }, + { + "grad_norm": 0.09932407736778259, + "learning_rate": 5.912540444859782e-07, + "loss": 0.0044, + "step": 19070 + }, + { + "grad_norm": 0.0911400094628334, + "learning_rate": 5.786451999229837e-07, + "loss": 0.007, + "step": 19080 + }, + { + "grad_norm": 0.0852760598063469, + "learning_rate": 5.661714714568722e-07, + "loss": 0.004, + "step": 19090 + }, + { + "grad_norm": 0.08842559158802032, + "learning_rate": 5.538328931903259e-07, + "loss": 0.0036, + "step": 19100 + }, + { + "grad_norm": 0.07818997651338577, + "learning_rate": 5.416294988565551e-07, + "loss": 0.0033, + "step": 19110 + }, + { + "grad_norm": 0.059740930795669556, + "learning_rate": 5.29561321819172e-07, + "loss": 0.004, + "step": 19120 + }, + { + "grad_norm": 0.07137617468833923, + "learning_rate": 5.176283950721061e-07, + "loss": 0.0034, + "step": 19130 + }, + { + "grad_norm": 0.08400587737560272, + "learning_rate": 5.058307512395332e-07, + "loss": 0.0037, + "step": 19140 + }, + { + "grad_norm": 0.19625791907310486, + "learning_rate": 4.941684225757526e-07, + "loss": 0.0035, + "step": 19150 + }, + { + "grad_norm": 0.03737051039934158, + "learning_rate": 4.826414409651314e-07, + "loss": 0.0031, + "step": 19160 + }, + { + "grad_norm": 0.1145489290356636, + "learning_rate": 4.712498379219943e-07, + "loss": 0.0033, + "step": 19170 + }, + { + "grad_norm": 0.10778439044952393, + "learning_rate": 4.599936445905506e-07, + "loss": 0.0047, + "step": 19180 + }, + { + "grad_norm": 0.056398406624794006, + "learning_rate": 4.4887289174480594e-07, + "loss": 0.0035, + "step": 19190 + }, + { + "grad_norm": 0.05212453380227089, + "learning_rate": 4.378876097884621e-07, + "loss": 0.0036, + "step": 19200 + }, + { + "grad_norm": 0.0517825148999691, + "learning_rate": 4.2703782875487264e-07, + "loss": 0.0027, + "step": 19210 + }, + { + "grad_norm": 0.04213884100317955, + "learning_rate": 4.163235783069208e-07, + "loss": 0.0025, + "step": 19220 + }, + { + "grad_norm": 0.0969507098197937, + "learning_rate": 4.057448877369585e-07, + "loss": 0.0041, + "step": 19230 + }, + { + "grad_norm": 0.05608559772372246, + "learning_rate": 3.9530178596672295e-07, + "loss": 0.0037, + "step": 19240 + }, + { + "grad_norm": 0.056393902748823166, + "learning_rate": 3.849943015472479e-07, + "loss": 0.0047, + "step": 19250 + }, + { + "grad_norm": 0.10194101184606552, + "learning_rate": 3.748224626588137e-07, + "loss": 0.0031, + "step": 19260 + }, + { + "grad_norm": 0.0955602377653122, + "learning_rate": 3.647862971108307e-07, + "loss": 0.0035, + "step": 19270 + }, + { + "grad_norm": 0.06831927597522736, + "learning_rate": 3.5488583234179473e-07, + "loss": 0.0046, + "step": 19280 + }, + { + "grad_norm": 0.044883936643600464, + "learning_rate": 3.4512109541920413e-07, + "loss": 0.0041, + "step": 19290 + }, + { + "grad_norm": 0.046192340552806854, + "learning_rate": 3.354921130394706e-07, + "loss": 0.0026, + "step": 19300 + }, + { + "grad_norm": 0.04406873881816864, + "learning_rate": 3.259989115278639e-07, + "loss": 0.0029, + "step": 19310 + }, + { + "grad_norm": 0.07073112577199936, + "learning_rate": 3.1664151683843403e-07, + "loss": 0.0042, + "step": 19320 + }, + { + "grad_norm": 0.05474907159805298, + "learning_rate": 3.074199545539447e-07, + "loss": 0.0052, + "step": 19330 + }, + { + "grad_norm": 0.047355033457279205, + "learning_rate": 2.983342498857955e-07, + "loss": 0.0043, + "step": 19340 + }, + { + "grad_norm": 0.10388905555009842, + "learning_rate": 2.893844276739499e-07, + "loss": 0.003, + "step": 19350 + }, + { + "grad_norm": 0.10018634796142578, + "learning_rate": 2.8057051238688514e-07, + "loss": 0.0038, + "step": 19360 + }, + { + "grad_norm": 0.03913424164056778, + "learning_rate": 2.71892528121509e-07, + "loss": 0.0032, + "step": 19370 + }, + { + "grad_norm": 0.07924854755401611, + "learning_rate": 2.633504986030988e-07, + "loss": 0.0043, + "step": 19380 + }, + { + "grad_norm": 0.09704583883285522, + "learning_rate": 2.549444471852347e-07, + "loss": 0.0055, + "step": 19390 + }, + { + "grad_norm": 0.05486253276467323, + "learning_rate": 2.4667439684974423e-07, + "loss": 0.0038, + "step": 19400 + }, + { + "grad_norm": 0.12703447043895721, + "learning_rate": 2.3854037020662467e-07, + "loss": 0.0056, + "step": 19410 + }, + { + "grad_norm": 0.06610122323036194, + "learning_rate": 2.3054238949399288e-07, + "loss": 0.0027, + "step": 19420 + }, + { + "grad_norm": 0.08677727729082108, + "learning_rate": 2.2268047657802993e-07, + "loss": 0.0037, + "step": 19430 + }, + { + "grad_norm": 0.06518864631652832, + "learning_rate": 2.149546529529034e-07, + "loss": 0.0039, + "step": 19440 + }, + { + "grad_norm": 0.11082996428012848, + "learning_rate": 2.0736493974071736e-07, + "loss": 0.004, + "step": 19450 + }, + { + "grad_norm": 0.036419276148080826, + "learning_rate": 1.9991135769145686e-07, + "loss": 0.003, + "step": 19460 + }, + { + "grad_norm": 0.0653953105211258, + "learning_rate": 1.9259392718293245e-07, + "loss": 0.0042, + "step": 19470 + }, + { + "grad_norm": 0.09075991064310074, + "learning_rate": 1.8541266822072467e-07, + "loss": 0.0056, + "step": 19480 + }, + { + "grad_norm": 0.06219499185681343, + "learning_rate": 1.7836760043811184e-07, + "loss": 0.0033, + "step": 19490 + }, + { + "grad_norm": 0.11633436381816864, + "learning_rate": 1.7145874309604792e-07, + "loss": 0.0031, + "step": 19500 + }, + { + "grad_norm": 0.11897324025630951, + "learning_rate": 1.6468611508308474e-07, + "loss": 0.0054, + "step": 19510 + }, + { + "grad_norm": 0.055197566747665405, + "learning_rate": 1.5804973491532204e-07, + "loss": 0.0032, + "step": 19520 + }, + { + "grad_norm": 0.07045387476682663, + "learning_rate": 1.5154962073637424e-07, + "loss": 0.0043, + "step": 19530 + }, + { + "grad_norm": 0.06476310640573502, + "learning_rate": 1.4518579031730372e-07, + "loss": 0.0034, + "step": 19540 + }, + { + "grad_norm": 0.10052386671304703, + "learning_rate": 1.389582610565876e-07, + "loss": 0.0041, + "step": 19550 + }, + { + "grad_norm": 0.12584728002548218, + "learning_rate": 1.3286704998003995e-07, + "loss": 0.0034, + "step": 19560 + }, + { + "grad_norm": 0.04220246151089668, + "learning_rate": 1.2691217374080632e-07, + "loss": 0.005, + "step": 19570 + }, + { + "grad_norm": 0.0714455097913742, + "learning_rate": 1.2109364861929705e-07, + "loss": 0.0024, + "step": 19580 + }, + { + "grad_norm": 0.07059883326292038, + "learning_rate": 1.1541149052312628e-07, + "loss": 0.0045, + "step": 19590 + }, + { + "grad_norm": 0.05011598765850067, + "learning_rate": 1.0986571498710074e-07, + "loss": 0.0046, + "step": 19600 + }, + { + "grad_norm": 0.03681618720293045, + "learning_rate": 1.0445633717316438e-07, + "loss": 0.003, + "step": 19610 + }, + { + "grad_norm": 0.13256072998046875, + "learning_rate": 9.918337187034277e-08, + "loss": 0.0038, + "step": 19620 + }, + { + "grad_norm": 0.08195763826370239, + "learning_rate": 9.404683349472643e-08, + "loss": 0.0029, + "step": 19630 + }, + { + "grad_norm": 0.06682359427213669, + "learning_rate": 8.904673608940983e-08, + "loss": 0.005, + "step": 19640 + }, + { + "grad_norm": 0.11099401861429214, + "learning_rate": 8.418309332447471e-08, + "loss": 0.0044, + "step": 19650 + }, + { + "grad_norm": 0.06852664053440094, + "learning_rate": 7.945591849692902e-08, + "loss": 0.0033, + "step": 19660 + }, + { + "grad_norm": 0.10303331166505814, + "learning_rate": 7.486522453069578e-08, + "loss": 0.0044, + "step": 19670 + }, + { + "grad_norm": 0.11328206211328506, + "learning_rate": 7.041102397655208e-08, + "loss": 0.0048, + "step": 19680 + }, + { + "grad_norm": 0.053033504635095596, + "learning_rate": 6.609332901210685e-08, + "loss": 0.0031, + "step": 19690 + }, + { + "grad_norm": 0.06228000670671463, + "learning_rate": 6.191215144178419e-08, + "loss": 0.0028, + "step": 19700 + }, + { + "grad_norm": 0.06171086058020592, + "learning_rate": 5.786750269675678e-08, + "loss": 0.0041, + "step": 19710 + }, + { + "grad_norm": 0.11801750957965851, + "learning_rate": 5.395939383494031e-08, + "loss": 0.0035, + "step": 19720 + }, + { + "grad_norm": 0.0650201365351677, + "learning_rate": 5.018783554095463e-08, + "loss": 0.0061, + "step": 19730 + }, + { + "grad_norm": 0.06774918735027313, + "learning_rate": 4.655283812610156e-08, + "loss": 0.0037, + "step": 19740 + }, + { + "grad_norm": 0.05833545699715614, + "learning_rate": 4.305441152831491e-08, + "loss": 0.0036, + "step": 19750 + }, + { + "grad_norm": 0.061846163123846054, + "learning_rate": 3.9692565312171584e-08, + "loss": 0.0026, + "step": 19760 + }, + { + "grad_norm": 0.11921104788780212, + "learning_rate": 3.6467308668824975e-08, + "loss": 0.0042, + "step": 19770 + }, + { + "grad_norm": 0.0699174776673317, + "learning_rate": 3.3378650416004964e-08, + "loss": 0.0039, + "step": 19780 + }, + { + "grad_norm": 0.05499502271413803, + "learning_rate": 3.042659899797906e-08, + "loss": 0.0035, + "step": 19790 + }, + { + "grad_norm": 0.050804853439331055, + "learning_rate": 2.76111624855524e-08, + "loss": 0.003, + "step": 19800 + }, + { + "grad_norm": 0.08485697954893112, + "learning_rate": 2.4932348576017784e-08, + "loss": 0.0034, + "step": 19810 + }, + { + "grad_norm": 0.11028487235307693, + "learning_rate": 2.239016459314458e-08, + "loss": 0.0039, + "step": 19820 + }, + { + "grad_norm": 0.11637432128190994, + "learning_rate": 1.9984617487173174e-08, + "loss": 0.004, + "step": 19830 + }, + { + "grad_norm": 0.07264947891235352, + "learning_rate": 1.7715713834776105e-08, + "loss": 0.0038, + "step": 19840 + }, + { + "grad_norm": 0.0735667273402214, + "learning_rate": 1.5583459839046964e-08, + "loss": 0.0037, + "step": 19850 + }, + { + "grad_norm": 0.09691016376018524, + "learning_rate": 1.3587861329489304e-08, + "loss": 0.0048, + "step": 19860 + }, + { + "grad_norm": 0.08705077320337296, + "learning_rate": 1.1728923761994415e-08, + "loss": 0.0046, + "step": 19870 + }, + { + "grad_norm": 0.06617732346057892, + "learning_rate": 1.0006652218819135e-08, + "loss": 0.0049, + "step": 19880 + }, + { + "grad_norm": 0.060623060911893845, + "learning_rate": 8.421051408596947e-09, + "loss": 0.0052, + "step": 19890 + }, + { + "grad_norm": 0.10678309947252274, + "learning_rate": 6.972125666299123e-09, + "loss": 0.0043, + "step": 19900 + }, + { + "grad_norm": 0.11491920053958893, + "learning_rate": 5.659878953229169e-09, + "loss": 0.0058, + "step": 19910 + }, + { + "grad_norm": 0.0382302887737751, + "learning_rate": 4.48431485701728e-09, + "loss": 0.0035, + "step": 19920 + }, + { + "grad_norm": 0.10758187621831894, + "learning_rate": 3.4454365916203322e-09, + "loss": 0.0036, + "step": 19930 + }, + { + "grad_norm": 0.11459647119045258, + "learning_rate": 2.5432469972830332e-09, + "loss": 0.0041, + "step": 19940 + }, + { + "grad_norm": 0.10030755400657654, + "learning_rate": 1.7777485405601203e-09, + "loss": 0.0031, + "step": 19950 + }, + { + "grad_norm": 0.0468946136534214, + "learning_rate": 1.1489433142941597e-09, + "loss": 0.004, + "step": 19960 + }, + { + "grad_norm": 0.12162579596042633, + "learning_rate": 6.568330376210963e-10, + "loss": 0.0043, + "step": 19970 + }, + { + "grad_norm": 0.051232777535915375, + "learning_rate": 3.0141905594249787e-10, + "loss": 0.0031, + "step": 19980 + }, + { + "grad_norm": 0.05374916270375252, + "learning_rate": 8.270234094776008e-11, + "loss": 0.0028, + "step": 19990 + }, + { + "grad_norm": 0.07233867049217224, + "learning_rate": 6.834906085551041e-13, + "loss": 0.0039, + "step": 20000 + } + ], + "logging_steps": 10, + "max_steps": 20000, + "num_input_tokens_seen": 0, + "num_train_epochs": 9223372036854775807, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 48, + "trial_name": null, + "trial_params": null +}